Andy Negati, s1740199 August 2012

(1)

Value of Mergers and Acquisitions: in

respect of legal environment and

investor protection

Master Thesis

Andy Negati, s1740199

August 2012

Supervisor: Dr. ing. N. Brunia

Co-assessor: Dr. W. Westerman

University of Groningen

Faculty of Economics and Business MSc International Financial Management

Uppsala University

(2)

Abstract

The thesis is an addition to the M&A literature within the field of shareholder protection and legal environment. The effects of shareholder protection, legal environment and concentrated ownership together with traditional control variables are researched in order to analyze short-term wealth effects for all shareholders by performing an event study. The results show that M&A are motivated by synergies. The empirical findings support the idea that acquirer and target company shareholders are better off when: 1) shareholder protection is higher in domestic deals; 2) firms are involved in common law deals rather than in civil law deals3) ownership concentration is lower.

Key words: mergers and acquisitions, bidding and target company shareholder wealth effects,

(3)

Acknowledgments

I am thankful to my supervisor, Dr. ing. N. Brunia for his valuable help and constant support. His professional behaviour and knowledge in the field helped me to broaden my scientific understanding of this research subject. I am grateful for is guidance and feedback during the complete master thesis process.

I thank also Dr. W.Westerman who was supporting me from day one of the master program. He was always there with a good advice when I needed one. Thank you! It goes without saying that successful fulfilment of this master program would not been possible without the help of my family and friends. I am especially thankful to Friedemann J. Barthel and Alexandra Stroe.

(4)

1. Introduction

The research on Mergers and Acquisitions (M&A) is vast and extended, but it used to be focused dominantly on the US and the UK, due to the high volume of deals. However, in the last decade continental Europe’s M&A activity increased immensely. In the last merger wave in the 1990s, continental Europe became an important player as did other developed and emerging nations (Lowinski et al. 2004; Martynova & Renneboog, 2011a).

Studies have shown that target’s return at the announcement period are generally positive (Martynova & Renneboog, 2011a), however the picture about acquirer’s return and the complete deal is not that clear. There are studies which say that the return of acquirers are positive, some evidence is found that the return is negative or even that there is a neutral wealth effect (Bruner, 2002). The motivation for managers to do a M&A is also not completely clear, Berkovitch & Narayanan (1993) found support in their empirical study that deals are supported by agency conflict or hubris, whereas Goergen & Renneboog (2004) and Martynova & Renneboog (2011a) found empirical evidence that deals are value creating and motivated by synergies.

(6)

Djankov et. al. (2003) concludes in their study that countries with different legal traditions find unique ways to control shareholders’ interests. The strategies, how they increase control of businesses differs between strong shareholder protection countries (common law) and weak shareholder protection countries (civil law). For example, countries with high level of shareholder protection use generally private contracts, whereas countries with low level of shareholder protection prefer government regulation and concentrated ownership. Especially this mechanism of concentrated ownership produced contradicting results. Faccio & Stolin, (2006) conclude in their study that concentrated ownership in civil law countries has a negative effect on the value creation of M&A, whereas Bris & Cabolis (2008) find in their empirical study results in line with Djankov et. al (2003) prediction, that the majority shareholders of the target company in civil law countries have higher bargaining power and demand higher premium for their shares. To bring clarity to this discussion, concentrated ownership variable was used in the comparison between common law and civil law countries. Furthermore, findings might also support investor’s choice of whether to invest in firms with concentrated ownership or rather opt for firms with dispersed ownership in civil law countries.

In the past 15 years, cross-border M&A occurred more frequently and next to the traditional sources of synergy in cross-border acquisitions, such as tax advantages, market access, labor and market recourses (Moeller & Schlingemann, 2005), La Porta et. al (1998) identified an additional synergy factor; namely spillover effects due to differences in corporate governance systems, legal systems and consequently shareholder protection. Triggered by La Porta et. al (1998) work, the economical effect of shareholder protection received a large amount of attention by researchers. A new stream of research was introduced in the beginning of the 21st century, focusing on the impact of law and finance on mergers and acquisitions (Martynova & Renneboog, 2011b).

(7)

(2010) investigated the effect of shareholder protection on the return of bidding firms. They used a sample of American bidders acquiring cross-border targets. Hangendorff et al. (2007) studied M&As between financial institution in different legal environments Goergen & Renneboog (2004) and Martynova & Renneboog, (2011a) in their empirical researched compared the return of targets, acquirers and the complete deal in M&As taking place in different legal environment within Europe between 1993 and 2001, whereas, Bris & Cabolis (2008) and Kuipers et al, (2009) investigated a similar time period as the previous authors but focused on worldwide M&As. They compared as well acquirers’, targets’ and complete deal returns.

These studies presented above describe the relationship between shareholder protection and the value effect on M&As, but come to different results. On the one hand,, Goergen & Renneboog (2004); Moeller & Schlingemann (2005); Kuipers et al. (2009) and Martynova & Renneboog, (2011a) find that the acquirer’s return is higher when the target is located in a country with high level of shareholder protection. They explained that finding, by higher level of transparency, lower agency and transaction costs. These findings confirm the theorized relationship of La Porta et. al (1998). On the other hand, Hangendorff et al. (2007); Francis et al. (2008) and John et.al. (2010) find that the acquirers benefit more when they acquire a target located in a country with lower level of shareholder protection. The main argument here is, that in a country with lower shareholder protection the M&A activity is lower and there are less bidding wars, due to less competition between acquirer. So it is easier to find undervalued targets, and due to less bidding wars the premium will be ultimately lower.

As the short discussion above shows there is discrepancy in the literature and empirical findings are contradicting. One motivation for this master thesis is therefore to shed some light on this discrepancy and to find out what benefits acquiring and target’s shareholder the most.

(8)

“To what extend do legal environment and the degree of investor protection have a value enhancing effect on the complete deal, the acquirer’s value and the target’s value?”

In order to answer the aforementioned question, data from 332 public targets and 332 acquirers involved in domestic and cross-border M&As is used. The data is extracted from Zephyr database of Bureau van Dijk and covers the period from 2002 until 2011. The sample period was chosen due to the topicality and thus, no research has analyzed before this sample period in respect to the legal origin and shareholder protection effect on M&As. In addition to answering the research question, the master thesis will therefore, also shed more light on recent developments.

The main results of the thesis are that targets, acquirers and the complete deal have positive abnormal return at the announcement period. Targets outperform the acquirer. Secondly, common law acquirers, common law targets and common law deals have higher returns around the announcement date than civil law acquirers, targets and deals. Deals in German civil law countries create more value than French civil law country deals. There was no significant effect of concentrate ownership structure in common law countries, but civil law countries present a negative effect, which is in line with the findings of Faccio & Stolin, (2006).

In domestic M&As acquirer, targets and complete deal create more value in countries with high level of shareholder protection than in countries with low shareholder protection. In cross-border deals, acquirer return is higher if the firm acquires a target situated in a country with high level of shareholder protection. Targets receive a higher return at the announcement date, when the acquirer originates from a country with lower shareholder protection. The added value of these findings are twofold. Firstly, the previous research focused on Europe or US, whereas this thesis takes a broader sample including all OECD countries. Secondly, the sample period in the previous research was in 1990s till 2002. In this paper a more recent period 2002 till 2011 is investigated.

(9)

methodology. Section five will present and discuss the results of the empirical analysis. The final sixth section provides the conclusion and limitations of this paper.

2. Literature review and hypotheses

This segment begins with a general overview about the returns on the announcement date for acquirer, targets and the complete deal. Then, the effects of different legal environments of M&A’s return are analyzed. Next, a discussion of the relevant literature on the relationship of investor protection and M&As is presented. Hypothesis will be specified in turn, based on the existing theoretical and previously proven empirical relationships between the variables for the regression, whereby the focus lies on the change of shareholder wealth in the acquiring and in the target firms, as well as the effect on wealth changes in the complete deal. This section ends with the examination of relevant determinants for the cross-sectional analysis.

2.1 Acquisition return on the announcement date

This section analyzes empirical findings of acquirer’s and target’s return, as well as the wealth effect on the complete deal. Table 1 below presents the empirical findings about abnormal returns of the acquiring and target firms as well as from the complete deal around the M&A’s announcement date. At the end of this paragraph hypothesis are built.

(10)

TABLE 1EMPIRICAL RESEARCH THAT SHOWED WEALTH EFFECT ON ACQUIRER, TARGET AND COMPLETE DEAL IN THE ANNOUNCEMENT PERIOD

Study Period Event

window

Sample size

Acquirer origin

Target origin Acquirer return Target return Complete Deal Goergen & Renneboog (2004) 1993-2000 (-1;0) 228 Europe Europe +0.7% 9.01% 4.00% Moeller & Schlingeman (2005) 1985-1995 (-1;+1) 4430 US UK, Canada, France Germany -0.87% N/A N/A

Campa & Hernando (2005)

1998- 2000

(-1;+1) 262 Europe Europe -1.48% N/A N/A

John et. al. (2010) 1984– 2005

(-1;+1) 1525 US firms Worldwide non- US firm 0.33% N/A N/A Martynova, Renneboog ( 2008) 1993-2001

(-1;+1) 2419 Europe Europe 0.83% 12.55% N/A

Martynova, Renneboog ( 2011a)

1993-2001

(-1;+1) 2419 Europe Europe 0.72% 12.47% Positive

Mentz & Schiereck (2008)

1981- 2004

(-1;+1) 192 Worldwide Worldwide 1.26% 20.13% N/A

Feito-Ruiz, Requejo (2009)

2002-2006

(-1;+1) 469 Europe Worldwide 0,99% N/A N/A

Starks & Wei (2004) 1980-1998

(-5;+5) 1781 Worldwide US firms 1.84% 23.45% N/A

Kuipers et al, (2009) 1982-1991

(-1;0) 181 Worldwide US firms -0.92 23.07% 2.99%

Bris & Cabolis(2008) 1989- 2002

(-1;+1) 506 Worldwide Worldwide -1.12% 14.2% N/A

(11)

On average, the margin ranged between 10 percent till 30 percent higher returns than the pre-acquisition announcement price (Martynova & Renneboog, 2011a). Additionally, the research of Mentz & Schiereck (2008) investigated 191 acquisitions in the period of 1981 till 2004 and found an 18 percent higher target than acquirer return. Furthermore, Goergen & Renneboog (2004) studied the European market and concluded that target shareholder gained more than acquirer shareholder. However, they showed as well that the complete M&A return was positive. Finally, Martynova & Renneboog (2011a) came to the same conclusion as their peers. In their study on the European M&A market from 1993 till 2001, they found that target firms had 10 percent higher returns than the acquirer company, but also the overall return of target and acquirer was positive. Finally, Goergen & Renneboog (2004) and Martynova & Renneboog (2011a) come to the conclusion that on the European market, the majority M&As are motivated by synergies rather than agency problems or hubris. Based on the above discussion the following hypotheses are built:

Hypothesis 1a: Acquirer will experience a positive abnormal return around the

announcement date of a merger and acquisition.

Hypothesis 1b: Targets will have a higher abnormal return than acquirer around the

announcement date of a merger and acquisition.

Hypothesis 1c: The complete deal will be positive around the announcement date of a

merger and acquisition.

2.2 Merger and Acquisition in different legal environments

This sub-section of the paper deals with the different legal environments of the sample. The relation between legal system and M&A return is presented. The relevant literature is discussed and based on this discussion hypotheses are built.

(12)

law and stresses rules, protecting all stakeholders: shareholders as well as creditors, employees, and others (Hall & Soskice, 2001). In domestic and cross-border acquisitions, variations in institutional environments, such as legislations and regulations are a cause of difference in acquirer shareholder wealth gains (Kuipers et al.,2002).

La Porta et al. (1997) performed an empirical study in 49 countries on the level of investor protection and creditor protection. They categorized the 49 countries in four groups: common law, Scandinavian civil law, German civil law and French civil law. On the one hand, they conclude that common law regimes provide a better investor and creditor protection than the civil law regimes. Further, they find that managers’ interest are better aligned with shareholders’ interests in the common law system than in the civil law system. But there are differences between civil law countries too. Scandinavian civil law offers better protection than German, and German better than French civil law. On the other hand, it is also theorized by La Porta et al. (1998) that countries with a lower investor protection may substitute this disadvantage with different instruments of corporate governance. Therefore, they predict higher ownership concentration in civil law countries.

Hypothesis 2a: German civil law countries have higher deal returns than French civil

law countries.1

Faccio & Lang (2002) confirm this hypothesis with their study of ownership concentration, using a sample of 5232 corporations. According to Faccio & Lang (2002), most firms, which originate from civil law regimes, are mainly owned by a few big shareholders, whereas in the common law countries, firms are characterized by a large amount of minority shareholders. Goergen & Renneboog (2004) found in their empirical study that United States firms have a higher spread ownership compared to firms from Continental Europe. La Porta et. al. (2006) developed a country index for ownership concentration. They defined ownership concentration as the average percentage of

(13)

common shares hold by the top three investors. In this index, common law countries have an average concentrated ownership of 44 percent, whereas civil law countries have an average concentrated ownership of 49 percent. Due to poorer legal protection of minority shareholder rights in civil law countries, large block-holders may use acquisitions as a tool to transfer wealth from minority shareholders to themselves (Faccio & Stolin, 2006). The likely reaction of minority shareholders is fear of potential expropriation; therefore the market may react negatively on the announcements of M&As by civil law country firms with concentrated ownership. In contrast, in common law countries minority shareholder rights are well protected by law, and block-holders are seen as an instrument to monitor the management decisions on M&As and prevent M&As to happen, which are motivated by managerial hubris (Martynova & Renneboog, 2009).

Hypothesis 2b: Concentrated ownership in Common law deals will create value for the

complete deal.

Hypothesis 2c: Concentrated ownership in civil law deals will destroy value for the

complete deal.

(14)

civil law countries. Dombret et al (2008) researched 5149 domestic and cross-border M&As in common law countries (i.e. UK and USA) with 549 domestic and cross-border M&As in civil law countries (i.e. Germany and France) in the period 1996 till 2004. They found that bidder from common law countries have higher abnormal return than bidders from civil law countries. Based on the discussion above the following hypothesis can be formed:

Hypothesis 2d: Common law targets have higher abnormal returns than civil law

targets.

Hypothesis 2e: Common law acquirers have higher abnormal returns than civil law

acquirers

Hypothesis 2f: Deals between Common law targets and Common law acquirer have

higher abnormal returns than deals between civil law targets and civil law acquirer.

2.3 Merger and Acquisition and shareholder protection

(15)

protection economies, it is more likely to have bidding wars, eventually driving up target returns and potential gains for acquirers are bid away. Francis et al. (2008) investigated worldwide M&As focusing on companies from the US in the period of 1990 till 2003. The results indicate that the bidder return was higher when the target originated from a country with lower investor protection. Conn et al. (2005) examine a global sample of 4,344 domestic and cross-border M&As between 1984 and 1998 done by an acquiring firms from the UK. They find that acquirer return is higher in countries with lower shareholder protection. Hagendorff et al. (2007). analyze domestic and cross-border M&As of public companies listed in the USA and Europe. The authors report that acquirers realize higher returns when the target is situated in a country with a lower level of legal shareholder protection. The explanation they give for their finding is that due to higher risk of expropriation by insiders (agency costs), which they encounter in poor shareholder protection environments, bidders need to be compensated. John et.al. (2010) performed an empirical study of 1525 cross-border M&As by US firms in the time period from 1985 till 2005. They researched the announcement period acquirer returns and target values. The emphasis of the study laid on the degree of minority shareholder protection in the target country. As a result, John et al (2010) state that acquirer’s returns are significantly negative when the shareholder protection is high in the target country and the acquirer’s returns are significantly positive when the targets originate from low shareholder protection countries. These findings were confirmed by a cross-sectional analysis with other firm variables. Feito-Ruiz & Requejo (2009) conducted a research on 469 worldwide, domestic and cross-border M&As over the period 2002 till 2006. They find that higher levels of shareholder protection in the acquirer’s country, than in the target’s country, result in significantly positive acquirer’s abnormal return. Pablo (2008) analyzed a sample of 952 domestic and cross-border acquisition in which at least one of the two parties was from Latin America in the period of 1998 and 2004. He finds a significant positive effect on bidders’ return, when the target company is located in a country with lower level of shareholder protection.

(16)

company. As La Porta et al (2002) show in their empirical research, also target companies can benefit from that. Higher shareholder protection translates in lower agency costs and therefore leads to higher target company valuations. Moeller & Schlingemann (2005) discovered that firms, which acquire targets from countries with high level of legal protection for shareholders, gain higher abnormal returns on the announcement date of the M&A compared to targets from countries with low level shareholder protection. A supporting empirical result to the previous three studies is found by Martynova & Renneboog (2008), who researched 2,419 domestic and cross-border M&As in 29 European countries during the time period 1993 till 2001. They ﬁnd significant evidence that target firms’ cumulative abnormal return increases with higher level of shareholder protection in the target country. Furthermore, the acquiring company, which originated in a lower institutional economy than the target company, will adopt voluntarily the higher corporate governance practices of the target company, resulting in higher value for shareholders of the bidder company (Martynova & Renneboog, 2008). Starks & Wei (2004) investigated the announcement effects of cross-border takeovers of US targets. They ﬁnd that target returns are higher, when the acquirer is located in a country which offers lower minority shareholder protection. Furthermore, if the acquirer comes from a country with lower level of shareholder protection than the target, and the method of payment in the deal is equity the shareholders of the target company demand for a higher premium to compensate for poorer corporate governance after the acquisition is finalized.

(17)

for targets, when the acquirer comes from a country with a higher level of shareholder protection than from a country with a lower level shareholder protection. They explain this finding by the concentrated ownership of the target company in these countries. The majority shareholders of the target company have higher bargaining power and demand higher premium for their shares.

TABLE 2: HYPOTHESES FOR SHAREHOLDER PROTECTION

Hypothesis Support in literature

Hypothesis 3a: If the target originates from a country

with a higher (lower) shareholder protection, the acquiring firm’s shareholders’ return will be higher (lower) at the announcement of the M&A.

Dahlquist et al. (2003); Kuipers et al, (2009) Martynova & Renneboog (2011a) Moeller &

Schlingemann (2005)

Hypothesis 3b: If the acquirer comes from a country

with a weaker (stronger) legal and institutional environment, the target shareholders’ return will be higher (lower) at the announcement of the M&A.

La Porta et al (2002); Starks & Wei (2004) Martynova & Renneboog (2008); Martynova & Renneboog (2011a) Hypothesis 3c: If the target and the acquirer both come

from a country with a strong (weak) shareholder protection, the complete deal value will be higher (lower) at the announcement of the M&A.

Kuipers et al, (2009) Martynova & Renneboog (2011a)

Hypothesis 3d: If the target and the acquirer both come

from a country with a strong shareholder protection, the returns of the acquirer will be higher at the announcement of the M&A than if both come from a country with low shareholder protection.

La Porta et al (2002); Bris & Cabolis (2008)

Hypothesis 3e: If the target and the acquirer both come

from a country with a strong shareholder protection, the returns of the target will be higher at the announcement of the M&A than if both come from a country with low shareholder protection.

(18)

2.4 The determinants of cross-sectional analysis

In this section of the paper, a variety of determinants, which have been found theoretically and empirically relevant in the M&A literature to influence takeover announcements return, will be introduced. The definition and measurement for each control variable is provided in the data section. Furthermore, in Table 3 the determinants, the related theory, the expected effect, as well as the relevant literature are presented.

To begin with, bidder size is negatively correlated with the short-term wealth effect captured by acquiring firm’s shareholder. Moeller et al. (2004) analyzed a sample of 12,023 M&As in the time period 1980-2001. They found that larger acquiring firms pay higher premiums and make complete deals produce negative synergies. On the contrary, smaller bidders achieve gains for the complete deal and generate about two percentage points higher returns than larger acquirer. Consequently, target companies would benefit from larger bidders and would earn higher returns.

Relative size of the target is as well a characteristic which might influence announcement returns of the participants in the deal. Hansen (1987) states that with larger relative size of the target company, the probability of a potential revaluation increases. This probability of revaluation increases the uncertainty about the true market value of the target company and that ultimately leads to negative effect on acquiring firm’s announcement return.

Furthermore, larger targets have higher post-acquisition integration costs; therefore the acquirer’s announcement return is expected to decrease. Contrary, the assumption of higher synergy effects and lower post-acquisition integration costs makes the acquirer to pay a higher premium and therefore, leads to higher target company’s announcement return.

(19)

TABLE 3:RELEVANT CONTROL VARIABLES

Determined Relevant theory Exp.

effect Target Exp. effect Acquirer Exp. effect deal Support in Literature

Bidder Size Hubris (Roll, 1986), empire building

+ - - Moeller et

al. (2004) Relative

size

Higher uncertainty about the target's true market value; complex managerial

structure; Inflated integration costs + - - John et. al (2010) Martynova, Renneboog (2011a) Acquirer Market to book ratio

Hubris (Roll, 1986) + - - Martynova &

Renneboog (2008;2011a) Target

Market to book ratio

Hubris (Roll, 1986) + - - Martynova,

Renneboog (2008;2011a) Acquirer

Leverage

Theory of capital structure Free cash flow hypothesis Restriction to financing with Cash

- + + Lang et

al.(1991)

Target leverage

Theory of capital structure Free cash flow hypothesis Restriction to financing with Cash - + + Martynova & Renneboog (2011a) Transaction Value Agency problem, misevaluation - - - Alexandridis et al. (2010) Method of payment Pecking-order theory Signaling theory Asymmetric information hypothesis + + + Goergen & Renneboog (2004) Related Acquisition Focus vs. diversification strategy of the bidder

+ - - Martynova,

Renneboog (2008; 2011) Cross

border

synergies, which are not available to domestic deals

+ - - Conn et al.

(2005) Degree of

economic Freedom

Lower transaction cost, less market restrictions, lower taxes + + + Feito-Ruiz, Requejo, (2009). Initial stake in target Transparency, more information available.

Premium for lower number of shares.

- + + Betton &

(20)

Goergen & Renneboog (2004) find that the market-to-book ratio of the target matters in terms of the bid premium, a high target’s market-to-book ratio is correlated with higher returns for the target and lower returns for the acquirer.

Lang et al. (1991) studied the relationship between free cash flow and leverage empirically. The authors find that companies which have low leverage and high free cash flow face significantly more agency costs between management and shareholders, as well as agency costs between majority and minority shareholders. As Pinkowitz et. al. (2006) argue, liquid assets are worth less to minority shareholders, because it is easier for majority shareholder to use these liquid assets for private benefits. Masulis et al. (2007) show that high leverage will lead to lower future free cash flows and therefore reduce agency costs. Furthermore, they continue that high leverage gives the manager incentive to improve firm performance, due to high monitoring by creditors. Besides acquirer’s leverage also target leverage plays an important role on the wealth effect of a M&A. Martynova & Renneboog (2011a) state that bidders are likely to pay higher premiums for a target with lower leverage and higher cash flows, because targets will have less obligations towards creditors and more liquid assets which the new acquirer takes control over.

Alexandridis et al. (2010) report in their findings a significant negative relationship between transaction value and the premium paid for the target. Despite the lower premium paid for the targets, they document that large transactions lead to acquirer’s shareholder value decrease. The explanation the authors give is that investors evaluate large deals as risky projects with high levels of uncertainty. The risk is due to high potential synergies that cannot be realized, because of high post integration costs. Moeller et al. (2005) as well as Martynova & Renneboog (2011a) conducted empirical research on that matter and found a value reducing relationship for the acquirer. The reason they mentioned was that high integration costs and complex management capabilities are needed.

(21)

Andrade et al., 2001; Moeller et al, 2004; Martynova & Renneboog, 2009). This finding is in line with Myers & Majuf (1984) finding, who state that managers offer equity of the firm when they think that the firm’s equity is overvalued, and they offer cash when they think that the firm’s equity is undervalued.

To acquire a target firm within the same industry is a focus strategy, whereas diversification is to acquire a company outside of the own industry. In general it is assumed that a diversification strategy destroys value, and a focus strategy creates value for the bidder (Doukas & Kan, 2004; Martynova & Renneboog, 2006). Contrary, the return for the shareholders of the target firm is likely to be positively related to the diversification strategy of the acquirer. The reason is that acquirers with diversification strategies are more likely to bid more aggressively and therefore, pay higher premiums for the targets (Martynova & Renneboog, 2011a).

Acquiring firms and target firms can both benefit from cross-border M&As, due to synergies, which are not available to domestic deals. These synergies might be international product and capital markets (Hymer, 1976), or R&D capacity of a foreign firm (Eun et. al., 1996). Goergen & Renneboog (2004) performed an empirical study on domestic and cross-border M&As and found that cross-border targets receive higher abnormal returns than domestic targets. However, Schoenberg (1999) theorized that cultural difference between acquiring and target country can lead to difficulties in the integration process. Therefore, the author expects M&A’s gains to decrease. Conn et al. (2005) and Moeller & Schlingemann (2005) found confirming empirical evidence for that hypothesis. In these two empirical studies, the domestic acquirer and domestic deal had higher returns than the cross-border acquirer and deals.

Economic freedom is an index created by the Heritage foundation. It measures the degree of economic freedom within a country. Several researchers use this variable in their study about M&As (e.g. Moeller & Schlingemann, 2005; Francis et al., 2008; Feito-Ruiz, Requejo, 2009). They find a positive relation between high economic freedom and target, bidder and deal return.

(22)

the other hand the bigger the prior stake is, the lower the premium, because they have to pay a premium on less shares. Betton & Eckbo (2000) conducted an empirical study and found evidence that shareholder of US target firms receive significant lower premium, when the acquirer had a prior toehold in the target.

3. Data

This part of the thesis describes the data used in the empirical research. Selection criteria and data source are show in section 3.1. The empirical framework of the study which describes the data can be found in section 3.2. Section 3.3 gives an overview of the variables used in the thesis.

3.1 Data source

The sample for this thesis has been obtained from the Zephyr database of Bureau van Dijk. M&As are included that were announced and completed between 01.01.2002 and 31.12.2011 between listed companies located in the OECD countries. This time period was chosen, because it is the most recent period with available data. Furthermore, a number of related studies (e.g. Kuipers et al, (2009); Bris & Cabolis (2008); John et.al. (2010) Martynova & Renneboog (2011a)) researched the time period 1989 till 2002. The database DataStream 5.1 by Thomson Reuters provided data for variables needed in cross-sectional analysis, as well as for the calculation of daily stock returns and market index returns. For the selection of the sample the criteria in Table 4 have been used:

- Both the acquiring and target firm were listed on stock exchanges and one of them had a headquarter in an OECD member country;

(23)

- On DataStream 5.1 the 574 events were cross-checked if they provided price and return data for the acquiring, as well as the target firms. Eventually, data for 250 trading days before and to 20 trading days after the announcement of the acquisition and merger was needed;

- After the cross-checking with DataStream, a sample size of 332 was left. The characteristics of the sample are outlined in the upcoming section.

TABLE 4SAMPLE SELECTION CRITERIA

Listed Acquirer and Listed Target 55,927

Deal type: Merger & Acquisition 4,623

Deal status: completed 884

Time period : 01.01.2002 till 31.12.2011 828

World region: OECD 574

Cross-checking information with

DataStream 5.1 and heritage foundation, indices

332

3.2 Sample characteristics

(24)

and target come from different legal systems. Civil law and common law deals are evenly spread; nonetheless, only 18 out of 332 deals are mixed deals. In the first period from 2002 till 2006 only 2 deals were between the two legal systems but in the second period there have been 16 deals.

TABLE 5SAMPLE DESCRIPTION DEAL VALUE AND NUMBER OF DEALS

Year Deals total Civil law deals Common law deals Mixed deals Cross-border (domestic) deals Deal Value (million EUR) Average deal value (in million EUR) Median Deal Value (in Maximum deal value (in Millions EUR) 2002 13 9 4 0 2 (11) 680 52.11 13.59 329 2003 22 11 11 0 6 (16) 940 38.96 9.77 292 2004 29 11 17 1 9 (20) 12,031 414.95 10.37 8,130 2005 29 13 16 0 9 (20) 14,410 496.62 1.94 5,048 2006 22 12 9 1 5 (17) 3,410 154.85 22.37 1,941 2007 48 28 17 3 16 (32) 25,690 535.15 11.08 17,266 2008 38 16 18 4 9 (29) 5,690 149.95 16.41 1,290 2009 37 19 16 2 7 (30) 1,180 31.85 6.95 268 2010 31 19 10 2 5 (26) 1,720 55.40 7.25 492 2011 63 28 30 5 12 (51) 41,880 664.80 84.12 6,462 2002-2011 332 166 148 18 80(252) 108.86 327.91 16.87 17,266

TABLE 6 LEGAL SYSTEMS OF TARGET AND ACQUIRER

Target Acquirer French civil law German civil law Scandinavian civil law Common law Total

French civil law 44 8 2 5 59

German civil law 9 78 1 4 92

Scand. civil law 0 0 12 0 12

Common law 2 6 2 159 169

Total 55 92 17 168 332

(25)

the most active between the different legal systems. It has 15 cross-legal acquisition, the German civil law system has 14, the common law system 10 and the Scandinavian civil law system none.

3.3 Variables

This thesis uses two dependent variables. One is chosen to represent the stockholder wealth changes around the announcement date, namely the average cumulative abnormal return. The other dependent variable is cumulative abnormal return for each event used when making the regression analysis. The descriptive statistics of the daily average abnormal return in the event period are presented in Appendix C. It can be seen that the Average abnormal returns for targets and acquirer during the announcement period are significant at the five percent significance level.

(26)

the highest level of shareholder protection, followed by Scandinavian civil law. German civil law and French civil law countries have the lowest level of shareholder protection. This study naturally also uses control variables which are: bidder size (Size), relative size of the target to the bidder (Rel), bidder market to book ratio (AMTB), Target market to book ratio (TMTB), acquirer leverage (ALev), target leverage (TARLev), transaction value (Deal_size), method of payment (Paym), focus or diversification (foc) cross-border versus domestic acquisition(CBA), Target economic freedom (ECO) as well as prior stake in the target (stake). Control variables help to more precisely identify which reasons and motivation are behind the changes in stockholder‘s wealth. Appendix A provides information about the calculation and estimations, as well as the gathering of data and sources of the control variables methodological calculations.

TABLE 7DESCRIPTIVE STATISTICS OF CONTROL VARIALBLES

Control variables Mean Median Standard deviation

Bidder size nominal in Million EUR 524.65 26.17 2,514 Bidder size 13.7 13.8 3.89 Relative size 0.44 0.46 0.99 Acquirer Market to book value 0.17 0.13 0.40 Target Market to book value 0.23 0.17 0.49 Acquirer Leverage 1.15 0.21 0.36 Target Leverage 1.31 0,17 0,31 Transaction value in Million EUR 327.91 16.87 1,306 Method of payment 0.39 0 0.49 Focus vs diversification 0.67 1 0.47 Cross-border vs Domestic 0.24 0 0.43 Target Economic freedom 72.82 74.12 7.58

Initial stake in target firm

(27)

The descriptive statistics are highlighted in Table 7 above and outlined in turn. To begin with, there exist a large difference between mean (524.65 million EUR) and median (26.17) million EUR) values of acquirer’s size. Consequently, log of pre-merger market capitalization of the acquirer is used. The target companies are smaller and have a market capitalization of 45 percent of the one of the acquirer, which can be seen in the variable relative size. It is observed that the acquirer have a lower market to book value (0.17) than the targets (0.23). That means there are more value acquirers than value targets in this sample. The targets leverage with 61 percent is almost double as high as the acquirer’s leverage of 31percent. The deal value mean of 327.91 million EUR differs remarkably from the deal value median of 16.87 million EUR. The variable method of payment shows that in the sample 39% of the deals are paid with only cash, and the other 61% are pure equity payments or a mixture of different means of payments. Two-thirds of the acquisitions are made within the same industry and 252 acquisitions are domestic acquisitions and just 80 acquisitions are cross-border deals. The average economic freedom in the sample is 72.81, which is a relative high value and shows a great deal of freedom and development of the markets, but which was to expect, because the sample firms were chosen from OECD countries. Finally, just 29% of the acquirer had a premerger stake in the target company; this is in line with the findings of Martynova & Renneboog (2011a).

4. Methodology

(28)

Ruback, 1983). As McWilliams & Siegel (1997) state, an alternative would be to look at accounting based measures, which could be easily manipulated by subjective choices of different accounting methods. Furthermore, Bruner (2002) concludes that the event study’s measurement of market reaction demonstrates the imminent amount of value created for shareholders and is the best possible estimated outlook for future value creation. The basis of the execution of the event study, as well as the calculation of the abnormal return at the announcement date, are the studies of Brown & Warner (1985), McKinley (1997), and Bruner (2002).

4.1 Estimation and event windows

The timeline which is used in this study of the M&A announcements of the 332 events is demonstrated in Figure 1 The day of announcement is treated as day zero in each event (MacKinley, 1997). The estimation period will be as in the study of Starks & Wei (2004), Mentz & Schiereck (2008) and Feito-Ruiz & Requejo (2009) from 250 days before the deal till 20 days before the announcement of the deal. Consequently, this period is used as estimation for the market parameters, for identification of normal movement in the stock, as well as correlation with an outside stock index, the MSCI world index (Brown & Warner, 1985). The actual event window starts 20 days prior to the announcement and finishes 20 days after the event. In this event window, the stock prices of the target and acquiring company are investigated.

FIGURE 1. TIME FRAME OF THE STUDY

(29)

4.2 Benchmark and abnormal returns

The next step in the event study is to choose an appropriate benchmark. In this thesis, the MSCI World Index by Morgan Stanley Capital International was chosen. The index covers stock data from 24 countries, which are all represented in the data set. Furthermore, 79% of this sample targets and acquirers originate in one of these 24 countries.

The following step, according to MacKinley (1997) is to determine the abnormal return of the acquirer and the target company (see Equation 1). That is calculated by subtracting the expected stock return in the estimation period from the actual return stock return. The calculation for the actual return is demonstrated in Equation 2, and the calculation for the expected return is shown in Equation 3. The closing share prices from the companies are derived from the database DataStream. As suggested by Henderson & Glenn (1990), this thesis uses daily continuously compounded stock returns instead of simple stock returns because it is assumed they conform better to the normality assumption.

( )

(1)

Where , and ( ) are abnormal, actual and expected normal return of the company i for the day t respectively.

(2)

Where stands for company’ i closing stock price on day t and represents company i stock price of the trading day prior to day t.

(30)

Following, other researchers such as Martynova & Renneboog (2011a), Bris& Cabolis (2008), Conn et. al. (2005) and the suggestion of Brown &Warner (1985), the focal thesis measures the expected return of the companies, using the residual market model. Equation 3 highlights the statistics of the expected normal return.

( )

(3)

Where and are company specific coefficients estimated with capital asset pricing model for 230 trading days (-250;-20 days prior to deal announcement). is an intercept coefficient, is a slope coefficient that captures company’s systematic risk and denotes the total return on the MSCI world index at day t.

(4)

Where is covariance between firm share and the MSCI world index and represents the market portfolio variance.

(31)

∑ (6) ̅̅̅̅̅̅ _∑ (7) 4.3 Tests

According to Henderson & Glenn (1990), the following assumptions are the cornerstone for the event study‘s methodology and help to understand the analysis of the regression models. Residuals of regression models are assumed to be normally distributed, secondly, not to be serially correlated, which is not to expect in a cross-sectional analysis; thirdly to show a constant variance; and lastly not to show any correlation with explanatory variables. In this thesis the Jarque-Bera test is used to check the distribution of the residuals. One parametric and one non-parametric test are employed, because the abnormal and the cumulative abnormal return by themselves are not satisfactory enough to draw reliable conclusions.

4.3.1 Parametric tests

The assumptions in this thesis for the standardized traditional test are the following; the stock residuals are uncorrelated and insignificant event induced variance, as well as that the abnormal returns have the same variance (Brooks, 2008). That is accomplished by dividing each company’s abnormal residual, measured in the event window, by its standard deviation, obtained from the estimation window (see Equation 8). The Equation 9 demonstrates the testing of AR and the Equation 10 describes the examination of CARs.

(32)

̅̅̅̅̅̅ ̅̅̅̅̅̅̅̅̅̅ ( ) ∑ ̅̅̅̅̅̅̅̅ (9) ∑ ̿̿̿̿̿ √ ∑ ̅̅̅̅̅̅ (10)

Some of the hypotheses in this thesis are related to differences between the mean differences of the cumulative abnormal returns of different sub-samples. To test if these differences are significant, this paper follows Dodd & Warner (1983). The formula is highlighted in Equation 11, where CAR1 and CAR2 stands for the different sub-sample cumulative abnormal returns, L stands for event window period and N stands for number of observations.

̅̅̅̅̅̅ ̅̅̅̅̅̅ √

(11)

4.3.2 Non-parametric test

(33)

To each daily return in this period a rank will be assigned. The rank of the abnormal return in the time series is symbolized as , whereby the lowest return will be ranked 1 and the highest return will be ranked 291, the medium rank is logically 146. With Equation 12, the rank test statistic is measured for each cumulative abnormal return.

̅̅̅ [∑ ̅̅̅ ⁄

]

(12) Where KL represents the average rank across the n shares and days l of the event window (1≤l≤41) and Kt signifies the average rank across the n stocks on day t of the combined estimation and event period (291 days).

4.4 Cross-sectional analysis

The multiple cross-section regression analysis is used to examine the hypothesizes (Equation 13). It will check upon the effect of the independent variables, as well as what effect control variables might have.

CARi=α0+ α1Common + α2*ANTI + α3*DISC+ α4*LIAB+ α5*PRO+ α6*CBA + α7*OWNER+ α8*Foc + α9*stake + α10*Size+ α11*ACQ_ Lev + α12*TAR_Lev + α13*Deal_size + α14*Rel + α15*Paym + α16*ACQ_MTB + α17*TAR_MTB +α18*ECO εi

(13)

(34)

cross-border and set to 0 when domestic M&A; Owner is a dummy variable that equals 1 when the company comes from a country with high ownership concentration and otherwise 0; foc is a dummy variable set to 1 when acquisition is in same industry otherwise 0; stake is a dummy variable set to 1 when the acquirer had a stake in the target company prior to the deal otherwise 0; Size is log of pre-merger market capitalization of the companies; ACQ_Lev is acquirer’s ratio of total book value of debt to total book value of assets, both measured at the year prior to the deal announcement; TAR_Lev is target’s ratio of total book value of debt to total book value of assets, both measured at the year prior to the deal announcement; Deal_size is the price paid for the target; Rel is target’s market capitalization divided by the acquirer’s market capitalization; Paym is a dummy variable set to 1 when deal is entirely financed with cash and 0 otherwise; ACQ_MTB is acquirer’s market value of equity plus book value of debt over the sum of book value of equity plus book value of debt prior to the bid; TAR_MTB is target’s market value of equity plus book value of debt over the sum of book value of equity plus book value of debt prior to the bid; ECO is a dummy variable that equals 1 when the economic freedom is higher than the median, otherwise 0.

The results of ordinary least squares regressions are accompanied with several statistical terms to support the analysis. To begin with, the R-Squared statistics will show how reliable the explanation of the relationship between the variables and the CAR is. F-statistics whose calculation results from the sample variances demonstrate how the variability of the model is described. The multicollinearity problem of the regression is identified by the Pearson correlation factor.

5. Empirical results

(35)

effects in sub-samples. In the last section the effects of the control and independent variables in the OLS-regressions will be presented.

5.1 Shareholder wealth changes surrounding M&A announcement

In figure 3 the results of the average daily abnormal returns of the target and the acquirer are presented. The calculation for the abnormal returns in this figure can be found in subsection 4.2 and the results for the standardized traditional test and for the Corrado Rank test can be found in the Appendix C, as well as the parametric and non-parametric tests. Figure 2 demonstrates that the return of the targets is higher than the ones of the acquirer. On the day of the announcement the target has a significant positive abnormal return of 4.9%, whereas the acquirer achieves a positive abnormal return of 1.8%.

FIGURE 2:AVERAGE DAILY ABNORMAL RETURNS OF TARGET AND ACQUIRER IN EVENT WINDOW

In the whole event window from 20 days prior to the announcement of the deal till 20 days after the announcement of the deal the cumulative average daily abnormal return

-2.00% -1.00% 0.00% 1.00% 2.00% 3.00% 4.00% 5.00% 6.00% AR-20 AR-19 AR-18 AR-17 AR-16 AR-15 AR-14 AR-13 AR-12 AR-11 AR-10 AR-9 AR-8 AR-7 AR-6 AR-5 AR-4 AR-3 AR-2 AR-1 AR-0

AR1 AR2 AR3 AR4 AR5 AR6 AR7 AR8 AR9 _AR1

0 AR1 1 AR1 2 AR1 3 AR1 4 AR1 5 AR 1 6 AR1 7 AR1 8 AR1 9 AR2 0

Average Daily Abnormal Return

(36)

is positive. That shows that the deals were value-creating deals for target’s and acquirer’s shareholders. Furthermore, out of these 41 days event window the targets and the acquirer have on 67.6% of the days and respectively 56.9% of the days a positive abnormal return.

In this research an event window of three days is chosen to test the hypotheses. The reason is that in the time period -1 day till +1 day, each daily average abnormal return is significant (see Appendix C ). To test if there are any pre-announcement effects, the event window (-20;-2) is chosen. Considering to investigate post-announcement effects, the event window (+2; +20) is elected. The results of the analysis of these three windows are presented in table 8 for targets and in table 9 for acquirers.

TABLE 8: TARGET CAR DURING DIFFERENT EVENT WINDOWS

Event Window Average CAR Standardized T-test Jarque- Bera Percent positive Median CAR Rank test CAR -1;+1 6.67% 6.23* 4,419* 61,1% 0.88% 6.14* CAR -20;-2 3.64% 2.39** 4,850* 49,2% 0.15% 0.73 CAR +2;+20 1.37% 1.02 890,* 42,2% 0.08% 1.39 CAR -20;+20 16.81% 7.57* 888* 67.6% 7.04% 7.93*

Note: * 1% significance level; ** 5% significance level; *** 10% significance level.

(37)

Martynova & Renneboog (2011a) state M&As come in waves and the fifth wave in the 1990s were characterized by big deals. Moeller & Schlingemann (2004) find that large deals might benefit the targets.

TABLE 9ACQUIRER CAR DURING DIFFERENT EVENT WINDOWS

Event Window Average CAR Standardized T-test Jarque- Bera Percent positive Median

CAR Rank test

CAR -1;+1 2.63 % 3.05* 20,416* 52.1% 0.12% 1.86*** CAR -20;-2 2.3% 1.74*** 14,132* 47.8% 0.12% 0.28 CAR +2;+20 0.92% 0.77 1,327* 46.4% 0.01% 1.38 CAR -20;+20 10.4% 4.91* 1,884 56.9% 2.51% 3.46

Note: * 1% significance level; ** 5% significance level; *** 10% significance level.

(38)

1.74), and no significant support from the non-parametric tests (rank test 0.28). A positive post-announcement drift can be observed, but there is no statistical support for it.

5.2 Announcement effects and independent variables

The results of the univariate analysis for the target, acquirer and complete deal for the whole sample and various sub-samples at the announcement period are shown in table 10 and table 11. Supporting statistical analysis data and calculation can be found in Appendix D. It provides average and median cumulative abnormal returns, as well as the parametrical and non-parametrical statistical tests for each sub-sample. It can be observed in Table 10 that the complete deal is positive with a return of 3.91% ( median 0.51%), both are significant on the 5% level, which proves hypothesis 1b and is in line with results of Goergen & Renneboog (2004) and Martynova & Renneboog (2011a) .However, this sample extends their research, because it is not limited to the European market, but focuses on all OECD countries. The result leads to the same conclusion as the previous mentioned research, that deals are rather motivated by synergy than by agency costs or hubris. Moving to table 11, panel A demonstrates that the returns of the targets are significantly higher than the returns of the Acquirer. That confirms hypothesis 1a.

TABLE 10AVERAGE AND MEDIAN CAR-1;+1 OF TARGET, ACQUIRER AND DEAL. Panel Sample of the independent variables Deals Average CAR Standar. T-test Percent positive CAR Median Rank test A Acquirer 332 2.63% 3.05* 52.1% 0.12% 1.86*** Target 332 6.67% test 6.23* 61.1% 0.88% 6.12* All Deals 332 2.81% 3.61* 57.2% 0.51% 2.93*

(39)

Panel B provides information for hypothesis 2d which confirms that common law targets have a significant higher abnormal return than civil law targets. This finding is in line with M&A literature and the results of Martynova & Renneboog (2011a); Goergen & Renneboog (2004), who focused on legal differences in Europe . Furthermore, this is in line with the argument of La Porta et al (2002) who state that targets located in common law countries provide better transparency and less agency costs and therefore receive a higher premium. Panel C compares common law acquirer with civil law acquirer. As, expected the thesis finds evidence that common law acquirer have significant higher returns around the announcement period than civil law acquirers. That is in line with Dombret et al (2008) who find similar results in Europe and Moeller & Schlingemann (2005) who find the same result for the US. It also shows support for the argument of Kuipers et al., (2002) that in domestic and cross-border acquisitions, variations in institutional environment such as legislations and regulations are a cause of difference in acquirer shareholder wealth gains.

In Panel D it is shown that common law deals capture significant higher returns than civil law deals, which confirms the hypothesis 2f, and represents the logical consequence of the two previous findings. Naturally, when common law targets and common law acquirers both receive higher returns, the positive wealth effect for the complete deal should be also higher for common law deals than for civil law deals. The findings are in line with Kuipers et al. (2009) and Martynova & Renneboog (2011a). The main arguments are higher transparency, lower agency and transaction costs.

Panel E compares German civil law deals with French Civil law deals. As expected, German civil law deals achieve a higher value creation for shareholders; however the mean difference test is not significant. Therefore the hypothesis is rejected.

(40)

voluntarily the better corporate government system from the target, which will have a positive synergy effect.

TABLE 11MEAN DIFFERENCES OF VARIOUS SUB-SAMPLES DURING ANNOUNCEMENT PERIOD

Panel Sub-Sample Deals Degrees of

Freedom Difference in average CAR -1;+1 Standardized T-test A Target Acquirer 332 634 4.04% 2.95* 332

B Common law target

Civil law target

159

215 6.52% 2.99*

173

C Common law acquirer

Civil law acquirer

159

177 5.67% 3.19*

173

D Common law deals

Civil law deals

159 184 5.06% 3.41* 154 E Domestic German Law

Domestic French Law

70 62 0.18% 0.12 29 F Cross-border Acquirer low liability Cross-border Acquirer high liability 49 52 1.92% 1.66*** 31 G Cross-border target high liability Cross-border target low liability 31 77 4.56% 1.28 49 H Domestic Acquirer high liability

Domestic Acquirer low liability

107

123 6.06% 2.39*

145

I

Domestic Target high liability

Domestic Target low liability

107

138 8.01% 2.84*

145

J Domestic High liability

Domestic Low Liability

107

130 4.64% 2.21**

145

(41)

A further possible explanation is the conclusion of Dahlquist et al. (2003) who theorized if the target company is situated in a strong legal environment, shareholders of acquiring companies might benefit from lower agency and transaction costs, as well as from more transparency in the target company, which allows a more precise valuation of the target and that reduces uncertainty, which leads to a more positive market reaction in case of a merger and acquisition. In panel G hypothesis 3b was tested. The coefficient point in the expected direction, but there is no significant statistical support found. Therefore, the hypothesis is rejected.

Hypothesis 3c is confirmed in panel H. If a domestic deal happens in a country with high shareholder protection, the acquirer return on the announcement day is higher than if the domestic deal happens in a country with low level of shareholder protection. As mentioned before, transparency is higher in countries with higher shareholder protection. This transparency will allow the acquirer to identify possible synergies more accurately, which will decrease uncertainty about revaluations in the integration process. Consequently, higher certainty and lower risk affects the market valuation of the M&A deal positively. According, to La Porta et. al. (2006) and Djankov et. al. (2008) it is expected that ownership concentration in a country with lower shareholder protection is higher. As (Faccio & Stolin, 2006) and Martynova & Renneboog (2008) state, the market may react negatively on the announcements of M&As by low shareholder protection country firms with concentrated ownership, because minority shareholders are afraid of potential expropriation in the deal. Bris & Cabolis (2008) extend that ownership argument and conclude that higher ownership concentration in the target company will give them higher bargaining power, and therefore demand a higher merger and acquisition premium.

(42)

shareholder protection countries. A lower level of agency and transaction costs is present in high level shareholder countries, which benefits both targets and acquirers (Dahlquist et al., 2003)

To summarize the findings of the unvariate analysis, firstly, targets benefit significantly higher in merger and acquisition than acquirers. However, also acquirer’s return is positive, and consequently, the deals done in this sample are on average value enhancing deals. Panel B till Panel J demonstrates clearly, that high legislative environment and high shareholder protection as well as an increase in shareholder protection (panel F) benefits acquires and complete deal.

5.3 Multivariate cross-sectional analysis

(43)

with hypothesis 3d. The control variable cross-border deal is also significant. It has a negative effect on the target returns, which means that domestic targets have higher returns than cross-border targets. This thesis expected the opposite. It also contradicts to the findings of Goergen & Renneboog (2004), who found a positive relationship. However, Schoenberg (1999) theorized that cultural difference between acquiring and target country can lead to difficulties in the integration process. Therefore, the author expects M&A’s gains to decrease. Conn et al. (2005) and Moeller & Schlingemann (2005) found confirming empirical evidence for that hypothesis. In these two empirical studies, the domestic acquirer and domestic deal had higher return than the cross-border acquirer and deals. Therefore, a possible explanation could be that shareholders of acquiring firms are more cautious when they value cross-border targets. They might pay lower premiums due to higher uncertainty about possible increased follow-up costs in the integration process.

(44)

TABLE 12RESULTS OF THE OLS-REGRESSION DURING THE ANNOUNCEMENT PERIOD

Independent and

control Variable Dependent Variable at CAR -1; +1 Target Acquirer Deal Common

law deal Civil law Deal Constant (t-stat) 0.05 (2.88*) 0.09 (2.82*) 0.07 (2.55**) 0.15 (2.59**) 0.01 (1.24) Target anti-self (t-stat) 0.08 (2.67*) 0.04 (2.25**) 0.04 (2.66*) 0.05 (1.74***) 0.07 (2.02**) Target Concentrated ownership (t-stat) -0.07 (-2.86*)

Acquirer high versus low protection (t-stat) 0.03 (3.61*) Cross-border (t-stat) -0.02 (-1.87***)

Target Market to book (t-stat) -0.05 (-1.93***) Acquirer Market to book (t-stat) -0.05 (-2.53**) -0.06 (-3.65*) Acquirer size (t-stat) -0.006 (-2.55**) -0.003 (-2.08**) -0.01 (-2.43**) Target leverage (t-stat) 0.009 (3.28*) 0.01 (4.30*) 0.01 (2.16**) Acquirer leverage (t-stat) -0.003 (-2.67*) -0.002 (-2,67*) -0,004 (-6.39*) R-squared**** 0.08 0.12 0.17 0.11 0.26 F-statistic***** 7.26* 7.88* 11.1* 5.15* 10.8* Durbin-Wastson****** 1.88 2.06 2.02 1.87 1.78

Note: * 1% significance level; ** 5% significance level; *** 10% significance level. ****stands for the proportion of variability in a data set that is accounted for by the statistical model; ***** test the overall significance of the regression model; ***** if the Durbin-Watson statistic is substantially less than 2, there is evidence of positive serial correlation

(45)

sample it has a negative effect. A possible explanation might be, as Masulis et al. (2007) show that high leverage will lead to lower future free cash flows and that the market expects that the acquirer will run in the long-run into financial distress.

The last variable which has a significant impact on acquirer and deal return at the announcement period is target leverage. Also in this case, the results contradict the expectations. As Martynova & Renneboog (2011a) state, bidders are likely to pay higher premiums for a target with lower leverage and higher cash flows, because targets will have less obligations towards creditors and more liquid assets which the new acquirer takes control over. Nonetheless, an explanation given by Masulis et al. (2007) for this finding could be that targets with high leverage are well monitored and controlled by creditors. That reduces the risk of agency problems, such as manager hubris. Furthermore, the target company might have good relation with creditors and banks, and gets good credit conditions, which might have a value for an acquirer. An argument derived from agency theory comes from Pinkowitz et. al. (2006) who argue liquid assets are worth less to minority shareholders, because it is easier for majority shareholder to use these liquid assets for private benefits. Therefore, high leverage in target could increase value.

The next column presents the variables influencing the return of common law deals at the announcement period. Higher target shareholder protection has a significant positive effect on common law deal return at the 10% significance level.