M&A announcement effects to bidding company shareholders: the case of European Union during 2004-2011

(1)

M&A announcement effects to bidding company shareholders:

the case of European Union during 2004-2011

Master thesis

Karolis Andriuskevicius s2045478 29th February, 2012

Supervisor: Dr. ing. N. Brunia Co-assessor: Dr. W. Westerman

University of Groningen Faculty of Economics and Business

(2)

2 Abstract

The thesis is an addition to the M&A literature within the framework of European Union. The effects of international expansion, type of an acquisition, cultural and institutional distance together with traditional control variables are investigated in order to explain short term wealth effects for acquiring company shareholders by performing a standard event study. Statistical support is provided to the statement that capital market approaches acquisition announcements as value increasing activity. The obtained empirical results support the idea that bidding company shareholders are better off when: 1) domestic rather than cross-border acquisition is announced; 2) firms involved dispose similar cultures and institutional backgrounds; 3) bidding company is located in the country with a common law. Acquiring company shareholder wealth gains are found to be positively related with bidder size and relative size of the target to the bidder, whereas large and related transactions are found to result into wealth losses.

(3)

3 Acknowledgments

I am grateful to my supervisor, Dr. ing. N. Brunia for his inestimable help and constant support. His professionalism in the field helped me to gain valuable knowledge during the research. Furthermore, his personal qualities contributed to the effective and pleasant communication.

(4)

4

1. Introduction

Internal growth and acquisitions are the two main modes of corporate expansion. Corporations compare the features of internal growth (e.g. long time for the accrual of returns, high needs for investment and business development, incremental nature of the internal development process) with the characteristics of external growth (lack of potential acquisition candidates, premium needed to pay for the target shareholders, transactions costs) when evaluating organizational expansion and growth strategies (Singh and Montgomery, 1987). When information about acquisitions is disclosed, capital market participants adjust their expectation of future earnings which is reflected in the stock price of the firms. Considering the priority of shareholder wealth maximization in MNEs and assuming the prevalence of efficient capital markets, M&A should occur when management has confidence in creating market value higher than the investor could obtain himself by diversifying his own portfolio (Salter and Weinhold, 1978). Successful acquisition has different meaning for target and bidding company. Target company benefits when receiving acquisition premium, while shareholders of the bidding company benefit from synergies greater than the premium paid to the target. Empirical research (Capron and Pistre, 2002; Campa and Hernando, 2004; von Eije and Wiegerinck, 2010) documents that the shareholders of the target companies receive large premiums, while bidders either gain a small statistically insignificant amount or lose a small significant amount from the announcement of the bid. Summarizing, the numbers of event studies have demonstrated that takeovers seem to create shareholder value with most of the gains accruing to the shareholders of the target company (Andrade et al. 2001). Therefore, even though traditional literature extensively defines motivational factors and strategies behind the acquisitions, their credibility is questioned by the capital markets.

(6)

6 number of CBAs, international industry characterization, and friendly negotiations between the acquirer and target.

Confronting theories and disputes exist regarding motives, strategies, expected- and factual M&A performance. Majority of the existing empirical studies find significantly positive returns to the shareholders of target firms. However, effects captured by acquiring company shareholders are less conclusive. On the one hand, some researches (Andrade et al, 2001; Jarell and Poulsen, 1989; Walker, 2000; Campa and Hernando, 2004; Martynova and Renneboog, 2006) identify negative or non-existent wealth changes to bidding company shareholders. On the other hand, Faccio et al. (2006), Eije and Wiegerinck (2010), Chari et al. (2010) and other academics provide support for the concept that M&A are bidders’ shareholder value increasing activity.

Considering the prevalence of M&A and the opposing results of the existing findings the main question of the thesis is stated as follows: Do capital markets assess merger and acquisitions as value creating or destroying activities?

Lall (2002) recognizes that even though M&A are most active within the framework of industrial world, they are also increasing in importance in the developing world especially in the form of acquisitions. However, most of the existing literature reflects the findings of the studies performed in the developed world. The analysis of M&A activities between European companies disposes unique contexts as both domestic and cross-border deals are prevalent, as companies with different origins cooperate, cultural-, economic- and socio-economic differences often exist between the parties involved in the M&A. Little is known about European takeover market as most of the researches done focus on the US (Martynova and Renneboog, 2011, Corhay and Rad, 2000, Lowinski et al. 2004). Considering the lack of empirical studies with European focus, the thesis aims to investigate whether cross-border acquisitions create value for the firms undertaking such transaction within Europe. Furthermore, the increasing trend towards acquisitions from developed economies to developing ones creates a need to address to what extent the acquisitions in this direction differ from cross border acquisitions between companies from developed economies.

(7)

7 as well as other areas of corporate finance are discussed as having impact on short term shareholder wealth effects.

Empirical research has been is based on a sample of acquisitions by companies with the origin in one of the EU27 countries (Belgium, Greece, Luxembourg, Denmark, Spain, Netherlands, Germany, France, Portugal, Ireland, Italy, United Kingdom, Austria, Finland, Sweden, Poland, Czech Republic, Cyprus, Latvia, Lithuania, Slovenia, Estonia, Slovakia, Hungary, Malta, Bulgaria and Romania). The research of this sample allows to investigate the gains and losses on bidding firm shareholders as influenced by the consequences of deregulation of national markets towards a single European market (Lisbon strategy) and a reduced number of trade barriers.

The main question of the thesis is examined throughout the whole paper by focusing on the following sub-questions:

- Does the recent trend of European companies expansion to New Member States result in a positive shareholder wealth effect?

- Which variables significantly explain the shareholder wealth effects? How can the variation in the distribution of any value created be explained?

- How do economic theories explain acquisition returns as well as the influence of company- and takeover-specific attributes?

The thesis contributes to the literature in the following ways. Firstly, the thesis aims to contribute to the existing international acquisition literature as the findings regarding the acquisition performance in the New Member States remain an open question (Bhagat et al. 2011). Specifically, it is intended to investigate to what extent acquisitions in New Member States by companies from EU 15 countries show similar returns as investigated in empirical studies with different domain of the target. Secondly, considering recent shift from planned to market economy, political and legal changes in the region, the thesis investigates whether the proxies chosen in the cross-sectional analysis can explain shareholder wealth effects. Thirdly, considering the lack of research on acquisitions of private targets as noted by Capron and Shen (2007), the thesis acknowledges the effect of public listing.

(8)

8 to their clients about how specific characteristics of the bidding-, target-company or takeover in broad sense are reflected in the market reactions. It is aimed to explain, why specific shareholder wealth changes are experienced.

(9)

9

2. Literature review and hypotheses

This section starts with motivational forces of the M&A’s as they help to understand why the takeovers occur and how stock markets react to the M&A announcements. Afterwards there follows the review of the previous empirical studies that analyzed acquisitions announcement effects of the bidding company shareholders. The section continues with developing hypotheses and focusing on bidding company shareholder wealth changes in European Union. Finally, control variables and the conceptual model of the research are presented.

2.1 Strategies and motives behind the acquisitions

Gammelgaard (2004) performs an empirical study of the emerging motives of M&A’s. The author states that rationale behind M&A’s is unique for each deal and its motives have changed over time. Elaborating on his work, the common motives of M&A’s could be categorized into market-, cost-, diversification- and financial synergy exploitation as well as synergy- and competence exploration (see Table 1).

Table 1 Motivation of M&A’s

Key motive Purpose Subordinated motives

Synergy exploration Restructuring of productive and

administrative resources

Pooling of complementary resources

Financial synergy exploitation

Reduction of financial cost Shareholder value, debt/equity ratio, tax, capital cost

Diversification exploitation

Minimization of fluctuation in revenues

Risk minimizing, alternative use of resources

Market exploitation Extension of sales opportunities

Growth, access to new markets or products, market product.

Cost exploitation Reduction of production and administration costs

Economies of scale and scope, access to low-cost resources

Competence exploration

Gain access to competences

Access to non-tradable, unique and inimitable resources deriving sustained competitive positions in the market

(10)

10 entity shares them and ability of price discrimination as companies are capable to discriminate between markets and set different prices. Andrade et al. (2001) and Walker (2000) relates synergy and financial synergy exploitation with the attempts to increase market power through the formation of monopolies and oligopolies. Financial synergy motive assumes benefits of tax credit utilization and capital cost reduction through the reduction of systematic risk (diversifying deal), access to cheaper capital (increased company’s size) or establishment of an internal capital market (Trautwein, 1990). The motive of the smoothing results of the company which in the long term gives more confidence in investing in the company lies behind the diversification goal. Andrade et al. (2001) add that new opportunities of diversification potentially arise from the exploitation of internal capital markets and management risk of undiversified managers. Similarly to synergy and increased market power motive, market exploitation motive is closely associated with changes in the demand-side. Cost exploitation is discussed by Walker (2000) as an aspect benefiting the company through the creation of economies of scale and disciplining inefficient managers. Gammelgaard (2004) holds that exploring competences through by the target possessed unique knowledge, employee skills and organizational motives are the key sources of synergy exploration and competence development in the 21st century.

Mukherjee et al. (2004) focused on managerial perspective of M&A’s and investigated motives of the M&A’s from the practical point of view. Specifically, 1200 acquisitions completed during 1990-2001 were analyzed through the surveys of the chief financial officers of the US companies. The results of their study show that following M&A’s motives are considered as most important: synergy (37.3% of responses), diversification (29.3%), achieving a specific organizational form as part of an ongoing restructuring program (10.7%), acquiring a firm with below its replacement cost (8.0%), using excess free cash (5.3%), tax reduction experienced due to tax losses of the acquired company (2.7%). Comparing the results with the above discussed categorization imply that motives behind the takeovers vary from a deal to deal and are complex and in most of the cases composite.

(11)

11 Efficiency, market power and corporate control theories are approached by shareholders when considering sources of economic value underlying M&A’s. According to the definition used by Trautwein (1990), theory of efficiency treats takeover activities as a rational choice when they are planned and executed to capture financial (e.g. reduced costs of capital), operational (e.g. reduced business costs or unique products/services) and managerial (e.g. superior planning and monitoring abilities) synergies. However, assumptions underlying this theory are criticized by academics. On the one hand, the existence of efficient capital markets disables exploitation of financial synergies. On the other hand, operational and managerial synergies are criticized as being often stated but rarely realized. Market power theory suggests increased monopolistic abilities emerging due to the M&A. For instance, firms can use profits in one market to sustain a fight for market share in another market and imply the so called cross-subsidization strategy. Deliberate reduction of supply and deterring potential market entrants are the other sources of market power theory applications. Manne (1965) developed the theory of corporate control with the key assumption of the potential shareholder value increase that emerges through the replacement of inefficient management. Management of the target company is assumed to lack knowledge or qualification and therefore is not able to fully capture company’s potential.

Table 2 Shareholder value increasing and decreasing theories underlying M&A’s

Shareholder value increasing theory Shareholder value decreasing theory Theory of efficiency Theory of managerial hubris

Market power theory Theory of managerial discretion Theory of corporate control Managerial entrenchment theory

Theory of empire building

(12)

12 and subsequently poor performance of the combined companies. Li and Tang (2010) suggest that existing empirical research has provided support for hubris affected decision makers as paying higher premiums, relying on internal rather than external financing, missing their own forecast of earning as well as undertaking more value destroying takeovers. The authors add to the research by empirically proving the positive relationship between CEO hubris and firm’s risk tolerance. The concept of theory of managerial discretion as introduced by Jensen and Meckling (1976) assumes differing objectives between shareholders and managers where managers dispose the latitude of decision-making. Jensen (1986) asserts that the interests mismatch is compounded when excess liquidity and/or free cash flow stand for the disposal of managers. These conditions strengthen the negative impact to M&A performance. Managerial entrenchment theory is based on an agency problem and is closely related to the previous theories. Shleifer and Vishny (1989) elaborate on this theory as managers making specific investments that reduce their probability of being harmed. For instance, the completed M&A’s increase fungibility of managers, reallocate wages and perquisites to the favour of managers. Jensen (1986) advocates that empire building theory is concisely related to the conflict of interest between shareholders and managers. “Managers have incentives to cause their firms to grow beyond the optimal size. Growth increases managers’ power by increasing the resources under their control which is associated with increased compensation as it is directly linked to the sales growth”.

2.2 Acquisition announcement returns

(13)

13 Table 3 Panel A: Empirical research that found either non-existing or negative shareholder wealth effects that occurred due to the acquisition announcement

Study Period Event

window Sample size Bidders origin Targets origin Bidder returns Bidder returns positively related

Bidder returns negatively related Corhay and Rad (2000) 1990-1996 (-5;+5) 111 Netherlands Western Europe

+0.14% Corporate control Foreign exposure, relatedness Walker (2000)

1980-1986

(-2;+2) 278 -0.8% Relatedness, tender offer,

cash payment

Diversification through operational overlap DeLong (2001)

1988-1995

(-1;+1) 280 US -1.7% Relative target to bidder

size Pre-merger performance of the target Campa and Hernando (2005) 1998-2000 (-1;+1) 262 European Union European Union

-1.48% Relative size of the acquisition

Acquisitions in government regulated industries,

financial industries, cross-border acquisition Moeller and Schlingeman (2005) 1985-1995 (-1;+1) 4430 US US, UK, Canada, France, Germany

-0.87 % Takeover activity in the target country, legal system offering better shareholder rights

Increase in global and industrial diversification, bidders’ country’s degree of economic restrictiveness Martynova and Renneboog (2008) 1993-2001 (-1;+1) 2419 Continental Europe and UK Continental Europe and UK

Private status of the target

Deal hostility, equity payment

Masulis et al. (2007)

1990-2003

(-2;+2) 3333 US +0.22% Financing with cash,

target being private

Financing with stock, target being public, large number of antitakeover provisions Karels et al. (2011) 1995-2007 (-1;+1) 678 US India -0.41% Weitzel and McCarthy (2009) 1996-2007 17137 US, Western Europe US, Western Europe

(14)

14 Table 3 (continued) Panel B: Empirical research that revealed positive shareholder wealth effects that occurred due to the acquisition announcement

Study Period Event

window Sample size Bidders origin Targets origin Bidder returns Bidder returns positively related

Bidder returns negatively related Seth et al. (2002) 1981-1990 (-10;+10) 100 US Canada, France, UK, Japan, Germany +7.57% Reverse internalization, asset transfer, financial diversification, bank oriented system

Market seeing motive ( GDP growth of the target country), relative deal size Goergen and Renneboog (2004) 1993-2000 (-2;+2) 228 Continental Europe and UK Continental Europe and UK +1.2% Friendly acquisitions, payment with cash, development of corporate governance Facio et al. (2006) 1996-2001

(-5;+5) 4429 Europe +1.53% Target being unlisted Eije and

Wiegerinck (2010)

1997-2008

(-2;+2) 821 EU-15 US +2.1% Relative target to bidder size, target being in emerging market Chari et al. (2010) 1986-2006 (-1;+1) 678 Developed countries: Canada, France, Germany, Italy, Japan, Netherland, Spain, UK, US. Emerging countries: Argentina, Brazil, Chile, China, , Costa Rica, Indonesia, Malaysia, Mexico, Taiwan, Thailand. +1.16% Target in emerging market, weakness of the contracting environment

(15)

15 The authors explain the predicted wealth gains through future synergies and wealth redistribution among stakeholders even though the calculations have proven that the acquirer’s cumulative abnormal returns are null on average. Moschieri and Campa (2009) analyzed the business climate change process undergone by European firms together with the takeover characteristics within EU. The authors highlighted the fact that the volume of M&A’s towards European firms surpassed that towards US companies which are traditionally assumed a dominant market for corporate takeovers. The authors acknowledge the positive influence of European Commission as this institution fosters standardization and increases transparency in the development of a single European market for M&A. Specifically, the authors state that the following implemented changes leads to shareholders wealth gains: increased use of cash as a payment method, increase in cross-border deals, industry consolidation, decrease in the time needed to execute the M&A. Tuch and Sullivan (2007) remark in their literature review that recent M&A deals tend to result in bidders shareholder wealth gains especially in the countries other than UK and US. Based on the presented discussion above together with the implicit assumption of validity of the efficiency-, market power- and corporate control theoriesefficiency-, the following hypothesis has been developed.

H1: Bidding company shareholders experience short-term wealth gains around the announcement day.

(16)

16 assets and operations of merging firms. However, no support for the “organizational fit” theory which is defined as benefiting merging companies due to lower integration costs when involved firms dispose similar cultures, structures, procedures and systems was found. Uhlenbruck and Castro (2000) determine that existing organizational fit differences between Western and CEE countries are too large to integrate the firms efficiently. Therefore the authors interpret “organizational fit” theory as not applicable to the CEE region. Chari et al. (2010) identified significantly positive abnormal return of 1.16% in their study about the returns to shareholders of developed country firms that acquired company from emerging during 1986-2006. The authors identify that bidding company shareholders gain through the acquisition of majority control of a firm in emerging market as contracting environment in a target country of origin is weaker. Bidder from developed country shares better institutional and corporate governance practices (e.g. legal and accounting standards) with the target company which in return benefits the acquiring company. It is remarkable that bidders have more incentive of practice sharing when disposing intangible assets production capabilities. Taking argumentation above into account, the following hypothesis is developed:

H2: Shareholders of a bidding company capture larger wealth gains when announced M&A’s involve a company with an origin in New rather than Old Member State of the EU.

2.3 M&A’s in the context of EU

(17)

17 Existing researches that investigated differences between shareholder wealth gains in domestic and cross-border acquisitions provide opposing results. Empirical results by Goergen and Renneboog (2004) contrast theories that postulate benefits of cross-border deals. Specifically, the authors measure statistically significant wealth gain changes of 2.7% between domestic and cross-border deals. Empirical studies by Campa and Hernando (2004) and Lowinski et al. (2004) do not find any significant differences in acquirers stock performance between domestic and cross-border deals. Lowinski et al. (2004) relate the non-existing differences between domestic and cross-border acquisitions to highly integrated capital markets. Hughes et al. (1975) explain that there are no diversification benefits inherent to international firms that could not be obtained by individual investors making direct investments in the countries in which the international companies would operate otherwise. Corhay and Rad (2000) measure that cross-border acquisitions are beneficial for acquiring company shareholders as statistically significant average cumulative abnormal return for cross-border acquisitions equals 1.44% during 5 days event window. The authors argue that bidders potentially gain from imperfections in factor, product and capital markets as mispricing and trade barriers are still common in Europe especially in the former Soviet Union countries. Their argumentation shows the validity of diversification motive as firms seek to reduce earnings volatility. When analyzing over 800 cross-border M&A that occurred during 1991-2004, Chakrabarti et al. (2009) surveyed country differences with the aim to explains short-term M&A performance. Their empirical results evidence that existing economic differences between bidding and target company country of origin serve as an alternative to distinguish between cross-border and domestic acquisition.. Chakrabarti et al. (2009) have shown that larger country economic differences lead to higher short-term M&A performance which is explained by larger potential of synergy gains. Taking relevance of the comparison between domestic and cross-border acquisitions as well insights of the discussion above into account, the following hypothesis is formulated:

H3: Shareholders of an acquiring company that participates in cross-border M&A’s are better off than the ones that participated in domestic one.

(18)

18 Announcement returns associated with M&A’s in the countries with low cultural differences are measured to amount to statistically insignificant 0.19%, whereas announcement of M&As in culturally distant companies result in significantly positive returns of 0.76%. Further analysis has shown that the initial findings are not well founded as reversed results are being proven over a longer period. Morosini et al. (1998) provide further empirical support for the benefits of cultural distances and questions the reliability of the theory which relates cross-border acquisitions with increasing costs of integration. Based on the sample of 52 cross-border acquisitions that occurred in 1987-1992 the empirical conclusions are drawn that cross-border M&As perform better when routines and repertoires of the target country of origin are more distant than those of the bidding company. Following argumentation by Morosini et al. (1998), inventiveness, innovation, entrepreneurship and decision making practices are closely related to culture and therefore are complicated to develop and adjust across different national cultures. Thus, a differing set of practices and routines have the potential to increase bidding, target and combined company performance:

H4: The greater the national cultural distance between the acquirer’s and target’s country of origin, the greater would be the acquisition announcement effects to bidding company shareholders.

(19)

19 empirically proven that the conjunction of institutional distance, market structure and power dependencies explains integration effects. On the one hand, large national institutional distances accompanied by highly localized industry market structure force to adopt locally driven integration strategy. On the other hand, globalized industry market and mirror institutional differences favour performance of globally driven integration strategy. La Porta et al. (2000) suggest that corporate governance improvements may result in acquisitions being more efficient if the acquirer is located in a common rather than in the civil law system as common law protects investors better. Common law is usually created, implemented and promoted by judges who are given legal rights to adjust existing laws and norms to new situations, whereas civil law is made by legislatures and judges are not supposed to go beyond the statutes. La Porta et al. (2000) find empirical support for positive relationship between strong investor protection and effective corporate governance which is characterized by broad financial markets, dispersed ownership shares, efficient allocation of capital across firms. The discussion above motivates the following hypotheses:

H5: Acquisitions between institutionally distant firms lead to short-term shareholder wealth gains.

H6: Shareholders of a bidding company with the origin in common law country capture larger gains than shareholders of the firm originated in civil law country.

2.4 Determinants of cross-sectional analysis (control variables)

Based on the previously provided information and reviewed studies, a set of control variables is found to be common for most of the studies. Appendix A elaborates on definition, measurement and source of them, whereas Table 4 summarizes the set of into the thesis incorporated variables which aim to provide additional insights into the factors behind the experienced stock price changes.

(20)

20 small firms make deals with small dollar gains (9 billion $ during 1980-2001), whereas deals by large MNCs lead to large absolute dollar losses (312 billion $ during 1980-2001). The authors identify that small acquirers gain roughly two percentage points higher returns than large companies.

Table 4 Control variables and their relevance

Control variable Expected effect

Relative theory Literature

support Bidder size

(Size)

-

Hubris (Roll, 1986), empire building

Faccio and Masulis (2005), Vishny and Shleifer (1989), Moeller et al. (2004) Bidder leverage (Lev) +

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

Lang et al.(1991), Masulis et al. (2007), Harrison and Oler (2008), Faccio and Masulis (2005)

Transaction value (Deal_size)

-

Agency problem, misevaluation Eun et al. (1996), Petmezas (2009), Alexandridis et al. (2010) M&A’s of private vs. public acquisitions (List) + Bhagat et al. (2011) Capron and Shen (2007) Relative size of the

target to the bidder

(RMV) +

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

Inflated integration costs

Eun et al. (1996) Asquith et al. (1983) Related vs. diversifying acquisition (Rel) + Focus vs. diversification strategy of the bidder

Bhagat et al. (2011)

Eun et al. (1996) Corhay and Rad (2000) DeLong (2001) Method of payment (Paym) + Pecking-order theory Signaling theory Asymmetric information hypothesis DeLong (2001), Petmezas (2009)

Bidders book to market ratio

(BTM)

+

Hubris (Roll, 1986) Faccio and Masulis (2005), Petmezas (2009)

Announcement day equals rumor date (ANN)

(21)

21 This phenomenon is explained as small firms are found to acquire more than half of the private companies whose acquisitions are empirically proven to be more profitable, whereas large MNCs tend to participate in the deals with public targets. Furthermore, small companies are found to pay for the deal more likely in cash than equity. Travlos (1987) gives empirical support for the statement that M&A’s by public companies paid with equity result to lower announcement returns than deals financed with cash. Theories related to agency provide additional expectation for the predicted negative relationship between bidder size and M&A performance. Moeller et al. (2004) empirically support that managerial hubris, equity signalling hypothesis (e.g. issue of equity by a company signals that the market overvalues its assets (Meyers and Majluf, 1984), empire-building theory are more common between large companies: 1) managers of large firms are found to dispose lower ownership parts than management of small companies and therefore their goals are less aligned with shareholders; 2) managers of large companies have usually succeeded in growing the firm and therefore are more prone to hubris; 3) large firms are more likely to dispose exhausted internal growth opportunities and therefore seek M&A.

(22)

22 motivation above, it is predicted that higher pre-merger bidding company leverage relates to higher abnormal returns during M&A announcement period.

Alexandridis et al. (2010) document a robust negative relationship between deal value and the premium paid in M&A. Despite the lower M&A premium, large deals result in shareholder value destruction and increased return uncertainty suggesting that capital market values large deals as uncertain projects. Alexandridis et al. (2010) explain that the negative relationship between stock price performance and premium paid occurs because large deals struggles to realize potential synergies due to problems and costs associated with the post-deal integration process. Similarly, Moeller et al. (2005) analyze the takeover period of 1998-2001 and assert that bidding company shareholders loose on average 12 cents per dollar spent on M&A around the deal announcement day. Their empirical research has shown that large transaction values are related with large losses. Martynova and Renneboog (2011) find negative relationship between transaction value and acquisition performance. The authors reason that this direction relationship occurs because acquiring large companies requires complex management capabilities, additional integration costs. Furthermore, Martynova and Renneboog (2011) ascertain that revaluation of the acquired assets may lead to substantial losses during the post-announcement period.

Capron and Shen (2007) examined the relationship between acquiring public and private firms. The authors empirically support that acquiring private targets leads to higher shareholder wealth gains than participating in the deal with public target. Acquirers are found to have more value-creating opportunities from exploiting private information, whereas fewer opportunities exist when acquiring public targets as information is processed and assets are evaluated by an active market for corporate control. Furthermore, private firms are identified to be traded on “discount” which means that bidder gets larger split of the value amiong merging firms. Fuller et al. (2002) used a sample of 3135 deals that occurred during 1990-2000 and further supported shareholder gains in the case of private target acquisition and losses in the case of public target acquisition. The researchers indicate lower price of private firms due to lower liquidity as the main reason for the observed phenomena. Faccio et al. (2006) measured that shareholders of the acquiring company face an insignificant loss of 0.38 % in the case of listed target acquisition and significant gain of 1.48% in the case of private company takeover. However, the researchers notice that the reasons behind the listing effect remain elusive.

(23)

23 Swary (1990) treat the relative size of the acquisition as decreasing acquirer returns due to the addtitional complexity and larger integration costs. Kooli et al. (2003) justify greater relative size as providing greater bargaining power to the target which subsequently enables targets to negotiate more favorable deal conditions. On the other hand, the authors approve that higher relative size facilitates greater potential for recombination and scale economies. Insights by Moeller et al. (2004) suggest that smaller acquirers are more careful when announcing acquisitions as it has relatively high influence on performance in comparison with a large acquirer.

(24)

24 Furthermore, Walker assumes corporate strategic objectives being reflected in related or unrelated takeovers.

Book-to-market ratio is reflected in the hubris hypothesis. Hubris is more common for "glamour" than for “value” firms which mean that markets attach less value. Rau and Vermaelen (1998) indicate in their long-horizon event study that bidders underperform the market. The authors link this loss to the poor post-acquisition performance of low book-to-market ("glamour") firms. Rau and Vermaelen elaborate on this finding as the evidence that market and management overextrapolate the bidder's past performance (as reflected in the bidder's book-to-market ratio) when they assess the potential of an acquisition. Similarly, Goergen and Renneboog (2004) show that even though, the market-to-book ratio of the target matters in terms of the bid premium, a high market-to-book ratio for the target leads to a higher bid premium combined with a negative abnormal return for the bidder.

The last control variable used in this thesis considers the possibility of leaking information. Occurrence of rumor date before the announcement day as determined by illegal or insider trading potentially misrepresents the results and their explanatory power. Announcement day should not have any significant stock price changes in the presence of information leakage as efficient market and active market participants have already revised their expectations regarding future profitability of the company and incorporated the acquisition intentions into their calculations. The use of Zephyr database is helpful in this case as it provides both the rumor and announcement days of the acquisition.

2.5 Conceptual model

(25)

25

Independent variables

Company’s with the origin in New Member State of the EU involvement Geographical expansion (cross border vs. domestic M&A) Cultural distance Institutional distance

Company’s with the origin in common law involvement ( vs. civil law)

Dependent variable

Acquisition performance: changes of bidding company’s share price around the announcement day

Control variables Bidders’ size Bidder’s pre-announce ment leverage Transactio n value Acquiring private company (vs. public) Relative deal size Related deal (vs. diversifyin g) Bidder pre book to market ratio Announce ment day matches rumor date ( or mismatch)

* Note: Specifies hypothesis’ direction + Hypothesized positive effect - Hypothesized negative effect

+ (-) Hypothesized positive (negative effect) Figure 1. Theoretical model of the research

+ _{+ (-)} ₊ ₊ _{+ (-)}

(26)

26

3. Data

This section of the paper presents the data used in the empirical analysis. Section 3.1 deals with the data source and selection criteria, whereas section 3.2 describes and presents the data that serves as the empirical framework for the research. Finally, section 3.3 provides the overview of the variables used in the research.

3.1 Data sources

Considering the motivation and main questions of the study, a combination of sources was used to obtain the data. To begin with, Zephyr a database of Bureau van Dijk electronic publishing was employed to identify the sample of publicly listed European companies that took over another company in the period 2004-2011. Furthermore, this database provides deal characteristics (e.g. announcement and rumor dates, transaction value, method of payment) as well as initial bidding and target company specific information (e.g. size, listing status of the target, geographical scope, industries companies are active in, etc.). The data used for the calculations of daily stock and market index returns as well as other missing variables used for the cross-sectional analysis were obtained through DataStream 5.1 database by Thomson Reuters. Finally, the websites of CIA World Fact Book, Eurostat, Heritage foundation and Geert Hofstede were employed to gather information regarding bidding and target companies country of origin specific characteristics that were further used in the empirical analysis.

(27)

27 minority stakes do not deliver significant acquirer returns. The other selection criterion considers industries of the target and bidding companies. Specifically, acquisitions related to banking, insurance and unknown industries are excluded as they have different motivational forces. Furthermore, these industries are known as dealing with special regulatory environment and accounting related issues which might skew the results and misrepresent the total sample. Finally, the last criterion selects M&A that were successfully completed. Overall, 746 events that meet the criteria raised are selected.

At the end, after cross-checking for the announcement date and eliminating cases when company or its country of origin specific information is missing on DataStream 5.1, CIA World Fact Book, Eurostat, Heritage foundation and Geert Hofstede databases, 572 events build the total sample used in the empirical analysis.

Table 5 Selection criteria

Criteria Number of events Difference

Time period: 01/01/2004-01/11/2011 465200

Origin of the target and bidder is EU-27 228264 236936 Public acquirer; private and public targets 103821 124443 Method of payment is known: cash, shares, mix of

cash and shares

63031 40790

Minimum of final stake in the acquired company: 50%

14639 48392

Industries of the acquiring and target company are known and excludes banking and insurance industries.

1296 13343

Current deal status: completed 746 550

Cross-checking information with DataStream 5.1 database

572 174

3.2 Sample characteristics

(28)

28 currently experienced are circumstanced by the threats to the global economy, worries over the health of government finances in Europe, declined stock market and lack of credit.

Table 6 Yearly breakdown of the M&A sample

Period No. of deals No. of domestic (cross-border) acquisiti ons No. of acquisi tions with a compa-ny that has origins in NMS of the EU No. of acquisiti ons with target being private ( public) Total deal value (billion EUR) Average deal value (million EUR) Median deal value (million EUR) Maxi-mum deal value (billion EUR) 2002* 2 0 (2) 2 2 (0) 0.01 6.40 6.40 0.01 2003* 5 2 (3) 1 5 (0) 0.25 49.22 18.32 0.13 2004 86 60 (26) 14 81 (5) 19.43 225.88 38.99 3.46 2005 118 74 (44) 8 112 (6) 24.98 211.67 15.61 3.90 2006 82 59 (23) 5 74 (8) 43.89 535.28 19.84 17.23 2007 115 74 (41) 3 112 (3) 28.43 247.24 57.00 11.80 2008 48 40 (8) 4 46 (2) 1.67 34.78 17.46 0.60 2009 37 26 (11) 9 35 (2) 19.34 522.74 7.39 9.63 2010 44 29 (15) 2 44 (0) 5.62 127.82 12.03 0.85 2011 35 24 (11) 1 30 (5) 2.54 72.44 31.69 0.37 2004-2011** 572 389 (183) 49 541 (31) 146.31 256.87 24.48 17.23 Note: * The years of 2002 and 2003 are included in the table as acquisitions were announced during these years, whereas they were completed in 2004-2011

** 2004 – 2011 is M&A completion period, whereas calculations also include 2002 and 2003 as these years mark acquisition announcement

(29)

29 up to 146.16 billion EUR. The average deal value presented in the seventh column of the table suggests that almost 257 million EUR were paid on average for the target company. However, this measure needs to be interpreted with caution as it is biased towards large deals. For instance, large deals (2006: EADS NV 2.8 billion EUR, IBERDROLA SA 17.3 billion EUR; 2009: ENEL SPA 9.63 billion EUR, RWE AG 7.3 billion EUR) caused the average acquisition size exceeding 520 million EUR in 2006 and 2009. Median values of the deals presented in the eight column aims to reduce the effect of large M&As. Interestingly it is inferred that the median value amounts to 24.48 million EUR and is more than 10 times lower than the average deal value.

(30)

30 Table 7 Distribution of the sample across the EU members

Company with the origin of the country is acquirer Company with the origin of the

(31)

31 Table 7 (continued) Distribution of the sample across the EU members

Company with the origin of the country is acquirer Company with the origin of the

country is target Country of origin of the company No. of deals Total deal value (bil. EUR) Averag e deal value (mil. EUR) Median deal value (mil. EUR) No. of domestic (cross-border) deals No. of deals with target being private (public) No. of deals with a company that has origins in NMS of the EU* No. of deals Total deal value (bil. EUR) Averag e deal value (mil. EUR) Median deal value ( mil. EUR) Portugal 4 0.96 238.95 56.20 3 (1) 4 (0) 0 4 0.94 312.80 426.00 Romania 2 0.01 7.34 7.34 2 (0) 0 (2) 2 4 0.90 224.77 30.57 Slovakia - - - - - - 1 1 0.01 12.70 12.70 Slovenia 2 0.11 55.42 55.42 1 (1) 2 (0) 2 1 0.01 13.85 13.85 Spain 36 46.49 384..38 194.35 23 (13) 34 (2) 2 35 32.50 928.66 132.40 Sweden 28 2.27 81.17 8.49 12 (16) 28 (0) 0 20 1.92 96.01 13.46 EU-27 572 146.31 255.78 150.34 389 (183) 541 (31) 49 572 146.31 255.78 150.34 EU-15** 533 144.18 347.46 227.20 358 (175) 510 (23) 11 523 140.94 271.05 25.87 EU-12*** 39 2.04 71.09 65.27 31 (8) 31 (8) 38 49 5.21 39.06 38.97

(32)

32 Besides that, it becomes evident that companies from Old Member States of the EU initiated M&As with an average (median) value 347.46 (227.2) million EUR whereas average (median) deal value paid by companies from New Member States is almost 5 times smaller and amounts to 71.09 (65.27) million EUR.

Furthermore, the gathered data suggests considering percentage terms countries from New Member States are tended more to acquire publicly listed firms than companies from Old Member States of the EU. Finally, it becomes evident that companies originated in EU-15 only announced and successfully completed only 11 takeovers of the companies situated in the New Member States of the EU. Specifically, 49 deals with the total value of 5.21 billion EUR were completed in New Member States.

On the other hand, following insights can be drawn up considering the deals when the company with the origin of the specific country is the target. Firstly, the total number and deal value of British companies being targets is the highest. Secondly, it becomes evident that companies from Old Member States were targets more often than they were acquirers. Furthermore, the average (39.09 million EUR) and median (38.97 million EUR) deal values paid for the firms located in the New Member States of the EU are smaller than the values paid by the companies in these countries. Finally, the gathered data indicates no M&A activated initiated neither by nor towards companies originated in Latvia, Estonia, Hungary and Malta.

3.3 Variables

Two dependent variables are used in the empirical research. On the one hand, average cumulative abnormal returns are chosen as dependent variable in the simple mean comparison tests and represents shareholder wealth changes around the announcement day. On the other hand, cumulative abnormal returns of each event are employed when performing multiple regression analysis.

(33)

33 company country is used as an alternative to test the third hypothesis. Hofstede_diff is the independent variable used in the research to proxy for the fourth hypothesis. Similar to Chakrabarti et al. (2009) a composite index (see Equation 2) that uses numerical characteristics of bidding and target countries is calculated based on the four national cultural dimensions (power distance; individualism; masculinity; uncertainty avoidance) introduced by Hofstede (1991). Appendix B provides country specific information used for calculating GDP per capita and cultural differences between countries of origin of target and bidding companies. A dummy variable (Legal_diff) is created to measure whether legal systems of an acquiring and acquired company match and to test the fifth hypothesis. Specifically, a dummy equals 1 if acquiring and acquired firms country of origin dispose the same legal system (either common or civil law countries), whereas variable 0 follows when legal systems of the firms environment differ. Finally, a dummy variable Law that equals 1 is used if acquirers legal system is English common law (UK, Ireland, Malta and Cyprus), and 0 in the case of civil law.

PCI_diff=

Per capita GDP of acquirer nation – Per capita GDP of target nation

(1) Per capita GDP of acquirer nation + Per capita GDP of target nation

4 ) ( _ 4 1 2



   j SAj STj diff Hofstede (2)

Where SAj is a value of acquirers nation cultural dimension j, whereas STj is targets nation

cultural dimension j.

Control variables used in the thesis include: bidder size (Size), leverage (Lev), transaction value (Deal_size), legal status of the target (List), relative size of the target to the bidder (RMV), relatedness of the acquisition (Rel), method of payment (Paym), bidder book to market ratio (BTM) as well as cross-check of conjunction between announcement and rumor dates (ANN) all together aim to explain the reasons and motivation behind the induced shareholder wealth changes. Information regarding methodological calculations and data gathering approaches of control variables is provided in Appendix A, whereas Table 8 provides descriptive statistics of control variable used in the research.

(34)

34 value of acquiring company’s assets is used. Furthermore, the data suggest that acquiring company’s liabilities amounts on average to 12% of its assets. Remarkable differences between mean (256.87 million EUR) and median (24.75 million) of transaction values are observed. Moreover, it is observed that 95% of M&A’s are directed to private targets, 17% are driven by diversification purpose, 68% are paid entirely with cash and information leakage is found in more than a quarter of events. Besides that, the mean value of relative size infers that bidders tend to acquire targets that are twice smaller than they. Finally, the acquirers’ book to market ratio averaging 0.71 suggests that M&As are commonly initiated by overvalued firms as book value of equity and debt tend to be lower than market value of equity and debt.

Table 8 Descriptive statistics of control variables

Control variable Mean Median Std. dev.

Bidder size (nominal, million EUR) 385.32 21.08 2031.33

Bidder size (Size) 9.66 9.96 3.10

Bidder leverage (Lev) 0.12 0.13 0.27

Transaction value (Deal_size, million EUR) 256.87 24.75 1290.75 Private vs. public M&A’s (List) 0.95 n.a. n.a. Relative size of the target to the bidder

(RMV)

0.51 0.09 2.24

Related vs. diversifying M&A’s (Rel) 0.83 n.a. n.a. Cash vs. other method of payment (Paym) 0.68 n.a. n.a.

Bidder book to market ratio (BTM) 0.71 0.5 2.31

Announcement day equals rumor date (ANN)

(35)

35

4. Event study methodology

Event study methodology is applied in the thesis with the aim to determine whether there exists an abnormal stock price effect associated with an M&A announcement. Considering the reviewed literature, this methodology is relevant for several reasons. Firstly, implementation of event study is based on stock prices which reflect the true value of firms. Therefore results cannot be manipulated like accounting-based measures of profit by choosing specific accounting procedures (McWilliams and Siegel, 1997). Secondly, measuring short-term value changes represent the best estimate of the expected present value to shareholders generated by the transaction (Jensen and Ruback, 1983). Finally, the measured market based returns are direct measures of value created for investors and dispose a forward looking perspective of value creation as stock prices are assumed to present the value of expected future cash flows (Bruner, 2002). In order to execute the study correctly, Brown and Warner (1985), MacKinley (1997) and Bruner (2002) were consulted when measuring abnormal returns experienced by the shareholders of the companies that announce M&A and testing the hypotheses discussed before.

4.1 Estimation and event windows

Figure 2 provides a timeline for a total set of 572 events (M&A announcements) used in the research. Following suggestions by MacKinley (1997), the announcement day of each M&A is identified as day zero. Following Brown and Warner (1985), the period between250 and 10 days prior the M&A announcement is identified as estimation period and is used to estimate the market parameters, to measure the stock’s normal movements and correlation with an outside index. 10 days prior and 30 days after the M&A announcement are treated as an event window during which the security prices of the acquiring companies are examined.

Figure 2. Time window of the study Days

(36)

36 4.2 Benchmark and abnormal returns

The choice of benchmark that is used in the estimation period is the next stage of the event study. The data set used in this thesis covers companies from 27 countries of the European Union. Considering coverage of 17 countries and representation of 75% of European equity market capitalization, broad market exposure in Europe as stated by Eije (2010) and consisting of large and mid-cap firms that are most active members of M&A market in EU as found by Martynova and Renneboog (2008), the S&P 350 index is treated as a benchmark in this thesis.

Following MacKinley (1997), the next step in performing event study is measuring abnormal returns (see Equation 3) that are defined as the difference between the actual security’s return (see Equation 4) over the period of the event window and the expected return without conditioning on the event taking place. This terminology allows the appraisal of the announcement impact. Database Datastream allows collecting stock prices that are adjusted for stock splits and other capital events. Closing firms i stock prices are used to calculate actual returns. Continuously compounded returns are employed as they are assumed to conform the normality assumptions of the regressions better than simple returns and to eliminate negative values (Henderson and Glenn, 1990).

ARi,t = Ri,t – E(Ri,t) (3)

Where ARi,t , Ri,t and E(Ritj) are abnormal, actual and expected normal return of the company

i for the time period t respectively.

1 , , 1 , , , ln ln ln      _i_t _i_t t i t i t i P P P P R (4)

Where Pi,t represents firms’ i closing stock price on day t and Pi,t-1 stands for firms i stock

price of the trading before day t.

Similarly to Eije and Wiegerinck (2010), Chang (1998), Conn et al. (2005) and other researchers, each firm’s expected normal returns are measured using residual market model as discussed by Brown and Warner (1985). Equation 5 provides a statistical relationship used in the estimations of the expected normal returns.

(37)

37 Where αi and βi are firm specific variables calculated using CAPM model for 240 trading

days length estimation window (-250;-10 days before deal announcement). αi is an intercept

coefficient, βi is a slope coefficient that captures firm’s systematic risk and Rm,t represents a

S&P350 return on day t.

βi=

cov(Ri,Rm) (6)

var (Rm)

Where cov(Ri,Rm) is covariance between individual stock and S&P350 and var(Rm) stands

for the variance of market portfolio.

αi = avg(Ri) - βi * Rm (7)

Having abnormal returns identified for each firm during the event period (-10 days before and +30 days after the M&A announcement) allows to measure cumulative abnormal returns (CARi,T) for different event windows (T) as well as cumulative average abnormal

returns ( ) for the total sample of N events as follows:

(8) (9)

(38)

38 4.3 Tests

When discussing the methodology used in event studies, Henderson and Glenn (1990) highlighted the fact that a set of statistical assumptions lies behind the regression models and provided equations. Residuals of regression models are assumed: 1) to be normally distributed; 2) not to be serially correlated; 3) dispose a constant variance; 4) not to be correlated with explanatory variables. Jarque-Bera test is employed when investigating distribution of the residuals. Furthermore, I employ two parametric and two non-parametric tests the retrieved abnormal and cumulative abnormal returns as by themselves they are not sufficient enough to draw solid conclusions and reliable implications.

4.3.1 Parametric tests

Traditional test (see Equation 10 for ARs and Equation 11 for the case of CARs ) used in the research assumes uncorrelated security residuals and insignificant event induced variance. When forming traditional test, sum of the event period abnormal returns is divided by the square root of the sum of all securities estimation period residual variance.

(10) (11)

Where l stands for the days in the event window that are used for the CAR calculations.

Similarly to the above discussed traditional test, the standardized residual test assumes uncorrelated securities and insignificant event induced variance. The key of standardized residual test is ensuring that each abnormal return will have same variance. This is obtained through standardization: each firm’s abnormal residual measured in the event window is divided by its standard deviation retrieved from the estimation window (see Equation 12). Equation 13 describes the formula of standardized residual test for the case of abnormal returns, whereas Equation 14 documents the case of cumulative abnormal returns’ testing.

(39)

39 (13) (14)

Similarly to Dodd and Warner (1983) the thesis incorporates z-test (see Equation 15) when testing whether differences between cumulative average abnormal returns ( ) of different subsamples exists.

z=

(15)

Where N stands for number of observation and L is event window period.

4.3.2 Non-parametric test

Non-parametric tests supplement the empirical analysis as they are well-specified and powerful in detecting fall null hypotheses of no abnormal returns. Generalized sign and rank test are employed in the thesis as they are identified by Kolari, and Pynnönen (2010) as commonly used and most reliable tests of events effects. Originally, Corrado (1989) and Corrado and Zivney (1992) suggested to apply non-parametric sign and rank tests as robustness checks in the event studies. The authors argue that these tests are valid robust against event-induced volatility and cross-correlation.

Following definition of generalized sign test by Cowan (1992), this test examines whether the frequency of positive cumulative abnormal returns in the event window exceeds the number expected in the absence of M&A announcement. Equations 16-18 provides the steps involved in the performance of generalized sign test statistic.

=

(16)

= (17)

(40)

40 Where N stands for the number of firm’s j; L1 and L2 limits the beginning and the end of the estimation window; expected fraction of positive returns; w is the number of stocks in the event window for which CAR are found positive.

Following rank test procedure explained by Corrado (1989), the 240 days length estimation period and the 41 days length event period for each company j is treated as a single time series, where rank is assigned to each daily return. Kjt stands for the rank of

abnormal return in the time series of 281 days. Rank 1 indicates the lowest return, rank 281 the highest, whereas the mean rank is 141. Equation 19 is used to measure the rank test statistic of cumulative abnormal returns during the event window T that consists of days L1 to

Lt.

(19)

Where signifies the average rank across the n stocks and days l of the event window (1≤l≤41) and represents the average rank across the n stocks on day t of the combined estimation and event period (281 days).

4.4 Cross-sectional analysis

The developed hypotheses are tested using multiple cross-section regression analysis (Equation 9). Hypotheses 2-6 are supported when coefficients α2-α6 are found to be significantly positive while considering the effects of including control variables.

CARi=α0+ α1*NMS + α2*CBA + α3*PCI_diff + α4*Hofstede_diff + α5*Legal_diff

+ α6*Common + α7*Size+ α8*Lev + α9*Deal_size + α10*List + α11*RMV +

α12*Rel + α13*Paym + α14*BTM + α15*ANN + εi

(9)

Where CARi is acquirers i cumulated abnormal return for the event window T; NMS is a

(41)

41 1 when the origin of the firm is common law and 0 otherwise; Size is log of pre-merger book value of acquiring company’s assets; Lev is acquirers 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; List is a dummy variable set to 1 when target is listed company and 0 otherwise; RMV is the market value of targets equity divided by market value of the acquirers’ equity, in the case target is not publicly listed, the Euro value of the acquisition was used as the market value; Rel is a dummy variable set to 1 when target and bidder operate in related industries and 0 otherwise; Paym is a dummy variable set to 1 when deal is entirely financed with cash and 0 otherwise; BTM is acquirers book value of equity plus book value of debt over the sum of market value of equity plus book value of debt prior to the bid; ANN is a dummy variable set to 1 when announcement and rumor dates match whereas 0 follows otherwise.

(42)

42

5. Empirical results

This part of the thesis provides the results of analysis where shareholder wealth changes stand in the discussion center. Firstly, stock price changes in the announcement, pre- and post-announcement periods are discussed. Secondly, section 5.2 provides the results of the univariate analysis which investigated whether differences between different subsamples relevant to hypotheses 2-6 exist. Finally, results of the ordinary least squares regression supplements the discussion regarding independent variables and reviews the effects of the chosen control variables.

5.1 Shareholder wealth changes surrounding M&A announcement

Based on the observed average abnormal returns and values of the T-tests, generalized sign and rank tests of the average abnormal returns for the whole sample provided in Appendix C, Figure 3 visualizes the estimates of abnormal and cumulative abnormal returns in the event period for the whole sample of companies. On average, acquisition announcement leads to statistically significant 1.92 % increase in stock price on the event day, whereas cumulative abnormal returns throughout the whole event window (-10;+30 days) accumulate a statistically significant value of 1.85%.

Figure 3. M&A announcement effects

Even though positive announcement effects occurred on the announcement day and accumulated throughout the whole event window of 41 days, negative abnormal returns dominated the sample as they were identified in 23 days (56%). Furthermore, the dotted line in Figure 3 which stands for cumulative abnormal returns in the event window suggests that

-0.50% 0.00% 0.50% 1.00% 1.50% 2.00% 2.50% 3.00% 3.50% -0.75% -0.50% -0.25% 0.00% 0.25% 0.50% 0.75% 1.00% 1.25% 1.50% 1.75% 2.00% 2.25% -10 -7 -4 -1 2 5 8 11 14 17 20 23 26 29 Average CAR (%) Average AR (%)

Days around the announcement

(43)

43 information leakage potentially exists in the market and inside information may cause significant price changes before the official M&A announcement. In line with the assumption of existing efficient market and informational efficiency, the price change on (0;1) is higher in percentage value and significance terms than that on (-1;0). Besides that, transitory period during which abnormal returns reverse and subsequently accumulated abnormal return shrink is observed after day one.

Considering the significant abnormal returns during 3 days length event window around the announcement day (see Appendix C), this window is used for hypotheses testing. Furthermore, as suggested by the dynamics of cumulative abnormal returns following Figure 3, the event window (-10;-2) is used to test whether significant pre-announcement effects exist while the event window (+2; +30) is used to investigate the post-announcement effects. Table 9 reports the results of univariate analysis of cumulative abnormal returns during announcement, pre- and post-announcement event windows.

Table 9 Cumulative abnormal returns during different event windows

Event window Average CAR T-test Standar dized T-test Jarque-beta % of positive Genera-lized sign test Rank test (-1;+1) 2.91% 6.32* 27.31* 78872.38 65.56% 7.40* 8.76* (-10;-2) -0.13% -0.47 -3.45* 615.91 45.45% 2.13** 2.02** (+2;+30) -0.90% -1.48 -2.01** 594.25 46.68% 1.55 1.20

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

M&A announcement effects to bidding company shareholders: the case of European Union during 2004-2011