cross-border bank flows through foreign branches: evidence...

Cross-Border Bank Flows through Foreign Branches: Evidence from Korea

Youngjin Yun*

The views expressed herein are those of the author and do not necessarily reflect the official views of the Bank of Korea. When reporting or citing this paper, the author’s name should always be explicitly stated.

* Economist, Economic Research Institute, The Bank of Korea, Tel: +82-2-759-5471, E-mail: [email protected]

I thank Hosung Jung, Jongho Park, and Dongsoo Kang for carefully reading the drafts and providing valuable comments. I also thank the seminar participants at the Korea and World Economy XVII Conference for helpful discussions. I gratefully acknowledge the excellent research assistance provided by Hyeonseo Lee.

Contents

Ⅰ. Introduction ················································································ 1

Ⅱ. Foreign Bank Branches in Korea ····································· 4

Ⅲ. Empirical Framework ······························································ 8

Ⅳ. Monetary Policy ······································································ 12

Ⅴ. Macroprudential Policy ······················································· 21

Ⅵ. Conclusion ················································································· 25

References ························································································· 26

Cross-Border Bank Flows through Foreign Branches: Evidence from Korea

Global banks play an important role in international monetary transmission by allocating funds across the world through their foreign affiliates. Using monthly data on individual foreign bank branches in Korea from 2004 to 2018, this paper investigates the effects of foreign monetary policies and Korean macroprudential policy on the cross-border capital flows between global banks’ headquarters and their Korean branches. I find that foreign branches reduce borrowing from their headquarters by 2.4% of their assets after a one percentage point hike in the home-country policy rates. The effect is more significant for the branches with higher loan-to-asset ratios as their asset maturities are longer. Korea introduced leverage caps on banks’ FX derivative positions in 2010, and has been adjusting the cap depending on the macroeconomic situation. I find that lowering the cap makes foreign branches increase capital by receiving long-term capital from headquarters. The branches with higher bond-to-asset ratios respond more as they trade heavily in FX derivatives.

Keywords: Foreign bank, Bank flows, Monetary policy, Macroprudential policy

JEL Classification Numbers: G21, F34, F38

1 BOK Working Paper No. 2018-23

Ⅰ. Introduction

In the past decade following the Global Financial Crisis, the literature has

revealed that banking has become global, and international banks are now

the key to understand the international transmission of shocks. Internationally

active banks operate branches and subsidiaries in many countries, and

they function as a channel of international shock propagation. The

literature has focused on the role of global banks and their affiliates in

international capital flows, and many researchers have concluded that we

need to better understand the behavior of foreign banks over the business

cycle.

This paper dissects the operation of foreign bank branches in Korea to

learn how monetary policy and macroprudential policy affect the branches’

borrowing and lending with their headquarters. Borrowing and lending

between a global bank head office and its branch are financial transactions

within a banking group, but they are also cross-border capital flows that

may transmit shocks from one country to another. This paper considers

whether the effect of foreign monetary policy is imported through foreign

bank branches, and examines whether macroprudential policy is an effective

tool to cope with international spillovers.

Foreign bank branches play an important role in Korea by providing

foreign capital. They are major suppliers of foreign currency liquidity in

the Korean financial market. Their foreign borrowing in the form of

net-due-to accounts for around 20% of the banking sector’s total external

liabilities and 10% of the whole nation’s. The funds they channel from

abroad are eventually utilized in Korea as lending to firms, banks and the

government, thus bringing real impacts. Not only is their presence

significant, but they also provide a useful laboratory to study international

capital flows. More than 40 foreign bank branches from 17 different

countries around the world are operating in Korea at any given time. One

can exploit the regional differences in monetary policies to tease out the

effects on capital flows.

Cross-Border Bank Flows through Foreign Branches: Evidence from Korea 2

I use regulatory data on individual foreign bank branches in Korea.

The data cover the 15 years from March 2004 to February 2018 in

monthly frequency. The primary variable of concern is the net cross-border

borrowing of the branches through the internal capital markets of global

banking groups. First, I estimate the effects of home (origin) country

monetary policy on the branches’ borrowing from mother banks, exploiting

the differences in the national origins of each branch. The identification,

then, comes in a given month from the differences of monetary policies

among the countries from which the branches originate. We can also

count on the exogeneity of the origin country monetary policy from the

foreign branch operation in Korea. In doing this analysis, I group the

foreign branches by different business models and investigate bank

heterogeneity. I posit that banks with long-term assets, like loans, would

respond to the policy rate more than banks with short-term assets.

Next, I examine the effect of macroprudential policy. Korea introduced

FX derivative position leverage cap regulation in October 2010. Banks are

mandated to keep their FX derivative positions below several multiples of

their capital bases. I ask how this policy influences foreign bank branches’

capital (or long-term borrowing from parent banks). The identification in

this analysis comes from a differences-in-differences framework. I estimate

different responses of banks with different bond-to-asset ratios before the

policy. For branches focused on security trading, FX derivative transactions

are essential. Those banks would make more efforts to recapitalize compared

to banks with less bonds and more long-term assets.

The main findings are: First, a foreign bank branch responds to

monetary policy tightening in its home country by reducing borrowing

from the head office. Second, lowering the leverage cap on FX derivative

positions makes foreign branches capitalize more by receiving long-term

capital from head offices. Lastly, there is substantial underlying

heterogeneity among branches with different business models: loan-making

branches are sensitive to policy rate changes, while macroprudential policy

has greater effects on branches specialized in security trading.


A number of recent studies have analyzed foreign bank affiliates’ role

in the cross-border bank flows. Cetorelli and Goldberg (2012) is the first

to document the importance of internal capital markets between the global

bank head offices and their foreign affiliates. They analyze data on U.S.

banks and their affiliates outside the U.S. to find that the cross-border

internal capital markets are actively used to satisfy the funding needs of

home and foreign offices of global banks. Borrowing and lending among

the head office and its branches contribute to international shock

spillovers. Hoggarth et al. (2013) analyze foreign bank branches in the

U.K. and find that they rely on fickle forms of funding from abroad,

amplifying domestic credit cycles. Kwan et al. (2014) find that global

banks used their branches in Hong Kong as a funding source during

financial crises in the U.S. and Europe, so they transmitted the credit

crunch to Hong Kong. All these papers find that the internal capital

markets within global banking groups play an important role in

international spillover effects by exporting and importing shocks across

borders. The current paper contributes to this literature by presenting new

direct evidence on the significance of cross-border borrowing and lending

by foreign branches with their parent banks.

This paper is also related to the literature of the cross-border spillover

effects of monetary policy. This literature does find significant international

spillovers of monetary policy, but different studies find different directions

of such spillovers, and the channels through which the spillovers occur are

not well understood either. Using BIS banking data, Bruno and Shin

(2015) find that contractionary U.S. monetary policy leads to a decrease in

cross-border bank flows as banks’ financing costs rise. In contrast, Correa

et al. (2015) find from the same data that tighter monetary policy in a

country makes banks in that country allocate more funds abroad. They

explain this as an effect of the portfolio channel. Monetary tightening

makes domestic borrowers riskier, and hence banks invest more abroad.

Avdjiev and Hale (2018) reconcile the two findings and argue that the

direction of flows depends on international capital flow regimes. They


argue that in the boom regime, macroeconomic fundamentals reflected in

policy rate moves in concert with international lending, and researchers

see a positive relationship between policy rates and international lending.

However in a stagnant regime, the other components of the policy rate

dominate the relationship, and one sees a negative correlation between

policy rates and lending. The International Banking Research Network

also delves into the spillover effects of monetary policy. See Buch et al.

(2018) and other contributing papers in this project. Their conclusions

also vary: some countries observe positive effects of the foreign monetary

policy rate on bank lending, but some others find negative effects. My

paper contributes to this literature by adding another result which is in

line more with Bruno and Shin (2015). As funding costs rise after policy

rate hikes, foreign bank branches reduce borrowing from headquarters. My

paper also documents a specific channel of the spillover effects, foreign

bank branches.

The remainder of the paper is organized as follows. Section Ⅱ discusses the role of foreign bank branches in Korea, and explains

macroprudential policy targeting them. Section Ⅲ describes the data and lays

out the empirical framework. Section Ⅳ analyzes the effect of monetary

policy, and Section Ⅴ investigates the effect of macroprudential policy.

Finally, Section Ⅵ concludes.

Ⅱ. Foreign Bank Branches in Korea Foreign bank branches play an important role in the Korean financial

market by bringing foreign capital into Korea. Although they are small in

terms of asset size, their cross-border borrowing is significant. The total

asset size of foreign branches as of December 2016 is 278.1 trillion won,

which is 12.4% of that of domestic banks. However, the branches’

borrowing from head offices (net-due-to) sum up to 40 billion dollars,

accounting for 23% of all external debt of depository banks, and 10% of

the national external debt.


We focus on the net-due position of the foreign branches in this paper.

Net-due-to occurs when the branch borrows from its head office, and

net-due-from occurs when the branch lends to the head office. Foreign

branches fund their investment in Korea through net-due-to accounts.

They bring the funds they receive from the parent banks into Korea,

swaps them with Korean won, and eventually invest them in securities or

loans. Foreign branches prefer risk-free public bonds, and loans are mostly

directed to firms from their origin countries. Many foreign branches

engage in risk-free arbitrage in this fashion, and hence FX derivative

trading is a very important part of their operations.

Figure 1 shows the fluctuations of net-due-to and net-due-from

separately. Net-due-to is significantly larger than net-due-from, and hence

one can see that foreign branches mainly supply foreign capital to Korea

rather than borrowing from Korea. Also, we see that the fluctuation is

massive. For instance, just within the last five years, net-due-to went on a

large swing from 40 trillion to 60 trillion and then back to 40 trillion

won. I examine how this fluctuation is associated with the monetary

policies of the home and host countries, and whether macroprudential

policy is an effective tool to deal with the fluctuation.

Figure 1: Net-due-to and Net-due-from

Notes: Data is from FAIRS. The two lines show the sum of net-due-to and net-due-from of all foreign bank branches. The unit is in trillion Korean won.


Figure 2 shows the size of net-due-to in comparison with Korea's

liabilities measured in the locational banking statistics (LBS) of the BIS.

The left panel shows the break-downs of the liabilities by country at the

end of 2017. The blue bar is the size of Korea’s liabilities to foreign

banks recorded in the LBS, and the red bar is sum of net-due-to as

reported in the Korean branches’ balance sheets. For instance, Japanese

foreign bank branches’ borrowing from their parent banks in Japan

accounts for around half of Korea’s total liabilities to the Japanese banking

sector. The relative sizes of net-due-to are small in financial hub countries

like the U.S. or U.K, but for some other countries like Switzerland or the

Netherlands, net-due-to accounts for almost all of Korea’s liabilities to the

banks in the corresponding countries. The right panel shows the

time-series variation of the relative size of net-due-to compared with

Korea’s external liabilities as shown in the LBS. In general, around 20%

of Korea’s external liabilities to foreign banks are in the form of

net-due-to of foreign bank branches, and the fluctuation is also large.

Figure 2: Net-due-to and LBS Claims

(a) Comparison by Country (Dec. 2017) (b) Comparison over the Sample Period

Notes: Data are from the FAIRS and the BIS LBS. The figures compare net-due-to from the FAIRS with claims to Korea in the LBS. Panel (a) compares them by country at the end of 2017. The figures are in billion USD. Panel (b) shows the relative size of total net-due-to compared to the LBS claims over the sample period.


At any given time, more than 40 branches are operating in Korea.

Table 1 shows the list of countries and the number of global banks that

originate from them and have branches in Korea over the sample period.

There are a total of 55 foreign bank branches from 17 countries around

the world. Many of the branches are from countries linked tightly with the

Korean economy, such as the U.S. (15), China (6), and Japan (5). The

monthly observations of all branches sum up to 6,285.

Table 1: Number of Banks and Observations by Origin Country

Origin Banks Observations Origin Banks Observations

U.S. 15 1,275 Switzerland 2 336

China 6 762 Canada 1 164

France 5 681 Indonesia 1 28

Japan 5 689 Iran 1 70

U.K. 5 523 Netherlands 1 169

Australia 3 280 Pakistan 1 168

Singapore 3 509 Philippines 1 170

Germany 2 292 Spain 1 69

India 2 192 Sum 55 6,377

Notes: The origin countries of banks are identified using the data from Claessens and Van Horen (2015). All foreign bank branches that operated during the sample period (March 2004 - February 2018) are included.

In 2010, Korea introduced a series of macroprudential policies to make

its economy resilient to capital flow reversal in the banking sector.1)

Among the measures introduced in 2010, the policy that directly affects

foreign bank branches is the leverage cap on the foreign exchange

derivative contracts (mainly currency swaps and forwards). Beginning in

October 2010, foreign bank branches were required to maintain their net

1) Those measures and their effect on capital flows are well summarized in Bruno and Shin (2014) and Kang et al. (2016).


FX derivative position below a level of 2.5 times their capital. Later the

limit was adjusted to 2.0, 1.5, and then back to 2.0 in July 2011, January

2013, and July 2016, respectively. This regulation works against the

business of foreign branches as they use FX swaps to convert the funds

they borrow from headquarters to Korean won and invest in Korean

currency denominated assets.

In principle, the branches could maintain their asset sizes and

businesses by increasing capital. The leverage cap regulation is applied to

the sum of Tier 1 capital (Capital A) and Tier 2 capital (Capital B).

Capital A is the usual paid-in capital, and it is not easy to adjust it

frequently due to legal restrictions. Capital B, however, is essentially a

long-term borrowing from the head office. If the branch borrows from its

mother bank in the long term (longer than one year), then the fund is

treated as Capital B on which the branch can leverage. Therefore, this

regulation must have significant implications for foreign branches’

long-term borrowing from their mother banks. Later, we estimate the

effect of this regulation on Capital B in the Section Ⅴ.

Ⅲ. Empirical Framework

Our main source of data is the Financial Analysis Information Retrieval

System (FAIRS) of the Bank of Korea. FAIRS provides detailed monthly

balance sheet information of financial intermediaries in Korea. The data in

FAIRS is originally collected by the Bank of Korea or by the Financial

Supervisory Service for the purpose of monetary policy and bank

supervision. The available data start from March 2004, and the analysis

sets the sample period as the 15 years from March 2004 to February 2018.

I also rely on other sources of data for macroeconomic variables for the

17 countries in the sample. The industrial production index data is obtained

from Bloomberg, and the policy rates of the countries are from CEIC data

and the central banks of each country. The macroeconomic data on Korea are

sourced from the Bank of Korea. To identify the origin countries of each bank,


I used the bank ownership database of Claessens and Van Horen (2015).

In the regression analyses following this section, I control for banking

crises by including a dummy variable. It is documented in Kwan et al.

(2014) that foreign branches withdraw funds from the host country during

crisis in their home country. It is also common for central banks to lower

policy rates during a banking crisis. Therefore, while in a banking crisis, a

positive correlation between net-due-to and policy rates would be observed,

but the causation is coming from the crisis, not from policy rates. The

normal time effects of home policy rates on net-due are likely to be

negative but would be obscured if one did not control for the crisis

properly. Over the sample period, we have the Global Financial Crisis

(GFC) and European Debt Crisis (EDC). The GFC was global as its name

suggests, so every country in the sample lowered its policy rate, while the

EDC affected mainly European countries. I set dummy variable

to be 1 for every branch during GFC, and set it to be 1 for only those

branches from European countries during the EDC. The GFC period is

set to be one year following the Lehman bankruptcy (from September

2008 to August 2009) and the EDC period is set to be the one year

preceding the introduction of the European Central Bank’s long-term

refinancing operation (from January 2011 to December 2011).

Foreign bank branches have different business models, according to

which they hold different assets. Some branches specialize in making

loans, while others are focused more on trading securities. This results in

different maturity of assets among banks. Typically, loans have longer

maturity than securities, and loan-making branches tend to hold bonds for

the long term whereas security trading branches rarely hold bonds to

maturity. Therefore, loan-type branches’ assets are of longer term than

those of security-type branches. These different branch types may respond

differently to monetary policy and macroprudential policy changes.

To analyze business model heterogeneity, I set two subsamples, L group

and B group, based on branches’ loan-to-asset and bond-to-asset ratios

averaged over the sample period. L group comprises the 14 banks with


above-median loan-to-asset ratios and below-median bond-to-asset ratios. B

group includes the 13 banks with below-median loan-to-asset ratios and

above-median bond-to-asset ratios. Figure 3 displays a foreign bank branch

as a dot on the loan-to-asset and bond-to-asset plane. The banks are

notated with the country abbreviations. There is a clear negative correlation

between the two ratios. The red squares are the L group banks, and the

blue diamonds are the B group banks. The asset compositions of L group

are very different from those of B group.

Figure 3: Loan- and Bond-to-Asset Ratios of Branches

Notes: The loan-to-asset ratio and bond-to-asset ratio are the averages over the sample periods.(March 2004 - February 2018) Banks are notated with the name of the countries where their head offices are located. Squares in red color are the 14 banks with an above-median loan-to-asset ratio and below-median bond-to-asset ratio. Diamonds in blue color are the 13 banks with a below-median loan-to-asset ratio and above-median bond-to-asset ratio.

Summary statistics for key variables are given in Table 2. We can see

the differences in balance sheet composition between L group and B

group clearly. The mean loan amount is 0.35 trillion won in L group,

while it is 0.03 trillion won in B group. The mean public bond holdings

are 2.0 trillion in B group and 0.19 trillion in L group. Also, B group

trades FX currency derivatives heavily compared to L group.


Table 2: Summary Statistics

All Sample (n=55)

Mean Median St.Dev. Mean Median St.Dev.

net-due-to 1.21 0.50 1.77 Loans 0.16 0.02 0.35

net-due-from 0.14 0.00 0.59 Public Bonds 1.08 0.27 1.72

Capital A 0.13 0.07 0.16 Total Assets 6.20 3.98 6.97

Capital B 0.28 0.07 0.56 Deriv/Assets 3.49 2.16 3.75

L Group (n=14)


net-due-to 1.87 0.62 2.49 Loans 0.35 0.08 0.55




B Group (n=13)


net-due-to 1.15 0.74 1.31 Loans 0.03 0.00 0.07




Notes: Data are from the FAIRS and cover from March 2004 to February 2018. The figures are in trillion Korean won. The 14 banks in L group have an above-median loan-to-asset ratio and below-median bond-to-asset ratio. The 13 banks in B group have an above-median bond-to-asset ratio and below-median loan-to-asset ratio. Public bonds include treasury bonds and central bank bonds. The derivatives are currency related only.

Table 3 presents the average security holdings of different groups of

banks decomposed by maturity. For B group, most of their securities are

short-term trading securities (85.8%). Held-to-maturity securities account

for only 0.6% of holdings. This is in stark contrast with L group. Trading

securities account for only 0.1%, and the L group banks hold 32% of

their securities to maturity. Given that L group holds a lot of loans and B

group does not, we can conclude that the effective maturity of bank assets

is much longer for L group than B group.


Table 3: Security Holdings by Maturity

Trading Security Avaiable for Sale Held to Maturity Sum

All 0.842 0.550 0.110 1.502(n=55) (56.1%) (36.6%) (7.3%) (100.0%)

L group 0.001 0.725 0.340 1.066(n=14) (0.1%) (68.0%) (31.9%) (100.0%)

B group 1.756 0.279 0.012 2.047(n=13) (85.8%) (13.6%) (0.6%) (100.0%)

Notes: Data are from FAIRS. The unit is trillion won in 2015 prices. The figures are the averages over the banks and the sample period.

I drop the observations with an asset growth rate of below -50% or

above 200%, and also exclude the outlying 1% (the top 0.5% and bottom

0.5%) of key variables to avoid influences of possible outliers. The balance

sheet items are deflated using the Korean CPI and are measured in 2015

Korean won. All regressions are weighted by bank sizes to prevent a large

number of small banks from altering the regression results in opposition

to the aggregate effects. Standard errors are clustered at the bank level to

allow arbitrary serial correlations of error terms within a bank. (Bertrand

et al., 2004)

Ⅳ. Monetary Policy

In this section, I study how foreign monetary policies cross borders

through foreign bank branches. We are mainly interested in the foreign

monetary policy spillover, but we also examine the effect of Korean

monetary policy after dropping time fixed effects.


1. Origin Country Monetary Policy

I begin by asking how home country monetary policy affects the

cross-border borrowing of foreign bank branches from their mother banks.

The baseline model includes bank fixed effects as well as monthly fixed

effects which control for most of the unobserved macroeconomic factors.

The origins of the foreign bank branches are widely spread throughout

the world as presented in Table 1. I exploit the differences in monetary

policies in origin countries in a given month. The main dependent

variable is , which I define as net-due-to minus net-due-from.

Therefore, a positive means that the branch owes funds to its

head office. The following is the baseline estimating equation:

∆

∆ × ′ (1)

where subscript indexes banks, and indexes months. The regressand

is the change in normalized by the beginning of the period total

assets. More than 5% of is negative in the sample. Rather than

taking the differences of log , I normalize the change in

with the bank’s asset size of the previous month.2) The terms and

denote bank fixed effects and monthly fixed effects. s are the

autoregressive coefficients.3)

The main regressor ∆ is the change in the monetary policy

rate of the country where the headquarters of branch is located. The

data is monthly and I get quarterly effects by including two lags of the

policy rate changes and doing an F-test on the sum of all three

2) This form of regressand was used in other studies, such as those published by the International Banking Research Network.

3) We use monthly data for 15 years and = 180. There is little concern of the Nickell bias.


coefficients. The null hypothesis is that the sum is equal to zero, and I

report p-value from the test. I do the same for the other regressors with

lags. A negative and significant sum of implies that foreign bank

branches reduce borrowing from parent banks after home country

monetary tightening. is the dummy variable for banking crises

as explained in the previous section. Its interaction term with the policy

rate changes is included to control for banking crises. is a vector of

control variables, which includes growth rates in industrial production of

origin countries and its two lagged terms, bank capital ratio and log of

bank total assets.

Table 4: Effects of Origin Country Monetary Policy

(1) (2) (3) (4)

Group All All L group B group

HQ MP rate -2.43* -1.54** -7.55* 1.24(0.076) (0.023) (0.099) (0.562)

HQ MP rate × Crisis 5.41** 4.55*** 13.43* -1.06(0.011) (0.001) (0.071) (0.736)

Crisis -0.793 -0.587 -1.245(0.216) (0.344) (0.283)

Lags of -0.17*** -0.17*** 0.04 -0.29**

(0.003) (0.003) (0.370) (0.011)

Observations 5,631 5,631 1,675 1,796

Number of banks 53 53 14 13

0.078 0.078 0.188 0.202

Bank F.E. yes yes yes yes

Monthly F.E. yes yes yes yes

Shadow rate no yes no no

Notes: The sample period is from March 2004 to February 2018. The dependent variable is the change in net net-due-to relative to assets at the beginning of the period. The top and bottom 0.5% are trimmed. Variables are in 2015 Korean won. The reported coefficients are the sum of all coefficients of each respective variable and its lags. Reported in brackets are the p-values from the test of the null hypothesis that the sum of the corresponding coefficients is zero. The growth in the home country industrial production index, capital ratio and log asset size are included as covariates, but their coefficients are not reported. Regressions are weighted by bank asset size. Standard errors are clustered at the bank level. ***, **, and * indicate statistical significance at the 10%, 5% and 1% level, respectively.


Table 4 presents the results. Columns (1) and (2) cover the full sample.

There is less variation in the monetary policy rates of advanced countries

after the GFC due to quantitative easing. Therefore, in column (2) I use

Wu-Xia shadow rates for the U.S., U.K., and Eurozone countries as a

robustness check. (Wu and Xia, 2016) In both of the regressions, the sums

of coefficients on home country monetary policy changes are negative and

significant, meaning that branches reduce borrowing from head offices

after policy rate hikes in the home countries. The size of the coefficient is

also economically significant. After a one percentage point hike in the home

policy rate, the branches reduce borrowing by 2.4% of their assets over the

following three months. Applied to the sum of all branches’ assets at the

end of 2016, this means 5.6 billion USD reduction.4) The results confirm

that Korea imports foreign monetary policy effects via foreign bank branches.

The sum of coefficients on the interaction term is positive and

significant. Also, the coefficient is larger in absolute value than the

coefficients on home country monetary policy changes. Therefore, during

banking crises, the coefficients on home monetary policy changes are

overall positive. We know that central banks lowered policy rates during

crises, and hence the results imply that foreign branches reduced

borrowing from head offices significantly during crises. This is consistent

with the findings of Kwan et al. (2014).

Columns (3) and (4) show the results from the same regression on the

subsamples, L group and B group. The policy rate coefficients are

negative and significant for L group but not for B group. The branches

making significant loans are sensitive to the borrowing cost changes from

home, while the branches concentrating on security trading are not

sensitive to home country policy rate changes. As discussed with Table 3,

loan-making branches have longer effective maturities of assets. Interest

rate changes directly affect their profits. Security-trading branches tend to

hold bonds only for short periods, and they trade heavily in FX

4) This is the effect of the shock that all 17 countries raise their interest rate by a one percentage point. For the effect of a specific country, one need to consider the bank heterogeneity.


derivatives. It seems that their profits are less sensitive to policy rate

changes for this reason.

2. Korean Monetary Policy

The next question is on the effect of host country monetary policy.

Korean monetary policy is common to all branches in the sample, so time

fixed effects compete with Korean monetary policy in regressions. To

estimate the effect of Korean monetary policy, I drop all time fixed

effects. Instead, I control further for macroeconomic changes in Korea.

The growth rate of the Korean industrial production index and its two

lagged terms are included to control for the host country business cycle.

Swap rates play an important role in the foreign bank branches’ financial

intermediation as explained in the section Ⅱ. Changes in the monthly

average of daily swap rates are included in the regression.

We also need to control for the macroprudential policy introduced in

2010. Since the government has been adjusting the leverage cap

depending on the macroeconomic circumstances, I generate a time-series

vector rather than a dummy variable out of it. The policy changes were

announced several months earlier than their effective dates, and the

branches adjusted their balance sheets before the policies became effective.

Therefore, I use the announcement dates of the leverage cap changes

rather than the effective dates. The average FX derivative leverage ratio was

301% (April 2010) right before the first policy announcement, and it is

reported that the leverage ratios of foreign branches rarely exceeded 400%

before the introduction of the cap. Hence, I set the leverage cap variable,

, to be 4.0 before the introduction of the policy. Including the

change of and its two lags as covariates, the following

estimating equation is derived:


∆

∆ ×

∆

∆ ×

∆

∆ ′ (2)

where ∆ is the change in Korean monetary policy. These terms

are also interacted with the crisis dummy.

Table 5 shows the results. In columns (1) and (2), the coefficients on

home country monetary policy changes are negative as in the previous

regressions, but the p-values slightly exceed 0.10. The sum of coefficients

on Korean monetary policy changes are positive and significant. This

means that Korean policy rate hikes induce foreign bank branches to

borrow more from mother banks. After a one percentage point rise in the

Korean policy rate, the branches increase their borrowing from

headquarters over the next quarter by 4.8% of their assets. This effect is

null during banking crises in home countries. The coefficients to changes

in swap rates are negative, meaning that the branches borrow more when

the swap rates are more favorable to them.

The results are heterogeneous among different banks. Columns (3) and

(4) show that L group banks are sensitive to the policy rate changes of

home and host countries, but B group banks are not. It also shows that

following the tightening of the leverage cap, L group banks increase their

borrowing from mother banks, but B group banks do not.


Table 5: Effects of Host Country Policy

(1) (2) (3) (4)


HQ MP rate -1.82 -1.01 -7.18* 1.82(0.106) (0.116) (0.057) (0.321)

HQ MP rate × Crisis 4.99*** 3.76*** 13.86** -0.51(0.006) (0.002) (0.036) (0.882)

KR MP rate 4.82*** 4.62*** 4.66* 3.36(0.005) (0.008) (0.060) (0.394)

KR MP rate × Crisis -4.86** -4.53** 0.17 -3.45(0.015) (0.018) (0.970) (0.484)

Crisis 0.69** 0.78*** 2.18* 0.327(0.014) (0.009) (0.070) (0.405)

Macropru 0.74 0.74 -2.65** -0.92(0.531) (0.531) (0.043) (0.684)

Swap Rate -3.77*** -3.60** -7.97** -3.83**

(0.007) (0.011) (0.045) (0.046)

Lags of -0.16*** -0.16*** 0.03 -0.29**

(0.003) (0.003) (0.551) (0.016)

Observations 5,631 5,631 1,675 1,796

Number of Banks 53 53 14 13

0.029 0.029 0.074 0.072


Shadow Rate no yes no no

Notes: The sample period is from March 2004 to February 2018. The dependent variable is the change in net net-due-to relative to assets at the beginning of the period. The top and bottom 0.5% are trimmed. Variables are in 2015 Korean won. The reported coefficients are the sum of all coefficients of each respective variable and its lags. Reported in brackets are the p-values from the test of the null hypothesis that the sum of the corresponding coefficients is zero. The growth rate in industrial production index of home and host countries, capital ratio and log asset size are included as covariates, but their coefficients are not reported. Regressions are weighted by bank asset size. Standard errors are clustered at the bank level. ***, **, and * indicate statistical significance at the 10%, 5% and 1% level, respectively.


3. Interest Rate Differentials

Since the foreign branches are lending in Korea and borrowing from their

headquarters, it might be the interest rate differential that matters for

them in the end. The coefficients in Table 5 also imply this possibility:

the coefficients on liability-side interest rate are negative and the

coefficients on the asset-side interest rate are positive. This can be tested

by setting a restriction that the coefficients for the policy rates of the

home and host countries sum up to zero. From the estimating equation

(2), the null hypothesis can be written as + . Table 6 shows the

test results based on the regressions in Table 5. The numbers are the

p-values from the test of the null hypothesis. Tests from the regressions

with normal policy rates do not reject the null, implying that the interest

rate differential is the variable to which the foreign branches are

responding. Only the test from the regression with shadow rates rejects

the null. This means that the effects of home and Korean policy rate are

different in magnitudes.

Table 6: F-test for Interest Rate Differential Restriction

(1) (2) (3) (4)


: HQMP+KRMP=0 0.142 0.048 0.417 0.181

Notes: The figures show the p-values from the test of the null hypothesis that the coefficients for headquarter monetary policy rate and Korean policy rate sum up to zero in the regression of the table 5.

I proceed to the regressions using interest rate gaps instead of home

and host country monetary policies. Every other setting is the same as

with the regressions in Table 5, but home and host country monetary

policies are now replaced with interest rate differentials between Korea and

the branches’ origin countries.


Table 7: Interest Rate Differentials

(1) (2) (3) (4)


Interest Rate Differential 2.66*** 1.56** 5.59** 0.01(0.006) (0.014) (0.028) (0.995)

Interest Rate Differential × Crisis -4.01*** -3.24*** -4.41 0.15(0.003) (0.001) (0.324) (0.962)

Crisis 0.283 0.37* 0.357 0.191(0.189) (0.092) (0.632) (0.643)

Macropru 0.34 0.15 -2.43* -1.51(0.777) (0.902) (0.077) (0.516)

Swap Rate -3.82*** -3.96*** -7.78** -4.37*

(0.006) (0.005) (0.020) (0.060)

Lags of -0.16*** -0.16*** 0.03 -0.29**

(0.003) (0.003) (0.582) (0.017)

Observations 5,631 5,631 1,675 1,796

Number of Banks 53 53 14 13

0.028 0.027 0.060 0.069


Shadow Rate no yes no no

Notes: The sample period is from March 2004 to February 2018. The dependent variable is the change in net net-due-to relative to assets at the beginning of the period. The top and bottom 0.5% are trimmed. Variables are in 2015 Korean won. The reported coefficients are the sum of all coefficients of each respective variable and its lags. Reported in brackets are the p-values from the test of the null hypothesis that the sum of the corresponding coefficients is zero. The growth rate in industrial production index of home and host countries, capital ratio and log asset size are included as covariates, but their coefficients are not reported. Regressions are weighted by bank asset size. Standard errors are clustered at the bank level. ***, **, and * indicate statistical significance at the 10%, 5% and 1% level, respectively.

Table 7 presents the results, which are not very different from the

results in Table 5. Column (1) indicates that, following a one percentage

point increase in the interest rate gap, foreign branches increase their

borrowing from parent banks by 2.7% of their assets. The impact is large

for L group, but B group is not responsive to interest rate differentials at

all.


Ⅴ. Macroprudential Policy

Beginning in October 2010, foreign bank branches were mandated to

maintain their FX derivative positions below 250% of their capital. In

response to this regulation, the foreign bank branches could increase their

Capital B by borrowing from the headquarters in the long term. Figure 2

shows Capital B and the leverage cap together. For comparison, the

leverage cap is set to 4.0 before the introduction of the regulation in

2010.5) The figure shows that the foreign branches significantly increased

Capital B prior to the introduction of the leverage cap regulation. The

figure also hints at a negative correlation of the aggregate Capital B with

the follow-up changes in the cap.

Figure 4: Capital B and the Leverage Cap Regulation

Notes: Data is from FAIRS. The unit is in trillion Korean won. The leverage cap is shown as 4.0 before its introduction in October 2010.

5) As explained earlier, even before the introduction of the policy, the FX derivative leverage ratios of foreign branches rarely exceeded 400%. The average ratio before the announcement of the policy was 301% (April 2010).


To evaluate the effect of macroprudential policy on the long-term

borrowing of branches, I regress the log growth rates of Capital B on the

change in the leverage cap.6) I adjust the sample period to be after the

GFC only: from September 2009 to February 2018. The main regressor is

the changes in the leverage cap recorded at the announced dates and its

two lagged terms. First, I quantify the direct effects of macroprudential

policy without time fixed effects, controlling for the swap rate, interest

rate differential, and its interaction with the crisis dummy. I do the same

regression on the subsamples as before.

To be more rigorous, I set up a differences-in-differences specification. I

estimate different changes in Capital B after the macroprudential policy

changes by banks with different pre-policy bond-to-asset ratios. The bond

business branches trade public bonds frequently and they tend to trade

lots of FX derivatives along with the bonds. Therefore, I posit that the

behavior of bond business branches and loan business branches would be

different after the regulation. Since the macroprudential policy is first

introduced in October 2010, I use the average bond-to-asset ratio of 2009

to avoid simultaneity. Monthly fixed effects are included in this

specification. To allow for different responses to the changes in policy

rates, is also interacted with the bond-to-asset ratio. Banking crisis

is controlled by including triple interaction terms and all the necessary

two-variable interaction terms. This yields the following regression equation:

∆ ln

∆ ×

∆ ××

∆ × ×

∆ln ′ (3)

6) I winsorize the top and bottom 2% of the Capital B growth rate given its long tail and fewer observations.


where is the bond-to-asset ratio of bank . The covariate

includes the triple interaction term × × , and all

the necessary auxiliary terms. We are mostly interested in the sign of

in this regression.

Table 8 presents the results. Since some of the branches do not have

any Capital B, the observations and number of banks shrink to 2,414 and

36, respectively. In columns (1) and (2), the sum of coefficients on the

leverage cap is negative and significant for the entire sample. When the

government lowers the leverage cap by 100 percentage points, foreign

bank branches increase their Capital B growth rate by 5.8 percentage

points. Columns (3) and (4) show the results from the subsamples. It is

mainly B group banks which increase their Capital B by borrowing more

in the long term from parent banks. L group banks’ Capital B do not

respond to the leverage cap changes.

Column (5) is the direct estimation of the equation (3). L group banks

have small bond-to-asset ratios and column (3) shows that they do not

adjust their Capital B responding to the macroprudential policy. Hence,

the column (5) regression is essentially setting L group as a control group

and estimate the effect of having higher bond-to-asset ratio on the

adjustments of Capital B following leverage cap changes. The coefficient

to the interaction term is negative and significant, meaning that high

bond-to-asset ratio banks increased Capital B significantly more than low

bond-to-asset ratio banks after the leverage cap is lowered. The

interquartile range of is 17.7 and the bottom 25 percentile

bank’s is zero. Therefore, the results means that Capital B

growth rate of a bank with a top 25 percentile bond-to-asset ratio is 7

percentage points (=17.7×0.40) higher than a bank with zero bond holdings

after the leverage cap is lowered by 100%. The economic magnitude is

similar with columns (1) and (4).

The leverage cap regulation is aiming to reduce the procyclicality and

volatility of cross-border bank liabilities. The previous results show that the

policy induces branches to fund their investment in Korea from more stable


Table 8: Macroprudential Policy and Capital B

(1) (2) (3) (4) (5)

Group All All L group B group All

Macropru -5.80** -5.86** 2.20 -8.59**

(0.047) (0.041) (0.758) (0.015)

Macropru×BondRatio -0.40**

(0.043)

Interest Rate Differential 3.85 2.50 0.39 0.72(0.372) (0.205) (0.974) (0.934)

HQMP 2.57(0.523)

HQMP×BondRatio 0.20(0.145)

Swap Rate -0.92 -1.86 -1.68 -2.85(0.912) (0.811) (0.881) (0.855)

Lagged Capital B -0.02 -0.01 0.00 0.00 -0.03(0.795) (0.842) (0.964) (0.993) (0.653)

Observations 2,414 2,414 825 773 2,287

Number of Banks 36 36 12 12 32

0.016 0.017 0.036 0.028 0.076

Bank F.E. yes yes yes yes yes

Monthly F.E. no no no no yes

Shadow Rate no yes no no no

Notes: The sample period is from September 2009 to February 2018. The dependent variable is the change in net net-due-to relative to assets at the beginning of the period. The top and bottom 2% are winsorized. Variables are in 2015 Korean won. The reported coefficients are the sum of all coefficients of each respective variable and its lags. Reported in brackets are the p-values from the test of the null hypothesis that the sum of the corresponding coefficients is zero. BondRatio is the average public-bond-to-asset ratio of 2009, and it is demeaned. The interactions of banking crises dummy with Macropru×BondRatio and HQMP×BondRatio are included. All the necessary auxiliary terms are also included, but their coefficients are not reported. The growth in industrial production index of home countries, capital ratio and log asset size are included as covariates, but their coefficients are not reported. Regressions are weighted by bank asset size. Standard errors are clustered at the bank level. ***, **, and * indicate statistical significance at the 10%, 5% and 1% level, respectively.


long-term borrowing. This might help mitigate the volatility of

cross-border bank flows. Indeed, Bruno and Shin (2014) analyze capital

flows into Korea before and after the policy and conclude that it became

less sensitive to global factors.

Ⅵ. Conclusion

This paper documents the significance of the internal capital markets

between global banks’ head offices and their Korean branches. The

amounts of cross-border borrowing and lending change actively in

response to policy changes, thus working as an important channel of

international shock spillovers. I find that home monetary policy tightening

reduces a foreign branch’s borrowing from its mother bank. Foreign bank

branches contribute to the Korean economy by supplying foreign capital to

the needy domestic capital market, but such capital is accompanied by

foreign monetary policy effects. Macroprudential policy may be used to

deal with these spillover effects. I find that the leverage cap regulation

induces foreign branches to recapitalize by receiving long-term capital

from headquarters.

The results highlight that bank heterogeneity plays a crucial role in

both monetary policy spillover effects and macroprudential policy effects.

Branches with different business models respond differently to these

shocks. While branches focused on lending are sensitive to policy rate

changes, branches specialized in security trading respond more to

macroprudential policy. The loan business branches have longer investment

horizons, and therefore their profits are more sensitive to policy rate

changes. On the other hand, the bond business branches tend to hold

bonds for short periods, presumably generating profits by exploiting

short-term deviations from the market equilibrium. As FX derivative

transactions are essential in their arbitrage business, the bond business

branches increase capital after the leverage cap is lowered. The bank

heterogeneity need to be considered carefully in policy implementations.


References

Avdjiev, S. and Hale, G. (2018), “U.S. Monetary Policy and Fluctuations of International Bank Lending,” Federal Reserve Bank of San Francisco Working Paper.

Bertrand, M., Duflo, E., Mullainathan, S., et al. (2004), “How Much Should We Trust Differences-in-Differences Estimates?” The Quarterly Journal of Economics, Vol. 119(1), pp. 249-275.

Bruno, V. and Shin, H. S. (2014), “Assessing Macroprudential Policies: Case of South Korea,” The Scandinavian Journal of Economics, Vol. 116(1), pp. 128-157.

Bruno, V. and Shin, H. S. (2015), “Capital Flows and the Risk-Taking Channel of Monetary Policy,” Journal of Monetary Economics, Vol. 71, pp. 119-132.

Buch, C. M., Bussiere, M., Goldberg, L., and Hills, R. (2018), “The International Transmission of Monetary Policy,” NBER Working Paper.

Cetorelli, N. and Goldberg, L. S. (2012), “Banking Globalization and Monetary Transmission,” The Journal of Finance, Vol. 67(5), pp. 1811-1843.

Claessens, S. and Van Horen, N. (2015), “The Impact of the Global Financial Crisis on Banking Globalization,” IMF Economic Review, Vol. 63(4), pp. 868-918.

Correa, R., Paligorova, T., Sapriza, H., and Zlate, A. (2015), “Cross-Border Bank Flows and Monetary Policy,” Manuscript, Federal Reserve Board of Governors. Washington DC.

Hoggarth, G., Hooley, J., and Korniyenko, Y. (2013), “Which Way Do Foreign Branches Sway? Evidence from the Recent U.K. Domestic Credit Cycle,” Bank of England Financial Stability Paper, No. 22.


Kang, D., Jeong, D., et al. (2016), “Investment Patterns of Foreign Bank Branches in Korea and Their Role in the Foreign Exchange Market,” Macroprudential Regulation of International Finance: Managing Capital Flows and Exchange Rates, KDI/EWC series on Economic Policy, pp. 201.

Kwan, S., Wong, E. T., and Hui, C.-h. (2014), “The International Transmission of Shocks: Foreign Bank Branches in Hong Kong during Crises,” Federal Reserve Bank of San Francisco Working Paper, No. 2014-25.

Wu, J. C. and Xia, F. D. (2016), “Measuring the Macroeconomic Impact of Monetary Policy at the Zero Lower Bound,” Journal of Money, Credit and Banking, Vol. 48(2-3), pp. 253-291.

<Abstract in Korean>

외은지점을 통한 은행자본유출입: 한국의 사례

윤영진*

글로벌 은행은 해외의 자은행 및 지점으로 자금을 배분함으로써 국제적

인 통화정책 충격 전파에 중요한 역할을 한다. 본고는 2004년부터 2018년

까지 국내 외은지점의 월별 재무상태표 자료를 이용하여 글로벌 은행의 본

부와 한국지점 간의 자본흐름에 외국 통화정책과 한국 거시건전성정책이

어떤 영향을 미치는지 분석하였다. 국내 외은지점들은 본국의 금리가 1%p

상승할 때 자기 총자산의 2.4%만큼 본부로부터의 차입을 줄이는 것으로 나

타났다. 이 효과는 자산대비 대출비중이 높아 자산의 만기가 긴 지점들에게

더 강하게 나타났다. 한국은 2010년에 선물환포지션 한도규제를 도입하여

이후 거시경제여건에 따라 능동적으로 이를 조정해왔다. 선물환포지션 한

도 인하는 외은지점의 본부로부터의 장기자금조달을 늘리는 효과를 내는

것으로 분석되었다. 자산대비 채권비중이 높고 선물환 거래가 많은 지점들

이 더 강하게 반응하였다.

핵심 주제어: 외국은행, 자본이동, 통화정책, 거시건전성정책

JEL Classification: G21, E34, F38

* 한국은행 경제연구원 국제경제연구실 부연구위원(전화: 02-759-5471, E-mail: [email protected])

이 연구내용은 집필자의 개인의견이며 한국은행의 공식견해와는 무관합니다. 따라서 본 논문의 내용

을 보도하거나 인용할 경우에는 집필자명을 반드시 명시하여주시기 바랍니다.

BOK 경제연구 발간목록한국은행 경제연구원에서는 Working Paper인 『BOK 경제연구』를 수시로 발간하고 있습니다.BOK 경제연구』는 주요 경제 현상 및 정책 효과에 대한 직관적 설명 뿐 아니라 깊이 있는 이론 또는 실증 분석을 제공함으로써 엄밀한 논증에 초점을 두는 학술논문 형태의 연구이며 한국은행 직원 및 한국은행 연구용역사업의 연구 결과물이 수록되고 있습니다.BOK 경제연구』는 한국은행 경제연구원 홈페이지(http://imer.bok.or.kr)에서 다운로드하여 보실 수 있습니다.

제2015 -1 글로벌 금융위기 이후 주요국 통화정책 운영체계의 변화

김병기⋅김인수

2 미국 장기시장금리 변동이 우리나라 금리기간구조에 미치는 영향 분석 및 정책적 시사점

강규호⋅오형석

3 직간접 무역연계성을 통한 해외충격의 우리나라 수출입 파급효과 분석

최문정⋅김근영

4 통화정책 효과의 지역적 차이 김기호

5 수입중간재의 비용효과를 고려한 환율변동과 수출가격 간의 관계

김경민

6 중앙은행의 정책금리 발표가 주식시장 유동성에 미치는 영향

이지은

7 은행 건전성지표의 변동요인과 거시건전성 규제의 영향

강종구

8 Price Discovery and Foreign Participation in The Republic of Korea's Government Bond Futures and Cash Markets

Jaehun Choi⋅Hosung Lim⋅Rogelio Jr. Mercado⋅Cyn-Young Park

9 규제가 노동생산성에 미치는 영향: 한국의 산업패널 자료를 이용한 실증분석

이동렬⋅최종일⋅이종한

10 인구 고령화와 정년연장 연구(세대 간 중첩모형(OLG)을 이용한 정량 분석)

홍재화⋅강태수

11 예측조합 및 밀도함수에 의한 소비자물가 상승률 전망

김현학

12 인플레이션 동학과 통화정책 우준명

13 Failure Risk and the Cross-Section of Hedge Fund Returns

Jung-Min Kim

14 Global Liquidity and Commodity Prices Hyunju Kang⋅Bok-Keun Yu⋅Jongmin Yu

제2015 -15 Foreign Ownership, Legal System and Stock Market Liquidity

Jieun Lee⋅Kee H. Chung

16 바젤Ⅲ 은행 경기대응완충자본 규제의 기준지표에 대한 연구

서현덕⋅이정연

17 우리나라 대출 수요와 공급의 변동요인 분석 강종구⋅임호성

18 북한 인구구조의 변화 추이와 시사점 최지영

19 Entry of Non-financial Firms and Competition in the Retail Payments Market

Jooyong Jun

20 Monetary Policy Regime Change and Regional Inflation Dynamics: Looking through the Lens of Sector-Level Data for Korea

Chi-Young Choi⋅Joo Yong Lee⋅Roisin O'Sullivan

21 Costs of Foreign Capital Flows in Emerging Market Economies: Unexpected Economic Growth and Increased Financial Market Volatility

Kyoungsoo Yoon⋅Jayoung Kim

22 글로벌 금리 정상화와 통화정책 과제: 2015년 한국은행 국제컨퍼런스 결과보고서

한국은행 경제연구원

23 The Effects of Global Liquidity on Global Imbalances

Marie-Louise DJIGBENOU-KRE⋅Hail Park

24 실물경기를 고려한 내재 유동성 측정 우준명⋅이지은

25 Deflation and Monetary Policy Barry Eichengreen

26 Macroeconomic Shocks and Dynamics of Labor Markets in Korea

Tae Bong Kim⋅Hangyu Lee

27 Reference Rates and Monetary Policy Effectiveness in Korea

Heung Soon Jung⋅Dong Jin Lee⋅Tae Hyo Gwon⋅Se Jin Yun

28 Energy Efficiency and Firm Growth Bongseok Choi⋅Wooyoung Park⋅Bok-Keun Yu

29 An Analysis of Trade Patterns in East Asia and the Effects of the Real Exchange Rate Movements

Moon Jung Choi⋅Geun-Young Kim⋅Joo Yong Lee

30 Forecasting Financial Stress Indices in Korea: A Factor Model Approach

Hyeongwoo Kim⋅Hyun Hak Kim⋅Wen Shi

제2016 -1 The Spillover Effects of U.S. Monetary Policy on Emerging Market Economies: Breaks, Asymmetries and Fundamentals

Geun-Young Kim⋅Hail Park⋅Peter Tillmann

2 Pass-Through of Imported Input Prices to Domestic Producer Prices: Evidence from Sector-Level Data

JaeBin Ahn⋅Chang-Gui Park⋅Chanho Park

3 Spillovers from U.S. Unconventional Monetary Policy and Its Normalization to Emerging Markets: A Capital Flow Perspective

Sangwon Suh⋅Byung-Soo Koo

4 Stock Returns and Mutual Fund Flows in the Korean Financial Market: A System Approach

Jaebeom Kim⋅ Jung-Min Kim

5 정책금리 변동이 성별 ‧ 세대별 고용률에 미치는 영향

정성엽

6 From Firm-level Imports to

Aggregate Productivity: Evidence

from Korean Manufacturing Firms Data

JaeBin Ahn⋅Moon Jung Choi

7 자유무역협정(FTA)이 한국 기업의 기업내 무역에 미친 효과

전봉걸⋅김은숙⋅이주용

8 The Relation Between Monetary and Macroprudential Policy

Jong Ku Kang

9 조세피난처 투자자가 투자 기업 및 주식시장에 미치는 영향

정호성⋅김순호

10 주택실거래 자료를 이용한 주택부문 거시건전성 정책 효과 분석

정호성⋅이지은

11 Does Intra-Regional Trade Matter in Regional Stock Markets?: New Evidence from Asia-Pacific Region

Sei-Wan Kim⋅Moon Jung Choi

12 Liability, Information, and Anti-fraud Investment in a Layered Retail Payment Structure

Kyoung-Soo Yoon⋅Jooyong Jun

13 Testing the Labor Market Dualism in Korea

Sungyup Chung⋅Sunyoung Jung

14 북한 이중경제 사회계정행렬 추정을 통한 비공식부문 분석

최지영

제2016 -15 Divergent EME Responses to Global and Domestic Monetary Policy Shocks

Woon Gyu Choi⋅Byongju Lee⋅ Taesu Kang⋅Geun-Young Kim

16 Loan Rate Differences across Financial Sectors: A Mechanism Design Approach

Byoung-Ki Kim⋅Jun Gyu Min

17 근로자의 고용형태가 임금 및 소득 분포에 미치는 영향

최충⋅정성엽

18 Endogeneity of Inflation Target Soyoung Kim ⋅Geunhyung Yim

19 Who Are the First Users of a Newly-Emerging International Currency? A Demand-Side Study of Chinese Renminbi Internationalization

Hyoung-kyu Chey⋅Geun-Young Kim⋅Dong Hyun Lee

20 기업 취약성 지수 개발 및 기업 부실화에 대한 영향 분석

최영준

21 US Interest Rate Policy Spillover and International Capital Flow: Evidence from Korea

Jieun Lee⋅Jung-Min Kim⋅Jong Kook Shin

제2017 -1 가계부채가 소비와 경제성장에 미치는 영향- 유량효과와 저량효과 분석 -

강종구

2 Which Monetary Shocks Matter in Small Open Economies? Evidence from SVARs

Jongrim Ha⋅Inhwan So

3 FTA의 물가 안정화 효과 분석 곽노선⋅임호성

4 The Effect of Labor Market Polarization on the College Students’ Employment

Sungyup Chung

5 국내 자영업의 폐업률 결정요인 분석 남윤미

6 차주별 패널자료를 이용한 주택담보대출의 연체요인에 대한 연구

정호성

7 국면전환 확산과정모형을 이용한 콜금리행태 분석

최승문⋅김병국

제2017 -8 Behavioral Aspects of Household Portfolio Choice: Effects of Loss Aversion on Life Insurance Uptake and Savings

In Do Hwang

9 신용공급 충격이 재화별 소비에 미치는 영향 김광환⋅최석기

10 유가가 손익분기인플레이션에 미치는 영향 김진용⋅김준철⋅임형준

11 인구구조변화가 인플레이션의 장기 추세에 미치는 영향

강환구

12 종합적 상환여건을 반영한 과다부채 가계의 리스크 요인 분석

이동진⋅한진현

13 Crowding out in a Dual Currency Regime? Digital versus Fiat Currency

KiHoon Hong⋅Kyounghoon Park⋅Jongmin Yu

14 Improving Forecast Accuracy of Financial Vulnerability: Partial Least Squares Factor Model Approach

Hyeongwoo Kim⋅Kyunghwan Ko

15 Which Type of Trust Matters?:Interpersonal vs. Institutional vs. Political Trust

In Do Hwang

16 기업특성에 따른 연령별 고용행태 분석 이상욱⋅권철우⋅남윤미

17 Equity Market Globalization and Portfolio Rebalancing

Kyungkeun Kim⋅Dongwon Lee

18 The Effect of Market Volatility on Liquidity and Stock Returns in the Korean Stock Market

Jieun Lee⋅KeeH.Chung

19 Using Cheap Talk to Polarize or Unify a Group of Decision Makers

Daeyoung Jeong

20 패스트트랙 기업회생절차가 법정관리 기업의 이자보상비율에 미친 영향

최영준

21 인구고령화가 경제성장에 미치는 영향 안병권⋅김기호⋅육승환

22 고령화에 대응한 인구대책: OECD사례를 중심으로

김진일⋅박경훈

제2017 -23 인구구조변화와 경상수지 김경근⋅김소영

24 통일과 고령화 최지영

25 인구고령화가 주택시장에 미치는 영향 오강현⋅김솔⋅윤재준⋅안상기⋅권동휘

26 고령화가 대외투자에 미치는 영향 임진수⋅김영래

27 인구고령화가 가계의 자산 및 부채에 미치는 영향

조세형⋅이용민⋅김정훈

28 인구고령화에 따른 우리나라 산업구조 변화

강종구

29 인구구조 변화와 재정 송호신⋅허준영

30 인구고령화가 노동수급에 미치는 영향 이철희⋅이지은

31 인구 고령화가 금융산업에 미치는 영향 윤경수⋅차재훈⋅박소희⋅강선영

32 금리와 은행 수익성 간의 관계 분석 한재준⋅소인환

33 Bank Globalization and Monetary Policy Transmission in Small Open Economies

Inhwan So

34 기존 경영자 관리인(DIP) 제도의 회생기업 경영성과에 대한 영향

최영준

35 Transmission of Monetary Policy in Times of High Household Debt

Youngju Kim⋅Hyunjoon Lim

제2018 -1 4차 산업혁명과 한국의 혁신역량: 특허자료를 이용한 국가‧기술별 비교 분석, 1976-2015

이지홍⋅임현경⋅정대영

2 What Drives the Stock Market Comovements between Korea and China, Japan and the US?

Jinsoo Lee⋅Bok-Keun Yu

3 Who Improves or Worsens Liquidity in the Korean Treasury Bond Market?

Jieun Lee

제2018 -4 Establishment Size and Wage Inequality: The Roles of Performance Pay and Rent Sharing

Sang-yoon Song

5 가계대출 부도요인 및 금융업권별 금융취약성: 자영업 차주를 중심으로

정호성

6 직업훈련이 청년취업률 제고에 미치는 영향

최충⋅김남주⋅최광성

7 재고투자와 경기변동에 대한 동학적 분석 서병선⋅장근호

8 Rare Disasters and Exchange Rates: An Empirical Investigation of South Korean Exchange Rates under Tension between the Two Koreas

Cheolbeom Park⋅Suyeon Park

9 통화정책과 기업 설비투자 - 자산가격경로와 대차대조표경로 분석 -

박상준⋅육승환

10 Upgrading Product Quality:The Impact of Tariffs and Standards

Jihyun Eum

11 북한이탈주민의 신용행태에 관한 연구 정승호⋅민병기⋅김주원

12 Uncertainty Shocks and Asymmetric Dynamics in Korea: A Nonlinear Approach

Kevin Larcher⋅Jaebeom Kim⋅Youngju Kim

13 북한경제의 대외개방에 따른 경제적 후생 변화 분석

정혁⋅최창용⋅최지영

14 Central Bank Reputation and Inflation-Unemployment Performance: Empirical Evidence from an Executive Survey of 62 Countries

In Do Hwang

15 Reserve Accumulation and Bank Lending: Evidence from Korea

Youngjin Yun

16 The Banks' Swansong: Banking and the Financial Markets under Asymmetric Information

Jungu Yang

제2018 -17 E-money: Legal Restrictions Theory and Monetary Policy

Ohik Kwon⋅Jaevin Park

18 글로벌 금융위기 전․후 외국인의 채권투자 결정요인 변화 분석: 한국의 사례

유복근

19 설비자본재 기술진보가 근로유형별 임금 및 고용에 미치는 영향

김남주

20 Fixed-Rate Loans and the Effectiveness of Monetary Policy

Sung Ho Park

21 Leverage, Hand-to-Mouth Households, and MPC Heterogeneity

Sang-yoon Song

22 선진국 수입수요가 우리나라 수출에 미치는 영향

최문정⋅김경근

23 Cross-Border Bank Flows through Foreign Branches: Evidence from Korea

Youngjin Yun

cross-border bank flows through foreign branches: evidence...

Documents