Impact of financial market integration on cross border acquisitions in European Union transition countries

(1)

Impact of financial market integration on cross

border acquisitions in European Union transition

countries

________________________________________________________________________

Roelof van Dijk

University of Groningen

Faculty of Business Administration and Economics, Finance department

April 2008

(2)

Summary

In this research it is found that both European Union (EU) and Non-EU firms create positive announcement returns when undertaking acquisitions in EU transition countries. Moreover, it is found that there is a significant difference between EU bidder shareholder return and Non-EU bidder shareholder return when acquiring in EU transition countries. By referring to the literature of Francis et al. (2008), it is not found that increasing worldwide and European financial integration of EU transition countries creates different bidder shareholders return between EU and Non-EU acquirers. The results indicate that EU acquirers are not negatively affected by worldwide and European financial integration of EU transition countries, but Non-EU acquirers do. Furthermore, the fact that EU acquirers are not negatively affected, despite of higher European financial integration, suggest that European acquirers create cost advantages when acquiring in EU transition countries.

JEL Classification: G14, G34

(3)

1. Introduction

In the last decade interest in the theory of mergers and acquisitions has grown increasingly. One of the reasons is that since the middle of the 1990s there has been an enormous increase in merger and acquisition activity throughout the world. This increase is in part due to the fact that firms have gone away from traditional “Greenfield”investments. It is also due to the growth in international financial markets, which has allowed firms to pursue investment opportunities both at home and abroad (Francis et al., 2008).

At this moment the sixth merger and acquisition wave is occurring worldwide and in Europe the fifth since 19001. The current merger wave started approximately in 2002. One of the first M&A waves in Europe was driven by a focus for economics of scale and since 1990 to get a more European identity. By enlarging the European Union (hereinafter “EU”) and opening up the financial markets, a more international focus was noticeable (Martinez Torre-Enciso & Bilbao Garcia, 1996).

Literature about the effect of cross-border M&As on bidder shareholders is mixed. Doukas and Travlos (1988) and Kiymaz (2004) found out that cross border M&As are value enhancing for bidder shareholders. However, Moeller and Schlingemann (2005) and Denis et al. (2002) found out that cross border M&As decrease bidder shareholders return. The contribution of this research is that it examines cross border acquisitions in a more recent time period. Furthermore, it examines bidder shareholders gains or losses from acquisitions involving targets from transition markets (Central and Eastern Europe countries, CEE). Transition markets are markets which are moving from a planned economy to a more free economy.

In the late 1980s and early 1990s CEE countries and the Baltic States (Estonia, Latvia and Lithuania) instituted reforms that liberalized their financial markets, thereby allowing firms to acquire domestic firms. Errunza and Miller (2000) pointed out that despite liberalization, most firms from these markets still face a higher cost of capital compared to firms from integrated markets. That most firms from transition countries face a higher cost of capital means that acquirers can provide funding to financially constrained firms either through internal capital markets or indirectly through access to external capital markets. When this funding result in the fact that these firms can undertake positive NPV projects and stock markets are efficient, these benefits should reflect in positive abnormal returns following the M&A announcement.

Although financial integration takes time, it is noticeable that European transition markets get more financially integrated into European and worldwide capital markets. Therefore, by applying the theory of Francis et al. (2008), who state that undertaking M&As in financially integrated markets result in

1

(6)

lower bidder shareholder value, it is expected that both EU and Non-EU acquirers gain less shareholder value when undertaking acquisitions in EU transition countries.

Furthermore, it is expected that EU acquirers (which are as well part of EU integration) are more affected by European financial integration compared to Non-EU acquirers when undertaking acquisitions in EU transition countries. Although both EU and Non-EU acquirers will probably create positive bidder shareholder returns when undertaking acquisitions in transition / emerging markets2, it is hypothesized that EU bidder shareholders create different announcement returns compared to Non-EU bidder shareholders, due to European financial market integration of Non-EU transition countries.

Using a sample of 191 EU cross border acquisitions and 92 Non-EU cross border acquisitions in EU transition countries for the period 1997-2007, we found out that both EU and Non-EU bidder shareholders create positive announcement returns, however insignificant. Moreover, the difference between the announcement return of Non-EU and EU bidder shareholders is significant and therefore supports the hypothesis that EU bidder shareholder create different announcement returns compared to Non-EU bidder shareholders when undertaking acquisitions in EU transition countries. When the sample 1997-2007 is separated into 1997-2002 and 2002-2008 sub-periods, it is found that EU bidder shareholders create significant higher bidder shareholder value in the period 2002-2008. However, Non-EU bidder shareholders create significantly lower bidder shareholder value in the period 2002-2008. That EU acquirers are positively affected by higher financial integration of EU transition countries compared to Non-EU acquirers was not expected. However, it was expected that Non-EU acquirers would create less shareholder return because of worldwide financial integration. The difference in the period 2002-2008 between EU and Non-EU bidder shareholder return is not significant. Because of the insignificant outcome these results do not support the hypothesis that increasing financial integration of EU transition countries does result in different announcement returns between EU and Non-EU bidder shareholder value when undertaking cross border acquisitions in EU transition countries.

In sum, it is found that EU bidder shareholders create significant different shareholder value compared to Non-EU bidder shareholders when undertaking cross border acquisitions in EU transition countries. However, it can not be said that increasing financial integration of EU transition countries does result in significant different announcement returns between EU and Non-EU bidder shareholder returns when undertaking cross border acquisitions in EU transition countries. It can be said that Non-EU acquirers create less shareholder return when undertaking acquisitions in EU transition countries, because of worldwide financial integration of these countries. Moreover, it can be said that EU

(7)

acquirers create more shareholder return when undertaking acquisitions in EU transition countries. This is probably due to the fact that European financial integration leads to costs advantages for European acquirers when undertaking acquisitions in EU transition countries.

This paper makes several contributions to the cross border literature. First, it is found that EU bidder shareholders create different announcement returns compared to Non-EU bidder shareholders when undertaking cross border acquisitions in EU transition countries. Secondly, that EU acquirers create significant more shareholder value when undertaking acquisitions in the period 2002-2008 compared to the period 1997-2002. This in contrast to the literature of Francis et al. (2008) who state that higher financial integration results in lower shareholder value when undertaking M&As.

(8)

2 Literature overview

2.1 Determinants of share price reactions

M&A literature has shown that a variety of characteristics affect the value of bidding firms surrounding the announcement of acquisitions.

2.1.1 Hostile or friendly

According to Gregory (1997) when a tender offer or hostile acquisition occurs, it generates higher bidder returns compared to friendly M&As. These findings are consistent with the view that unwanted suitors are entrepreneurs who have uncovered special value-creating insights about the target firm. By making an unsolicited bid, the buyer seeks to retain value for itself, rather than give it up in a negotiation (Bruner, 2002). On the other hand, Healey, Palepu and Ruback (1997) found that hostile deals were associated with insignificant improvements in cash flow returns, owing possibly to the payment of higher acquisition premiums.

2.1.2 Method of payment

Many literature has been published about the influence of the method of payment on the acquirers return. The usual argument for the method of payment effect is that cash bids do better for bidder shareholders, because the market takes this as a positive signal of bidder expectations of future returns (Travlos, 1987). The positive signal a cash payment gives to shareholders about the confidence of the acquirer is not relevant for cross border deals. Because, uncertainty about the information of the acquirer may force acquiring firms to pay with equity. This may be especially true for private overseas deals where the information can be even more imperfect. Moreover, the neutralised effect of the method of payment may be caused as well by the fact that sometimes target shareholders force bidder management to pay in cash instead of accepting foreign equity (Gaughan, 2002).

2.1.3 Public or private

(9)

Thirdly, the information in private deals is mostly not exposed to codes and rules and therefore better and easier accessible for shareholders. Finally, in private bids the payments of the significant premium, which characterise contested public acquisitions are less likely and bidder returns are therefore more likely to be positive as a result of the discount.

2.1.4 Relative size

The relative size of the bidder and target influences the M&A success. Beitel et al. (2004) found out that the acquisitions of smaller targets are less complex and, although the scale effects might be smaller, the realization of potential for value creation may be easier. Furthermore, Moeller, Schlingemann and Stulz (2004) found out that smaller bidding firms create two percentage points more return than larger bidding firms. They sum up several explanation why that is the case, but they only found significant evidence that managerial hubris influences decisions of large firms.

2.1.5 Excess cash

It is well known that acquiring firms with excess cash destroys value because of overbidding. This is mostly true because companies with excess cash pay too much money for overvalued acquisitions. Several papers have unearthed evidence that free cash flow (Jensen, 1986) is frequently used for managerial empire building (see e.g. Servaes, 1991; Lang et al., 1991). These managers acquire these overvalued companies, because they are overconfident and not because they want to generate shareholder value. Moreover, Georgen and Renneboog (2004) show as well that excess cash is value destructing, because managers pay too much for their acquisitions.

2.1.6 Equity stake bidding management

Healy et al. (1997) and Agrawal and Mandelker (1987) show that when the acquirers management holds large equity stakes, the share price reaction of bidding firms are higher. When the bidder management does not own equity, the agency problem is arising. Shareholders of the bidding firms believe that when management does not hold equity, they will undertake value destroying activities rather than maximizing shareholder value.

2.1.7 The acquisition of control

(10)

acquirer split these gains in a constant ratio. Furthermore, acquirers could pay different prices when they acquire control of the target.

2.1.8 Experience bidder

Former findings found out that more experienced bidders are able to generate higher synergies and therefore generate higher returns (DeYoung, 1997 and Zollo and Leshchinkskii, 2000). The experience of the bidder can be measured in two ways. Firstly, the M&A frequency of earlier cross border deals. Secondly, if the bidder had already a minority stake in the target firm. Kaufmann (1988) found out that bidding firms with a prior ownership interest in the target firm, pays a smaller premium as the bidder’s increases its stake of ownership. The fact that majority control creates value is also shown by Chari et al. (2004) regarding emerging markets.

2.1.9 Industrial diversification

Most of the M&A literature published, state that it is better for bidder shareholder value when acquirer and target are related in some way, for instance in the same industry. Berger and Ofek (1995) found an average loss in value from diversification of between 13% and 15%. The degree of relatedness between the businesses of the buyer and seller is positively associated with higher returns. Intuitively, this makes sense if synergies or savings arise from the economics of the two firms. In particular, conglomerate deals (i.e., deals between firms with unrelated lines of business) are associated with the poorest returns. Furthermore, Denis et al. (2002) found out as well that industrial diversification is associated with a significant firm discount.

2.1.10 Industry

(11)

2.1.11 Target country characteristics

The institutional characteristics of the target country influences the return of bidder shareholders. Moeller and Schlingemann (2005) type these characteristics as trade policy, government intervention and capital restrictions. These characteristics may affect the efficiency of the market for corporate control, cost of capital, costs of exporting and importing. Moeller and Schlingemann (2005) argue that when a company acquirers a firm in a less restrictive institutional environment, they experience higher gains because of reduced agency problems and less asymmetric information. However, Francis et al. (2008) found out the opposite, which means that there would be a positive relationship between a country’s economic restrictiveness and the acquirer’s abnormal returns. By using the index of Economic freedom of the Heritage foundation ( Gwartney et al., 1996) as a variable for the univariate analysis something can be said about the influence of the restrictiveness of the target country. Furthermore, Moeller and Schlingemann (2005) state that the economic restrictiveness says something about the targets financial market integration.

2.1.12 Time period

This research tries to find out whether (European) financial market integration influences announcement returns. Therefore, the sample is divided into two time periods. The first time period is from 1 January 1997 until 31 December 2001 and the second period from 1 Janaury 2002 until 31 December 2007. In 2002 the Euro was introduced as a common currency throughout the European Union. Furthermore, in 2004 (10) and in 2007 (2) new member states joined the European Union. It is expected (based on the literature of Francis et al., 2008) that due to the increasing financial market integration throughout time, EU acquirers create different shareholder returns compared to Non-EU acquirers in the period 2002 until 2008.

2.2 Value creation of cross border acquisitions in emerging markets

In this section it is explained why cross border acquisitions should create more value in emerging markets instead of developed countries. The target countries used in this research are developing and/or emerging markets. Which means that these countries have sustained economic growth over the years and exhibit good economic potential. Furthermore, foreign participation was fairly restricted till the 1990s, when many emerging market countries deregulated their capital markets to foreign entry (Henry, 2000).

(12)

returns in emerging markets were inherent by gaining a majority of control of the target firm. Through the transfer of control to the developed market acquirer, the boundaries of the firm are extended into an environment where contracting and property rights are weak, the M&A transaction completes the market and there are substantial gains for shareholders of acquiring firms.

Chari et al. (2006) described three circumstances where the acquirer should gain more return in emerging markets. Firstly, a better bargaining position for the acquirer could generate a positive return, because in emerging markets there are presumably less bidders who compete for the target. The ability of developed market acquirers to provide access to finance may be particularly valuable for targets located in capital-scarce emerging markets. This situation increases in importance during financial crisis when cash-strapped targets have liquidity needs. Chari et al. (2006) measured bargaining power by including acquirer and target size as possible indicators of firm bargaining power. Secondly, if the target is uncertain about its own true stand-alone value, the target firm may undervalue its assets. This is creating possibilities for the acquiring firm when they have better information about the stand-alone value of the target firm. Moreover, the stock price in emerging markets is often viewed as a noise estimate of true firm values. This will create additional information asymmetry, which create advantages for acquiring firms who are better capable of valuing firms. Thirdly, developed acquirers from developed countries gain more return if they are able to acquire a majority control of the target firm. The majority control enables acquirers to shift the boundary and transfer technology or invest in the target. If the transfer of control leads to an increase in investment and transfer of technology, joint returns should increase with control.

To summarize, there are a limited of studies that focus especially on acquisitions involving emerging market targets. Chari et al. (2006) hypothesize that shareholders generate positive returns with cross border acquisitions in emerging markets if they have better bargaining power, are able to form better estimates of the true stand alone value of the target and when they acquire a majority of control of the target firm. Furthermore, Waheed and Mathur (1995) and Kiymaz (2004) show as well that cross border acquisitions in developing countries create value for the bidder shareholders. They conclude that the diversification benefits, in conjunction with the advantages of lower competition in developing countries, outweigh the political risk associated with expansion in such economies.

2.3 Value creation in EU transition countries

(13)

Firstly, CEE and Baltic countries are moving from a planned economy to a market economy. Although there are still some problems with bureaucracy and political uncertainty, the common European legal environment and subsidy payments have aided transition. Furthermore, in terms of the transaction cycle, the CEE region is moving closer to Western Europe regarding transparency and speed. More bureaucratic state owned companies become private and it is possible for foreign investors to trade in shares and to invest in those companies. Secondly, the CEE region and the Baltic States have become a bridge, for acquirers, to new emerging markets like Turkey and Russia. The central heart of the EU is moving eastwards since the admission of the new EU member states in 2004 and 2007. Thirdly, the wages are relatively low, the workforce is well educated and the GDP growth is high the last years. This factors explain why, especially EU transition countries, are attractive regions for foreign investors to build production sites and service centres (Ernst & Young, 2006).

2.4 Financial market integration of EU transition countries

In this section theoretical background is given why Non-EU acquirers should create different shareholder value compared to EU acquirers, when undertaking acquisitions in EU transition countries.

In the 1980s and early 1990s, CEE markets instituted reforms that liberalized their financial markets, thereby allowing foreign firms to acquire domestic firms (Francis et al., 2008). Although liberalization is synonymous with integration, these financial markets were at best only partially integrated into world capital markets. As Francis et al. (2008) stated it takes time to be fully integrated into world capital markets. Errunza and Miller (2000) state that despite liberalization, firms from these markets still have higher cost of capital compared to firms from integrated markets. Francis et al. (2008) found out that acquisitions in financially integrated markets create less shareholder value compared to acquisitions in segmented markets. In addition they found out that for the period 1996 until 2003 U.S. acquirers gained more shareholder return when they acquired in segmented markets. Reason for that is to provide funding to financially constrained firms either through internal capital markets or indirectly through access to external markets. When these firms where financial healthy again they could undertake positive NPV projects, which otherwise was not possible. These benefits should be reflected in positive abnormal returns following the M&A announcement.

(14)

fuelled by the creation of liquid European capital market which provides companies with new sources of financing (Martynova and Renneboog, 2006). According to Ernst and Young (2006) the EU transition countries are moving closer to Western Europe and are more integrated into the European financial market.

The article of Francis et al. (2008) argues that in the period 1997 until 2002 many countries that undertook liberalization, had still high cost of capital and were not integrated markets yet. Therefore, it was possible for both EU and Non-EU companies to undertake value enhancing acquisitions in these segmented markets. However, since 2002 the European financial integration is going faster and EU transition countries get more worldwide and European financially integrated. This development, referring to the article of Francis et al. (2008), means that EU acquirers should create relatively lower shareholder return, because they are now undertaking acquisitions in a more financially integrated and less segmented market. In addition, since Non-EU acquirers are less directly part of the European financial market integration compared to EU acquirers, it is expected due to the increasing speed of European financial integration that Non-EU acquirers should create different shareholder value in the period 2002 until 2008.

Bjorvatn (2004) states that European integration will result in increased competition between companies and that could lead to both shareholder creation and shareholder destruction regarding cross-border M&A. An advantage, according to Bjorvatn (2004), is that economic integration may intensify the pre-merger competition in the market and thereby reducing the reservation price of the target firm. In addition, economic integration in the form of lower trade costs may reduce the post-merger business stealing effect as the outside firm chooses exports rather than Greenfield investment. However, disadvantages could be that due to the increasing competition in the market for corporate control, a smaller part of the gains are for the acquirers. Furthermore, when trading costs are lower it would be less expensive for domestic firms to export their products and thereby enforcing the competition with firms who undertook a cross-border acquisition (which is called “the business stealing effect”). In addition, an increase in market integration and a reduction in costs of cross-border acquisitions could result in managerial hubris and agency problems (Denis et al., 2002). Managerial hubris means that managers get too overconfident and mostly pay too much for acquisitions which will not lead to synergistic gains.

(15)

3 Hypothesis development

The amount of acquisitions in CEE and the Baltic States, by both EU and Non-EU companies, show that these regions are attractive for foreign companies to undertake acquisitions. One of the reasons for this popularity is the emerging environment of the EU transition countries. Which means that these countries show a stable or developing macroeconomic situation with low inflation, low unemployment and high GDP indicators. Chari et al. (2006) showed that developed market acquirers earn significant more bidder shareholder return when acquiring in emerging markets instead of developed markets. Positive returns for bidder shareholders to emerging market acquisitions are created by bargaining power for bidders, advantage of information asymmetry and access of capital. These positive returns are strengthened, according to Chari et al. (2006), when the acquiring company gains a majority control of the target company. By referring to the article of Chari et al. (2006) it is expected that EU and Non-EU bidder shareholder create different announcement returns when acquiring in EU transition countries. Therefore, the first hypothesis is conducted as follows:

:

0

H

EU bidder shareholder do not create different announcement returns compared to Non-EU bidder shareholders when undertaking cross border acquisitions in EU transition countries

:

1

H

EU bidder shareholder do create different announcement returns compared to Non-EU bidder shareholders when undertaking cross border acquisitions in EU transition countries (1)

(16)

:

0

H

increasing financial integration of EU transition countries does not result in different announcement returns between EU and Non-EU bidder shareholder returns when undertaking cross border acquisitions in EU transition countries

:

1

H

increasing financial integration of EU transition countries does result in different announcement returns between EU and Non-EU bidder shareholder returns when undertaking cross border

acquisitions in EU transition countries (2)

4. Data

4.1 Data selection

This research will focus on acquisitions in European Union transition countries3 for the period 1 January 1997 until 31 December 2007. The reason for starting from 1997 is because the database of Zephyr (Bureau Van Dijk, 2007) does not collect data before 1997. The ten years time period is chosen, because there was a growing amount of cross border acquisitions noticeable throughout the years. Furthermore, in 2002 the Euro was introduced and the EU admitted ten new member countries in 2004 and 2007. Only Cyprus is left out of the sample, because that is not a country in the transition phase. To keep practical relevance the period of ten years will be split up in two sub periods. The first period is from 1 January 1997 until 31 December 2001 and the second period from 1 January 2002 until 31 December 2007. By splitting up the research in these two time periods, it can be measured whether there is a difference between the announcement returns of EU and Non-EU bidder shareholders when undertaking acquisitions in EU transition countries.

Secondly, this research will not focus on a particular industry. There has not been any research yet which investigated acquirer return of cross-border M&As in European transition countries in general. However, there have been papers who investigated the announcement return of bank M&As in emerging markets (Soussa and Wheeler, 2006), of which European emerging markets as well. Moreover, Chari et al. (2006) used stock price data for firms in a range of industries to assess the benefits of emerging markets acquisitions by developed firms. To measure the impact of different target industries, the sample will be controlled for the sectors consumer goods, consumer services, financials and industrials. These industries are controlled for, because most acquisitions were announced in these industries.

Thirdly, only acquisitions are considered. The reason for this is that only a few mergers occurred in the time period investigated. By omitting the mergers, it is possible to conduct a valid conclusion only for acquisitions.

(17)

Finally, this research is focused on public companies from EU4 and Non-EU countries. To calculate the abnormal return for bidder shareholders around the announcement date, stock prices need to be available. The target company can be either public or private, but has to be situated in one the EU transition countries, as stated earlier.

4.2 Sample selection

The data of the two samples which will be used for this research are collected from the Zephyr database of Bureau van Dijk. This database describes announcement dates, deal values and other characteristics of M&As from 1997. The two samples are conducted as follows:

1. Companies located in one of the 17 European Union countries that announced an acquisition with a company in an European Union transition country in the period 1 January 1997 until 31 December 2007;

2. Companies located in Non-European Union countries that announced an acquisition with a company in an European Union transition country in the period 1 January 1997 until 31 December 2007.

To make the samples suitable for the event study, some eliminations will have to be made: 1. The private bidders are excluded; only public bidders are considered;

2. The rumor date should equal the announcement date; the rumor date should not influence the announcement return;

3. Bidders without an ISIN code are excluded; the ISIN codes are necessary for collecting data from Datastream (Thomson Financial, 2007).

After these eliminations the first sample consists of 206 deals and the second sample of 103 deals. An important note is that half of the sample acquirers undertook multiple acquisitions during the sample period 1997 until 2008. Multiple acquisitions by the same acquirer over a short period may lead to biases, because with multiple acquisitions it is hard to distinguish the effect of the shocks from each other. Therefore, the market adjusted return model (Brown and Warner, 1985) is used in this research. How the market adjusted return model takes into account the overlapping observations is discussed in paragraph 5.1.2.

4.3 Sample description

Table 1 highlights which EU transition countries are most popular to acquire in for both samples. As can been seen Non-EU companies are acquiring more (in percentage) in the Czech Republic (21,4%),

(18)

Hungary (18,4%), Romania (13,6%) and Slovenia (7,8%) compared to EU companies. Noticeable is that Non-EU companies favor to invest in the European emerging markets The Czech Republic (21,4%), Hungary (18,4%) and Poland (24,3%). A possible explanation for their preference in emerging countries, is that these countries show good economic development for several years now. Furthermore, there is more information about these markets and makes it more secure to acquire companies. Remarkable is that EU acquirers undertake more acquisitions in the Baltic States and South Eastern Europe countries. This is probably because Scandinavian EU countries are familiar with the investment climate in the Baltic States.

Table 1 Number of acquisitions in EU transition countries

This tables below presents the amount of acquisitions that were undertaken in the different EU transition countries. The amount of acquisitions is shown as # and the percentage relative to the total amount is shown as %. Both the EU sample and Non-EU sample are investigated.

Non-EU sam ple EU sam ple

# % # % Target country Bulgaria 4 3,9% 20 9,7% Czech Republic 22 21,4% 26 12,6% Estonia 6 5,8% 11 5,3% Hungary 19 18,4% 28 13,6% Latvia 1 1,0% 5 2,4% Lithuania 1 1,0% 9 4,4% Poland 25 24,3% 70 34,0% Romania 14 13,6% 21 10,2% Slovenia 8 7,8% 4 1,9% Slovakia 3 2,9% 12 5,8% Total 103 100% 206 100%

Table 2 and 3 show the geographical dispersion of acquirers by country. U.S. companies are the most aggressive acquirers for the Non-EU sample, namely 51 acquisitions. The Czech Republic (22 deals), Hungary (19 deals) and Poland (25 deals) were the most favorable targets. Furthermore, in table 3 can been seen that Austria (33 deals), France (30 deals) and Germany (27 deals) are the most frequent acquirers looking at the EU sample. Again The Czech Republic (26 deals), Hungary (28 deals) and Poland (70 deals) are favorable targets.

(19)

Table 6 shows that for both samples almost 75% of the method of payment is unknown. The deals for which the method of payment is known are for both samples almost all paid in cash (19,9% for the EU sample and 17,5% for the Non-EU sample). For both samples companies mostly acquired a majority stake (84,5% for the EU sample and 78,6% for the Non-EU sample). The percentage of deals that were completed is higher for Non-EU companies (81,6% compared to 73,8%). Therefore, the percentage of announcements is higher looking at the EU sample (10,7% compared to 19,4%).

Table 6 shows as well a breakdown of the acquisitions by industry5. Looking at the Non-EU sample most acquisitions are made in the Industrials sector (23,3%). The European sample shows that most acquisitions are made in the Industrials (23,8%) Consumer Goods (23,3%) sector. Furthermore, more acquisitions were made in the financial sector (18,9%) regarding the EU sample and Non-EU acquirers undertook more acquisitions in the Basic Materials sector (14,6%).

(20)

Table 2: Amount of acquisitions by acquirer in EU transition countries (Non-EU sample)

This table presents the total amount of M&As undertaken by each acquirer country in the period 1 January 1997 until 31 December 2007. All acquirer countries are Non-European Union countries. The total amount of acquisitions is 103 for the Non-EU sample.

Non-EU sam ple Target

country Bulgaria Czech Republic Estonia Hungary Latvia Lithuania Poland Romania Slovenia Slovakia Total Acquirer country Australia 1 1 Bermuda 2 1 3 Canada 3 2 3 8 Cayman Islands 1 1 Croatia 1 1 2 Iceland 1 1 India 2 2 Israel 1 1 2 Japan 1 1 1 3 Republic of Korea 1 1 Norw ay 1 2 1 1 1 1 1 8 Philippines 1 1 Russian Federation 2 1 1 2 4 10 Sw itzerland 1 2 3 3 9 Ukraine 0

United States of America 1 12 2 8 19 3 3 3 51

(21)

Table 3: Amount of acquisitions by acquirer country in EU transition countries (EU sample)

This table presents the total amount of M&As undertaken by each acquirer country in the period 1 January 1997 until 31 December 2007. All acquirer countries are European Union countries. The total amount of acquisitions is 206 for the EU sample.

EU sam ple Target

(22)

Table 4 Amount of acquisitions by time period in EU transition countries (Non-EU sample)

This tables shows the amount of M&As undertaken in EU transition countries by year. The time period investigated is from 1 January 1997 until 31 December 2007. This tables presents the Non-EU sample.

Non-EU sam ple

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Total Target countries Bulgaria 2 2 4 Czech Republic 1 1 4 2 1 1 5 1 6 22 Estonia 1 1 2 1 1 6 Hungary 1 1 3 1 2 1 3 4 1 2 19 Latvia 1 1 Lithuania 1 1 Poland 3 6 1 2 3 4 1 1 3 1 25 Romania 1 2 1 2 2 2 3 1 14 Slovenia 1 1 1 3 Slovakia 1 1 1 1 2 1 1 8 Total 5 9 8 9 10 9 5 10 18 6 14 103

Table 5 Amount of acquisitions by time period in EU transition countries (EU sample)

This tables shows the amount of M&As undertaken in EU transition countries by year. The time period investigated is from 1 January 1997 until 31 December 2007. This tables presents the EU sample.

(23)

Table 6 Takeover characteristics

This table presents the takeover characteristics by both samples investigated. The characteristics investigated are the method of payment, deals status, deal type and industry. The amount of M&A is presented by # and the percentage of this amount relative to the total amount is denoted by %.

Acquirer European sam ple Non-European sam ple

# % # %

Method of paym ent

Cash 41 19,9% 18 17,5%

Converted Debt 1 0,5% 3 2,9%

Debt 0 0,0% 0 0,0%

Def erred payment 0 0,0% 1 1,0%

Earn out 0 0,0% 1 1,0% Loan notes 0 0,0% 0 0,0% Shares 3 1,5% 4 3,9% Unknow n payment 161 78,2% 76 73,8% Total 206 100% 103 100% Deal status Announced 40 19,4% 11 10,7% Completed 152 73,8% 84 81,6% Pending (regulating 11 5,3% 5 4,9% approval) 0,0% Pending (shareholder 0 0,0% 0 0,0% approval) 0,0%

Pending (no reason 1 0,5% 1 1,0%

specif ied) 0,0% Postponed 0 0,0% 0 0,0% Withdraw n 2 1,0% 2 1,9% Total 206 100% 103 100% Deal type Majority stake 183 88,8% 88 85,4% Joint Ventures 0 0,0% 0 0,0% Minority stake 23 11,2% 15 14,6% IPO 0 0,0% 0 0,0% Merger 0 0,0% 0 0,0% Institutional buy-out 0 0,0% 0 0,0% Total 206 100% 103 100% Industry Basis Materials 16 7,8% 15 14,6% Consumer Goods 48 23,3% 20 19,4% Consumer Services 34 16,5% 20 19,4% Financials 39 18,9% 7 6,8% Healthcare 2 1,0% 2 1,9% Industrials 49 23,8% 24 23,3%

Oil and Gas 1 0,5% 1 1,0%

Technology 9 4,4% 8 7,8%

Telecommunications 3 1,5% 3 2,9%

Utilities 5 2,4% 3 2,9%

(24)

5. Methodology

5.1 Event study

The event study is an important research tool in economics and finance. The goal of an event study is to measure the effects of an unexpected event on the value of firms (MacKinlay, 1997). In this case the event study measures the abnormal returns that are caused by the announcement of an acquisition. The event study methods exploit the fact that, given rationality in the marketplace, the effects of an event will be reflected immediately in security prices. Thereby, this event study has only a short term focus and does not measure long term returns.

5.1.1 Event study methodology

To calculate the abnormal returns for the event study, first the return of the share has to be calculated. This can be done by the following formula:





1 1 ,  





t t t t i

I

R

(4)

WhereR_i_,_t is equal to the return of the share of firm i on time t.

I

_tand

I

_t_₁ are equal to the total return index of the used share on respectively time t and time t-1. The total return index used in this study is given as the Return on Investment (RI) in Datastream (Thomson financial, 2007). By using the RI the dividend payments will not cause any difference between the shares prices of the companies in the sample.

The abnormal return is calculated using the following equation:

 

it t i t i R E R AR,  ,  ,

(5)

Where for time period t means that AR_i,_t the abnormal return for time period t is, Ri,tis the actual

return, and E

 

R_i_,_t the normal return.

(25)

However, when there is uncertainty about the event date then it is necessary to take a longer time period. The advantage of a longer event window is that it takes bid revisions and competition into account. In this research the event window of 21 days and 3 days are taken into account, as can been seen in figure 1.

Figure 1: estimation period and the event window

Period

t0 = -110 τ1= -10 -1 0 1 τ2 = 10

| | | | | |

Estimation Period Event Window

To calculate the normal returns, a period of 100 days before the event window should be taken into account (the estimation period), according to Armitage (1995). The normal return is calculated using a statistical model. According to Mackinlay (1997) the normal return is the expected return without the occurrence of the event. Thereby, it is assumed that the pre-acquisition strategy of the acquiring company continues and the normal return is the return when there were no intentions for an acquisition at all.

5.1.2 The market adjusted model

In this event study the market adjusted model of Brown and Warner (1985) is used. This model is chosen instead of the Ordinary Least Squares market model for the following two reasons. Firstly, it has been shown that for short-window event studies weighting the market return by the firm’s beta does not significantly improve estimation (Brown and Warner, 1980). Secondly, since the sample of acquirers make frequent acquisitions and there is a high probability that previous takeover attempts would be included in the estimation period, it will make beta estimation less meaningful (Conn et al., 2005).

The adjusted market model assumes that the ex-ante expected returns will be equal between the companies, but not necessary equal for a given company i. In other words, the adjusted market model assumes that every company has the same systematic risk as the market.

it mt it

R







(6)

(26)

The ex-post abnormal return of a company i is equal to the difference between the return of the company and the return from the market portfolio. The abnormal returns (

AR

_it) are expressed in this model as follows:

mt it

it

R

AR





(7)

The index, used as the market return, for the EU sample is the S&P 350. The S&P Europe 350 is an equity index drawn from 17 major European markets, covering approximately 70% of the region’s market capitalization. Index constituents are leading companies from each of the 10 Sectors of the Global Industry Classification Standard (GICS) and from the 17 major European markets6.

The index, used as the market return, for the Non-EU sample is the World Datastream Market Total Return Index (Thompson Financial, 2007). This index gives information about the return of companies all over the world. That is necessary, because the Non-EU sample is worldwide.

5.1.3 Average Abnormal Returns (AARs) and Cumulative Average Abnormal Returns (CAARS)

To calculate the impact of the acquisitions, the average abnormal returns (AARs) and cumulative average abnormal returns (CAARs) has to be calculated. The abnormal returns of the acquisitions are accumulated over two dimensions, namely by time and by acquisitions. The abnormal returns which are accumulated over time are the AARs. By calculating the AARs it has to be assumed that the returns of every trading day are independent of each other (Mackinlay, 1997).

The event window is calculated from



₁ until

. This result in the following expression for the

₂ ‘Average Abnormal Returns’:



  N i it t AR N AR 1 1 (8)

The

AR

_t (= AAR) is the average abnormal return on a given day with relation to the event. Where N is the amount of events. If

AR

_tsignificant differs from zero, this can be seen as the fact that the market does react on the news of the event at that time.

(27)



  2 1 ) , ( 1 2   



AR_t CAR (9)

The

AR

_t represents a reaction on a special time, while CAAR gives a reaction for a certain time period.

5.1.4 Testing for significance (T-test)

Now it should be tested whether the AARs and the CAARs during the event window significantly differ from zero. To test for this significance, a test is used (Kothari and Warner, 1997). Before a T-test can be calculated, the returns must be T-tested for normal distribution by using Jarque Bera (JB). JB measures the difference of the skewness and kurtosis of the series. Skewness measures the extent to which a distribution is not symmetric about its mean value and Kurtosis measures how fat the tails of the distribution are (Brooks, 2002).

)

4 )

3 (

(

6

2 2

_





N

K

S

K

Jarquebera

(10)

Where, S is skewness and K is kurtosis. The JB value will be tested on the 5% significance level. If the values are significant, normal distribution can be assumed.

The AARs and CAARs were tested for normal distribution by using JB and the results show that both samples are normally distributed. By knowing this, significance can be tested by using the standard parametric T-test described by Kothari and Warner (1997). The formula of the T-Test is as follows:





)

(

AAR

_t

SD

AAR

T

 



(11)







SD

AAR

L



CAAR

T

  

)



(

(12)

Where, SD is standard deviation and L is the amount of event days.

(28)

Table 7 Critical values for a two-sided T-test

In the table below the critical values of the Non-EU sample (v = 92) and the EU sample (v = 191) are shown. The level of significance is looked up for in the table of critical value for the student T-test in Brooks (2002). v denotes the degrees of freedom.

Critical value αat Critical value αat

Significance v = 92 v = 191

10% 1,662 1,653 5% 1,986 1,972 1% 2,630 2,602

Besides the standard parametric T-test, another T-test is used to calculate significance of the difference between two independent samples. The T-test for two independent samples is conducted as follows:

 

c c t t c t

N

y

x

y

x

T

var

,







(13)

Where

x

_t and

y

_c are the averages of the CAARs for the two samples which are compared. N is the size of the samples and

var

_t the variance. The critical values in the distribution table for a one-sided t-test (Brooks, 2002) show whether the difference between two samples is significant.

5.2 Univariate analysis

In this section the univariate analysis is used to get a more clear view of the various characteristics described in paragraph 2.1. The univariate analysis takes sub samples of the two main samples to calculate the CAARs of various characteristics. The characteristics which are focused on in this research are whether the target is public or private, the type of acquisitions (Minority or Majority), the industry (based on the industry classification benchmark) in which the target is active, the corporate strategy (focus or diversification) and the time period of acquisition. The time period is divided in two periods as described earlier (from 1997 until 2002 and from 2002 until 2008). The industries which are calculated for in the univariate analysis are the industries in which it is most popular to undertake an acquisition, for both samples. The theory of why some of these characteristics could influence bidder returns are discussed in paragraph 2.1.

(29)

Some characteristics described in paragraph 2.1. are not taken into account in this univariate analysis. There was not enough data available of deal values, enterprise values and method of payment in the database of Zephyr (Bureau van Dijk, 2007) to calculate the influence of these characteristics on the acquirer returns.

5.3 Cross sectional analysis

To calculate what the influence is of different characteristics on the announcement returns, a cross sectional analysis is applied. In theory, a cross sectional analysis measures the influence of some independent variables on a dependent variable. The independent variables used for the cross sectional analysis are almost the same characteristics as used for the univariate analysis, namely whether the target is public or private, the type of acquisition (Minority or Majority), the industry (based on the industry classification benchmark) in which the target is active, the corporate strategy (focus or diversification) and the time period of the acquisition. The dependent variable used is the cumulative abnormal return CAR of each acquisition and not over all acquisitions.

The CAR is defined as the sum of the abnormal returns for each acquisition in the event window. To measure the influences of the independent variables, dummy variables has to be used. Dummy variables measure the expected difference in the CARs between two categories (0 or 1) within one group. The cross sectional analysis is conducted as follows:











)

(

)

(

)

(

)

(

)

(

)

(

)

(

)

(

)

2 (

)

(

)

(

)

(

12 11 10 9 8 7 6 5 4 3 2 1

Dpoland

Def

Dip

Dfocus

ls

Dindustria

s

Dfinancial

ervices

Dconsumers

oods

Dconsumerg

Dperiod

Dmajority

Dprivate

Deurope

CAR

(14)

CAR = 3 day cumulative abnormal return or 21 day cumulative abnormal return

Deurope = dummy variable for the acquirer (1 = European Union, 0 = Non-European Union) Dprivate = dummy variable for the legal status of the acquirer (1 = private, 0 = public) Dmajority = dummy variable for the type of acquisition (1 = majority, 0 = other)

2

Dperiod = dummy variable for the announcement period (1 = 2002 until 2008, 0 = other) oods

Dconsumerg = dummy variable for the type of target industry (1 = consumer goods, 0 = others) ervices

Dconsumers = dummy variable for the type of target industry (1 = consumer services, 0 = others)

s

Dfinancial = dummy variable for the type of target industry (1 = financials, 0 = others) ls

(30)

Dfocus = dummy variable for the corporate strategy of the bidder (1 = focus, 0 = diversification) Dip= dummy variable for investor protection (1= high protection, 0 = low protection)

Def = dummy variable for economic freedom of the target country (1= high economic freedom, 0 = low economic freedom)

Dpoland= dummy variable for the target country (1 = Poland, 0 = others)

The parameter



measures the influence of the independent variables (dummy) on the dependent (CAR) variable. The intercept is αand measures the influence on the CAR when all dummy variables would be zero.

are the error terms.

There are four industries that have to be controlled for, namely consumer goods, consumer services, financials and industrials. Furthermore, the sample is divided in two announcement periods. To avoid exact co linearity, which means that the sum of the dummy would be one, Dperiod1 is left out. This dummy variable functions as the reference group and measures the expected differences relative to the intercept coefficient.

5.3.1 Variables for target country characteristics

The formula of the cross sectional analysis is extended with a dummy for investor protection and the degree of economic freedom.

The dummy of investor protection is included to see whether the differences between the corporate governance systems of the bidder and the target have an influence on the CAR. The data is collected from the World Bank7, which gives grades from zero till ten for 175 countries8. These grades (presented in table 17) are an average of three other indexes; the disclosure index, the director liability index and the shareholder suits index. The investor protection index measures of how good the minority shareholders are protected by their country against misuse of corporate assets by directors for their personal gains. It is difficult for a country with a low investor protection to undertake a successful acquisition with an emerging country, because the target cannot take advantage of the better investor protection to be provided by the new owner of the company. The difference between the acquirer index score and the target index score is low when it is negative and is high when it is positive.

The economic freedom index is collected from the Heritage foundation9and it measures the degree of economic freedom within a country. It is based on ten specific economic freedom related criteria, namely trade policy, fiscal burden, capital flows, foreign investment, property rights and wages and

7_{These grades of the world bank can be found on the website;}_{www.doingbusiness.org}

(31)

prices. It ranks countries in terms of percentages, and higher value indicates less restrictiveness of the country’s economic environment and vice versa. Moeller and Schlingemann (2005) show that a target country’s economic freedom is positively correlated with acquirers abnormal return. That is, the more restrictive is the country economically, the less is the abnormal return associated with the transaction. Francis et al. (2008) showed the opposite and found out that there is a positive relationship between a country’s economic restrictiveness and the acquirers abnormal returns. To measure the economic freedom for the ten EU transition countries, the averages are taken of the country index scores over the period 1997 until 2008. The average of these country index scores is 60%. Therefore, a score higher than 60% is presented by high economic freedom and lower than 60% is presented by low economic freedom. This is converted into a dummy variable. The percentages are shown in figure 3.

5.3.2 Additional variables

Formula (14) is used as a cross sectional analysis for the total sample. For the EU sample and the Non-EU sample there are some extra dummies used. For both samples the dummy Poland is included, because most of the target companies are located in Poland. Furthermore, for the EU sample the most frequent acquirers are included as dummy variables, namely Austria, France and Germany. For the Non-EU sample the U.S. dummy is included. These extra dummies are included to see whether it influences the CAR when the acquisition is undertaken in a certain target country or undertaken by a certain acquirer country.

5.3.3 Testing for significance (F-test)

To test for the significance of the dummy variables a T-test is conducted. However, to measure the significance for the total model, a F-test is used. The null hypothesis for the F-test is that all coefficients are zero. The alternative hypothesis is that at least one of the coefficients is different from zero. The formula is expressed as follows:

w b statistic

MS

F



(15)

MSb is the mean square within the groups and MSwis the mean square between the groups.

5.3.4 Multicollinearity

(32)

table 18. When the correlation between two variables is higher than 0.8, multicollinearity is found (Brooks, 2002). As can been seen in table 18 there is no multicollinearity between the variables used in this research.

6. Results

This section will discuss the results of the event study, univariate analysis and the cross sectional analysis. The samples used for financial statistics are smaller compared to the samples used for the takeover characteristics in paragraph 4.3. The reason for this is that Datastream (Thompson Financials, 2007) could not collect all data. The sample sizes used for the financial statistics are 191 acquisitions for the first sample and 92 acquisitions for the second sample.

6.1 Results event study

In table 8 on the next page the results of the cumulative average abnormal returns (CAARs) for bidder shareholders are presented shortly. In table 19 in appendix 2 the average abnormal returns (AARs) and the CAARs per event window are given. The results of the CAARs are divided in four different event windows, namely [-10;10], [-1;1], [-10;-1] and [0;10].

Looking at the 3 days event window both samples, the EU and the Non-EU sample, show no significant positive CAARs around the announcement date when undertaking acquisitions in EU transition countries. For the EU sample the CAARs around the announcement date show a positive CAAR of 0,51%. The reaction for the Non-EU sample is stronger around the announcement date, the CAARs show a positive reaction of 0,83%. Looking at the difference for the 3 days event window between the EU and Non-EU sample, it can be concluded that the CAARs significantly differ from each other at a 10% significance level. Therefore, support is founded for hypothesis 1.

(33)

shows that the AARs for the EU sample are increasing after the announcement date and the AARs for the Non-EU sample are decreasing. This polynomial trend line graphically strengthens our results as just described.

Table 8 Results of CAARs for bidder shareholders for different event windows

This table below presents the CAARs for bidder shareholders of the EU sample, the Non-EU sample and the total sample. Statistical significance is calculated by using the T-test. Differences between the samples are calculated by the T-test for independent variables. Statistical significant of 10%, 5% and 1% denotes *,** or ***. SD = standard deviation and N = the amount of events

CAAR CAAR CAAR CAAR 21 days 3 days 10 days 10 days [-10;+10] [-1;+1] [-10;-1] [0;+10] EU (N=191) CAAR 0,95% 0,51% -0,02% 0,98% T-test 1,022 1,456 -0,038 1,448 Non-EU (N=92) CAAR 2,87% 0,83% 1,62% 1,24% T-test 1,840 * 1,416 1,511 1,102 Difference EU-Non EU -1,92% -0,32% -1,64% -0,26% T-test 4,823 *** 1,661 * 2,548 ** 2,539 **

To conclude, the results in table 8 show that both EU and Non-EU bidder shareholders create value around the announcement date for the 3 and 21 days event window, however insignificant. Moreover, it is found that Non-EU bidder shareholder create significant different announcement returns compared to EU bidder shareholders when undertaking acquisition in EU transition countries.

6.2 Results univariate analysis

In this section it is investigated whether factors, which are known to influence acquisition outcomes, have significant CAARs within the 3 days [-1;1] and 21 days event window [-10;10]. Therefore, sub samples of various characteristics are taken from the origin sample and it is calculated whether these characteristics show significant results. The results from the univariate analysis are shown below. Figure 4 until figure 13 in appendix 3, show the movement of the CAARs for different characteristics in line graphs.

6.2.1 Public or private

(34)

Table 9 Results of bidder CAARs by legal status of the target for the event window of 21 days and 3 days. The following table presents the bidder shareholders CAAR by legal status of the target company. The results are presented for the European sample, Non-European sample and for the total sample. Furthermore, the differences in the CAARs between the sub samples is calculated by using the T-test for independent variables. Statistical significance of 10%, 5% and 1% is denoted as *,** or ***. N is the number of events.

Event window 21 Event window 3

Difference Difference Private - Private -Private Public Public Private Public Public EU sample N= 177 N= 14 N= 177 N= 14 CAAR [-10;10] 1,17% -1,83% 3,00% 0,54% 0,13% 0,41% StDEV 0,32% 1,16% 0,32% 0,31% T-test 1,229 -0,506 3,972 *** 1,506 0,093 0,271 Non-EU sample N= 89 N= 3 N= 89 N= 3 CAAR [-10;10] 2,64% 9,68% -7,04% 0,80% 1,17% -0,37% StDEV 0,88% 3,44% 0,27% 0,32% T-test 1,676 * 1,614 2,593 ** 1,340 1,020 0,748 All N=266 N=17 N=266 N=17 CAAR [-10;10] 1,66% 0,20% 1,46% 0,63% 0,44% 0,19% StDEV 0,42% 0,72% 0,28% 0,37% T-test 2,065 ** 0,061 1,488 2,064 ** 0,359 1,077 Difference EU and Non-EU -1,47% -11,51% -0,26% -1,04% T-test 4,401 *** 3,711 *** 1,612 1,228

Results for the 21 days event window show that bidder shareholders of Non-EU companies acquiring private targets in EU transition countries create a significant positive CAAR of 2,64%. The EU sample shows a positive but insignificant reaction for acquiring private targets of 1,17% and an insignificant negative reaction of 1,83% for acquiring public targets. Remarkable is that for the total sample acquiring private targets leads to a significant CAAR of 1,66%. The difference between the Non-EU and EU sample when acquiring private targets is significant at a 1% significance level.

The results for the 3 days event window show only significant results for the total sample acquiring private targets. Furthermore, no significant differences are found between the EU and Non-EU sample.

(35)

6.2.2 Types of acquisition

In this section a sub sample is conducted of majority acquisitions (acquiring > 50%) and minority acquisitions (acquiring < 50%). Table 10 shows the results for the CAARs per type of acquisition and figure 4 and 5 show the movement of the CAARs in line graphs.

Table 10 Results of bidder CAARs by type of acquisition for the event window of 21 days and 3 days. The following table presents the bidder shareholders CAARs by the type of acquisition. The results are presented for the European sample, Non-European sample and for the total sample. Furthermore, the differences in the CAARs between the sub samples is calculated by using the T-test for independent variables. Majority acquisitions is acquiring >50% and a minority acquisition < 50%. Statistical significance of 10%, 5% and 1% is denoted as *,** or ***. N is the number of events.

Difference Difference Majority - Majority -Majority Minority Minority Majority Minority Minority EU sample N= 170 N= 21 N= 170 N= 21 CAAR [-10;10] 0,85% -1,02% 1,87% 0,42% -0,02% 0,44% StDEV 0,29% 0,93% 0,28% 0,38% T-test 0,922 -0,311 4,689 *** 1,197 0,014 0,096 Non-EU sample N= 81 N= 11 N= 81 N= 11 CAAR [-10;10] 3,45% 3,39% 0,06% 0,93% -0,02% 0,95% StDEV 1,13% 1,29% 0,30% 0,49% T-test 2,097 ** 0,853 1,372 1,496 -0,014 2,423 ** All N=251 N=32 N=251 N=32 CAAR [-10;10] 1,68% 0,50% 1,18% 0,58% -0,02% 0,60% StDEV 0,44% 0,47% 0,23% 0,31% T-test 2,118 ** 0,202 3,792 *** 1,942 * -0,020 1,603 Difference EU -Non-EU -2,60% -4,41% -0,51% 0,00% T-test 4,734 *** 7,536 *** 2,248 ** 1,526

Looking at the 21 days event window the results in table 10 show that acquiring a majority control for the Non-EU sample leads to a significant positive CAAR of 3,45%. Furthermore, regarding the EU sample acquiring a majority control results in a positive CAAR of 0,85%, however insignificant. Looking at the total sample, it can been seen that acquiring a majority control leads to a significant CAAR of 1,68%. Moreover, the differences between the EU and Non-EU sample are significant. It can be concluded that Non-EU bidders shareholders create significant more return when acquiring both a majority or a minority stake compared to EU bidder shareholders when undertaking acquisitions in EU transition countries.

(36)

Chari et al. (2006) found out that acquiring a majority control in emerging markets created significant more value instead of acquiring a minority control. Therefore, it can be said that the results of Chari et al. (2006) are in line with the results in this research. Arguments of Chari et al. (2006) why a majority control leads to positive bidder shareholder returns are explained in paragraph 2.1.

6.2.3 Time period

The time period investigated is divided into two sub periods. The first period is from 1 January 1997 until 31 December 2001 and the second period is from 1 January 2002 until 31 December 2007. By measuring the returns of the EU sample as well for the Non-EU sample before and after 2002, something can be said about the influence of worldwide and European financial integration on bidder shareholder returns for both samples. Furthermore, by measuring the differences of the CAARs between the EU sample and Non-EU sample after 2002, it can be said if European financial market integration causes different shareholder returns for European acquirers compared to Non-EU acquirers (hypothesis 2). Table 11 shows the results for the CAARs by time period and figure 8 and 9 in appendix 3 show the movement of the CAARs in line graphs.

Table 11 Results of bidder CAARs by time period for the event window of 21 days and 3 days.

The following table presents the bidder shareholders CAAR by time period. The results are presented for the European sample, Non-European sample and for the total sample. Furthermore, the differences in the CAARs between the subsamples is calculated by using the T-test for independent variables. Time period is split in two period to control European integration. The first Period is from 1 January 1997 until 31 December 2001. The second period is from 1 January 2002 until 31 December 2008. Statistical significance of 10%, 5% and 1% is denoted as *,** or ***. N is the number of events.

(37)

For the event window of 21 days the results in table 11 show that EU acquisitions created more value in period 2 compared with period 1. In period 2 the CAARs for bidder shareholder were significant positive with 2,57% compared to -0,47% for the first period. Furthermore, the difference for both the EU and Non-EU sample between period 1 and period 2 is significant. These results indicate that higher financial integration of EU transition countries do not result in lower bidder shareholder return when undertaking acquisitions in EU transition countries. Looking at the results of the Non-EU sample the opposite is true. Bidder shareholders from the Non-EU sample created more value in the first period compared to the second period. In the first period the CAAR is 3,69% and in the second period the CAAR is positive with 2,36%, however insignificant. Moreover, the difference between period 1 and 2 is significant. The results show that worldwide financial market integration leads to lower bidder shareholder return for Non-EU acquirers when undertaking acquisitions in EU transition countries.

For the event window of three days it is found that EU bidder shareholders create significantly more return in period 2 compared to period 1 as well. Reasons for that are summed up by Bjorvatn (2004) in paragraph 2.3. Furthermore, Non-EU bidder shareholders create less return in period 2 compared to period 1. This result is in line with the expectation that worldwide financial integration of EU transition countries leads to lower bidder shareholder returns the Non-EU sample. The differences between period 1 and period 2 are for both samples not significant.

It can be concluded that EU bidder shareholder create significant more return in period 2 compared to period 1 and Non-EU bidder shareholders less. This means that worldwide and European financial market integration does not negatively affect EU bidder shareholders. However, it does for Non-EU bidder shareholder. The significant difference between the EU and Non-EU sample is not found in period 2 and therefore does not support hypothesis 2.

6.2.4 Focus or diversification

Impact of financial market integration on cross border acquisitions in European Union transition countries