Mergers and acquisitions in Europe: The effects of
payment methods and financing decisions
Student number: s3273121
Name: David Bastiaan Kruizenga
Study Programme: MSc International Financial Management (IFM)
Field Key Words: Mergers and acquisitions, Event study, Financing decisions,
Payment method, Share price performance
JEL codes: G12, G32, G34
Supervisor: Prof. Dr. W. (Wolfgang) Bessler
Co-Assessor: Dr. W. (Wim) Westerman
Date: June 7, 2019
Abstract
2
1. Introduction
Mergers and acquisitions (M&A) play a fundamental role in corporate finance. In the
last decade, M&A activity grew rapidly and was popular among corporations that seek
opportunities to expand their businesses. However, the public’s resistance to large
corporations grows1, which occurs mainly when the M&A announcement is shared with the
public. When a merger or acquisition is publicly announced, a considerable amount of
information is confessed about the potential transaction. The effects of the announcement
represent the capital markets’ expectation of the possible takeover benefits (Asquith, 1983).
One recent acquisition in Europe that had impact on business and society, was the
takeover of Monsanto by Bayer AG. German based Bayer AG one of the largest
pharmaceutical corporations in the world, announced in 2016 his interest in Monsanto
which is a leading producer of genetically engineered crops with headquarters in the United
States. On May 24, 2016 Bayer AG publicly announced the merger plans in reaction to
market speculations and stakeholder inquiries. The incentive of the German corporation is to
acquire Monsanto in becoming the market leader in agriculture. The management of Bayer
AG expected synergies of USD 1.5 billion after three years of operation. The value of
Monsanto is estimated USD 62 billion, whereby Bayer AG would pay USD 128 per share in an
all-cash offer2. The announcement led to a mixed reaction in the markets. The shares of
Bayer AG fall by 8% compared to 6.2% raise in the stock price of Monsanto. Shareholders in
the German firm claimed they invest in the pharmaceutical business and not for the
3
The past growth in mergers force academic research to analyses the drivers of
corporations and its impact on firms and capital markets (Andrade, Mitchell and Stafford,
2001). Most studies in literature are focused on the United States and the United Kingdom,
since the corporate markets plays a formerly role in the mergers and acquisitions. The
pattern of M&A activity in Continental Europe is relatively unexplored, although in recent
years there is a more pronounced appearance of the European market in merger activity.
Hence, more research studies concentrate on the effect of M&As in Europe (see e.g.
Martynova and Renneboog, 2009; Mateev, 2017).
Most studies on merger announcements in corporate finance follow the event study
methodology (see e.g. MacKinlay, 1997), in which the impact of events is analysed on the
value of the firms. This study performs an event study by investigating the impact of
financing decisions and payment methods on the stock price movements around the merger
announcement. Special attention is paid to investigate the impact of the financing sources
and method of payment separately. Bessler, Drobetz and Zimmerman (2011) argued that
both determinants are independent from each other and the impact on the firm’s value can
be viewed individually. There are a few studies in finance that are coherent with the
discussion of Bessler et al. (2011), and these are from Bharadwaj and Shivdasani (2003),
Martynova and Renneboog (2009) and Fischer (2017). Fischer (2017) discussed limitations in
the sources of financing, and mentioned that further research should look in the
pre-dominant source of financing since the information provided was relatively unclear.
Therefore, this research assumes that the determinants of the financing source and method
of payment could only take one factor. In other terms, the transactions do not consider
mixed financing sources or payment methods. The takeovers in this study are covered by
4
others, distinctions are made between firms from the United Kingdom and Continental
Europe and domestic and cross-border bids. This because country specific elements affect
the decisions made in corporate takeovers (Huang, Officer and Powell, 2016). Further
attention is paid to the efficient market hypothesis of Fama (1965). Fama (1965) claim that
the capital markets react directly to public announcements, therefore investors cannot
benefit from overvalued or undervalued shares. The hypothesis will be tested whether the
market efficiency holds or not.
The remaining part of this report consists of several sections. Section two provides an
informative background of multiple theories in corporate finance that are used for this
research. In addition, empirical analyses from prior research behind the payment methods,
financing decisions and cross-border mergers are discussed. The theories in corporate
finance in synthesis with prior results from previous studies forms the hypothesis
development of this study. Multiple hypotheses are constructed and presented in section
three. Section four discusses the data that is used for the research. The data restrictions are
mentioned, followed by the deal distribution along the European countries and the summary
statistics of the sample. The methodology section explains which approaches are used, how
the stock price performance is measured, and which tests are performed. The results of this
study are covered in section six. Additionally, the conclusion, discussion and limitations of
my research are shown. The report ends with the references and appendices. In the
5
2. Literature review
This section starts with the discussion of common theories in corporate finance that
are used in this research. The theories are explained in this research, since the theoretical
framework seems to provide evidence on past studies in mergers and acquisitions. The
theories are the Efficient Market Hypothesis from Fama (1965, 1991), the pecking-order
theory (Myers and Majluf, 1984), and the foreign direct investment theory (Hymer, 1976).
The literature section is followed by possible motives for capital structure decisions and
corporate takeovers. And this chapter ends with the empirical review on analyses regarding
financing decisions, method of payments and cross-border deals in corporate takeovers.
These information is gathered to understand and explain the research topic.
2.1 Theories in corporate finance
2.1.1 Efficient Market Hypothesis
In 1965 Eugene Fama published the Efficient Market Hypothesis (EMH), which since
then is revised multiple times (see e.g. Fama, 1971; Fama, 1991). The theory discusses that
efficient markets - these are markets without frictions - reflect all available information
directly in the stock price of a given asset. The theory consists of three variants, where the
semi-strong form adheres to the execution of event studies in corporate finance. In the
semi-strong form stock prices adjust immediately to publicly available information. The
other two variants, the weak and strong form, rely respectively on historical prices and
private information (Fama, 1965).
A merger announcement that is made public will reflect the new information in the
6
(MacKinlay, 1997). Event studies on daily data seem to adjust in one day to event
announcements. And this quick adjustment is consistent with market efficiency, where no
under- or overvalued securities are available. Fama (1991) argues in this case that
companies’ investment strategies cannot earn abnormal returns, because all stock prices
mirror the true value. However, when the market reacts slowly to the merger
announcement, event studies may face the joint-hypothesis problem. The joint-hypothesis
explains that testing for market efficiency becomes difficult when the asset pricing models
are inaccurate. In this case companies may overpay for targets and the market realizes this
slowly, hence the market is inefficient and valuation of asset prices is hard (Roll, 1986).
Prior researchers in corporate finance critically reviewed the efficient market
hypothesis (see e.g. Ball, 1978; Grossman and Stiglitz, 1980; Ross, 1986). However, Fama
(1991) concluded that event studies like merger announcements support the market
efficiency. This study tests the effect of the merger announcements on the stock price
performance, therefore follows the efficient market hypothesis of Fama (1965). See Fig. 1 for
an indication how the efficient market works3.
Figure 1
Efficient market hypothesis.
3
7 2.1.2 Pecking order theory
The pecking-order theory infers that the cost of financing increases with
asymmetric information. Asymmetric information means one party has better information
than the other (Myers and Majluf, 1984). This is widely studied in the principal-agent
relation and is fundamental to the signaling theory of Spence (1973). Myers and Majluf
(1984) suggest in their work that management is better informed about the firm’s value than
potential investors. The pecking order theory assumes that companies prioritize their
financing decisions based on the cost of financing. Therefore, companies use internal funds
first, followed by debt and equity issues (Myers and Majluf, 1984).
If companies issue new debt or equity the cost of capital increases (Modigliani
and Miller, 1958). Issuing new equity is observed by potential investors that managers think
their shares are overvalued, hence companies may take advantages of the over-valuation.
Therefore, investors may place a lower value on the issuance, or ask a higher rate of return
(Myers and Majluf, 1984). On the contrary, debt issuance signals the companies’ confidence
that the investment is profitable and the stock price is undervalued. Thus, debt issues are
preferred over equity issues. Since mergers and acquisitions are viewed as enormous
investments, financing decisions may be explained by the pecking order theory of Myers and
Majluf (1984).
2.1.3 Foreign direct investment theory
The foreign direct investment theory (FDI) is originally from Hymer (1976) who
presented the theory as the subject of his PhD dissertation. Hymer (1976) made the
distinction between financial investments across borders and foreign direct investments.
According to his theory the foreign direct investments gave firms control over business
8
Bank (1996) categorize FDIs as cross-border investments where an investor and/or business
in one country establishes a form of controlling ownership in an entity based in another
country. Investments which establishes ownership control are mergers and acquisitions,
intra company loans, building new facilities and reinvesting profits earned from overseas
operations (World Bank, 2012). According to the OECD (2002), FDI’s are key elements in
international economic integration since it creates stable and long-term relationships
between countries. Additionally, Froot and Stein (1991) mention that firms can profit from
imperfect capital markets with asymmetric information by doing cross-border investments,
followed by other advantages of FDIs such as cost and tax advantages, market differentiation
and economies of scale. The GailFosler Group (2018) studied the interaction between global
M&As and FDIs and found some interesting results. Specifically, the rise of FDIs are followed
by the increase of cross-border mergers. Consequently, companies may expand their
business globally to benefit from the advantages underlying the foreign direct investment
theory. See Fig. 2 for the relation between global M&A and FDI, performed by The GailFosler
Group (2018).
Figure 2
9
2.2 Mergers and acquisitions
2.2.1 M&A motivesThe fundamental objective of mergers and acquisitions is the creation of synergies, to
develop corporate growth, increase market power, boost profitability and improve
shareholders’ wealth (Alexandris, Petmezas and Travlos, 2010). Additionally, Andrade et al.
(2001) mention that economic theory provide many reasons for companies engaging in
merger activities. The possible reasons are built on efficiency-related purposes involving
economies of scale, attempts to create market power and take advantages of diversification
opportunities. Mitchell and Mulherin (1996) assume that mergers occur in waves, or within
waves. The authors suggest that mergers might appear as a reaction to unexpected shocks in
industries. Studies such as Andrade et al. (2001) and Mitchell and Mulherin (1996) provide
evidence that merger activity clusters by industry. Among others, synergies could also be
realized by economies of scale, vertical integration and adoption of more efficient
technology (Jensen and Ruback, 1983). In contrary to prior research conducted by Alexandris
et al. (2010) and Andrade et al. (2010), Jensen and Ruback (1983) found evidence that
takeover gains do not come from market power development.
The empirical literature in corporate finance presents differences in takeover gains
between target and bidding firms. Studies such as Jensen and Ruback (1983) and Andrade et
al. (2001), found evidence that the shareholders of target firms benefit the most from
mergers. These dissimilarities in takeover gains may be faced by information asymmetry and
uncertainty in mergers and acquisitions. Information asymmetry and uncertainty in mergers
affects deal characteristics and wealth creation for both parties (Moeller, Schlingemann and
Stulz, 2005; Luypaert and Van Caneghem, 2017). Luypaert and Van Cangehem (2017) argue
10
value and synergistic effects. Contrary, the wealth effects for target companies depend on
the acquirer value and potential synergistic gains. Additionally, Alexandris, Fuller, Terhaar
and Travlos (2013) found that the returns of acquirers rely on concerns about the strategic
potential and complexity of the deal.
2.2.2 Financing decisions
Recent empirical literature has observed the method of payment and the financing
sources in corporate takeovers. In most articles (see e.g. Vermaelen and Xu, 2014; Faccio
and Masulis, 2005; Harford, Klasa and Walcott, 2009), the ‘method of payment’ is
considered as synonyms for the ‘sources of takeover financing’. Bessler et al. (2011) argue
that the method of payment and the sources of financing are independent and therefore can
be viewed separately4. The sources of finance separated from the payment method in
acquisitions, is a subject in academic finance that is investigated briefly. There are a few
studies consistent with the discussion of Bessler et al. (2011), among them the studies from
Bharadwaj and Shivdasani (2003), Martynova and Renneboog (2009), and Fischer (2017).
This study follows the approach of Bessler et al. (2011) and investigates the three research
studies of Bharadwaj and Shivdasani (2003), Martynova and Renneboog (2009), and Fischer
(2017).
Bharadwaj and Shivdasani (2003) examine a sample of 115 cash tender offers
between 1990 and 1996. Their paper differs from prior studies since the authors focus on
the source of financing apart from the method of payment. However, their study only
consists of cash tender offers so it is unable to differentiate between financing decisions and
4
11
payment methods. The study of Martynova and Renneboog (2009) filled this gap by linking
the payment method in acquisitions to the sources of financing (Fischer, 2017).
The work of Martynova and Renneboog (2009) contributed to the field of corporate
finance. The empirical study of Martynova and Renneboog (2009) focused on a European
sample and is the first that study both the payment methods and financing sources in
corporate takeovers. The authors argue that the motives underlying the method of
payments may lead to incorrect conclusions about the validity of the theories that explain
the firm’s financing decisions. Therefore, they classified takeovers to the sources of
financing, to test for predictions determined by financing theories (Martynova and
Renneboog, 2009). The theories mentioned are, among others: pecking order and market
timing (Myers and Majluf, 1984), regulatory environment (La Porta et al., 1997), debt
overhang (Myers, 1977) and the agency cost of debt and equity (Jensen and Meckling, 1976).
Fischer (2017) continues with the subject by expanding the analysis. He claims that
the ‘connected model’ of Martynova and Renneboog (2009) explain the source of financing
dominated by the method of payment. Therefore, the author suggests that internal funds
are used for smaller takeovers, additionally external funds are needed for larger bids. Hence,
Fischer (2017) used a different approach and argues that the source of financing is decided
in a two-stage process. In this process internal funds are preferred over external funds,
which is similar to the conclusions of the pecking-order theory (Myers and Majluf, 1984). In
addition, the second stage considers debt and equity issues if necessary (Fischer, 2017).
2.2.3 Payment methods
Martynova and Renneboog (2009) mention the following means of payments in
12
European sample, 63% of the takeovers are paid with cash and 37% is borrowed to financing
the offer. The authors find evidence that the choice of payments is influenced by the
bidder’s strategic preferences. The analysis reveals that the payment method depends on
the degree to which the bidder’s shareholders wish to retain control after the transaction.
Another example caused by the payment method, is whether the bidder’s shareholders wish
to share the risk with the target shareholders. These elements do not influence the financing
decision (Martynova and Renneboog, 2009).
The choice of payment is fundamental since cash and stock offers differ regarding
transaction risks. The differences may be faced by information asymmetry and disparities in
the pricing mechanisms (Bessler et al., 2011). The bidder could be overpaying the target or
face the concerns that the offer is too low, therefore targets may reject the offer or attract
competing bidders into the transaction. Hence, the acquirer should decide on the method of
payment before making an offer (Bessler et al., 2011). When the shares of the bidder seem
to be overvalued or when risk-sharing motives are essential for the acquirer’s shareholders,
then stock payments are preferred over cash payments (Bessler et al., 2011; Martynova and
Renneboog, 2009). If other factors play a role in the payment decision, then cash offers
might be preferred (Luypaert and Van Caneghem, 2017). Among others prior research
showed that due information asymmetry and valuation uncertainty around a stock
acquisition, the market’s perspective threat a stock payment less favourable than a cash
payment (see e.g. Fuller, Netter and Stegemoller, 2002; Moeller et al. 2005; Luypaert and
Van Caneghem, 2017). Furthermore, cash hoarding acquirers can strengthen the negative
signal if they finance their deal with stock according to Lie and Liu (2018). Ismail and Krause
(2010) argue that there is still a significant gap in understanding the determinants of the
13
related to the target, acquirer, or deal. Since other variables that may impact the payment
method and stock return could arise from the environment. The authors mention the
general conditions of the stock market, ownership structure of the companies involved and
the bidding process of the merger (Ismail and Krause, 2010).
Prior research from Huang and Walkling (1987) investigated the effect of the target
abnormal returns on the acquisitions announcements. They analysed whether there are
differences in abnormal returns between cash and stock offers, tender and merger offers,
and resisted and not resisted bids. Huang and Walkling (1987) reported that abnormal
returns differ significantly with the method of payment. The factors that may influence the
payment method are taxes, regulatory environment, agency problems, compensation
effects, and accounting treatment.
2.2.4 Internationalization
The method of payments has more implications for cross-border bids than domestic
mergers according to Huang et al. (2016). The authors expect that cross-border deals involve
factors that are not considered by domestic acquisitions. They claim that the method of
payments in cross-border M&As received little attention in the past literature. Huang et al.
(2016) argue that the transparency, corporate governance, and/or the institutional quality of
the country where the target is located impact the method of payment. Therefore, the
cross-border deal is riskier for the acquirer, when the target is located in a country that exhibits
weaker corporate governance practices, weaker shareholder protection, or less
transparency. They find evidence that the country-level risk factors significantly influence
the choice of payment in cross-border transactions (Huang et al., 2016).
14
domestic and cross-border takeovers. The sample investigated in their research consists of
187 merger announcements from Continental Europe and the United Kingdom. The authors
found strong evidence that the means of payments and the location of the bidder and target
has impact on the share price. Goergen and Renneboog (2004) found interesting results:
deals involving firms from the United Kingdom obtain higher abnormal returns compared to
deals between both Continental European firms; cash offers generate higher abnormal
returns than equity offers in cross-border bids; and domestic deals create higher shareholder
wealth than cross-border acquisitions. The authors assume that higher returns for
companies from the United Kingdom are derived by the more developed market for
corporate control.
Another study finds no evidence in the acquirer’s abnormal returns regarding
payment methods between domestic and cross-border deals (Dutta, Saadi and Zhu, 2013).
Dutta et al. (2013) investigated a large sample of Canadian firms competing in cross-border
deals, despite finding differences between wealth effects. However, their insignificant
results may be caused by the relative size of the cross-border deals. Since the majority of the
transactions occurred with firms from the United States the sample variety is low. Though,
Mateev (2017) found significant evidence that the short-term performance of domestic and
cross-border deals differs. He investigated almost 3,000 European acquisitions and
discovered that domestic acquisitions earn higher abnormal returns than cross-border deals.
When controlling for firm characteristics like type of transaction, activity relatedness
between bidder and target, or the public status of the target, then domestic acquisitions
15
3. Hypothesis
This chapter provides the hypothesis development, where the structure relies on the
empirical analysis of past researchers in the field of mergers and acquisitions. Since we
assume the Efficient Market Hypothesis of Fama (1965, 1991) holds then the announcement
of a merger will adjust directly in the stock prices of the companies. Consequently, the
research investigates the stock price performance during the event period of the merger
announcement. Both single and multiple regression models are used to explain the effects of
the payment method and financing decisions on the wealth creation of the shareholders.
Chapter three is structured per topic, where results from prior studies built the expectations
for this research.
3.1 Financing decisions
Regarding the financing sources independently from the payment methods, three
studies are considered for the development of expectations. First, the study of Bharadwai
and Shivdasani (2003) investigates in which circumstances acquisitions are bank financed
before doing any transaction offer. Bank financing is superior when an acquirer has a low
cash reserve, or the relative size of the takeover is large. The study shows that the abnormal
returns around the announcement are higher for acquisitions financed with bank debt
compared with acquisitions financed with internal funds.
Secondly, Martynova and Renneboog (2009) link the method of payment to the
source of financing. This enables them to measure the additional performance of companies
that choose the source of finance and payment method independently. They claim that:”
Investors take into account the information signaled by the choices of both the payment
16
of the takeover announcement,” (Martynova and Renneboog, 2009, p. 28). The authors
show that debt financing outperforms internal financing. Hence, debt financing sends a
positive signal to the market that the stock of the bidder is not overvalued.
The final study of Fischer (2017) follows the pecking order theory (Majluf and Myers,
1984), and is therefore consistent with the research of Bharadwai and Shivdasani (2003),
and Martynova and Renneboog (2009). The theory assumes that companies prioritize their
sources of financing based on the cost of capital. Therefore, companies finance their
investment with internal funds followed by debt financing and equity financing (Myers and
Majluf, 1984). Following the pecking order theory, Fischer (2017) expects that M&As using
internal funds perform better than M&As financed with debt and credit.
According to the pecking order theory of Majluf and Myers (1984), we expect that
debt financing achieves better abnormal returns than equity financing around the merger
announcement.
H1: debt financing outperforms equity financing in the short run.
3.2 Payment method
In addition to the signaling theory of Spence (1973), Lie and Liu (2018) suggest that
cash hoarding acquirers can strengthen the negative signal if these companies finance their
deal with stock. The authors assume that acquirers with large amounts of cash have greater
flexibility in choosing a payment method. This can be viewed in accordance with the pecking
order theory, where capital investors assume cash offers as a favorable signal and equity
offers as an unfavorable signal (Majluf and Myers, 1984).
Empirical results from Fuller et al. (2002), and Moeller et al. (2005) support the
17
equity offers. The results of both studies show higher abnormal returns for acquisitions
financed with internal funds. Moreover, other studies (see e.g. Fischer, 2017; Mateev, 2017;
Martynova and Renneboog, 2009) follow the procedure of the pecking order theory in their
research studies. The authors expect that cash offers generate higher abnormal returns than
equity offers around the merger announcement. Hence, we expect that cash payments
outperform stock payments for bidders.
H2a: Cash payments outperforms stock payments in the short run.
Studies such as Lie and Liu (2018), and Ismail and Krause (2010) show that targets
earn positive abnormal returns after the announcement, which holds for both stock and
cash payments. The researchers argue that cash payments are favorable for target’s
shareholders, which is in line with the theory of Majluf and Myers (1984). Vermaelen and Xu
(2015) shed an interesting view on the payment method. They assume among others, that a
cash payment may indicate that the acquirer’s stock is undervalued. Although this would be
irrelevant to the target firm because a cash transaction is not sensitive to the acquirer’s
stock price. The management of the target has the objective to maximize the transaction
price. However, the target firm knows less about the true value of the acquirer hence
information asymmetry might occur (Vermaelen and Xu, 2015). Asymmetric information can
be measured by the Tobin’s Q, which is the ratio of market value to book value (Ismail and
Krause, 2010). We want to examine the effect of the acquirer’s market to book ratio on the
target returns in cash paid takeovers. Since literature assume bidders pay with cash if their
stock is undervalued, capital markets expect these companies have low market to book
ratios (e.g. M/B = <1). Hence, we expect that acquirers with high market to book ratio in
18
acquirers are overvalued and may paying more for their targets (see e.g. Ismael and Krause,
2010; Vermaelen and Xu, 2015).
H2b: High M/B ratio of the bidder create more wealth for targets, compared to low M/B ratio of the bidder
3.3 Internationalization
Goergen and Renneboog (2004) reported that both UK bidders and targets earn
significant higher returns than firms from Continental Europe. Regarding the method of
payment, the pecking order theory (Majluf and Myers, 1984) holds in their sample of
cross-border acquisitions where cash offers generate higher returns than equity offers. Secondly,
Goergen and Renneboog (2004) did not find evidence of the foreign direct investment
theory (Hymer, 1976) since the domestic deals generated higher abnormal returns
compared to cross-border bids.
On the contrary, Mateev (2017) found that the foreign direct investment theory
holds to a certain extent for his study on bidding firms. He studied European domestic and
border mergers and found that stock financed mergers react more positively in
cross-border takeovers. In contrast, cash financed deals shows no significant difference in
abnormal returns between domestic and cross-border acquisitions. This outcome is
consistent with Dutta, Saadi and Zhu (2013), who found evidence that stock financed
mergers generate higher wealth effects than cash payments in cross-border deals. They
mention that stock payments can be perceived as a solution to reducing information
asymmetry and lowering corporate governance risk in cross-border transactions.
Regarding the next analysis we want to examine the effects of the pecking order
19
coherent with the pecking order theory, although the results of Dutta et al. (2013), and
Mateev (2017) show opposite effects. Huang et al. (2016) argued that cross-border
transactions are riskier due different country characteristics. This would influence the
evaluation of (potential) investors around the merger announcement and therefore affect
the stock price performance. Hence, their work does not follow the foreign direct
investment theory. We expect that domestic transactions paid in cash outperform equity
payments and cross-border bids.
H3a: Cash payments in domestic bids generate the highest abnormal returns.
Most studies (see e.g. Dutta et al, 2013; Huang et al., 2016; Mateev, 2017)
investigated the stock price movements of bidders in domestic and cross-border
transactions. Goergen and Renneboog (2004) also analysed the announcement effects of
target companies. Their results show that the announcement of a takeover bid causes
positive abnormal returns for the shareholders of the target. Abnormal returns of 9% are
even realized on the event day. They found strong evidence that the method of payment for
the target’s shares impacts the share price reaction. Cash payments trigger substantially
higher abnormal returns, than equity offers and/or combined offers (Goergen and
Renneboog, 2004). This is confirmed by the study of Huang and Walkling (1987), who found
that target abnormal returns are significantly higher when acquisitions are paid with cash.
Since cross-border acquisitions are risky due to different country characteristics (Huang et
al., 2016), we expect that cash payments in domestic mergers outperform stock bids and
cross-border takeovers.
20
4. Data
Chapter four include the data analysis, which discusses the data description and
explains the sources that are used to gather the information. Moreover, this section
introduces the dependent, independent and control variables followed by the sample
distribution.
4.1 Data description
The transaction data of the deals are obtained from the Zephyr database from
Bureau van Dijk. The deals include mergers or acquisitions, which are announced and
completed between 1 January 1997 and 31 December 2016. Only European transactions are
included in the sample, therefore the headquarters of the acquirer and target is located in
Continental Europe (CE) and/or the United Kingdom/Ireland (UK). The stock price
performance of the acquirer and target is investigated, so acquirers must be publicly listed.
The sample of target companies include both public and private firms. The initial stake of the
takeover company is maximum 50% and the minimum acquired stake is 50%. Furthermore,
the transaction value has a minimum of 1 million euro and multiple deals from the same
companies are allowed but do not overlap in estimation periods. Regarding the payment
method and sources of financing, the following information is required by Zephyr:
information on the method of payment, financing source, and the deal value in euro. Hence,
the initial sample includes 1,780 transaction observations.
Datastream from Thomson Reuters is used to retrieve the historical stock prices per
observation. The ISIN identifier is applied in Datastream to collect the share prices of the
events. Since a small size of observations were not applicable for the analysis of the share
21
include 717 transactions that were found sufficient. This is equal to the sample of total
acquirers, but differs for the target sample. Most of the acquisitions involved private targets,
hence the event study is not applicable to those companies. The final sample of target
companies contain 102 public companies, where from 92 companies the stock price
movements can be investigated around the events. Table 1 in Appendix A displays the
selection criteria for the sample restrictions.
4.2 Dependent and independent variables
The study considers the impact of the financing decisions, payment methods, and
cross-border deals on the stock price performance of bidders and targets. An event study is
conducted to measure the reaction of the stock prices, therefore the dependent variable is
the abnormal return. Consequently, the independent variables are the method of payment,
financing sources, market to book of the acquirer, and whether the transaction is a domestic
or cross-border acquisition. Dummy variables are formed for the independent variables, and
are specified below:
Financing source = [Debt = 1, Equity = 0].
Payment method = [Cash = 1, Shares = 0].
Internationalization = [Cross-border = 1, Domestic = 0].
Market to Book = [M/B < 1 = 0, M/B > 0 =1].
4.3 Control variables
We include a set of control variables for the regressions. The aim is to analyse
22
transaction specific characteristics; (ii) firm specific characteristics; (iii) country specific
characteristics. The variety of variables are from multiple studies (see e.g. Mateev, 2017;
Dutta et al., 2013; Martynova and Renneboog, 2009; Huang et al., 2016). The accounting
variables and stock price returns are winsorized at the lower 1% and upper 99% level to
remove outliers.
The transaction specific variables other than the dependent variables include the deal
value and the deal allocation per industry. The industry classification benchmark of FTSE
Russell is used for the distribution per industry, which include ten categories. The FTSE
Russell is a benchmark adopted globally for categorizing companies across industries,
sectors, and subsectors5. Further, we set dummy variables whether the transactions include
firms from the utilities and financial industry. Companies from the utility sector are often
isolated because of their regulated character (Kahle and Walkling, 1996). In addition,
financials should be treated different because of their dissimilarities in the financial
character as compared to other industries, such as high leverage levels (Fama and French,
1992; Kahle and Walkling, 1996).
The firm specific characteristics consist of several balance sheet indicators, such as
firm size, asset tangibility, profitability, and cash. These variables may affect the payment
method, financing decisions and stock price performance around the merger
announcement. The last set of control variables are covered by the country aspects. Specific
dummies are created for the legal system, legal rights, country transparency, and GDP per
capita. The legal system around Europe consists of mainly Common Law (United
Kingdom/Ireland) and Civil Law (Continental Europe) countries. Other country variables such
5
23
as legal rights, country transparency, GDP per capita, are obtained from the World Bank
database6. The dummy variables per country indicator contain 0 and 1 values to distinguish
between country characteristics. The median is calculated per variable to identify the “high”
and “low” ranks. Table 2 in Appendix A presents all variable definitions.
4.4 Sample distribution
Table 3 demonstrates the sample distribution per country. Most of the transactions
are domestic, and around 60% of these occur between companies located in the United
Kingdom/Ireland (UK). With respect to the deal characteristics, their mainly paid in cash and
financed with debt (see Table 4 in Appendix A). Almost 70% of the deals are from companies
operating in the same industry, where the consumer services industry gathers the highest
frequency in merger announcements. Around 85% of all transactions involve private targets
and this indicates 615 target companies in total. See Fig. 3 in Appendix B for the distribution
of deals over the years. The summary statistics of the dependent and independent variables
are given in Table 5 displayed in Appendix A. These figures shows the numbers before
removing any outliers. Table 6 in Appendix A displays the t-statistic and corresponding
p-values of the abnormal returns over seven days surrounding the event before removing any
outliers.
Table 3
Distribution per country.
Acquirer Target
Country Frequency Percent Country Frequency Percent
Austria 2 0,28% Austria 2 0,28% Belgium 9 1,26% Belgium 10 1,39% Switzerland 15 2,09% Switzerland 13 1,81% Cyprus 1 0,14% Cyprus 1 0,14% 6
24 Germany 20 2,79% Germany 38 5,30% Denmark 4 0,56% Denmark 9 1,26% Spain 20 2,79% Spain 20 2,79% Finland 22 3,07% Finland 13 1,81% France 31 4,32% France 33 4,60%
United Kingdom 456 63,60% United Kingdom 429 59,83%
Greece 2 0,28% Croatia 1 0,14% Croatia 1 0,14% Ireland 11 1,53% Hungary 1 0,14% Iceland 1 0,14% Ireland 16 2,23% Italy 21 2,93% Iceland 5 0,70% Luxembourg 5 0,70% Italy 19 2,65% Serbia/Montenegro 1 0,14% Netherlands 17 2,37% Malta 1 0,14% Norway 11 1,53% Netherlands 34 4,74% Poland 14 1,95% Norway 10 1,39% Russia 3 0,42% Poland 13 1,81% Sweden 48 6,69% Portugal 1 0,14% Romania 2 0,28% Russia 3 0,42% Sweden 44 6,14% Turkey 1 0,14% Total 717 100% Total 717 100%
5. Methodology
This study focuses on the effects of the payment method and sources of financing, on
the wealth of acquirers and targets surrounding merger announcements. The Capital Asset
Pricing Model of Markowitz (1952) and Sharpe (1964) is applied in the valuation of the stock
prices. Further, the event study methodologies of MacKinlay (1997) and Brown and Warner
(1985) are used to analyse the impact of certain events. This research adopts the short-term
event study, which use the cumulative abnormal returns (CAR) as the measurement. Other
25
5.1 Capital Asset Pricing Model
The aim of the event study is to obtain the valuation effects as reaction to merger
announcements. The asset pricing models from Markowitz (1952) and Sharpe (1964) are
adopted for the valuation of the stocks. Since the theory of Fama (1965) suggests that
efficiency should be tested by asset pricing models, the (ab)normal returns in this study
follow the procedures of the Capital Asset Pricing Model (CAPM). The CAPM is used to
determine the expected rate of return of a given asset. The model was built by Markowitz
(1952) and during the years enhanced by other researchers such as Sharpe (1964).
The CAPM model (Markowitz, 1952; Sharpe, 1964) provides the scope for investors to
evaluate the trade-off between risk and returns for diversified portfolios and individual
assets. The rate of return can be obtained by the risk-free rate plus the beta of the asset
multiplied by the market premium. A proxy for the risk-free rate is the government interest
rate or a market index. The asset beta is considered as the systematic risk of a stock,
whereas the market premium is the difference between the market return and the risk-free
rate.
5.2 Event study methodology
The standard event study methodologies of MacKinlay (1997) and Brown and Warner
(1985) are used to capture the abnormal returns around the announcement. The abnormal
return is calculated by the difference between the actual return (Rit) and the expected
return ((E(Rit)). The expected return is measured over the period before the announcement,
and is labeled as the estimation window. This study uses an estimation window of 201 days,
and is covered by the period from 260 days prior to 60 prior the event. Since the estimation
26
rumors (MacKinlay, 1997; Fischer, 2017). There are two models for estimating the expected
returns, the constant mean return model and the market model (MacKinlay, 1997). We use
the latter one and adopt the STOXX EUROPE 600 as the market model. The index includes
600 European organizations across large, medium and small size capitalized companies from
17 European countries. This index is applied since it represents the research data set, is
available in the euro currency, and covers the sample period from 1997-2016.
The event window exists of 61 days to capture the abnormal returns, and the period
starts from 30 days prior to 30 days post the announcement. We consider alternative event
windows between the (-30,30) day interval to secure additional pre- and
post-announcements effects, which is also performed by Martynova and Renneboog (2009).
Special attention is paid to the short-term event window that covers the three days
surrounding - which is one day before to one day after - the merger announcement. Andrade
et al. (2001) claim that the most statistically reliable evidence on whether mergers create
value for the shareholders comes from short-term windows. In the period of three days, the
average abnormal stock market reaction of merger announcements is normally used as a
benchmark of value creation or destruction (Andrade et al., 2001).
Brown and Warner (1985) examined the properties of daily stock returns and how
particular characteristics of these data affect the share price impact of firm-specific events.
Their paper extends earlier work on event study methodologies on monthly returns (Brown
and Warner, 1980). The authors argue that the parameter estimation of daily data is
problematic by non-synchronous trading, since the actual returns and expected returns are
calculated over different trading intervals (Brown and Warner, 1985). According to Scholes
and Williams (1977) and Brown and Warner (1985), the ordinary least squares (OLS) of the
27
Nonetheless, this study captures the daily stock returns since most researchers in corporate
finance adopt the methodology of day-to-day stock returns. In addition, daily stock returns
follow a better distribution compared to the normal distribution of monthly returns (Fama,
1970).
The actual returns for any given security in the event, are calculated as follows:
(1)
= Period t-return on security i,
= Period t-return on the market model,
= Intercept coefficient of the market model,
= Slope coefficient of the market model,
= Zero mean disturbance term.
The abnormal return that is calculated for any given security, is computed as follows:
= – E( ) (2)
= Abnormal return in period t on security i,
= Actual return in period t on security i,
E( ) = Normal return in period t on security i.
To test the significance for the event of interest, the abnormal returns are aggregated
into average abnormal returns (AAR) and cumulative average abnormal returns (CAAR). This
to make interpretations and test if any hypothesis hold. The AAR and CAAR are calculated as
28
= 1/N (3)
= (4)
The calculation of the T-statistic is shown below:
= CAARt / (5)
The study tests regressions using the cumulative average abnormal returns (CAARs)
as the dependent variable. Thereby, a range of independent variables – continuous,
categorical, and indicator variables - are used to measure additional effects and test multiple
hypothesis. The equation is displayed as:
= + β₁ Financing source + β₂ Payment method + β₃ M/B + β₄ Transaction
value + β₅ Cash + β₆ Firm size + β₇ Profitability + β₈ Asset tangibility + β₉ Industries +
β₁₀ Transparency + β₁₁ Legal rights + β₁₂ GDP Per Capita + β₁₃ Cross-border deals + β₁₄
Status target + β₁₅ Financials/Utilities (6)
5.3 Research models
The effect of the explanatory variables on the abnormal returns of the acquirer and
target will be analysed by single and multiple regression models. The single regressions
translate the effect of one variable on the abnormal returns. In contrast to multiple
regressions, where the effect on the abnormal returns are measured by several explanatory
variables. This study used two types of linear regression models, whereby attention should
be paid to the interpretations of the model and to the regression output. For simplicity, the
regressions are illustrated by the financing sources. This hypothesis (H1) analyse the effects
29
equation investigates the impact of debt and equity financing on the cumulative abnormal
returns. Since the variable contains a dummy indicator, one could only choose debt or equity
as financing source.
= + β₁ Financing source (7)
The multiple regression model included multiple explanatory variables that translate
each effect on the abnormal returns. See Eq. (8) for the mathematical representation:
= + β₁ Financing source + βn (Explanatory variables) (8)
Whether the group of explanatory variables influence transactions that are financed
with debt or equity, the regression model should be transformed and the interpretation
change. Equation 9 shows the single regression of debt or equity and the abnormal returns.
Since there are no other variables that explain the model, the outcome is coherent with the
outcome of Equation 7.
|Debt or Equity = (9)
However, if we extend the model to a multiple regression model the economical
explanation is considerable different than the regression model illustrated by Equation 8.
Equation 10 displays the mathematical description if we extend the regression model with
multiple explanatory variables. Consequently, the explanatory variables measure the impact
on the abnormal returns that are financed with debt or equity.
| Debt or Equity = + βn (Explanatory variables) (10)
In addition, supplementary tests are performed to verifying the validity of the
30
(White standard errors), where heteroskedastic residuals are approved in the model. The
procedure was introduced by Huber (1967) and further developed by researchers such as
White (1980). Next, the Mann-Whitney test is used to compare the different values of
multiple populations. The Mann-Whitney test (or Wilcoxon rank-sum test) is a
non-parametric test that compares the populations, whether one sample will be significant less
than or greater than a randomly selected value of the other sample (Mann and Whitney,
1947; Wilcoxon, 1945). Finally, the Welch’s t-test will be used to measure the means of two
populations that have unequal variances and differ in sample size (Welch, 1947). The normal
student t-test assumes that two populations have normal distributions with equal variances,
though the Welch’s t-test is depicted for unequal variances (Welch, 1947). The Welch’s t-test
is considered to have more robustness than the normal student t-test (Rasch, Teuscher and
Guiard, 2007). Hence, Rasch, Teuscher and Guiard (2007) recommend applying the Welch
t-test for comparisons between populations with different sample sizes.
Since the sample of targets is rather low, the Mann-Whitney test and Welch’s t-test
will be performed on the target mean populations. Regression analyses on a few
observations may led to highly inaccurate parameters and biased estimates (Potter, 2005;
Maiti and Pradhan, 2009). Therefore, we choose for mean comparison tests and prevent for
inaccurate outcomes of regression analyses. Besides, Lapron and Shen (2007) argued that
due to the lack of research it is not clear whether the key findings in M&A hold for
acquisitions of private firms. Therefore, any evidence found on the population of public
31
6. Results
Chapter six shows the results of the study and contains three sections. Section one
demonstrates the descriptive statistics, which include tests whether or not the merger
announcements create value for the shareholders. Section 6.2 presents the findings of the
analysis on the sample of acquirers. These consists of three hypotheses – H1, H2a, H3a – and
additional tests. Per hypothesis the single and multiple regression output will be interpret. In
other terms, the regression follows the procedures from the equations mentioned in the
previous chapter. Hence, per hypothesis we test the impact of the main dependent variable
on the abnormal returns (see Eq. (7)) followed by the influence of multiple explanatory
variables on the returns of the acquirer (see Eq. (8)). In addition, we investigate the impact
of the explanatory variables on the abnormal returns controlled by population groups (see
Eq. (10)). Subsequently, robust standard errors are used for the multiple regression models
and the Mann-Whitney test is performed to compare the population means. This structure
holds for section 6.2, though the analysis will deviate if otherwise stated. Section 6.3 include
the tests on the sample of targets. Since the target sample size is rather low, the
expectations are demonstrated by the Welch’s t-test and Mann-Whitney test (see
hypotheses H2b and H3b).
6.1 Descriptive statistics
Table 7 perform the one sample t-test of abnormal returns over different interval
windows. The abnormal returns for the acquirer and target samples are virtually in most
event windows statistically significant different from zero. This is consistent with studies
such as (Goergen and Renneboog, 2004; Martynova and Renneboog, 2009; Alexandris et al.,
32
mean returns of the target companies are considerable higher compared to the means of
acquirers. This indicate that shareholders of targets benefit the most from merger
announcements, which is coherent with the findings Goergen and Renneboog (2004) and
Martynova and Renneboog (2009).
Regarding the market hypothesis of Fama (1965), there is no leakage effect in the
(-10, -1) days before the announcement date of the acquirers’ sample. Whether the market
reacts slowly to the merger announcement is covered by the (1,10) interval window. In both
cases – the acquirer and target sample – the abnormal returns are significant positive and
different from zero. The Efficient Market Hypothesis (Fama, 1965), consider that
shareholders could not realize abnormal returns since under-valued and over-valued
securities are not available after the merger is publicly announced. Despite, it is hard to
argue whether the market is inefficient based on these figures. Therefore the (cumulative)
abnormal returns over the event window (-30, +30) are shown in Figure 4 displayed in
Appendix B. On the announcement date (day 0) both graphs show a substantial rise in the
stock prices, which mean that the market reacts positive to the announcements. In the days
after the announcement the stock prices follow a normal pattern. This pattern is consistent
with the graphically explanation of the market efficiency hypothesis of Fama (1965). See Fig.
33
Table 7
One sample t-test of the cumulative abnormal returns.
Panel A Acquirer [-30,30] [-10,10] [-3,3] [-1,1] [0] [-10,-1] [1,10] [0,5] Mean 0.0296 0.0337 0.0312 0.0271 0.0173 0.0031 0.0116 0.0311 T-stat 3.663 6.984 9.465 9.978 8.770 1.090 3.803 9.462 P-value 0.0001 *** 0.0000 *** 0.0000 *** 0.0000 *** 0.0000 *** 0.1378 0.0001 *** 0.0000 *** Panel B Target [-30,30] [-10,10] [-3,3] [-1,1] [0] [-10,-1] [1,10] [0,5] Mean 0.2708 0.1999 0.1703 0.1540 0.1199 0.0571 0.0229 0.1425 T-stat 8.565 8.629 8.479 7.706 6.574 4.555 2.331 7.366 P-value 0.0000 *** 0.0000 *** 0.0000 *** 0.0000 *** 0.0000 *** 0.0000 *** 0.0110 ** 0.0000 ***
Table 7 presents the cumulative average abnormal returns (CAAR) of the sample of acquirers and targets. The number of observations are respectively 717 and 92. Per event window is the mean, t-statistic and p-value given. The values are winsorized at the lower 1% and upper 99%. The significance levels are: *,**,*** which stands for 10%,5%, and the 1% level.
Other results show that the acquirers from Continental Europe achieve on average
higher means (1.94%) than their counterparts from the United Kingdom (1.65%). Financial
and utility companies that expand their business by acquisitions score lower (0.26%)
compared to non-financials and non-utilities (1.95%). In addition, bidders that acquire public
targets have on average a negative mean (-0.92%) around the announcement date. This is
significant different from the stock price performance when bidders take over private
targets (2.16%). Further, targets from the United Kingdom perform on average better
(13.08%) than firms from Continental Europe (10.29%).
6.2 Acquirers: cross-sectional regression analysis
6.2.1 Financing sources
Table 8 displays the cumulative abnormal returns for the acquirer sample. Multiple
interval windows are provided in which the mean and median values are given for equity,
34
returns compared to parties who issue equity. This is consistent with the findings of
Martynova and Renneboog (2009) and Fischer (2017).
Table 8
Cumulative abnormal returns of the financing sources.
[-10,+10] [-3,+3] [-1,+1] [0] Obs. Equity 0.0207 0.0197 0.0209 0.0129 167 (0.0158) (0.0078) (0.0096) (0.0084) Debt 0.0375 0.0347 0.0290 0.0185 550 (0.0236) (0.0255) (0.0196) (0.0061) Total 0.0336 0.0312 0.0271 0.0172 717 (0.0214) (0.0208) (0.0154) (0.0044)
This table shows the average abnormal returns over different interval windows. The mean and median (in parentheses) values are given for equity, debt and total financing. Obs. stands for the number of
observations.
Table 9 in Appendix A presents the outcomes of the single and multiple regression
models. The single regression models test the impact of equity and debt financing on the
abnormal returns. The single regression models evaluate the impact of debt or equity
financing on the abnormal returns over two different event windows (see Model 1 and
Model 4). The multiple regression model includes control variables that may influence the
abnormal returns of acquirers. See Eq. (7) and Eq. (8) for the mathematical illustrations.
The results from the regressions show us that debt financing have a positive influence
on the cumulative abnormal returns. A similar conclusion was found by Martynova and
Renneboog (2009) who showed that debt financing outperforms equity financing in the
capital markets. The single regression in Model 1 presents the intercept (2.07%) of the
cumulative abnormal returns if a company finance their M&A with equity. This coefficient is
significant and different from zero at the 5% level, holding other variables constant. When
the takeover is financed with debt the abnormal returns can increase with 1.68% to the level
of 3.75%. The economical meaning shows us that the magnitude of the explanatory variables
35
is positive but has no statistically significance. Therefore, based on this model, we cannot
imply that debt or equity financing has a significant effect on the abnormal returns of
acquirers.
The first column in Table 9 presents a similar pattern with the outcome of the single
regression, so the equations can be illustrated as followed:
Equity = 0.0207 + 0.0168 * Debt (11)
Debt = 0.0375 – 0.0168 * Equity (12)
This pattern occurs since the financing source is a binary variable, which take the
values of “0” or “1”. Hence, if a company use equity as their financing source, they cannot
choose debt as their capital funding and visa versa. Therefore, one single regression model is
sufficient to illustrate the effect of the financing sources on the cumulative abnormal
returns. The pattern holds for other single regression models such as Model 4 in Table 9. The
regression outcome is consistent with the corresponding mean values from Table 8.
The multiple regression models from Table 9 reveal that the market to book ratio has
a negative effect on the returns of acquirers. The outcome is statistically and economically
significant on the cumulative abnormal returns in an event window of 21 days surrounding
the announcement. The economically relevance imply that if a companies’ market to book
ratio increase from two to three, the abnormal returns can decrease with 0.461%. This
magnitude holds also for other control variables such as the firm size and profitability.
Interesting results are shown from the target’s status which is presented by the public target
dummy. The regression model demonstrates that the shareholders’ confidence in public
targets is considerable lower compared to the confidence in private targets. The effects of
36
the public targets contain a binary variable, the private targets show opposite results. The
decreasing rate of returns of public targets is substantial, although one should consider that
this impact only plays a role if we hold the other variables constant. Therefore, the
economical relevance is less observable because in M&As multiple indicators play a role in
the wealth effects of the shareholders. Furthermore, the r-squared in the regression models
is quite low so it is problematic to make precise estimations.
Table 10 in the Appendix covers the regression models per financing source. The
outcomes imply the effect of explanatory variables on the abnormal returns of debt or
equity financing (see Eq. (10)). The single regression models prove that debt financing
outperform equity financing significantly, which is consistent with Martynova and
Renneboog (2009). The differences are 1.69% and 0.81% for the event windows of
respectively 21 days and three days around the merger announcement. Model 7 shows that
the coefficient of debt financing is statistically significant at the 1% level, implying that the
abnormal returns of debt financing in the short-term window could raise to 12.5%. The
dependent variable is measured in cumulative percentages, so considerable high returns are
economically meaningful. Nonetheless this outcome only holds when the explanatory
variables take a value of zero. Besides, the r-squared is 0.081 explaining that other variables
outside the model have more influence on the abnormal returns of debt financing. The
effect of public targets has a negative effect on the abnormal returns for both debt and
equity financing. These findings are in accordance with the research of Martynova and
Renneboog (2009) and Fischer (2017). This may imply that shareholders from the bidder are
37
In addition, robustness tests are performed whether the means are statistically
different. The Mann-Whitney test shown in Table 11 in Appendix A displays the Z-score and
P-value of the corresponding event windows. Additionally, the tests are also observed at the
announcement date (day 0). Based on the figures in this research the means of debt and
equity financing are statistically different from zero in the short run. The Z-score at the
announcement date is 2.083 which has a significance level of 5%, compared to the Z-score
on the three-day interval (-1, +1 days) of 1.737 and a corresponding significance level of
10%. The Mann-Whitney test implies that the population mean of debt financing
outperforms respectively the mean of equity financing in the three days surrounding the
merger announcement.
6.2.2 Payment methods
Table 12 presents the cumulative abnormal returns over different interval windows.
The windows cover the mean and median abnormal return values for cash paid, stock paid,
and total payments. Firms who pay corporate takeovers with cash have considerable higher
returns around the merger announcement. Studies such as Martynova and Renneboog,
(2009), Fuller et al. (2002), and Moeller et al. (2005) show similar results.
Table 12
Cumulative abnormal returns of the payment determinants.
[-10,+10] [-3,+3] [-1,+1] [0] Obs. Cash 0.0348 0.0336 0.0292 0.0192 560 (0.0210) (0.0249) (0.0169) (0.0048) Shares 0.0298 0.0228 0.0197 0.0104 157 (0.0219) (0.0078) (0.0125) (0.0027) Total 0.0336 0.0312 0.0271 0.0172 717 (0.0214) (0.0208) (0.0154) (0.0044)
38
Tables 13 and 14 in the Appendix presents the outcomes of the single and multiple
regressions model controlling for cash and stock paid acquisitions. Table 13 performs the
single and multiple regression of the explanatory variable(s) on the abnormal returns. See
Eq. (7) and Eq. (8) that illustrate the regressions mathematically.
The cash payment coefficients in Model 5 and 6 of Table 13 show us that cash
payments have a statistically significant influence on the abnormal returns calculated in the
short-term window. This indicator implies that with a 5% significance level, abnormal returns
may increase by 1.73% if they are paid in cash. This could be economical relevant, since the
cumulative abnormal returns can increase by 1.73% in three days around the announcement
if other variables held constant. The same models present positive and significant
coefficients of 10.3%, which is quite high around the announcement date for acquirers
Hence, the economical relevance is reasonable, because this may occur by accident.
The payment method is a binary variable, indicating whether companies paid their
transactions with cash or equity. Hence, we can illustrate the single regression
mathematically on the short-term window which is consistent with the figures in the third
column of Table 12.
Stock = 0.0197 + 0.0095 * Cash (11)
Cash = 0.0292 – 0.0095 * Equity (12)
In Table 14 presented in Appendix A the single and multiple regression analysis are
showed by the method of payment. The single regression conveys that cash payments
generate statistically significant higher abnormal returns than stock paid acquisitions. This
39
payments are a favourable signal to the market (Fischer, 2017; Majluf and Myers, 1984).
Profitability expressed as the EBITDA divided by the total assets reveal opposite
reactions on cash and stock paid acquisitions. The models assume an increase in profitability
reduce the abnormal returns for cash paid takeovers. The effect is significant at the 10%
level (see Model 3), indicating that an increase in the profitability ratio from three to four
can decrease the abnormal returns by 0.12%. The binary variable that specify the financial
and utility companies, show us that acquire companies outside the financial and utility
industry lead to significant higher abnormal returns in the window of (-10, +10 days).
Moreover, cash paid acquirers who takeover public targets face considerable and significant
depreciation of the market value. Hence, the significant effect of public targets on the
payment method is driven by the cash paid acquisitions. In addition, the total amount of
cash is significant and negative for companies paying with stock (see Model 4). It remains
unclear to predict the effect, since the relative size is not observable. Though, this consistent
with the pecking order theory (Majluf and Myers, 1984), implying that firms who have
sufficient cash levels would use cash for investments.
Robustness tests are included as well to compare the means of the populations. The
Mann-Whitney test shows that the population mean between the payment determinants
are statistically different from zero around the announcement date. The Z-score is 1.713 and
the corresponding P-value is 0.0867. Hence, with a 1% significance level we can assume the
population means of cash and equity payments differs significantly. In other terms, the
population mean of cash significantly outperform the equity mean around the
40 6.2.3 Internationalization
The figures from Table 15 demonstrate that domestic acquisitions generate on
average higher abnormal returns than cross-border takeovers. A similar finding was reached
by the studies from Goergen and Renneboog (2004), Huang et al. (2016), and Mateev (2017).
Other results show that bidders from the United Kingdom earn on average higher returns
than companies from Continental Europe, surrounding the seven days of the merger
announcement. This is contrary to the results from Goergen and Renneboog (2004).
Table 16 in Appendix A illustrate the single and multiple regressions of the
international deals on the abnormal returns of the acquirers. The variable of the
international deals is a binary and is followed by explanatory variables in the regression
model.
The single regressions in Table 16 unveil that cash payments in domestic transactions
have positive effects on the abnormal returns of bidders. Hence, equity offers and
cross-border acquisitions will reduce the stock price performance holding other variables constant. Table 15
Acquirer’s cumulative abnormal returns in international context.
[-10,+10] [-3,+3] [-1,+1] [0] Obs. Domestic 0.0361 0.0320 0.0293 0.0182 (0.0267) (0.0194) (0.0176) (0.0055) 524 CB 0.0271 0.0289 0.0212 0.0129 (0.0098) (0.0233) (0.0098) (0.0019) 193 CE 0.0255 0.0388 0.0313 0.0193 (0.0174) (0.0270) (0.0228) (0.0055) 245 UK 0.0379 0.0298 0.0249 0.0161 (0.0264) (0.0147) (0.0133) (0.0039) 472 Total 0.0336 0.0312 0.0271 0.0172 (0.0214) (0.0208) (0.0154) (0.0044) 717