trade shock transmission: a study of selected african ... · jonathan e. ogbuabor phd department of...
TRANSCRIPT
1
Trade Shock Transmission A Study of Selected African Economies the BRIC
and the Rest of the Global Economy
Ekeocha Patterson C PhD
Central Bank of Nigeria Research Department
pekeocha04fulbrightmailorg pcekeochacbngovng
+2348035487968
amp
Jonathan E Ogbuabor PhD
Department of Economics University of Nigeria Nsukka Nigeria
jonathanogbuaborunnedung
+2348035077722 Abstract
This study investigated trade shock transmission between Africa the BRIC and the rest of the global
economy with a view to understanding the likely disposition of African economies towards trade shocks The
study extended the network approach of Diebold and Yilmaz (2009) by constructing generalized trade
linkage measures at various degrees of aggregation The results indicate that the trade linkage between
Africa and the rest of the global economy is quite substantial with the total trade linkage index having an
average value of 87 We find also that China USA UK Japan EU and Canada dominate Africarsquos trade
and therefore have the potential to spread trade shocks to it The results further indicate that apart from the
BRIC other regional trading blocs such as Asia the Americas and Europe play influential roles in Africarsquos
trade Overall the findings show that African economies are predominantly net receivers of trade shocks
originating from the aforementioned dominant sources The study therefore concludes that the patterns of
cross-country trade shock spillovers obtained in this study would likely influence Africarsquos move from
globalization to regional integration or multi-polarity in the years to come
Keywords Globalization Trade Shock Transmission Global VAR Africa the Global Economy
JEL Classification F42 F43 C32 N77 N70
Submitted to the 22nd
Annual Conference on Global Economic Analysis (GTAP) on the Theme
ldquoChallenges to Global Social and Economic Growthrdquo June 19-21 2019 Warsaw Poland
2
Introduction
Intra-Africalsquos trade performance (import and export) is very infinitesimal compared to the rest of
the world (African Trade Statistics Yearbook 2017) The same trend reflects the characteristics of
African countries intra and extra - African trade For instance Nigerialsquos Intra-Africa import trade is
far below its trade with the rest of the world from 2010 ndash 2016 While Nigerialsquos import from the
rest of the world averaged US$4264277 million or 948 per cent during the same period it was
US$229385 million or 52 per cent within Africa Similarly Nigerialsquos intra-African export trade
averaged US$1046565 million or 120 per cent of total export during 2010 to 2016 while the
export to the rest of the world averaged US$8124003 or 880 per cent of total export The low
intra-Africa trade may have necessitated the move for Africalsquos trade integration to fast-track its
development Little wonder Free Trade Agreements (FTAs) have become important elements in
economic integration amongst most countries of the world It provides avenues for nations to enter
into bilateral or regional trade agreements to eliminate tariffs and non-tariff barriers to trade among
member countries African is currently establishing the African Continental Free Trade Area
(AfCFTA) treaty to create a single continental market for goods and services in member nations of
the African Union (AU) with free movement of business persons and investments using a single
currency
The AfCFTA is gathering momentum because globalization remains a contentious subject due to
the mixed benefits and opportunities it presents to participating countries In particular it makes
countries vulnerable to external shocks Granted that shocks often occur and granted further that
African countries largely have significant trade linkage with the rest of the world it is important to
model the sources of foreign influence on domestic economies particularly the BRIC (Brazil
Russia India and China) economies whose trade linkage with Africa is enormous In fact not only
does Africa trade significantly with the BRIC countries the latter also represent a model of
economic development exemplified by strong economic growth and an enormous capacity to
compete in a globalized world Besides trade linkage has been adjudged an important feature of
global economic integration between countries However following literature there is no common
view on whether intense global trade linkages would lead to more or less business cycle
harmonization Again World Banklsquos Global Economic Prospects report posits that growth in
emerging market and developing economies is expected to remain flat in 2019 Growth in many
other economies is also anticipated to decelerate with risks growing that growth could be even
3
weaker than anticipated This phenomenon may invariably worsen the growth prospects of
developing economies in the event of trade shocks
Broadly this study therefore examines trade shock propagation between Africa the BRIC
economies and the rest of the global economy Specifically the study seeks to (i) measure the
degree of trade linkage between Africa the BRIC and the rest of the global economy (ii) determine
the BRIC countries and other countries in the rest of the world that are dominating Africalsquos trade
and therefore have the potential to spread trade shocks to Africa and thus minimize its growth (iii)
determine the African economies that are most susceptible to trade shocks originating from the
BRIC countries and the rest of the global economy (iv) determine other trade blocs outside the
BRIC that exert dominant influence on Africa and therefore have the prospects of spreading trade
shocks to it
To achieve the above objectives the paper employs the normalized generalized forecast error
variance decompositions (NGFEVDs) distilled from an underlying global vector autoregressive
model (Global VAR) to build measures of trade linkage and shocks transmission between Africa
the BRICS countries and the rest of the global economy over the period 1970Q1 ndash 2017Q4 The
underlying model is estimated using the total trade statistics of the countries The time series
properties of the data are examined using the Augmented Dickey-Fuller (ADF) and Phillips-Perron
unit root tests as well as the Johansen cointegration test in order to determine the form of the
model to be estimated If the variables are overwhelmingly stationary at levels then a basic VAR
would be appropriate However if the variables are nonstationary then a VAR in first differences
becomes tenable in the absence of equilibrium relationship otherwise a vector error correction
model (VECM) would be estimated A key innovation in this study would be the use of the
NGFEVDs to build various measures of trade linkages between Africa and the rest of the global
economy at different levels of aggregation The generalized nature of the FEVDs means that they
would be invariant to the reordering of the variables in the system The choice of the global VAR
framework is mainly due to its ability to explicitly model the source of foreign influences on
domestic economies in Africa and the contributions of these domestic economies to the rest of the
world in a transparent and coherent manner It allows for the impact of shocks to be consistently
quantified aggregated and assessed so that interactions between countriesregions can be analyzed
Simply put our proposed framework measures both the direction and the strength of linkages
4
among entities in the system while identifying systemically important or vulnerable entities within
the system Several studies in the literature have exploited this framework in the study of
macroeconomic linkages with great success Examples include Pesaran Schuermann and Weiner
(2004) Dees di Mauro Pesaran and Smith (2007) Dees Holly Pesaran and Smith (2007) and
more recently Greenwood-Nimmo Nguyen and Shin (2015) We are adopting same framework
using some selected African economies and since the data for the study are quarterly the measures
of trade linkages would be computed over a 24-period horizon so that the long-run results are
better captured
The rationale for assessing the impact of trade between Africa the BRIC and the rest of the world is
to bring to the fore the potential trade shock impact between Africa and other trade blocks with
which it has significant trade linkages even as it makes concerted efforts towards signing of the
African Continental Free Trade Area (AfCFTA) through a treaty by all the countries concerned
Nigeria which is Africalsquos largest economy by size of GDP is currently understudying the tenets of
AfCFTA treaty preparatory to signing it The AfCFTA is to create a single continental market for
goods and services in member nations of the African Union (AU) with free movement of business
persons and investments using a single currency The AfCFTA is intended to be the external cordon
protecting the African countrieslsquo domestic markets Thus while AU member states are signing the
treaty it is important to be aware of the potential trade shocks that may emanate from the
significant trade linkages between Africa and the rest of the world
The rest of the paper is organized as follows Section 2 undertakes a brief review of relevant
literature Section 3 presents the data and methodology detailing some preliminary data analysis
while Section 4 discusses the empirical results Section 5 concludes the study and provides some
policy implications that the patterns of cross-country trade shock spillovers obtained in this study
would likely influence Africalsquos move from globalization to regional integration or multi-polarity in
the years to come
2 Brief Review of Literature
The impact of trade shocks in large advanced countries on emerging market economies mdash dubbed
spillovers mdash is hotly debated in global and national policy circles Kose and Reizman (1999)
examined the role of external shocks in explaining macroeconomic fluctuations in African
5
countries They constructed a quantitative stochastic dynamic multi-sector equilibrium model of a
small open economy calibrated to represent a typical African economy In their framework external
shocks consist of trade shocks modeled as fluctuations in the prices of exported primary
commodities imported capital goods and intermediate inputs and a financial shock modeled as
fluctuations in the world real interest rate Results from their analysis indicated that while trade
shocks accounted for roughly 45 percent of economic fluctuations in aggregate output financial
shocks play only a minor role They also found that adverse trade shocks induce prolonged
recessions
Ccedilakir and Kabundi (2011) examined the trade linkages between South Africa and the BRIC (Brazil
Russia India and China) countries They applied the global vector autoregressive model (global
VAR) to investigate the degree of trade linkages and shock transmission between South Africa and
the BRIC countries over the period 1995Q1-2009Q4 The model contained 32 countries and had
two different estimations the first one consists of 24 countries and one region with the 8 countries
in the euro area treated as a single economy and the second estimation contains 20 countries and
two regions with the BRIC and the euro area countries respectively treated as a single economy
The results suggest that trade linkages exist between our focus economies however the magnitude
differs between countries Shocks from each BRIC country are shown to have considerable impact
on South African real imports and output
Greenwood-Nimmo et al (2015) developed a technique to evaluate macroeconomic connectedness
among entities in sophisticated multi-country and global macroeconomic models Their
methodology is highly adaptable and may be applied to any model with an approximate VAR
representation They applied their technique to a global vector autoregressive model containing 169
macroeconomic and financial variables for 25 countries They derived vivid representations of the
connectedness of the system and found that the US the Eurozone and the crude oil market exert a
dominant influence over conditions in the global macroeconomy and that China and Brazil are also
globally significant economies Recursive analysis over the period of the global financial crisis
shows that shocks to global equity markets are rapidly and forcefully transmitted to real trade flows
and real GDP
Pesaran et al (2004) posit that financial institutions are ultimately exposed to macroeconomic
fluctuations in the global economy and thus they built a compact global model capable of
6
generating forecasts for a core set of macroeconomic factors (or variables) across a number of
countries Their model explicitly allows for the interdependencies that exist between national and
international factors Individual region-specific vector error-correcting models were estimated in
which the domestic variables were related to corresponding foreign variables constructed
exclusively to match the international trade pattern of the country under consideration The
individual country models were then linked in a consistent and cohesive manner to generate
forecasts for all of the variables in the world economy simultaneously Their global model was
estimated for 25 countries grouped into 11 regions using quarterly data over 1979Q1ndash1999Q1 The
degree of regional interdependencies was investigated via generalized impulse responses where the
effects of shocks to a given variable in a given country on the rest of the world were provided The
model was then used to investigate the effects of various global risk scenarios on a banklsquos loan
portfolio
Dees et al (2007) presented a quarterly global model combining individual country vector error-
correcting models in which the domestic variables were related to the country-specific foreign
variables Their global VAR (GVAR) model was estimated for 26 countries the Euro area being
treated as a single economy over the period 1979ndash2003 They provided a theoretical framework
where the GVAR was derived as an approximation to a global unobserved common factor model
Using the average pair-wise cross-section error correlations their GVAR approach was shown to be
quite effective in dealing with the common factor interdependencies and international co-
movements of business cycles They developed a sieve bootstrap procedure for simulation of the
GVAR as a whole which is then used in testing the structural stability of the parameters and for
establishing bootstrap confidence bounds for the impulse responses Finally in addition to
generalized impulse responses their paper considered the use of the GVAR for structurallsquo impulse
response analysis with focus on external shocks for the Euro area economy particularly in response
to shocks to the US
In another study Dees et al (2007) focused on testing the long run macroeconomic relations for
interest rates equity prices and exchange rates suggested by arbitrage in financial and
goods markets They used the global vector autoregressive (GVAR) model to test
for long run restrictions in each countryregion conditioning on the rest of the
world They employed Bootstrapping to compute both the empirical distribution of
7
the impulse responses and the log-likelihood ratio statistic for over-identifying
restrictions Their paper also examined the speed with which adjustments to the
long run relations would take place via the persistence profiles They found strong evidence
in favour of the uncovered interest parity (UIP) and to a lesser extent the Fisher equation across a
number of countries but their results for the purchasing power parity (PPP) were much weaker
Also the transmission of shocks and subsequent adjustments in financial markets were much faster
than those in goods markets
Ogbuabor et al (2016) examined the real and financial connectedness of selected African
economies with the global economy using a network approach They found that the connectedness
of African economies with the global economy is quite sizable with the global financial crisis
increasing the connectedness measures above their pre-crisis levels Their results show that US
EU and Canada dominate Africalsquos equity markets while China India and Japan dominate Africalsquos
real activities Their results suggest that African economies are predominantly small open
economies deeply interconnected but systemically unimportant and vulnerable to headwinds
emanating from the dominant economies in the overall global economy
Lubik and Teo (2003) found that world interest rate shocks are the main driving forces of business
cycles in small open economies while terms of trade shocks are not Thus they challenged the
existing results on the contribution of terms of trade and world interest rate shocks to output
fluctuations in small open economies which had been found to range from less than 10 per cent to
almost 90 per cent They argue that an identification problems lies at the heart of existing vastly
different results They overcame this by estimating a DSGE model using a structural Bayesian
estimation approach applying their methodology to five developed and developing economies The
approach enabled them to efficiently exploit cross-equation restrictions implied by the structural
model
Huidrom et al (2017) provide empirical estimates of the cross border spillovers from Seven largest
emerging market economies (China India Brazil Russia Mexico Indonesia and Turkey) using a
Bayesian vector autoregression model Their results indicate that first spillovers from EM7 are
sizeable a 1 percentage point increase in EM7 growth is associated with a 09 percentage point
increase in growth in other emerging and frontier markets and a 06 percentage point increase in
world growth at the end of three years Second sizeable as they are spillovers from EM7 are still
8
smaller than those from G7 countries (Group of Seven of advanced economies) Specifically
growth in other emerging and frontier markets and the global economy would increase by one-half
to three times more due to a similarly sized increase in G7 growth Third among the EM7
spillovers from China are the largest and permeate globally Their study was based on the
phenomenon that their growth could have significant cross-border spillovers based on their size and
integration The seven largest emerging market economies - China India Brazil Russia Mexico
Indonesia and Turkey - was adjudged to constitute more than one-quarter of global output and
more than half of global output growth during 2010ndash15
This study contributes to this emerging literature in several ways First it quantifies the degree of
trade linkage between Africa the BRIC and the rest of the global economy Two apart from
exposing the patterns of trade shock spillovers between Africa and the rest of the world using total
trade statistics this study also disaggregates the trade shock spillovers using exports and imports
statistics This provides deeper insights into the trade linkages between Africa and the rest of the
world Besides the study extended the empirical method by constructing trade linkage measures in
a coherent and transparent manner for ease of replication
3 Data and Methodology
The data for this study consists of the log of exports imports and total trade for the period 1990Q1-
2016Q4 The choice of this period is based on data availability for some of the African economies
and for Russia whose trade data in the World Development Indicators started in 1990 The BRICS
economies in this study are Brazil Russia India China and South Africa The selected African
economies included in this study are DR Congo Egypt Ghana Morocco Nigeria Tanzania and
Tunisia while the rest of the global economy is accounted for by Australia Canada the Eurozone
(EU) Indonesia Japan Malaysia United States of America (USA) and United Kingdom (UK)
These economies account for the larger chunk of Africalsquos trade while the selected African
economies together with South Africa account substantially for Africalsquos GDP The entire data were
taken from the World Development Indicators (WDI) based on the following indicator names
exports of goods and services (constant 2010 US$) as measure for exports imports of goods and
services (constant 2010 US$) as measure for imports and exports plus imports as measure for total
trade (or simply trade)
9
To ensure that there are enough observations for the analysis the data were converted from annual
to quarterly using Eviewlsquos quadratic match average option This is in line with the methodology
used in compiling the Global VAR database used by Greenwood-Nimmo Nguyen and Shin (2015)
the connectedness of the global economy Other studies that have successfully used the same
methodology are Ogbuabor Eigbiremolen Aneke and Manasseh (2018) and Ogbuabor Orji
Aneke and Erdene-Urnukh (2016) Also to reduce noise and ensure uniform scaling the entire data
were converted into indices (2010Y = 100) and logged prior to estimation Appendix 1 summarizes
the descriptive statistics of the data based on the log transformation of the data Focusing on Panel 1
which reports for total trade we find that among the African economies Ghana DR Congo and
Tanzania are the most volatile as seen from their standard deviations while China India and Russia
are the most volatile among the BRICS economies Among all the countries in the sample Japan
Canada and UK are the least volatile and also recorded the highest mean values The maximum and
minimum values do not suggest the presence of outliers in the data The time series plots of the data
for all the economies in the sample are presented in Appendix 2 based on the log transformation of
the data A close examination of the plots show close comovement among all the countries
indicating that the series track themselves closely The implication of this is that the data may be
cointegrated which will in turn affect the form of the underlying model to be estimated in this
study Thus we shall subject the series to cointegration test as part of the empirical procedures in
this study
31 The Choice of Diebold and Yilmaz (2009) Network Framework
A number of methodologies have been employed in the study of macroeconomic linkages among
entities in the global economy particularly in the study of shock propagation among countries For
instance cross-country correlations-based measures have been used to characterize macroeconomic
linkages among countries (Kehoe et al 1995 Kose et al 2003 Bollerslev 1990 Engle et al
1990 Mantegna 1999 Tumminello et al 2005 Taylor 2007 Gray amp Malone 2008 Engle 2009
Engle amp Kelly 2012) The pitfalls of this approach are twofold namely correlation is simply a
pairwise measure of association and it is non-directional This means that correlation-based
approach cannot handle such questions as ―what is the degree of trade linkage between Africa the
BRICS and the rest of the global economy Unlike the correlation-based measures of
macroeconomic linkage the network approach of Diebold and Yilmaz (2009) is non-pairwise yet
10
directional Granger Causality measures have also been used characterize networks so that the
macroeconomic linkages among entities in the global economy can be described and understood
(Caraiani 2013 Hiemstra amp Jones 1994 Dahlhans amp Eichler 2003 Shojaie amp Michailidis 2010
Billio et al 2012) The main weakness of the Granger Causality approach is that it captures only
pairwise relations and may not be useful in answering important questions like ―what is the degree
of trade linkage between Africa the BRICS and the rest of the global economy1
Furthermore as noted by Diebold and Yilmaz (2016) these alternative methodologies generally
dwell exclusively on testing rather than measurement and estimation of macroeconomic linkages
which are the key issues in this paper This study therefore follows the network approach of
Diebold and Yilmaz (2009) based on its ability to transparently use the size and direction of shocks
to build both directional and non-directional trade linkage measures over a given forecast horizon
According to Ogbuabor et al (2016) studies using this approach have four common features
namely (i) they are generally based on connectedness measures distilled from forecast error
variance decompositions (FEVDs) of an approximating vector autoregressive (VAR) model (ii)
they measure the direction and strength of linkages among entities in the system (iii) they can
identify systemically important entities in the system and (iv) they can study the dynamic nature of
shock propagation among entities in the system In what follows the underlying VAR model for
this study and the construction of the generalized trade linkages measures (GTLMs) are presented
to guide the ensuing analysis
32 Model Specification The Vector Autoregression (VAR) Model
The broad objective of this study is to examine the propagation of trade shock between Africa the
BRICS economies and the rest of the global economy Let 119937120061 be the log of total trade for all the
countries selected for this study so that 119937119947120061 stands for the logged total trade of the j-th country in the
system with 119947 = 120783 120784 hellip 119925 and 119925 is the number of countries selected for the study (which is 20)
Following Diebold and Yilmaz (2009) the trade linkage measures for this study are based on the
normalized generalized forecast error variance decompositions (NGFEVDs) of an underlying 119953-th
1 Other techniques have also been used in the literature for the study of macroeconomic linkages Ogbuabor et al
(2018) provides an overview of such alternative methodologies such as the dynamic latent factor models of Kose et al (2008) and Canova et al (2007) the CoVaR approach of Adrian and Brunnermeier (2011) and the marginal expected shortfall (MES) approach of Acharya et al (2017) and Brownlees and Engle (2012)
11
order VAR model for the 119925 119961 120783 vector of endogenous variables 119937120061 The VAR(119953) model is specified
as follows
119937120061 = 120630119963 + 120509120051119937120061minus120051120057120051=120783 + 120634120061 (1)
where 120630 is 119925119961120783 vector of intercepts 120509120051 is 119925 119961 119925 coefficient matrix 120057 is the lag order and the
residuals 120634120050120061 ~ 119946119946119941 120782 120634119946119946 so that 120634120061 ~ 120782 120634 where 120634 is positive definite covariance matrix
The optimal VAR lag order selected by Schwarz Information Criterion (SIC) for this study is one
(see Appendix 5) Using the Woldlsquos Representation Theorem the model in equation (1) is
expressed as an infinite order vector moving average representation given by
119937120061 = 120634119957 + 120495120783120518119853minus120783 + 120495120784120518119853minus120784 + hellip = 120495120051120634120061minus120051infin120051=120782 (2)
where 120495120782 = 119920119925 120495120051 = 120509120051 120051 = 120783 120784 hellip and 119920119925 stands for an 119925 times 119925 identity matrix in which all
the principal diagonal elements are ones and all other elements are zeros
The network approach of Diebold and Yilmaz (2009) requires that after estimating the underlying
VAR model the forecast error variance decompositions (FEVDs) are then generated and used to
build linkage measures In this study the interest is in the shocks to the disturbances 120634119947120061 in the
country-specific equations Hence following Pesaran and Shin (1998) Diebold and Yilmaz (2016)
and Greenwood-Nimmo et al (2015) this study adopts the order-invariant generalized forecast error
variance decompositions (GFEVDs) defined as
119918119917119916119933119915 119937119946120061 120634119947120061 119919 = 119941119946119947119944119919
= 120648120634119947119947minus120783 119942119946
prime120495119945120506120634119942119947 sup2119919minus120783119945=120782
119942119946prime120495119945120506120634120495119945
prime 119942119946 119919minus120783119945=120782
(3)
where 119894 119895 = 1 hellip 119873 119867 = 1 2 hellip is the forecast horizon 119890119894 119890119895 is 119873 119909 1 selection vector whose i-th
element (j-th element) is unity with zeros elsewhere Θℎ is the coefficient matrix multiplying the h-
lagged shock vector in the infinite moving-average representation of the non-orthogonalized VAR
120634 is the covariance matrix of the shock vector in the non-orthogonalized VAR and 120648120634119947119947 is the j-th
diagonal element of 120634 (ie the standard deviation of 120576119895 ) We adopted a maximum forecast horizon
of 16 quarters in order to ensure that the long-run results are better captured It must be stressed that
the choice of GFEVDs for this study rather than the orthogonalized forecast error variance
decompositions (OFEVDs) of Diebold and Yilmaz (2009) is particularly based on the fact that the
12
OFEVDs depend on the reordering of the variables in the system such that once the order of
variables in the VAR is reshuffled a different outcome results
Diebold and Yilmaz (2014) explain that shocks are rarely orthogonal in the GFEVD environment
so that sums of forecast error variance contributions are not necessarily unity that is row sums of
the generalized variance decomposition matrix 119863119892119867 are not necessarily unity This renders the
interpretation of the GFEVDs complicated Thus to restore a percentage interpretation of the
GFEVDs this study follows Diebold and Yilmaz (2014) to define the normalized GFEVDs
(NGFEVDs) given by2
119863 119892 = 119889 119894119895119892 where 119889 119894119895
119892=
119889119894119895119892
119889119894119895119892119873
119895=1
119889119894119895119892
= 119918119917119916119933119915 119937119946120061 120634119947120061 119919 (4)
By construction 119889 119894119895119892119873
119895=1 = 1 and 119889 119894119895119892119873
119894 119895=1 = 119873 so that the total sum of the generalized forecast
error variance share of each variable is normalized to 100
33 Construction of the Generalized Trade Linkage Measures (GTLMs)
The various trade linkage measures that are relevant for the ensuing analysis are defined in this sub-
section The intuition behind this framework is quite simple Variance decomposition permits the
splitting of the forecast error variances of each variable in the VAR system into parts attributable to
the various system shocks By so doing it becomes easy to answer the question What fraction of
the h-step-ahead error variance in forecasting 119937120783119957 is due to shocks to 119937120783119957 Shocks to 119937120784119957 Similarly
what fraction of the h-step-ahead error variance in forecasting 119937120784119957 is due to shocks to 119937120783119957 Shocks
to 119937120784119957 And in general what fraction of the h-step-ahead error variance in forecasting 119937119947119957 is due to
shocks to 119937119946119957 119946 = 120783 120784 hellip 119925 Thus the approach marries VAR variance decomposition theory and
network topology theory by recognizing that variance decompositions of VARs form networks and
also characterizing linkages in those variance decomposition networks This in turn characterizes
trade linkages of the variables in our VAR system This is the intuition behind this framework
which we now exploit in the ensuing analysis Diebold and Yilmaz (2015) authoritatively document
this framework and its relation to network theory
2 In what follows and without loss of generality the superscript 119867 is dropped whenever it is not needed for clarity so
that 119863119892119867 and 119889119894119895119892119867
are simply written as 119863119892 and 119889119894119895119892
respectively
13
Table 1 Trade Linkage Schematic
Variables 119937120783 119937120784 ⋯ 119937119925 From Others
119937120783
⋯ 1198891119895
119873
119895=1 119895 ne 1
119937120784
⋯ 1198892119895
119873
119895=1 119895 ne 2
⋮ ⋮ ⋮ ⋱
⋮ ⋮
119937119925
⋯ 119889119873119895
119873
119895=1 119895 ne 119873
To Others
1198891198941
119873
119894=1
119894 ne 1
1198891198942
119873
119894=1
119894 ne 2
⋯ 119889119894119873
119873
119894=1
119894 ne N
1
119873 119889119894119895
119873
119894 119895=1 119894 ne 119895
Source Adapted from Diebold and Yilmaz (2014) and Ogbuabor et al (2016) Note For
simplicity each time series variable in this table 119937119947119957 is written as 119937119947 119947 = 120783 120784 hellip 119925
To construct the GTLMs for this study let us denote the H-step ahead NGFEVDs for the 119925 119961 120783
vector of endogenous variables 119937120061 obtained from equation (4) by 119941119946119947 By cross-tabulating 119941119946119947 the
trade linkage table shown in Table 1 is formed The sum of each row in Table 1 is normalized to
100 in line with equation (4) This table is now used to define the various GTLMs and their
relationships The diagonal entries in Table 1 measure own variance shares (or own-effect) while
the off-diagonal entries measure variance shares arising from shocks to other variables in the
system and are therefore referred to as pairwise directional linkage Accordingly the own-effect
(119919119947) also known as the heatwave is defined as
119919119947 = 119941119947119947 (5)
The total cross-variable variance share (119917119947) captures the spillovers from all other variables to 119937119947120061 as
fractions of the H-step-ahead error variance in the forecasts of 119937119947120061 resulting from 119937119946120061 where
119946 = 120783 120784 hellip 119925 and 119946 ne 119947 This measures the total directional linkage from other variables
(countries) in the system (ie the from-effect) to Zj120009 This means that the from-effect can be used to
capture the role each individual economy in the system plays in a given African economy and it is
14
computed in this study by aggregating the spillovers from all the economies in the system to a
given African economy across all horizons Hence the economy contributing the highest of such
aggregate spillover is deemed to play a dominant role in the particular African economy This study
therefore defines 119917119947 as
119917119947 = 119941119947119946119925119946=120783119946ne119947 (6)
By construction 119919119947 + 119917119947 = 120783 forall 119947
This study also defines the total spillover or total contributions of 119937119947120061 to all other variables (denoted
by 119931119947) as
119931119947 = 119941119946119947119925119946=120783119946ne119947 (7)
By construction 119931119947 measures the total directional linkage from 119937119947120061 to other variables in the system
(ie the to-effect) In other words the to-effect measures the directional linkage from a given
economy (for instance a given African economy) to other economies in the system thereby
showing the impact or influence of that particular African economy on other economies in the VAR
system The net directional linkage (or simply net-effect) of 119937119947120061 is therefore defined as
119925119947 = 119931119947 minus 119917119947 (8)
Since there are N economies in the VAR system it also follows that there are 2N total directional
trade linkages N capturing the total shocks transmitted to others (ie spillover to) and N capturing
the total shocks received from others (ie spillover from) These measures are aptly reflect the
bilateral trade patterns between the African economies and other economies in the system since
they mirror the total exports and total imports for each of the N economies in the system and will
therefore facilitate deeper understanding of the inter-linkages among the African economies We
shall utilize the net-effects to establish the net transmittersreceivers of shocks in the system over
time By construction 119925119947119925119947=120783 = 120782
The most aggregative (non-directional) trade linkage measure in this study which will be used to
evaluate the degree of trade linkage Africa the BRICS and the rest of the global economy is known
as the total trade linkage index (119931119931119923119920) or total-effects and it is defined as
119931119931119923119920 = 120783
119925 119917119947
119925119947=120783 =
120783
119925 119931119947
119925119947=120783 (9)
15
This measure captures the grand total of the off-diagonal elements in Table 1 that is the sum of the
―From Others column or ―To Others row There is only one total trade linkage measure which is
analogous to total global imports or total global exports since the two are identical
An important objective of this study is to determine which of the countries among the BRICS and
in the rest of the global economy exert the most dominant trade influence on African economies
and therefore have the potential to spread trade shocks to Africa To achieve this objective this
study defines two indices dependence and influence indices These indices are necessary to
determine the dependence of the j-th variable (or j-th economy) on external shocks and the
influence of the j-th variable (or j-th economy) on the system as a whole The dependence index is
defined as
119926119947119919 =
119917119947
119919119947+ 119917119947 forall 119947 = 120783 120784 hellip 119925 (10)
where 120782 le 119926119947119919 le 120783 This index expresses the relative importance of external shocks for the j-th
economy in the VAR system such that if 119926119947119919 rarr 120783 then conditions in the j-th economy is open
deeply interlinked and sensitive to external conditions but if 119926119947119919 rarr 120782 then the j-th economy is less
sensitive to external shocks
Similarly the influence index is expressed as
119920119947119919 =
119925119947
119931119947+119917119947 forall 119947 = 120783 120784 hellip 119925 (11)
where minus120783 le 119920119947119919 le 120783 For a given horizon H the j-th economy is a net receiver of trade shocks if
minus120783 le 119920119947119919 lt 0 that is if the index has a negative value a net transmitter of trade shocks if 120782 lt
119920119947119919 le 120783 that is if the index takes a positive value and neither a net receiver or transmitter of trade
shocks if IjH = 0 Thus the influence index measures the extent to which the j-th economy in the
system influences or is influenced by external shocks Overall the coordinate pair (119926119947119919 119920119947
119919) in the
dependence-influence space provides a good representation of the j-th economylsquos role in global real
activities A priori expectation is that African economies which are relatively small open
economies would be located close to the point (1-1) while overwhelmingly open but highly
influential and dominant economies like the USA China and Japan would be located close to the
point (11)
4 Results and Discussions
16
We started this empirical analysis by examining the time series properties of the data The Phillips-
Perron unit root tests showed that all the series are I(1) for all the countries (Please see Appendix 3
which reports for total trade of all the countries To conserve space we do not report separately for
exports and imports since the patterns are similar that is all the series are I(1)) However the test
for long-run or equilibrium relationship using the Johansen System Cointegration test showed that
both the Trace and Maximum Eigenvalue statistics returned full rank indicating that the data are
not cointegrated (Please see Appendix 4 which reports the Trace test in Panel 1 and the Max-
Eigenvalue test in Panel 2) Therefore a VAR in first differences was estimated rather than a vector
error correction model (VECM) Thereafter the GTLMs were computed for all forecasting
horizons 119867 = 12 hellip 16 The average values of the GTLMs over all the horizons are reported
since Diebold and Yilmaz (2016) had shown that the measures follow similar patterns regardless of
the choice of window lengths and forecast horizons As earlier stated we set the maximum forecast
horizon at 16 quarters in order to capture the long-run results better As a robustness check we also
estimated the underlying model and computed the GTLMs separately for exports and imports As
shall soon see the results follow the same qualitative pattern
41 Measuring the degree of trade linkage between Africa the BRICS and the rest of the
global economy
The first specific objective of this study is to measure the degree of trade linkage between African
economies the BRICS and the rest of the global economy In addition we examine here how this
index changed from the short-run (ie from horizon 1) through the long-run (ie until horizon 16)
To achieve this objective we estimated the underlying VAR model of equation (1) and used the
NGFEVDs distilled from this estimation based on equation (4) to compute the Total Trade Linkage
Index (TTLI) based on equation (9) We report the plots of this index in Figure 1 based on the total
trade data while the plots based on the exports and imports data are reported in Appendix 5 This
figure shows how the most aggregated generalized trade linkage measure in this study evolved from
the short-run through the long-run We find that the index rose smoothly from 84 at horizons 1
and 2 to 90 at horizons 15 and 16 In other words the total trade linkage index is higher in the
long-run than in the short-run indicating African economies become more interlinked in the long-
run as the business cycles become more synchronized This finding is consistent with the trend of
global macroeconomic interlinkage reported by Greenwood-Nimmo et al (2015) Diebold and
Yilmaz (2016) and Ogbuabor et al (2018) that the ongoing globalization process is engendering
17
more significant comovement in industrial production fluctuations Furthermore we find that the
TTLI recorded an average value of 87 which shows that the trade linkage between Africa the
BRICS and the rest of the global economy is quite substantial
As shown in Appendix 5 the patterns in this index remained qualitatively unchanged when the
underlying model is estimated using the exports and imports data respectively However a closer
look at panels 1 and 2 of this Appendix reveals that the index recorded higher values across all the
horizons for exports than for imports This aptly captures the pattern of trade between Africa and
the rest of the world which shows that Africalsquos export trade in dollar terms at least in the last
decade is higher than its import trade (see table 1 in appendix 1 of African Trade Statistics
Yearbook 2017) This finding clearly indicates that Africa may be vulnerable to trade shocks
which underlines the motivation behind this study
Figure 1 Total Trade Linkage Index
Source Authors Notes The Total trade linkage Index reported here is the most aggregated non-
directional trade linkage index computed for all horizons following equation (9) Notice that the index rises
smoothly from the short-run (ie horizon 1) towards the long-run (ie horizon 16) indicating that African
economies become more interlinked in the long-run
42 Determining the BRICS countries and other countries in the rest of the world that are
dominating Africarsquos trade and therefore have the potential to spread trade shocks to Africa
An important objective of this study is to determine the BRICS member countries and other
countries in the rest of the world that are dominating Africalsquos trade and therefore have the potential
to spread trade shocks to Africa To achieve this objective we report the from-effect trade linkage
from all the economies in the system to each African economy following equation (6) We include
the heatwave (or own-effect) of equation (5) in this table so that it sums up to 100 We aggregate
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
2
Introduction
Intra-Africalsquos trade performance (import and export) is very infinitesimal compared to the rest of
the world (African Trade Statistics Yearbook 2017) The same trend reflects the characteristics of
African countries intra and extra - African trade For instance Nigerialsquos Intra-Africa import trade is
far below its trade with the rest of the world from 2010 ndash 2016 While Nigerialsquos import from the
rest of the world averaged US$4264277 million or 948 per cent during the same period it was
US$229385 million or 52 per cent within Africa Similarly Nigerialsquos intra-African export trade
averaged US$1046565 million or 120 per cent of total export during 2010 to 2016 while the
export to the rest of the world averaged US$8124003 or 880 per cent of total export The low
intra-Africa trade may have necessitated the move for Africalsquos trade integration to fast-track its
development Little wonder Free Trade Agreements (FTAs) have become important elements in
economic integration amongst most countries of the world It provides avenues for nations to enter
into bilateral or regional trade agreements to eliminate tariffs and non-tariff barriers to trade among
member countries African is currently establishing the African Continental Free Trade Area
(AfCFTA) treaty to create a single continental market for goods and services in member nations of
the African Union (AU) with free movement of business persons and investments using a single
currency
The AfCFTA is gathering momentum because globalization remains a contentious subject due to
the mixed benefits and opportunities it presents to participating countries In particular it makes
countries vulnerable to external shocks Granted that shocks often occur and granted further that
African countries largely have significant trade linkage with the rest of the world it is important to
model the sources of foreign influence on domestic economies particularly the BRIC (Brazil
Russia India and China) economies whose trade linkage with Africa is enormous In fact not only
does Africa trade significantly with the BRIC countries the latter also represent a model of
economic development exemplified by strong economic growth and an enormous capacity to
compete in a globalized world Besides trade linkage has been adjudged an important feature of
global economic integration between countries However following literature there is no common
view on whether intense global trade linkages would lead to more or less business cycle
harmonization Again World Banklsquos Global Economic Prospects report posits that growth in
emerging market and developing economies is expected to remain flat in 2019 Growth in many
other economies is also anticipated to decelerate with risks growing that growth could be even
3
weaker than anticipated This phenomenon may invariably worsen the growth prospects of
developing economies in the event of trade shocks
Broadly this study therefore examines trade shock propagation between Africa the BRIC
economies and the rest of the global economy Specifically the study seeks to (i) measure the
degree of trade linkage between Africa the BRIC and the rest of the global economy (ii) determine
the BRIC countries and other countries in the rest of the world that are dominating Africalsquos trade
and therefore have the potential to spread trade shocks to Africa and thus minimize its growth (iii)
determine the African economies that are most susceptible to trade shocks originating from the
BRIC countries and the rest of the global economy (iv) determine other trade blocs outside the
BRIC that exert dominant influence on Africa and therefore have the prospects of spreading trade
shocks to it
To achieve the above objectives the paper employs the normalized generalized forecast error
variance decompositions (NGFEVDs) distilled from an underlying global vector autoregressive
model (Global VAR) to build measures of trade linkage and shocks transmission between Africa
the BRICS countries and the rest of the global economy over the period 1970Q1 ndash 2017Q4 The
underlying model is estimated using the total trade statistics of the countries The time series
properties of the data are examined using the Augmented Dickey-Fuller (ADF) and Phillips-Perron
unit root tests as well as the Johansen cointegration test in order to determine the form of the
model to be estimated If the variables are overwhelmingly stationary at levels then a basic VAR
would be appropriate However if the variables are nonstationary then a VAR in first differences
becomes tenable in the absence of equilibrium relationship otherwise a vector error correction
model (VECM) would be estimated A key innovation in this study would be the use of the
NGFEVDs to build various measures of trade linkages between Africa and the rest of the global
economy at different levels of aggregation The generalized nature of the FEVDs means that they
would be invariant to the reordering of the variables in the system The choice of the global VAR
framework is mainly due to its ability to explicitly model the source of foreign influences on
domestic economies in Africa and the contributions of these domestic economies to the rest of the
world in a transparent and coherent manner It allows for the impact of shocks to be consistently
quantified aggregated and assessed so that interactions between countriesregions can be analyzed
Simply put our proposed framework measures both the direction and the strength of linkages
4
among entities in the system while identifying systemically important or vulnerable entities within
the system Several studies in the literature have exploited this framework in the study of
macroeconomic linkages with great success Examples include Pesaran Schuermann and Weiner
(2004) Dees di Mauro Pesaran and Smith (2007) Dees Holly Pesaran and Smith (2007) and
more recently Greenwood-Nimmo Nguyen and Shin (2015) We are adopting same framework
using some selected African economies and since the data for the study are quarterly the measures
of trade linkages would be computed over a 24-period horizon so that the long-run results are
better captured
The rationale for assessing the impact of trade between Africa the BRIC and the rest of the world is
to bring to the fore the potential trade shock impact between Africa and other trade blocks with
which it has significant trade linkages even as it makes concerted efforts towards signing of the
African Continental Free Trade Area (AfCFTA) through a treaty by all the countries concerned
Nigeria which is Africalsquos largest economy by size of GDP is currently understudying the tenets of
AfCFTA treaty preparatory to signing it The AfCFTA is to create a single continental market for
goods and services in member nations of the African Union (AU) with free movement of business
persons and investments using a single currency The AfCFTA is intended to be the external cordon
protecting the African countrieslsquo domestic markets Thus while AU member states are signing the
treaty it is important to be aware of the potential trade shocks that may emanate from the
significant trade linkages between Africa and the rest of the world
The rest of the paper is organized as follows Section 2 undertakes a brief review of relevant
literature Section 3 presents the data and methodology detailing some preliminary data analysis
while Section 4 discusses the empirical results Section 5 concludes the study and provides some
policy implications that the patterns of cross-country trade shock spillovers obtained in this study
would likely influence Africalsquos move from globalization to regional integration or multi-polarity in
the years to come
2 Brief Review of Literature
The impact of trade shocks in large advanced countries on emerging market economies mdash dubbed
spillovers mdash is hotly debated in global and national policy circles Kose and Reizman (1999)
examined the role of external shocks in explaining macroeconomic fluctuations in African
5
countries They constructed a quantitative stochastic dynamic multi-sector equilibrium model of a
small open economy calibrated to represent a typical African economy In their framework external
shocks consist of trade shocks modeled as fluctuations in the prices of exported primary
commodities imported capital goods and intermediate inputs and a financial shock modeled as
fluctuations in the world real interest rate Results from their analysis indicated that while trade
shocks accounted for roughly 45 percent of economic fluctuations in aggregate output financial
shocks play only a minor role They also found that adverse trade shocks induce prolonged
recessions
Ccedilakir and Kabundi (2011) examined the trade linkages between South Africa and the BRIC (Brazil
Russia India and China) countries They applied the global vector autoregressive model (global
VAR) to investigate the degree of trade linkages and shock transmission between South Africa and
the BRIC countries over the period 1995Q1-2009Q4 The model contained 32 countries and had
two different estimations the first one consists of 24 countries and one region with the 8 countries
in the euro area treated as a single economy and the second estimation contains 20 countries and
two regions with the BRIC and the euro area countries respectively treated as a single economy
The results suggest that trade linkages exist between our focus economies however the magnitude
differs between countries Shocks from each BRIC country are shown to have considerable impact
on South African real imports and output
Greenwood-Nimmo et al (2015) developed a technique to evaluate macroeconomic connectedness
among entities in sophisticated multi-country and global macroeconomic models Their
methodology is highly adaptable and may be applied to any model with an approximate VAR
representation They applied their technique to a global vector autoregressive model containing 169
macroeconomic and financial variables for 25 countries They derived vivid representations of the
connectedness of the system and found that the US the Eurozone and the crude oil market exert a
dominant influence over conditions in the global macroeconomy and that China and Brazil are also
globally significant economies Recursive analysis over the period of the global financial crisis
shows that shocks to global equity markets are rapidly and forcefully transmitted to real trade flows
and real GDP
Pesaran et al (2004) posit that financial institutions are ultimately exposed to macroeconomic
fluctuations in the global economy and thus they built a compact global model capable of
6
generating forecasts for a core set of macroeconomic factors (or variables) across a number of
countries Their model explicitly allows for the interdependencies that exist between national and
international factors Individual region-specific vector error-correcting models were estimated in
which the domestic variables were related to corresponding foreign variables constructed
exclusively to match the international trade pattern of the country under consideration The
individual country models were then linked in a consistent and cohesive manner to generate
forecasts for all of the variables in the world economy simultaneously Their global model was
estimated for 25 countries grouped into 11 regions using quarterly data over 1979Q1ndash1999Q1 The
degree of regional interdependencies was investigated via generalized impulse responses where the
effects of shocks to a given variable in a given country on the rest of the world were provided The
model was then used to investigate the effects of various global risk scenarios on a banklsquos loan
portfolio
Dees et al (2007) presented a quarterly global model combining individual country vector error-
correcting models in which the domestic variables were related to the country-specific foreign
variables Their global VAR (GVAR) model was estimated for 26 countries the Euro area being
treated as a single economy over the period 1979ndash2003 They provided a theoretical framework
where the GVAR was derived as an approximation to a global unobserved common factor model
Using the average pair-wise cross-section error correlations their GVAR approach was shown to be
quite effective in dealing with the common factor interdependencies and international co-
movements of business cycles They developed a sieve bootstrap procedure for simulation of the
GVAR as a whole which is then used in testing the structural stability of the parameters and for
establishing bootstrap confidence bounds for the impulse responses Finally in addition to
generalized impulse responses their paper considered the use of the GVAR for structurallsquo impulse
response analysis with focus on external shocks for the Euro area economy particularly in response
to shocks to the US
In another study Dees et al (2007) focused on testing the long run macroeconomic relations for
interest rates equity prices and exchange rates suggested by arbitrage in financial and
goods markets They used the global vector autoregressive (GVAR) model to test
for long run restrictions in each countryregion conditioning on the rest of the
world They employed Bootstrapping to compute both the empirical distribution of
7
the impulse responses and the log-likelihood ratio statistic for over-identifying
restrictions Their paper also examined the speed with which adjustments to the
long run relations would take place via the persistence profiles They found strong evidence
in favour of the uncovered interest parity (UIP) and to a lesser extent the Fisher equation across a
number of countries but their results for the purchasing power parity (PPP) were much weaker
Also the transmission of shocks and subsequent adjustments in financial markets were much faster
than those in goods markets
Ogbuabor et al (2016) examined the real and financial connectedness of selected African
economies with the global economy using a network approach They found that the connectedness
of African economies with the global economy is quite sizable with the global financial crisis
increasing the connectedness measures above their pre-crisis levels Their results show that US
EU and Canada dominate Africalsquos equity markets while China India and Japan dominate Africalsquos
real activities Their results suggest that African economies are predominantly small open
economies deeply interconnected but systemically unimportant and vulnerable to headwinds
emanating from the dominant economies in the overall global economy
Lubik and Teo (2003) found that world interest rate shocks are the main driving forces of business
cycles in small open economies while terms of trade shocks are not Thus they challenged the
existing results on the contribution of terms of trade and world interest rate shocks to output
fluctuations in small open economies which had been found to range from less than 10 per cent to
almost 90 per cent They argue that an identification problems lies at the heart of existing vastly
different results They overcame this by estimating a DSGE model using a structural Bayesian
estimation approach applying their methodology to five developed and developing economies The
approach enabled them to efficiently exploit cross-equation restrictions implied by the structural
model
Huidrom et al (2017) provide empirical estimates of the cross border spillovers from Seven largest
emerging market economies (China India Brazil Russia Mexico Indonesia and Turkey) using a
Bayesian vector autoregression model Their results indicate that first spillovers from EM7 are
sizeable a 1 percentage point increase in EM7 growth is associated with a 09 percentage point
increase in growth in other emerging and frontier markets and a 06 percentage point increase in
world growth at the end of three years Second sizeable as they are spillovers from EM7 are still
8
smaller than those from G7 countries (Group of Seven of advanced economies) Specifically
growth in other emerging and frontier markets and the global economy would increase by one-half
to three times more due to a similarly sized increase in G7 growth Third among the EM7
spillovers from China are the largest and permeate globally Their study was based on the
phenomenon that their growth could have significant cross-border spillovers based on their size and
integration The seven largest emerging market economies - China India Brazil Russia Mexico
Indonesia and Turkey - was adjudged to constitute more than one-quarter of global output and
more than half of global output growth during 2010ndash15
This study contributes to this emerging literature in several ways First it quantifies the degree of
trade linkage between Africa the BRIC and the rest of the global economy Two apart from
exposing the patterns of trade shock spillovers between Africa and the rest of the world using total
trade statistics this study also disaggregates the trade shock spillovers using exports and imports
statistics This provides deeper insights into the trade linkages between Africa and the rest of the
world Besides the study extended the empirical method by constructing trade linkage measures in
a coherent and transparent manner for ease of replication
3 Data and Methodology
The data for this study consists of the log of exports imports and total trade for the period 1990Q1-
2016Q4 The choice of this period is based on data availability for some of the African economies
and for Russia whose trade data in the World Development Indicators started in 1990 The BRICS
economies in this study are Brazil Russia India China and South Africa The selected African
economies included in this study are DR Congo Egypt Ghana Morocco Nigeria Tanzania and
Tunisia while the rest of the global economy is accounted for by Australia Canada the Eurozone
(EU) Indonesia Japan Malaysia United States of America (USA) and United Kingdom (UK)
These economies account for the larger chunk of Africalsquos trade while the selected African
economies together with South Africa account substantially for Africalsquos GDP The entire data were
taken from the World Development Indicators (WDI) based on the following indicator names
exports of goods and services (constant 2010 US$) as measure for exports imports of goods and
services (constant 2010 US$) as measure for imports and exports plus imports as measure for total
trade (or simply trade)
9
To ensure that there are enough observations for the analysis the data were converted from annual
to quarterly using Eviewlsquos quadratic match average option This is in line with the methodology
used in compiling the Global VAR database used by Greenwood-Nimmo Nguyen and Shin (2015)
the connectedness of the global economy Other studies that have successfully used the same
methodology are Ogbuabor Eigbiremolen Aneke and Manasseh (2018) and Ogbuabor Orji
Aneke and Erdene-Urnukh (2016) Also to reduce noise and ensure uniform scaling the entire data
were converted into indices (2010Y = 100) and logged prior to estimation Appendix 1 summarizes
the descriptive statistics of the data based on the log transformation of the data Focusing on Panel 1
which reports for total trade we find that among the African economies Ghana DR Congo and
Tanzania are the most volatile as seen from their standard deviations while China India and Russia
are the most volatile among the BRICS economies Among all the countries in the sample Japan
Canada and UK are the least volatile and also recorded the highest mean values The maximum and
minimum values do not suggest the presence of outliers in the data The time series plots of the data
for all the economies in the sample are presented in Appendix 2 based on the log transformation of
the data A close examination of the plots show close comovement among all the countries
indicating that the series track themselves closely The implication of this is that the data may be
cointegrated which will in turn affect the form of the underlying model to be estimated in this
study Thus we shall subject the series to cointegration test as part of the empirical procedures in
this study
31 The Choice of Diebold and Yilmaz (2009) Network Framework
A number of methodologies have been employed in the study of macroeconomic linkages among
entities in the global economy particularly in the study of shock propagation among countries For
instance cross-country correlations-based measures have been used to characterize macroeconomic
linkages among countries (Kehoe et al 1995 Kose et al 2003 Bollerslev 1990 Engle et al
1990 Mantegna 1999 Tumminello et al 2005 Taylor 2007 Gray amp Malone 2008 Engle 2009
Engle amp Kelly 2012) The pitfalls of this approach are twofold namely correlation is simply a
pairwise measure of association and it is non-directional This means that correlation-based
approach cannot handle such questions as ―what is the degree of trade linkage between Africa the
BRICS and the rest of the global economy Unlike the correlation-based measures of
macroeconomic linkage the network approach of Diebold and Yilmaz (2009) is non-pairwise yet
10
directional Granger Causality measures have also been used characterize networks so that the
macroeconomic linkages among entities in the global economy can be described and understood
(Caraiani 2013 Hiemstra amp Jones 1994 Dahlhans amp Eichler 2003 Shojaie amp Michailidis 2010
Billio et al 2012) The main weakness of the Granger Causality approach is that it captures only
pairwise relations and may not be useful in answering important questions like ―what is the degree
of trade linkage between Africa the BRICS and the rest of the global economy1
Furthermore as noted by Diebold and Yilmaz (2016) these alternative methodologies generally
dwell exclusively on testing rather than measurement and estimation of macroeconomic linkages
which are the key issues in this paper This study therefore follows the network approach of
Diebold and Yilmaz (2009) based on its ability to transparently use the size and direction of shocks
to build both directional and non-directional trade linkage measures over a given forecast horizon
According to Ogbuabor et al (2016) studies using this approach have four common features
namely (i) they are generally based on connectedness measures distilled from forecast error
variance decompositions (FEVDs) of an approximating vector autoregressive (VAR) model (ii)
they measure the direction and strength of linkages among entities in the system (iii) they can
identify systemically important entities in the system and (iv) they can study the dynamic nature of
shock propagation among entities in the system In what follows the underlying VAR model for
this study and the construction of the generalized trade linkages measures (GTLMs) are presented
to guide the ensuing analysis
32 Model Specification The Vector Autoregression (VAR) Model
The broad objective of this study is to examine the propagation of trade shock between Africa the
BRICS economies and the rest of the global economy Let 119937120061 be the log of total trade for all the
countries selected for this study so that 119937119947120061 stands for the logged total trade of the j-th country in the
system with 119947 = 120783 120784 hellip 119925 and 119925 is the number of countries selected for the study (which is 20)
Following Diebold and Yilmaz (2009) the trade linkage measures for this study are based on the
normalized generalized forecast error variance decompositions (NGFEVDs) of an underlying 119953-th
1 Other techniques have also been used in the literature for the study of macroeconomic linkages Ogbuabor et al
(2018) provides an overview of such alternative methodologies such as the dynamic latent factor models of Kose et al (2008) and Canova et al (2007) the CoVaR approach of Adrian and Brunnermeier (2011) and the marginal expected shortfall (MES) approach of Acharya et al (2017) and Brownlees and Engle (2012)
11
order VAR model for the 119925 119961 120783 vector of endogenous variables 119937120061 The VAR(119953) model is specified
as follows
119937120061 = 120630119963 + 120509120051119937120061minus120051120057120051=120783 + 120634120061 (1)
where 120630 is 119925119961120783 vector of intercepts 120509120051 is 119925 119961 119925 coefficient matrix 120057 is the lag order and the
residuals 120634120050120061 ~ 119946119946119941 120782 120634119946119946 so that 120634120061 ~ 120782 120634 where 120634 is positive definite covariance matrix
The optimal VAR lag order selected by Schwarz Information Criterion (SIC) for this study is one
(see Appendix 5) Using the Woldlsquos Representation Theorem the model in equation (1) is
expressed as an infinite order vector moving average representation given by
119937120061 = 120634119957 + 120495120783120518119853minus120783 + 120495120784120518119853minus120784 + hellip = 120495120051120634120061minus120051infin120051=120782 (2)
where 120495120782 = 119920119925 120495120051 = 120509120051 120051 = 120783 120784 hellip and 119920119925 stands for an 119925 times 119925 identity matrix in which all
the principal diagonal elements are ones and all other elements are zeros
The network approach of Diebold and Yilmaz (2009) requires that after estimating the underlying
VAR model the forecast error variance decompositions (FEVDs) are then generated and used to
build linkage measures In this study the interest is in the shocks to the disturbances 120634119947120061 in the
country-specific equations Hence following Pesaran and Shin (1998) Diebold and Yilmaz (2016)
and Greenwood-Nimmo et al (2015) this study adopts the order-invariant generalized forecast error
variance decompositions (GFEVDs) defined as
119918119917119916119933119915 119937119946120061 120634119947120061 119919 = 119941119946119947119944119919
= 120648120634119947119947minus120783 119942119946
prime120495119945120506120634119942119947 sup2119919minus120783119945=120782
119942119946prime120495119945120506120634120495119945
prime 119942119946 119919minus120783119945=120782
(3)
where 119894 119895 = 1 hellip 119873 119867 = 1 2 hellip is the forecast horizon 119890119894 119890119895 is 119873 119909 1 selection vector whose i-th
element (j-th element) is unity with zeros elsewhere Θℎ is the coefficient matrix multiplying the h-
lagged shock vector in the infinite moving-average representation of the non-orthogonalized VAR
120634 is the covariance matrix of the shock vector in the non-orthogonalized VAR and 120648120634119947119947 is the j-th
diagonal element of 120634 (ie the standard deviation of 120576119895 ) We adopted a maximum forecast horizon
of 16 quarters in order to ensure that the long-run results are better captured It must be stressed that
the choice of GFEVDs for this study rather than the orthogonalized forecast error variance
decompositions (OFEVDs) of Diebold and Yilmaz (2009) is particularly based on the fact that the
12
OFEVDs depend on the reordering of the variables in the system such that once the order of
variables in the VAR is reshuffled a different outcome results
Diebold and Yilmaz (2014) explain that shocks are rarely orthogonal in the GFEVD environment
so that sums of forecast error variance contributions are not necessarily unity that is row sums of
the generalized variance decomposition matrix 119863119892119867 are not necessarily unity This renders the
interpretation of the GFEVDs complicated Thus to restore a percentage interpretation of the
GFEVDs this study follows Diebold and Yilmaz (2014) to define the normalized GFEVDs
(NGFEVDs) given by2
119863 119892 = 119889 119894119895119892 where 119889 119894119895
119892=
119889119894119895119892
119889119894119895119892119873
119895=1
119889119894119895119892
= 119918119917119916119933119915 119937119946120061 120634119947120061 119919 (4)
By construction 119889 119894119895119892119873
119895=1 = 1 and 119889 119894119895119892119873
119894 119895=1 = 119873 so that the total sum of the generalized forecast
error variance share of each variable is normalized to 100
33 Construction of the Generalized Trade Linkage Measures (GTLMs)
The various trade linkage measures that are relevant for the ensuing analysis are defined in this sub-
section The intuition behind this framework is quite simple Variance decomposition permits the
splitting of the forecast error variances of each variable in the VAR system into parts attributable to
the various system shocks By so doing it becomes easy to answer the question What fraction of
the h-step-ahead error variance in forecasting 119937120783119957 is due to shocks to 119937120783119957 Shocks to 119937120784119957 Similarly
what fraction of the h-step-ahead error variance in forecasting 119937120784119957 is due to shocks to 119937120783119957 Shocks
to 119937120784119957 And in general what fraction of the h-step-ahead error variance in forecasting 119937119947119957 is due to
shocks to 119937119946119957 119946 = 120783 120784 hellip 119925 Thus the approach marries VAR variance decomposition theory and
network topology theory by recognizing that variance decompositions of VARs form networks and
also characterizing linkages in those variance decomposition networks This in turn characterizes
trade linkages of the variables in our VAR system This is the intuition behind this framework
which we now exploit in the ensuing analysis Diebold and Yilmaz (2015) authoritatively document
this framework and its relation to network theory
2 In what follows and without loss of generality the superscript 119867 is dropped whenever it is not needed for clarity so
that 119863119892119867 and 119889119894119895119892119867
are simply written as 119863119892 and 119889119894119895119892
respectively
13
Table 1 Trade Linkage Schematic
Variables 119937120783 119937120784 ⋯ 119937119925 From Others
119937120783
⋯ 1198891119895
119873
119895=1 119895 ne 1
119937120784
⋯ 1198892119895
119873
119895=1 119895 ne 2
⋮ ⋮ ⋮ ⋱
⋮ ⋮
119937119925
⋯ 119889119873119895
119873
119895=1 119895 ne 119873
To Others
1198891198941
119873
119894=1
119894 ne 1
1198891198942
119873
119894=1
119894 ne 2
⋯ 119889119894119873
119873
119894=1
119894 ne N
1
119873 119889119894119895
119873
119894 119895=1 119894 ne 119895
Source Adapted from Diebold and Yilmaz (2014) and Ogbuabor et al (2016) Note For
simplicity each time series variable in this table 119937119947119957 is written as 119937119947 119947 = 120783 120784 hellip 119925
To construct the GTLMs for this study let us denote the H-step ahead NGFEVDs for the 119925 119961 120783
vector of endogenous variables 119937120061 obtained from equation (4) by 119941119946119947 By cross-tabulating 119941119946119947 the
trade linkage table shown in Table 1 is formed The sum of each row in Table 1 is normalized to
100 in line with equation (4) This table is now used to define the various GTLMs and their
relationships The diagonal entries in Table 1 measure own variance shares (or own-effect) while
the off-diagonal entries measure variance shares arising from shocks to other variables in the
system and are therefore referred to as pairwise directional linkage Accordingly the own-effect
(119919119947) also known as the heatwave is defined as
119919119947 = 119941119947119947 (5)
The total cross-variable variance share (119917119947) captures the spillovers from all other variables to 119937119947120061 as
fractions of the H-step-ahead error variance in the forecasts of 119937119947120061 resulting from 119937119946120061 where
119946 = 120783 120784 hellip 119925 and 119946 ne 119947 This measures the total directional linkage from other variables
(countries) in the system (ie the from-effect) to Zj120009 This means that the from-effect can be used to
capture the role each individual economy in the system plays in a given African economy and it is
14
computed in this study by aggregating the spillovers from all the economies in the system to a
given African economy across all horizons Hence the economy contributing the highest of such
aggregate spillover is deemed to play a dominant role in the particular African economy This study
therefore defines 119917119947 as
119917119947 = 119941119947119946119925119946=120783119946ne119947 (6)
By construction 119919119947 + 119917119947 = 120783 forall 119947
This study also defines the total spillover or total contributions of 119937119947120061 to all other variables (denoted
by 119931119947) as
119931119947 = 119941119946119947119925119946=120783119946ne119947 (7)
By construction 119931119947 measures the total directional linkage from 119937119947120061 to other variables in the system
(ie the to-effect) In other words the to-effect measures the directional linkage from a given
economy (for instance a given African economy) to other economies in the system thereby
showing the impact or influence of that particular African economy on other economies in the VAR
system The net directional linkage (or simply net-effect) of 119937119947120061 is therefore defined as
119925119947 = 119931119947 minus 119917119947 (8)
Since there are N economies in the VAR system it also follows that there are 2N total directional
trade linkages N capturing the total shocks transmitted to others (ie spillover to) and N capturing
the total shocks received from others (ie spillover from) These measures are aptly reflect the
bilateral trade patterns between the African economies and other economies in the system since
they mirror the total exports and total imports for each of the N economies in the system and will
therefore facilitate deeper understanding of the inter-linkages among the African economies We
shall utilize the net-effects to establish the net transmittersreceivers of shocks in the system over
time By construction 119925119947119925119947=120783 = 120782
The most aggregative (non-directional) trade linkage measure in this study which will be used to
evaluate the degree of trade linkage Africa the BRICS and the rest of the global economy is known
as the total trade linkage index (119931119931119923119920) or total-effects and it is defined as
119931119931119923119920 = 120783
119925 119917119947
119925119947=120783 =
120783
119925 119931119947
119925119947=120783 (9)
15
This measure captures the grand total of the off-diagonal elements in Table 1 that is the sum of the
―From Others column or ―To Others row There is only one total trade linkage measure which is
analogous to total global imports or total global exports since the two are identical
An important objective of this study is to determine which of the countries among the BRICS and
in the rest of the global economy exert the most dominant trade influence on African economies
and therefore have the potential to spread trade shocks to Africa To achieve this objective this
study defines two indices dependence and influence indices These indices are necessary to
determine the dependence of the j-th variable (or j-th economy) on external shocks and the
influence of the j-th variable (or j-th economy) on the system as a whole The dependence index is
defined as
119926119947119919 =
119917119947
119919119947+ 119917119947 forall 119947 = 120783 120784 hellip 119925 (10)
where 120782 le 119926119947119919 le 120783 This index expresses the relative importance of external shocks for the j-th
economy in the VAR system such that if 119926119947119919 rarr 120783 then conditions in the j-th economy is open
deeply interlinked and sensitive to external conditions but if 119926119947119919 rarr 120782 then the j-th economy is less
sensitive to external shocks
Similarly the influence index is expressed as
119920119947119919 =
119925119947
119931119947+119917119947 forall 119947 = 120783 120784 hellip 119925 (11)
where minus120783 le 119920119947119919 le 120783 For a given horizon H the j-th economy is a net receiver of trade shocks if
minus120783 le 119920119947119919 lt 0 that is if the index has a negative value a net transmitter of trade shocks if 120782 lt
119920119947119919 le 120783 that is if the index takes a positive value and neither a net receiver or transmitter of trade
shocks if IjH = 0 Thus the influence index measures the extent to which the j-th economy in the
system influences or is influenced by external shocks Overall the coordinate pair (119926119947119919 119920119947
119919) in the
dependence-influence space provides a good representation of the j-th economylsquos role in global real
activities A priori expectation is that African economies which are relatively small open
economies would be located close to the point (1-1) while overwhelmingly open but highly
influential and dominant economies like the USA China and Japan would be located close to the
point (11)
4 Results and Discussions
16
We started this empirical analysis by examining the time series properties of the data The Phillips-
Perron unit root tests showed that all the series are I(1) for all the countries (Please see Appendix 3
which reports for total trade of all the countries To conserve space we do not report separately for
exports and imports since the patterns are similar that is all the series are I(1)) However the test
for long-run or equilibrium relationship using the Johansen System Cointegration test showed that
both the Trace and Maximum Eigenvalue statistics returned full rank indicating that the data are
not cointegrated (Please see Appendix 4 which reports the Trace test in Panel 1 and the Max-
Eigenvalue test in Panel 2) Therefore a VAR in first differences was estimated rather than a vector
error correction model (VECM) Thereafter the GTLMs were computed for all forecasting
horizons 119867 = 12 hellip 16 The average values of the GTLMs over all the horizons are reported
since Diebold and Yilmaz (2016) had shown that the measures follow similar patterns regardless of
the choice of window lengths and forecast horizons As earlier stated we set the maximum forecast
horizon at 16 quarters in order to capture the long-run results better As a robustness check we also
estimated the underlying model and computed the GTLMs separately for exports and imports As
shall soon see the results follow the same qualitative pattern
41 Measuring the degree of trade linkage between Africa the BRICS and the rest of the
global economy
The first specific objective of this study is to measure the degree of trade linkage between African
economies the BRICS and the rest of the global economy In addition we examine here how this
index changed from the short-run (ie from horizon 1) through the long-run (ie until horizon 16)
To achieve this objective we estimated the underlying VAR model of equation (1) and used the
NGFEVDs distilled from this estimation based on equation (4) to compute the Total Trade Linkage
Index (TTLI) based on equation (9) We report the plots of this index in Figure 1 based on the total
trade data while the plots based on the exports and imports data are reported in Appendix 5 This
figure shows how the most aggregated generalized trade linkage measure in this study evolved from
the short-run through the long-run We find that the index rose smoothly from 84 at horizons 1
and 2 to 90 at horizons 15 and 16 In other words the total trade linkage index is higher in the
long-run than in the short-run indicating African economies become more interlinked in the long-
run as the business cycles become more synchronized This finding is consistent with the trend of
global macroeconomic interlinkage reported by Greenwood-Nimmo et al (2015) Diebold and
Yilmaz (2016) and Ogbuabor et al (2018) that the ongoing globalization process is engendering
17
more significant comovement in industrial production fluctuations Furthermore we find that the
TTLI recorded an average value of 87 which shows that the trade linkage between Africa the
BRICS and the rest of the global economy is quite substantial
As shown in Appendix 5 the patterns in this index remained qualitatively unchanged when the
underlying model is estimated using the exports and imports data respectively However a closer
look at panels 1 and 2 of this Appendix reveals that the index recorded higher values across all the
horizons for exports than for imports This aptly captures the pattern of trade between Africa and
the rest of the world which shows that Africalsquos export trade in dollar terms at least in the last
decade is higher than its import trade (see table 1 in appendix 1 of African Trade Statistics
Yearbook 2017) This finding clearly indicates that Africa may be vulnerable to trade shocks
which underlines the motivation behind this study
Figure 1 Total Trade Linkage Index
Source Authors Notes The Total trade linkage Index reported here is the most aggregated non-
directional trade linkage index computed for all horizons following equation (9) Notice that the index rises
smoothly from the short-run (ie horizon 1) towards the long-run (ie horizon 16) indicating that African
economies become more interlinked in the long-run
42 Determining the BRICS countries and other countries in the rest of the world that are
dominating Africarsquos trade and therefore have the potential to spread trade shocks to Africa
An important objective of this study is to determine the BRICS member countries and other
countries in the rest of the world that are dominating Africalsquos trade and therefore have the potential
to spread trade shocks to Africa To achieve this objective we report the from-effect trade linkage
from all the economies in the system to each African economy following equation (6) We include
the heatwave (or own-effect) of equation (5) in this table so that it sums up to 100 We aggregate
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
3
weaker than anticipated This phenomenon may invariably worsen the growth prospects of
developing economies in the event of trade shocks
Broadly this study therefore examines trade shock propagation between Africa the BRIC
economies and the rest of the global economy Specifically the study seeks to (i) measure the
degree of trade linkage between Africa the BRIC and the rest of the global economy (ii) determine
the BRIC countries and other countries in the rest of the world that are dominating Africalsquos trade
and therefore have the potential to spread trade shocks to Africa and thus minimize its growth (iii)
determine the African economies that are most susceptible to trade shocks originating from the
BRIC countries and the rest of the global economy (iv) determine other trade blocs outside the
BRIC that exert dominant influence on Africa and therefore have the prospects of spreading trade
shocks to it
To achieve the above objectives the paper employs the normalized generalized forecast error
variance decompositions (NGFEVDs) distilled from an underlying global vector autoregressive
model (Global VAR) to build measures of trade linkage and shocks transmission between Africa
the BRICS countries and the rest of the global economy over the period 1970Q1 ndash 2017Q4 The
underlying model is estimated using the total trade statistics of the countries The time series
properties of the data are examined using the Augmented Dickey-Fuller (ADF) and Phillips-Perron
unit root tests as well as the Johansen cointegration test in order to determine the form of the
model to be estimated If the variables are overwhelmingly stationary at levels then a basic VAR
would be appropriate However if the variables are nonstationary then a VAR in first differences
becomes tenable in the absence of equilibrium relationship otherwise a vector error correction
model (VECM) would be estimated A key innovation in this study would be the use of the
NGFEVDs to build various measures of trade linkages between Africa and the rest of the global
economy at different levels of aggregation The generalized nature of the FEVDs means that they
would be invariant to the reordering of the variables in the system The choice of the global VAR
framework is mainly due to its ability to explicitly model the source of foreign influences on
domestic economies in Africa and the contributions of these domestic economies to the rest of the
world in a transparent and coherent manner It allows for the impact of shocks to be consistently
quantified aggregated and assessed so that interactions between countriesregions can be analyzed
Simply put our proposed framework measures both the direction and the strength of linkages
4
among entities in the system while identifying systemically important or vulnerable entities within
the system Several studies in the literature have exploited this framework in the study of
macroeconomic linkages with great success Examples include Pesaran Schuermann and Weiner
(2004) Dees di Mauro Pesaran and Smith (2007) Dees Holly Pesaran and Smith (2007) and
more recently Greenwood-Nimmo Nguyen and Shin (2015) We are adopting same framework
using some selected African economies and since the data for the study are quarterly the measures
of trade linkages would be computed over a 24-period horizon so that the long-run results are
better captured
The rationale for assessing the impact of trade between Africa the BRIC and the rest of the world is
to bring to the fore the potential trade shock impact between Africa and other trade blocks with
which it has significant trade linkages even as it makes concerted efforts towards signing of the
African Continental Free Trade Area (AfCFTA) through a treaty by all the countries concerned
Nigeria which is Africalsquos largest economy by size of GDP is currently understudying the tenets of
AfCFTA treaty preparatory to signing it The AfCFTA is to create a single continental market for
goods and services in member nations of the African Union (AU) with free movement of business
persons and investments using a single currency The AfCFTA is intended to be the external cordon
protecting the African countrieslsquo domestic markets Thus while AU member states are signing the
treaty it is important to be aware of the potential trade shocks that may emanate from the
significant trade linkages between Africa and the rest of the world
The rest of the paper is organized as follows Section 2 undertakes a brief review of relevant
literature Section 3 presents the data and methodology detailing some preliminary data analysis
while Section 4 discusses the empirical results Section 5 concludes the study and provides some
policy implications that the patterns of cross-country trade shock spillovers obtained in this study
would likely influence Africalsquos move from globalization to regional integration or multi-polarity in
the years to come
2 Brief Review of Literature
The impact of trade shocks in large advanced countries on emerging market economies mdash dubbed
spillovers mdash is hotly debated in global and national policy circles Kose and Reizman (1999)
examined the role of external shocks in explaining macroeconomic fluctuations in African
5
countries They constructed a quantitative stochastic dynamic multi-sector equilibrium model of a
small open economy calibrated to represent a typical African economy In their framework external
shocks consist of trade shocks modeled as fluctuations in the prices of exported primary
commodities imported capital goods and intermediate inputs and a financial shock modeled as
fluctuations in the world real interest rate Results from their analysis indicated that while trade
shocks accounted for roughly 45 percent of economic fluctuations in aggregate output financial
shocks play only a minor role They also found that adverse trade shocks induce prolonged
recessions
Ccedilakir and Kabundi (2011) examined the trade linkages between South Africa and the BRIC (Brazil
Russia India and China) countries They applied the global vector autoregressive model (global
VAR) to investigate the degree of trade linkages and shock transmission between South Africa and
the BRIC countries over the period 1995Q1-2009Q4 The model contained 32 countries and had
two different estimations the first one consists of 24 countries and one region with the 8 countries
in the euro area treated as a single economy and the second estimation contains 20 countries and
two regions with the BRIC and the euro area countries respectively treated as a single economy
The results suggest that trade linkages exist between our focus economies however the magnitude
differs between countries Shocks from each BRIC country are shown to have considerable impact
on South African real imports and output
Greenwood-Nimmo et al (2015) developed a technique to evaluate macroeconomic connectedness
among entities in sophisticated multi-country and global macroeconomic models Their
methodology is highly adaptable and may be applied to any model with an approximate VAR
representation They applied their technique to a global vector autoregressive model containing 169
macroeconomic and financial variables for 25 countries They derived vivid representations of the
connectedness of the system and found that the US the Eurozone and the crude oil market exert a
dominant influence over conditions in the global macroeconomy and that China and Brazil are also
globally significant economies Recursive analysis over the period of the global financial crisis
shows that shocks to global equity markets are rapidly and forcefully transmitted to real trade flows
and real GDP
Pesaran et al (2004) posit that financial institutions are ultimately exposed to macroeconomic
fluctuations in the global economy and thus they built a compact global model capable of
6
generating forecasts for a core set of macroeconomic factors (or variables) across a number of
countries Their model explicitly allows for the interdependencies that exist between national and
international factors Individual region-specific vector error-correcting models were estimated in
which the domestic variables were related to corresponding foreign variables constructed
exclusively to match the international trade pattern of the country under consideration The
individual country models were then linked in a consistent and cohesive manner to generate
forecasts for all of the variables in the world economy simultaneously Their global model was
estimated for 25 countries grouped into 11 regions using quarterly data over 1979Q1ndash1999Q1 The
degree of regional interdependencies was investigated via generalized impulse responses where the
effects of shocks to a given variable in a given country on the rest of the world were provided The
model was then used to investigate the effects of various global risk scenarios on a banklsquos loan
portfolio
Dees et al (2007) presented a quarterly global model combining individual country vector error-
correcting models in which the domestic variables were related to the country-specific foreign
variables Their global VAR (GVAR) model was estimated for 26 countries the Euro area being
treated as a single economy over the period 1979ndash2003 They provided a theoretical framework
where the GVAR was derived as an approximation to a global unobserved common factor model
Using the average pair-wise cross-section error correlations their GVAR approach was shown to be
quite effective in dealing with the common factor interdependencies and international co-
movements of business cycles They developed a sieve bootstrap procedure for simulation of the
GVAR as a whole which is then used in testing the structural stability of the parameters and for
establishing bootstrap confidence bounds for the impulse responses Finally in addition to
generalized impulse responses their paper considered the use of the GVAR for structurallsquo impulse
response analysis with focus on external shocks for the Euro area economy particularly in response
to shocks to the US
In another study Dees et al (2007) focused on testing the long run macroeconomic relations for
interest rates equity prices and exchange rates suggested by arbitrage in financial and
goods markets They used the global vector autoregressive (GVAR) model to test
for long run restrictions in each countryregion conditioning on the rest of the
world They employed Bootstrapping to compute both the empirical distribution of
7
the impulse responses and the log-likelihood ratio statistic for over-identifying
restrictions Their paper also examined the speed with which adjustments to the
long run relations would take place via the persistence profiles They found strong evidence
in favour of the uncovered interest parity (UIP) and to a lesser extent the Fisher equation across a
number of countries but their results for the purchasing power parity (PPP) were much weaker
Also the transmission of shocks and subsequent adjustments in financial markets were much faster
than those in goods markets
Ogbuabor et al (2016) examined the real and financial connectedness of selected African
economies with the global economy using a network approach They found that the connectedness
of African economies with the global economy is quite sizable with the global financial crisis
increasing the connectedness measures above their pre-crisis levels Their results show that US
EU and Canada dominate Africalsquos equity markets while China India and Japan dominate Africalsquos
real activities Their results suggest that African economies are predominantly small open
economies deeply interconnected but systemically unimportant and vulnerable to headwinds
emanating from the dominant economies in the overall global economy
Lubik and Teo (2003) found that world interest rate shocks are the main driving forces of business
cycles in small open economies while terms of trade shocks are not Thus they challenged the
existing results on the contribution of terms of trade and world interest rate shocks to output
fluctuations in small open economies which had been found to range from less than 10 per cent to
almost 90 per cent They argue that an identification problems lies at the heart of existing vastly
different results They overcame this by estimating a DSGE model using a structural Bayesian
estimation approach applying their methodology to five developed and developing economies The
approach enabled them to efficiently exploit cross-equation restrictions implied by the structural
model
Huidrom et al (2017) provide empirical estimates of the cross border spillovers from Seven largest
emerging market economies (China India Brazil Russia Mexico Indonesia and Turkey) using a
Bayesian vector autoregression model Their results indicate that first spillovers from EM7 are
sizeable a 1 percentage point increase in EM7 growth is associated with a 09 percentage point
increase in growth in other emerging and frontier markets and a 06 percentage point increase in
world growth at the end of three years Second sizeable as they are spillovers from EM7 are still
8
smaller than those from G7 countries (Group of Seven of advanced economies) Specifically
growth in other emerging and frontier markets and the global economy would increase by one-half
to three times more due to a similarly sized increase in G7 growth Third among the EM7
spillovers from China are the largest and permeate globally Their study was based on the
phenomenon that their growth could have significant cross-border spillovers based on their size and
integration The seven largest emerging market economies - China India Brazil Russia Mexico
Indonesia and Turkey - was adjudged to constitute more than one-quarter of global output and
more than half of global output growth during 2010ndash15
This study contributes to this emerging literature in several ways First it quantifies the degree of
trade linkage between Africa the BRIC and the rest of the global economy Two apart from
exposing the patterns of trade shock spillovers between Africa and the rest of the world using total
trade statistics this study also disaggregates the trade shock spillovers using exports and imports
statistics This provides deeper insights into the trade linkages between Africa and the rest of the
world Besides the study extended the empirical method by constructing trade linkage measures in
a coherent and transparent manner for ease of replication
3 Data and Methodology
The data for this study consists of the log of exports imports and total trade for the period 1990Q1-
2016Q4 The choice of this period is based on data availability for some of the African economies
and for Russia whose trade data in the World Development Indicators started in 1990 The BRICS
economies in this study are Brazil Russia India China and South Africa The selected African
economies included in this study are DR Congo Egypt Ghana Morocco Nigeria Tanzania and
Tunisia while the rest of the global economy is accounted for by Australia Canada the Eurozone
(EU) Indonesia Japan Malaysia United States of America (USA) and United Kingdom (UK)
These economies account for the larger chunk of Africalsquos trade while the selected African
economies together with South Africa account substantially for Africalsquos GDP The entire data were
taken from the World Development Indicators (WDI) based on the following indicator names
exports of goods and services (constant 2010 US$) as measure for exports imports of goods and
services (constant 2010 US$) as measure for imports and exports plus imports as measure for total
trade (or simply trade)
9
To ensure that there are enough observations for the analysis the data were converted from annual
to quarterly using Eviewlsquos quadratic match average option This is in line with the methodology
used in compiling the Global VAR database used by Greenwood-Nimmo Nguyen and Shin (2015)
the connectedness of the global economy Other studies that have successfully used the same
methodology are Ogbuabor Eigbiremolen Aneke and Manasseh (2018) and Ogbuabor Orji
Aneke and Erdene-Urnukh (2016) Also to reduce noise and ensure uniform scaling the entire data
were converted into indices (2010Y = 100) and logged prior to estimation Appendix 1 summarizes
the descriptive statistics of the data based on the log transformation of the data Focusing on Panel 1
which reports for total trade we find that among the African economies Ghana DR Congo and
Tanzania are the most volatile as seen from their standard deviations while China India and Russia
are the most volatile among the BRICS economies Among all the countries in the sample Japan
Canada and UK are the least volatile and also recorded the highest mean values The maximum and
minimum values do not suggest the presence of outliers in the data The time series plots of the data
for all the economies in the sample are presented in Appendix 2 based on the log transformation of
the data A close examination of the plots show close comovement among all the countries
indicating that the series track themselves closely The implication of this is that the data may be
cointegrated which will in turn affect the form of the underlying model to be estimated in this
study Thus we shall subject the series to cointegration test as part of the empirical procedures in
this study
31 The Choice of Diebold and Yilmaz (2009) Network Framework
A number of methodologies have been employed in the study of macroeconomic linkages among
entities in the global economy particularly in the study of shock propagation among countries For
instance cross-country correlations-based measures have been used to characterize macroeconomic
linkages among countries (Kehoe et al 1995 Kose et al 2003 Bollerslev 1990 Engle et al
1990 Mantegna 1999 Tumminello et al 2005 Taylor 2007 Gray amp Malone 2008 Engle 2009
Engle amp Kelly 2012) The pitfalls of this approach are twofold namely correlation is simply a
pairwise measure of association and it is non-directional This means that correlation-based
approach cannot handle such questions as ―what is the degree of trade linkage between Africa the
BRICS and the rest of the global economy Unlike the correlation-based measures of
macroeconomic linkage the network approach of Diebold and Yilmaz (2009) is non-pairwise yet
10
directional Granger Causality measures have also been used characterize networks so that the
macroeconomic linkages among entities in the global economy can be described and understood
(Caraiani 2013 Hiemstra amp Jones 1994 Dahlhans amp Eichler 2003 Shojaie amp Michailidis 2010
Billio et al 2012) The main weakness of the Granger Causality approach is that it captures only
pairwise relations and may not be useful in answering important questions like ―what is the degree
of trade linkage between Africa the BRICS and the rest of the global economy1
Furthermore as noted by Diebold and Yilmaz (2016) these alternative methodologies generally
dwell exclusively on testing rather than measurement and estimation of macroeconomic linkages
which are the key issues in this paper This study therefore follows the network approach of
Diebold and Yilmaz (2009) based on its ability to transparently use the size and direction of shocks
to build both directional and non-directional trade linkage measures over a given forecast horizon
According to Ogbuabor et al (2016) studies using this approach have four common features
namely (i) they are generally based on connectedness measures distilled from forecast error
variance decompositions (FEVDs) of an approximating vector autoregressive (VAR) model (ii)
they measure the direction and strength of linkages among entities in the system (iii) they can
identify systemically important entities in the system and (iv) they can study the dynamic nature of
shock propagation among entities in the system In what follows the underlying VAR model for
this study and the construction of the generalized trade linkages measures (GTLMs) are presented
to guide the ensuing analysis
32 Model Specification The Vector Autoregression (VAR) Model
The broad objective of this study is to examine the propagation of trade shock between Africa the
BRICS economies and the rest of the global economy Let 119937120061 be the log of total trade for all the
countries selected for this study so that 119937119947120061 stands for the logged total trade of the j-th country in the
system with 119947 = 120783 120784 hellip 119925 and 119925 is the number of countries selected for the study (which is 20)
Following Diebold and Yilmaz (2009) the trade linkage measures for this study are based on the
normalized generalized forecast error variance decompositions (NGFEVDs) of an underlying 119953-th
1 Other techniques have also been used in the literature for the study of macroeconomic linkages Ogbuabor et al
(2018) provides an overview of such alternative methodologies such as the dynamic latent factor models of Kose et al (2008) and Canova et al (2007) the CoVaR approach of Adrian and Brunnermeier (2011) and the marginal expected shortfall (MES) approach of Acharya et al (2017) and Brownlees and Engle (2012)
11
order VAR model for the 119925 119961 120783 vector of endogenous variables 119937120061 The VAR(119953) model is specified
as follows
119937120061 = 120630119963 + 120509120051119937120061minus120051120057120051=120783 + 120634120061 (1)
where 120630 is 119925119961120783 vector of intercepts 120509120051 is 119925 119961 119925 coefficient matrix 120057 is the lag order and the
residuals 120634120050120061 ~ 119946119946119941 120782 120634119946119946 so that 120634120061 ~ 120782 120634 where 120634 is positive definite covariance matrix
The optimal VAR lag order selected by Schwarz Information Criterion (SIC) for this study is one
(see Appendix 5) Using the Woldlsquos Representation Theorem the model in equation (1) is
expressed as an infinite order vector moving average representation given by
119937120061 = 120634119957 + 120495120783120518119853minus120783 + 120495120784120518119853minus120784 + hellip = 120495120051120634120061minus120051infin120051=120782 (2)
where 120495120782 = 119920119925 120495120051 = 120509120051 120051 = 120783 120784 hellip and 119920119925 stands for an 119925 times 119925 identity matrix in which all
the principal diagonal elements are ones and all other elements are zeros
The network approach of Diebold and Yilmaz (2009) requires that after estimating the underlying
VAR model the forecast error variance decompositions (FEVDs) are then generated and used to
build linkage measures In this study the interest is in the shocks to the disturbances 120634119947120061 in the
country-specific equations Hence following Pesaran and Shin (1998) Diebold and Yilmaz (2016)
and Greenwood-Nimmo et al (2015) this study adopts the order-invariant generalized forecast error
variance decompositions (GFEVDs) defined as
119918119917119916119933119915 119937119946120061 120634119947120061 119919 = 119941119946119947119944119919
= 120648120634119947119947minus120783 119942119946
prime120495119945120506120634119942119947 sup2119919minus120783119945=120782
119942119946prime120495119945120506120634120495119945
prime 119942119946 119919minus120783119945=120782
(3)
where 119894 119895 = 1 hellip 119873 119867 = 1 2 hellip is the forecast horizon 119890119894 119890119895 is 119873 119909 1 selection vector whose i-th
element (j-th element) is unity with zeros elsewhere Θℎ is the coefficient matrix multiplying the h-
lagged shock vector in the infinite moving-average representation of the non-orthogonalized VAR
120634 is the covariance matrix of the shock vector in the non-orthogonalized VAR and 120648120634119947119947 is the j-th
diagonal element of 120634 (ie the standard deviation of 120576119895 ) We adopted a maximum forecast horizon
of 16 quarters in order to ensure that the long-run results are better captured It must be stressed that
the choice of GFEVDs for this study rather than the orthogonalized forecast error variance
decompositions (OFEVDs) of Diebold and Yilmaz (2009) is particularly based on the fact that the
12
OFEVDs depend on the reordering of the variables in the system such that once the order of
variables in the VAR is reshuffled a different outcome results
Diebold and Yilmaz (2014) explain that shocks are rarely orthogonal in the GFEVD environment
so that sums of forecast error variance contributions are not necessarily unity that is row sums of
the generalized variance decomposition matrix 119863119892119867 are not necessarily unity This renders the
interpretation of the GFEVDs complicated Thus to restore a percentage interpretation of the
GFEVDs this study follows Diebold and Yilmaz (2014) to define the normalized GFEVDs
(NGFEVDs) given by2
119863 119892 = 119889 119894119895119892 where 119889 119894119895
119892=
119889119894119895119892
119889119894119895119892119873
119895=1
119889119894119895119892
= 119918119917119916119933119915 119937119946120061 120634119947120061 119919 (4)
By construction 119889 119894119895119892119873
119895=1 = 1 and 119889 119894119895119892119873
119894 119895=1 = 119873 so that the total sum of the generalized forecast
error variance share of each variable is normalized to 100
33 Construction of the Generalized Trade Linkage Measures (GTLMs)
The various trade linkage measures that are relevant for the ensuing analysis are defined in this sub-
section The intuition behind this framework is quite simple Variance decomposition permits the
splitting of the forecast error variances of each variable in the VAR system into parts attributable to
the various system shocks By so doing it becomes easy to answer the question What fraction of
the h-step-ahead error variance in forecasting 119937120783119957 is due to shocks to 119937120783119957 Shocks to 119937120784119957 Similarly
what fraction of the h-step-ahead error variance in forecasting 119937120784119957 is due to shocks to 119937120783119957 Shocks
to 119937120784119957 And in general what fraction of the h-step-ahead error variance in forecasting 119937119947119957 is due to
shocks to 119937119946119957 119946 = 120783 120784 hellip 119925 Thus the approach marries VAR variance decomposition theory and
network topology theory by recognizing that variance decompositions of VARs form networks and
also characterizing linkages in those variance decomposition networks This in turn characterizes
trade linkages of the variables in our VAR system This is the intuition behind this framework
which we now exploit in the ensuing analysis Diebold and Yilmaz (2015) authoritatively document
this framework and its relation to network theory
2 In what follows and without loss of generality the superscript 119867 is dropped whenever it is not needed for clarity so
that 119863119892119867 and 119889119894119895119892119867
are simply written as 119863119892 and 119889119894119895119892
respectively
13
Table 1 Trade Linkage Schematic
Variables 119937120783 119937120784 ⋯ 119937119925 From Others
119937120783
⋯ 1198891119895
119873
119895=1 119895 ne 1
119937120784
⋯ 1198892119895
119873
119895=1 119895 ne 2
⋮ ⋮ ⋮ ⋱
⋮ ⋮
119937119925
⋯ 119889119873119895
119873
119895=1 119895 ne 119873
To Others
1198891198941
119873
119894=1
119894 ne 1
1198891198942
119873
119894=1
119894 ne 2
⋯ 119889119894119873
119873
119894=1
119894 ne N
1
119873 119889119894119895
119873
119894 119895=1 119894 ne 119895
Source Adapted from Diebold and Yilmaz (2014) and Ogbuabor et al (2016) Note For
simplicity each time series variable in this table 119937119947119957 is written as 119937119947 119947 = 120783 120784 hellip 119925
To construct the GTLMs for this study let us denote the H-step ahead NGFEVDs for the 119925 119961 120783
vector of endogenous variables 119937120061 obtained from equation (4) by 119941119946119947 By cross-tabulating 119941119946119947 the
trade linkage table shown in Table 1 is formed The sum of each row in Table 1 is normalized to
100 in line with equation (4) This table is now used to define the various GTLMs and their
relationships The diagonal entries in Table 1 measure own variance shares (or own-effect) while
the off-diagonal entries measure variance shares arising from shocks to other variables in the
system and are therefore referred to as pairwise directional linkage Accordingly the own-effect
(119919119947) also known as the heatwave is defined as
119919119947 = 119941119947119947 (5)
The total cross-variable variance share (119917119947) captures the spillovers from all other variables to 119937119947120061 as
fractions of the H-step-ahead error variance in the forecasts of 119937119947120061 resulting from 119937119946120061 where
119946 = 120783 120784 hellip 119925 and 119946 ne 119947 This measures the total directional linkage from other variables
(countries) in the system (ie the from-effect) to Zj120009 This means that the from-effect can be used to
capture the role each individual economy in the system plays in a given African economy and it is
14
computed in this study by aggregating the spillovers from all the economies in the system to a
given African economy across all horizons Hence the economy contributing the highest of such
aggregate spillover is deemed to play a dominant role in the particular African economy This study
therefore defines 119917119947 as
119917119947 = 119941119947119946119925119946=120783119946ne119947 (6)
By construction 119919119947 + 119917119947 = 120783 forall 119947
This study also defines the total spillover or total contributions of 119937119947120061 to all other variables (denoted
by 119931119947) as
119931119947 = 119941119946119947119925119946=120783119946ne119947 (7)
By construction 119931119947 measures the total directional linkage from 119937119947120061 to other variables in the system
(ie the to-effect) In other words the to-effect measures the directional linkage from a given
economy (for instance a given African economy) to other economies in the system thereby
showing the impact or influence of that particular African economy on other economies in the VAR
system The net directional linkage (or simply net-effect) of 119937119947120061 is therefore defined as
119925119947 = 119931119947 minus 119917119947 (8)
Since there are N economies in the VAR system it also follows that there are 2N total directional
trade linkages N capturing the total shocks transmitted to others (ie spillover to) and N capturing
the total shocks received from others (ie spillover from) These measures are aptly reflect the
bilateral trade patterns between the African economies and other economies in the system since
they mirror the total exports and total imports for each of the N economies in the system and will
therefore facilitate deeper understanding of the inter-linkages among the African economies We
shall utilize the net-effects to establish the net transmittersreceivers of shocks in the system over
time By construction 119925119947119925119947=120783 = 120782
The most aggregative (non-directional) trade linkage measure in this study which will be used to
evaluate the degree of trade linkage Africa the BRICS and the rest of the global economy is known
as the total trade linkage index (119931119931119923119920) or total-effects and it is defined as
119931119931119923119920 = 120783
119925 119917119947
119925119947=120783 =
120783
119925 119931119947
119925119947=120783 (9)
15
This measure captures the grand total of the off-diagonal elements in Table 1 that is the sum of the
―From Others column or ―To Others row There is only one total trade linkage measure which is
analogous to total global imports or total global exports since the two are identical
An important objective of this study is to determine which of the countries among the BRICS and
in the rest of the global economy exert the most dominant trade influence on African economies
and therefore have the potential to spread trade shocks to Africa To achieve this objective this
study defines two indices dependence and influence indices These indices are necessary to
determine the dependence of the j-th variable (or j-th economy) on external shocks and the
influence of the j-th variable (or j-th economy) on the system as a whole The dependence index is
defined as
119926119947119919 =
119917119947
119919119947+ 119917119947 forall 119947 = 120783 120784 hellip 119925 (10)
where 120782 le 119926119947119919 le 120783 This index expresses the relative importance of external shocks for the j-th
economy in the VAR system such that if 119926119947119919 rarr 120783 then conditions in the j-th economy is open
deeply interlinked and sensitive to external conditions but if 119926119947119919 rarr 120782 then the j-th economy is less
sensitive to external shocks
Similarly the influence index is expressed as
119920119947119919 =
119925119947
119931119947+119917119947 forall 119947 = 120783 120784 hellip 119925 (11)
where minus120783 le 119920119947119919 le 120783 For a given horizon H the j-th economy is a net receiver of trade shocks if
minus120783 le 119920119947119919 lt 0 that is if the index has a negative value a net transmitter of trade shocks if 120782 lt
119920119947119919 le 120783 that is if the index takes a positive value and neither a net receiver or transmitter of trade
shocks if IjH = 0 Thus the influence index measures the extent to which the j-th economy in the
system influences or is influenced by external shocks Overall the coordinate pair (119926119947119919 119920119947
119919) in the
dependence-influence space provides a good representation of the j-th economylsquos role in global real
activities A priori expectation is that African economies which are relatively small open
economies would be located close to the point (1-1) while overwhelmingly open but highly
influential and dominant economies like the USA China and Japan would be located close to the
point (11)
4 Results and Discussions
16
We started this empirical analysis by examining the time series properties of the data The Phillips-
Perron unit root tests showed that all the series are I(1) for all the countries (Please see Appendix 3
which reports for total trade of all the countries To conserve space we do not report separately for
exports and imports since the patterns are similar that is all the series are I(1)) However the test
for long-run or equilibrium relationship using the Johansen System Cointegration test showed that
both the Trace and Maximum Eigenvalue statistics returned full rank indicating that the data are
not cointegrated (Please see Appendix 4 which reports the Trace test in Panel 1 and the Max-
Eigenvalue test in Panel 2) Therefore a VAR in first differences was estimated rather than a vector
error correction model (VECM) Thereafter the GTLMs were computed for all forecasting
horizons 119867 = 12 hellip 16 The average values of the GTLMs over all the horizons are reported
since Diebold and Yilmaz (2016) had shown that the measures follow similar patterns regardless of
the choice of window lengths and forecast horizons As earlier stated we set the maximum forecast
horizon at 16 quarters in order to capture the long-run results better As a robustness check we also
estimated the underlying model and computed the GTLMs separately for exports and imports As
shall soon see the results follow the same qualitative pattern
41 Measuring the degree of trade linkage between Africa the BRICS and the rest of the
global economy
The first specific objective of this study is to measure the degree of trade linkage between African
economies the BRICS and the rest of the global economy In addition we examine here how this
index changed from the short-run (ie from horizon 1) through the long-run (ie until horizon 16)
To achieve this objective we estimated the underlying VAR model of equation (1) and used the
NGFEVDs distilled from this estimation based on equation (4) to compute the Total Trade Linkage
Index (TTLI) based on equation (9) We report the plots of this index in Figure 1 based on the total
trade data while the plots based on the exports and imports data are reported in Appendix 5 This
figure shows how the most aggregated generalized trade linkage measure in this study evolved from
the short-run through the long-run We find that the index rose smoothly from 84 at horizons 1
and 2 to 90 at horizons 15 and 16 In other words the total trade linkage index is higher in the
long-run than in the short-run indicating African economies become more interlinked in the long-
run as the business cycles become more synchronized This finding is consistent with the trend of
global macroeconomic interlinkage reported by Greenwood-Nimmo et al (2015) Diebold and
Yilmaz (2016) and Ogbuabor et al (2018) that the ongoing globalization process is engendering
17
more significant comovement in industrial production fluctuations Furthermore we find that the
TTLI recorded an average value of 87 which shows that the trade linkage between Africa the
BRICS and the rest of the global economy is quite substantial
As shown in Appendix 5 the patterns in this index remained qualitatively unchanged when the
underlying model is estimated using the exports and imports data respectively However a closer
look at panels 1 and 2 of this Appendix reveals that the index recorded higher values across all the
horizons for exports than for imports This aptly captures the pattern of trade between Africa and
the rest of the world which shows that Africalsquos export trade in dollar terms at least in the last
decade is higher than its import trade (see table 1 in appendix 1 of African Trade Statistics
Yearbook 2017) This finding clearly indicates that Africa may be vulnerable to trade shocks
which underlines the motivation behind this study
Figure 1 Total Trade Linkage Index
Source Authors Notes The Total trade linkage Index reported here is the most aggregated non-
directional trade linkage index computed for all horizons following equation (9) Notice that the index rises
smoothly from the short-run (ie horizon 1) towards the long-run (ie horizon 16) indicating that African
economies become more interlinked in the long-run
42 Determining the BRICS countries and other countries in the rest of the world that are
dominating Africarsquos trade and therefore have the potential to spread trade shocks to Africa
An important objective of this study is to determine the BRICS member countries and other
countries in the rest of the world that are dominating Africalsquos trade and therefore have the potential
to spread trade shocks to Africa To achieve this objective we report the from-effect trade linkage
from all the economies in the system to each African economy following equation (6) We include
the heatwave (or own-effect) of equation (5) in this table so that it sums up to 100 We aggregate
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
4
among entities in the system while identifying systemically important or vulnerable entities within
the system Several studies in the literature have exploited this framework in the study of
macroeconomic linkages with great success Examples include Pesaran Schuermann and Weiner
(2004) Dees di Mauro Pesaran and Smith (2007) Dees Holly Pesaran and Smith (2007) and
more recently Greenwood-Nimmo Nguyen and Shin (2015) We are adopting same framework
using some selected African economies and since the data for the study are quarterly the measures
of trade linkages would be computed over a 24-period horizon so that the long-run results are
better captured
The rationale for assessing the impact of trade between Africa the BRIC and the rest of the world is
to bring to the fore the potential trade shock impact between Africa and other trade blocks with
which it has significant trade linkages even as it makes concerted efforts towards signing of the
African Continental Free Trade Area (AfCFTA) through a treaty by all the countries concerned
Nigeria which is Africalsquos largest economy by size of GDP is currently understudying the tenets of
AfCFTA treaty preparatory to signing it The AfCFTA is to create a single continental market for
goods and services in member nations of the African Union (AU) with free movement of business
persons and investments using a single currency The AfCFTA is intended to be the external cordon
protecting the African countrieslsquo domestic markets Thus while AU member states are signing the
treaty it is important to be aware of the potential trade shocks that may emanate from the
significant trade linkages between Africa and the rest of the world
The rest of the paper is organized as follows Section 2 undertakes a brief review of relevant
literature Section 3 presents the data and methodology detailing some preliminary data analysis
while Section 4 discusses the empirical results Section 5 concludes the study and provides some
policy implications that the patterns of cross-country trade shock spillovers obtained in this study
would likely influence Africalsquos move from globalization to regional integration or multi-polarity in
the years to come
2 Brief Review of Literature
The impact of trade shocks in large advanced countries on emerging market economies mdash dubbed
spillovers mdash is hotly debated in global and national policy circles Kose and Reizman (1999)
examined the role of external shocks in explaining macroeconomic fluctuations in African
5
countries They constructed a quantitative stochastic dynamic multi-sector equilibrium model of a
small open economy calibrated to represent a typical African economy In their framework external
shocks consist of trade shocks modeled as fluctuations in the prices of exported primary
commodities imported capital goods and intermediate inputs and a financial shock modeled as
fluctuations in the world real interest rate Results from their analysis indicated that while trade
shocks accounted for roughly 45 percent of economic fluctuations in aggregate output financial
shocks play only a minor role They also found that adverse trade shocks induce prolonged
recessions
Ccedilakir and Kabundi (2011) examined the trade linkages between South Africa and the BRIC (Brazil
Russia India and China) countries They applied the global vector autoregressive model (global
VAR) to investigate the degree of trade linkages and shock transmission between South Africa and
the BRIC countries over the period 1995Q1-2009Q4 The model contained 32 countries and had
two different estimations the first one consists of 24 countries and one region with the 8 countries
in the euro area treated as a single economy and the second estimation contains 20 countries and
two regions with the BRIC and the euro area countries respectively treated as a single economy
The results suggest that trade linkages exist between our focus economies however the magnitude
differs between countries Shocks from each BRIC country are shown to have considerable impact
on South African real imports and output
Greenwood-Nimmo et al (2015) developed a technique to evaluate macroeconomic connectedness
among entities in sophisticated multi-country and global macroeconomic models Their
methodology is highly adaptable and may be applied to any model with an approximate VAR
representation They applied their technique to a global vector autoregressive model containing 169
macroeconomic and financial variables for 25 countries They derived vivid representations of the
connectedness of the system and found that the US the Eurozone and the crude oil market exert a
dominant influence over conditions in the global macroeconomy and that China and Brazil are also
globally significant economies Recursive analysis over the period of the global financial crisis
shows that shocks to global equity markets are rapidly and forcefully transmitted to real trade flows
and real GDP
Pesaran et al (2004) posit that financial institutions are ultimately exposed to macroeconomic
fluctuations in the global economy and thus they built a compact global model capable of
6
generating forecasts for a core set of macroeconomic factors (or variables) across a number of
countries Their model explicitly allows for the interdependencies that exist between national and
international factors Individual region-specific vector error-correcting models were estimated in
which the domestic variables were related to corresponding foreign variables constructed
exclusively to match the international trade pattern of the country under consideration The
individual country models were then linked in a consistent and cohesive manner to generate
forecasts for all of the variables in the world economy simultaneously Their global model was
estimated for 25 countries grouped into 11 regions using quarterly data over 1979Q1ndash1999Q1 The
degree of regional interdependencies was investigated via generalized impulse responses where the
effects of shocks to a given variable in a given country on the rest of the world were provided The
model was then used to investigate the effects of various global risk scenarios on a banklsquos loan
portfolio
Dees et al (2007) presented a quarterly global model combining individual country vector error-
correcting models in which the domestic variables were related to the country-specific foreign
variables Their global VAR (GVAR) model was estimated for 26 countries the Euro area being
treated as a single economy over the period 1979ndash2003 They provided a theoretical framework
where the GVAR was derived as an approximation to a global unobserved common factor model
Using the average pair-wise cross-section error correlations their GVAR approach was shown to be
quite effective in dealing with the common factor interdependencies and international co-
movements of business cycles They developed a sieve bootstrap procedure for simulation of the
GVAR as a whole which is then used in testing the structural stability of the parameters and for
establishing bootstrap confidence bounds for the impulse responses Finally in addition to
generalized impulse responses their paper considered the use of the GVAR for structurallsquo impulse
response analysis with focus on external shocks for the Euro area economy particularly in response
to shocks to the US
In another study Dees et al (2007) focused on testing the long run macroeconomic relations for
interest rates equity prices and exchange rates suggested by arbitrage in financial and
goods markets They used the global vector autoregressive (GVAR) model to test
for long run restrictions in each countryregion conditioning on the rest of the
world They employed Bootstrapping to compute both the empirical distribution of
7
the impulse responses and the log-likelihood ratio statistic for over-identifying
restrictions Their paper also examined the speed with which adjustments to the
long run relations would take place via the persistence profiles They found strong evidence
in favour of the uncovered interest parity (UIP) and to a lesser extent the Fisher equation across a
number of countries but their results for the purchasing power parity (PPP) were much weaker
Also the transmission of shocks and subsequent adjustments in financial markets were much faster
than those in goods markets
Ogbuabor et al (2016) examined the real and financial connectedness of selected African
economies with the global economy using a network approach They found that the connectedness
of African economies with the global economy is quite sizable with the global financial crisis
increasing the connectedness measures above their pre-crisis levels Their results show that US
EU and Canada dominate Africalsquos equity markets while China India and Japan dominate Africalsquos
real activities Their results suggest that African economies are predominantly small open
economies deeply interconnected but systemically unimportant and vulnerable to headwinds
emanating from the dominant economies in the overall global economy
Lubik and Teo (2003) found that world interest rate shocks are the main driving forces of business
cycles in small open economies while terms of trade shocks are not Thus they challenged the
existing results on the contribution of terms of trade and world interest rate shocks to output
fluctuations in small open economies which had been found to range from less than 10 per cent to
almost 90 per cent They argue that an identification problems lies at the heart of existing vastly
different results They overcame this by estimating a DSGE model using a structural Bayesian
estimation approach applying their methodology to five developed and developing economies The
approach enabled them to efficiently exploit cross-equation restrictions implied by the structural
model
Huidrom et al (2017) provide empirical estimates of the cross border spillovers from Seven largest
emerging market economies (China India Brazil Russia Mexico Indonesia and Turkey) using a
Bayesian vector autoregression model Their results indicate that first spillovers from EM7 are
sizeable a 1 percentage point increase in EM7 growth is associated with a 09 percentage point
increase in growth in other emerging and frontier markets and a 06 percentage point increase in
world growth at the end of three years Second sizeable as they are spillovers from EM7 are still
8
smaller than those from G7 countries (Group of Seven of advanced economies) Specifically
growth in other emerging and frontier markets and the global economy would increase by one-half
to three times more due to a similarly sized increase in G7 growth Third among the EM7
spillovers from China are the largest and permeate globally Their study was based on the
phenomenon that their growth could have significant cross-border spillovers based on their size and
integration The seven largest emerging market economies - China India Brazil Russia Mexico
Indonesia and Turkey - was adjudged to constitute more than one-quarter of global output and
more than half of global output growth during 2010ndash15
This study contributes to this emerging literature in several ways First it quantifies the degree of
trade linkage between Africa the BRIC and the rest of the global economy Two apart from
exposing the patterns of trade shock spillovers between Africa and the rest of the world using total
trade statistics this study also disaggregates the trade shock spillovers using exports and imports
statistics This provides deeper insights into the trade linkages between Africa and the rest of the
world Besides the study extended the empirical method by constructing trade linkage measures in
a coherent and transparent manner for ease of replication
3 Data and Methodology
The data for this study consists of the log of exports imports and total trade for the period 1990Q1-
2016Q4 The choice of this period is based on data availability for some of the African economies
and for Russia whose trade data in the World Development Indicators started in 1990 The BRICS
economies in this study are Brazil Russia India China and South Africa The selected African
economies included in this study are DR Congo Egypt Ghana Morocco Nigeria Tanzania and
Tunisia while the rest of the global economy is accounted for by Australia Canada the Eurozone
(EU) Indonesia Japan Malaysia United States of America (USA) and United Kingdom (UK)
These economies account for the larger chunk of Africalsquos trade while the selected African
economies together with South Africa account substantially for Africalsquos GDP The entire data were
taken from the World Development Indicators (WDI) based on the following indicator names
exports of goods and services (constant 2010 US$) as measure for exports imports of goods and
services (constant 2010 US$) as measure for imports and exports plus imports as measure for total
trade (or simply trade)
9
To ensure that there are enough observations for the analysis the data were converted from annual
to quarterly using Eviewlsquos quadratic match average option This is in line with the methodology
used in compiling the Global VAR database used by Greenwood-Nimmo Nguyen and Shin (2015)
the connectedness of the global economy Other studies that have successfully used the same
methodology are Ogbuabor Eigbiremolen Aneke and Manasseh (2018) and Ogbuabor Orji
Aneke and Erdene-Urnukh (2016) Also to reduce noise and ensure uniform scaling the entire data
were converted into indices (2010Y = 100) and logged prior to estimation Appendix 1 summarizes
the descriptive statistics of the data based on the log transformation of the data Focusing on Panel 1
which reports for total trade we find that among the African economies Ghana DR Congo and
Tanzania are the most volatile as seen from their standard deviations while China India and Russia
are the most volatile among the BRICS economies Among all the countries in the sample Japan
Canada and UK are the least volatile and also recorded the highest mean values The maximum and
minimum values do not suggest the presence of outliers in the data The time series plots of the data
for all the economies in the sample are presented in Appendix 2 based on the log transformation of
the data A close examination of the plots show close comovement among all the countries
indicating that the series track themselves closely The implication of this is that the data may be
cointegrated which will in turn affect the form of the underlying model to be estimated in this
study Thus we shall subject the series to cointegration test as part of the empirical procedures in
this study
31 The Choice of Diebold and Yilmaz (2009) Network Framework
A number of methodologies have been employed in the study of macroeconomic linkages among
entities in the global economy particularly in the study of shock propagation among countries For
instance cross-country correlations-based measures have been used to characterize macroeconomic
linkages among countries (Kehoe et al 1995 Kose et al 2003 Bollerslev 1990 Engle et al
1990 Mantegna 1999 Tumminello et al 2005 Taylor 2007 Gray amp Malone 2008 Engle 2009
Engle amp Kelly 2012) The pitfalls of this approach are twofold namely correlation is simply a
pairwise measure of association and it is non-directional This means that correlation-based
approach cannot handle such questions as ―what is the degree of trade linkage between Africa the
BRICS and the rest of the global economy Unlike the correlation-based measures of
macroeconomic linkage the network approach of Diebold and Yilmaz (2009) is non-pairwise yet
10
directional Granger Causality measures have also been used characterize networks so that the
macroeconomic linkages among entities in the global economy can be described and understood
(Caraiani 2013 Hiemstra amp Jones 1994 Dahlhans amp Eichler 2003 Shojaie amp Michailidis 2010
Billio et al 2012) The main weakness of the Granger Causality approach is that it captures only
pairwise relations and may not be useful in answering important questions like ―what is the degree
of trade linkage between Africa the BRICS and the rest of the global economy1
Furthermore as noted by Diebold and Yilmaz (2016) these alternative methodologies generally
dwell exclusively on testing rather than measurement and estimation of macroeconomic linkages
which are the key issues in this paper This study therefore follows the network approach of
Diebold and Yilmaz (2009) based on its ability to transparently use the size and direction of shocks
to build both directional and non-directional trade linkage measures over a given forecast horizon
According to Ogbuabor et al (2016) studies using this approach have four common features
namely (i) they are generally based on connectedness measures distilled from forecast error
variance decompositions (FEVDs) of an approximating vector autoregressive (VAR) model (ii)
they measure the direction and strength of linkages among entities in the system (iii) they can
identify systemically important entities in the system and (iv) they can study the dynamic nature of
shock propagation among entities in the system In what follows the underlying VAR model for
this study and the construction of the generalized trade linkages measures (GTLMs) are presented
to guide the ensuing analysis
32 Model Specification The Vector Autoregression (VAR) Model
The broad objective of this study is to examine the propagation of trade shock between Africa the
BRICS economies and the rest of the global economy Let 119937120061 be the log of total trade for all the
countries selected for this study so that 119937119947120061 stands for the logged total trade of the j-th country in the
system with 119947 = 120783 120784 hellip 119925 and 119925 is the number of countries selected for the study (which is 20)
Following Diebold and Yilmaz (2009) the trade linkage measures for this study are based on the
normalized generalized forecast error variance decompositions (NGFEVDs) of an underlying 119953-th
1 Other techniques have also been used in the literature for the study of macroeconomic linkages Ogbuabor et al
(2018) provides an overview of such alternative methodologies such as the dynamic latent factor models of Kose et al (2008) and Canova et al (2007) the CoVaR approach of Adrian and Brunnermeier (2011) and the marginal expected shortfall (MES) approach of Acharya et al (2017) and Brownlees and Engle (2012)
11
order VAR model for the 119925 119961 120783 vector of endogenous variables 119937120061 The VAR(119953) model is specified
as follows
119937120061 = 120630119963 + 120509120051119937120061minus120051120057120051=120783 + 120634120061 (1)
where 120630 is 119925119961120783 vector of intercepts 120509120051 is 119925 119961 119925 coefficient matrix 120057 is the lag order and the
residuals 120634120050120061 ~ 119946119946119941 120782 120634119946119946 so that 120634120061 ~ 120782 120634 where 120634 is positive definite covariance matrix
The optimal VAR lag order selected by Schwarz Information Criterion (SIC) for this study is one
(see Appendix 5) Using the Woldlsquos Representation Theorem the model in equation (1) is
expressed as an infinite order vector moving average representation given by
119937120061 = 120634119957 + 120495120783120518119853minus120783 + 120495120784120518119853minus120784 + hellip = 120495120051120634120061minus120051infin120051=120782 (2)
where 120495120782 = 119920119925 120495120051 = 120509120051 120051 = 120783 120784 hellip and 119920119925 stands for an 119925 times 119925 identity matrix in which all
the principal diagonal elements are ones and all other elements are zeros
The network approach of Diebold and Yilmaz (2009) requires that after estimating the underlying
VAR model the forecast error variance decompositions (FEVDs) are then generated and used to
build linkage measures In this study the interest is in the shocks to the disturbances 120634119947120061 in the
country-specific equations Hence following Pesaran and Shin (1998) Diebold and Yilmaz (2016)
and Greenwood-Nimmo et al (2015) this study adopts the order-invariant generalized forecast error
variance decompositions (GFEVDs) defined as
119918119917119916119933119915 119937119946120061 120634119947120061 119919 = 119941119946119947119944119919
= 120648120634119947119947minus120783 119942119946
prime120495119945120506120634119942119947 sup2119919minus120783119945=120782
119942119946prime120495119945120506120634120495119945
prime 119942119946 119919minus120783119945=120782
(3)
where 119894 119895 = 1 hellip 119873 119867 = 1 2 hellip is the forecast horizon 119890119894 119890119895 is 119873 119909 1 selection vector whose i-th
element (j-th element) is unity with zeros elsewhere Θℎ is the coefficient matrix multiplying the h-
lagged shock vector in the infinite moving-average representation of the non-orthogonalized VAR
120634 is the covariance matrix of the shock vector in the non-orthogonalized VAR and 120648120634119947119947 is the j-th
diagonal element of 120634 (ie the standard deviation of 120576119895 ) We adopted a maximum forecast horizon
of 16 quarters in order to ensure that the long-run results are better captured It must be stressed that
the choice of GFEVDs for this study rather than the orthogonalized forecast error variance
decompositions (OFEVDs) of Diebold and Yilmaz (2009) is particularly based on the fact that the
12
OFEVDs depend on the reordering of the variables in the system such that once the order of
variables in the VAR is reshuffled a different outcome results
Diebold and Yilmaz (2014) explain that shocks are rarely orthogonal in the GFEVD environment
so that sums of forecast error variance contributions are not necessarily unity that is row sums of
the generalized variance decomposition matrix 119863119892119867 are not necessarily unity This renders the
interpretation of the GFEVDs complicated Thus to restore a percentage interpretation of the
GFEVDs this study follows Diebold and Yilmaz (2014) to define the normalized GFEVDs
(NGFEVDs) given by2
119863 119892 = 119889 119894119895119892 where 119889 119894119895
119892=
119889119894119895119892
119889119894119895119892119873
119895=1
119889119894119895119892
= 119918119917119916119933119915 119937119946120061 120634119947120061 119919 (4)
By construction 119889 119894119895119892119873
119895=1 = 1 and 119889 119894119895119892119873
119894 119895=1 = 119873 so that the total sum of the generalized forecast
error variance share of each variable is normalized to 100
33 Construction of the Generalized Trade Linkage Measures (GTLMs)
The various trade linkage measures that are relevant for the ensuing analysis are defined in this sub-
section The intuition behind this framework is quite simple Variance decomposition permits the
splitting of the forecast error variances of each variable in the VAR system into parts attributable to
the various system shocks By so doing it becomes easy to answer the question What fraction of
the h-step-ahead error variance in forecasting 119937120783119957 is due to shocks to 119937120783119957 Shocks to 119937120784119957 Similarly
what fraction of the h-step-ahead error variance in forecasting 119937120784119957 is due to shocks to 119937120783119957 Shocks
to 119937120784119957 And in general what fraction of the h-step-ahead error variance in forecasting 119937119947119957 is due to
shocks to 119937119946119957 119946 = 120783 120784 hellip 119925 Thus the approach marries VAR variance decomposition theory and
network topology theory by recognizing that variance decompositions of VARs form networks and
also characterizing linkages in those variance decomposition networks This in turn characterizes
trade linkages of the variables in our VAR system This is the intuition behind this framework
which we now exploit in the ensuing analysis Diebold and Yilmaz (2015) authoritatively document
this framework and its relation to network theory
2 In what follows and without loss of generality the superscript 119867 is dropped whenever it is not needed for clarity so
that 119863119892119867 and 119889119894119895119892119867
are simply written as 119863119892 and 119889119894119895119892
respectively
13
Table 1 Trade Linkage Schematic
Variables 119937120783 119937120784 ⋯ 119937119925 From Others
119937120783
⋯ 1198891119895
119873
119895=1 119895 ne 1
119937120784
⋯ 1198892119895
119873
119895=1 119895 ne 2
⋮ ⋮ ⋮ ⋱
⋮ ⋮
119937119925
⋯ 119889119873119895
119873
119895=1 119895 ne 119873
To Others
1198891198941
119873
119894=1
119894 ne 1
1198891198942
119873
119894=1
119894 ne 2
⋯ 119889119894119873
119873
119894=1
119894 ne N
1
119873 119889119894119895
119873
119894 119895=1 119894 ne 119895
Source Adapted from Diebold and Yilmaz (2014) and Ogbuabor et al (2016) Note For
simplicity each time series variable in this table 119937119947119957 is written as 119937119947 119947 = 120783 120784 hellip 119925
To construct the GTLMs for this study let us denote the H-step ahead NGFEVDs for the 119925 119961 120783
vector of endogenous variables 119937120061 obtained from equation (4) by 119941119946119947 By cross-tabulating 119941119946119947 the
trade linkage table shown in Table 1 is formed The sum of each row in Table 1 is normalized to
100 in line with equation (4) This table is now used to define the various GTLMs and their
relationships The diagonal entries in Table 1 measure own variance shares (or own-effect) while
the off-diagonal entries measure variance shares arising from shocks to other variables in the
system and are therefore referred to as pairwise directional linkage Accordingly the own-effect
(119919119947) also known as the heatwave is defined as
119919119947 = 119941119947119947 (5)
The total cross-variable variance share (119917119947) captures the spillovers from all other variables to 119937119947120061 as
fractions of the H-step-ahead error variance in the forecasts of 119937119947120061 resulting from 119937119946120061 where
119946 = 120783 120784 hellip 119925 and 119946 ne 119947 This measures the total directional linkage from other variables
(countries) in the system (ie the from-effect) to Zj120009 This means that the from-effect can be used to
capture the role each individual economy in the system plays in a given African economy and it is
14
computed in this study by aggregating the spillovers from all the economies in the system to a
given African economy across all horizons Hence the economy contributing the highest of such
aggregate spillover is deemed to play a dominant role in the particular African economy This study
therefore defines 119917119947 as
119917119947 = 119941119947119946119925119946=120783119946ne119947 (6)
By construction 119919119947 + 119917119947 = 120783 forall 119947
This study also defines the total spillover or total contributions of 119937119947120061 to all other variables (denoted
by 119931119947) as
119931119947 = 119941119946119947119925119946=120783119946ne119947 (7)
By construction 119931119947 measures the total directional linkage from 119937119947120061 to other variables in the system
(ie the to-effect) In other words the to-effect measures the directional linkage from a given
economy (for instance a given African economy) to other economies in the system thereby
showing the impact or influence of that particular African economy on other economies in the VAR
system The net directional linkage (or simply net-effect) of 119937119947120061 is therefore defined as
119925119947 = 119931119947 minus 119917119947 (8)
Since there are N economies in the VAR system it also follows that there are 2N total directional
trade linkages N capturing the total shocks transmitted to others (ie spillover to) and N capturing
the total shocks received from others (ie spillover from) These measures are aptly reflect the
bilateral trade patterns between the African economies and other economies in the system since
they mirror the total exports and total imports for each of the N economies in the system and will
therefore facilitate deeper understanding of the inter-linkages among the African economies We
shall utilize the net-effects to establish the net transmittersreceivers of shocks in the system over
time By construction 119925119947119925119947=120783 = 120782
The most aggregative (non-directional) trade linkage measure in this study which will be used to
evaluate the degree of trade linkage Africa the BRICS and the rest of the global economy is known
as the total trade linkage index (119931119931119923119920) or total-effects and it is defined as
119931119931119923119920 = 120783
119925 119917119947
119925119947=120783 =
120783
119925 119931119947
119925119947=120783 (9)
15
This measure captures the grand total of the off-diagonal elements in Table 1 that is the sum of the
―From Others column or ―To Others row There is only one total trade linkage measure which is
analogous to total global imports or total global exports since the two are identical
An important objective of this study is to determine which of the countries among the BRICS and
in the rest of the global economy exert the most dominant trade influence on African economies
and therefore have the potential to spread trade shocks to Africa To achieve this objective this
study defines two indices dependence and influence indices These indices are necessary to
determine the dependence of the j-th variable (or j-th economy) on external shocks and the
influence of the j-th variable (or j-th economy) on the system as a whole The dependence index is
defined as
119926119947119919 =
119917119947
119919119947+ 119917119947 forall 119947 = 120783 120784 hellip 119925 (10)
where 120782 le 119926119947119919 le 120783 This index expresses the relative importance of external shocks for the j-th
economy in the VAR system such that if 119926119947119919 rarr 120783 then conditions in the j-th economy is open
deeply interlinked and sensitive to external conditions but if 119926119947119919 rarr 120782 then the j-th economy is less
sensitive to external shocks
Similarly the influence index is expressed as
119920119947119919 =
119925119947
119931119947+119917119947 forall 119947 = 120783 120784 hellip 119925 (11)
where minus120783 le 119920119947119919 le 120783 For a given horizon H the j-th economy is a net receiver of trade shocks if
minus120783 le 119920119947119919 lt 0 that is if the index has a negative value a net transmitter of trade shocks if 120782 lt
119920119947119919 le 120783 that is if the index takes a positive value and neither a net receiver or transmitter of trade
shocks if IjH = 0 Thus the influence index measures the extent to which the j-th economy in the
system influences or is influenced by external shocks Overall the coordinate pair (119926119947119919 119920119947
119919) in the
dependence-influence space provides a good representation of the j-th economylsquos role in global real
activities A priori expectation is that African economies which are relatively small open
economies would be located close to the point (1-1) while overwhelmingly open but highly
influential and dominant economies like the USA China and Japan would be located close to the
point (11)
4 Results and Discussions
16
We started this empirical analysis by examining the time series properties of the data The Phillips-
Perron unit root tests showed that all the series are I(1) for all the countries (Please see Appendix 3
which reports for total trade of all the countries To conserve space we do not report separately for
exports and imports since the patterns are similar that is all the series are I(1)) However the test
for long-run or equilibrium relationship using the Johansen System Cointegration test showed that
both the Trace and Maximum Eigenvalue statistics returned full rank indicating that the data are
not cointegrated (Please see Appendix 4 which reports the Trace test in Panel 1 and the Max-
Eigenvalue test in Panel 2) Therefore a VAR in first differences was estimated rather than a vector
error correction model (VECM) Thereafter the GTLMs were computed for all forecasting
horizons 119867 = 12 hellip 16 The average values of the GTLMs over all the horizons are reported
since Diebold and Yilmaz (2016) had shown that the measures follow similar patterns regardless of
the choice of window lengths and forecast horizons As earlier stated we set the maximum forecast
horizon at 16 quarters in order to capture the long-run results better As a robustness check we also
estimated the underlying model and computed the GTLMs separately for exports and imports As
shall soon see the results follow the same qualitative pattern
41 Measuring the degree of trade linkage between Africa the BRICS and the rest of the
global economy
The first specific objective of this study is to measure the degree of trade linkage between African
economies the BRICS and the rest of the global economy In addition we examine here how this
index changed from the short-run (ie from horizon 1) through the long-run (ie until horizon 16)
To achieve this objective we estimated the underlying VAR model of equation (1) and used the
NGFEVDs distilled from this estimation based on equation (4) to compute the Total Trade Linkage
Index (TTLI) based on equation (9) We report the plots of this index in Figure 1 based on the total
trade data while the plots based on the exports and imports data are reported in Appendix 5 This
figure shows how the most aggregated generalized trade linkage measure in this study evolved from
the short-run through the long-run We find that the index rose smoothly from 84 at horizons 1
and 2 to 90 at horizons 15 and 16 In other words the total trade linkage index is higher in the
long-run than in the short-run indicating African economies become more interlinked in the long-
run as the business cycles become more synchronized This finding is consistent with the trend of
global macroeconomic interlinkage reported by Greenwood-Nimmo et al (2015) Diebold and
Yilmaz (2016) and Ogbuabor et al (2018) that the ongoing globalization process is engendering
17
more significant comovement in industrial production fluctuations Furthermore we find that the
TTLI recorded an average value of 87 which shows that the trade linkage between Africa the
BRICS and the rest of the global economy is quite substantial
As shown in Appendix 5 the patterns in this index remained qualitatively unchanged when the
underlying model is estimated using the exports and imports data respectively However a closer
look at panels 1 and 2 of this Appendix reveals that the index recorded higher values across all the
horizons for exports than for imports This aptly captures the pattern of trade between Africa and
the rest of the world which shows that Africalsquos export trade in dollar terms at least in the last
decade is higher than its import trade (see table 1 in appendix 1 of African Trade Statistics
Yearbook 2017) This finding clearly indicates that Africa may be vulnerable to trade shocks
which underlines the motivation behind this study
Figure 1 Total Trade Linkage Index
Source Authors Notes The Total trade linkage Index reported here is the most aggregated non-
directional trade linkage index computed for all horizons following equation (9) Notice that the index rises
smoothly from the short-run (ie horizon 1) towards the long-run (ie horizon 16) indicating that African
economies become more interlinked in the long-run
42 Determining the BRICS countries and other countries in the rest of the world that are
dominating Africarsquos trade and therefore have the potential to spread trade shocks to Africa
An important objective of this study is to determine the BRICS member countries and other
countries in the rest of the world that are dominating Africalsquos trade and therefore have the potential
to spread trade shocks to Africa To achieve this objective we report the from-effect trade linkage
from all the economies in the system to each African economy following equation (6) We include
the heatwave (or own-effect) of equation (5) in this table so that it sums up to 100 We aggregate
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
5
countries They constructed a quantitative stochastic dynamic multi-sector equilibrium model of a
small open economy calibrated to represent a typical African economy In their framework external
shocks consist of trade shocks modeled as fluctuations in the prices of exported primary
commodities imported capital goods and intermediate inputs and a financial shock modeled as
fluctuations in the world real interest rate Results from their analysis indicated that while trade
shocks accounted for roughly 45 percent of economic fluctuations in aggregate output financial
shocks play only a minor role They also found that adverse trade shocks induce prolonged
recessions
Ccedilakir and Kabundi (2011) examined the trade linkages between South Africa and the BRIC (Brazil
Russia India and China) countries They applied the global vector autoregressive model (global
VAR) to investigate the degree of trade linkages and shock transmission between South Africa and
the BRIC countries over the period 1995Q1-2009Q4 The model contained 32 countries and had
two different estimations the first one consists of 24 countries and one region with the 8 countries
in the euro area treated as a single economy and the second estimation contains 20 countries and
two regions with the BRIC and the euro area countries respectively treated as a single economy
The results suggest that trade linkages exist between our focus economies however the magnitude
differs between countries Shocks from each BRIC country are shown to have considerable impact
on South African real imports and output
Greenwood-Nimmo et al (2015) developed a technique to evaluate macroeconomic connectedness
among entities in sophisticated multi-country and global macroeconomic models Their
methodology is highly adaptable and may be applied to any model with an approximate VAR
representation They applied their technique to a global vector autoregressive model containing 169
macroeconomic and financial variables for 25 countries They derived vivid representations of the
connectedness of the system and found that the US the Eurozone and the crude oil market exert a
dominant influence over conditions in the global macroeconomy and that China and Brazil are also
globally significant economies Recursive analysis over the period of the global financial crisis
shows that shocks to global equity markets are rapidly and forcefully transmitted to real trade flows
and real GDP
Pesaran et al (2004) posit that financial institutions are ultimately exposed to macroeconomic
fluctuations in the global economy and thus they built a compact global model capable of
6
generating forecasts for a core set of macroeconomic factors (or variables) across a number of
countries Their model explicitly allows for the interdependencies that exist between national and
international factors Individual region-specific vector error-correcting models were estimated in
which the domestic variables were related to corresponding foreign variables constructed
exclusively to match the international trade pattern of the country under consideration The
individual country models were then linked in a consistent and cohesive manner to generate
forecasts for all of the variables in the world economy simultaneously Their global model was
estimated for 25 countries grouped into 11 regions using quarterly data over 1979Q1ndash1999Q1 The
degree of regional interdependencies was investigated via generalized impulse responses where the
effects of shocks to a given variable in a given country on the rest of the world were provided The
model was then used to investigate the effects of various global risk scenarios on a banklsquos loan
portfolio
Dees et al (2007) presented a quarterly global model combining individual country vector error-
correcting models in which the domestic variables were related to the country-specific foreign
variables Their global VAR (GVAR) model was estimated for 26 countries the Euro area being
treated as a single economy over the period 1979ndash2003 They provided a theoretical framework
where the GVAR was derived as an approximation to a global unobserved common factor model
Using the average pair-wise cross-section error correlations their GVAR approach was shown to be
quite effective in dealing with the common factor interdependencies and international co-
movements of business cycles They developed a sieve bootstrap procedure for simulation of the
GVAR as a whole which is then used in testing the structural stability of the parameters and for
establishing bootstrap confidence bounds for the impulse responses Finally in addition to
generalized impulse responses their paper considered the use of the GVAR for structurallsquo impulse
response analysis with focus on external shocks for the Euro area economy particularly in response
to shocks to the US
In another study Dees et al (2007) focused on testing the long run macroeconomic relations for
interest rates equity prices and exchange rates suggested by arbitrage in financial and
goods markets They used the global vector autoregressive (GVAR) model to test
for long run restrictions in each countryregion conditioning on the rest of the
world They employed Bootstrapping to compute both the empirical distribution of
7
the impulse responses and the log-likelihood ratio statistic for over-identifying
restrictions Their paper also examined the speed with which adjustments to the
long run relations would take place via the persistence profiles They found strong evidence
in favour of the uncovered interest parity (UIP) and to a lesser extent the Fisher equation across a
number of countries but their results for the purchasing power parity (PPP) were much weaker
Also the transmission of shocks and subsequent adjustments in financial markets were much faster
than those in goods markets
Ogbuabor et al (2016) examined the real and financial connectedness of selected African
economies with the global economy using a network approach They found that the connectedness
of African economies with the global economy is quite sizable with the global financial crisis
increasing the connectedness measures above their pre-crisis levels Their results show that US
EU and Canada dominate Africalsquos equity markets while China India and Japan dominate Africalsquos
real activities Their results suggest that African economies are predominantly small open
economies deeply interconnected but systemically unimportant and vulnerable to headwinds
emanating from the dominant economies in the overall global economy
Lubik and Teo (2003) found that world interest rate shocks are the main driving forces of business
cycles in small open economies while terms of trade shocks are not Thus they challenged the
existing results on the contribution of terms of trade and world interest rate shocks to output
fluctuations in small open economies which had been found to range from less than 10 per cent to
almost 90 per cent They argue that an identification problems lies at the heart of existing vastly
different results They overcame this by estimating a DSGE model using a structural Bayesian
estimation approach applying their methodology to five developed and developing economies The
approach enabled them to efficiently exploit cross-equation restrictions implied by the structural
model
Huidrom et al (2017) provide empirical estimates of the cross border spillovers from Seven largest
emerging market economies (China India Brazil Russia Mexico Indonesia and Turkey) using a
Bayesian vector autoregression model Their results indicate that first spillovers from EM7 are
sizeable a 1 percentage point increase in EM7 growth is associated with a 09 percentage point
increase in growth in other emerging and frontier markets and a 06 percentage point increase in
world growth at the end of three years Second sizeable as they are spillovers from EM7 are still
8
smaller than those from G7 countries (Group of Seven of advanced economies) Specifically
growth in other emerging and frontier markets and the global economy would increase by one-half
to three times more due to a similarly sized increase in G7 growth Third among the EM7
spillovers from China are the largest and permeate globally Their study was based on the
phenomenon that their growth could have significant cross-border spillovers based on their size and
integration The seven largest emerging market economies - China India Brazil Russia Mexico
Indonesia and Turkey - was adjudged to constitute more than one-quarter of global output and
more than half of global output growth during 2010ndash15
This study contributes to this emerging literature in several ways First it quantifies the degree of
trade linkage between Africa the BRIC and the rest of the global economy Two apart from
exposing the patterns of trade shock spillovers between Africa and the rest of the world using total
trade statistics this study also disaggregates the trade shock spillovers using exports and imports
statistics This provides deeper insights into the trade linkages between Africa and the rest of the
world Besides the study extended the empirical method by constructing trade linkage measures in
a coherent and transparent manner for ease of replication
3 Data and Methodology
The data for this study consists of the log of exports imports and total trade for the period 1990Q1-
2016Q4 The choice of this period is based on data availability for some of the African economies
and for Russia whose trade data in the World Development Indicators started in 1990 The BRICS
economies in this study are Brazil Russia India China and South Africa The selected African
economies included in this study are DR Congo Egypt Ghana Morocco Nigeria Tanzania and
Tunisia while the rest of the global economy is accounted for by Australia Canada the Eurozone
(EU) Indonesia Japan Malaysia United States of America (USA) and United Kingdom (UK)
These economies account for the larger chunk of Africalsquos trade while the selected African
economies together with South Africa account substantially for Africalsquos GDP The entire data were
taken from the World Development Indicators (WDI) based on the following indicator names
exports of goods and services (constant 2010 US$) as measure for exports imports of goods and
services (constant 2010 US$) as measure for imports and exports plus imports as measure for total
trade (or simply trade)
9
To ensure that there are enough observations for the analysis the data were converted from annual
to quarterly using Eviewlsquos quadratic match average option This is in line with the methodology
used in compiling the Global VAR database used by Greenwood-Nimmo Nguyen and Shin (2015)
the connectedness of the global economy Other studies that have successfully used the same
methodology are Ogbuabor Eigbiremolen Aneke and Manasseh (2018) and Ogbuabor Orji
Aneke and Erdene-Urnukh (2016) Also to reduce noise and ensure uniform scaling the entire data
were converted into indices (2010Y = 100) and logged prior to estimation Appendix 1 summarizes
the descriptive statistics of the data based on the log transformation of the data Focusing on Panel 1
which reports for total trade we find that among the African economies Ghana DR Congo and
Tanzania are the most volatile as seen from their standard deviations while China India and Russia
are the most volatile among the BRICS economies Among all the countries in the sample Japan
Canada and UK are the least volatile and also recorded the highest mean values The maximum and
minimum values do not suggest the presence of outliers in the data The time series plots of the data
for all the economies in the sample are presented in Appendix 2 based on the log transformation of
the data A close examination of the plots show close comovement among all the countries
indicating that the series track themselves closely The implication of this is that the data may be
cointegrated which will in turn affect the form of the underlying model to be estimated in this
study Thus we shall subject the series to cointegration test as part of the empirical procedures in
this study
31 The Choice of Diebold and Yilmaz (2009) Network Framework
A number of methodologies have been employed in the study of macroeconomic linkages among
entities in the global economy particularly in the study of shock propagation among countries For
instance cross-country correlations-based measures have been used to characterize macroeconomic
linkages among countries (Kehoe et al 1995 Kose et al 2003 Bollerslev 1990 Engle et al
1990 Mantegna 1999 Tumminello et al 2005 Taylor 2007 Gray amp Malone 2008 Engle 2009
Engle amp Kelly 2012) The pitfalls of this approach are twofold namely correlation is simply a
pairwise measure of association and it is non-directional This means that correlation-based
approach cannot handle such questions as ―what is the degree of trade linkage between Africa the
BRICS and the rest of the global economy Unlike the correlation-based measures of
macroeconomic linkage the network approach of Diebold and Yilmaz (2009) is non-pairwise yet
10
directional Granger Causality measures have also been used characterize networks so that the
macroeconomic linkages among entities in the global economy can be described and understood
(Caraiani 2013 Hiemstra amp Jones 1994 Dahlhans amp Eichler 2003 Shojaie amp Michailidis 2010
Billio et al 2012) The main weakness of the Granger Causality approach is that it captures only
pairwise relations and may not be useful in answering important questions like ―what is the degree
of trade linkage between Africa the BRICS and the rest of the global economy1
Furthermore as noted by Diebold and Yilmaz (2016) these alternative methodologies generally
dwell exclusively on testing rather than measurement and estimation of macroeconomic linkages
which are the key issues in this paper This study therefore follows the network approach of
Diebold and Yilmaz (2009) based on its ability to transparently use the size and direction of shocks
to build both directional and non-directional trade linkage measures over a given forecast horizon
According to Ogbuabor et al (2016) studies using this approach have four common features
namely (i) they are generally based on connectedness measures distilled from forecast error
variance decompositions (FEVDs) of an approximating vector autoregressive (VAR) model (ii)
they measure the direction and strength of linkages among entities in the system (iii) they can
identify systemically important entities in the system and (iv) they can study the dynamic nature of
shock propagation among entities in the system In what follows the underlying VAR model for
this study and the construction of the generalized trade linkages measures (GTLMs) are presented
to guide the ensuing analysis
32 Model Specification The Vector Autoregression (VAR) Model
The broad objective of this study is to examine the propagation of trade shock between Africa the
BRICS economies and the rest of the global economy Let 119937120061 be the log of total trade for all the
countries selected for this study so that 119937119947120061 stands for the logged total trade of the j-th country in the
system with 119947 = 120783 120784 hellip 119925 and 119925 is the number of countries selected for the study (which is 20)
Following Diebold and Yilmaz (2009) the trade linkage measures for this study are based on the
normalized generalized forecast error variance decompositions (NGFEVDs) of an underlying 119953-th
1 Other techniques have also been used in the literature for the study of macroeconomic linkages Ogbuabor et al
(2018) provides an overview of such alternative methodologies such as the dynamic latent factor models of Kose et al (2008) and Canova et al (2007) the CoVaR approach of Adrian and Brunnermeier (2011) and the marginal expected shortfall (MES) approach of Acharya et al (2017) and Brownlees and Engle (2012)
11
order VAR model for the 119925 119961 120783 vector of endogenous variables 119937120061 The VAR(119953) model is specified
as follows
119937120061 = 120630119963 + 120509120051119937120061minus120051120057120051=120783 + 120634120061 (1)
where 120630 is 119925119961120783 vector of intercepts 120509120051 is 119925 119961 119925 coefficient matrix 120057 is the lag order and the
residuals 120634120050120061 ~ 119946119946119941 120782 120634119946119946 so that 120634120061 ~ 120782 120634 where 120634 is positive definite covariance matrix
The optimal VAR lag order selected by Schwarz Information Criterion (SIC) for this study is one
(see Appendix 5) Using the Woldlsquos Representation Theorem the model in equation (1) is
expressed as an infinite order vector moving average representation given by
119937120061 = 120634119957 + 120495120783120518119853minus120783 + 120495120784120518119853minus120784 + hellip = 120495120051120634120061minus120051infin120051=120782 (2)
where 120495120782 = 119920119925 120495120051 = 120509120051 120051 = 120783 120784 hellip and 119920119925 stands for an 119925 times 119925 identity matrix in which all
the principal diagonal elements are ones and all other elements are zeros
The network approach of Diebold and Yilmaz (2009) requires that after estimating the underlying
VAR model the forecast error variance decompositions (FEVDs) are then generated and used to
build linkage measures In this study the interest is in the shocks to the disturbances 120634119947120061 in the
country-specific equations Hence following Pesaran and Shin (1998) Diebold and Yilmaz (2016)
and Greenwood-Nimmo et al (2015) this study adopts the order-invariant generalized forecast error
variance decompositions (GFEVDs) defined as
119918119917119916119933119915 119937119946120061 120634119947120061 119919 = 119941119946119947119944119919
= 120648120634119947119947minus120783 119942119946
prime120495119945120506120634119942119947 sup2119919minus120783119945=120782
119942119946prime120495119945120506120634120495119945
prime 119942119946 119919minus120783119945=120782
(3)
where 119894 119895 = 1 hellip 119873 119867 = 1 2 hellip is the forecast horizon 119890119894 119890119895 is 119873 119909 1 selection vector whose i-th
element (j-th element) is unity with zeros elsewhere Θℎ is the coefficient matrix multiplying the h-
lagged shock vector in the infinite moving-average representation of the non-orthogonalized VAR
120634 is the covariance matrix of the shock vector in the non-orthogonalized VAR and 120648120634119947119947 is the j-th
diagonal element of 120634 (ie the standard deviation of 120576119895 ) We adopted a maximum forecast horizon
of 16 quarters in order to ensure that the long-run results are better captured It must be stressed that
the choice of GFEVDs for this study rather than the orthogonalized forecast error variance
decompositions (OFEVDs) of Diebold and Yilmaz (2009) is particularly based on the fact that the
12
OFEVDs depend on the reordering of the variables in the system such that once the order of
variables in the VAR is reshuffled a different outcome results
Diebold and Yilmaz (2014) explain that shocks are rarely orthogonal in the GFEVD environment
so that sums of forecast error variance contributions are not necessarily unity that is row sums of
the generalized variance decomposition matrix 119863119892119867 are not necessarily unity This renders the
interpretation of the GFEVDs complicated Thus to restore a percentage interpretation of the
GFEVDs this study follows Diebold and Yilmaz (2014) to define the normalized GFEVDs
(NGFEVDs) given by2
119863 119892 = 119889 119894119895119892 where 119889 119894119895
119892=
119889119894119895119892
119889119894119895119892119873
119895=1
119889119894119895119892
= 119918119917119916119933119915 119937119946120061 120634119947120061 119919 (4)
By construction 119889 119894119895119892119873
119895=1 = 1 and 119889 119894119895119892119873
119894 119895=1 = 119873 so that the total sum of the generalized forecast
error variance share of each variable is normalized to 100
33 Construction of the Generalized Trade Linkage Measures (GTLMs)
The various trade linkage measures that are relevant for the ensuing analysis are defined in this sub-
section The intuition behind this framework is quite simple Variance decomposition permits the
splitting of the forecast error variances of each variable in the VAR system into parts attributable to
the various system shocks By so doing it becomes easy to answer the question What fraction of
the h-step-ahead error variance in forecasting 119937120783119957 is due to shocks to 119937120783119957 Shocks to 119937120784119957 Similarly
what fraction of the h-step-ahead error variance in forecasting 119937120784119957 is due to shocks to 119937120783119957 Shocks
to 119937120784119957 And in general what fraction of the h-step-ahead error variance in forecasting 119937119947119957 is due to
shocks to 119937119946119957 119946 = 120783 120784 hellip 119925 Thus the approach marries VAR variance decomposition theory and
network topology theory by recognizing that variance decompositions of VARs form networks and
also characterizing linkages in those variance decomposition networks This in turn characterizes
trade linkages of the variables in our VAR system This is the intuition behind this framework
which we now exploit in the ensuing analysis Diebold and Yilmaz (2015) authoritatively document
this framework and its relation to network theory
2 In what follows and without loss of generality the superscript 119867 is dropped whenever it is not needed for clarity so
that 119863119892119867 and 119889119894119895119892119867
are simply written as 119863119892 and 119889119894119895119892
respectively
13
Table 1 Trade Linkage Schematic
Variables 119937120783 119937120784 ⋯ 119937119925 From Others
119937120783
⋯ 1198891119895
119873
119895=1 119895 ne 1
119937120784
⋯ 1198892119895
119873
119895=1 119895 ne 2
⋮ ⋮ ⋮ ⋱
⋮ ⋮
119937119925
⋯ 119889119873119895
119873
119895=1 119895 ne 119873
To Others
1198891198941
119873
119894=1
119894 ne 1
1198891198942
119873
119894=1
119894 ne 2
⋯ 119889119894119873
119873
119894=1
119894 ne N
1
119873 119889119894119895
119873
119894 119895=1 119894 ne 119895
Source Adapted from Diebold and Yilmaz (2014) and Ogbuabor et al (2016) Note For
simplicity each time series variable in this table 119937119947119957 is written as 119937119947 119947 = 120783 120784 hellip 119925
To construct the GTLMs for this study let us denote the H-step ahead NGFEVDs for the 119925 119961 120783
vector of endogenous variables 119937120061 obtained from equation (4) by 119941119946119947 By cross-tabulating 119941119946119947 the
trade linkage table shown in Table 1 is formed The sum of each row in Table 1 is normalized to
100 in line with equation (4) This table is now used to define the various GTLMs and their
relationships The diagonal entries in Table 1 measure own variance shares (or own-effect) while
the off-diagonal entries measure variance shares arising from shocks to other variables in the
system and are therefore referred to as pairwise directional linkage Accordingly the own-effect
(119919119947) also known as the heatwave is defined as
119919119947 = 119941119947119947 (5)
The total cross-variable variance share (119917119947) captures the spillovers from all other variables to 119937119947120061 as
fractions of the H-step-ahead error variance in the forecasts of 119937119947120061 resulting from 119937119946120061 where
119946 = 120783 120784 hellip 119925 and 119946 ne 119947 This measures the total directional linkage from other variables
(countries) in the system (ie the from-effect) to Zj120009 This means that the from-effect can be used to
capture the role each individual economy in the system plays in a given African economy and it is
14
computed in this study by aggregating the spillovers from all the economies in the system to a
given African economy across all horizons Hence the economy contributing the highest of such
aggregate spillover is deemed to play a dominant role in the particular African economy This study
therefore defines 119917119947 as
119917119947 = 119941119947119946119925119946=120783119946ne119947 (6)
By construction 119919119947 + 119917119947 = 120783 forall 119947
This study also defines the total spillover or total contributions of 119937119947120061 to all other variables (denoted
by 119931119947) as
119931119947 = 119941119946119947119925119946=120783119946ne119947 (7)
By construction 119931119947 measures the total directional linkage from 119937119947120061 to other variables in the system
(ie the to-effect) In other words the to-effect measures the directional linkage from a given
economy (for instance a given African economy) to other economies in the system thereby
showing the impact or influence of that particular African economy on other economies in the VAR
system The net directional linkage (or simply net-effect) of 119937119947120061 is therefore defined as
119925119947 = 119931119947 minus 119917119947 (8)
Since there are N economies in the VAR system it also follows that there are 2N total directional
trade linkages N capturing the total shocks transmitted to others (ie spillover to) and N capturing
the total shocks received from others (ie spillover from) These measures are aptly reflect the
bilateral trade patterns between the African economies and other economies in the system since
they mirror the total exports and total imports for each of the N economies in the system and will
therefore facilitate deeper understanding of the inter-linkages among the African economies We
shall utilize the net-effects to establish the net transmittersreceivers of shocks in the system over
time By construction 119925119947119925119947=120783 = 120782
The most aggregative (non-directional) trade linkage measure in this study which will be used to
evaluate the degree of trade linkage Africa the BRICS and the rest of the global economy is known
as the total trade linkage index (119931119931119923119920) or total-effects and it is defined as
119931119931119923119920 = 120783
119925 119917119947
119925119947=120783 =
120783
119925 119931119947
119925119947=120783 (9)
15
This measure captures the grand total of the off-diagonal elements in Table 1 that is the sum of the
―From Others column or ―To Others row There is only one total trade linkage measure which is
analogous to total global imports or total global exports since the two are identical
An important objective of this study is to determine which of the countries among the BRICS and
in the rest of the global economy exert the most dominant trade influence on African economies
and therefore have the potential to spread trade shocks to Africa To achieve this objective this
study defines two indices dependence and influence indices These indices are necessary to
determine the dependence of the j-th variable (or j-th economy) on external shocks and the
influence of the j-th variable (or j-th economy) on the system as a whole The dependence index is
defined as
119926119947119919 =
119917119947
119919119947+ 119917119947 forall 119947 = 120783 120784 hellip 119925 (10)
where 120782 le 119926119947119919 le 120783 This index expresses the relative importance of external shocks for the j-th
economy in the VAR system such that if 119926119947119919 rarr 120783 then conditions in the j-th economy is open
deeply interlinked and sensitive to external conditions but if 119926119947119919 rarr 120782 then the j-th economy is less
sensitive to external shocks
Similarly the influence index is expressed as
119920119947119919 =
119925119947
119931119947+119917119947 forall 119947 = 120783 120784 hellip 119925 (11)
where minus120783 le 119920119947119919 le 120783 For a given horizon H the j-th economy is a net receiver of trade shocks if
minus120783 le 119920119947119919 lt 0 that is if the index has a negative value a net transmitter of trade shocks if 120782 lt
119920119947119919 le 120783 that is if the index takes a positive value and neither a net receiver or transmitter of trade
shocks if IjH = 0 Thus the influence index measures the extent to which the j-th economy in the
system influences or is influenced by external shocks Overall the coordinate pair (119926119947119919 119920119947
119919) in the
dependence-influence space provides a good representation of the j-th economylsquos role in global real
activities A priori expectation is that African economies which are relatively small open
economies would be located close to the point (1-1) while overwhelmingly open but highly
influential and dominant economies like the USA China and Japan would be located close to the
point (11)
4 Results and Discussions
16
We started this empirical analysis by examining the time series properties of the data The Phillips-
Perron unit root tests showed that all the series are I(1) for all the countries (Please see Appendix 3
which reports for total trade of all the countries To conserve space we do not report separately for
exports and imports since the patterns are similar that is all the series are I(1)) However the test
for long-run or equilibrium relationship using the Johansen System Cointegration test showed that
both the Trace and Maximum Eigenvalue statistics returned full rank indicating that the data are
not cointegrated (Please see Appendix 4 which reports the Trace test in Panel 1 and the Max-
Eigenvalue test in Panel 2) Therefore a VAR in first differences was estimated rather than a vector
error correction model (VECM) Thereafter the GTLMs were computed for all forecasting
horizons 119867 = 12 hellip 16 The average values of the GTLMs over all the horizons are reported
since Diebold and Yilmaz (2016) had shown that the measures follow similar patterns regardless of
the choice of window lengths and forecast horizons As earlier stated we set the maximum forecast
horizon at 16 quarters in order to capture the long-run results better As a robustness check we also
estimated the underlying model and computed the GTLMs separately for exports and imports As
shall soon see the results follow the same qualitative pattern
41 Measuring the degree of trade linkage between Africa the BRICS and the rest of the
global economy
The first specific objective of this study is to measure the degree of trade linkage between African
economies the BRICS and the rest of the global economy In addition we examine here how this
index changed from the short-run (ie from horizon 1) through the long-run (ie until horizon 16)
To achieve this objective we estimated the underlying VAR model of equation (1) and used the
NGFEVDs distilled from this estimation based on equation (4) to compute the Total Trade Linkage
Index (TTLI) based on equation (9) We report the plots of this index in Figure 1 based on the total
trade data while the plots based on the exports and imports data are reported in Appendix 5 This
figure shows how the most aggregated generalized trade linkage measure in this study evolved from
the short-run through the long-run We find that the index rose smoothly from 84 at horizons 1
and 2 to 90 at horizons 15 and 16 In other words the total trade linkage index is higher in the
long-run than in the short-run indicating African economies become more interlinked in the long-
run as the business cycles become more synchronized This finding is consistent with the trend of
global macroeconomic interlinkage reported by Greenwood-Nimmo et al (2015) Diebold and
Yilmaz (2016) and Ogbuabor et al (2018) that the ongoing globalization process is engendering
17
more significant comovement in industrial production fluctuations Furthermore we find that the
TTLI recorded an average value of 87 which shows that the trade linkage between Africa the
BRICS and the rest of the global economy is quite substantial
As shown in Appendix 5 the patterns in this index remained qualitatively unchanged when the
underlying model is estimated using the exports and imports data respectively However a closer
look at panels 1 and 2 of this Appendix reveals that the index recorded higher values across all the
horizons for exports than for imports This aptly captures the pattern of trade between Africa and
the rest of the world which shows that Africalsquos export trade in dollar terms at least in the last
decade is higher than its import trade (see table 1 in appendix 1 of African Trade Statistics
Yearbook 2017) This finding clearly indicates that Africa may be vulnerable to trade shocks
which underlines the motivation behind this study
Figure 1 Total Trade Linkage Index
Source Authors Notes The Total trade linkage Index reported here is the most aggregated non-
directional trade linkage index computed for all horizons following equation (9) Notice that the index rises
smoothly from the short-run (ie horizon 1) towards the long-run (ie horizon 16) indicating that African
economies become more interlinked in the long-run
42 Determining the BRICS countries and other countries in the rest of the world that are
dominating Africarsquos trade and therefore have the potential to spread trade shocks to Africa
An important objective of this study is to determine the BRICS member countries and other
countries in the rest of the world that are dominating Africalsquos trade and therefore have the potential
to spread trade shocks to Africa To achieve this objective we report the from-effect trade linkage
from all the economies in the system to each African economy following equation (6) We include
the heatwave (or own-effect) of equation (5) in this table so that it sums up to 100 We aggregate
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
6
generating forecasts for a core set of macroeconomic factors (or variables) across a number of
countries Their model explicitly allows for the interdependencies that exist between national and
international factors Individual region-specific vector error-correcting models were estimated in
which the domestic variables were related to corresponding foreign variables constructed
exclusively to match the international trade pattern of the country under consideration The
individual country models were then linked in a consistent and cohesive manner to generate
forecasts for all of the variables in the world economy simultaneously Their global model was
estimated for 25 countries grouped into 11 regions using quarterly data over 1979Q1ndash1999Q1 The
degree of regional interdependencies was investigated via generalized impulse responses where the
effects of shocks to a given variable in a given country on the rest of the world were provided The
model was then used to investigate the effects of various global risk scenarios on a banklsquos loan
portfolio
Dees et al (2007) presented a quarterly global model combining individual country vector error-
correcting models in which the domestic variables were related to the country-specific foreign
variables Their global VAR (GVAR) model was estimated for 26 countries the Euro area being
treated as a single economy over the period 1979ndash2003 They provided a theoretical framework
where the GVAR was derived as an approximation to a global unobserved common factor model
Using the average pair-wise cross-section error correlations their GVAR approach was shown to be
quite effective in dealing with the common factor interdependencies and international co-
movements of business cycles They developed a sieve bootstrap procedure for simulation of the
GVAR as a whole which is then used in testing the structural stability of the parameters and for
establishing bootstrap confidence bounds for the impulse responses Finally in addition to
generalized impulse responses their paper considered the use of the GVAR for structurallsquo impulse
response analysis with focus on external shocks for the Euro area economy particularly in response
to shocks to the US
In another study Dees et al (2007) focused on testing the long run macroeconomic relations for
interest rates equity prices and exchange rates suggested by arbitrage in financial and
goods markets They used the global vector autoregressive (GVAR) model to test
for long run restrictions in each countryregion conditioning on the rest of the
world They employed Bootstrapping to compute both the empirical distribution of
7
the impulse responses and the log-likelihood ratio statistic for over-identifying
restrictions Their paper also examined the speed with which adjustments to the
long run relations would take place via the persistence profiles They found strong evidence
in favour of the uncovered interest parity (UIP) and to a lesser extent the Fisher equation across a
number of countries but their results for the purchasing power parity (PPP) were much weaker
Also the transmission of shocks and subsequent adjustments in financial markets were much faster
than those in goods markets
Ogbuabor et al (2016) examined the real and financial connectedness of selected African
economies with the global economy using a network approach They found that the connectedness
of African economies with the global economy is quite sizable with the global financial crisis
increasing the connectedness measures above their pre-crisis levels Their results show that US
EU and Canada dominate Africalsquos equity markets while China India and Japan dominate Africalsquos
real activities Their results suggest that African economies are predominantly small open
economies deeply interconnected but systemically unimportant and vulnerable to headwinds
emanating from the dominant economies in the overall global economy
Lubik and Teo (2003) found that world interest rate shocks are the main driving forces of business
cycles in small open economies while terms of trade shocks are not Thus they challenged the
existing results on the contribution of terms of trade and world interest rate shocks to output
fluctuations in small open economies which had been found to range from less than 10 per cent to
almost 90 per cent They argue that an identification problems lies at the heart of existing vastly
different results They overcame this by estimating a DSGE model using a structural Bayesian
estimation approach applying their methodology to five developed and developing economies The
approach enabled them to efficiently exploit cross-equation restrictions implied by the structural
model
Huidrom et al (2017) provide empirical estimates of the cross border spillovers from Seven largest
emerging market economies (China India Brazil Russia Mexico Indonesia and Turkey) using a
Bayesian vector autoregression model Their results indicate that first spillovers from EM7 are
sizeable a 1 percentage point increase in EM7 growth is associated with a 09 percentage point
increase in growth in other emerging and frontier markets and a 06 percentage point increase in
world growth at the end of three years Second sizeable as they are spillovers from EM7 are still
8
smaller than those from G7 countries (Group of Seven of advanced economies) Specifically
growth in other emerging and frontier markets and the global economy would increase by one-half
to three times more due to a similarly sized increase in G7 growth Third among the EM7
spillovers from China are the largest and permeate globally Their study was based on the
phenomenon that their growth could have significant cross-border spillovers based on their size and
integration The seven largest emerging market economies - China India Brazil Russia Mexico
Indonesia and Turkey - was adjudged to constitute more than one-quarter of global output and
more than half of global output growth during 2010ndash15
This study contributes to this emerging literature in several ways First it quantifies the degree of
trade linkage between Africa the BRIC and the rest of the global economy Two apart from
exposing the patterns of trade shock spillovers between Africa and the rest of the world using total
trade statistics this study also disaggregates the trade shock spillovers using exports and imports
statistics This provides deeper insights into the trade linkages between Africa and the rest of the
world Besides the study extended the empirical method by constructing trade linkage measures in
a coherent and transparent manner for ease of replication
3 Data and Methodology
The data for this study consists of the log of exports imports and total trade for the period 1990Q1-
2016Q4 The choice of this period is based on data availability for some of the African economies
and for Russia whose trade data in the World Development Indicators started in 1990 The BRICS
economies in this study are Brazil Russia India China and South Africa The selected African
economies included in this study are DR Congo Egypt Ghana Morocco Nigeria Tanzania and
Tunisia while the rest of the global economy is accounted for by Australia Canada the Eurozone
(EU) Indonesia Japan Malaysia United States of America (USA) and United Kingdom (UK)
These economies account for the larger chunk of Africalsquos trade while the selected African
economies together with South Africa account substantially for Africalsquos GDP The entire data were
taken from the World Development Indicators (WDI) based on the following indicator names
exports of goods and services (constant 2010 US$) as measure for exports imports of goods and
services (constant 2010 US$) as measure for imports and exports plus imports as measure for total
trade (or simply trade)
9
To ensure that there are enough observations for the analysis the data were converted from annual
to quarterly using Eviewlsquos quadratic match average option This is in line with the methodology
used in compiling the Global VAR database used by Greenwood-Nimmo Nguyen and Shin (2015)
the connectedness of the global economy Other studies that have successfully used the same
methodology are Ogbuabor Eigbiremolen Aneke and Manasseh (2018) and Ogbuabor Orji
Aneke and Erdene-Urnukh (2016) Also to reduce noise and ensure uniform scaling the entire data
were converted into indices (2010Y = 100) and logged prior to estimation Appendix 1 summarizes
the descriptive statistics of the data based on the log transformation of the data Focusing on Panel 1
which reports for total trade we find that among the African economies Ghana DR Congo and
Tanzania are the most volatile as seen from their standard deviations while China India and Russia
are the most volatile among the BRICS economies Among all the countries in the sample Japan
Canada and UK are the least volatile and also recorded the highest mean values The maximum and
minimum values do not suggest the presence of outliers in the data The time series plots of the data
for all the economies in the sample are presented in Appendix 2 based on the log transformation of
the data A close examination of the plots show close comovement among all the countries
indicating that the series track themselves closely The implication of this is that the data may be
cointegrated which will in turn affect the form of the underlying model to be estimated in this
study Thus we shall subject the series to cointegration test as part of the empirical procedures in
this study
31 The Choice of Diebold and Yilmaz (2009) Network Framework
A number of methodologies have been employed in the study of macroeconomic linkages among
entities in the global economy particularly in the study of shock propagation among countries For
instance cross-country correlations-based measures have been used to characterize macroeconomic
linkages among countries (Kehoe et al 1995 Kose et al 2003 Bollerslev 1990 Engle et al
1990 Mantegna 1999 Tumminello et al 2005 Taylor 2007 Gray amp Malone 2008 Engle 2009
Engle amp Kelly 2012) The pitfalls of this approach are twofold namely correlation is simply a
pairwise measure of association and it is non-directional This means that correlation-based
approach cannot handle such questions as ―what is the degree of trade linkage between Africa the
BRICS and the rest of the global economy Unlike the correlation-based measures of
macroeconomic linkage the network approach of Diebold and Yilmaz (2009) is non-pairwise yet
10
directional Granger Causality measures have also been used characterize networks so that the
macroeconomic linkages among entities in the global economy can be described and understood
(Caraiani 2013 Hiemstra amp Jones 1994 Dahlhans amp Eichler 2003 Shojaie amp Michailidis 2010
Billio et al 2012) The main weakness of the Granger Causality approach is that it captures only
pairwise relations and may not be useful in answering important questions like ―what is the degree
of trade linkage between Africa the BRICS and the rest of the global economy1
Furthermore as noted by Diebold and Yilmaz (2016) these alternative methodologies generally
dwell exclusively on testing rather than measurement and estimation of macroeconomic linkages
which are the key issues in this paper This study therefore follows the network approach of
Diebold and Yilmaz (2009) based on its ability to transparently use the size and direction of shocks
to build both directional and non-directional trade linkage measures over a given forecast horizon
According to Ogbuabor et al (2016) studies using this approach have four common features
namely (i) they are generally based on connectedness measures distilled from forecast error
variance decompositions (FEVDs) of an approximating vector autoregressive (VAR) model (ii)
they measure the direction and strength of linkages among entities in the system (iii) they can
identify systemically important entities in the system and (iv) they can study the dynamic nature of
shock propagation among entities in the system In what follows the underlying VAR model for
this study and the construction of the generalized trade linkages measures (GTLMs) are presented
to guide the ensuing analysis
32 Model Specification The Vector Autoregression (VAR) Model
The broad objective of this study is to examine the propagation of trade shock between Africa the
BRICS economies and the rest of the global economy Let 119937120061 be the log of total trade for all the
countries selected for this study so that 119937119947120061 stands for the logged total trade of the j-th country in the
system with 119947 = 120783 120784 hellip 119925 and 119925 is the number of countries selected for the study (which is 20)
Following Diebold and Yilmaz (2009) the trade linkage measures for this study are based on the
normalized generalized forecast error variance decompositions (NGFEVDs) of an underlying 119953-th
1 Other techniques have also been used in the literature for the study of macroeconomic linkages Ogbuabor et al
(2018) provides an overview of such alternative methodologies such as the dynamic latent factor models of Kose et al (2008) and Canova et al (2007) the CoVaR approach of Adrian and Brunnermeier (2011) and the marginal expected shortfall (MES) approach of Acharya et al (2017) and Brownlees and Engle (2012)
11
order VAR model for the 119925 119961 120783 vector of endogenous variables 119937120061 The VAR(119953) model is specified
as follows
119937120061 = 120630119963 + 120509120051119937120061minus120051120057120051=120783 + 120634120061 (1)
where 120630 is 119925119961120783 vector of intercepts 120509120051 is 119925 119961 119925 coefficient matrix 120057 is the lag order and the
residuals 120634120050120061 ~ 119946119946119941 120782 120634119946119946 so that 120634120061 ~ 120782 120634 where 120634 is positive definite covariance matrix
The optimal VAR lag order selected by Schwarz Information Criterion (SIC) for this study is one
(see Appendix 5) Using the Woldlsquos Representation Theorem the model in equation (1) is
expressed as an infinite order vector moving average representation given by
119937120061 = 120634119957 + 120495120783120518119853minus120783 + 120495120784120518119853minus120784 + hellip = 120495120051120634120061minus120051infin120051=120782 (2)
where 120495120782 = 119920119925 120495120051 = 120509120051 120051 = 120783 120784 hellip and 119920119925 stands for an 119925 times 119925 identity matrix in which all
the principal diagonal elements are ones and all other elements are zeros
The network approach of Diebold and Yilmaz (2009) requires that after estimating the underlying
VAR model the forecast error variance decompositions (FEVDs) are then generated and used to
build linkage measures In this study the interest is in the shocks to the disturbances 120634119947120061 in the
country-specific equations Hence following Pesaran and Shin (1998) Diebold and Yilmaz (2016)
and Greenwood-Nimmo et al (2015) this study adopts the order-invariant generalized forecast error
variance decompositions (GFEVDs) defined as
119918119917119916119933119915 119937119946120061 120634119947120061 119919 = 119941119946119947119944119919
= 120648120634119947119947minus120783 119942119946
prime120495119945120506120634119942119947 sup2119919minus120783119945=120782
119942119946prime120495119945120506120634120495119945
prime 119942119946 119919minus120783119945=120782
(3)
where 119894 119895 = 1 hellip 119873 119867 = 1 2 hellip is the forecast horizon 119890119894 119890119895 is 119873 119909 1 selection vector whose i-th
element (j-th element) is unity with zeros elsewhere Θℎ is the coefficient matrix multiplying the h-
lagged shock vector in the infinite moving-average representation of the non-orthogonalized VAR
120634 is the covariance matrix of the shock vector in the non-orthogonalized VAR and 120648120634119947119947 is the j-th
diagonal element of 120634 (ie the standard deviation of 120576119895 ) We adopted a maximum forecast horizon
of 16 quarters in order to ensure that the long-run results are better captured It must be stressed that
the choice of GFEVDs for this study rather than the orthogonalized forecast error variance
decompositions (OFEVDs) of Diebold and Yilmaz (2009) is particularly based on the fact that the
12
OFEVDs depend on the reordering of the variables in the system such that once the order of
variables in the VAR is reshuffled a different outcome results
Diebold and Yilmaz (2014) explain that shocks are rarely orthogonal in the GFEVD environment
so that sums of forecast error variance contributions are not necessarily unity that is row sums of
the generalized variance decomposition matrix 119863119892119867 are not necessarily unity This renders the
interpretation of the GFEVDs complicated Thus to restore a percentage interpretation of the
GFEVDs this study follows Diebold and Yilmaz (2014) to define the normalized GFEVDs
(NGFEVDs) given by2
119863 119892 = 119889 119894119895119892 where 119889 119894119895
119892=
119889119894119895119892
119889119894119895119892119873
119895=1
119889119894119895119892
= 119918119917119916119933119915 119937119946120061 120634119947120061 119919 (4)
By construction 119889 119894119895119892119873
119895=1 = 1 and 119889 119894119895119892119873
119894 119895=1 = 119873 so that the total sum of the generalized forecast
error variance share of each variable is normalized to 100
33 Construction of the Generalized Trade Linkage Measures (GTLMs)
The various trade linkage measures that are relevant for the ensuing analysis are defined in this sub-
section The intuition behind this framework is quite simple Variance decomposition permits the
splitting of the forecast error variances of each variable in the VAR system into parts attributable to
the various system shocks By so doing it becomes easy to answer the question What fraction of
the h-step-ahead error variance in forecasting 119937120783119957 is due to shocks to 119937120783119957 Shocks to 119937120784119957 Similarly
what fraction of the h-step-ahead error variance in forecasting 119937120784119957 is due to shocks to 119937120783119957 Shocks
to 119937120784119957 And in general what fraction of the h-step-ahead error variance in forecasting 119937119947119957 is due to
shocks to 119937119946119957 119946 = 120783 120784 hellip 119925 Thus the approach marries VAR variance decomposition theory and
network topology theory by recognizing that variance decompositions of VARs form networks and
also characterizing linkages in those variance decomposition networks This in turn characterizes
trade linkages of the variables in our VAR system This is the intuition behind this framework
which we now exploit in the ensuing analysis Diebold and Yilmaz (2015) authoritatively document
this framework and its relation to network theory
2 In what follows and without loss of generality the superscript 119867 is dropped whenever it is not needed for clarity so
that 119863119892119867 and 119889119894119895119892119867
are simply written as 119863119892 and 119889119894119895119892
respectively
13
Table 1 Trade Linkage Schematic
Variables 119937120783 119937120784 ⋯ 119937119925 From Others
119937120783
⋯ 1198891119895
119873
119895=1 119895 ne 1
119937120784
⋯ 1198892119895
119873
119895=1 119895 ne 2
⋮ ⋮ ⋮ ⋱
⋮ ⋮
119937119925
⋯ 119889119873119895
119873
119895=1 119895 ne 119873
To Others
1198891198941
119873
119894=1
119894 ne 1
1198891198942
119873
119894=1
119894 ne 2
⋯ 119889119894119873
119873
119894=1
119894 ne N
1
119873 119889119894119895
119873
119894 119895=1 119894 ne 119895
Source Adapted from Diebold and Yilmaz (2014) and Ogbuabor et al (2016) Note For
simplicity each time series variable in this table 119937119947119957 is written as 119937119947 119947 = 120783 120784 hellip 119925
To construct the GTLMs for this study let us denote the H-step ahead NGFEVDs for the 119925 119961 120783
vector of endogenous variables 119937120061 obtained from equation (4) by 119941119946119947 By cross-tabulating 119941119946119947 the
trade linkage table shown in Table 1 is formed The sum of each row in Table 1 is normalized to
100 in line with equation (4) This table is now used to define the various GTLMs and their
relationships The diagonal entries in Table 1 measure own variance shares (or own-effect) while
the off-diagonal entries measure variance shares arising from shocks to other variables in the
system and are therefore referred to as pairwise directional linkage Accordingly the own-effect
(119919119947) also known as the heatwave is defined as
119919119947 = 119941119947119947 (5)
The total cross-variable variance share (119917119947) captures the spillovers from all other variables to 119937119947120061 as
fractions of the H-step-ahead error variance in the forecasts of 119937119947120061 resulting from 119937119946120061 where
119946 = 120783 120784 hellip 119925 and 119946 ne 119947 This measures the total directional linkage from other variables
(countries) in the system (ie the from-effect) to Zj120009 This means that the from-effect can be used to
capture the role each individual economy in the system plays in a given African economy and it is
14
computed in this study by aggregating the spillovers from all the economies in the system to a
given African economy across all horizons Hence the economy contributing the highest of such
aggregate spillover is deemed to play a dominant role in the particular African economy This study
therefore defines 119917119947 as
119917119947 = 119941119947119946119925119946=120783119946ne119947 (6)
By construction 119919119947 + 119917119947 = 120783 forall 119947
This study also defines the total spillover or total contributions of 119937119947120061 to all other variables (denoted
by 119931119947) as
119931119947 = 119941119946119947119925119946=120783119946ne119947 (7)
By construction 119931119947 measures the total directional linkage from 119937119947120061 to other variables in the system
(ie the to-effect) In other words the to-effect measures the directional linkage from a given
economy (for instance a given African economy) to other economies in the system thereby
showing the impact or influence of that particular African economy on other economies in the VAR
system The net directional linkage (or simply net-effect) of 119937119947120061 is therefore defined as
119925119947 = 119931119947 minus 119917119947 (8)
Since there are N economies in the VAR system it also follows that there are 2N total directional
trade linkages N capturing the total shocks transmitted to others (ie spillover to) and N capturing
the total shocks received from others (ie spillover from) These measures are aptly reflect the
bilateral trade patterns between the African economies and other economies in the system since
they mirror the total exports and total imports for each of the N economies in the system and will
therefore facilitate deeper understanding of the inter-linkages among the African economies We
shall utilize the net-effects to establish the net transmittersreceivers of shocks in the system over
time By construction 119925119947119925119947=120783 = 120782
The most aggregative (non-directional) trade linkage measure in this study which will be used to
evaluate the degree of trade linkage Africa the BRICS and the rest of the global economy is known
as the total trade linkage index (119931119931119923119920) or total-effects and it is defined as
119931119931119923119920 = 120783
119925 119917119947
119925119947=120783 =
120783
119925 119931119947
119925119947=120783 (9)
15
This measure captures the grand total of the off-diagonal elements in Table 1 that is the sum of the
―From Others column or ―To Others row There is only one total trade linkage measure which is
analogous to total global imports or total global exports since the two are identical
An important objective of this study is to determine which of the countries among the BRICS and
in the rest of the global economy exert the most dominant trade influence on African economies
and therefore have the potential to spread trade shocks to Africa To achieve this objective this
study defines two indices dependence and influence indices These indices are necessary to
determine the dependence of the j-th variable (or j-th economy) on external shocks and the
influence of the j-th variable (or j-th economy) on the system as a whole The dependence index is
defined as
119926119947119919 =
119917119947
119919119947+ 119917119947 forall 119947 = 120783 120784 hellip 119925 (10)
where 120782 le 119926119947119919 le 120783 This index expresses the relative importance of external shocks for the j-th
economy in the VAR system such that if 119926119947119919 rarr 120783 then conditions in the j-th economy is open
deeply interlinked and sensitive to external conditions but if 119926119947119919 rarr 120782 then the j-th economy is less
sensitive to external shocks
Similarly the influence index is expressed as
119920119947119919 =
119925119947
119931119947+119917119947 forall 119947 = 120783 120784 hellip 119925 (11)
where minus120783 le 119920119947119919 le 120783 For a given horizon H the j-th economy is a net receiver of trade shocks if
minus120783 le 119920119947119919 lt 0 that is if the index has a negative value a net transmitter of trade shocks if 120782 lt
119920119947119919 le 120783 that is if the index takes a positive value and neither a net receiver or transmitter of trade
shocks if IjH = 0 Thus the influence index measures the extent to which the j-th economy in the
system influences or is influenced by external shocks Overall the coordinate pair (119926119947119919 119920119947
119919) in the
dependence-influence space provides a good representation of the j-th economylsquos role in global real
activities A priori expectation is that African economies which are relatively small open
economies would be located close to the point (1-1) while overwhelmingly open but highly
influential and dominant economies like the USA China and Japan would be located close to the
point (11)
4 Results and Discussions
16
We started this empirical analysis by examining the time series properties of the data The Phillips-
Perron unit root tests showed that all the series are I(1) for all the countries (Please see Appendix 3
which reports for total trade of all the countries To conserve space we do not report separately for
exports and imports since the patterns are similar that is all the series are I(1)) However the test
for long-run or equilibrium relationship using the Johansen System Cointegration test showed that
both the Trace and Maximum Eigenvalue statistics returned full rank indicating that the data are
not cointegrated (Please see Appendix 4 which reports the Trace test in Panel 1 and the Max-
Eigenvalue test in Panel 2) Therefore a VAR in first differences was estimated rather than a vector
error correction model (VECM) Thereafter the GTLMs were computed for all forecasting
horizons 119867 = 12 hellip 16 The average values of the GTLMs over all the horizons are reported
since Diebold and Yilmaz (2016) had shown that the measures follow similar patterns regardless of
the choice of window lengths and forecast horizons As earlier stated we set the maximum forecast
horizon at 16 quarters in order to capture the long-run results better As a robustness check we also
estimated the underlying model and computed the GTLMs separately for exports and imports As
shall soon see the results follow the same qualitative pattern
41 Measuring the degree of trade linkage between Africa the BRICS and the rest of the
global economy
The first specific objective of this study is to measure the degree of trade linkage between African
economies the BRICS and the rest of the global economy In addition we examine here how this
index changed from the short-run (ie from horizon 1) through the long-run (ie until horizon 16)
To achieve this objective we estimated the underlying VAR model of equation (1) and used the
NGFEVDs distilled from this estimation based on equation (4) to compute the Total Trade Linkage
Index (TTLI) based on equation (9) We report the plots of this index in Figure 1 based on the total
trade data while the plots based on the exports and imports data are reported in Appendix 5 This
figure shows how the most aggregated generalized trade linkage measure in this study evolved from
the short-run through the long-run We find that the index rose smoothly from 84 at horizons 1
and 2 to 90 at horizons 15 and 16 In other words the total trade linkage index is higher in the
long-run than in the short-run indicating African economies become more interlinked in the long-
run as the business cycles become more synchronized This finding is consistent with the trend of
global macroeconomic interlinkage reported by Greenwood-Nimmo et al (2015) Diebold and
Yilmaz (2016) and Ogbuabor et al (2018) that the ongoing globalization process is engendering
17
more significant comovement in industrial production fluctuations Furthermore we find that the
TTLI recorded an average value of 87 which shows that the trade linkage between Africa the
BRICS and the rest of the global economy is quite substantial
As shown in Appendix 5 the patterns in this index remained qualitatively unchanged when the
underlying model is estimated using the exports and imports data respectively However a closer
look at panels 1 and 2 of this Appendix reveals that the index recorded higher values across all the
horizons for exports than for imports This aptly captures the pattern of trade between Africa and
the rest of the world which shows that Africalsquos export trade in dollar terms at least in the last
decade is higher than its import trade (see table 1 in appendix 1 of African Trade Statistics
Yearbook 2017) This finding clearly indicates that Africa may be vulnerable to trade shocks
which underlines the motivation behind this study
Figure 1 Total Trade Linkage Index
Source Authors Notes The Total trade linkage Index reported here is the most aggregated non-
directional trade linkage index computed for all horizons following equation (9) Notice that the index rises
smoothly from the short-run (ie horizon 1) towards the long-run (ie horizon 16) indicating that African
economies become more interlinked in the long-run
42 Determining the BRICS countries and other countries in the rest of the world that are
dominating Africarsquos trade and therefore have the potential to spread trade shocks to Africa
An important objective of this study is to determine the BRICS member countries and other
countries in the rest of the world that are dominating Africalsquos trade and therefore have the potential
to spread trade shocks to Africa To achieve this objective we report the from-effect trade linkage
from all the economies in the system to each African economy following equation (6) We include
the heatwave (or own-effect) of equation (5) in this table so that it sums up to 100 We aggregate
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
7
the impulse responses and the log-likelihood ratio statistic for over-identifying
restrictions Their paper also examined the speed with which adjustments to the
long run relations would take place via the persistence profiles They found strong evidence
in favour of the uncovered interest parity (UIP) and to a lesser extent the Fisher equation across a
number of countries but their results for the purchasing power parity (PPP) were much weaker
Also the transmission of shocks and subsequent adjustments in financial markets were much faster
than those in goods markets
Ogbuabor et al (2016) examined the real and financial connectedness of selected African
economies with the global economy using a network approach They found that the connectedness
of African economies with the global economy is quite sizable with the global financial crisis
increasing the connectedness measures above their pre-crisis levels Their results show that US
EU and Canada dominate Africalsquos equity markets while China India and Japan dominate Africalsquos
real activities Their results suggest that African economies are predominantly small open
economies deeply interconnected but systemically unimportant and vulnerable to headwinds
emanating from the dominant economies in the overall global economy
Lubik and Teo (2003) found that world interest rate shocks are the main driving forces of business
cycles in small open economies while terms of trade shocks are not Thus they challenged the
existing results on the contribution of terms of trade and world interest rate shocks to output
fluctuations in small open economies which had been found to range from less than 10 per cent to
almost 90 per cent They argue that an identification problems lies at the heart of existing vastly
different results They overcame this by estimating a DSGE model using a structural Bayesian
estimation approach applying their methodology to five developed and developing economies The
approach enabled them to efficiently exploit cross-equation restrictions implied by the structural
model
Huidrom et al (2017) provide empirical estimates of the cross border spillovers from Seven largest
emerging market economies (China India Brazil Russia Mexico Indonesia and Turkey) using a
Bayesian vector autoregression model Their results indicate that first spillovers from EM7 are
sizeable a 1 percentage point increase in EM7 growth is associated with a 09 percentage point
increase in growth in other emerging and frontier markets and a 06 percentage point increase in
world growth at the end of three years Second sizeable as they are spillovers from EM7 are still
8
smaller than those from G7 countries (Group of Seven of advanced economies) Specifically
growth in other emerging and frontier markets and the global economy would increase by one-half
to three times more due to a similarly sized increase in G7 growth Third among the EM7
spillovers from China are the largest and permeate globally Their study was based on the
phenomenon that their growth could have significant cross-border spillovers based on their size and
integration The seven largest emerging market economies - China India Brazil Russia Mexico
Indonesia and Turkey - was adjudged to constitute more than one-quarter of global output and
more than half of global output growth during 2010ndash15
This study contributes to this emerging literature in several ways First it quantifies the degree of
trade linkage between Africa the BRIC and the rest of the global economy Two apart from
exposing the patterns of trade shock spillovers between Africa and the rest of the world using total
trade statistics this study also disaggregates the trade shock spillovers using exports and imports
statistics This provides deeper insights into the trade linkages between Africa and the rest of the
world Besides the study extended the empirical method by constructing trade linkage measures in
a coherent and transparent manner for ease of replication
3 Data and Methodology
The data for this study consists of the log of exports imports and total trade for the period 1990Q1-
2016Q4 The choice of this period is based on data availability for some of the African economies
and for Russia whose trade data in the World Development Indicators started in 1990 The BRICS
economies in this study are Brazil Russia India China and South Africa The selected African
economies included in this study are DR Congo Egypt Ghana Morocco Nigeria Tanzania and
Tunisia while the rest of the global economy is accounted for by Australia Canada the Eurozone
(EU) Indonesia Japan Malaysia United States of America (USA) and United Kingdom (UK)
These economies account for the larger chunk of Africalsquos trade while the selected African
economies together with South Africa account substantially for Africalsquos GDP The entire data were
taken from the World Development Indicators (WDI) based on the following indicator names
exports of goods and services (constant 2010 US$) as measure for exports imports of goods and
services (constant 2010 US$) as measure for imports and exports plus imports as measure for total
trade (or simply trade)
9
To ensure that there are enough observations for the analysis the data were converted from annual
to quarterly using Eviewlsquos quadratic match average option This is in line with the methodology
used in compiling the Global VAR database used by Greenwood-Nimmo Nguyen and Shin (2015)
the connectedness of the global economy Other studies that have successfully used the same
methodology are Ogbuabor Eigbiremolen Aneke and Manasseh (2018) and Ogbuabor Orji
Aneke and Erdene-Urnukh (2016) Also to reduce noise and ensure uniform scaling the entire data
were converted into indices (2010Y = 100) and logged prior to estimation Appendix 1 summarizes
the descriptive statistics of the data based on the log transformation of the data Focusing on Panel 1
which reports for total trade we find that among the African economies Ghana DR Congo and
Tanzania are the most volatile as seen from their standard deviations while China India and Russia
are the most volatile among the BRICS economies Among all the countries in the sample Japan
Canada and UK are the least volatile and also recorded the highest mean values The maximum and
minimum values do not suggest the presence of outliers in the data The time series plots of the data
for all the economies in the sample are presented in Appendix 2 based on the log transformation of
the data A close examination of the plots show close comovement among all the countries
indicating that the series track themselves closely The implication of this is that the data may be
cointegrated which will in turn affect the form of the underlying model to be estimated in this
study Thus we shall subject the series to cointegration test as part of the empirical procedures in
this study
31 The Choice of Diebold and Yilmaz (2009) Network Framework
A number of methodologies have been employed in the study of macroeconomic linkages among
entities in the global economy particularly in the study of shock propagation among countries For
instance cross-country correlations-based measures have been used to characterize macroeconomic
linkages among countries (Kehoe et al 1995 Kose et al 2003 Bollerslev 1990 Engle et al
1990 Mantegna 1999 Tumminello et al 2005 Taylor 2007 Gray amp Malone 2008 Engle 2009
Engle amp Kelly 2012) The pitfalls of this approach are twofold namely correlation is simply a
pairwise measure of association and it is non-directional This means that correlation-based
approach cannot handle such questions as ―what is the degree of trade linkage between Africa the
BRICS and the rest of the global economy Unlike the correlation-based measures of
macroeconomic linkage the network approach of Diebold and Yilmaz (2009) is non-pairwise yet
10
directional Granger Causality measures have also been used characterize networks so that the
macroeconomic linkages among entities in the global economy can be described and understood
(Caraiani 2013 Hiemstra amp Jones 1994 Dahlhans amp Eichler 2003 Shojaie amp Michailidis 2010
Billio et al 2012) The main weakness of the Granger Causality approach is that it captures only
pairwise relations and may not be useful in answering important questions like ―what is the degree
of trade linkage between Africa the BRICS and the rest of the global economy1
Furthermore as noted by Diebold and Yilmaz (2016) these alternative methodologies generally
dwell exclusively on testing rather than measurement and estimation of macroeconomic linkages
which are the key issues in this paper This study therefore follows the network approach of
Diebold and Yilmaz (2009) based on its ability to transparently use the size and direction of shocks
to build both directional and non-directional trade linkage measures over a given forecast horizon
According to Ogbuabor et al (2016) studies using this approach have four common features
namely (i) they are generally based on connectedness measures distilled from forecast error
variance decompositions (FEVDs) of an approximating vector autoregressive (VAR) model (ii)
they measure the direction and strength of linkages among entities in the system (iii) they can
identify systemically important entities in the system and (iv) they can study the dynamic nature of
shock propagation among entities in the system In what follows the underlying VAR model for
this study and the construction of the generalized trade linkages measures (GTLMs) are presented
to guide the ensuing analysis
32 Model Specification The Vector Autoregression (VAR) Model
The broad objective of this study is to examine the propagation of trade shock between Africa the
BRICS economies and the rest of the global economy Let 119937120061 be the log of total trade for all the
countries selected for this study so that 119937119947120061 stands for the logged total trade of the j-th country in the
system with 119947 = 120783 120784 hellip 119925 and 119925 is the number of countries selected for the study (which is 20)
Following Diebold and Yilmaz (2009) the trade linkage measures for this study are based on the
normalized generalized forecast error variance decompositions (NGFEVDs) of an underlying 119953-th
1 Other techniques have also been used in the literature for the study of macroeconomic linkages Ogbuabor et al
(2018) provides an overview of such alternative methodologies such as the dynamic latent factor models of Kose et al (2008) and Canova et al (2007) the CoVaR approach of Adrian and Brunnermeier (2011) and the marginal expected shortfall (MES) approach of Acharya et al (2017) and Brownlees and Engle (2012)
11
order VAR model for the 119925 119961 120783 vector of endogenous variables 119937120061 The VAR(119953) model is specified
as follows
119937120061 = 120630119963 + 120509120051119937120061minus120051120057120051=120783 + 120634120061 (1)
where 120630 is 119925119961120783 vector of intercepts 120509120051 is 119925 119961 119925 coefficient matrix 120057 is the lag order and the
residuals 120634120050120061 ~ 119946119946119941 120782 120634119946119946 so that 120634120061 ~ 120782 120634 where 120634 is positive definite covariance matrix
The optimal VAR lag order selected by Schwarz Information Criterion (SIC) for this study is one
(see Appendix 5) Using the Woldlsquos Representation Theorem the model in equation (1) is
expressed as an infinite order vector moving average representation given by
119937120061 = 120634119957 + 120495120783120518119853minus120783 + 120495120784120518119853minus120784 + hellip = 120495120051120634120061minus120051infin120051=120782 (2)
where 120495120782 = 119920119925 120495120051 = 120509120051 120051 = 120783 120784 hellip and 119920119925 stands for an 119925 times 119925 identity matrix in which all
the principal diagonal elements are ones and all other elements are zeros
The network approach of Diebold and Yilmaz (2009) requires that after estimating the underlying
VAR model the forecast error variance decompositions (FEVDs) are then generated and used to
build linkage measures In this study the interest is in the shocks to the disturbances 120634119947120061 in the
country-specific equations Hence following Pesaran and Shin (1998) Diebold and Yilmaz (2016)
and Greenwood-Nimmo et al (2015) this study adopts the order-invariant generalized forecast error
variance decompositions (GFEVDs) defined as
119918119917119916119933119915 119937119946120061 120634119947120061 119919 = 119941119946119947119944119919
= 120648120634119947119947minus120783 119942119946
prime120495119945120506120634119942119947 sup2119919minus120783119945=120782
119942119946prime120495119945120506120634120495119945
prime 119942119946 119919minus120783119945=120782
(3)
where 119894 119895 = 1 hellip 119873 119867 = 1 2 hellip is the forecast horizon 119890119894 119890119895 is 119873 119909 1 selection vector whose i-th
element (j-th element) is unity with zeros elsewhere Θℎ is the coefficient matrix multiplying the h-
lagged shock vector in the infinite moving-average representation of the non-orthogonalized VAR
120634 is the covariance matrix of the shock vector in the non-orthogonalized VAR and 120648120634119947119947 is the j-th
diagonal element of 120634 (ie the standard deviation of 120576119895 ) We adopted a maximum forecast horizon
of 16 quarters in order to ensure that the long-run results are better captured It must be stressed that
the choice of GFEVDs for this study rather than the orthogonalized forecast error variance
decompositions (OFEVDs) of Diebold and Yilmaz (2009) is particularly based on the fact that the
12
OFEVDs depend on the reordering of the variables in the system such that once the order of
variables in the VAR is reshuffled a different outcome results
Diebold and Yilmaz (2014) explain that shocks are rarely orthogonal in the GFEVD environment
so that sums of forecast error variance contributions are not necessarily unity that is row sums of
the generalized variance decomposition matrix 119863119892119867 are not necessarily unity This renders the
interpretation of the GFEVDs complicated Thus to restore a percentage interpretation of the
GFEVDs this study follows Diebold and Yilmaz (2014) to define the normalized GFEVDs
(NGFEVDs) given by2
119863 119892 = 119889 119894119895119892 where 119889 119894119895
119892=
119889119894119895119892
119889119894119895119892119873
119895=1
119889119894119895119892
= 119918119917119916119933119915 119937119946120061 120634119947120061 119919 (4)
By construction 119889 119894119895119892119873
119895=1 = 1 and 119889 119894119895119892119873
119894 119895=1 = 119873 so that the total sum of the generalized forecast
error variance share of each variable is normalized to 100
33 Construction of the Generalized Trade Linkage Measures (GTLMs)
The various trade linkage measures that are relevant for the ensuing analysis are defined in this sub-
section The intuition behind this framework is quite simple Variance decomposition permits the
splitting of the forecast error variances of each variable in the VAR system into parts attributable to
the various system shocks By so doing it becomes easy to answer the question What fraction of
the h-step-ahead error variance in forecasting 119937120783119957 is due to shocks to 119937120783119957 Shocks to 119937120784119957 Similarly
what fraction of the h-step-ahead error variance in forecasting 119937120784119957 is due to shocks to 119937120783119957 Shocks
to 119937120784119957 And in general what fraction of the h-step-ahead error variance in forecasting 119937119947119957 is due to
shocks to 119937119946119957 119946 = 120783 120784 hellip 119925 Thus the approach marries VAR variance decomposition theory and
network topology theory by recognizing that variance decompositions of VARs form networks and
also characterizing linkages in those variance decomposition networks This in turn characterizes
trade linkages of the variables in our VAR system This is the intuition behind this framework
which we now exploit in the ensuing analysis Diebold and Yilmaz (2015) authoritatively document
this framework and its relation to network theory
2 In what follows and without loss of generality the superscript 119867 is dropped whenever it is not needed for clarity so
that 119863119892119867 and 119889119894119895119892119867
are simply written as 119863119892 and 119889119894119895119892
respectively
13
Table 1 Trade Linkage Schematic
Variables 119937120783 119937120784 ⋯ 119937119925 From Others
119937120783
⋯ 1198891119895
119873
119895=1 119895 ne 1
119937120784
⋯ 1198892119895
119873
119895=1 119895 ne 2
⋮ ⋮ ⋮ ⋱
⋮ ⋮
119937119925
⋯ 119889119873119895
119873
119895=1 119895 ne 119873
To Others
1198891198941
119873
119894=1
119894 ne 1
1198891198942
119873
119894=1
119894 ne 2
⋯ 119889119894119873
119873
119894=1
119894 ne N
1
119873 119889119894119895
119873
119894 119895=1 119894 ne 119895
Source Adapted from Diebold and Yilmaz (2014) and Ogbuabor et al (2016) Note For
simplicity each time series variable in this table 119937119947119957 is written as 119937119947 119947 = 120783 120784 hellip 119925
To construct the GTLMs for this study let us denote the H-step ahead NGFEVDs for the 119925 119961 120783
vector of endogenous variables 119937120061 obtained from equation (4) by 119941119946119947 By cross-tabulating 119941119946119947 the
trade linkage table shown in Table 1 is formed The sum of each row in Table 1 is normalized to
100 in line with equation (4) This table is now used to define the various GTLMs and their
relationships The diagonal entries in Table 1 measure own variance shares (or own-effect) while
the off-diagonal entries measure variance shares arising from shocks to other variables in the
system and are therefore referred to as pairwise directional linkage Accordingly the own-effect
(119919119947) also known as the heatwave is defined as
119919119947 = 119941119947119947 (5)
The total cross-variable variance share (119917119947) captures the spillovers from all other variables to 119937119947120061 as
fractions of the H-step-ahead error variance in the forecasts of 119937119947120061 resulting from 119937119946120061 where
119946 = 120783 120784 hellip 119925 and 119946 ne 119947 This measures the total directional linkage from other variables
(countries) in the system (ie the from-effect) to Zj120009 This means that the from-effect can be used to
capture the role each individual economy in the system plays in a given African economy and it is
14
computed in this study by aggregating the spillovers from all the economies in the system to a
given African economy across all horizons Hence the economy contributing the highest of such
aggregate spillover is deemed to play a dominant role in the particular African economy This study
therefore defines 119917119947 as
119917119947 = 119941119947119946119925119946=120783119946ne119947 (6)
By construction 119919119947 + 119917119947 = 120783 forall 119947
This study also defines the total spillover or total contributions of 119937119947120061 to all other variables (denoted
by 119931119947) as
119931119947 = 119941119946119947119925119946=120783119946ne119947 (7)
By construction 119931119947 measures the total directional linkage from 119937119947120061 to other variables in the system
(ie the to-effect) In other words the to-effect measures the directional linkage from a given
economy (for instance a given African economy) to other economies in the system thereby
showing the impact or influence of that particular African economy on other economies in the VAR
system The net directional linkage (or simply net-effect) of 119937119947120061 is therefore defined as
119925119947 = 119931119947 minus 119917119947 (8)
Since there are N economies in the VAR system it also follows that there are 2N total directional
trade linkages N capturing the total shocks transmitted to others (ie spillover to) and N capturing
the total shocks received from others (ie spillover from) These measures are aptly reflect the
bilateral trade patterns between the African economies and other economies in the system since
they mirror the total exports and total imports for each of the N economies in the system and will
therefore facilitate deeper understanding of the inter-linkages among the African economies We
shall utilize the net-effects to establish the net transmittersreceivers of shocks in the system over
time By construction 119925119947119925119947=120783 = 120782
The most aggregative (non-directional) trade linkage measure in this study which will be used to
evaluate the degree of trade linkage Africa the BRICS and the rest of the global economy is known
as the total trade linkage index (119931119931119923119920) or total-effects and it is defined as
119931119931119923119920 = 120783
119925 119917119947
119925119947=120783 =
120783
119925 119931119947
119925119947=120783 (9)
15
This measure captures the grand total of the off-diagonal elements in Table 1 that is the sum of the
―From Others column or ―To Others row There is only one total trade linkage measure which is
analogous to total global imports or total global exports since the two are identical
An important objective of this study is to determine which of the countries among the BRICS and
in the rest of the global economy exert the most dominant trade influence on African economies
and therefore have the potential to spread trade shocks to Africa To achieve this objective this
study defines two indices dependence and influence indices These indices are necessary to
determine the dependence of the j-th variable (or j-th economy) on external shocks and the
influence of the j-th variable (or j-th economy) on the system as a whole The dependence index is
defined as
119926119947119919 =
119917119947
119919119947+ 119917119947 forall 119947 = 120783 120784 hellip 119925 (10)
where 120782 le 119926119947119919 le 120783 This index expresses the relative importance of external shocks for the j-th
economy in the VAR system such that if 119926119947119919 rarr 120783 then conditions in the j-th economy is open
deeply interlinked and sensitive to external conditions but if 119926119947119919 rarr 120782 then the j-th economy is less
sensitive to external shocks
Similarly the influence index is expressed as
119920119947119919 =
119925119947
119931119947+119917119947 forall 119947 = 120783 120784 hellip 119925 (11)
where minus120783 le 119920119947119919 le 120783 For a given horizon H the j-th economy is a net receiver of trade shocks if
minus120783 le 119920119947119919 lt 0 that is if the index has a negative value a net transmitter of trade shocks if 120782 lt
119920119947119919 le 120783 that is if the index takes a positive value and neither a net receiver or transmitter of trade
shocks if IjH = 0 Thus the influence index measures the extent to which the j-th economy in the
system influences or is influenced by external shocks Overall the coordinate pair (119926119947119919 119920119947
119919) in the
dependence-influence space provides a good representation of the j-th economylsquos role in global real
activities A priori expectation is that African economies which are relatively small open
economies would be located close to the point (1-1) while overwhelmingly open but highly
influential and dominant economies like the USA China and Japan would be located close to the
point (11)
4 Results and Discussions
16
We started this empirical analysis by examining the time series properties of the data The Phillips-
Perron unit root tests showed that all the series are I(1) for all the countries (Please see Appendix 3
which reports for total trade of all the countries To conserve space we do not report separately for
exports and imports since the patterns are similar that is all the series are I(1)) However the test
for long-run or equilibrium relationship using the Johansen System Cointegration test showed that
both the Trace and Maximum Eigenvalue statistics returned full rank indicating that the data are
not cointegrated (Please see Appendix 4 which reports the Trace test in Panel 1 and the Max-
Eigenvalue test in Panel 2) Therefore a VAR in first differences was estimated rather than a vector
error correction model (VECM) Thereafter the GTLMs were computed for all forecasting
horizons 119867 = 12 hellip 16 The average values of the GTLMs over all the horizons are reported
since Diebold and Yilmaz (2016) had shown that the measures follow similar patterns regardless of
the choice of window lengths and forecast horizons As earlier stated we set the maximum forecast
horizon at 16 quarters in order to capture the long-run results better As a robustness check we also
estimated the underlying model and computed the GTLMs separately for exports and imports As
shall soon see the results follow the same qualitative pattern
41 Measuring the degree of trade linkage between Africa the BRICS and the rest of the
global economy
The first specific objective of this study is to measure the degree of trade linkage between African
economies the BRICS and the rest of the global economy In addition we examine here how this
index changed from the short-run (ie from horizon 1) through the long-run (ie until horizon 16)
To achieve this objective we estimated the underlying VAR model of equation (1) and used the
NGFEVDs distilled from this estimation based on equation (4) to compute the Total Trade Linkage
Index (TTLI) based on equation (9) We report the plots of this index in Figure 1 based on the total
trade data while the plots based on the exports and imports data are reported in Appendix 5 This
figure shows how the most aggregated generalized trade linkage measure in this study evolved from
the short-run through the long-run We find that the index rose smoothly from 84 at horizons 1
and 2 to 90 at horizons 15 and 16 In other words the total trade linkage index is higher in the
long-run than in the short-run indicating African economies become more interlinked in the long-
run as the business cycles become more synchronized This finding is consistent with the trend of
global macroeconomic interlinkage reported by Greenwood-Nimmo et al (2015) Diebold and
Yilmaz (2016) and Ogbuabor et al (2018) that the ongoing globalization process is engendering
17
more significant comovement in industrial production fluctuations Furthermore we find that the
TTLI recorded an average value of 87 which shows that the trade linkage between Africa the
BRICS and the rest of the global economy is quite substantial
As shown in Appendix 5 the patterns in this index remained qualitatively unchanged when the
underlying model is estimated using the exports and imports data respectively However a closer
look at panels 1 and 2 of this Appendix reveals that the index recorded higher values across all the
horizons for exports than for imports This aptly captures the pattern of trade between Africa and
the rest of the world which shows that Africalsquos export trade in dollar terms at least in the last
decade is higher than its import trade (see table 1 in appendix 1 of African Trade Statistics
Yearbook 2017) This finding clearly indicates that Africa may be vulnerable to trade shocks
which underlines the motivation behind this study
Figure 1 Total Trade Linkage Index
Source Authors Notes The Total trade linkage Index reported here is the most aggregated non-
directional trade linkage index computed for all horizons following equation (9) Notice that the index rises
smoothly from the short-run (ie horizon 1) towards the long-run (ie horizon 16) indicating that African
economies become more interlinked in the long-run
42 Determining the BRICS countries and other countries in the rest of the world that are
dominating Africarsquos trade and therefore have the potential to spread trade shocks to Africa
An important objective of this study is to determine the BRICS member countries and other
countries in the rest of the world that are dominating Africalsquos trade and therefore have the potential
to spread trade shocks to Africa To achieve this objective we report the from-effect trade linkage
from all the economies in the system to each African economy following equation (6) We include
the heatwave (or own-effect) of equation (5) in this table so that it sums up to 100 We aggregate
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
8
smaller than those from G7 countries (Group of Seven of advanced economies) Specifically
growth in other emerging and frontier markets and the global economy would increase by one-half
to three times more due to a similarly sized increase in G7 growth Third among the EM7
spillovers from China are the largest and permeate globally Their study was based on the
phenomenon that their growth could have significant cross-border spillovers based on their size and
integration The seven largest emerging market economies - China India Brazil Russia Mexico
Indonesia and Turkey - was adjudged to constitute more than one-quarter of global output and
more than half of global output growth during 2010ndash15
This study contributes to this emerging literature in several ways First it quantifies the degree of
trade linkage between Africa the BRIC and the rest of the global economy Two apart from
exposing the patterns of trade shock spillovers between Africa and the rest of the world using total
trade statistics this study also disaggregates the trade shock spillovers using exports and imports
statistics This provides deeper insights into the trade linkages between Africa and the rest of the
world Besides the study extended the empirical method by constructing trade linkage measures in
a coherent and transparent manner for ease of replication
3 Data and Methodology
The data for this study consists of the log of exports imports and total trade for the period 1990Q1-
2016Q4 The choice of this period is based on data availability for some of the African economies
and for Russia whose trade data in the World Development Indicators started in 1990 The BRICS
economies in this study are Brazil Russia India China and South Africa The selected African
economies included in this study are DR Congo Egypt Ghana Morocco Nigeria Tanzania and
Tunisia while the rest of the global economy is accounted for by Australia Canada the Eurozone
(EU) Indonesia Japan Malaysia United States of America (USA) and United Kingdom (UK)
These economies account for the larger chunk of Africalsquos trade while the selected African
economies together with South Africa account substantially for Africalsquos GDP The entire data were
taken from the World Development Indicators (WDI) based on the following indicator names
exports of goods and services (constant 2010 US$) as measure for exports imports of goods and
services (constant 2010 US$) as measure for imports and exports plus imports as measure for total
trade (or simply trade)
9
To ensure that there are enough observations for the analysis the data were converted from annual
to quarterly using Eviewlsquos quadratic match average option This is in line with the methodology
used in compiling the Global VAR database used by Greenwood-Nimmo Nguyen and Shin (2015)
the connectedness of the global economy Other studies that have successfully used the same
methodology are Ogbuabor Eigbiremolen Aneke and Manasseh (2018) and Ogbuabor Orji
Aneke and Erdene-Urnukh (2016) Also to reduce noise and ensure uniform scaling the entire data
were converted into indices (2010Y = 100) and logged prior to estimation Appendix 1 summarizes
the descriptive statistics of the data based on the log transformation of the data Focusing on Panel 1
which reports for total trade we find that among the African economies Ghana DR Congo and
Tanzania are the most volatile as seen from their standard deviations while China India and Russia
are the most volatile among the BRICS economies Among all the countries in the sample Japan
Canada and UK are the least volatile and also recorded the highest mean values The maximum and
minimum values do not suggest the presence of outliers in the data The time series plots of the data
for all the economies in the sample are presented in Appendix 2 based on the log transformation of
the data A close examination of the plots show close comovement among all the countries
indicating that the series track themselves closely The implication of this is that the data may be
cointegrated which will in turn affect the form of the underlying model to be estimated in this
study Thus we shall subject the series to cointegration test as part of the empirical procedures in
this study
31 The Choice of Diebold and Yilmaz (2009) Network Framework
A number of methodologies have been employed in the study of macroeconomic linkages among
entities in the global economy particularly in the study of shock propagation among countries For
instance cross-country correlations-based measures have been used to characterize macroeconomic
linkages among countries (Kehoe et al 1995 Kose et al 2003 Bollerslev 1990 Engle et al
1990 Mantegna 1999 Tumminello et al 2005 Taylor 2007 Gray amp Malone 2008 Engle 2009
Engle amp Kelly 2012) The pitfalls of this approach are twofold namely correlation is simply a
pairwise measure of association and it is non-directional This means that correlation-based
approach cannot handle such questions as ―what is the degree of trade linkage between Africa the
BRICS and the rest of the global economy Unlike the correlation-based measures of
macroeconomic linkage the network approach of Diebold and Yilmaz (2009) is non-pairwise yet
10
directional Granger Causality measures have also been used characterize networks so that the
macroeconomic linkages among entities in the global economy can be described and understood
(Caraiani 2013 Hiemstra amp Jones 1994 Dahlhans amp Eichler 2003 Shojaie amp Michailidis 2010
Billio et al 2012) The main weakness of the Granger Causality approach is that it captures only
pairwise relations and may not be useful in answering important questions like ―what is the degree
of trade linkage between Africa the BRICS and the rest of the global economy1
Furthermore as noted by Diebold and Yilmaz (2016) these alternative methodologies generally
dwell exclusively on testing rather than measurement and estimation of macroeconomic linkages
which are the key issues in this paper This study therefore follows the network approach of
Diebold and Yilmaz (2009) based on its ability to transparently use the size and direction of shocks
to build both directional and non-directional trade linkage measures over a given forecast horizon
According to Ogbuabor et al (2016) studies using this approach have four common features
namely (i) they are generally based on connectedness measures distilled from forecast error
variance decompositions (FEVDs) of an approximating vector autoregressive (VAR) model (ii)
they measure the direction and strength of linkages among entities in the system (iii) they can
identify systemically important entities in the system and (iv) they can study the dynamic nature of
shock propagation among entities in the system In what follows the underlying VAR model for
this study and the construction of the generalized trade linkages measures (GTLMs) are presented
to guide the ensuing analysis
32 Model Specification The Vector Autoregression (VAR) Model
The broad objective of this study is to examine the propagation of trade shock between Africa the
BRICS economies and the rest of the global economy Let 119937120061 be the log of total trade for all the
countries selected for this study so that 119937119947120061 stands for the logged total trade of the j-th country in the
system with 119947 = 120783 120784 hellip 119925 and 119925 is the number of countries selected for the study (which is 20)
Following Diebold and Yilmaz (2009) the trade linkage measures for this study are based on the
normalized generalized forecast error variance decompositions (NGFEVDs) of an underlying 119953-th
1 Other techniques have also been used in the literature for the study of macroeconomic linkages Ogbuabor et al
(2018) provides an overview of such alternative methodologies such as the dynamic latent factor models of Kose et al (2008) and Canova et al (2007) the CoVaR approach of Adrian and Brunnermeier (2011) and the marginal expected shortfall (MES) approach of Acharya et al (2017) and Brownlees and Engle (2012)
11
order VAR model for the 119925 119961 120783 vector of endogenous variables 119937120061 The VAR(119953) model is specified
as follows
119937120061 = 120630119963 + 120509120051119937120061minus120051120057120051=120783 + 120634120061 (1)
where 120630 is 119925119961120783 vector of intercepts 120509120051 is 119925 119961 119925 coefficient matrix 120057 is the lag order and the
residuals 120634120050120061 ~ 119946119946119941 120782 120634119946119946 so that 120634120061 ~ 120782 120634 where 120634 is positive definite covariance matrix
The optimal VAR lag order selected by Schwarz Information Criterion (SIC) for this study is one
(see Appendix 5) Using the Woldlsquos Representation Theorem the model in equation (1) is
expressed as an infinite order vector moving average representation given by
119937120061 = 120634119957 + 120495120783120518119853minus120783 + 120495120784120518119853minus120784 + hellip = 120495120051120634120061minus120051infin120051=120782 (2)
where 120495120782 = 119920119925 120495120051 = 120509120051 120051 = 120783 120784 hellip and 119920119925 stands for an 119925 times 119925 identity matrix in which all
the principal diagonal elements are ones and all other elements are zeros
The network approach of Diebold and Yilmaz (2009) requires that after estimating the underlying
VAR model the forecast error variance decompositions (FEVDs) are then generated and used to
build linkage measures In this study the interest is in the shocks to the disturbances 120634119947120061 in the
country-specific equations Hence following Pesaran and Shin (1998) Diebold and Yilmaz (2016)
and Greenwood-Nimmo et al (2015) this study adopts the order-invariant generalized forecast error
variance decompositions (GFEVDs) defined as
119918119917119916119933119915 119937119946120061 120634119947120061 119919 = 119941119946119947119944119919
= 120648120634119947119947minus120783 119942119946
prime120495119945120506120634119942119947 sup2119919minus120783119945=120782
119942119946prime120495119945120506120634120495119945
prime 119942119946 119919minus120783119945=120782
(3)
where 119894 119895 = 1 hellip 119873 119867 = 1 2 hellip is the forecast horizon 119890119894 119890119895 is 119873 119909 1 selection vector whose i-th
element (j-th element) is unity with zeros elsewhere Θℎ is the coefficient matrix multiplying the h-
lagged shock vector in the infinite moving-average representation of the non-orthogonalized VAR
120634 is the covariance matrix of the shock vector in the non-orthogonalized VAR and 120648120634119947119947 is the j-th
diagonal element of 120634 (ie the standard deviation of 120576119895 ) We adopted a maximum forecast horizon
of 16 quarters in order to ensure that the long-run results are better captured It must be stressed that
the choice of GFEVDs for this study rather than the orthogonalized forecast error variance
decompositions (OFEVDs) of Diebold and Yilmaz (2009) is particularly based on the fact that the
12
OFEVDs depend on the reordering of the variables in the system such that once the order of
variables in the VAR is reshuffled a different outcome results
Diebold and Yilmaz (2014) explain that shocks are rarely orthogonal in the GFEVD environment
so that sums of forecast error variance contributions are not necessarily unity that is row sums of
the generalized variance decomposition matrix 119863119892119867 are not necessarily unity This renders the
interpretation of the GFEVDs complicated Thus to restore a percentage interpretation of the
GFEVDs this study follows Diebold and Yilmaz (2014) to define the normalized GFEVDs
(NGFEVDs) given by2
119863 119892 = 119889 119894119895119892 where 119889 119894119895
119892=
119889119894119895119892
119889119894119895119892119873
119895=1
119889119894119895119892
= 119918119917119916119933119915 119937119946120061 120634119947120061 119919 (4)
By construction 119889 119894119895119892119873
119895=1 = 1 and 119889 119894119895119892119873
119894 119895=1 = 119873 so that the total sum of the generalized forecast
error variance share of each variable is normalized to 100
33 Construction of the Generalized Trade Linkage Measures (GTLMs)
The various trade linkage measures that are relevant for the ensuing analysis are defined in this sub-
section The intuition behind this framework is quite simple Variance decomposition permits the
splitting of the forecast error variances of each variable in the VAR system into parts attributable to
the various system shocks By so doing it becomes easy to answer the question What fraction of
the h-step-ahead error variance in forecasting 119937120783119957 is due to shocks to 119937120783119957 Shocks to 119937120784119957 Similarly
what fraction of the h-step-ahead error variance in forecasting 119937120784119957 is due to shocks to 119937120783119957 Shocks
to 119937120784119957 And in general what fraction of the h-step-ahead error variance in forecasting 119937119947119957 is due to
shocks to 119937119946119957 119946 = 120783 120784 hellip 119925 Thus the approach marries VAR variance decomposition theory and
network topology theory by recognizing that variance decompositions of VARs form networks and
also characterizing linkages in those variance decomposition networks This in turn characterizes
trade linkages of the variables in our VAR system This is the intuition behind this framework
which we now exploit in the ensuing analysis Diebold and Yilmaz (2015) authoritatively document
this framework and its relation to network theory
2 In what follows and without loss of generality the superscript 119867 is dropped whenever it is not needed for clarity so
that 119863119892119867 and 119889119894119895119892119867
are simply written as 119863119892 and 119889119894119895119892
respectively
13
Table 1 Trade Linkage Schematic
Variables 119937120783 119937120784 ⋯ 119937119925 From Others
119937120783
⋯ 1198891119895
119873
119895=1 119895 ne 1
119937120784
⋯ 1198892119895
119873
119895=1 119895 ne 2
⋮ ⋮ ⋮ ⋱
⋮ ⋮
119937119925
⋯ 119889119873119895
119873
119895=1 119895 ne 119873
To Others
1198891198941
119873
119894=1
119894 ne 1
1198891198942
119873
119894=1
119894 ne 2
⋯ 119889119894119873
119873
119894=1
119894 ne N
1
119873 119889119894119895
119873
119894 119895=1 119894 ne 119895
Source Adapted from Diebold and Yilmaz (2014) and Ogbuabor et al (2016) Note For
simplicity each time series variable in this table 119937119947119957 is written as 119937119947 119947 = 120783 120784 hellip 119925
To construct the GTLMs for this study let us denote the H-step ahead NGFEVDs for the 119925 119961 120783
vector of endogenous variables 119937120061 obtained from equation (4) by 119941119946119947 By cross-tabulating 119941119946119947 the
trade linkage table shown in Table 1 is formed The sum of each row in Table 1 is normalized to
100 in line with equation (4) This table is now used to define the various GTLMs and their
relationships The diagonal entries in Table 1 measure own variance shares (or own-effect) while
the off-diagonal entries measure variance shares arising from shocks to other variables in the
system and are therefore referred to as pairwise directional linkage Accordingly the own-effect
(119919119947) also known as the heatwave is defined as
119919119947 = 119941119947119947 (5)
The total cross-variable variance share (119917119947) captures the spillovers from all other variables to 119937119947120061 as
fractions of the H-step-ahead error variance in the forecasts of 119937119947120061 resulting from 119937119946120061 where
119946 = 120783 120784 hellip 119925 and 119946 ne 119947 This measures the total directional linkage from other variables
(countries) in the system (ie the from-effect) to Zj120009 This means that the from-effect can be used to
capture the role each individual economy in the system plays in a given African economy and it is
14
computed in this study by aggregating the spillovers from all the economies in the system to a
given African economy across all horizons Hence the economy contributing the highest of such
aggregate spillover is deemed to play a dominant role in the particular African economy This study
therefore defines 119917119947 as
119917119947 = 119941119947119946119925119946=120783119946ne119947 (6)
By construction 119919119947 + 119917119947 = 120783 forall 119947
This study also defines the total spillover or total contributions of 119937119947120061 to all other variables (denoted
by 119931119947) as
119931119947 = 119941119946119947119925119946=120783119946ne119947 (7)
By construction 119931119947 measures the total directional linkage from 119937119947120061 to other variables in the system
(ie the to-effect) In other words the to-effect measures the directional linkage from a given
economy (for instance a given African economy) to other economies in the system thereby
showing the impact or influence of that particular African economy on other economies in the VAR
system The net directional linkage (or simply net-effect) of 119937119947120061 is therefore defined as
119925119947 = 119931119947 minus 119917119947 (8)
Since there are N economies in the VAR system it also follows that there are 2N total directional
trade linkages N capturing the total shocks transmitted to others (ie spillover to) and N capturing
the total shocks received from others (ie spillover from) These measures are aptly reflect the
bilateral trade patterns between the African economies and other economies in the system since
they mirror the total exports and total imports for each of the N economies in the system and will
therefore facilitate deeper understanding of the inter-linkages among the African economies We
shall utilize the net-effects to establish the net transmittersreceivers of shocks in the system over
time By construction 119925119947119925119947=120783 = 120782
The most aggregative (non-directional) trade linkage measure in this study which will be used to
evaluate the degree of trade linkage Africa the BRICS and the rest of the global economy is known
as the total trade linkage index (119931119931119923119920) or total-effects and it is defined as
119931119931119923119920 = 120783
119925 119917119947
119925119947=120783 =
120783
119925 119931119947
119925119947=120783 (9)
15
This measure captures the grand total of the off-diagonal elements in Table 1 that is the sum of the
―From Others column or ―To Others row There is only one total trade linkage measure which is
analogous to total global imports or total global exports since the two are identical
An important objective of this study is to determine which of the countries among the BRICS and
in the rest of the global economy exert the most dominant trade influence on African economies
and therefore have the potential to spread trade shocks to Africa To achieve this objective this
study defines two indices dependence and influence indices These indices are necessary to
determine the dependence of the j-th variable (or j-th economy) on external shocks and the
influence of the j-th variable (or j-th economy) on the system as a whole The dependence index is
defined as
119926119947119919 =
119917119947
119919119947+ 119917119947 forall 119947 = 120783 120784 hellip 119925 (10)
where 120782 le 119926119947119919 le 120783 This index expresses the relative importance of external shocks for the j-th
economy in the VAR system such that if 119926119947119919 rarr 120783 then conditions in the j-th economy is open
deeply interlinked and sensitive to external conditions but if 119926119947119919 rarr 120782 then the j-th economy is less
sensitive to external shocks
Similarly the influence index is expressed as
119920119947119919 =
119925119947
119931119947+119917119947 forall 119947 = 120783 120784 hellip 119925 (11)
where minus120783 le 119920119947119919 le 120783 For a given horizon H the j-th economy is a net receiver of trade shocks if
minus120783 le 119920119947119919 lt 0 that is if the index has a negative value a net transmitter of trade shocks if 120782 lt
119920119947119919 le 120783 that is if the index takes a positive value and neither a net receiver or transmitter of trade
shocks if IjH = 0 Thus the influence index measures the extent to which the j-th economy in the
system influences or is influenced by external shocks Overall the coordinate pair (119926119947119919 119920119947
119919) in the
dependence-influence space provides a good representation of the j-th economylsquos role in global real
activities A priori expectation is that African economies which are relatively small open
economies would be located close to the point (1-1) while overwhelmingly open but highly
influential and dominant economies like the USA China and Japan would be located close to the
point (11)
4 Results and Discussions
16
We started this empirical analysis by examining the time series properties of the data The Phillips-
Perron unit root tests showed that all the series are I(1) for all the countries (Please see Appendix 3
which reports for total trade of all the countries To conserve space we do not report separately for
exports and imports since the patterns are similar that is all the series are I(1)) However the test
for long-run or equilibrium relationship using the Johansen System Cointegration test showed that
both the Trace and Maximum Eigenvalue statistics returned full rank indicating that the data are
not cointegrated (Please see Appendix 4 which reports the Trace test in Panel 1 and the Max-
Eigenvalue test in Panel 2) Therefore a VAR in first differences was estimated rather than a vector
error correction model (VECM) Thereafter the GTLMs were computed for all forecasting
horizons 119867 = 12 hellip 16 The average values of the GTLMs over all the horizons are reported
since Diebold and Yilmaz (2016) had shown that the measures follow similar patterns regardless of
the choice of window lengths and forecast horizons As earlier stated we set the maximum forecast
horizon at 16 quarters in order to capture the long-run results better As a robustness check we also
estimated the underlying model and computed the GTLMs separately for exports and imports As
shall soon see the results follow the same qualitative pattern
41 Measuring the degree of trade linkage between Africa the BRICS and the rest of the
global economy
The first specific objective of this study is to measure the degree of trade linkage between African
economies the BRICS and the rest of the global economy In addition we examine here how this
index changed from the short-run (ie from horizon 1) through the long-run (ie until horizon 16)
To achieve this objective we estimated the underlying VAR model of equation (1) and used the
NGFEVDs distilled from this estimation based on equation (4) to compute the Total Trade Linkage
Index (TTLI) based on equation (9) We report the plots of this index in Figure 1 based on the total
trade data while the plots based on the exports and imports data are reported in Appendix 5 This
figure shows how the most aggregated generalized trade linkage measure in this study evolved from
the short-run through the long-run We find that the index rose smoothly from 84 at horizons 1
and 2 to 90 at horizons 15 and 16 In other words the total trade linkage index is higher in the
long-run than in the short-run indicating African economies become more interlinked in the long-
run as the business cycles become more synchronized This finding is consistent with the trend of
global macroeconomic interlinkage reported by Greenwood-Nimmo et al (2015) Diebold and
Yilmaz (2016) and Ogbuabor et al (2018) that the ongoing globalization process is engendering
17
more significant comovement in industrial production fluctuations Furthermore we find that the
TTLI recorded an average value of 87 which shows that the trade linkage between Africa the
BRICS and the rest of the global economy is quite substantial
As shown in Appendix 5 the patterns in this index remained qualitatively unchanged when the
underlying model is estimated using the exports and imports data respectively However a closer
look at panels 1 and 2 of this Appendix reveals that the index recorded higher values across all the
horizons for exports than for imports This aptly captures the pattern of trade between Africa and
the rest of the world which shows that Africalsquos export trade in dollar terms at least in the last
decade is higher than its import trade (see table 1 in appendix 1 of African Trade Statistics
Yearbook 2017) This finding clearly indicates that Africa may be vulnerable to trade shocks
which underlines the motivation behind this study
Figure 1 Total Trade Linkage Index
Source Authors Notes The Total trade linkage Index reported here is the most aggregated non-
directional trade linkage index computed for all horizons following equation (9) Notice that the index rises
smoothly from the short-run (ie horizon 1) towards the long-run (ie horizon 16) indicating that African
economies become more interlinked in the long-run
42 Determining the BRICS countries and other countries in the rest of the world that are
dominating Africarsquos trade and therefore have the potential to spread trade shocks to Africa
An important objective of this study is to determine the BRICS member countries and other
countries in the rest of the world that are dominating Africalsquos trade and therefore have the potential
to spread trade shocks to Africa To achieve this objective we report the from-effect trade linkage
from all the economies in the system to each African economy following equation (6) We include
the heatwave (or own-effect) of equation (5) in this table so that it sums up to 100 We aggregate
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
9
To ensure that there are enough observations for the analysis the data were converted from annual
to quarterly using Eviewlsquos quadratic match average option This is in line with the methodology
used in compiling the Global VAR database used by Greenwood-Nimmo Nguyen and Shin (2015)
the connectedness of the global economy Other studies that have successfully used the same
methodology are Ogbuabor Eigbiremolen Aneke and Manasseh (2018) and Ogbuabor Orji
Aneke and Erdene-Urnukh (2016) Also to reduce noise and ensure uniform scaling the entire data
were converted into indices (2010Y = 100) and logged prior to estimation Appendix 1 summarizes
the descriptive statistics of the data based on the log transformation of the data Focusing on Panel 1
which reports for total trade we find that among the African economies Ghana DR Congo and
Tanzania are the most volatile as seen from their standard deviations while China India and Russia
are the most volatile among the BRICS economies Among all the countries in the sample Japan
Canada and UK are the least volatile and also recorded the highest mean values The maximum and
minimum values do not suggest the presence of outliers in the data The time series plots of the data
for all the economies in the sample are presented in Appendix 2 based on the log transformation of
the data A close examination of the plots show close comovement among all the countries
indicating that the series track themselves closely The implication of this is that the data may be
cointegrated which will in turn affect the form of the underlying model to be estimated in this
study Thus we shall subject the series to cointegration test as part of the empirical procedures in
this study
31 The Choice of Diebold and Yilmaz (2009) Network Framework
A number of methodologies have been employed in the study of macroeconomic linkages among
entities in the global economy particularly in the study of shock propagation among countries For
instance cross-country correlations-based measures have been used to characterize macroeconomic
linkages among countries (Kehoe et al 1995 Kose et al 2003 Bollerslev 1990 Engle et al
1990 Mantegna 1999 Tumminello et al 2005 Taylor 2007 Gray amp Malone 2008 Engle 2009
Engle amp Kelly 2012) The pitfalls of this approach are twofold namely correlation is simply a
pairwise measure of association and it is non-directional This means that correlation-based
approach cannot handle such questions as ―what is the degree of trade linkage between Africa the
BRICS and the rest of the global economy Unlike the correlation-based measures of
macroeconomic linkage the network approach of Diebold and Yilmaz (2009) is non-pairwise yet
10
directional Granger Causality measures have also been used characterize networks so that the
macroeconomic linkages among entities in the global economy can be described and understood
(Caraiani 2013 Hiemstra amp Jones 1994 Dahlhans amp Eichler 2003 Shojaie amp Michailidis 2010
Billio et al 2012) The main weakness of the Granger Causality approach is that it captures only
pairwise relations and may not be useful in answering important questions like ―what is the degree
of trade linkage between Africa the BRICS and the rest of the global economy1
Furthermore as noted by Diebold and Yilmaz (2016) these alternative methodologies generally
dwell exclusively on testing rather than measurement and estimation of macroeconomic linkages
which are the key issues in this paper This study therefore follows the network approach of
Diebold and Yilmaz (2009) based on its ability to transparently use the size and direction of shocks
to build both directional and non-directional trade linkage measures over a given forecast horizon
According to Ogbuabor et al (2016) studies using this approach have four common features
namely (i) they are generally based on connectedness measures distilled from forecast error
variance decompositions (FEVDs) of an approximating vector autoregressive (VAR) model (ii)
they measure the direction and strength of linkages among entities in the system (iii) they can
identify systemically important entities in the system and (iv) they can study the dynamic nature of
shock propagation among entities in the system In what follows the underlying VAR model for
this study and the construction of the generalized trade linkages measures (GTLMs) are presented
to guide the ensuing analysis
32 Model Specification The Vector Autoregression (VAR) Model
The broad objective of this study is to examine the propagation of trade shock between Africa the
BRICS economies and the rest of the global economy Let 119937120061 be the log of total trade for all the
countries selected for this study so that 119937119947120061 stands for the logged total trade of the j-th country in the
system with 119947 = 120783 120784 hellip 119925 and 119925 is the number of countries selected for the study (which is 20)
Following Diebold and Yilmaz (2009) the trade linkage measures for this study are based on the
normalized generalized forecast error variance decompositions (NGFEVDs) of an underlying 119953-th
1 Other techniques have also been used in the literature for the study of macroeconomic linkages Ogbuabor et al
(2018) provides an overview of such alternative methodologies such as the dynamic latent factor models of Kose et al (2008) and Canova et al (2007) the CoVaR approach of Adrian and Brunnermeier (2011) and the marginal expected shortfall (MES) approach of Acharya et al (2017) and Brownlees and Engle (2012)
11
order VAR model for the 119925 119961 120783 vector of endogenous variables 119937120061 The VAR(119953) model is specified
as follows
119937120061 = 120630119963 + 120509120051119937120061minus120051120057120051=120783 + 120634120061 (1)
where 120630 is 119925119961120783 vector of intercepts 120509120051 is 119925 119961 119925 coefficient matrix 120057 is the lag order and the
residuals 120634120050120061 ~ 119946119946119941 120782 120634119946119946 so that 120634120061 ~ 120782 120634 where 120634 is positive definite covariance matrix
The optimal VAR lag order selected by Schwarz Information Criterion (SIC) for this study is one
(see Appendix 5) Using the Woldlsquos Representation Theorem the model in equation (1) is
expressed as an infinite order vector moving average representation given by
119937120061 = 120634119957 + 120495120783120518119853minus120783 + 120495120784120518119853minus120784 + hellip = 120495120051120634120061minus120051infin120051=120782 (2)
where 120495120782 = 119920119925 120495120051 = 120509120051 120051 = 120783 120784 hellip and 119920119925 stands for an 119925 times 119925 identity matrix in which all
the principal diagonal elements are ones and all other elements are zeros
The network approach of Diebold and Yilmaz (2009) requires that after estimating the underlying
VAR model the forecast error variance decompositions (FEVDs) are then generated and used to
build linkage measures In this study the interest is in the shocks to the disturbances 120634119947120061 in the
country-specific equations Hence following Pesaran and Shin (1998) Diebold and Yilmaz (2016)
and Greenwood-Nimmo et al (2015) this study adopts the order-invariant generalized forecast error
variance decompositions (GFEVDs) defined as
119918119917119916119933119915 119937119946120061 120634119947120061 119919 = 119941119946119947119944119919
= 120648120634119947119947minus120783 119942119946
prime120495119945120506120634119942119947 sup2119919minus120783119945=120782
119942119946prime120495119945120506120634120495119945
prime 119942119946 119919minus120783119945=120782
(3)
where 119894 119895 = 1 hellip 119873 119867 = 1 2 hellip is the forecast horizon 119890119894 119890119895 is 119873 119909 1 selection vector whose i-th
element (j-th element) is unity with zeros elsewhere Θℎ is the coefficient matrix multiplying the h-
lagged shock vector in the infinite moving-average representation of the non-orthogonalized VAR
120634 is the covariance matrix of the shock vector in the non-orthogonalized VAR and 120648120634119947119947 is the j-th
diagonal element of 120634 (ie the standard deviation of 120576119895 ) We adopted a maximum forecast horizon
of 16 quarters in order to ensure that the long-run results are better captured It must be stressed that
the choice of GFEVDs for this study rather than the orthogonalized forecast error variance
decompositions (OFEVDs) of Diebold and Yilmaz (2009) is particularly based on the fact that the
12
OFEVDs depend on the reordering of the variables in the system such that once the order of
variables in the VAR is reshuffled a different outcome results
Diebold and Yilmaz (2014) explain that shocks are rarely orthogonal in the GFEVD environment
so that sums of forecast error variance contributions are not necessarily unity that is row sums of
the generalized variance decomposition matrix 119863119892119867 are not necessarily unity This renders the
interpretation of the GFEVDs complicated Thus to restore a percentage interpretation of the
GFEVDs this study follows Diebold and Yilmaz (2014) to define the normalized GFEVDs
(NGFEVDs) given by2
119863 119892 = 119889 119894119895119892 where 119889 119894119895
119892=
119889119894119895119892
119889119894119895119892119873
119895=1
119889119894119895119892
= 119918119917119916119933119915 119937119946120061 120634119947120061 119919 (4)
By construction 119889 119894119895119892119873
119895=1 = 1 and 119889 119894119895119892119873
119894 119895=1 = 119873 so that the total sum of the generalized forecast
error variance share of each variable is normalized to 100
33 Construction of the Generalized Trade Linkage Measures (GTLMs)
The various trade linkage measures that are relevant for the ensuing analysis are defined in this sub-
section The intuition behind this framework is quite simple Variance decomposition permits the
splitting of the forecast error variances of each variable in the VAR system into parts attributable to
the various system shocks By so doing it becomes easy to answer the question What fraction of
the h-step-ahead error variance in forecasting 119937120783119957 is due to shocks to 119937120783119957 Shocks to 119937120784119957 Similarly
what fraction of the h-step-ahead error variance in forecasting 119937120784119957 is due to shocks to 119937120783119957 Shocks
to 119937120784119957 And in general what fraction of the h-step-ahead error variance in forecasting 119937119947119957 is due to
shocks to 119937119946119957 119946 = 120783 120784 hellip 119925 Thus the approach marries VAR variance decomposition theory and
network topology theory by recognizing that variance decompositions of VARs form networks and
also characterizing linkages in those variance decomposition networks This in turn characterizes
trade linkages of the variables in our VAR system This is the intuition behind this framework
which we now exploit in the ensuing analysis Diebold and Yilmaz (2015) authoritatively document
this framework and its relation to network theory
2 In what follows and without loss of generality the superscript 119867 is dropped whenever it is not needed for clarity so
that 119863119892119867 and 119889119894119895119892119867
are simply written as 119863119892 and 119889119894119895119892
respectively
13
Table 1 Trade Linkage Schematic
Variables 119937120783 119937120784 ⋯ 119937119925 From Others
119937120783
⋯ 1198891119895
119873
119895=1 119895 ne 1
119937120784
⋯ 1198892119895
119873
119895=1 119895 ne 2
⋮ ⋮ ⋮ ⋱
⋮ ⋮
119937119925
⋯ 119889119873119895
119873
119895=1 119895 ne 119873
To Others
1198891198941
119873
119894=1
119894 ne 1
1198891198942
119873
119894=1
119894 ne 2
⋯ 119889119894119873
119873
119894=1
119894 ne N
1
119873 119889119894119895
119873
119894 119895=1 119894 ne 119895
Source Adapted from Diebold and Yilmaz (2014) and Ogbuabor et al (2016) Note For
simplicity each time series variable in this table 119937119947119957 is written as 119937119947 119947 = 120783 120784 hellip 119925
To construct the GTLMs for this study let us denote the H-step ahead NGFEVDs for the 119925 119961 120783
vector of endogenous variables 119937120061 obtained from equation (4) by 119941119946119947 By cross-tabulating 119941119946119947 the
trade linkage table shown in Table 1 is formed The sum of each row in Table 1 is normalized to
100 in line with equation (4) This table is now used to define the various GTLMs and their
relationships The diagonal entries in Table 1 measure own variance shares (or own-effect) while
the off-diagonal entries measure variance shares arising from shocks to other variables in the
system and are therefore referred to as pairwise directional linkage Accordingly the own-effect
(119919119947) also known as the heatwave is defined as
119919119947 = 119941119947119947 (5)
The total cross-variable variance share (119917119947) captures the spillovers from all other variables to 119937119947120061 as
fractions of the H-step-ahead error variance in the forecasts of 119937119947120061 resulting from 119937119946120061 where
119946 = 120783 120784 hellip 119925 and 119946 ne 119947 This measures the total directional linkage from other variables
(countries) in the system (ie the from-effect) to Zj120009 This means that the from-effect can be used to
capture the role each individual economy in the system plays in a given African economy and it is
14
computed in this study by aggregating the spillovers from all the economies in the system to a
given African economy across all horizons Hence the economy contributing the highest of such
aggregate spillover is deemed to play a dominant role in the particular African economy This study
therefore defines 119917119947 as
119917119947 = 119941119947119946119925119946=120783119946ne119947 (6)
By construction 119919119947 + 119917119947 = 120783 forall 119947
This study also defines the total spillover or total contributions of 119937119947120061 to all other variables (denoted
by 119931119947) as
119931119947 = 119941119946119947119925119946=120783119946ne119947 (7)
By construction 119931119947 measures the total directional linkage from 119937119947120061 to other variables in the system
(ie the to-effect) In other words the to-effect measures the directional linkage from a given
economy (for instance a given African economy) to other economies in the system thereby
showing the impact or influence of that particular African economy on other economies in the VAR
system The net directional linkage (or simply net-effect) of 119937119947120061 is therefore defined as
119925119947 = 119931119947 minus 119917119947 (8)
Since there are N economies in the VAR system it also follows that there are 2N total directional
trade linkages N capturing the total shocks transmitted to others (ie spillover to) and N capturing
the total shocks received from others (ie spillover from) These measures are aptly reflect the
bilateral trade patterns between the African economies and other economies in the system since
they mirror the total exports and total imports for each of the N economies in the system and will
therefore facilitate deeper understanding of the inter-linkages among the African economies We
shall utilize the net-effects to establish the net transmittersreceivers of shocks in the system over
time By construction 119925119947119925119947=120783 = 120782
The most aggregative (non-directional) trade linkage measure in this study which will be used to
evaluate the degree of trade linkage Africa the BRICS and the rest of the global economy is known
as the total trade linkage index (119931119931119923119920) or total-effects and it is defined as
119931119931119923119920 = 120783
119925 119917119947
119925119947=120783 =
120783
119925 119931119947
119925119947=120783 (9)
15
This measure captures the grand total of the off-diagonal elements in Table 1 that is the sum of the
―From Others column or ―To Others row There is only one total trade linkage measure which is
analogous to total global imports or total global exports since the two are identical
An important objective of this study is to determine which of the countries among the BRICS and
in the rest of the global economy exert the most dominant trade influence on African economies
and therefore have the potential to spread trade shocks to Africa To achieve this objective this
study defines two indices dependence and influence indices These indices are necessary to
determine the dependence of the j-th variable (or j-th economy) on external shocks and the
influence of the j-th variable (or j-th economy) on the system as a whole The dependence index is
defined as
119926119947119919 =
119917119947
119919119947+ 119917119947 forall 119947 = 120783 120784 hellip 119925 (10)
where 120782 le 119926119947119919 le 120783 This index expresses the relative importance of external shocks for the j-th
economy in the VAR system such that if 119926119947119919 rarr 120783 then conditions in the j-th economy is open
deeply interlinked and sensitive to external conditions but if 119926119947119919 rarr 120782 then the j-th economy is less
sensitive to external shocks
Similarly the influence index is expressed as
119920119947119919 =
119925119947
119931119947+119917119947 forall 119947 = 120783 120784 hellip 119925 (11)
where minus120783 le 119920119947119919 le 120783 For a given horizon H the j-th economy is a net receiver of trade shocks if
minus120783 le 119920119947119919 lt 0 that is if the index has a negative value a net transmitter of trade shocks if 120782 lt
119920119947119919 le 120783 that is if the index takes a positive value and neither a net receiver or transmitter of trade
shocks if IjH = 0 Thus the influence index measures the extent to which the j-th economy in the
system influences or is influenced by external shocks Overall the coordinate pair (119926119947119919 119920119947
119919) in the
dependence-influence space provides a good representation of the j-th economylsquos role in global real
activities A priori expectation is that African economies which are relatively small open
economies would be located close to the point (1-1) while overwhelmingly open but highly
influential and dominant economies like the USA China and Japan would be located close to the
point (11)
4 Results and Discussions
16
We started this empirical analysis by examining the time series properties of the data The Phillips-
Perron unit root tests showed that all the series are I(1) for all the countries (Please see Appendix 3
which reports for total trade of all the countries To conserve space we do not report separately for
exports and imports since the patterns are similar that is all the series are I(1)) However the test
for long-run or equilibrium relationship using the Johansen System Cointegration test showed that
both the Trace and Maximum Eigenvalue statistics returned full rank indicating that the data are
not cointegrated (Please see Appendix 4 which reports the Trace test in Panel 1 and the Max-
Eigenvalue test in Panel 2) Therefore a VAR in first differences was estimated rather than a vector
error correction model (VECM) Thereafter the GTLMs were computed for all forecasting
horizons 119867 = 12 hellip 16 The average values of the GTLMs over all the horizons are reported
since Diebold and Yilmaz (2016) had shown that the measures follow similar patterns regardless of
the choice of window lengths and forecast horizons As earlier stated we set the maximum forecast
horizon at 16 quarters in order to capture the long-run results better As a robustness check we also
estimated the underlying model and computed the GTLMs separately for exports and imports As
shall soon see the results follow the same qualitative pattern
41 Measuring the degree of trade linkage between Africa the BRICS and the rest of the
global economy
The first specific objective of this study is to measure the degree of trade linkage between African
economies the BRICS and the rest of the global economy In addition we examine here how this
index changed from the short-run (ie from horizon 1) through the long-run (ie until horizon 16)
To achieve this objective we estimated the underlying VAR model of equation (1) and used the
NGFEVDs distilled from this estimation based on equation (4) to compute the Total Trade Linkage
Index (TTLI) based on equation (9) We report the plots of this index in Figure 1 based on the total
trade data while the plots based on the exports and imports data are reported in Appendix 5 This
figure shows how the most aggregated generalized trade linkage measure in this study evolved from
the short-run through the long-run We find that the index rose smoothly from 84 at horizons 1
and 2 to 90 at horizons 15 and 16 In other words the total trade linkage index is higher in the
long-run than in the short-run indicating African economies become more interlinked in the long-
run as the business cycles become more synchronized This finding is consistent with the trend of
global macroeconomic interlinkage reported by Greenwood-Nimmo et al (2015) Diebold and
Yilmaz (2016) and Ogbuabor et al (2018) that the ongoing globalization process is engendering
17
more significant comovement in industrial production fluctuations Furthermore we find that the
TTLI recorded an average value of 87 which shows that the trade linkage between Africa the
BRICS and the rest of the global economy is quite substantial
As shown in Appendix 5 the patterns in this index remained qualitatively unchanged when the
underlying model is estimated using the exports and imports data respectively However a closer
look at panels 1 and 2 of this Appendix reveals that the index recorded higher values across all the
horizons for exports than for imports This aptly captures the pattern of trade between Africa and
the rest of the world which shows that Africalsquos export trade in dollar terms at least in the last
decade is higher than its import trade (see table 1 in appendix 1 of African Trade Statistics
Yearbook 2017) This finding clearly indicates that Africa may be vulnerable to trade shocks
which underlines the motivation behind this study
Figure 1 Total Trade Linkage Index
Source Authors Notes The Total trade linkage Index reported here is the most aggregated non-
directional trade linkage index computed for all horizons following equation (9) Notice that the index rises
smoothly from the short-run (ie horizon 1) towards the long-run (ie horizon 16) indicating that African
economies become more interlinked in the long-run
42 Determining the BRICS countries and other countries in the rest of the world that are
dominating Africarsquos trade and therefore have the potential to spread trade shocks to Africa
An important objective of this study is to determine the BRICS member countries and other
countries in the rest of the world that are dominating Africalsquos trade and therefore have the potential
to spread trade shocks to Africa To achieve this objective we report the from-effect trade linkage
from all the economies in the system to each African economy following equation (6) We include
the heatwave (or own-effect) of equation (5) in this table so that it sums up to 100 We aggregate
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
10
directional Granger Causality measures have also been used characterize networks so that the
macroeconomic linkages among entities in the global economy can be described and understood
(Caraiani 2013 Hiemstra amp Jones 1994 Dahlhans amp Eichler 2003 Shojaie amp Michailidis 2010
Billio et al 2012) The main weakness of the Granger Causality approach is that it captures only
pairwise relations and may not be useful in answering important questions like ―what is the degree
of trade linkage between Africa the BRICS and the rest of the global economy1
Furthermore as noted by Diebold and Yilmaz (2016) these alternative methodologies generally
dwell exclusively on testing rather than measurement and estimation of macroeconomic linkages
which are the key issues in this paper This study therefore follows the network approach of
Diebold and Yilmaz (2009) based on its ability to transparently use the size and direction of shocks
to build both directional and non-directional trade linkage measures over a given forecast horizon
According to Ogbuabor et al (2016) studies using this approach have four common features
namely (i) they are generally based on connectedness measures distilled from forecast error
variance decompositions (FEVDs) of an approximating vector autoregressive (VAR) model (ii)
they measure the direction and strength of linkages among entities in the system (iii) they can
identify systemically important entities in the system and (iv) they can study the dynamic nature of
shock propagation among entities in the system In what follows the underlying VAR model for
this study and the construction of the generalized trade linkages measures (GTLMs) are presented
to guide the ensuing analysis
32 Model Specification The Vector Autoregression (VAR) Model
The broad objective of this study is to examine the propagation of trade shock between Africa the
BRICS economies and the rest of the global economy Let 119937120061 be the log of total trade for all the
countries selected for this study so that 119937119947120061 stands for the logged total trade of the j-th country in the
system with 119947 = 120783 120784 hellip 119925 and 119925 is the number of countries selected for the study (which is 20)
Following Diebold and Yilmaz (2009) the trade linkage measures for this study are based on the
normalized generalized forecast error variance decompositions (NGFEVDs) of an underlying 119953-th
1 Other techniques have also been used in the literature for the study of macroeconomic linkages Ogbuabor et al
(2018) provides an overview of such alternative methodologies such as the dynamic latent factor models of Kose et al (2008) and Canova et al (2007) the CoVaR approach of Adrian and Brunnermeier (2011) and the marginal expected shortfall (MES) approach of Acharya et al (2017) and Brownlees and Engle (2012)
11
order VAR model for the 119925 119961 120783 vector of endogenous variables 119937120061 The VAR(119953) model is specified
as follows
119937120061 = 120630119963 + 120509120051119937120061minus120051120057120051=120783 + 120634120061 (1)
where 120630 is 119925119961120783 vector of intercepts 120509120051 is 119925 119961 119925 coefficient matrix 120057 is the lag order and the
residuals 120634120050120061 ~ 119946119946119941 120782 120634119946119946 so that 120634120061 ~ 120782 120634 where 120634 is positive definite covariance matrix
The optimal VAR lag order selected by Schwarz Information Criterion (SIC) for this study is one
(see Appendix 5) Using the Woldlsquos Representation Theorem the model in equation (1) is
expressed as an infinite order vector moving average representation given by
119937120061 = 120634119957 + 120495120783120518119853minus120783 + 120495120784120518119853minus120784 + hellip = 120495120051120634120061minus120051infin120051=120782 (2)
where 120495120782 = 119920119925 120495120051 = 120509120051 120051 = 120783 120784 hellip and 119920119925 stands for an 119925 times 119925 identity matrix in which all
the principal diagonal elements are ones and all other elements are zeros
The network approach of Diebold and Yilmaz (2009) requires that after estimating the underlying
VAR model the forecast error variance decompositions (FEVDs) are then generated and used to
build linkage measures In this study the interest is in the shocks to the disturbances 120634119947120061 in the
country-specific equations Hence following Pesaran and Shin (1998) Diebold and Yilmaz (2016)
and Greenwood-Nimmo et al (2015) this study adopts the order-invariant generalized forecast error
variance decompositions (GFEVDs) defined as
119918119917119916119933119915 119937119946120061 120634119947120061 119919 = 119941119946119947119944119919
= 120648120634119947119947minus120783 119942119946
prime120495119945120506120634119942119947 sup2119919minus120783119945=120782
119942119946prime120495119945120506120634120495119945
prime 119942119946 119919minus120783119945=120782
(3)
where 119894 119895 = 1 hellip 119873 119867 = 1 2 hellip is the forecast horizon 119890119894 119890119895 is 119873 119909 1 selection vector whose i-th
element (j-th element) is unity with zeros elsewhere Θℎ is the coefficient matrix multiplying the h-
lagged shock vector in the infinite moving-average representation of the non-orthogonalized VAR
120634 is the covariance matrix of the shock vector in the non-orthogonalized VAR and 120648120634119947119947 is the j-th
diagonal element of 120634 (ie the standard deviation of 120576119895 ) We adopted a maximum forecast horizon
of 16 quarters in order to ensure that the long-run results are better captured It must be stressed that
the choice of GFEVDs for this study rather than the orthogonalized forecast error variance
decompositions (OFEVDs) of Diebold and Yilmaz (2009) is particularly based on the fact that the
12
OFEVDs depend on the reordering of the variables in the system such that once the order of
variables in the VAR is reshuffled a different outcome results
Diebold and Yilmaz (2014) explain that shocks are rarely orthogonal in the GFEVD environment
so that sums of forecast error variance contributions are not necessarily unity that is row sums of
the generalized variance decomposition matrix 119863119892119867 are not necessarily unity This renders the
interpretation of the GFEVDs complicated Thus to restore a percentage interpretation of the
GFEVDs this study follows Diebold and Yilmaz (2014) to define the normalized GFEVDs
(NGFEVDs) given by2
119863 119892 = 119889 119894119895119892 where 119889 119894119895
119892=
119889119894119895119892
119889119894119895119892119873
119895=1
119889119894119895119892
= 119918119917119916119933119915 119937119946120061 120634119947120061 119919 (4)
By construction 119889 119894119895119892119873
119895=1 = 1 and 119889 119894119895119892119873
119894 119895=1 = 119873 so that the total sum of the generalized forecast
error variance share of each variable is normalized to 100
33 Construction of the Generalized Trade Linkage Measures (GTLMs)
The various trade linkage measures that are relevant for the ensuing analysis are defined in this sub-
section The intuition behind this framework is quite simple Variance decomposition permits the
splitting of the forecast error variances of each variable in the VAR system into parts attributable to
the various system shocks By so doing it becomes easy to answer the question What fraction of
the h-step-ahead error variance in forecasting 119937120783119957 is due to shocks to 119937120783119957 Shocks to 119937120784119957 Similarly
what fraction of the h-step-ahead error variance in forecasting 119937120784119957 is due to shocks to 119937120783119957 Shocks
to 119937120784119957 And in general what fraction of the h-step-ahead error variance in forecasting 119937119947119957 is due to
shocks to 119937119946119957 119946 = 120783 120784 hellip 119925 Thus the approach marries VAR variance decomposition theory and
network topology theory by recognizing that variance decompositions of VARs form networks and
also characterizing linkages in those variance decomposition networks This in turn characterizes
trade linkages of the variables in our VAR system This is the intuition behind this framework
which we now exploit in the ensuing analysis Diebold and Yilmaz (2015) authoritatively document
this framework and its relation to network theory
2 In what follows and without loss of generality the superscript 119867 is dropped whenever it is not needed for clarity so
that 119863119892119867 and 119889119894119895119892119867
are simply written as 119863119892 and 119889119894119895119892
respectively
13
Table 1 Trade Linkage Schematic
Variables 119937120783 119937120784 ⋯ 119937119925 From Others
119937120783
⋯ 1198891119895
119873
119895=1 119895 ne 1
119937120784
⋯ 1198892119895
119873
119895=1 119895 ne 2
⋮ ⋮ ⋮ ⋱
⋮ ⋮
119937119925
⋯ 119889119873119895
119873
119895=1 119895 ne 119873
To Others
1198891198941
119873
119894=1
119894 ne 1
1198891198942
119873
119894=1
119894 ne 2
⋯ 119889119894119873
119873
119894=1
119894 ne N
1
119873 119889119894119895
119873
119894 119895=1 119894 ne 119895
Source Adapted from Diebold and Yilmaz (2014) and Ogbuabor et al (2016) Note For
simplicity each time series variable in this table 119937119947119957 is written as 119937119947 119947 = 120783 120784 hellip 119925
To construct the GTLMs for this study let us denote the H-step ahead NGFEVDs for the 119925 119961 120783
vector of endogenous variables 119937120061 obtained from equation (4) by 119941119946119947 By cross-tabulating 119941119946119947 the
trade linkage table shown in Table 1 is formed The sum of each row in Table 1 is normalized to
100 in line with equation (4) This table is now used to define the various GTLMs and their
relationships The diagonal entries in Table 1 measure own variance shares (or own-effect) while
the off-diagonal entries measure variance shares arising from shocks to other variables in the
system and are therefore referred to as pairwise directional linkage Accordingly the own-effect
(119919119947) also known as the heatwave is defined as
119919119947 = 119941119947119947 (5)
The total cross-variable variance share (119917119947) captures the spillovers from all other variables to 119937119947120061 as
fractions of the H-step-ahead error variance in the forecasts of 119937119947120061 resulting from 119937119946120061 where
119946 = 120783 120784 hellip 119925 and 119946 ne 119947 This measures the total directional linkage from other variables
(countries) in the system (ie the from-effect) to Zj120009 This means that the from-effect can be used to
capture the role each individual economy in the system plays in a given African economy and it is
14
computed in this study by aggregating the spillovers from all the economies in the system to a
given African economy across all horizons Hence the economy contributing the highest of such
aggregate spillover is deemed to play a dominant role in the particular African economy This study
therefore defines 119917119947 as
119917119947 = 119941119947119946119925119946=120783119946ne119947 (6)
By construction 119919119947 + 119917119947 = 120783 forall 119947
This study also defines the total spillover or total contributions of 119937119947120061 to all other variables (denoted
by 119931119947) as
119931119947 = 119941119946119947119925119946=120783119946ne119947 (7)
By construction 119931119947 measures the total directional linkage from 119937119947120061 to other variables in the system
(ie the to-effect) In other words the to-effect measures the directional linkage from a given
economy (for instance a given African economy) to other economies in the system thereby
showing the impact or influence of that particular African economy on other economies in the VAR
system The net directional linkage (or simply net-effect) of 119937119947120061 is therefore defined as
119925119947 = 119931119947 minus 119917119947 (8)
Since there are N economies in the VAR system it also follows that there are 2N total directional
trade linkages N capturing the total shocks transmitted to others (ie spillover to) and N capturing
the total shocks received from others (ie spillover from) These measures are aptly reflect the
bilateral trade patterns between the African economies and other economies in the system since
they mirror the total exports and total imports for each of the N economies in the system and will
therefore facilitate deeper understanding of the inter-linkages among the African economies We
shall utilize the net-effects to establish the net transmittersreceivers of shocks in the system over
time By construction 119925119947119925119947=120783 = 120782
The most aggregative (non-directional) trade linkage measure in this study which will be used to
evaluate the degree of trade linkage Africa the BRICS and the rest of the global economy is known
as the total trade linkage index (119931119931119923119920) or total-effects and it is defined as
119931119931119923119920 = 120783
119925 119917119947
119925119947=120783 =
120783
119925 119931119947
119925119947=120783 (9)
15
This measure captures the grand total of the off-diagonal elements in Table 1 that is the sum of the
―From Others column or ―To Others row There is only one total trade linkage measure which is
analogous to total global imports or total global exports since the two are identical
An important objective of this study is to determine which of the countries among the BRICS and
in the rest of the global economy exert the most dominant trade influence on African economies
and therefore have the potential to spread trade shocks to Africa To achieve this objective this
study defines two indices dependence and influence indices These indices are necessary to
determine the dependence of the j-th variable (or j-th economy) on external shocks and the
influence of the j-th variable (or j-th economy) on the system as a whole The dependence index is
defined as
119926119947119919 =
119917119947
119919119947+ 119917119947 forall 119947 = 120783 120784 hellip 119925 (10)
where 120782 le 119926119947119919 le 120783 This index expresses the relative importance of external shocks for the j-th
economy in the VAR system such that if 119926119947119919 rarr 120783 then conditions in the j-th economy is open
deeply interlinked and sensitive to external conditions but if 119926119947119919 rarr 120782 then the j-th economy is less
sensitive to external shocks
Similarly the influence index is expressed as
119920119947119919 =
119925119947
119931119947+119917119947 forall 119947 = 120783 120784 hellip 119925 (11)
where minus120783 le 119920119947119919 le 120783 For a given horizon H the j-th economy is a net receiver of trade shocks if
minus120783 le 119920119947119919 lt 0 that is if the index has a negative value a net transmitter of trade shocks if 120782 lt
119920119947119919 le 120783 that is if the index takes a positive value and neither a net receiver or transmitter of trade
shocks if IjH = 0 Thus the influence index measures the extent to which the j-th economy in the
system influences or is influenced by external shocks Overall the coordinate pair (119926119947119919 119920119947
119919) in the
dependence-influence space provides a good representation of the j-th economylsquos role in global real
activities A priori expectation is that African economies which are relatively small open
economies would be located close to the point (1-1) while overwhelmingly open but highly
influential and dominant economies like the USA China and Japan would be located close to the
point (11)
4 Results and Discussions
16
We started this empirical analysis by examining the time series properties of the data The Phillips-
Perron unit root tests showed that all the series are I(1) for all the countries (Please see Appendix 3
which reports for total trade of all the countries To conserve space we do not report separately for
exports and imports since the patterns are similar that is all the series are I(1)) However the test
for long-run or equilibrium relationship using the Johansen System Cointegration test showed that
both the Trace and Maximum Eigenvalue statistics returned full rank indicating that the data are
not cointegrated (Please see Appendix 4 which reports the Trace test in Panel 1 and the Max-
Eigenvalue test in Panel 2) Therefore a VAR in first differences was estimated rather than a vector
error correction model (VECM) Thereafter the GTLMs were computed for all forecasting
horizons 119867 = 12 hellip 16 The average values of the GTLMs over all the horizons are reported
since Diebold and Yilmaz (2016) had shown that the measures follow similar patterns regardless of
the choice of window lengths and forecast horizons As earlier stated we set the maximum forecast
horizon at 16 quarters in order to capture the long-run results better As a robustness check we also
estimated the underlying model and computed the GTLMs separately for exports and imports As
shall soon see the results follow the same qualitative pattern
41 Measuring the degree of trade linkage between Africa the BRICS and the rest of the
global economy
The first specific objective of this study is to measure the degree of trade linkage between African
economies the BRICS and the rest of the global economy In addition we examine here how this
index changed from the short-run (ie from horizon 1) through the long-run (ie until horizon 16)
To achieve this objective we estimated the underlying VAR model of equation (1) and used the
NGFEVDs distilled from this estimation based on equation (4) to compute the Total Trade Linkage
Index (TTLI) based on equation (9) We report the plots of this index in Figure 1 based on the total
trade data while the plots based on the exports and imports data are reported in Appendix 5 This
figure shows how the most aggregated generalized trade linkage measure in this study evolved from
the short-run through the long-run We find that the index rose smoothly from 84 at horizons 1
and 2 to 90 at horizons 15 and 16 In other words the total trade linkage index is higher in the
long-run than in the short-run indicating African economies become more interlinked in the long-
run as the business cycles become more synchronized This finding is consistent with the trend of
global macroeconomic interlinkage reported by Greenwood-Nimmo et al (2015) Diebold and
Yilmaz (2016) and Ogbuabor et al (2018) that the ongoing globalization process is engendering
17
more significant comovement in industrial production fluctuations Furthermore we find that the
TTLI recorded an average value of 87 which shows that the trade linkage between Africa the
BRICS and the rest of the global economy is quite substantial
As shown in Appendix 5 the patterns in this index remained qualitatively unchanged when the
underlying model is estimated using the exports and imports data respectively However a closer
look at panels 1 and 2 of this Appendix reveals that the index recorded higher values across all the
horizons for exports than for imports This aptly captures the pattern of trade between Africa and
the rest of the world which shows that Africalsquos export trade in dollar terms at least in the last
decade is higher than its import trade (see table 1 in appendix 1 of African Trade Statistics
Yearbook 2017) This finding clearly indicates that Africa may be vulnerable to trade shocks
which underlines the motivation behind this study
Figure 1 Total Trade Linkage Index
Source Authors Notes The Total trade linkage Index reported here is the most aggregated non-
directional trade linkage index computed for all horizons following equation (9) Notice that the index rises
smoothly from the short-run (ie horizon 1) towards the long-run (ie horizon 16) indicating that African
economies become more interlinked in the long-run
42 Determining the BRICS countries and other countries in the rest of the world that are
dominating Africarsquos trade and therefore have the potential to spread trade shocks to Africa
An important objective of this study is to determine the BRICS member countries and other
countries in the rest of the world that are dominating Africalsquos trade and therefore have the potential
to spread trade shocks to Africa To achieve this objective we report the from-effect trade linkage
from all the economies in the system to each African economy following equation (6) We include
the heatwave (or own-effect) of equation (5) in this table so that it sums up to 100 We aggregate
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
11
order VAR model for the 119925 119961 120783 vector of endogenous variables 119937120061 The VAR(119953) model is specified
as follows
119937120061 = 120630119963 + 120509120051119937120061minus120051120057120051=120783 + 120634120061 (1)
where 120630 is 119925119961120783 vector of intercepts 120509120051 is 119925 119961 119925 coefficient matrix 120057 is the lag order and the
residuals 120634120050120061 ~ 119946119946119941 120782 120634119946119946 so that 120634120061 ~ 120782 120634 where 120634 is positive definite covariance matrix
The optimal VAR lag order selected by Schwarz Information Criterion (SIC) for this study is one
(see Appendix 5) Using the Woldlsquos Representation Theorem the model in equation (1) is
expressed as an infinite order vector moving average representation given by
119937120061 = 120634119957 + 120495120783120518119853minus120783 + 120495120784120518119853minus120784 + hellip = 120495120051120634120061minus120051infin120051=120782 (2)
where 120495120782 = 119920119925 120495120051 = 120509120051 120051 = 120783 120784 hellip and 119920119925 stands for an 119925 times 119925 identity matrix in which all
the principal diagonal elements are ones and all other elements are zeros
The network approach of Diebold and Yilmaz (2009) requires that after estimating the underlying
VAR model the forecast error variance decompositions (FEVDs) are then generated and used to
build linkage measures In this study the interest is in the shocks to the disturbances 120634119947120061 in the
country-specific equations Hence following Pesaran and Shin (1998) Diebold and Yilmaz (2016)
and Greenwood-Nimmo et al (2015) this study adopts the order-invariant generalized forecast error
variance decompositions (GFEVDs) defined as
119918119917119916119933119915 119937119946120061 120634119947120061 119919 = 119941119946119947119944119919
= 120648120634119947119947minus120783 119942119946
prime120495119945120506120634119942119947 sup2119919minus120783119945=120782
119942119946prime120495119945120506120634120495119945
prime 119942119946 119919minus120783119945=120782
(3)
where 119894 119895 = 1 hellip 119873 119867 = 1 2 hellip is the forecast horizon 119890119894 119890119895 is 119873 119909 1 selection vector whose i-th
element (j-th element) is unity with zeros elsewhere Θℎ is the coefficient matrix multiplying the h-
lagged shock vector in the infinite moving-average representation of the non-orthogonalized VAR
120634 is the covariance matrix of the shock vector in the non-orthogonalized VAR and 120648120634119947119947 is the j-th
diagonal element of 120634 (ie the standard deviation of 120576119895 ) We adopted a maximum forecast horizon
of 16 quarters in order to ensure that the long-run results are better captured It must be stressed that
the choice of GFEVDs for this study rather than the orthogonalized forecast error variance
decompositions (OFEVDs) of Diebold and Yilmaz (2009) is particularly based on the fact that the
12
OFEVDs depend on the reordering of the variables in the system such that once the order of
variables in the VAR is reshuffled a different outcome results
Diebold and Yilmaz (2014) explain that shocks are rarely orthogonal in the GFEVD environment
so that sums of forecast error variance contributions are not necessarily unity that is row sums of
the generalized variance decomposition matrix 119863119892119867 are not necessarily unity This renders the
interpretation of the GFEVDs complicated Thus to restore a percentage interpretation of the
GFEVDs this study follows Diebold and Yilmaz (2014) to define the normalized GFEVDs
(NGFEVDs) given by2
119863 119892 = 119889 119894119895119892 where 119889 119894119895
119892=
119889119894119895119892
119889119894119895119892119873
119895=1
119889119894119895119892
= 119918119917119916119933119915 119937119946120061 120634119947120061 119919 (4)
By construction 119889 119894119895119892119873
119895=1 = 1 and 119889 119894119895119892119873
119894 119895=1 = 119873 so that the total sum of the generalized forecast
error variance share of each variable is normalized to 100
33 Construction of the Generalized Trade Linkage Measures (GTLMs)
The various trade linkage measures that are relevant for the ensuing analysis are defined in this sub-
section The intuition behind this framework is quite simple Variance decomposition permits the
splitting of the forecast error variances of each variable in the VAR system into parts attributable to
the various system shocks By so doing it becomes easy to answer the question What fraction of
the h-step-ahead error variance in forecasting 119937120783119957 is due to shocks to 119937120783119957 Shocks to 119937120784119957 Similarly
what fraction of the h-step-ahead error variance in forecasting 119937120784119957 is due to shocks to 119937120783119957 Shocks
to 119937120784119957 And in general what fraction of the h-step-ahead error variance in forecasting 119937119947119957 is due to
shocks to 119937119946119957 119946 = 120783 120784 hellip 119925 Thus the approach marries VAR variance decomposition theory and
network topology theory by recognizing that variance decompositions of VARs form networks and
also characterizing linkages in those variance decomposition networks This in turn characterizes
trade linkages of the variables in our VAR system This is the intuition behind this framework
which we now exploit in the ensuing analysis Diebold and Yilmaz (2015) authoritatively document
this framework and its relation to network theory
2 In what follows and without loss of generality the superscript 119867 is dropped whenever it is not needed for clarity so
that 119863119892119867 and 119889119894119895119892119867
are simply written as 119863119892 and 119889119894119895119892
respectively
13
Table 1 Trade Linkage Schematic
Variables 119937120783 119937120784 ⋯ 119937119925 From Others
119937120783
⋯ 1198891119895
119873
119895=1 119895 ne 1
119937120784
⋯ 1198892119895
119873
119895=1 119895 ne 2
⋮ ⋮ ⋮ ⋱
⋮ ⋮
119937119925
⋯ 119889119873119895
119873
119895=1 119895 ne 119873
To Others
1198891198941
119873
119894=1
119894 ne 1
1198891198942
119873
119894=1
119894 ne 2
⋯ 119889119894119873
119873
119894=1
119894 ne N
1
119873 119889119894119895
119873
119894 119895=1 119894 ne 119895
Source Adapted from Diebold and Yilmaz (2014) and Ogbuabor et al (2016) Note For
simplicity each time series variable in this table 119937119947119957 is written as 119937119947 119947 = 120783 120784 hellip 119925
To construct the GTLMs for this study let us denote the H-step ahead NGFEVDs for the 119925 119961 120783
vector of endogenous variables 119937120061 obtained from equation (4) by 119941119946119947 By cross-tabulating 119941119946119947 the
trade linkage table shown in Table 1 is formed The sum of each row in Table 1 is normalized to
100 in line with equation (4) This table is now used to define the various GTLMs and their
relationships The diagonal entries in Table 1 measure own variance shares (or own-effect) while
the off-diagonal entries measure variance shares arising from shocks to other variables in the
system and are therefore referred to as pairwise directional linkage Accordingly the own-effect
(119919119947) also known as the heatwave is defined as
119919119947 = 119941119947119947 (5)
The total cross-variable variance share (119917119947) captures the spillovers from all other variables to 119937119947120061 as
fractions of the H-step-ahead error variance in the forecasts of 119937119947120061 resulting from 119937119946120061 where
119946 = 120783 120784 hellip 119925 and 119946 ne 119947 This measures the total directional linkage from other variables
(countries) in the system (ie the from-effect) to Zj120009 This means that the from-effect can be used to
capture the role each individual economy in the system plays in a given African economy and it is
14
computed in this study by aggregating the spillovers from all the economies in the system to a
given African economy across all horizons Hence the economy contributing the highest of such
aggregate spillover is deemed to play a dominant role in the particular African economy This study
therefore defines 119917119947 as
119917119947 = 119941119947119946119925119946=120783119946ne119947 (6)
By construction 119919119947 + 119917119947 = 120783 forall 119947
This study also defines the total spillover or total contributions of 119937119947120061 to all other variables (denoted
by 119931119947) as
119931119947 = 119941119946119947119925119946=120783119946ne119947 (7)
By construction 119931119947 measures the total directional linkage from 119937119947120061 to other variables in the system
(ie the to-effect) In other words the to-effect measures the directional linkage from a given
economy (for instance a given African economy) to other economies in the system thereby
showing the impact or influence of that particular African economy on other economies in the VAR
system The net directional linkage (or simply net-effect) of 119937119947120061 is therefore defined as
119925119947 = 119931119947 minus 119917119947 (8)
Since there are N economies in the VAR system it also follows that there are 2N total directional
trade linkages N capturing the total shocks transmitted to others (ie spillover to) and N capturing
the total shocks received from others (ie spillover from) These measures are aptly reflect the
bilateral trade patterns between the African economies and other economies in the system since
they mirror the total exports and total imports for each of the N economies in the system and will
therefore facilitate deeper understanding of the inter-linkages among the African economies We
shall utilize the net-effects to establish the net transmittersreceivers of shocks in the system over
time By construction 119925119947119925119947=120783 = 120782
The most aggregative (non-directional) trade linkage measure in this study which will be used to
evaluate the degree of trade linkage Africa the BRICS and the rest of the global economy is known
as the total trade linkage index (119931119931119923119920) or total-effects and it is defined as
119931119931119923119920 = 120783
119925 119917119947
119925119947=120783 =
120783
119925 119931119947
119925119947=120783 (9)
15
This measure captures the grand total of the off-diagonal elements in Table 1 that is the sum of the
―From Others column or ―To Others row There is only one total trade linkage measure which is
analogous to total global imports or total global exports since the two are identical
An important objective of this study is to determine which of the countries among the BRICS and
in the rest of the global economy exert the most dominant trade influence on African economies
and therefore have the potential to spread trade shocks to Africa To achieve this objective this
study defines two indices dependence and influence indices These indices are necessary to
determine the dependence of the j-th variable (or j-th economy) on external shocks and the
influence of the j-th variable (or j-th economy) on the system as a whole The dependence index is
defined as
119926119947119919 =
119917119947
119919119947+ 119917119947 forall 119947 = 120783 120784 hellip 119925 (10)
where 120782 le 119926119947119919 le 120783 This index expresses the relative importance of external shocks for the j-th
economy in the VAR system such that if 119926119947119919 rarr 120783 then conditions in the j-th economy is open
deeply interlinked and sensitive to external conditions but if 119926119947119919 rarr 120782 then the j-th economy is less
sensitive to external shocks
Similarly the influence index is expressed as
119920119947119919 =
119925119947
119931119947+119917119947 forall 119947 = 120783 120784 hellip 119925 (11)
where minus120783 le 119920119947119919 le 120783 For a given horizon H the j-th economy is a net receiver of trade shocks if
minus120783 le 119920119947119919 lt 0 that is if the index has a negative value a net transmitter of trade shocks if 120782 lt
119920119947119919 le 120783 that is if the index takes a positive value and neither a net receiver or transmitter of trade
shocks if IjH = 0 Thus the influence index measures the extent to which the j-th economy in the
system influences or is influenced by external shocks Overall the coordinate pair (119926119947119919 119920119947
119919) in the
dependence-influence space provides a good representation of the j-th economylsquos role in global real
activities A priori expectation is that African economies which are relatively small open
economies would be located close to the point (1-1) while overwhelmingly open but highly
influential and dominant economies like the USA China and Japan would be located close to the
point (11)
4 Results and Discussions
16
We started this empirical analysis by examining the time series properties of the data The Phillips-
Perron unit root tests showed that all the series are I(1) for all the countries (Please see Appendix 3
which reports for total trade of all the countries To conserve space we do not report separately for
exports and imports since the patterns are similar that is all the series are I(1)) However the test
for long-run or equilibrium relationship using the Johansen System Cointegration test showed that
both the Trace and Maximum Eigenvalue statistics returned full rank indicating that the data are
not cointegrated (Please see Appendix 4 which reports the Trace test in Panel 1 and the Max-
Eigenvalue test in Panel 2) Therefore a VAR in first differences was estimated rather than a vector
error correction model (VECM) Thereafter the GTLMs were computed for all forecasting
horizons 119867 = 12 hellip 16 The average values of the GTLMs over all the horizons are reported
since Diebold and Yilmaz (2016) had shown that the measures follow similar patterns regardless of
the choice of window lengths and forecast horizons As earlier stated we set the maximum forecast
horizon at 16 quarters in order to capture the long-run results better As a robustness check we also
estimated the underlying model and computed the GTLMs separately for exports and imports As
shall soon see the results follow the same qualitative pattern
41 Measuring the degree of trade linkage between Africa the BRICS and the rest of the
global economy
The first specific objective of this study is to measure the degree of trade linkage between African
economies the BRICS and the rest of the global economy In addition we examine here how this
index changed from the short-run (ie from horizon 1) through the long-run (ie until horizon 16)
To achieve this objective we estimated the underlying VAR model of equation (1) and used the
NGFEVDs distilled from this estimation based on equation (4) to compute the Total Trade Linkage
Index (TTLI) based on equation (9) We report the plots of this index in Figure 1 based on the total
trade data while the plots based on the exports and imports data are reported in Appendix 5 This
figure shows how the most aggregated generalized trade linkage measure in this study evolved from
the short-run through the long-run We find that the index rose smoothly from 84 at horizons 1
and 2 to 90 at horizons 15 and 16 In other words the total trade linkage index is higher in the
long-run than in the short-run indicating African economies become more interlinked in the long-
run as the business cycles become more synchronized This finding is consistent with the trend of
global macroeconomic interlinkage reported by Greenwood-Nimmo et al (2015) Diebold and
Yilmaz (2016) and Ogbuabor et al (2018) that the ongoing globalization process is engendering
17
more significant comovement in industrial production fluctuations Furthermore we find that the
TTLI recorded an average value of 87 which shows that the trade linkage between Africa the
BRICS and the rest of the global economy is quite substantial
As shown in Appendix 5 the patterns in this index remained qualitatively unchanged when the
underlying model is estimated using the exports and imports data respectively However a closer
look at panels 1 and 2 of this Appendix reveals that the index recorded higher values across all the
horizons for exports than for imports This aptly captures the pattern of trade between Africa and
the rest of the world which shows that Africalsquos export trade in dollar terms at least in the last
decade is higher than its import trade (see table 1 in appendix 1 of African Trade Statistics
Yearbook 2017) This finding clearly indicates that Africa may be vulnerable to trade shocks
which underlines the motivation behind this study
Figure 1 Total Trade Linkage Index
Source Authors Notes The Total trade linkage Index reported here is the most aggregated non-
directional trade linkage index computed for all horizons following equation (9) Notice that the index rises
smoothly from the short-run (ie horizon 1) towards the long-run (ie horizon 16) indicating that African
economies become more interlinked in the long-run
42 Determining the BRICS countries and other countries in the rest of the world that are
dominating Africarsquos trade and therefore have the potential to spread trade shocks to Africa
An important objective of this study is to determine the BRICS member countries and other
countries in the rest of the world that are dominating Africalsquos trade and therefore have the potential
to spread trade shocks to Africa To achieve this objective we report the from-effect trade linkage
from all the economies in the system to each African economy following equation (6) We include
the heatwave (or own-effect) of equation (5) in this table so that it sums up to 100 We aggregate
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
12
OFEVDs depend on the reordering of the variables in the system such that once the order of
variables in the VAR is reshuffled a different outcome results
Diebold and Yilmaz (2014) explain that shocks are rarely orthogonal in the GFEVD environment
so that sums of forecast error variance contributions are not necessarily unity that is row sums of
the generalized variance decomposition matrix 119863119892119867 are not necessarily unity This renders the
interpretation of the GFEVDs complicated Thus to restore a percentage interpretation of the
GFEVDs this study follows Diebold and Yilmaz (2014) to define the normalized GFEVDs
(NGFEVDs) given by2
119863 119892 = 119889 119894119895119892 where 119889 119894119895
119892=
119889119894119895119892
119889119894119895119892119873
119895=1
119889119894119895119892
= 119918119917119916119933119915 119937119946120061 120634119947120061 119919 (4)
By construction 119889 119894119895119892119873
119895=1 = 1 and 119889 119894119895119892119873
119894 119895=1 = 119873 so that the total sum of the generalized forecast
error variance share of each variable is normalized to 100
33 Construction of the Generalized Trade Linkage Measures (GTLMs)
The various trade linkage measures that are relevant for the ensuing analysis are defined in this sub-
section The intuition behind this framework is quite simple Variance decomposition permits the
splitting of the forecast error variances of each variable in the VAR system into parts attributable to
the various system shocks By so doing it becomes easy to answer the question What fraction of
the h-step-ahead error variance in forecasting 119937120783119957 is due to shocks to 119937120783119957 Shocks to 119937120784119957 Similarly
what fraction of the h-step-ahead error variance in forecasting 119937120784119957 is due to shocks to 119937120783119957 Shocks
to 119937120784119957 And in general what fraction of the h-step-ahead error variance in forecasting 119937119947119957 is due to
shocks to 119937119946119957 119946 = 120783 120784 hellip 119925 Thus the approach marries VAR variance decomposition theory and
network topology theory by recognizing that variance decompositions of VARs form networks and
also characterizing linkages in those variance decomposition networks This in turn characterizes
trade linkages of the variables in our VAR system This is the intuition behind this framework
which we now exploit in the ensuing analysis Diebold and Yilmaz (2015) authoritatively document
this framework and its relation to network theory
2 In what follows and without loss of generality the superscript 119867 is dropped whenever it is not needed for clarity so
that 119863119892119867 and 119889119894119895119892119867
are simply written as 119863119892 and 119889119894119895119892
respectively
13
Table 1 Trade Linkage Schematic
Variables 119937120783 119937120784 ⋯ 119937119925 From Others
119937120783
⋯ 1198891119895
119873
119895=1 119895 ne 1
119937120784
⋯ 1198892119895
119873
119895=1 119895 ne 2
⋮ ⋮ ⋮ ⋱
⋮ ⋮
119937119925
⋯ 119889119873119895
119873
119895=1 119895 ne 119873
To Others
1198891198941
119873
119894=1
119894 ne 1
1198891198942
119873
119894=1
119894 ne 2
⋯ 119889119894119873
119873
119894=1
119894 ne N
1
119873 119889119894119895
119873
119894 119895=1 119894 ne 119895
Source Adapted from Diebold and Yilmaz (2014) and Ogbuabor et al (2016) Note For
simplicity each time series variable in this table 119937119947119957 is written as 119937119947 119947 = 120783 120784 hellip 119925
To construct the GTLMs for this study let us denote the H-step ahead NGFEVDs for the 119925 119961 120783
vector of endogenous variables 119937120061 obtained from equation (4) by 119941119946119947 By cross-tabulating 119941119946119947 the
trade linkage table shown in Table 1 is formed The sum of each row in Table 1 is normalized to
100 in line with equation (4) This table is now used to define the various GTLMs and their
relationships The diagonal entries in Table 1 measure own variance shares (or own-effect) while
the off-diagonal entries measure variance shares arising from shocks to other variables in the
system and are therefore referred to as pairwise directional linkage Accordingly the own-effect
(119919119947) also known as the heatwave is defined as
119919119947 = 119941119947119947 (5)
The total cross-variable variance share (119917119947) captures the spillovers from all other variables to 119937119947120061 as
fractions of the H-step-ahead error variance in the forecasts of 119937119947120061 resulting from 119937119946120061 where
119946 = 120783 120784 hellip 119925 and 119946 ne 119947 This measures the total directional linkage from other variables
(countries) in the system (ie the from-effect) to Zj120009 This means that the from-effect can be used to
capture the role each individual economy in the system plays in a given African economy and it is
14
computed in this study by aggregating the spillovers from all the economies in the system to a
given African economy across all horizons Hence the economy contributing the highest of such
aggregate spillover is deemed to play a dominant role in the particular African economy This study
therefore defines 119917119947 as
119917119947 = 119941119947119946119925119946=120783119946ne119947 (6)
By construction 119919119947 + 119917119947 = 120783 forall 119947
This study also defines the total spillover or total contributions of 119937119947120061 to all other variables (denoted
by 119931119947) as
119931119947 = 119941119946119947119925119946=120783119946ne119947 (7)
By construction 119931119947 measures the total directional linkage from 119937119947120061 to other variables in the system
(ie the to-effect) In other words the to-effect measures the directional linkage from a given
economy (for instance a given African economy) to other economies in the system thereby
showing the impact or influence of that particular African economy on other economies in the VAR
system The net directional linkage (or simply net-effect) of 119937119947120061 is therefore defined as
119925119947 = 119931119947 minus 119917119947 (8)
Since there are N economies in the VAR system it also follows that there are 2N total directional
trade linkages N capturing the total shocks transmitted to others (ie spillover to) and N capturing
the total shocks received from others (ie spillover from) These measures are aptly reflect the
bilateral trade patterns between the African economies and other economies in the system since
they mirror the total exports and total imports for each of the N economies in the system and will
therefore facilitate deeper understanding of the inter-linkages among the African economies We
shall utilize the net-effects to establish the net transmittersreceivers of shocks in the system over
time By construction 119925119947119925119947=120783 = 120782
The most aggregative (non-directional) trade linkage measure in this study which will be used to
evaluate the degree of trade linkage Africa the BRICS and the rest of the global economy is known
as the total trade linkage index (119931119931119923119920) or total-effects and it is defined as
119931119931119923119920 = 120783
119925 119917119947
119925119947=120783 =
120783
119925 119931119947
119925119947=120783 (9)
15
This measure captures the grand total of the off-diagonal elements in Table 1 that is the sum of the
―From Others column or ―To Others row There is only one total trade linkage measure which is
analogous to total global imports or total global exports since the two are identical
An important objective of this study is to determine which of the countries among the BRICS and
in the rest of the global economy exert the most dominant trade influence on African economies
and therefore have the potential to spread trade shocks to Africa To achieve this objective this
study defines two indices dependence and influence indices These indices are necessary to
determine the dependence of the j-th variable (or j-th economy) on external shocks and the
influence of the j-th variable (or j-th economy) on the system as a whole The dependence index is
defined as
119926119947119919 =
119917119947
119919119947+ 119917119947 forall 119947 = 120783 120784 hellip 119925 (10)
where 120782 le 119926119947119919 le 120783 This index expresses the relative importance of external shocks for the j-th
economy in the VAR system such that if 119926119947119919 rarr 120783 then conditions in the j-th economy is open
deeply interlinked and sensitive to external conditions but if 119926119947119919 rarr 120782 then the j-th economy is less
sensitive to external shocks
Similarly the influence index is expressed as
119920119947119919 =
119925119947
119931119947+119917119947 forall 119947 = 120783 120784 hellip 119925 (11)
where minus120783 le 119920119947119919 le 120783 For a given horizon H the j-th economy is a net receiver of trade shocks if
minus120783 le 119920119947119919 lt 0 that is if the index has a negative value a net transmitter of trade shocks if 120782 lt
119920119947119919 le 120783 that is if the index takes a positive value and neither a net receiver or transmitter of trade
shocks if IjH = 0 Thus the influence index measures the extent to which the j-th economy in the
system influences or is influenced by external shocks Overall the coordinate pair (119926119947119919 119920119947
119919) in the
dependence-influence space provides a good representation of the j-th economylsquos role in global real
activities A priori expectation is that African economies which are relatively small open
economies would be located close to the point (1-1) while overwhelmingly open but highly
influential and dominant economies like the USA China and Japan would be located close to the
point (11)
4 Results and Discussions
16
We started this empirical analysis by examining the time series properties of the data The Phillips-
Perron unit root tests showed that all the series are I(1) for all the countries (Please see Appendix 3
which reports for total trade of all the countries To conserve space we do not report separately for
exports and imports since the patterns are similar that is all the series are I(1)) However the test
for long-run or equilibrium relationship using the Johansen System Cointegration test showed that
both the Trace and Maximum Eigenvalue statistics returned full rank indicating that the data are
not cointegrated (Please see Appendix 4 which reports the Trace test in Panel 1 and the Max-
Eigenvalue test in Panel 2) Therefore a VAR in first differences was estimated rather than a vector
error correction model (VECM) Thereafter the GTLMs were computed for all forecasting
horizons 119867 = 12 hellip 16 The average values of the GTLMs over all the horizons are reported
since Diebold and Yilmaz (2016) had shown that the measures follow similar patterns regardless of
the choice of window lengths and forecast horizons As earlier stated we set the maximum forecast
horizon at 16 quarters in order to capture the long-run results better As a robustness check we also
estimated the underlying model and computed the GTLMs separately for exports and imports As
shall soon see the results follow the same qualitative pattern
41 Measuring the degree of trade linkage between Africa the BRICS and the rest of the
global economy
The first specific objective of this study is to measure the degree of trade linkage between African
economies the BRICS and the rest of the global economy In addition we examine here how this
index changed from the short-run (ie from horizon 1) through the long-run (ie until horizon 16)
To achieve this objective we estimated the underlying VAR model of equation (1) and used the
NGFEVDs distilled from this estimation based on equation (4) to compute the Total Trade Linkage
Index (TTLI) based on equation (9) We report the plots of this index in Figure 1 based on the total
trade data while the plots based on the exports and imports data are reported in Appendix 5 This
figure shows how the most aggregated generalized trade linkage measure in this study evolved from
the short-run through the long-run We find that the index rose smoothly from 84 at horizons 1
and 2 to 90 at horizons 15 and 16 In other words the total trade linkage index is higher in the
long-run than in the short-run indicating African economies become more interlinked in the long-
run as the business cycles become more synchronized This finding is consistent with the trend of
global macroeconomic interlinkage reported by Greenwood-Nimmo et al (2015) Diebold and
Yilmaz (2016) and Ogbuabor et al (2018) that the ongoing globalization process is engendering
17
more significant comovement in industrial production fluctuations Furthermore we find that the
TTLI recorded an average value of 87 which shows that the trade linkage between Africa the
BRICS and the rest of the global economy is quite substantial
As shown in Appendix 5 the patterns in this index remained qualitatively unchanged when the
underlying model is estimated using the exports and imports data respectively However a closer
look at panels 1 and 2 of this Appendix reveals that the index recorded higher values across all the
horizons for exports than for imports This aptly captures the pattern of trade between Africa and
the rest of the world which shows that Africalsquos export trade in dollar terms at least in the last
decade is higher than its import trade (see table 1 in appendix 1 of African Trade Statistics
Yearbook 2017) This finding clearly indicates that Africa may be vulnerable to trade shocks
which underlines the motivation behind this study
Figure 1 Total Trade Linkage Index
Source Authors Notes The Total trade linkage Index reported here is the most aggregated non-
directional trade linkage index computed for all horizons following equation (9) Notice that the index rises
smoothly from the short-run (ie horizon 1) towards the long-run (ie horizon 16) indicating that African
economies become more interlinked in the long-run
42 Determining the BRICS countries and other countries in the rest of the world that are
dominating Africarsquos trade and therefore have the potential to spread trade shocks to Africa
An important objective of this study is to determine the BRICS member countries and other
countries in the rest of the world that are dominating Africalsquos trade and therefore have the potential
to spread trade shocks to Africa To achieve this objective we report the from-effect trade linkage
from all the economies in the system to each African economy following equation (6) We include
the heatwave (or own-effect) of equation (5) in this table so that it sums up to 100 We aggregate
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
13
Table 1 Trade Linkage Schematic
Variables 119937120783 119937120784 ⋯ 119937119925 From Others
119937120783
⋯ 1198891119895
119873
119895=1 119895 ne 1
119937120784
⋯ 1198892119895
119873
119895=1 119895 ne 2
⋮ ⋮ ⋮ ⋱
⋮ ⋮
119937119925
⋯ 119889119873119895
119873
119895=1 119895 ne 119873
To Others
1198891198941
119873
119894=1
119894 ne 1
1198891198942
119873
119894=1
119894 ne 2
⋯ 119889119894119873
119873
119894=1
119894 ne N
1
119873 119889119894119895
119873
119894 119895=1 119894 ne 119895
Source Adapted from Diebold and Yilmaz (2014) and Ogbuabor et al (2016) Note For
simplicity each time series variable in this table 119937119947119957 is written as 119937119947 119947 = 120783 120784 hellip 119925
To construct the GTLMs for this study let us denote the H-step ahead NGFEVDs for the 119925 119961 120783
vector of endogenous variables 119937120061 obtained from equation (4) by 119941119946119947 By cross-tabulating 119941119946119947 the
trade linkage table shown in Table 1 is formed The sum of each row in Table 1 is normalized to
100 in line with equation (4) This table is now used to define the various GTLMs and their
relationships The diagonal entries in Table 1 measure own variance shares (or own-effect) while
the off-diagonal entries measure variance shares arising from shocks to other variables in the
system and are therefore referred to as pairwise directional linkage Accordingly the own-effect
(119919119947) also known as the heatwave is defined as
119919119947 = 119941119947119947 (5)
The total cross-variable variance share (119917119947) captures the spillovers from all other variables to 119937119947120061 as
fractions of the H-step-ahead error variance in the forecasts of 119937119947120061 resulting from 119937119946120061 where
119946 = 120783 120784 hellip 119925 and 119946 ne 119947 This measures the total directional linkage from other variables
(countries) in the system (ie the from-effect) to Zj120009 This means that the from-effect can be used to
capture the role each individual economy in the system plays in a given African economy and it is
14
computed in this study by aggregating the spillovers from all the economies in the system to a
given African economy across all horizons Hence the economy contributing the highest of such
aggregate spillover is deemed to play a dominant role in the particular African economy This study
therefore defines 119917119947 as
119917119947 = 119941119947119946119925119946=120783119946ne119947 (6)
By construction 119919119947 + 119917119947 = 120783 forall 119947
This study also defines the total spillover or total contributions of 119937119947120061 to all other variables (denoted
by 119931119947) as
119931119947 = 119941119946119947119925119946=120783119946ne119947 (7)
By construction 119931119947 measures the total directional linkage from 119937119947120061 to other variables in the system
(ie the to-effect) In other words the to-effect measures the directional linkage from a given
economy (for instance a given African economy) to other economies in the system thereby
showing the impact or influence of that particular African economy on other economies in the VAR
system The net directional linkage (or simply net-effect) of 119937119947120061 is therefore defined as
119925119947 = 119931119947 minus 119917119947 (8)
Since there are N economies in the VAR system it also follows that there are 2N total directional
trade linkages N capturing the total shocks transmitted to others (ie spillover to) and N capturing
the total shocks received from others (ie spillover from) These measures are aptly reflect the
bilateral trade patterns between the African economies and other economies in the system since
they mirror the total exports and total imports for each of the N economies in the system and will
therefore facilitate deeper understanding of the inter-linkages among the African economies We
shall utilize the net-effects to establish the net transmittersreceivers of shocks in the system over
time By construction 119925119947119925119947=120783 = 120782
The most aggregative (non-directional) trade linkage measure in this study which will be used to
evaluate the degree of trade linkage Africa the BRICS and the rest of the global economy is known
as the total trade linkage index (119931119931119923119920) or total-effects and it is defined as
119931119931119923119920 = 120783
119925 119917119947
119925119947=120783 =
120783
119925 119931119947
119925119947=120783 (9)
15
This measure captures the grand total of the off-diagonal elements in Table 1 that is the sum of the
―From Others column or ―To Others row There is only one total trade linkage measure which is
analogous to total global imports or total global exports since the two are identical
An important objective of this study is to determine which of the countries among the BRICS and
in the rest of the global economy exert the most dominant trade influence on African economies
and therefore have the potential to spread trade shocks to Africa To achieve this objective this
study defines two indices dependence and influence indices These indices are necessary to
determine the dependence of the j-th variable (or j-th economy) on external shocks and the
influence of the j-th variable (or j-th economy) on the system as a whole The dependence index is
defined as
119926119947119919 =
119917119947
119919119947+ 119917119947 forall 119947 = 120783 120784 hellip 119925 (10)
where 120782 le 119926119947119919 le 120783 This index expresses the relative importance of external shocks for the j-th
economy in the VAR system such that if 119926119947119919 rarr 120783 then conditions in the j-th economy is open
deeply interlinked and sensitive to external conditions but if 119926119947119919 rarr 120782 then the j-th economy is less
sensitive to external shocks
Similarly the influence index is expressed as
119920119947119919 =
119925119947
119931119947+119917119947 forall 119947 = 120783 120784 hellip 119925 (11)
where minus120783 le 119920119947119919 le 120783 For a given horizon H the j-th economy is a net receiver of trade shocks if
minus120783 le 119920119947119919 lt 0 that is if the index has a negative value a net transmitter of trade shocks if 120782 lt
119920119947119919 le 120783 that is if the index takes a positive value and neither a net receiver or transmitter of trade
shocks if IjH = 0 Thus the influence index measures the extent to which the j-th economy in the
system influences or is influenced by external shocks Overall the coordinate pair (119926119947119919 119920119947
119919) in the
dependence-influence space provides a good representation of the j-th economylsquos role in global real
activities A priori expectation is that African economies which are relatively small open
economies would be located close to the point (1-1) while overwhelmingly open but highly
influential and dominant economies like the USA China and Japan would be located close to the
point (11)
4 Results and Discussions
16
We started this empirical analysis by examining the time series properties of the data The Phillips-
Perron unit root tests showed that all the series are I(1) for all the countries (Please see Appendix 3
which reports for total trade of all the countries To conserve space we do not report separately for
exports and imports since the patterns are similar that is all the series are I(1)) However the test
for long-run or equilibrium relationship using the Johansen System Cointegration test showed that
both the Trace and Maximum Eigenvalue statistics returned full rank indicating that the data are
not cointegrated (Please see Appendix 4 which reports the Trace test in Panel 1 and the Max-
Eigenvalue test in Panel 2) Therefore a VAR in first differences was estimated rather than a vector
error correction model (VECM) Thereafter the GTLMs were computed for all forecasting
horizons 119867 = 12 hellip 16 The average values of the GTLMs over all the horizons are reported
since Diebold and Yilmaz (2016) had shown that the measures follow similar patterns regardless of
the choice of window lengths and forecast horizons As earlier stated we set the maximum forecast
horizon at 16 quarters in order to capture the long-run results better As a robustness check we also
estimated the underlying model and computed the GTLMs separately for exports and imports As
shall soon see the results follow the same qualitative pattern
41 Measuring the degree of trade linkage between Africa the BRICS and the rest of the
global economy
The first specific objective of this study is to measure the degree of trade linkage between African
economies the BRICS and the rest of the global economy In addition we examine here how this
index changed from the short-run (ie from horizon 1) through the long-run (ie until horizon 16)
To achieve this objective we estimated the underlying VAR model of equation (1) and used the
NGFEVDs distilled from this estimation based on equation (4) to compute the Total Trade Linkage
Index (TTLI) based on equation (9) We report the plots of this index in Figure 1 based on the total
trade data while the plots based on the exports and imports data are reported in Appendix 5 This
figure shows how the most aggregated generalized trade linkage measure in this study evolved from
the short-run through the long-run We find that the index rose smoothly from 84 at horizons 1
and 2 to 90 at horizons 15 and 16 In other words the total trade linkage index is higher in the
long-run than in the short-run indicating African economies become more interlinked in the long-
run as the business cycles become more synchronized This finding is consistent with the trend of
global macroeconomic interlinkage reported by Greenwood-Nimmo et al (2015) Diebold and
Yilmaz (2016) and Ogbuabor et al (2018) that the ongoing globalization process is engendering
17
more significant comovement in industrial production fluctuations Furthermore we find that the
TTLI recorded an average value of 87 which shows that the trade linkage between Africa the
BRICS and the rest of the global economy is quite substantial
As shown in Appendix 5 the patterns in this index remained qualitatively unchanged when the
underlying model is estimated using the exports and imports data respectively However a closer
look at panels 1 and 2 of this Appendix reveals that the index recorded higher values across all the
horizons for exports than for imports This aptly captures the pattern of trade between Africa and
the rest of the world which shows that Africalsquos export trade in dollar terms at least in the last
decade is higher than its import trade (see table 1 in appendix 1 of African Trade Statistics
Yearbook 2017) This finding clearly indicates that Africa may be vulnerable to trade shocks
which underlines the motivation behind this study
Figure 1 Total Trade Linkage Index
Source Authors Notes The Total trade linkage Index reported here is the most aggregated non-
directional trade linkage index computed for all horizons following equation (9) Notice that the index rises
smoothly from the short-run (ie horizon 1) towards the long-run (ie horizon 16) indicating that African
economies become more interlinked in the long-run
42 Determining the BRICS countries and other countries in the rest of the world that are
dominating Africarsquos trade and therefore have the potential to spread trade shocks to Africa
An important objective of this study is to determine the BRICS member countries and other
countries in the rest of the world that are dominating Africalsquos trade and therefore have the potential
to spread trade shocks to Africa To achieve this objective we report the from-effect trade linkage
from all the economies in the system to each African economy following equation (6) We include
the heatwave (or own-effect) of equation (5) in this table so that it sums up to 100 We aggregate
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
14
computed in this study by aggregating the spillovers from all the economies in the system to a
given African economy across all horizons Hence the economy contributing the highest of such
aggregate spillover is deemed to play a dominant role in the particular African economy This study
therefore defines 119917119947 as
119917119947 = 119941119947119946119925119946=120783119946ne119947 (6)
By construction 119919119947 + 119917119947 = 120783 forall 119947
This study also defines the total spillover or total contributions of 119937119947120061 to all other variables (denoted
by 119931119947) as
119931119947 = 119941119946119947119925119946=120783119946ne119947 (7)
By construction 119931119947 measures the total directional linkage from 119937119947120061 to other variables in the system
(ie the to-effect) In other words the to-effect measures the directional linkage from a given
economy (for instance a given African economy) to other economies in the system thereby
showing the impact or influence of that particular African economy on other economies in the VAR
system The net directional linkage (or simply net-effect) of 119937119947120061 is therefore defined as
119925119947 = 119931119947 minus 119917119947 (8)
Since there are N economies in the VAR system it also follows that there are 2N total directional
trade linkages N capturing the total shocks transmitted to others (ie spillover to) and N capturing
the total shocks received from others (ie spillover from) These measures are aptly reflect the
bilateral trade patterns between the African economies and other economies in the system since
they mirror the total exports and total imports for each of the N economies in the system and will
therefore facilitate deeper understanding of the inter-linkages among the African economies We
shall utilize the net-effects to establish the net transmittersreceivers of shocks in the system over
time By construction 119925119947119925119947=120783 = 120782
The most aggregative (non-directional) trade linkage measure in this study which will be used to
evaluate the degree of trade linkage Africa the BRICS and the rest of the global economy is known
as the total trade linkage index (119931119931119923119920) or total-effects and it is defined as
119931119931119923119920 = 120783
119925 119917119947
119925119947=120783 =
120783
119925 119931119947
119925119947=120783 (9)
15
This measure captures the grand total of the off-diagonal elements in Table 1 that is the sum of the
―From Others column or ―To Others row There is only one total trade linkage measure which is
analogous to total global imports or total global exports since the two are identical
An important objective of this study is to determine which of the countries among the BRICS and
in the rest of the global economy exert the most dominant trade influence on African economies
and therefore have the potential to spread trade shocks to Africa To achieve this objective this
study defines two indices dependence and influence indices These indices are necessary to
determine the dependence of the j-th variable (or j-th economy) on external shocks and the
influence of the j-th variable (or j-th economy) on the system as a whole The dependence index is
defined as
119926119947119919 =
119917119947
119919119947+ 119917119947 forall 119947 = 120783 120784 hellip 119925 (10)
where 120782 le 119926119947119919 le 120783 This index expresses the relative importance of external shocks for the j-th
economy in the VAR system such that if 119926119947119919 rarr 120783 then conditions in the j-th economy is open
deeply interlinked and sensitive to external conditions but if 119926119947119919 rarr 120782 then the j-th economy is less
sensitive to external shocks
Similarly the influence index is expressed as
119920119947119919 =
119925119947
119931119947+119917119947 forall 119947 = 120783 120784 hellip 119925 (11)
where minus120783 le 119920119947119919 le 120783 For a given horizon H the j-th economy is a net receiver of trade shocks if
minus120783 le 119920119947119919 lt 0 that is if the index has a negative value a net transmitter of trade shocks if 120782 lt
119920119947119919 le 120783 that is if the index takes a positive value and neither a net receiver or transmitter of trade
shocks if IjH = 0 Thus the influence index measures the extent to which the j-th economy in the
system influences or is influenced by external shocks Overall the coordinate pair (119926119947119919 119920119947
119919) in the
dependence-influence space provides a good representation of the j-th economylsquos role in global real
activities A priori expectation is that African economies which are relatively small open
economies would be located close to the point (1-1) while overwhelmingly open but highly
influential and dominant economies like the USA China and Japan would be located close to the
point (11)
4 Results and Discussions
16
We started this empirical analysis by examining the time series properties of the data The Phillips-
Perron unit root tests showed that all the series are I(1) for all the countries (Please see Appendix 3
which reports for total trade of all the countries To conserve space we do not report separately for
exports and imports since the patterns are similar that is all the series are I(1)) However the test
for long-run or equilibrium relationship using the Johansen System Cointegration test showed that
both the Trace and Maximum Eigenvalue statistics returned full rank indicating that the data are
not cointegrated (Please see Appendix 4 which reports the Trace test in Panel 1 and the Max-
Eigenvalue test in Panel 2) Therefore a VAR in first differences was estimated rather than a vector
error correction model (VECM) Thereafter the GTLMs were computed for all forecasting
horizons 119867 = 12 hellip 16 The average values of the GTLMs over all the horizons are reported
since Diebold and Yilmaz (2016) had shown that the measures follow similar patterns regardless of
the choice of window lengths and forecast horizons As earlier stated we set the maximum forecast
horizon at 16 quarters in order to capture the long-run results better As a robustness check we also
estimated the underlying model and computed the GTLMs separately for exports and imports As
shall soon see the results follow the same qualitative pattern
41 Measuring the degree of trade linkage between Africa the BRICS and the rest of the
global economy
The first specific objective of this study is to measure the degree of trade linkage between African
economies the BRICS and the rest of the global economy In addition we examine here how this
index changed from the short-run (ie from horizon 1) through the long-run (ie until horizon 16)
To achieve this objective we estimated the underlying VAR model of equation (1) and used the
NGFEVDs distilled from this estimation based on equation (4) to compute the Total Trade Linkage
Index (TTLI) based on equation (9) We report the plots of this index in Figure 1 based on the total
trade data while the plots based on the exports and imports data are reported in Appendix 5 This
figure shows how the most aggregated generalized trade linkage measure in this study evolved from
the short-run through the long-run We find that the index rose smoothly from 84 at horizons 1
and 2 to 90 at horizons 15 and 16 In other words the total trade linkage index is higher in the
long-run than in the short-run indicating African economies become more interlinked in the long-
run as the business cycles become more synchronized This finding is consistent with the trend of
global macroeconomic interlinkage reported by Greenwood-Nimmo et al (2015) Diebold and
Yilmaz (2016) and Ogbuabor et al (2018) that the ongoing globalization process is engendering
17
more significant comovement in industrial production fluctuations Furthermore we find that the
TTLI recorded an average value of 87 which shows that the trade linkage between Africa the
BRICS and the rest of the global economy is quite substantial
As shown in Appendix 5 the patterns in this index remained qualitatively unchanged when the
underlying model is estimated using the exports and imports data respectively However a closer
look at panels 1 and 2 of this Appendix reveals that the index recorded higher values across all the
horizons for exports than for imports This aptly captures the pattern of trade between Africa and
the rest of the world which shows that Africalsquos export trade in dollar terms at least in the last
decade is higher than its import trade (see table 1 in appendix 1 of African Trade Statistics
Yearbook 2017) This finding clearly indicates that Africa may be vulnerable to trade shocks
which underlines the motivation behind this study
Figure 1 Total Trade Linkage Index
Source Authors Notes The Total trade linkage Index reported here is the most aggregated non-
directional trade linkage index computed for all horizons following equation (9) Notice that the index rises
smoothly from the short-run (ie horizon 1) towards the long-run (ie horizon 16) indicating that African
economies become more interlinked in the long-run
42 Determining the BRICS countries and other countries in the rest of the world that are
dominating Africarsquos trade and therefore have the potential to spread trade shocks to Africa
An important objective of this study is to determine the BRICS member countries and other
countries in the rest of the world that are dominating Africalsquos trade and therefore have the potential
to spread trade shocks to Africa To achieve this objective we report the from-effect trade linkage
from all the economies in the system to each African economy following equation (6) We include
the heatwave (or own-effect) of equation (5) in this table so that it sums up to 100 We aggregate
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
15
This measure captures the grand total of the off-diagonal elements in Table 1 that is the sum of the
―From Others column or ―To Others row There is only one total trade linkage measure which is
analogous to total global imports or total global exports since the two are identical
An important objective of this study is to determine which of the countries among the BRICS and
in the rest of the global economy exert the most dominant trade influence on African economies
and therefore have the potential to spread trade shocks to Africa To achieve this objective this
study defines two indices dependence and influence indices These indices are necessary to
determine the dependence of the j-th variable (or j-th economy) on external shocks and the
influence of the j-th variable (or j-th economy) on the system as a whole The dependence index is
defined as
119926119947119919 =
119917119947
119919119947+ 119917119947 forall 119947 = 120783 120784 hellip 119925 (10)
where 120782 le 119926119947119919 le 120783 This index expresses the relative importance of external shocks for the j-th
economy in the VAR system such that if 119926119947119919 rarr 120783 then conditions in the j-th economy is open
deeply interlinked and sensitive to external conditions but if 119926119947119919 rarr 120782 then the j-th economy is less
sensitive to external shocks
Similarly the influence index is expressed as
119920119947119919 =
119925119947
119931119947+119917119947 forall 119947 = 120783 120784 hellip 119925 (11)
where minus120783 le 119920119947119919 le 120783 For a given horizon H the j-th economy is a net receiver of trade shocks if
minus120783 le 119920119947119919 lt 0 that is if the index has a negative value a net transmitter of trade shocks if 120782 lt
119920119947119919 le 120783 that is if the index takes a positive value and neither a net receiver or transmitter of trade
shocks if IjH = 0 Thus the influence index measures the extent to which the j-th economy in the
system influences or is influenced by external shocks Overall the coordinate pair (119926119947119919 119920119947
119919) in the
dependence-influence space provides a good representation of the j-th economylsquos role in global real
activities A priori expectation is that African economies which are relatively small open
economies would be located close to the point (1-1) while overwhelmingly open but highly
influential and dominant economies like the USA China and Japan would be located close to the
point (11)
4 Results and Discussions
16
We started this empirical analysis by examining the time series properties of the data The Phillips-
Perron unit root tests showed that all the series are I(1) for all the countries (Please see Appendix 3
which reports for total trade of all the countries To conserve space we do not report separately for
exports and imports since the patterns are similar that is all the series are I(1)) However the test
for long-run or equilibrium relationship using the Johansen System Cointegration test showed that
both the Trace and Maximum Eigenvalue statistics returned full rank indicating that the data are
not cointegrated (Please see Appendix 4 which reports the Trace test in Panel 1 and the Max-
Eigenvalue test in Panel 2) Therefore a VAR in first differences was estimated rather than a vector
error correction model (VECM) Thereafter the GTLMs were computed for all forecasting
horizons 119867 = 12 hellip 16 The average values of the GTLMs over all the horizons are reported
since Diebold and Yilmaz (2016) had shown that the measures follow similar patterns regardless of
the choice of window lengths and forecast horizons As earlier stated we set the maximum forecast
horizon at 16 quarters in order to capture the long-run results better As a robustness check we also
estimated the underlying model and computed the GTLMs separately for exports and imports As
shall soon see the results follow the same qualitative pattern
41 Measuring the degree of trade linkage between Africa the BRICS and the rest of the
global economy
The first specific objective of this study is to measure the degree of trade linkage between African
economies the BRICS and the rest of the global economy In addition we examine here how this
index changed from the short-run (ie from horizon 1) through the long-run (ie until horizon 16)
To achieve this objective we estimated the underlying VAR model of equation (1) and used the
NGFEVDs distilled from this estimation based on equation (4) to compute the Total Trade Linkage
Index (TTLI) based on equation (9) We report the plots of this index in Figure 1 based on the total
trade data while the plots based on the exports and imports data are reported in Appendix 5 This
figure shows how the most aggregated generalized trade linkage measure in this study evolved from
the short-run through the long-run We find that the index rose smoothly from 84 at horizons 1
and 2 to 90 at horizons 15 and 16 In other words the total trade linkage index is higher in the
long-run than in the short-run indicating African economies become more interlinked in the long-
run as the business cycles become more synchronized This finding is consistent with the trend of
global macroeconomic interlinkage reported by Greenwood-Nimmo et al (2015) Diebold and
Yilmaz (2016) and Ogbuabor et al (2018) that the ongoing globalization process is engendering
17
more significant comovement in industrial production fluctuations Furthermore we find that the
TTLI recorded an average value of 87 which shows that the trade linkage between Africa the
BRICS and the rest of the global economy is quite substantial
As shown in Appendix 5 the patterns in this index remained qualitatively unchanged when the
underlying model is estimated using the exports and imports data respectively However a closer
look at panels 1 and 2 of this Appendix reveals that the index recorded higher values across all the
horizons for exports than for imports This aptly captures the pattern of trade between Africa and
the rest of the world which shows that Africalsquos export trade in dollar terms at least in the last
decade is higher than its import trade (see table 1 in appendix 1 of African Trade Statistics
Yearbook 2017) This finding clearly indicates that Africa may be vulnerable to trade shocks
which underlines the motivation behind this study
Figure 1 Total Trade Linkage Index
Source Authors Notes The Total trade linkage Index reported here is the most aggregated non-
directional trade linkage index computed for all horizons following equation (9) Notice that the index rises
smoothly from the short-run (ie horizon 1) towards the long-run (ie horizon 16) indicating that African
economies become more interlinked in the long-run
42 Determining the BRICS countries and other countries in the rest of the world that are
dominating Africarsquos trade and therefore have the potential to spread trade shocks to Africa
An important objective of this study is to determine the BRICS member countries and other
countries in the rest of the world that are dominating Africalsquos trade and therefore have the potential
to spread trade shocks to Africa To achieve this objective we report the from-effect trade linkage
from all the economies in the system to each African economy following equation (6) We include
the heatwave (or own-effect) of equation (5) in this table so that it sums up to 100 We aggregate
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
16
We started this empirical analysis by examining the time series properties of the data The Phillips-
Perron unit root tests showed that all the series are I(1) for all the countries (Please see Appendix 3
which reports for total trade of all the countries To conserve space we do not report separately for
exports and imports since the patterns are similar that is all the series are I(1)) However the test
for long-run or equilibrium relationship using the Johansen System Cointegration test showed that
both the Trace and Maximum Eigenvalue statistics returned full rank indicating that the data are
not cointegrated (Please see Appendix 4 which reports the Trace test in Panel 1 and the Max-
Eigenvalue test in Panel 2) Therefore a VAR in first differences was estimated rather than a vector
error correction model (VECM) Thereafter the GTLMs were computed for all forecasting
horizons 119867 = 12 hellip 16 The average values of the GTLMs over all the horizons are reported
since Diebold and Yilmaz (2016) had shown that the measures follow similar patterns regardless of
the choice of window lengths and forecast horizons As earlier stated we set the maximum forecast
horizon at 16 quarters in order to capture the long-run results better As a robustness check we also
estimated the underlying model and computed the GTLMs separately for exports and imports As
shall soon see the results follow the same qualitative pattern
41 Measuring the degree of trade linkage between Africa the BRICS and the rest of the
global economy
The first specific objective of this study is to measure the degree of trade linkage between African
economies the BRICS and the rest of the global economy In addition we examine here how this
index changed from the short-run (ie from horizon 1) through the long-run (ie until horizon 16)
To achieve this objective we estimated the underlying VAR model of equation (1) and used the
NGFEVDs distilled from this estimation based on equation (4) to compute the Total Trade Linkage
Index (TTLI) based on equation (9) We report the plots of this index in Figure 1 based on the total
trade data while the plots based on the exports and imports data are reported in Appendix 5 This
figure shows how the most aggregated generalized trade linkage measure in this study evolved from
the short-run through the long-run We find that the index rose smoothly from 84 at horizons 1
and 2 to 90 at horizons 15 and 16 In other words the total trade linkage index is higher in the
long-run than in the short-run indicating African economies become more interlinked in the long-
run as the business cycles become more synchronized This finding is consistent with the trend of
global macroeconomic interlinkage reported by Greenwood-Nimmo et al (2015) Diebold and
Yilmaz (2016) and Ogbuabor et al (2018) that the ongoing globalization process is engendering
17
more significant comovement in industrial production fluctuations Furthermore we find that the
TTLI recorded an average value of 87 which shows that the trade linkage between Africa the
BRICS and the rest of the global economy is quite substantial
As shown in Appendix 5 the patterns in this index remained qualitatively unchanged when the
underlying model is estimated using the exports and imports data respectively However a closer
look at panels 1 and 2 of this Appendix reveals that the index recorded higher values across all the
horizons for exports than for imports This aptly captures the pattern of trade between Africa and
the rest of the world which shows that Africalsquos export trade in dollar terms at least in the last
decade is higher than its import trade (see table 1 in appendix 1 of African Trade Statistics
Yearbook 2017) This finding clearly indicates that Africa may be vulnerable to trade shocks
which underlines the motivation behind this study
Figure 1 Total Trade Linkage Index
Source Authors Notes The Total trade linkage Index reported here is the most aggregated non-
directional trade linkage index computed for all horizons following equation (9) Notice that the index rises
smoothly from the short-run (ie horizon 1) towards the long-run (ie horizon 16) indicating that African
economies become more interlinked in the long-run
42 Determining the BRICS countries and other countries in the rest of the world that are
dominating Africarsquos trade and therefore have the potential to spread trade shocks to Africa
An important objective of this study is to determine the BRICS member countries and other
countries in the rest of the world that are dominating Africalsquos trade and therefore have the potential
to spread trade shocks to Africa To achieve this objective we report the from-effect trade linkage
from all the economies in the system to each African economy following equation (6) We include
the heatwave (or own-effect) of equation (5) in this table so that it sums up to 100 We aggregate
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
17
more significant comovement in industrial production fluctuations Furthermore we find that the
TTLI recorded an average value of 87 which shows that the trade linkage between Africa the
BRICS and the rest of the global economy is quite substantial
As shown in Appendix 5 the patterns in this index remained qualitatively unchanged when the
underlying model is estimated using the exports and imports data respectively However a closer
look at panels 1 and 2 of this Appendix reveals that the index recorded higher values across all the
horizons for exports than for imports This aptly captures the pattern of trade between Africa and
the rest of the world which shows that Africalsquos export trade in dollar terms at least in the last
decade is higher than its import trade (see table 1 in appendix 1 of African Trade Statistics
Yearbook 2017) This finding clearly indicates that Africa may be vulnerable to trade shocks
which underlines the motivation behind this study
Figure 1 Total Trade Linkage Index
Source Authors Notes The Total trade linkage Index reported here is the most aggregated non-
directional trade linkage index computed for all horizons following equation (9) Notice that the index rises
smoothly from the short-run (ie horizon 1) towards the long-run (ie horizon 16) indicating that African
economies become more interlinked in the long-run
42 Determining the BRICS countries and other countries in the rest of the world that are
dominating Africarsquos trade and therefore have the potential to spread trade shocks to Africa
An important objective of this study is to determine the BRICS member countries and other
countries in the rest of the world that are dominating Africalsquos trade and therefore have the potential
to spread trade shocks to Africa To achieve this objective we report the from-effect trade linkage
from all the economies in the system to each African economy following equation (6) We include
the heatwave (or own-effect) of equation (5) in this table so that it sums up to 100 We aggregate
0
20
40
60
80
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
18
the contributions from key regional trading blocs in order to reveal their influence on each African
economy The regional trading blocs accounted for here includes the BRICS (Brazil Russia India
China South Africa) the Asian Bloc (India China Indonesia Japan Malaysia Russia) Europe
(UK EU) and the Americas (Canada Brazil USA)
Table 2 The From-effect Linkage of African Economies based on Total Trade Data
Country South Africa
DR Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 71778 29440 61788 42154 74728 43927 18340 60843
Brazil 58032 19932 55790 44390 33250 42150 14919 67961
Russia 66697 31216 64970 49067 35979 32582 25549 72983
India 55713 60962 72008 64639 46561 32564 08157 46453
China 68649 57823 73921 90160 55913 54103 16375 75379
South Africa 80744 44883 80196 65987 54168 42197 06952 75993
Australia 27148 31751 25479 38758 25946 44026 26291 34948
Canada 65383 41541 57408 37902 73538 47009 14725 55939
EU 68368 15015 57468 38353 75445 42242 12341 63107
Indonesia 23779 16606 16213 15499 09009 26810 06512 23182
Japan 68181 31048 55894 40754 77014 55484 24794 61861
Malaysia 45584 42499 33993 16137 42698 30786 55850 29298
UK 80756 15008 76576 76385 96626 78466 69155 62834
DR Congo 18171 287238 16923 29821 07776 15582 103410 20364
Egypt 60237 30150 89778 75198 46710 27362 35750 64429
Ghana 33558 41383 52468 121599 42637 46661 27242 37347
Morocco 43059 10130 40407 33048 93997 68184 52822 45226
Nigeria 13664 95792 11543 67823 28771 186518 05394 16720
Tanzania 05952 80171 03841 15953 29832 65904 451662 08304
Tunisia 44544 17412 53335 36375 49400 17444 23759 76830
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 249093 214816 346886 314243 225871 203596 71952 338770
Total from Asia 328605 240154 317000 276255 267174 232328 137238 309157
Total from Europe 149124 30023 134044 114738 172072 120708 81496 125940
Total from the Americas 195193 90913 174986 124446 181516 133086 47984 184742
Source Authors Notes This table is a transpose of Table 1 It reports the from-effect of equation (6) for all
the African economies and includes the heatwave of equation (5) so that the total for each economy sums up
to 100 The table also aggregates the contributions from key regional trading blocs in the system to each
African economy In the case of South Africa the contribution from BRICS excludes South Africalsquos
heatwave Here BRICS includes Brazil Russia India China and South Africa Asia includes India China
Indonesia Japan Malaysia and Russia Europe includes UK and EU while the Americas include Canada
Brazil and USA
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
19
We find that among the BRICS China is the highest contributor to the NGFEVDs of the African
economies It contributed at least 5 to each African economy except for Tanzania India and
South Africa also made remarkable contributions India contributed 3 to Nigeria and at least 5
to all other African economies except for Tanzania while South Africa contributed at least 4 to
majority of the other African economies The roles of Brazil and Russia are only notable in South
Africa Egypt and Tunisia At this point we have established that among the BRICS China plays a
dominant role in Africalsquos trade As a trading bloc we find that the BRICS contribute at least 20
to each African economy except for Tanzania that showed 7 This shows that the BRICS play
dominant role in Africalsquos trade and therefore have the potential to spread trade shocks to it In the
rest of the global economy we find that UK USA Japan EU and Canada play important roles in
Africalsquos trade UK contributed at least 6 to all the African economies except for DR Congo
while each of USA Japan EU and Canada contributed at least 4 to majority of the African
economies Overall we find that China UK USA Japan EU and Canada dominate Africalsquos trade
and that the BRICS as a trading bloc also exert a dominant influence on Africa These findings are
consistent with the bulk of the established literature such as Greenwood-Nimmo et al (2015)
Diebold and Yilmaz (2016) Ogbuabor et al (2016) and Ogbuabor et al (2018) which shows that
China USA Japan EU and UK are important real activity shock transmitters and therefore
dominate global economic activities We stress that even though the results in Table 2 are based on
the total trade data they are nonetheless consistent with the results obtained by using the exports
and imports data in separate regressions (please see Appendix 6 for the details of these results)
43 Determining other trade blocs outside the BRICS that exert dominant influence on Africa
and therefore have the prospects of spreading trade shocks to it
Another specific objective of this study seeks to determine other trading blocs outside the BRICS
that exert dominant influence on Africa and therefore have the potential to spread trade shocks to it
To achieve this objective we consider the contributions from the regional trading blocs to the
respective African economies as shown in Table 2 We find that outside the BRICS the Asian
trading bloc exert the most dominant influence on Africa with China Japan and India accounting
for the bulk of the influence coming from this region to Africa The contribution of this region
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
20
ranged from 14 in Tanzania to 33 in South Africa The Asian bloc is followed by the Americas
whose influence ranged from 5 in Tanzania to 20 in South Africa with USA and Canada
accounting for the bulk of the influence emanating from this region We also find that the role of
the European trading bloc cannot be called negligible The European trading bloc comprises UK
and EU whose joint contribution to majority of the African economies is at least 12 Overall we
find that even though the Asian trading bloc is playing a dominant role in Africa the roles of the
Americas and Europe cannot be called unimportant The findings here are consistent with the ones
obtained by using the exports and imports data in separate regressions (please see Appendix 6)
44 Determining the African economies that are most susceptible to trade shocks originating
from the BRICS member countries and the rest of the global economy
Another important specific objective of this study seeks to determine the African economies that
are most vulnerable to trade shocks originating from the BRICS and the rest of the global economy
To achieve this objective we report the estimates of the dependence and influence indices
following equations (10) and (11) respectively To provide a robustness check on these estimates
we will also consider the results of the net-effect linkage of each African economy following
equation (8)
To begin consider the results of the dependence and influence indices as shown in Table 3 based on
the total trade data We find that except for South Africa and Egypt all the other African
economies in the system reported a negative influence index indicating that rather than being
influential they are all vulnerable or susceptible to trade shocks emanating from the influential
economies The notable influential economies in the system include China UK USA Japan
Canada EU and South Africa These economies have positive influence index of at least 5 and
thus dominate the system Among the BRICS China has the highest influence index of 16 and in
the rest of the global economy UK and USA recorded the highest influence index of 18 and 11
respectively We find that the African economies have dependence index ranging from 71 to
92 except for Tanzania which recorded 55 This indicates that the African economies are
considerably open to external influences particularly the trade shocks originating from the
dominant economies in the system The BRICS have dependence index ranging between 91 and
92 while the other countries in the rest of the global economy have dependence index ranging
between 87 and 92 showing that these economies are highly open economies These findings
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
21
are consistent with our a priori expectations They are also consistent with studies like Ogbuabor et
al (2016) which have established that African economies are predominantly small open economies
deeply interlinked with the rest of the world but systemically unimportant and vulnerable to
headwinds emanating from the dominant economies in the overall global economy Note that even
though the results in Table 3 are based on the total trade data they are quite consistent with the
results obtained by using the exports and imports data in separate regressions (please see Appendix
7)
Table 3 Dependence and Influence Indices Results based on Total Trade Data
Country Dependence Index Influence Index
USA 09225 01146
Brazil 09120 -00219
Russia 09112 00455
India 09078 -00084
China 09089 01571
South Africa 09193 01043
Australia 09204 -01804
Canada 09162 00942
EU 09242 00508
Indonesia 08702 -04052
Japan 09058 00983
Malaysia 08912 -00356
UK 08952 01750
DR Congo 07128 -01741
Egypt 09102 00250
Ghana 08784 -00938
Morocco 09060 -00343
Nigeria 08135 -01450
Tanzania 05483 -02800
Tunisia 09232 -01425 Source Authors Notes The dependence and influence indices were computed using equations (10) and
(11) respectively Please see Appendix 7 for the details of the results obtained by using the exports and
imports data in separate regressions which are consistent with the results reported in this Table 3
To see if the above findings are robust to the net-effect linkage of African economies let us
consider the net-effect results presented in Table 4 following equation (8) We find that while the
dominant economies have huge positive net-effects including USA China UK Japan Canada and
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
22
EU all the African economies have negative net-effects except for South Africa (which is also a
member of the BRICS) and Egypt (whose positive net-effect is quite marginal) Hence while the
aforementioned dominant economies are net transmitters of trade shocks majority of the African
economies are net receivers of trade shocks This is consistent with our earlier findings that African
economies are predominantly small open economies deeply interlinked with the rest of the world
but systemically unimportant and vulnerable to headwinds emanating from the dominant
economies These results still hold even when are subjected to robustness check by running
separate regressions using the exports and imports data (please see Appendix 8)
Table 4 Net-effect Linkage Results based on Total Trade Data
Country From-effect To-effect Net-effect
USA 922486 1167212 244726
Brazil 911992 874640 -37352
Russia 911225 1000613 89388
India 907825 898242 -09583
China 908903 1248026 339122
South Africa 919256 1136053 216797
Australia 920363 648580 -271782
Canada 916215 1112171 195956
EU 924230 1029565 105335
Indonesia 870173 369507 -500665
Japan 905765 1109993 204227
Malaysia 891174 833400 -57774
UK 895223 1279940 384716
DR Congo 712762 501596 -211166
Egypt 910222 960466 50244
Ghana 878401 774361 -104040
Morocco 906003 846578 -59425
Nigeria 813482 682922 -130560
Tanzania 548338 329206 -219132
Tunisia 923170 694139 -229031 Source Authors Notes The From-effects To-effects and Net-effects were computed following equations
(6) (7) and (8) respectively Please see Appendix 8 for the details of the results obtained by using the
exports and imports data in separate regressions which are consistent with the results reported in this Table
4
5 Conclusion and Policy Implications
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
23
This study examined the dynamics of trade shock propagation between Africa the BRICS and the
rest of the global economy using the network approach of Diebold and Yilmaz (2009) The paper
extended the empirical method by constructing generalized trade linkage measures at various levels
of aggregation The main findings are summarized as follows First we find that African economies
become more interlinked in the long-run as the business cycles become more synchronized and that
the trade linkage between Africa and the rest of the global economy is quite substantial with the
total trade linkage index having an average value of 87 Second we find that China USA UK
Japan EU and Canada dominate Africalsquos trade and therefore have the potential to spread output
shocks to it The results further indicate that apart from the BRICS other regional trade blocs (Asia
the Americas and Europe) play influential roles in Africalsquos trade Third we find that African
economies are predominantly open but vulnerable to global trade shocks especially those
originating from the aforementioned dominant sources These findings particularly provide support
for some earlier research findings which indicate that intra-Africalsquos trade has been relatively low as
a result of low level of structural complementarities among these economies Overall our findings
suggest that the roles of African economies in global economic activities can be systemically called
unimportant since these economies are predominantly net receivers of trade shocks rather than net
transmitters
The findings of this study have several policy implications because according to Greenwood-
Nimmo et al (2015) ―hellipglobalization makes it impossible for dominant economies to collapse in
isolation First policymakers in Africa are able to see that the stability of the continent depends
somewhat on the actions of the rest of the global economy which are generally outside the control
of the continent In other words the results of this study constitute an essential wake-up call to
policymakers in Africa to be mindful of the chances of adverse trade shocks emanating from the
aforementioned dominant sources particularly the Asian trade partners Policymakers and leaders
in Africa should therefore coordinate policies towards safeguarding the continent from future crisis
For example deliberate and well-coordinated policy efforts should target the diversification of
African economies as a form of insurance against future trade shocks To drive economic
diversification entrepreneurs and other investors should be encouraged through the provision of
basic infrastructure and adequate security for lives and property Second the results of this study
provide evidence to assist policymakers in Africa and the rest of the global economy in identifying
the likely sources of future trade shocks so that appropriate policy responses to such shocks can be
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
24
designed Such policies and strategies will in turn ensure that the Africalsquos common goal of shared
prosperity and enhanced living standards for all its citizens are achieved on a sustainable basis
Third the findings provide evidence that can assist policymakers across the globe in understanding
how measurement and evaluation of macroeconomic linkages can be used to improve risk
measurement and management public policy regulatory oversight and overall economic
integration This underlines the need for policymakers in Africa to view the results of this study as
part of the much needed early warning signals towards the evolution of well-coordinated policy
actions that can safeguard the continent from potential global trade shocks
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
25
References
Acharya V V Pedersen L H Philippon Tand Richardson M (2017) ―Measuring Systemic
Risk The Review of Financial Studies 30(1) 2ndash47 httpsdoiorg101093rfshhw088
Adrian T and Brunnermeier M K (2011) CoVaR National Bureau of Economic Research
Working Paper No 17454 httpwwwnberorgpapersw17454pdf
African Union African Trade Statistics Yearbook 2017
Billio M Getmansky M Lo AW and Pelizzon L (2012) Econometric Measures of
Connectedness and Systemic Risk in the Finance and Insurance Sectors Journal of
Financial Economics 104(3) 535-559 httpsdoiorg101016jjfineco201112010
Bollerslev T (1990) Modelling the Coherence in Short-run Nominal Exchange Rates A
Multivariate Generalized ARCH Model The Review of Economics and Statistics 72(3) 498
ndash 505 DOI 1023072109358
Brownlees C T Engle R (2012) Volatility Correlation and Tails for Systemic Risk
Measurement Available at httpsbfiuchicagoedusitesdefaultfilesresearchSSRN-
id1611229pdf accessed 15 March 2019
Ccedilakir M Y and Kabundi Alain (2011) ―Trade Shocks from BRIC to South Africa A Global
VAR Analysis University of Johannesburg Auckland Park Campus Johannesburg
South Africa
Canova F Ciccarelli M and Ortega E (2007) Similarities and Convergence in G-7 Cycles
Journal of Monetary Economics 54 (3) 850 ndash 878
Caraiani P (2013) Using Complex Networks to Characterize International Business Cycles
PLoS ONE 8(3) e58109 Doi101371journalpone0058109
Dahlhans R and Eichler M (2003) Causality and Graphical Models in Time Series Analysis In
P J Green N L Hjort S Richardson (eds) Highly Structured Stochastic Systems Oxford
Statistical Science Series 27 115 ndash 137
Dees S di Mauro F Pesaran MH and Smith LV (2007) ―Exploring the International
Linkages of the Euro Area a Global VAR Analysis Journal of Applied Econometrics J
Appl Econ 22 1ndash38 (2007) Published online in Wiley InterScience
Dees S Holly S Pesaran MH and Smith LV (2007) ldquoLong Run Macroeconomic Relations
in the Global Economy Working Paper Series No 750 May 2007 European Central
Bank EuroSystem
Diebold FX and Yilmaz K (2009) Measuring Financial Asset Return and Volatility Spillovers
with Application to Global Equity Markets The Economic Journal 119(534) 158-171
httpsdoiorg101111j1468-0297200802208x
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
26
Diebold F X and Yilmaz K (2014) On the Network Topology of Variance Decompositions
Measuring the Connectedness of Financial Firms Journal of Econometrics 182(1) 119-
134 httpsdoiorg101016jjeconom201404012
Diebold FX and Yilmaz K (2015) Financial and Macroeconomic Connectedness A Network
Approach to Measurement and Monitoring New York Oxford University Press ISBN
9780199338306
Diebold FX and Yilmaz K (2016) Measuring the Dynamics of Global Business Cycle
Connectedness In SJ Koopman and N Shephard (Eds) Unobserved Components and
Time Series Econometrics Essays in Honor of Andrew C Harvey Oxford Oxford
University Press 45 ndash 70 DOI 101093acprofoso97801996836660010001 ISBN-13
9780199683666
Engle R F (2009) Anticipating Correlations A New Paradigm for Risk Management
Princeton NJ Princeton University Press ISBN 9780691116419
Engle R F Ito T Lin W (1990) Meteor Showers or Heat Waves Heteroskedastic Intra-
daily Volatility in the Foreign Exchange Market Econometrica 58(3) 525 ndash 542 DOI
1023072938189
Engle R F and Kelly B (2012) Dynamic Equicorrelation Journal of Business and Economic
Statistics 30(2) 212 ndash 228 httpsdoiorg101080073500152011652048
Gray D F and Malone S W (2008) Macrofinancial Risk Analysis West Sussex England John
Wiley amp Sons ISBN9780470058312 DOI1010029781118467428
Greenwood-Nimmo M Nguyen V H and Shin Y (2015) Measuring the Connectedness of the
Global Economy Melbourne Institute Working Paper No 715 Melbourne Institute of
Applied Economic and Social Research The University of Melbourne Australia
httpdxdoiorg102139ssrn2586861
Hiemstra C and Jones J D (1994) Testing for Linear and Nonlinear Granger Causality in the
Stock Price-Volume Relation The Journal of Finance 49(5) 1639 ndash 1664
httpsdoiorg101111j1540-62611994tb04776x
Huidrom R Kose M A and Ohnsorge F L (2017) ―How Important are Spillovers from Major
Emerging Markets World Bank Group Policy Research Working Paper 8093
Kehoe P J Backus D K and Kydland F E (1995) International Business Cycles Theory vs
Evidence In T F Cooley (Ed) Frontiers of Business Cycle Research Princeton NJ
Princeton University Press
Kose MA Otrok C and Whiteman C H (2008) Understanding the Evolution of World
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
27
Business Cycles Journal of International Economics 75(1) 110 ndash 130
httpsdoiorg101016jjinteco200710002
Kose M A Prasad E S and Terrones M E (2003) How Does Globalization Affect the
Synchronization of Business Cycles American Economic Review 93(2) 57-62 DOI
101257000282803321946804
Kose M A and Riezman R (1999) ―Trade Shocks and Macroeconomic Fluctuations in
Africa CESifo Working Paper No 203 Center for Economic Studies and Ifo Institute
(CESifo) Munich ECONSTOR
Lubik Thomas and Teo Wing Leong (2005) ―Do world shocks drive domestic
business cycles Some evidence from structural estimation Working Paper No 522 The
Johns Hopkins University Department of Economics Baltimore MD ECONSTOR
Mantegna R N (1999) Hierarchical Structure in Financial Markets The European Physical
Journal B 11(1) 193 ndash 197 httpsdoiorg101007s100510050929
Ogbuabor J E Eigbiremolen G O Aneke G C and Manasseh C O (2018) Measuring the
Dynamics of APEC Output Connectedness Asian-Pacific Economic Literature 32(1) 29 ndash
43 doi 101111apel12218
Ogbuabor J E Orji A Aneke G C and Erdene-Urnukh O (2016) Measuring the Real and
Financial Connectedness of Selected African Economies with the Global Economy South
African Journal of Economics 84(3) 364 ndash 399 doi 101111saje12135
Pesaran M H and Shin Y (1998) Generalized Impulse Response Analysis in Linear
Multivariate Models Economics Letters 58(1) 17-29 httpsdoiorg101016S0165-
1765(97)00214-0
Pesaran M H Til Schuermann and Weiner SM (2004) ―Modeling Regional Interdependencies
Using a Global Error-Correcting Macroeconometric Model Journal of Business amp
Economic Statistics 2004 vol 22 129-162
Shojaie A and Michailidis G (2010) Discovering Graphical Granger Causality Using the
Truncating Lasso Penalty Bioinformatics 26(18) i517 ndash i523
doi101093bioinformaticsbtq377
Taylor S J (2007) Asset Price Dynamics Volatility and Prediction Princeton NJ Princeton
University Press ISBN 9780691134796
Tumminello M Aste T Di Matteo T and Mantegna R N (2005) A Tool for Filtering
Information in Complex Systems Proceedings of the National Academy of Sciences
102(30) 10421 ndash 10426 doi 101073pnas0500298102
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
28
Appendix 1 Descriptive statistics
Panel 1 Total trade
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 84011 81358 88127 67001 77455 80301 85688 74184 72914 84264
Median 84236 79800 90033 67500 83329 75169 86593 68642 71163 83149
Maximum 97207 94582 95355 101668 98959 96559 96925 102321 97017 96790
Minimum 68070 59661 74991 21419 52033 66311 71997 45609 41361 65780
Std Dev 09103 09923 06307 25939 15215 10771 07832 18510 18889 08659
Skewness -02557 -03265 -08339 -01564 -02113 01983 -03892 02254 -01552 -00788
Kurtosis 18436 20982 24076 16568 16486 13023 17853 16793 16301 18843
Jarque-Bera 71942 55790 140960 85592 90220 136781 93660 87637 88786 57136
Probability 00274 00615 00009 00138 00110 00011 00093 00125 00118 00575
Sum 9073191 8786664 9517664 7236112 8365173 8672560 9254352 8011865 7874676 9100564
Sum Sq Dev 886614 1053555 425632 7199027 2477109 1241421 656302 3666218 3817577 802300
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 87236 83778 83738 76289 82244 86661 78667 79968 86786 85707
Median 87044 84389 82930 73778 83385 86057 79551 77917 88201 86262
Maximum 95578 94754 98656 93881 95886 96149 101364 94617 95788 95813
Minimum 77730 59370 71394 59040 66530 73517 55142 63446 74210 70590
Std Dev 05861 09711 08697 11261 10013 07122 14826 10113 06881 07874
Skewness -01509 -07693 00520 00961 -00788 -03648 00245 00168 -05261 -05221
Kurtosis 16540 25582 15010 14196 14106 19515 16529 15297 18919 20504
Jarque-Bera 85624 115304 101596 114054 114789 73429 81769 97338 105077 89641
Probability 00138 00031 00062 00033 00032 00254 00168 00077 00052 00113
Sum 9421434 9048058 9043667 8239173 8882388 9359360 8495988 8636512 9372902 9256322
Sum Sq Dev 367539 1009027 809327 1356855 1072816 542746 2351912 1094302 506657 663408
Observations 108 108 108 108 108 108 108 108 108 108
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
29
Panel 2 Exports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 43295 41726 44997 33564 39203 39204 42884 37423 36712 41902
Median 43989 42513 46369 33563 40228 37215 43260 34893 36168 41063
Maximum 49267 47480 48047 50647 51078 48075 48591 52280 49347 48121
Minimum 35181 32658 38487 10999 28248 31282 35764 22685 20866 32299
Std Dev 03567 04627 02889 12951 07307 05704 03974 09473 09485 04404
Skewness -05126 -03390 -11139 -01593 01003 01444 -03788 02615 -01339 -00785
Kurtosis 24040 16695 28948 16237 17569 13333 18292 17687 15771 19101
Jarque-Bera 63277 100352 223840 89813 71350 128756 87513 80535 94336 54568
Probability 00423 00066 00000 00112 00282 00016 00126 00178 00089 00653
Sum 4675847 4506379 4859645 3624899 4233933 4234076 4631435 4041731 3964876 4525383
Sum Sq Dev 136178 229032 89333 1794823 571317 348181 168996 960184 962699 207557
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 42639 41893 42189 40448 42248 44177 39647 39949 43572 42808
Median 42484 42653 42154 38822 43018 44454 41144 39034 43987 42616
Maximum 47407 47004 48937 48585 47489 47584 52432 47544 47638 47884
Minimum 36895 30077 36326 33208 34602 38788 25067 31643 37575 35685
Std Dev 03435 04889 04158 05011 03967 02711 07820 04997 03222 03643
Skewness -00975 -08534 -00141 03327 -03128 -05993 -02009 -00408 -05455 -02672
Kurtosis 15123 25749 15233 15819 16434 21907 19619 15902 19928 19945
Jarque-Bera 101311 139237 98169 110416 100430 94126 55756 89743 99206 58349
Probability 00063 00009 00074 00040 00066 00090 00616 00113 00070 00541
Sum 4605043 4524443 4556463 4368366 4562831 4771111 4281869 4314453 4705765 4623307
Sum Sq Dev 126277 255735 184992 268627 168392 78665 654348 267198 111068 141975
Observations 108 108 108 108 108 108 108 108 108 108
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
30
Panel 3 Imports
AUS BRA CAN CHN CON EGY EU GHA IND INDO
Mean 40716 39632 43130 33437 38252 41097 42805 36761 36202 42363
Median 40293 38797 43425 33938 42632 38025 43333 33749 34995 42524
Maximum 48275 47789 47365 51059 47958 48484 48334 50553 48606 48670
Minimum 31854 27002 36458 10124 23431 34572 36232 22924 19763 33482
Std Dev 05623 05535 03511 13000 08121 05130 03861 09070 09426 04291
Skewness -01054 -02950 -05179 -01564 -04266 02822 -03998 01867 -01796 -00986
Kurtosis 16142 23579 19916 16996 17030 13341 17464 15966 16997 18396
Jarque-Bera 88414 34218 94030 80500 108455 139220 99495 94901 81891 62349
Probability 00120 01807 00091 00179 00044 00009 00069 00087 00167 00443
Sum 4397345 4280285 4658019 3611214 4131240 4438484 4622917 3970134 3909800 4575180
Sum Sq Dev 338298 327825 131914 1808338 705702 281581 159501 880160 950711 197015
Observations 108 108 108 108 108 108 108 108 108 108
JPN MAL MOR NIG RUS SAF TAN TUN UK USA
Mean 44596 41885 41548 35841 39996 42484 39020 40019 43214 42898
Median 44561 41759 40821 35106 40211 41620 38407 38883 44214 44011
Maximum 48381 47750 49720 46132 49229 48587 49650 47288 48210 47980
Minimum 40342 29293 35068 24090 30390 34645 29905 31802 36600 34622
Std Dev 02461 04854 04568 06666 06219 04457 07157 05136 03670 04301
Skewness -02384 -06748 01142 00167 00316 -02192 02042 00811 -05114 -06862
Kurtosis 19290 25254 14961 15951 14596 18055 14998 14947 18138 20792
Jarque-Bera 61848 92090 104121 88865 106955 72860 108786 103152 110391 122905
Probability 00454 00100 00055 00118 00048 00262 00043 00058 00040 00021
Sum 4816391 4523614 4487204 3870807 4319557 4588249 4214119 4322059 4667138 4633015
Sum Sq Dev 64784 252086 223300 475453 413775 212507 548102 282235 144087 197896
Observations 108 108 108 108 108 108 108 108 108 108
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
31
Appendix 2 Time series plots of the data
Panel 1 Total Trade
Panel 2 Exports
0
2
4
6
8
10
121
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Tota
l Tra
de
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
0
1
2
3
4
5
6
19
90
Q1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Exp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tanzania tunisia
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
32
Panel 3 Imports
Source Authors Notes These graphs plot the data over the full sample
0
1
2
3
4
5
61
99
0Q
1
19
91
Q1
19
92
Q1
19
93
Q1
19
94
Q1
19
95
Q1
19
96
Q1
19
97
Q1
19
98
Q1
19
99
Q1
20
00
Q1
20
01
Q1
20
02
Q1
20
03
Q1
20
04
Q1
20
05
Q1
20
06
Q1
20
07
Q1
20
08
Q1
20
09
Q1
20
10
Q1
20
11
Q1
20
12
Q1
20
13
Q1
20
14
Q1
20
15
Q1
20
16
Q1
Log
of
Imp
ort
s
Dateusa brazil russia india
china safrica australia canada
eu indonesia japan malaysia
uk drcongo egypt ghana
morocco nigeria tunisia tanzania
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
33
Appendix 3 Phillips-Perron Unit Root Test Results for Total Trade
Variables PP Test stat at
level Critical
value at 5
PP Test stat at
1st Diff
Critical value at 5
Order of Integration
Australia -2087649 -3452358 -5511474 -3452764 I(1)
Brazil -151073 -3452358 -4815699 -3452764 I(1)
Canada -1688886 -3452358 -4919794 -3452764 I(1)
China -0761713 -3452358 -4666041 -3452764 I(1)
DRCongo -2701397 -3452358 -4969701 -3452764 I(1)
Egypt -1110372 -3452358 -4587926 -3452764 I(1)
EU -2479035 -3452358 -4796753 -3452764 I(1)
Ghana -1789434 -3452358 -4749838 -3452764 I(1)
India -0176913 -3452358 -5580811 -3452764 I(1)
Indonesia -1940294 -3452358 -7723706 -3452764 I(1)
Japan -27304 -3452358 -4666113 -3452764 I(1)
Malaysia -1798537 -3452358 -5140808 -3452764 I(1)
Morocco -3427262 -3452358 -5251978 -3452764 I(1)
Nigeria -1885798 -3452358 -6573952 -3452764 I(1)
Russia -2427019 -3452358 -4584745 -3452764 I(1)
Safrica -2459347 -3452358 -5061753 -3452764 I(1)
Tanzania -2080362 -3452358 -4639207 -3452764 I(1)
Tunisia -2521464 -3452358 -4922059 -3452764 I(1)
UK -292419 -3452358 -4914527 -3452764 I(1)
USA -1015221 -3452358 -5002348 -3452764 I(1) Source Authors Notes This table reports for total trade data To conserve space we do not report for
exports and imports since the patterns are similar that is all the variables are I(1)
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
34
Appendix 4 Johansen system cointegration test results
Panel 1 Trace test
Hypothesized No of CE(s) Eigenvalue Trace Stat
005 Critical Value Prob
None 08119 20372650 7605409 00000
At most 1 07484 18618330 6950496 00000
At most 2 06943 17169350 6409570 00000
At most 3 06896 15925080 5945066 00000
At most 4 06813 14696730 5486505 00000
At most 5 06751 13496030 5038266 00000
At most 6 06619 12315570 4597583 00000
At most 7 06605 11176950 4172519 00000
At most 8 06570 10042670 3749076 00000
At most 9 06534 8919101 3220692 00000
At most 10 06439 7806611 2731889 00000
At most 11 06424 6722533 2282979 00001
At most 12 06365 5642802 1874701 00000
At most 13 06312 4580251 1505585 00000
At most 14 06183 3532980 1177082 00000
At most 15 05059 2521643 888038 00000
At most 16 04229 1781345 638761 00000
At most 17 03477 1204084 429153 00000
At most 18 03308 755525 258721 00000
At most 19 02723 333740 125180 00000
Panel 2 Max-Eigenvalue test
Hypothesized No of CE(s) Eigenvalue
Max-Eigen Stat
005 Critical Value Prob
None 08119 1754311 1262691 00000
At most 1 07484 1448984 1042928 00000
At most 2 06943 1244273 895584 00000
At most 3 06896 1228346 884120 00000
At most 4 06813 1200704 864224 00000
At most 5 06751 1180455 849650 00000
At most 6 06619 1138626 819543 00000
At most 7 06605 1134280 816415 00000
At most 8 06570 1123565 808703 00000
At most 9 06534 1112489 748375 00000
At most 10 06439 1084078 688121 00000
At most 11 06424 1079731 627522 00000
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
35
At most 12 06365 1062552 567052 00000
At most 13 06312 1047271 505999 00000
At most 14 06183 1011337 444972 00000
At most 15 05059 740298 383310 00000
At most 16 04229 577261 321183 00000
At most 17 03477 448558 258232 00001
At most 18 03308 421785 193870 00000
At most 19 02723 333740 125180 00000
Source Authors
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
36
Appendix 5 Plots of Total Trade Linkage Index (TTLI)
Panel 1 For Exports Data
Panel 2 For Imports Data
Source Authors Notes The graphs plot the Total Trade Linkage Index of equation (9) based on
the data for exports and imports
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
5)
Horizon
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Pe
rce
nta
ge (
)
Horizon
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
37
Appendix 6 The From-effect Linkage of African Economies
Panel 1 Based on Exports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tanzania Tunisia
USA 75340 31220 52535 24920 32451 80425 16142 58582
Brazil 70609 08813 89311 04341 118017 31421 34896 70440
Russia 77612 125327 110425 34878 37666 32531 20684 95616
India 37554 88615 28643 68745 14462 45156 27108 20011
China 72605 55905 65828 98491 44992 74646 19369 79720
South Africa 97203 36750 72784 35206 31078 59433 25869 68574
Australia 11614 44655 02851 08205 109096 02585 32873 15055
Canada 73007 73293 50845 17185 57174 54735 08255 60655
EU 84055 25762 66141 26268 48995 68512 12978 68320
Indonesia 19769 24314 01169 80320 18449 16285 14550 05139
Japan 85308 40028 55534 22141 62552 85835 16891 70244
Malaysia 56649 70935 37736 31557 26228 67378 19406 37956
UK 53469 16542 64573 110048 19054 107249 55380 45530
DR Congo 29744 215520 39112 71369 09432 26191 07574 32377
Egypt 43495 37493 140244 17884 31158 27479 80075 60913
Ghana 03332 39604 21476 261490 10461 04486 12166 24795
Morocco 26742 12945 28298 04647 232156 27901 47621 48632
Nigeria 30411 05806 22706 64205 14887 130388 42022 25243
Tanzania 03455 10693 03545 11170 24621 31812 501363 06036
Tunisia 48028 35778 46245 06928 57070 25551 04778 106163
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 258380 315410 366990 241662 246216 243188 127926 334361
Total from Asia 349497 405125 299335 336134 204349 321832 118008 308685
Total from Europe 137525 42304 130714 136316 68050 175761 68358 113850
Total from the Americas 218956 113326 192691 46445 207642 166581 59294 189677
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
38
Panel 2 Based on Imports Data
Country South Africa DR
Congo Egypt Ghana Morocco Nigeria Tunisia Tanzania
USA 84417 27346 64937 08671 58633 45316 50364 08786
Brazil 62986 17990 52082 67154 42216 23105 79339 53891
Russia 80149 22871 85737 59520 44582 03436 72180 08594
India 24108 88781 32734 41954 80329 59152 35292 31134
China 57369 67788 51214 97239 36961 03302 48212 15927
South Africa 96491 23300 70515 22174 37526 16513 58827 15241
Australia 29237 67747 29516 62979 69457 03181 43946 24635
Canada 68897 31125 55227 47717 82931 07959 61205 05466
EU 71382 41920 40932 09740 82114 44444 71835 16233
Indonesia 45443 03361 37592 03145 24308 04855 20374 05870
Japan 67811 27701 61483 42867 46307 37472 64150 53856
Malaysia 57080 08193 56191 27331 61632 42365 26281 25684
UK 79705 08899 57763 09308 18636 75911 39259 30355
DR Congo 09940 314128 04962 51581 29301 07936 32248 94814
Egypt 33338 15688 113752 57598 34201 18514 40609 16733
Ghana 14299 25910 45405 243033 14637 60390 36649 10144
Morocco 29828 91510 32439 50298 136523 28312 64231 34121
Nigeria 16423 06604 31183 17516 07315 390801 09295 68504
Tunisia 40701 34480 50979 63742 67218 35759 117458 38775
Tanzania 30396 74657 25357 16435 25175 91275 28247 441238
Total From-effect 100 100 100 100 100 100 100 100
Total from Regional Trading Blocs
Total from BRICS 224613 220729 292282 288040 241613 105508 293849 124787
Total from Asia 331961 218695 324952 272055 294119 150583 266489 141065
Total from Europe 151087 50819 98695 19048 100749 120355 111094 46588
Total from the Americas 216300 76461 172245 123541 183780 76381 190908 68143 Source Authors Notes The notes in Table 2 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
39
Appendix 7 Dependence and Influence Indices Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country Dependence Index Influence Index Dependence Index Influence Index
USA 09188 00959 09065 00856
Brazil 08344 01108 08742 00260
Russia 08388 02094 08924 00929
India 09000 -01139 08413 -00254
China 09102 01438 08912 00402
South Africa 09028 00416 09035 00097
Australia 07105 -02842 08673 -00439
Canada 08991 00946 08986 01122
EU 09121 01077 08971 00713
Indonesia 07591 -03108 08369 -02587
Japan 08998 01361 08803 00883
Malaysia 09000 00528 08364 00472
UK 08870 00649 08808 -00493
DR Congo 07845 -00328 06859 -00974
Egypt 08598 -00230 08862 -01441
Ghana 07385 -03746 07570 -01304
Morocco 07678 -00723 08635 01011
Nigeria 08696 -01538 06092 -02776
Tanzania 04986 -03598 05588 -00770
Tunisia 08938 -01993 08825 -00518 Source Authors Notes The notes in Table 3 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply
40
Appendix 8 Net-effect Linkage Results
Panel 1 Based on Exports Data Panel 2 Based on Imports Data
Country From-effect To-effect Net-effect From-effect To-effect Net-effect
USA 918838 1116203 197365 906516 1080837 174321
Brazil 834405 1048823 214418 874223 922305 48081
Russia 838838 1301518 462681 892421 1076102 183681
India 900031 715701 -184330 841338 842608 01270
China 910213 1216863 306650 891159 966746 75587
South Africa 902797 982282 79485 903509 923706 20197
Australia 710525 395694 -314831 867260 798447 -68813
Canada 899143 1089583 190440 898639 1131561 232923
EU 912144 1134632 222487 897115 1036992 139877
Indonesia 759059 405570 -353488 836906 492962 -343944
Japan 899771 1184811 285039 880287 1054533 174246
Malaysia 899995 1001143 101148 836384 922360 85975
UK 887018 1011048 124029 880782 799256 -81526
DR Congo 784480 738784 -45696 685872 576096 -109776
Egypt 859756 838059 -21696 886248 665117 -221131
Ghana 738510 364252 -374258 756967 607614 -149354
Morocco 767844 666083 -101761 863477 1066309 202831
Nigeria 869612 638121 -231491 609199 351031 -258168
Tanzania 498637 239716 -258921 558762 535112 -23650
Tunisia 893837 596566 -297271 882542 799916 -82626 Source Authors Notes The notes in Table 4 apply