the effect of acquired company ebitda on the deal value
TRANSCRIPT
The effect of acquired company EBITDA on the deal value within M&A context
A study on the Pharmaceutical sector
BACHELOR THESIS WITHIN: Economics
NUMBER OF CREDITS: 15 ECTS PROGRAMME OF STUDY: International Economics AUTHOR: Constantin Copãceanu, Armand-Valeriu Perianu JÖNKÖPING May 2019
i
Bachelor Thesis in Economics
Title: The effect of acquired company EBITDA on the deal value within M&A context
Authors: Constantin Copãceanu, Armand-Valeriu Perianu
Tutor: Michael Olsson
Date: 2019-05-20
Key terms: Merger, Acquisition, Pharmaceutical, EBITDA multiple, Deal value, Synergy
Abstract
This thesis examines the impact of the valuation multiple ‘earnings before interest, taxes,
depreciation and amortization’ (EBITDA) on the ‘merger and acquisition’ (M&A) activity and
deal value. For small firms, mergers are primarily an exit strategy for firms in financial trouble,
as indicated by few marketed products and low cash-sales ratios. Meanwhile, mergers and
acquisitions for large drug makers is a way to leverage their sales networks and benefit from
monopolies from patents. The paper analyses the impact size of EBITDA, assuming it is
positive and smaller than 8% of the deal value. This thesis examines 46 cases within European
Union for a period of 10 years between 2009 to 2018. The conclusion reached is that EBITDA
valuation multiple has a significant negative impact on the purchase price but with little effect
in the pharmaceutical industry as the focus is always put on the operational synergies that the
target’s assets can bring compared to its earnings and sale prowess.
ii
Table of Contents
1. Introduction ..................................................................................... 1
2. Literature Review ........................................................................... 5
2.1. Hypothesis formulation ....................................................................................... 8
2.2. Theories behind chosen control variables ........................................................... 9
3. Data ................................................................................................ 12
3.1. Variables ........................................................................................................... 13
3.1.1. Deal Value (DV) ............................................................................................... 14
3.1.2. EBITDA ............................................................................................................ 14
3.1.3. Size .................................................................................................................... 14
3.1.4. Hirschman-Herfindahl Index (HI) .................................................................... 14
3.1.5. STOXX® Europe TMI Pharmaceuticals Index (SI) ......................................... 15
3.1.6. Dummy variables .............................................................................................. 16
4. Method ........................................................................................... 18
5. Results ............................................................................................ 20
6. Conclusion ..................................................................................... 22
7. Reference list ................................................................................. 25
7.1. Internet sources ................................................................................................. 30
7.2. Databases........................................................................................................... 34
8. Appendix ........................................................................................ 36
8.1. Legend ............................................................................................................... 36
8.2. Extended calculations........................................................................................ 36
8.3. Extended regression equation ........................................................................... 36
8.4. Correlation Matrix ............................................................................................. 37
8.5. Regression diagnostics ...................................................................................... 37
8.6. Key stats of chosen model................................................................................. 40
8.7. Choice of model ............................................................................................... 40
8.8. Complementary figures ..................................................................................... 41
8.9. Other insights .................................................................................................... 41
Table of Figures
Figure 1: Scatter plot of residuals ................................................................................ ...38
Figure 2: Histogram of residuals ..................................................................................... 38 Figure 3: Path of Hirschman-Herfindahl Index and the STOXX Index ......................... 41
1
1. Introduction
Over the past decade established pharmaceutical companies have not been successful at
the costly task of finding new drugs. According to a report published by the accountancy
firm Deloitte in 2018, the average return on research and development (R&D) at 12 of
the world’s largest pharmaceutical companies fell to just 1.9% in 2018, the lowest level
in almost a decade, as the cost of discovering new drugs has risen sharply. This is well
below the cost of capital, the rate at which companies can borrow money.
The concentration in the pharmaceutical industry has increased rapidly over the past 20
years. The total value of mergers and acquisitions activities in pharmaceutical industry
has been over $500 billion, because of which the top 10 pharmaceutical firm’s market
shares went up from 20% in 1985 to almost 50% by the end of 2002 (Danzon, Nicholson,
& Pereira, 2005). For the pharmaceutical industry, M&A’s are important not just in their
profit enhancing abilities, but also for their survival in the highly competitive
pharmaceutical industry. Few unique characteristics of the pharmaceutical industry define
the meaning of mergers and acquisitions in this field in a slightly different manner when
compared to the other industries, as pharmaceutical companies rely on advanced
biological knowledge, testing and regulatory approval. The single most important driver
for changes in the pharmaceutical industry is the ever-increasing cost of drug
development combined with the time limit imposed by the patent duration. Most
companies can no longer afford to carry out R&D to find innovative compounds (Mishra,
2018).
Because investments in R&D have been so fruitless, many large drugmakers have
relatively few remedies in the pipeline to drive sales once patents run out on existing
drugs (Deloitte, 2018). With their portfolios being endangered, large pharmaceutical
firms are further looking to buy other companies as a measure of acquiring new expertise
and diversification. The established sales networks of the bigger firms grant them certain
advantages over smaller firms within the industry, hence developing benefits from
economies of scope (Berk & DeMarzo, 2017). Patents grant a monopoly over the sales of
a new drug to its developer or owner for a certain period, which gives firms a limited time
during which they can earn monopoly profits. But smaller companies cannot accelerate
2
sales as quickly as a larger firm with established distribution networks, which can earn
more revenue more quickly from a new therapy. This allows them to lure target firms
with deals that are tempting to both sides, despite the purchaser paying a premium on top
of the target’s market price (The Economist, 2019).
The measurement of M&A success involves the acquirer company profits (Mishra, 2018).
They are considering the bid premium, which is the price that the acquirer firm pays at
the time of the purchase proposal, to the target firm’s shareholder over and above the
market price of the target firm’s shares. Furthermore, R&D investment is a crucial factor
because firms with high research expenses may be interested in M&A, as they tend to
exhaust large sums of money in developing in-house technology. On the other hand, firms
may instead of spending a large sum on R&D, invest in expanding via merger or
acquisition to increase technological superiority by inorganic growth means (Vyas,
Narayanan & Ramanathan, 2012). This phenomenon is traditionally known as the ‘make
or buy’ strategy (Miyazaki, 2009).
As most of the papers in the field of M&A, such as the ones in our literature review, are
trying to measure success by focusing on acquirer return post takeover activity, we
consider that the analysis should also focus on the acquisition price or the estimation of
deal value. Berk and DeMarzo (2017) categorizes the company valuation in two
categories: by comparing the target with a similar company or by valuating the future
cash flows that the merger would generate in the future. The first approach makes it very
difficult to forecast operational improvements and other synergies, so the second category
is used by specialists in M&A valuation.
There are four commonly accepted valuation methods that should be considered when
valuing a pharmaceutical company (Fulcrum, 2019):
1. Asset-based valuation calculates a business’s equity value as the fair market value of
a company’s assets less the fair market value of its liabilities. This approach is rarely
used for a pharmaceutical company because its value is more closely related to
intangible assets, R&D expenditure, and cash flows.
3
2. Income approach (via capitalization of earnings): This method is most applicable to
companies that face predictable and constant growth in earnings and have an
established record of operations. The business value under this method is equal to the
cash flow projection for the next year divided by a capitalization rate (i.e. the
appropriate discount rate less the predicted growth rate).
3. Income approach (via discounted cash flow): The value of equity utilizing this method
is equal to the present value of free cash flows available to common equity holders
over the life of the business. This method works well for both established companies
with low growth rates and new companies with higher rates of growth, requiring to
forecast future cash flows to aid the investment decision.
4. Market approach: This method utilizes market indications of value based on metrics
from guideline publicly traded pharmaceutical companies or privately held
businesses. The financial metrics of public companies or those of private transactions,
such as Price-Earnings ratio (P/E), Price-to-Sales ratio (P/S), and Enterprise Value
over Earnings before Interest, Taxes, Depreciation and Amortization (EV/EBITDA),
can be used to generate valuation multiples that are then used to calculate business
value.
EBITDA is the most common valuation multiple because it can give an accurate overview
even when comparing firms with different amounts of leverage and also valuates the
cashflows accurately (Berk & DeMarzo, 2017). Still, one should be aware that this
measurement is not regulated by the Generally Accepted Accounting Principles (GAAP),
so slight differences in calculation might differ from company to company.
Specialists agree that EBITDA is highly important in valuation of a company, but there
is no consensus on how much of the acquisition price is determined by this multiple as
it’s not accounting for indirect elements of comparative advantage as high management
efficiency, patents and technology (Berk & DeMarzo, 2017). From the limitations
considered, stems our theory that the commonly used market value approach does not
entirely apply to the pharmaceutical industry. Therefore, we will test this theory on a
sample of 46 M&A deals in a timeframe of 10 years within the European pharmaceutical
and biotechnology industry.
4
Considering the R&D component that drives the M&A in the pharmaceutical industry
and the lack of inclusion of such factor in the traditional firm valuation strategies, we
raise two main questions that we will attempt to answer in this thesis within the following
sections: literature review on the topic, data presentation, then a discussion on the method
of analysis with the final results of the study and other considerations.
5
2. Literature Review
Mergers and acquisitions are defined as a combination of companies. When two
companies combine together to form one company, it is termed as ‘merger’ of companies.
While ‘acquisitions’ are when one company is taken over by another company. There are
four types of M&A activity:
1. Horizontal Mergers: when one company merges or takes over another company
that has similar products and services, which means that both the companies are
in the same industry;
2. Vertical Mergers: a combination of two companies that are in the same business
of producing the same goods and services, but the only difference is the stage of
production at which they are operating are different;
3. Concentric Mergers: are between firms that serve the same customers in a
particular industry, but the products and services offered are different;
4. Conglomerate Mergers: when two companies that operate in a completely
different industry combine their business together to form a new company (CFA
Institute, 2019; Evans, 2000).
The motives to engage in a takeover are diverse. Firms can merge with rivals in order to
attain monopolist status and increase profits by limiting competition, or to obtain
efficiency gains via elimination of inefficient management and employees. Also, one
company can enjoy economies of scale or savings by increasing production volume,
something that is not generally available to a small company. Another reason can be the
expertise in a general field. One firm could avoid dealing with shortage of skilled labor
or dealing with an unfamiliar new technology via merger and acquisition, instead of
investing resources into internal R&D effort that might as well fail (Berk & DeMarzo,
2017).
Economies of scope can provide gains from synergistic operations as production with
distribution or pairing of complementary goods. This goes hand in hand with the motive
of diversification, that can be seen from two different angles: portfolio development and
risk reduction, or increase of debt capacity and liquidity (Berk & DeMarzo, 2017). In the
pharmaceutical industry, the motives of portfolio diversification and expertise are used
6
when announcing the M&A initiative, both being not accounted in the EBITDA valuation
multiple thus potentially leading to assessment errors.
The empirical evidences in support of or against efficiency argument of M&A are
provided by several studies. Ravenscraft and Scherer (1987) emphasized that stock
market values mergers as positive events but Seth (1990) analyzed that financial synergies
do not create any value in related and unrelated mergers. Black (1989) postulates that
managers are highly optimistic about targets and they overpay for targets as their interest
differ from that of stockholders. Ravenscraft and Scherer (1987) also supported the
argument of manager’s empire building as a motive for acquiring new businesses, while
Roll (1986) considered the managerial over optimism in hubris hypothesis of mergers and
acquisition, leading to overestimation of synergies, agency problems or winner’s curse
(Diaz and Gutierrez, 2013). In many cases buyers have justified deals by citing
questionable synergies, but in the heat of a takeover battle, ego often played a greater role
than even such dubious logic (The Economist, 1999).
Healy, Palepu and Ruback (1997) also provide evidence for higher acquirer returns when
the acquirer company management owns large stakes in the target firm. On the contrary,
when the acquirer company management does not own enough target equity beforehand,
it signifies agency problems in the management that could lead the acquirer firm
shareholders to believe that the management prefers growth strategies including value-
destroying mergers over shareholder value maximization and this would in such a case
lead to lower acquirer returns post M&A announcement. For our study, we consider only
the cases where the bidder had no ownership stake prior to the merger or acquisition as
the decision to engage in M&A from an existing ownership share might lean towards
reasonings which do not represent the object of our thesis.
Hagedoorn and Duysters (2002) take into account another important aspect related to
M&A’s which is the strategic and organizational fit between the parties involved and
conclude that M&A’s are generally profitable for companies in a high-tech environment.
An alternative study by Rawani and Kettani (2010), suggests that there is a positive
market reaction after merger or acquisition announcement for both the target and the
bidder firms in the pharmaceutical industry.
7
Antoniou, Arbour and Zhao (2008) find evidence for the fact that the more benefits
acquirers expect to earn from a M&A, the more willing they will be to pay a higher price
for the acquisition. In such a case, the higher the potential synergies expected, the higher
the premium, and as a result the shareholders of the acquirer firm perceive it as a good
sign thereby increasing the abnormal stock returns. This phenomenon is traditionally
known as the ‘Synergy hypothesis’. Sirower (1997) on the other hand emphasizes more
on the flipside of the synergy hypothesis thereby explaining the reason behind negative
relationship between premium and abnormal returns for the acquirer post an M&A deal.
When the acquiring company pays a premium to the target firm, which is higher than the
market expected profits, it leads to a decline in the stock returns of the acquiring company.
Sirower (1997) terms this as the ‘Overpayment hypothesis’. Diaz, Azofra and Gutierrez
(2009) prove that the premium that does not exceed 40% becomes a sign of future synergy
and anything over will have a negative influence on the acquirer returns as result of
overpayment, thus connecting both Antoniou’s and Sirrower’s theories. This pair will
help towards formulating the interpretation of our results.
Shepherd (1986) suggests that size is directly correlated with market power which could
develop inefficiencies causing poor performance; therefore, size could affect in both
positive and negative direction concerning firm’s decision to grow. Duflos and Pfister
(2008) studied the technological determinants of acquisitions in pharmaceutical industry
and argue that motives for acquisitions would differ in relation to acquirer’s and target’s
size. Large firms tend to receive negative synergy even by paying larger acquisition
premium which is consistent with managerial hubris hypothesis. Also, on the topic of
larger firms, Moeller, Schlingemann and Stulz (2004) augment that the bigger the
acquirer is, the heftier the premium offered to the target shareholders in order to keep his
position consolidated. Another example of size interaction is given by Loderer and Martin
(1990) by concluding that acquiring firms experience more losses when buying large
target firms because they generally end up overpaying in such cases.
According to Danzon, Epstein and Nicholson (2007) among large firms, they find that
mergers are a response to excess capacity due to anticipated patent expirations and gaps
in a company’s product pipeline. For small firms, mergers are primarily an exit strategy
8
and sometimes a solution for firms in financial trouble. Danzon et al. (2007) also suggest
that excess capacity may be a reason for M&A activity, with the purpose to restructure
asset bases in industries that experience shocks due to technological change or
deregulation. In the pharmaceutical industry, this capacity-adjustment motive for merging
occurs because of the patent-driven nature of a research-based pharmaceutical company’s
revenues. Essentially, a fully-integrated pharmaceutical firm has two sets of production
activities. The first is research and development to develop new drugs and perform the
clinical trials that are required for regulatory approval. R&D investment is substantial but
by itself generates no revenue. The second set of activities is production, marketing and
sales, for which approved compounds obtained from internal R&D, in-licensing or
acquisition, are an essential input. Patent protection on new drugs on average lasts for
roughly 12 years after market approval. Once the patent expires, generic competitors
usually enter and rapidly erode the originator firm’s sales.
2.1. Hypothesis formulation
In regards to the deal value, Koeplin, Sarin and Shapiro (2000) argue that the valuation
of a target company is relevant to determine the price to be paid for its acquisition. Based
on this value, the acquirer will acquire at a premium (i.e. price for a target firm bigger
than average price paid for comparable companies) or discount (price for a target firm
smaller than average price paid for comparable companies). As Hassan, Patro, Tuckman
and Wang (2007) argue that financial synergies are not driving merger initiative in the
pharmaceutical industry, we would expect traditional valuation multiples like EBITDA
to have minimal influence on the deal value, because it is not an indicator of operational
synergies between the firms entering the M&A. This translates into our first null
hypothesis which aims to reject or fail to reject our expectation.
For the maximum threshold that we would like to test against, we attempt to quantify1
this upper limit for the coefficient of interest in our first hypothesis. So, we use the
arithmetic mean obtained from the estimated EBITDA multiples of the ‘Biotechnology
& Medical Research’ and the ‘Pharmaceuticals’ industries, as taken from the Equidam
(2018) platform. By dividing 100 with the resulted average multiple we obtain an
1 See Appendix 8.2 for full calculation.
9
approximate value of 8 which represents the maximum impact (in percentage terms) that
EBITDA can have on the M&A deal value, according to our theory.
Considering that EBITDA is a measurement of financial health based on earnings and
sale prowess, an increase in this indicator eventually leads to positive free-cash-flow
(FCF) which usually leads to positive net-present value (NPV), condition necessary to
engage into merger and acquisition (Berk & DeMarzo, 2017). Hence our second
hypothesis shapes up, stating that an increase in EBITDA should positively impact the
acquisition price.
All above leads to our proposal to analyze the effect of EBITDA upon the acquisition
price and hence the following hypotheses:
Hypothesis 1: The EBITDA of the target firm does not impact by more than 8% the
deal value of a M&A in the pharmaceutical industry (β1<8%).
Hypothesis 2: The EBITDA of the target firm does positively impact the deal value of
a M&A in the pharmaceutical industry (β1>0).
2.2. Theories behind chosen control variables
Following the hypotheses stated, the literature that contributes to the formation of our
model is presented. Pecking order theory (Myers & Maijluf, 1984) states that companies
prioritize their sources of financing by using internal funds first, then issue debt, and
finally raise equity as a last resort. This can be explained by the information asymmetry
between the internal and the external party of the company. Debt or equities must be
issued under a high discount rate which is lower than the standard refinancing rate due to
adverse selection problem. This financing process may limit a company from maximizing
their corporate value since the financing possibility or period is undecided. The
uncertainty will force them to choose another investment project other than the best
investment project with positive NPV. The pecking order theory explains that businesses
prefer internal financing in order to maximize corporate value.
10
Previous research examining the role of the method of payment in explaining
announcement returns to bidding firms in acquisitions, finds significant differences
between cash and stock transactions. Brown and Ryngaert (1991) report that returns to
bidders tend to be negative and significant in stock acquisitions, and slightly positive
though not significant in cash acquisitions. This empirical evidence of larger returns in
cash offers when compared to stock exchange offers implies that choice of exchange
medium has economic significance. One argument is based on the premise that the market
and management do not share the same information set and that the resulting information
asymmetry affects the choice of exchange medium. Georgen and Renneboog (2004) have
concluded in their study that paying the acquisition amount entirely in cash results in
negative returns for the acquirer. Jensen (1986) also argued that managers undertake
M&A activity to waste cash in order to avoid shareholders’ value maximization. This
allows them to increase their control on the firm in comparison to shareholders; therefore,
he further states that all mergers and acquisitions do not occur with the motive of
promoting efficiency.
In the presence of information asymmetry, management’s choice of financing conveys
information about the firm’s true value to the market. Market participants interpret a
stock-financed acquisition as a negative (positive for cash-financed) signal of the value
of the acquiring firm (Franks, Harris & Mayer, 1988). Acquisitions involve two primary
effects: the effect of capital investments, which may or may not be positive, and the effect
of financing. As studies consistently find that announcement returns to bidding firms
making cash offers are higher than those making stock offers, two schools of thought seek
to explain this phenomenon based on these two effects. The first school of thought is that
cash is likely to be used for positive NPV acquisitions, while the second school of thought
is that paying out funds (instead of using them—perhaps wastefully—elsewhere) or
issuing debt, benefits shareholders (Brown and Ryngaert, 1991).
Another major factor influencing the acquirer firm’s abnormal returns post an M&A
announcement is the geographical orientation of the companies involved in it. Both
domestic and cross-border M&A creates value for the parties involved, but a study created
by Bassen, Schiereck and Wübben (2010) by examining German acquisitions in U.S.A.,
showed positive effect of cross-border M&A for the acquiring company. This claim is
11
backed by Deng and Yang (2015) that find that cross-border M&A’s experience higher
post announcement returns compared to domestic. They describe that this is due to the
shareholders valuing more the market seeking and resource seeking objectives of acquirer
firms, who gain access to new markets, over the risk.
Older firms have a benefit of learning and can therefore, enjoy superior performance. But
at the same time, younger firms are far from inertia and prevented from bureaucratic
practices, therefore more flexible and responsive to adjust changing economic
circumstances (Marshall, 1920). Duflos and Pfister (2008) state in their results that
acquiring and target firms in pharmaceutical industry are younger than sample average.
The management of young firms wants to grow faster and the possibility of merging or
being acquired, provides this opportunity to them.
It has been found that an acquisition of a hostile nature or that by a tender offer generates
higher returns compared to friendly mergers or acquisitions (Gregory, 1997) while R&D
intensity can have both negative and positive impact on firm’s probability to undertake
M&A decision (Mishra, 2018).
12
3. Data
Our study will focus on the European market (EU-28) of which both acquirer and target
are part of. We did choose this geographical area as Mishra (2018) discovered in this
research that market capitalization has a significant impact on M&A bid premiums
compared to other markets.
Our sample consists at first of 69 observations after the primary filtering, but after
handpicking we remain only with 46 as some transactions were either ownership stake
increases from a value over 50% or acquisition of one R&D facility or product from the
target firm. An acquisition from a controlling position in the target firm and partial
purchase of portfolio imply different rationale than what we want to test, rendering these
cases not useful to our interest. Our main source is Zephyr for the general details related
to the M&A activity and Amadeus in conjunction with Orbis and DataStream for the
specific accounting details and market indexes. Any other missing data was manually
gathered and transformed from the annual reports of the companies involved in the
transaction, as well as from specialized associations as the European Federation of
Pharmaceutical Industries and Associations (EFPIA).
The deals that we are considering need to be “completed-confirmed” which means that
the parties involved in M&A have been confirmed as fully integrated within each other.
Also, the acquiror must get from 0% ownership stake at least a majority share of the target
as M&A in our context implies a change of majority ownership. Both bidder and target
must belong to the UK SIC 2007 industry classification of “Manufacture of basic
pharmaceutical products and preparations” as it represents the industry of study for our
theory. This industry was chosen due to the frequency of M&A activity and its high
dependence on R&D.
The timeframe of the study is 31/12/2009 – 31/12/2018, mainly due to our data
restrictions. The maximum time length of available data is ten years and in order to
accommodate our model this interval was chosen. The selection of data for both bidder
and acquirer has been done from the year pre-M&A announcement, as our literature
review suggests that the scouting and number crunching from both parties is based on the
last available annual reports before the decision. The year 2009 was picked as the lower
13
bound so that we could gather company data from 2008 which is the maximum available
length for information on most databases we used. Also, circumstantially this starting date
considers the market maturation after the EU monetary union formation in 1999
(Andrews & Varrasi, 2015) and also will synchronize with the period after the 2008
financial crisis, where the M&A market started to recover (Institute for Mergers,
Acquisitions and Alliances – IMAA, 2019). The deal specifics as method of payment or
cross-border and market data as concentration and risk were taken from the year of
announcement.
We filtered for the methods of payment by selecting only cash and shares as we want to
test if it would have significant effect as according to previous literature (Yook, 2003).
In the same time, we disregard the debt-based methods like ‘leveraged buyout’ or
‘deferred payment’ as Myers’s Pecking Order theory mentions that the firm prioritizes
internal resources first and then debt or equity raise as method of payment.
The step results can be seen in Table 1 after the application of the above-mentioned filters,
with the purpose of aiding any further researcher in duplicating our results.
Table 1: Stepwise filtering results via Amadeus database
Search Strategy Filter Result Total Result
World Regions (EU enlarged 28): Acquiror and Target
384,415 384,415
Deal Type: Acquisition and Merger 700,101 214,796 Methods of Payment: Shares, Cash 954,924 34,296 Time Period: 31/12/2009 – 31/12/2018 (status: completed-confirmed)
767,063 240
UK SIC: 21- Manufacture of basic pharmaceutical products and pharmaceutical preparations (Acquiror and Target)
8,374 69
3.1. Variables
Our variables were chosen in accordance with the models and theories of our literature
review and will account for the most important factors that can influence the acquisition
price. All variables except the dummy variables, Hirschman-Herfindahl Index (HI) and
STOXX Index (SI) are denominated in Euro as currency.
14
3.1.1. Deal Value (DV)
The dependent variable ‘Deal Value’ represents the acquisition price within M&A. Also,
it can be denominated as the summation of target company value plus bid premium, thus
helping with the interpretation of our literature review that is specialized in analyzing the
effects on bid premium alone.
3.1.2. EBITDA
Kim and Ritter (1999) used several multiples for the valuation of initial public offering
matching companies, like P/E and enterprise value to sales, but EBITDA came out as the
most precise, thus becoming the one of the focus variables in our study. In terms of
company valuation measures, Chirico and Granata (2010) recommend using EBITDA
over Earnings Before Interest and Taxes (EBIT) as their study confirms Lie and Lie’s
(2002) theory that EBITDA is generally a more suitable measure because depreciation
expenses distort the information value of earnings. Another reason for using EBITDA in
our model over EBIT is that this measure is more suitable for mature industries of which
the pharmaceutical one belongs to (Fernández, 2001).
We do introduce in our model this valuation multiple for both target and acquirer with
the respective shortcuts ‘ET’ and ‘EA’, but the main relationship of study is between the
dependent variable DV and ET. As per our hypothesis’s we do expect that any increase
in the target’s EBITDA to impact positively but under 8% the dependent variable DV.
3.1.3. Size
Some researches measure size by total sales, but it is not the case in our situation. Instead
our literature review suggests we measure size by total assets, as within the category
intangible assets are included. It is an important distinction as within intangible assets,
patents and intellectual property (IP) are incorporated, factors that we suspect should have
a higher impact on the acquisition price than EBITDA.
We do introduce in our model those measurements for both target and acquirer with the
respective shortcuts ‘ST’ and ‘SA’.
3.1.4. Hirschman-Herfindahl Index (HI)
The performance of an industry varies with market structure which can be measured by
the degree of concentration. The more concentrated a market is, the bigger the suspicion
of an oligopolistic format exists; same line of thought is applicable for the other direction
15
where a market with little concentration can be considered a highly competitive one
(Carlton & Perloff, 2015). Therefore, to account for market concentration, we attempted
to calculate the Hirschman-Herfindahl Index (HI) for the previous year corresponding to
every transaction in our sample. This additional variable was not present in any previous
model of the literature review, so we include it to test if it would bring a significant
improvement to our model. HI is calculated by summing up the squared market shares
without the percentage sign, of every firm in the industry of study.
HI = s12 + s2
2 + s32 +…+sn
2 (1)
The market share is calculated as the firm’s total sales in relation to the total sales number
of the industry, that in our case is the pharmaceutical market of EU-28.
Due to data availability, certain assumptions had to be made in order to gain an
approximation:
1. The number of firms in the market is constant. In 2019, according to Amadeus
database, there are 6428 firms in EU-28 within the industry. As there is no
available data for every year in the interval of study, this number is assumed
constant.
2. Any other firm outside the top 14 in sales, have equal market share with the rest
of the companies in the fringe. As the data is not available for every company in
the industry, we assume the fringe splits equally the rest of the pie.
3. The total sales number generated by the firms in top 14 is taken as a whole. With
no data on how the sales for each company are split for each geographical region,
we take it as a whole, even if the summation surpasses than the total amount of
sales within EU-28.
3.1.5. STOXX® Europe TMI Pharmaceuticals Index (SI)
As Berk and DeMarzo (2017) summarize, in the world of investment market risk has a
big role in determining the optimal choice in relation with expected return, most notable
model of such sort being the Capital Asset Pricing Model (CAPM). For our case in the
role of approximating market risk, we choose the STOXX Total Market Pharmaceutical
Index as it is the most accurate measure we could find in relation to the characteristics of
our sample. This additional variable was also not present in any previous model of the
16
literature review. Considering that the decision for company merger or acquisition is one
of the biggest investment decisions that a management team can make, we were surprised
not to see such factor accounted for in the previous literature; therefore we decided to
include it to test if it would bring any significant improvement to our model. In the
calculation of this index, multiple factors such as liquidity, market capitalization
weighted-indices, price-weighted indices, currency and listing are included. This index
might also be biased towards the top 20 companies in the European pharmaceutical
market, but it is the best alternative to risk measurement we could apply in our model.
3.1.6. Dummy variables
Researchers of previous literature, like Franks, Harris and Mayer (1988) or Bassen,
Schiereck and Wübben (2010), have found significant differences between domestic and
cross-border deals and various methods of payment, elements that became auto-include
factors in further analysis within the field of M&A. Thus, we control for the deal type via
two dummy variables; one for deals effectuated fully via cash (Dc) and one for domestic
M&A (Dd), becoming value one if the condition is met and zero otherwise.
The third dummy variable used has the role to differentiate between targets that have
R&D operations or not. Generally, a more meaningful measurement is added instead, as
R&D intensity (Mishra, 2018) or R&D expenditure, but we use the contribution of
Mataigne, De Maeseneire and Luypaert (2018) that there is a significant impact on bid
premium between targets that run R&D activity or not.
In the table below (Table 2) it is shown the distribution of deals within our sample,
considering the conditions:
Table 2: Distribution of deals
Domestic M&A 54.35 % of total sample
Deal paid fully in cash 45.65 % of total sample
Target pursues R&D 86.95 % of total sample
It can be seen our dataset is balanced when accounting for the first two conditions, but
there is a huge majority when it comes to the third category, as expected considering the
sector. One would ask how a company in pharmaceutical industry cannot have R&D. As
the data for the R&D dummy was handpicked in most of the cases, we observed that most
of the companies in this category did manufacture the medicinal products with license
17
from the patent holder, making the prospect of extending the supply chain a good
purchase opportunity.
Table 3 presents the univariate statistics of the dependent, independent and control
variables. All the non-ratio and non-index numbers are denominated in Euro as currency.
Notably can be seen that the Hirschman-Herfindahl Index indicates that the industry has
been at most in the interval of medium-high market concentration which corresponds to
the period right after the 2008 financial crisis, further rising up with the passing of time
to values correlated with little competition. In parallel the STOXX Index follows a similar
trajectory within the time interval of our study2.
Table 3: Univariate Analysis
Mean Median Min Max S.D.
Dependent Var. DV 1.43e+09 9771195 213510 1.1e+10 2.9e+09
Independent Var. ET 1.26e+08 5899713 -2.2e+07 3.1e+09 4.7e+08
Control Variables
Deal Level Dc 0.457 0 0 1 0.504
Dd 0.543 1 0 1 0.504
Acquirer Level EA 1.21e+09 33466500 -2.2e+08 1.4e+10 3.1e+09
SA 5.41e+09 3.06e+08 7248163 5.1e+10 1.2e+10
Target Level ST 8.97e+08 47575946 637000 2.1e+10 3.2e+09
Dr&dT 0.870 1 0 1 0.341
Market Level HI 2971.63 2781.0 2079.1 4302.1 747.9
SI 517.33 525.1 258.9 704.2 139.9
Nr. of Obs. 46
2 See Appendix 8.8 Figure 3.
18
4. Method
The regression is built in a log-lin format as the variable of interest ET cannot be
logarithmic due to the negative values that it can take. Same restriction applies to EA.
Between the rest of the control variables and DV we suspect that the relation is not linear,
so by application of natural logarithm we smoothen out the potential huge value
differences from one case to another and we get closer to a classic linear regression model
(CLRM). Dummy variables are not logarithmic as they can take only two values. This
actual form was further backed3 by the testing in parallel of other regression forms: linear,
lin-log, with/without “HI” variable and with/without “SI” variable sequentially and
simultaneously. The Akaike Information Criterion, the coefficient significance and R2
were the ultimate factors in choosing our current regression form. This type of model is
consistent with a number of papers in our literature review, most notable ones being
Mishra (2018) and Mataigne et al. (2018), and it is shown in equation (2).
lnDVi=β0+β1ETi+β2EAi+β3lnSAi+β4lnSTi+β5Ddi+β6Dci+β7Dr&dTi+β8lnHIi+β9lnSIi (2)
Our aim is to apply Ordinary Least Squares (OLS) method which is a popular and
powerful tool for estimating coefficients, but for the before we run OLS certain
assumptions have to be met:
1. No autocorrelation; The Durbin Watson statistic of 1.20 does reflect a similar
result as the Breusch-Pagan-Godfrey LM test: an uncertain one. For 1% level of
significance there is no autocorrelation but for any other significance level over
3%, autocorrelation exists. As our data set is a random cross-sectional sample
and does not include time dependent factors, there is no prior reason to believe
that the errors from one observation influence the errors of another. If such
correlation is observed in a cross-sectional sample, it is called spatial
autocorrelation and does not pose a problem as long as the data has some logic
or economic interest (Gujarati & Porter, 2009).
3 See Appendix 8.7 Table 9.
19
2. No heteroskedasticity; The test for heteroskedasticity was done through White’s
test4 and Park’s test and no heteroskedasticity was found at any given standard
level of significance.
3. The number of observations must be greater than the numbers of parameters
estimated; Even though the sample consists of only 46 observations, if we deduct
the number of estimated parameters and intercept, we obtain 36 degrees of
freedom. This number is close to the lower threshold for statistical significance,
but still can provide valid results.
4. No multicollinearity; The testing for multicollinearity was conducted via
Variance Inflation Factors (VIF) which confirmed that there are no severe
problems in this model.
5. Residuals are normally distributed; the residuals present normality in their
distribution, confirmed via histogram analysis and Jarque-Bera test.
By respecting the assumptions presented above within our model, with the application of
the Gauss-Markov theorem which states “Given the assumption of CLRM, the least-
squares estimators, in the class of unbiased linear estimators, have minimum variance,
that is, they are BLUE” (Gujarati & Porter, 2009, page 72), we can assume that our
estimators are also belonging to the ‘best linear unbiased estimators’ (BLUE) category.
4 White’s test was effectuated without the cross terms as the sample size is too small.
20
5. Results
By conducting the Ordinary Least Squares method on our model, we obtain the estimates
present in Table 4 which will be the base to our takeaways. We can certainly notice
consistency with different conclusions from our literature review, especially when
referring to the control variables used. As most of the research papers in the field of M&A
has cumulative abnormal returns (CAR) as dependent variable in contrast to our choice,
we will adapt our results to the base theories of section two and then draw conclusions.
The focus of our study is not the impact measure on the returns or picking a winner from
the parties involved in the M&A, but to prove that traditional valuation multiples do not
entirely apply in a R&D intensive industry.
Table 4: Multivariate Analysis
Coeff. Std. Error T-Statistic P-Value
Independent Var. ET -7.2e-10 4.2e-10 -1.703 0.097
Deal Level Dc -0.113 0.388 -0.292 0.772
Dd -0.032 0.313 -0.104 0.918
Acquirer Level EA 5.8e-11 5.9e-11 0.978 0.334
SA 0.186 0.111 1.672 0.103
Target Level ST 0.936 0.119 7.846 0.000
Dr&dT 0.127 0.481 0.265 0.793
Market Level HI -0.082 1.017 -0.085 0.936
S.I 0.425 0.837 0.507 0.615
Intercept -3.863 5.954 -0.649 0.521
As we want to see if EBITDA is a proper valuation measurement in the context of
pharmaceutical industry, we added this indicator for both target and acquirer in our model.
On one hand the results show that for the acquirer the variable is both very little and
insignificant. But on the other hand, an 1% increase in the target’s EBITDA leads to a
reduction of 0.09% in the deal value, taken at sample average5. The result for the target
valuation multiple is significant at 90% confidence interval and consistent with Koeplin
et al. (2000) argument that the valuation of a target company is relevant to determine the
5 Considering the log-lin relationship between “Deal Value” and the EBITDA variables,
the elasticity interpretation is calculated as 1% change in EBITDA variable = (βEBITDA *
Sample Mean EBITDA) % change in “Deal Value”.
21
price to be paid for its acquisition, thus rejecting the second hypothesis and failing to
reject the first one.
Further it is reassured by the significant results that the size of the targeted firm does
matter. As the size in our model is measured by total assets, within the value are also
included the intangible components of interest like intellectual property and patents. One
percent increase in the assets of the target, leads to ~0.93% increase in the deal value.
Danzon et al. (2007) argues that the chance to restructure the asset base, is generally the
main motive to engage in M&A activity within pharmaceutical and biotechnology
industry. This theory in conjuncture with our results convinces us to conclude that the
management of the bidder pinpoints the target’s assets as the main value generator and
thus becoming the biggest determinant of the acquisition price.
With the failure to reject the first hypothesis, comes the conclusion that the EBITDA
valuation multiple has a significant impact on the purchase price but with little effect in
the pharmaceutical industry as the focus it’s always put on the operational synergies that
the target’s assets can bring compared to its earnings and sale prowess.
By rejecting the second hypothesis we are more able to pick a side according to our
analysis. Loderer and Martin (1990) have argued that firms that acquire large targets end
up overpaying due to the bigger losses recorded post-M&A. Instead, Black (1989)
considers that the overpayment is generated by the optimism of the management, side
which we lean more towards. The inversely proportionality in our result dismisses the
direction Loderer and Martin (1990) proposes (bigger target results into bigger
probability of overpayment) but not entirely their theory. Instead by adding Black’s
(1989) conclusion into our construction we can assume that rather the smaller the target
firm, the bigger the probability to overpay becomes, as the managers tend to be overly
optimistic over the potential synergy. This conclusion is aligned with the results of
Alexandridis, Fuller, Terhaar and Travlos (2013) that studied the relation between market
capitalization and acquisition premium and would also check-up partially with the
‘Synergy hypothesis’ proposed by Antoniou et al. (2007).
22
6. Conclusion
The previous literature when studying the world of mergers and acquisitions is often
inconsistent as the ultimate decisions of engagement are still made by the management
boards of both parties, thus bringing subjectiveness. Also, the models are very different
from researcher to researcher, especially those before 1997 when MacKinlay set the
methodology for an event study that we mostly use today. We assume this reason in
conjuncture with the difference in the datasets used, generated all the contrary views and
inconsistency.
When talking about valuation, it is to be noted that even though EBITDA is the most used
and widely considered appropriate measure for company valuation, it is not a GAAP
approved indicator, hence potentially altering the validity of results.
The biggest barrier towards a more compelling study was the availability of data. Such
restrictive filtering leads to a limited number of observations that in conjuncture with our
cross-section selection, raises certain problems that we have anticipated. Further research
might avoid significance and generalization issues just by increasing the timespan for
more probes. Upon this case and having enough degrees of freedom, one should also
ideally control for the year and acquirer country as external macroeconomic factors might
influence the results as Mataigne, De Maeseneire and Lupayert (2018) used in their
model.
A scholar that researches this topic further shall consider including a differentiation
between the type of offers that are done by the acquirer. According to Gregory (1997) a
hostile bid or a tender offer has a different impact on the returns compared to a friendly
offer, so this might as well impact the amount of money that parties agree upon M&A
decision.
Also, an important consideration should be given to the ‘information asymmetry’
component. Upon availability of data, one should consider the method Chae, Chung and
Yang (2009) used to account for this factor and to resolve the ‘size effect puzzle’,
especially if the object of study is leaning more towards analyzing returns than deal value.
23
As ‘size’ is one of the most significant variables in M&A analysis, this suggestion could
help towards an easier interpretation of the resulted effect.
As all bidding companies in our sample conducted research and development at the
moment and pre-M&A, future studies should include a measure of R&D intensity. Mishra
(2018) suggests that this characteristic is important to determine the operational
effectiveness of the company within pharmaceutical industry and should be calculated as
the ratio of R&D of the previous year divided by the sales of the previous year, both pre-
M&A.
One of the main differences that our paper brings in comparison to literature review is the
addition of competition and market risk measurements. Even though these variables
turned out insignificant, they still improved our model and one should consider adding
them with the awareness of their limitations. As the S&P 500 Index, the chosen STOXX
Index also brings a weighting issue in favour of the big companies within its composition.
Almost in the same manner our estimation of the Hirschman-Herfindahl Index is built.
For further researches, we consider that the levels of market capitalization should be
introduced for both target and acquirer, especially as in Mishra’s (2018) research this
indicator has been significant in the European sample. Also factors as company age and
type of offer as Duflos and Pfister (2008) and Gregory (1997) suggest, might bring useful
insights that could explain better the settlement of acquisition price from the behavioral
point of view.
From our readings, a lot of analysis is made with the rigid accounting indicators instead
of focusing on trying to quantify the intangibles, task that is difficult but more relevant
for R&D intensive industries. Instead of using the amount invested in R&D, a scholar
should attempt to quantify the target’s company entire product portfolio as Girotra,
Terwiesch and Ulrich (2007) suggest, at least within pharmaceutical industry. Every
product in every stage of development should be accounted for, weighted by the
probability rate of reaching the market from development phase. According to the authors
the necessary data for such estimation exists in Springer’s ADIS Insight International
database.
24
The inclusion of year dummies would be recommended as well, as it would account for
possible biases resulted from hidden fixed effects as Andrade and Stafford (2004)
recommend. This is accounted for in many studies of our literature review, notably in the
studies of Alexandridis et al. (2013) and Mataigne et al. (2018). In the absence of enough
observations, we could not include these variables in our thesis.
25
7. Reference list
Alexandridis, G., Fuller, K. P., Terhaar, L., & Travlos, N. G. (2013). Deal size, acquisition
premia and shareholder gains. Journal of Corporate Finance, 20, 1-13.
Andrade, G., & Stafford, E. (2004). Investigating the economic role of mergers. Journal
of corporate finance, 10(1), 1-36.
Andrews, C., & Varrasi, E. (2015). Credit Rating Changes and Post-M&A Firm Value:
Assessing the importance of credit ratings changes as a motive for successful M&A.
Antoniou, A., Arbour, P., & Zhao, H. (2008). How much is too much: are merger
premiums too high?. European Financial Management, 14(2), 268-287.
Bassen, A., Schiereck, D., & Wübben, B. (2010). M&A success of German acquisitions
in the US–evidence from capital market and survey data. Applied Financial
Economics, 20(7), 543-559.
Benninga, S., & Sarig, O. (1997). Corporate Finance: A Valuation Approach. McGraw-
Hill. New York.
Berk, J., & DeMarzo, P. (2017). Corporate Finance. Fourth Edition. Pearson Education
Limited. Essex, England. ISBN: 978-0-13-408327-8.
Black, B. (1989). Bidder Overpayment in Takeovers. Stanford Law Review, 41(3), 597-
660. doi:10.2307/1228881
Brown, D. T., & Ryngaert, M. D. (1991). The mode of acquisition in takeovers: Taxes
and asymmetric information. The Journal of Finance, 46(2), 653-669.
Carlton, D. W., & Perloff, J. M. (2015). Industrial Organization. Fourth Edition. Global
Edition. Pearson Education Limited. Essex, England. ISBN: 978-0-321-18023-0.
26
Chae, J., Chung, J. Y., & Yang, C. W. (2009). Does information asymmetry affect merger
and acquisitions in an emerging market. Financial Management Association
International (FMA), 1-46.
Danzon, P. M., Epstein, A., & Nicholson, S. (2007). Mergers and acquisitions in the
pharmaceutical and biotech industries. Managerial and Decision Economics, 28(4‐
5), 307-328.
Danzon, P. M., Nicholson, S., & Pereira, N. S. (2005). Productivity in pharmaceutical–
biotechnology R&D: the role of experience and alliances. Journal of health
economics, 24(2), 317-339.
Deng, P., & Yang, M. (2015). Cross-border mergers and acquisitions by emerging market
firms: A comparative investigation. International Business Review, 24(1), 157-172.
Dessyllas, P., & Hughes, A. (2005). R & D and Patenting Activity and the Propensity to
Acquire in High Technology Industries. ESRC Centre for Business Research,
University of Cambridge.
Diaz Diaz, B., Sanfilippo Azofra, S., & Gutierrez C. L., (2009). Determinants of
premiums paid in European banking mergers and acquisitions. International
Journal of Banking, Accounting and Finance, 1(4), 358-380.
Duflos, G., & Pfister, E. (2008). Searching for innovations? The technological
determinants of acquisitions in the pharmaceutical industry. CES Working Papers,
Centre d’Economie de la Sorbonne, Paris, series 2008.57.
Fernández, P. (2002). Valuation Methods and Shareholder Value Creation. San Diego:
Academic Press.
Franks, J. R., Harris, R. S., & Mayer, C. (1988). Means of payment in takeovers: Results
for the United Kingdom and the United States. In Corporate takeovers: Causes and
consequences (pp. 221-264). University of Chicago Press.
27
Girotra, K., Terwiesch, C., & Ulrich, K. T. (2007). Valuing R&D projects in a portfolio:
Evidence from the pharmaceutical industry. Management Science, 53(9), 1452-
1466.
Goergen, M., & Renneboog, L. (2004). Shareholder wealth effects of European domestic
and cross‐border takeover bids. European Financial Management, 10(1), 9-45.
Granata, D., & Chirico, F. (2010). Measures of value in acquisitions: family versus
nonfamily firms. Family Business Review, 23(4), 341-354.
Gregory, A. (1997). An examination of the long run performance of UK acquiring
firms. Journal of Business Finance & Accounting, 24(7‐8), 971-1002.
Gujarati, D. N., & Porter, D. C. (2009). Basic Econometrics. Fifth Edition. International
Edition. McGraw-Hill. Singapore. ISBN: 978-007-127625-2.
Hagedoorn, J., & Duysters, G. (2002). External sources of innovative capabilities: the
preferences for strategic alliances or mergers and acquisitions. Journal of
management studies, 39(2), 167-188.
Harrison, J. S., Hitt, M. A., Hoskisson, R. E., & Ireland, R. D. (1991). Synergies and post-
acquisition performance: Differences versus similarities in resource
allocations. Journal of management, 17(1), 173-190.
Hassan, M., Patro, D. K., Tuckman, H., & Wang, X. (2007). Do mergers and acquisitions
create shareholder wealth in the pharmaceutical industry?. International Journal of
Pharmaceutical and Healthcare Marketing, 1(1), 58-78.
Healy, P. M., Palepu, K. G., & Ruback, R. S. (1997). Which takeovers are profitable?
Strategic or financial. MIT Sloan Management Review, 38(4), 45.
28
Hitt, M. A., Harrison, J. S., & Ireland, R. D. (2001). Mergers & acquisitions: A guide to
creating value for stakeholders. Oxford University Press.
Hitt, M. A., Hoskisson, R. E., & Ireland, R. D. (1990). Mergers and acquisitions and
managerial commitment to innovation in M-form firms. Strategic Management
Journal, 29-47.
Jensen, M. C. (1986). Agency costs of free cash flow, corporate finance, and
takeovers. The American economic review, 76(2), 323-329.
Kim, M., & Ritter, J. R. (1999). Valuing IPOs. Journal of financial economics, 53(3),
409-437.
Koeplin, J., Sarin, A., & Shapiro, A. C. (2000). The private company discount. Journal
of Applied Corporate Finance, 12(4), 94-101.
Kumar, N., & Siddharthan, N. S. (1994). Technology, firm size and export behaviour in
developing countries: the case of Indian enterprises. The Journal of Development
Studies, 31(2), 289-309.
Lie, E., & Lie, H. J. (2002). Multiples used to estimate corporate value. Financial
Analysts Journal, 58(2), 44-54.
Loderer, C., & Martin, K. (1997). Executive stock ownership and performance tracking
faint traces. Journal of Financial economics, 45(2), 223-255.
Marshall, A. (1920). Industry and trade. Third edition.
Mataigne, V., De Maeseneire, W., & Luypaert, M. (2018). The interplay between target
firm R&D, acquirer debt financing and takeover premia. Applied Economics
Letters, 25(7), 451-455.
29
Mishra, P. (2018). Effect of M&A announcement on acquirer stock prices in the
Pharmaceutical sector and the role of bid premium.
Miyazaki, H. (2009). An analysis of the relation between R&D and M&A in high-tech
industries. Applied Economics Letters, 16(2), 199-201.
Moeller, S. B., Schlingemann, F. P., & Stulz, R. M. (2005). Wealth destruction on a
massive scale? A study of acquiring‐firm returns in the recent merger wave. The
journal of finance, 60(2), 757-782.
Myers, S. C. (1977). Determinants of corporate borrowing. Journal of financial
economics, 5(2), 147-175.
Myers, S. C., & Majluf, N. S. (1984). Corporate financing and investment decisions when
firms have information that investors do not have. Journal of financial
economics, 13(2), 187-221.
Rawani, A., & Kettani, H. (2010). Humanities, Historical and Social Sciences. Social
Sciences, 26, 28.
Ravenscraft, D. J., & Scherer, F. M. (1987). Life after takeover. The Journal of Industrial
Economics, 147-156.
Roll, R. (1986). The hubris hypothesis of corporate takeovers. Journal of business, 197-
216.
Seth, A. (1990). Value creation in acquisitions: A re‐examination of performance
issues. Strategic management journal, 11(2), 99-115.
Shepherd, W. G. (1986). Tobin's q and the Structure-performance Relationship:
Comment. The American Economic Review, 76(5), 1205-1210.
30
Sirower, M. L. (1997). The synergy trap: How companies lose the acquisition game.
Simon and Schuster.
Wong, A., Cheung, K. Y., & Mun, T. (2009). The effects of merger and acquisition
announcements on the security prices of bidding firms and target firms in
Asia. International journal of economics and finance, 1(2), 274-283.
Vyas, V., Narayanan, K., & Ramanathan, A. (2012). Determinants of mergers and
acquisitions in Indian pharmaceutical industry. Eurasian Journal of Business and
economics, 5(9), 79-102.
Yook, K. C. (2003). Larger return to cash acquisitions: Signaling effect or leverage
effect?. The Journal of Business, 76(3), 477-498.
7.1. Internet sources
Advanced Accelerator Applications SA, (2016). Annual report 2016. Retrieved on 18
February 2019, from
http://investorrelations.adacap.com/phoenix.zhtml?c=253874&p=irol-
reportsAnnual_pf
Alivira Animal Health Limited, (2019). Annual report 2014. Retrieved on 18 February
2019, from http://alivira.co/Investor-Relations/Press-Release/sequent-annual-
report-16-17
ALK Group, (2009). Annual report 2009. Retrieved on 18 February 2019, from
https://ir.alk.net/static-files/c94617ce-b54b-4145-958a-01475c445f55
ALK Group, (2010). Annual report 2010. Retrieved on 18 February 2019, from
https://ir.alk.net/static-files/1ddb49e2-25a7-4418-882a-5baec0ed2f1a
Allabolag.se, (2019). Galderma Nordic AB. Retrieved on 18 February 2019, from
https://www.allabolag.se/5565620928/bokslut
31
Alliance Pharma PLC, (2015). Acquisition of MacuVision. Retrieved on 17 February
2019, from https://alliancepharmaceuticals.com/en-
gb/news/2015/february/acquisition-of-macuvision
Bloomberg L.P., (2019). Company overview of Alerpharma S.A.. Retrieved on 17
February 2019, from
https://www.bloomberg.com/research/stocks/private/snapshot.asp?privcapid=303
163639
Boehringer Ingelheim GmbH, (2016). Annual report 2016. Retrieved on 17 February
2019, from https://annualreport.boehringer-ingelheim.com
Bremer Pharma GmbH, (2018). Annual financial statement. Retrieved on 18 February
2019, from https://zyduscadila.com/public/pdf/financial/subsidaries(2017-
2018)/Bremer%20Pharma%20GMBH_Mar18.pdf
CFA Institute, WallStreetMojo, (2019). Mergers and Acquisitions. Retrieved on 17 May
2019, from https://www.wallstreetmojo.com/mergers-and-
acquisitions/?fbclid=IwAR1Dxhpy3Z2UJv6I-rdW-
KmYL7NFMUrNvQ7wK8h44wOYtl7hqstA6Nfqs6g
Clingen Group PLC, (2019). Quantum archive. Retrieved on 17 February 2019, from
https://www.clinigengroup.com/investors/quantum-archive/?year=2015&load=1
Companies House, (2019). Winchpharma (Consumer Healthcare) Ltd.. Retrieved on 17
February 2019, from
https://beta.companieshouse.gov.uk/company/08192249/filing-history
Deloitte Touche Tohmatsu Ltd., (2018). 2018 Global life sciences outlook. Retrieved on
17 May 2019, from
https://www2.deloitte.com/content/dam/Deloitte/global/Documents/Life-
Sciences-Health-Care/gx-lshc-ls-outlook-2018.pdf
32
Dr. Reddy’s Laboratories Ltd., (2011). Annual report 2011. Retrieved on 17 February
2019, from https://www.drreddys.com/media/123516/annualreport2010-11.pdf
Dr. Reddy’s Laboratories Ltd., (2019). Retrieved on 17 February 2019, from
https://www.drreddys.com/investors/reports-and-filings/
Elan Corporation PLC, (2012). Annual report 2012. Retrieved on 17 February 2019, from
https://www.google.com/search?ei=co7cXMiqN9WGk74P_rSgoAw&q=elan+cor
poration+plc+annual+report+2012&oq=el&gs_l=psy-
ab.1.0.35i39l2j0l8.41196.41389..43541...0.0..0.138.235.1j1......0....1..gws-
wiz.......0i71j0i67.DoLuUOGchuc
Equidam, Trevisan L., (2018). EBITDA multiples by industry. Retrieved on 11 March
2019, from https://www.equidam.com/ebitda-multiples-trbc-
industries/?fbclid=IwAR3wfpdZsbHHNkAhMFxWrvbTRwtNMw6C4S1J-
9t71J9clx7fprz4pFsMKHI
Evans, M.H., (2000). Excellence in financial management. Retrieved on 18 February
2019, from https://exinfm.com/training/pdfiles/course07-1.pdf
Fernández, P., (2001). Valuation using multiples: How do analysts reach their
conclusions?. Retrieved on 24 February 2019, from
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=274972
Fulcrum Financial Inquiry LLP, (2019). Valuation Guide: Pharmaceuticals. Retrieved on
18 May 2019, from https://www.fulcrum.com/valuation-guide-
pharmaceuticals/?fbclid=IwAR3ngOp_Zmi3jqalAliRTaXpuFMP36V-qdx9c0-
3j17zHP5oWjg1eq9QMyA
Kyowa Hakko Kirin Co Ltd., (2012). Annual report 2012. Retrieved on 17 February 2019,
from https://ir.kyowa-kirin.com/en/library/annualreport.html
33
Les Echos Capital Finance, (2019). Laboratoires Ineldea croque 2,5 M Euros. Retrieved
on 17 February 2019, from https://capitalfinance.lesechos.fr/deals-m-a/mid-
cap/laboratoires-ineldea-croque-25-m-euros-125278
Made Simple Group Ltd., (2019). Gentrix Limited. Retrieved on 17 February 2019, from
https://www.companysearchesmadesimple.com/company/uk/03257633/genitrix-
limited/#balance-sheet
Made Simple Group Ltd., (2019). WPG Wholesale Trading Ltd.. Retrieved on 17
February 2019, from
https://www.companysearchesmadesimple.com/company/uk/07726842/wpg-
wholesale-trading-ltd/#financials
Merck KgaA, (2014). Annual report 2014. Retrieved on 17 February 2019, from
https://www.merckgroup.com/en/investors/reports-and-financials.html
Nepentes Capital Group, (2009). Consolidated financial statements. Retrieved on 17
February 2019, from https://www.bankier.pl/static/att/emitent/2010-
04/Skons_Spr_finansowe_GK_NEPENTES_za_2009_201004120000141859.pdf
Novartis International AG, (2016). Annual report 2016-17. Retrieved on 18 February
2019, from https://www.novartis.com/sites/www.novartis.com/files/novartis-
annual-report-2016-en-low-res.pdf
OctoPlus N.V., (2011). Annual report Octoplus 2011. Retrieved on 17 February 2019,
from
https://www.google.com/search?q=octoplus+annual+report+2011&oq=octo&aqs
=chrome.0.69i59j69i57j69i59j0l3.2104j0j4&sourceid=chrome&ie=UTF-8
Omega Pharma Corporate N.V., (2013). Annual report 2013. Retrieved on 17 February
2019, from https://www.omega-
pharma.com/sites/all/pdf/OP_2013_Annual_Report.pdf
34
Patheon N.V., (2016). Annual report 2016. Retrieved on 17 February 2019, from
http://ir.patheon.com/~/media/Files/P/Patheon-IR/documents/2016-annual-
report.pdf
Reckitt Benckiser Froup PLC, (2009). Annual report and financial statements 2009.
Retrieved on 17 February 2019, from https://www.rb.com/media/1450/annual-
report-and-accounts-2009.pdf
Sinclair IS Pharma PLC, (2011). Annual report and accounts 2011. Retrieved on 17
February 2019, from https://www.zonebourse.com/SINCLAIR-PHARMA-PLC-
4004466/pdf/280606/Sinclair%20Pharma%20PLC_Rapport-annuel.pdf
The Economist Newspaper Ltd., (2019). A.F.. Why pharmaceutical companies are on a
shopping binge. Retrieved on 17 May 2019, from https://www.economist.com/the-
economist-explains/2019/01/14/why-pharmaceutical-companies-are-on-a-
shopping-binge
The Economist Newspaper Ltd., (1999). After the deal. Retrieved on 17 May 2019, from
The Economist Newspaper Ltd., (2019). A.F.. Why pharmaceutical companies are
on a shopping binge. Retrieved on 17 May 2019, from
https://www.economist.com/the-economist-explains/2019/01/14/why-
pharmaceutical-companies-are-on-a-shopping-binge
Thomson Financial, Institute for Mergers, Acquisitions and Alliances (IMAA), (2019).
Retrieved on 18 February 2019, from https://imaa-institute.org/mergers-and-
acquisitions-statistics/
7.2. Databases
Bureau van Dijk. Amadeus.
Bureau van Dijk. Orbis.
Bureau van Dijk. Zephyr
Thomson Reuters. Datastream-Eikon.
35
36
8. Appendix
8.1. Legend
Table 5: Legend
Shortcut Meaning
AIC Akaike Information Criterion Coeff. Coefficient D.f. Degrees of freedom Dc Cash Dummy Variable Dd Domestic Dummy Variable Dr&dT Target R&D Dummy Variable DV Deal Value EA EBITDA Acquirer ET EBITDA Target H0 Null Hypothesis H1 Alternative Hypothesis HI Hirschman-Herfindahl Index Nr. Obs. Number of Observations S.D. Standard Deviation SA Size Acquirer SI Pharmaceutical STOXX Index Europe ST Size Target Std. Standard Var. Variable VIF Variance Inflation Factors
8.2. Extended calculations
Upper limit (%) = 100
′𝐵𝑖𝑜𝑡𝑒𝑐ℎ𝑛𝑜𝑙𝑜𝑔𝑦 & 𝑀𝑒𝑑𝑖𝑐𝑎𝑙 𝑅𝑒𝑠𝑒𝑎𝑟𝑐ℎ′ 𝑀𝑢𝑙𝑡𝑖𝑝𝑙𝑒+′𝑃ℎ𝑎𝑟𝑚𝑎𝑐𝑒𝑢𝑡𝑖𝑐𝑎𝑙𝑠′ 𝑀𝑢𝑙𝑡𝑖𝑝𝑙𝑒
2
% =
10013.44+12.93
2
= ~7.58 % = 8 % (with rounding)
8.3. Extended regression equation
lnDealValuei = β0 + β1 EBITDAtargeti + β2 EBITDAacquireri + β3 lnSizeAcquireri + β4
lnSizeTargeti + β5 Ddomestici + β6 Dcashi + β7 DtargetR&Di + β8 lnHIi + β9 lnStoxxIndexi (3)
37
8.4. Correlation Matrix
Table 6: Correlation Matrix
DV EA ET SA ST HI SI
DV 1 0.469 0.419 0.721 0.930 0.151 0.110
EA 0.469 1 0.347 0.488 0.437 0.082 0.089
ET 0.419 0.347 1 0.450 0.533 -0.084 0.043
SA 0.721 0.488 0.450 1 0.705 -0.135 -0.137
ST 0.930 0.437 0.533 0.705 1 0.139 0.096
HI 0.151 0.082 -0.084 -0.135 0.139 1 0.761
SI 0.110 0.089 0.043 -0.137 0.096 0.761 1
8.5. Regression diagnostics
1) Testing for multicollinearity via Variance Inflation Factors.
Table 7: Results of VIF testing
Centered VIF
Independent Var. ET 1.863
Deal Level Dc 1.786
Dd 1.161
Acquirer Level EA 1.536
SA 2.937
Target Level ST 3.785
Dr&dT 1.255
Market Level HI 2.903
SI 2.797
Intercept N/A
If centered VIF < 10, then there is no serious issue of multicollinearity.
38
2) Residual plot
Figure 1: Scatter plot of residuals – Shows the pattern the residuals form with their
values
Figure 2: Histogram of residuals – The distribution the residuals form while grouped
within intervals
3) Testing that residuals are normally distributed via Jarque-Bera Normality Test.
From Figure 2 we can suspect that our residuals are normally distributed but we attempt
to confirm this via the normality test.
H0: The residuals are normally distributed;
H1: The residuals are not normally distributed.
-3.000
-2.000
-1.000
0.000
1.000
2.000
3.000
0 10 20 30 40 50
Val
ue
of
resi
dual
Residual association with observation
Scatter plot of residuals
39
Jarque-Bera = 0.132 > 0.01; 0.05; 0.1 => Fail to reject H0.
4) Testing the residuals for heteroskedasticity via White’s Heteroskedasticity Test
without cross-terms and Park’s Heteroskedasticity Test.
White’s Heteroskedasticity Test with cross terms cannot be used for our regression as it
does not have enough observations to generate an output.
White’s Heteroskedasticity Test without cross-terms
H0: Homoskedasticity;
H1: Heteroskedasticity.
Obs*R-squared = 13.033
Prob. Chi-Square(9) = 0.1611 > 0.01; 0.05; 0.1. => Fail to reject H0.
Park’s Heteroskedasticity Test
H0: Homoskedasticity;
H1: Heteroskedasticity.
“Q” is a generated variable that takes the value of the residuals generated by the original
regression.
Ln(Q^2) becomes the new dependent variable instead of Ln(Deal Value) while the rest
of the independent ones and dummies stay the same.
New R-Squared = 0.206
Obs*New R-Squared = 9.484 < X2(6 d.f.) = 12.591 => Fail to reject H0.
5) Autocorrelation
Autocorrelation should not be a problem as our data is cross-sectional. The Durbin-
Watson statistic of 1.20 it’s failing to point a certain conclusion but indicates a tendency
towards positive autocorrelation as visible in Figure 1.
The following hypothesis’ are tested:
H0: No autocorrelation;
40
H1: Autocorrelation of order 2.
The Breusch-Pagan-Godfrey Lagrange Multiplier test, on our data, at two lags gives the
following result:
Obs*R-squared = 7.56
Prob. Chi-Square(2) = 0.02 => We reject H0 at 0.1 and 0.05 significance levels and we
fail to reject for 0.01.
8.6. Key stats of chosen model
Table 8: Regression key indices
R-Squared 0.889
Adj. R-Squared 0.862
S.E. Regression 0.980
Sum Resid.^2 34.605
Log Likelihood -58.724
F-Statistic 32.193
Prob(F-Statistic) 0.000
Durbin-Watson 1.205
8.7. Choice of model 67
Table 9: Results of testing models with or without the HI and STOXX Index variables
Coeff. (1) Coeff. (2) Coeff. (3) Coeff. (4)
ET -7.2e-10* -6.51e-10 -7.06e-10* -6.69e-10*
Dc -0.113 -0.140 -0.110 -0.190
Dd -0.032 -0.074 -0.04 -0.066
EA 5.8e-11 5.96e-11 5.77e-11 6.37e-11
SA 0.186* 0.180 0.186* 0.171
ST 0.936*** 0.928*** 0.934*** 0.935***
Dr&dT 0.127 0.098 0.122 0.115
HI -0.082 0.299 - -
SI 0.425 - 0.375 -
Intercept -3.863 -3.969 -4.18 -1.51
R-Squared 0.889 0.888 0.889 0.888
AIC 2.988 2.952 2.944 2.913
6 (1) represents the model used in the thesis while (2), (3) and (4) are preliminary models
containing either one or none of the HI or STOXX Index variable. 7 ***, **, * indicate significance levels of 1%, 5% and 10%, respectively.
41
8.8. Complementary figures
Figure 3: Path of the calculated HI and the STOXX Index for the timespan 2010-2018
8.9. Other insights
Even if most of the control variables are insignificant in our case, it is still our duty to
acknowledge them and relate their effect to the previous literature that found them
significant. The obtained set of P-values is consistent with Mishra (2018) and Mataigne
et al. (2018) on the common variables of choice.
Our interpretation8 starts with the dummy variables and according to our results, choosing
to acquire fully with cash reduces the deal value by ~10.7%. This possibly it’s a sign that
multiple theories are checking out. According to Myers’s Pecking Order theory (1984) a
firm is willing to use its internal resources ahead of other methods of funding and it’s
maximizing corporate value by using cash only for positive NPV purchases. The
8 The formula for dummy interpretation as suggested by Halvorsen: [(eβdummy −1)*100]%.
Halvorsen, R. and R. Palmquist (1980). The interpretation of dummy variables in
semilogarithmic equations. American Economic Review, 70, 474–475.
0.00
100.00
200.00
300.00
400.00
500.00
600.00
700.00
800.00
0500
100015002000250030003500400045005000
01-D
ec-2
010
01-J
un-2
011
01-D
ec-2
011
01-J
un-2
012
01-D
ec-2
012
01-J
un-2
013
01-D
ec-2
013
01-J
un-2
014
01-D
ec-2
014
01-J
un-2
015
01-D
ec-2
015
01-J
un-2
016
01-D
ec-2
016
01-J
un-2
017
01-D
ec-2
017
01-J
un-2
018
01
-Dec
-20
18
Sto
xx
Index
Val
ue
HH
I V
alue
Timespan
Evolution of pharmaceutical market indices
Stoxx Index HHI
42
reduction in deal value could be interpreted that when using cash, the management is
more committed to reducing the information asymmetry, further being consistent with the
results of Brown and Ryngayert (1991) that cash financed deals generate slight positive
returns to the acquirer, rather than negative as per Goergen and Reneboog (2004).
As our results suggest, a target with R&D increases the deal value by ~13.5%. This
dummy behaves as expected considering the industry nature and M&A motives, as
portrayed by Danzon et al. (2007).
A merger or acquisition occurring in a domestic setup is ~3.1% lower in deal value than
a cross-border one. According to Deng and Yang (2015) the stock returns are higher post
announcement for a cross-border deal as the shareholders value more the new market
access and resource scouting, over the risk. In our setup, we can assume that the
management is also more willing to pay extra for the same reasons, in addition to general
higher cost caused by information asymmetry and integration, in cross-border M&A
scenario.
The control variables for the bidder were both insignificant in our model but our suspicion
is that at least the P-value for the size coefficient, should improve with an increase in
sample size. On one hand, an 1% increase in the bidder’s EBITDA would result in a
0.07% increase in the deal value, taken at sample average. For every 1% increase in the
assets of the bidder, the deal value increases by ~0.18%. This outcome is consistent with
the argument provided by Moeller, Schlingemann and Stulz (2004) that the bigger the
acquirer is, the heftier the premium offered in order to maintain its market share.
In the formulation of our model we decided to add a measurement of the concentration in
the market and another one for market risk, both customized for the EU pharmaceutical
and biotechnology industry. This has been proven to bring improvement over the original
model, as per R2, ultimately generating significant results even if there is no research in
our literature review that would back up our conclusions. One percent increase in HI
brings a -0.08% change in the deal value according to our findings. In theory, this makes
sense because a higher HI value denotes a high market concentration, thus little
competition. Hence, the less competitive the market is the lower the amount the acquirer
is willing to pay as he can take advantage of his generally more dominant market
43
positioning. In terms of risk if the market becomes less risky, equivalent with a STOXX
Index escalation, for every 1% increase the deal value increases by 0.42%. This result is
consistent with the M&A history, as per IMAA, that shows that during “good times” the
number of mergers and acquisitions together with the size of the deals increases, in
contrast with the time when the index decreases and the premium is not favorable enough
for the target to engage into merger.