DO EUROPEAN SHAREHOLDERS BELIEVE THAT INVESTING IN CHINA IS MORE VALUABLE?

(1)

THAT INVESTING IN CHINA IS MORE

VALUABLE?

An event study about the differences in the announcement effect

between acquisitions of Chinese and American companies.

Hélène F. M. Wiegerinck

University of Groningen

Faculty of Economics, Finance department

(2)

Do European Shareholders believe that investing in China is more

valuable?

Abstract

The cross-border acquisition market has changed a lot the past decades, as Europe and China have become more important players. In this event study I examine 114 deals with China and 116 deals with the USA over the period of 1999 till the first half of 2006. I hypothesize that the bidder returns for takeovers with Chinese targets are higher than for takeovers with American targets. This is not the case; for both target countries no significant abnormal bidder returns are found around the time of the announcement of the acquisition. I do find some other characteristics that influence the bidder returns. When the target is a public company or when the bidder is a manufacturer this has a negative effect on the bidder returns. For bidders from Continental Europe there is a significant difference in the returns between deals with a negative CAAR from China and deals with a positive CAAR from the USA.

JEL classification : G14, G34

Keywords : cross-border mergers and acquisitions, bidder returns, event study

(3)

1. Introduction

The integration of the world market during the past decades had many consequences for the cross-border acquisitions market. The fast developments in the product and capital market created new markets in emerging countries and made it for many companies important to expand globally. This caused a movement in the geographical distribution of mergers and acquisitions activity in the world. Besides the United States and the United Kingdom other countries and continents also became involved in the cross-border acquisition market.

During the last decade cross-border mergers and acquisitions (M&As) have become the most important way of inward foreign direct investment for developing countries (Chari, Ouimet and Tesar, 2004). This tendency towards more M&A activity can also been seen in China, which became an important target for companies looking to make a successful acquisition. China’s entering the world trade organization (WTO) in 2001 caused many of the entry barriers to disappear. For foreign companies this made it easier to take advantages of the opportunities that are available in China. In 2002 China received the largest amount on foreign direct investments of all countries in the world; a consequence of the remarkable economic growth and excellent economic prospects (www.evd.nl). M&A activity in China has been very important for the restructuring of the inefficient state-owned enterprises (SOEs) and is needed for the growth of many firms. The Asian financial crisis weakened the SOEs, who dealt with insolvencies and the inability to cover operating expenses. This made the Chinese government more open to foreign investors recognizing the skills and the capital they could bring. The government is still experimenting with different models for laws on mergers and acquisitions, and the existing laws are likely to change and are also starting to follow the international practice (Tan Lay Hong, 2003). Laws on trade and investment became more transparent and were published in English (www.evd.nl). The entering of China to the World Trade Organization also forced China to open the borders and raise the limits on the maximum foreign investment permitted (Norton and Chao 2001). China also agreed to eliminate prohibitions on foreign distribution activities. This generates investments possibilities also in distribution and retail activities (Norton and Chao 2001).

(5)

privatization of companies and the globalization process all influenced continental European firms to become active participants (Martynova and Renneboog, 2006 ; Gugler et. al. 2003). In this thesis I examine European companies that acquire Chinese or American companies. As the Chinese economy maybe gives more opportunities than entering in the more mature US market, I hypothesized that the returns to bidders will be larger with Chinese acquisition than with American acquisitions. I examine the announcement effect on European bidder returns and its influencing characteristics in order to see whether the European investor indeed values acquisitions in China higher than acquisitions in the USA.

Prior research showed that acquisitions of emerging market targets can be very successful (Chari, Ouimet and Tesar, 2004); this also supported the formulation of my hypothesis of larger bidder returns for Chinese acquisitions. This thesis is a contribution to the existing research on cross-border acquisitions because of its focus on China and the USA as targets. It also focuses only on acquirers from Europe, while prior research frequently uses the USA as headquartering the acquiring company. The comparison of European-Chinese and European-USA acquisitions is new and can show fresh insights on how investors value these two groups of acquisitions. This study is also an addition to the rather scarce literature on cross-border acquisitions with targets from developing countries.

However, the results of this event study show that for both acquisitions in China and the USA there are no significant abnormal returns for the bidders. Acquisitions with American targets provide larger positive returns to the bidder, than acquisitions with targets from China, thought the difference is not significant either. I do find other characteristics, like the legal status of the target and whether the bidder is a manufacturer or not, to have a significant influence on stock returns of the bidder. For bidders from Continental Europe there is a difference in their returns between undertaking an acquisition in China and the USA.

(6)

2. Background

In this section the literature on cross-border acquisitions will be examined. I discuss the value creation aspect of cross-border acquisitions, the characteristics influencing the announcement effect, cross-border takeovers involving a developing country target and M&A activity with China.

The empirical literature shows that takeovers do create value, but that there is a big difference between the stock returns for bidders and the stock returns for targets. Campo and Hernando (2004) present a literature overview which shows that for the bidders the announcement effect is not significantly different from zero or there is no agreement on whether there is a positive or negative share price return. On the other hand targets often do have large abnormal returns after the announcement of a takeover. Most studies on the value creation for the new combined firm find that there is a significant positive announcement effect (Campo and Hernando, 2004).

2.1. Announcement effect and takeover characteristics

The short-term wealth effect of a takeover is usually measured using the event study methodology. The announcement effect is the change measured in the shareholder returns in the period around the announcement date of the takeover; they are the so-called abnormal returns. The announcement effect has been extensively studied in the literature, often also investigating the effect of different characteristics that influence the abnormal returns. Some characteristics in the literature are widely accepted to have an influence on the abnormal return; these characteristics will now be shortly discussed.

2.1.1. Method of payment

(7)

not exist for cross-border acquisitions. On the other hand targets may not accept foreign equity. In that case there is no positive signaling effect for the use of cash either (Gaughan, 2002).

2.1.2. Focus or diversification strategy

It is assumed that when a firm undertakes a diversification strategy it generally destroys value (Martynova and Renneboog, 2005). Moreover companies should not try to diversify, because individual investors are able to create their own diversified portfolio. But corporate diversification can be motivated by the benefits of creating or expanding the internal capital market. Especially in the case of severe external capital market imperfections (which are the case for emerging markets) diversification could be beneficial (Khana and Palepu, 1997).

2.1.3. Relative size

There is evidence that the relative size of the target influences the potential to create value with a takeover. Moeller, Schlingemann and Stulz (2004a) show that the announcement returns for smaller bidding firms are two percentage points higher than for large bidding firms. They examine several explanations; firstly, it could be that the difference in public and private targets causes the difference. Secondly, small firms are more likely to pay with cash, which can cause a difference. Thirdly, small and big firms may differ in other characteristics like leverage and Tobin’s q. Finally managers of large firms are more sensitive to hubris and the incentives of managers of small firms are better aligned with those of the shareholders. A large firm may be large because its equity is highly valued: so a large firm is more likely to be overvalued. Moeller, Schlingemann and Stulz (2004a) only find evidence that the managerial hubris plays more of a role in the decisions of large firms. The results are robust for the other examined deal and firm characteristics.

2.1.4. Hostile or Friendly bid

(8)

2.1.5. Public or Private target

Conn et al. (2005) suggest that there are several reasons why private targets may cause higher bidder returns than public targets. Private bids are less visible for the general public so it is possible to end negotiations without the loss of face and bidders will not pay too much. Secondly, when paid out in shares, targets’ management can become significant blockholders in the post-merger firms. These generally manage in the best interest of the firm and have a monitoring role. Finally, the private and illiquid nature of this takeover market makes the bidding less competitive and makes it possible to disclose sensitive information to the bidder on the firm value. This last advantage is not possible for public takeovers because regulation requires that equal information is available to all shareholders.

2.1.6. Target country characteristics

For cross-border acquisitions the institutional character of the target country also influences the gains of the bidder. Examples of these characteristics are trade policy, government intervention, capital restrictions, market for corporate control and cost of capital. Moeller and Schlingemann, (2004) assume that when the target has a less restrictive institutional environment, this will reduce agency problems and asymmetric information and higher bidder gains can be expected. Moeller and Schlingemann show that with exception of the UK as target country there is a negative relationship between the bidder returns and the target country’s economic restrictiveness. China is becoming a less restrictive country, because of the changes in the trade policy and the elimination of trade barriers.

2.2. Cross border M&A activity and value creation

(9)

Moeller and Schlingemann (2004) examine if there are differences between cross-border and domestic acquisitions. They assume that there are several reasons why cross-border acquisitions may be more valuable than domestic acquisitions. Firstly, due to integration of international capital and product markets, relative costs and benefits of cross-border acquisitions may have changed. Secondly, overseas acquisitions may enable companies to profit from differences in tax systems (Seveas and Zenner, 1994). Thirdly, imperfect capital markets may make it possible to exploit exchange rate movements (Froot and Stein, 1991). Finally, cross-border M&A activity can be very valuable, when it causes an expansion of investment opportunities that generate possibilities for synergistic gains (Moeller and Schlingemann, 2004).

Moeller and Schlingemann (2004) also give some reasons why cross-border takeovers might not be successful; increased competition in the market for corporate control may have eliminated synergistic gains. When the acquirer is not fully informed it may become victim of the ‘winners curse’ by overpaying for the target (Eckbo, 1992) or pay an unwarranted premium for a foreign company just because of the desire to enter foreign markets (Shaked, Michel and McClain, 1991). Cross- border takeovers could also cause an increase in managerialism and agency problems (Dennis, Diane and Keven, 2002) and there are differences between domestic and cross-border acquisitions in terms of asymmetric information, culture, politics and the economy of a target country.

Another important consideration that has to be taken into account is that portfolio diversification for the individual investor has become less costly; this may imply that the benefits of an international firm may no longer outweigh the cost of cross-border acquisitions (Moeller and Schlingemann, 2004).

2.3. Cross-border takeovers with developed markets as target countries

(10)

Their results show that the relative transactions size for the cross-border transactions is smaller while the bidder is larger compared to domestics transactions. They also find evidence that the abnormal returns are smaller for the bidder when the acquisition strategy is global diversification and becomes even smaller when there is also industrial diversification. Overall they conclude that cross-border takeovers have statistically significant lower bidder gains than domestic takeovers. Martynova and Renneboog (2006) examine in their paper M&A activity in 30 European countries in the period 1993-2001. They also study both domestic and cross-border acquisitions, but only within Europe. The paper indicates that the increase in intra-European deals is a consequence of the development of the single European market and the introduction of the euro. Martynova and Renneboog (2006) find that the shareholder wealth effect strongly depends on several characteristics of the takeover. For the bidder they find a statistically significant positive announcement effect of 0.5 %. The type of takeover bid impacts on the announcement effect for the bidder, if the bidding is hostile the abnormal returns are negative (-0.4%), but when the bid is friendly it triggers an abnormal return of 0.8 %. The payment method and the takeover strategy also influence the announcement effect. All-cash offers lead to larger abnormal returns, for the bidder, than all-equity offers and the focus strategy generates significantly higher returns. Cross-border takeovers generate a lower announcement effect than domestic takeovers (0.4% vs. 0.6% respectively) A last finding is that acquisitions that took place in later phase of the wave are less successful, than in the beginning of the wave. This same effect is also found for M&A activity in the US (Moeller, Schlingemann and Stulz 2005). The results from Martynova and Renneboog differ from earlier findings on cross-border acquisitions from the UK, where there were significant negative returns for a bidder acquiring in continental Europe (Aw and Chatterjee, 2004). Contradicting with Moeller and Schlingemann (2005) results from Conn et al. (2005) show a zero announcement returns for public cross-border acquisitions, while domestic acquisitions generate a negative announcement effect. Overall we can say that there is little agreement in the literature on the announcement effect of takeovers in Europe.

2.4. Cross-border takeovers with emerging markets as target countries

(11)

capital caused by the acquisition of a developing country firm by a developed country firm can make projects, that otherwise would have been rejected in developing countries, possible, which may be value creating.

Secondly, bidders may have greater bargaining power, because in emerging markets there are presumably less bidders. This bargaining power increases in periods during which the target is in financial distress. Mody and Negishi (2002) find in their paper that in the restructuring process after the Asian crisis cross-border M&A activity occurred primarily in the most distressed sectors. Asian governments undertook several steps to encourage these mergers and acquisitions. However there is not enough evidence to identify so-called ‘emergency sales’ and that this M&A activity contributed immediately to the restructuring of the economies in crisis.

Thirdly, an obstacle in the value creation process may be that the stock prices in emerging countries are noisy and therefore not a good measure for the true firm value. This creates additional information asymmetry of which the bidder can take advantage if it gets the chance to estimate the fundamental firm value, or when the target is uncertain about its own value.

Fourthly, Chari, Quimet and Tesar (2004) suggests that the acquisition of control may be important in countries with poor protection and enforcement of the minority shareholders rights (La Porta, et. al. 1999) and may be value enhancing for the bidder and the target.

Chari, Quimet and Tesar (2004) show that the increase in acquisitions of companies from emerging markets in the late 1990s increased value for both bidder and target shareholders. The average monthly returns for the acquirer increased from 1.65 % to 3.05% and are robust to the inclusion of controls for industrial diversification, country, time and method of payment. The most important condition was that the acquirer gained majority control over the target firm in the emerging market.

Overall Chari, Quimet and Tesar (2004) show that the transfer of corporate ownership from an emerging country to a developed country leads to substantial gains, for both the bidder and the target.

(12)

2.5. Cross-border takeovers with China

In 1979 the Chinese government started its ‘Open Door Policy’. This policy caused a gradual increase in foreign direct investment. Foreign direct investment in China can be divided in three waves; first in the 1980s the foreign direct investment took the form of joint ventures, later in the 1990s many greenfield wholly foreign-owned enterprises (WFOEs) were created. At this moment there is a third wave of foreign direct investment that is based on cross-border mergers and acquisitions (Peng, 2006). As far as we know there has not been done much research on developed countries undertaking acquisitions in China. This may have to do with the fact that China only recently opened the gates for this kind of direct investment. Cheng (1998) finds a positive announcement effect of US-Chinese joint ventures, but cannot link this value enhancement to any of the characteristics of the US firms.

(13)

3. Data and Methodology

In this section the construction and some characteristics of the dataset will be described and the methodology of the event study, the univariate analysis and the cross-sectional analysis will be discussed.

3.1. Data selection

The dataset consists of companies from the EU-15 countries3_{that announced an acquisition with}

either a Chinese company or an American company in the period between 1999 and the first half of 2006. The Zephyr database from Bureau van Dijk is used to find the announcements dates and some other characteristics of the takeovers. The Zephyr database includes information on both public and private takeovers. However for this study I will only use public bidders. This means that the bidder must be listed on, at least one, European stock exchange. I find in total 456 European-Chinese announcements of M&A in China, when private bidders are excluded 279 deals remain. The European-American acquisitions are used as a control sample, to see whether a Chinese takeover causes a larger announcement effect than this group. European companies announced 3.579 takeovers with the USA, of which 2.161 announcements were with public bidders.

To make our dataset suitable for the event study, some adjustments will have to be made. As mentioned before, takeovers of a private bidder will be removed, because they do not have stock returns with which the announcement effect is measured. Secondly, takeovers where the rumor date differs from the announcement date are left out because it is not clear at which moment the reaction of the shareholder is reflected in the stock returns. Thirdly, I exclude deals of bidders that already did an acquisition in China during the year before the latest announcement, because it could lead to biases in the calculation of the normal and abnormal returns. If there are multiple announcements over a short amount of time it is not possible to distinguish the effects of those shocks from each other. Finally, bidders that do not have an ISIN code from Zephyr are removed; this ISIN code is needed to get the stock return data from Datastream. The dataset is divided in minority acquisitions, joint ventures and majority acquisitions in order to study if the type of acquisition also has an effect on the abnormal returns. A majority acquisition is defined as occurring when the bidder owns less than 50% of the targets voting shares before the takeover

3_{European Union 15 - Austria, Belgium, Denmark, Finland, France, Germany, Greece, Italy, Luxemburg, Netherlands, Portugal,}

(14)

and increases its ownership to at least 50% as a result of the takeover (Conn, 2005). Joint ventures arise when two or more companies create a new company. The new company is coded as the target and the investors are coded as the bidders. In the case of a joint venture I will follow Zephyr, on who is the acquirer and who is the target.

3.2. Sample description

In tables 4 and 5 of appendix 1 some of the characteristics of our data-set are presented. This is the data-set of all public bidders before various eliminations. Both tables show whether the differences between the observed characteristics of both sample groups are significant or not. Table 4 shows a breakdown of the dataset in European Union countries. The UK is the most frequent acquirer country for both groups; for China in 27 % of the acquisitions and for the USA in 42%. Other frequent acquirers are France, Germany and The Netherlands. Belgium, France and Italy acquire significantly more in China and Ireland and the UK more in the USA .Table 4 also present the semi-annual distribution of the acquisition. The number of deals between Europe and the China is growing, especially since 2001, when China joined the WTO. In the years 2002 through 2004 there is a downfall in the number of deals.

Table 5 also shows various characteristics: for almost 50% of all the deals the method of payment is not known. The deals for which the method of payment is known are almost all paid in cash, both in China and the USA. It is remarkable to see that other payment methods like debt and shares are used more often in the USA than in China. (6.1% debt and 16.1% shares vs. 0.7% debt and 7.0% shares, respectively). The type of deals also significantly differ from each other: in China there are more minority acquisitions and joint ventures. Therefore the share of majority acquisitions is smaller: 54.5 % in China versus 83.9 % in the USA. The percentage of deals that were actually completed is much larger for the USA, 76.8 % while in China only 48.0% reaches this status. In China more deals are pending for several reasons (regulating or shareholder approval) or are still in the announcement phase.

Table 5 also shows a breakdown of the acquisitions by industry is shown.4_{As can been seen}

most of the deals between Europe and China find place in the manufacturing sector, while in the

4

The following industry breakdown is used for main SIC industry classifications: Two-digit SIC code (01-09) Agriculture, Forestry, and Fishing, (10-14) Mining, (15-17) Construction, (20-39) Manufacturing, (40-49) Transportation, Communication, Electric, Gas, and Sanitary Services, (50-51) Wholesale Trade, (52-59) Retail Trade, (60-67) Finance, Insurance, and Real Estate, (70-89) Services, (91-97) Public Administration. (Moeller and Schlingeman, 2004)

(15)

USA most of the deals find place in both the manufacturing and service sector. More often the acquirers in China are in the finance, insurance and real estate industry. The total number of deals in table 3 is higher than the previous mentioned numbers of 279 deals for China and 2,161 for the USA. The reason is that many companies are in several classes of the SIC classification system and are therefore double counted in this table.

3.3. Matching the Chinese takeovers with the USA takeovers

Looking at our sample characteristics in table 4 and table 5 one can see that the sample of border bidders to China has specific characteristics when compared with the sample of cross-border bidders to the USA. Our sample of Chinese takeovers is smaller than our sample of USA takeovers. Therefore I select European public bidders who perform a takeover with an American company that have the same characteristics as the European companies that perform a takeover with a Chinese company. Every Chinese takeover is linked to an American takeover with similar characteristics. The selection of the matched companies will be based on three characteristics of the bidder:

• The industry (the same two digit SIC code)

• Type of acquisition (majority or minority acquisition or joint venture) • Time of the acquisition (within one year from each other)

The takeovers are matched on the type of acquisition to see whether the transfer of corporate ownership has the same influence on the bidder returns for deals with China as with the USA. I also match the deals for the time of the acquisition to be certain that the matched deals were announced under the same macro-economic circumstances.

(16)

differences between the sample of Chinese takeovers and USA takeovers. Finland acquires significantly more in China, while Sweden acquires more in the USA. Moreover there are significant more deals that are completed in the USA, while in China more bidders are pending for regulation approval.

Table 1: takeover characteristics of final sample

Statistical significance at 1%, 5% and 10% level is denoted with a, b and c.

China USA

No. % No. % Difference p-value

(17)

Table 2: takeover characteristics of final sample

Statistical significance at 1%, 5% and 10% level is denoted with a, b and c.

China USA

No. % No. % Difference P-value

Method of payment

Cash 42 36,84% 34 29,31% 7,53% 0,226

Converted Debt 0 0,00% 0 0,00% 0,00% n.a.

Debt 1 0,88% 1 0,86% 0,02% 0,992

Deferred payment 1 0,88% 1 0,86% 0,02% 0,992

Earn out 1 0,88% 0 0,00% 0,88% 0,316

Loan notes 0 0,00% 0 0,00% 0,00% n.a.

Shares 6 5,26% 6 5,17% 0,09% 0,976 Other 1 0,88% 4 3,45% -2,57% 0,180 Unknown method of payment _{62 54,39% 70 60,34%}_-5,96% Total 114 100,00% 116 100,00% deal type Acquisitions 75 65,79% 81 69,83% -4,04% 0,517 Joint Ventures 9 7,89% 7 6,03% 1,86% 0,582 Minority Shares 30 26,32% 28 24,14% 2,18% 0,704 IPO 0 0,00% 0 0,00% 0,00% n.a. Merger 0 0,00% 0 0,00% 0,00% n.a.

Institutional buy-out 0 0,00% 0 0,00% 0,00% n.a.

Total 114 100,00% 116 100,00% 0,00% n.a.

Deal status 0,00% 0,00% 0,00% n.a.

Announced 46 40,35% 34 29,31% 11,04% 0,784 Completed 55 48,25% 79 68,10% -19,86% 0,002 a Pending (regulating approval) _{10 8,77% 1 0,86% 7,91% 0,005 c} Pending (shareholder approval) 0 0,00% 0 0,00% 0,00% n.a.

Pending (reason not

specified) _{2 1,75% 1 0,86%}_{0,89% 0,555} Postponed _{1 0,88% 1 0,86%}_{0,02% 0,992} Withdrawn _{1 0,88% 1 0,86%}_{0,02% 0,992} Total _{114 100,00% 116 100,00%} Industry Agriculture 0 0,00% 0 0,00% 0,00% n.a. Mining 1 0,88% 1 0,86% 0,02% 0,992 Construction 0 0,00% 0 0,00% 0,00% n.a. Manufacturing 61 53,51% 64 55,17% -1,66% 0,803 Transportation 8 7,02% 9 7,76% -0,74% 0,834 Wholesale Trade 2 1,75% 2 1,72% 0,03% 0,992 Retail Trade 6 5,26% 6 5,17% 0,09% 0,976 Financial 13 11,40% 13 11,21% 0,20% 0,968 Service 23 20,18% 21 18,10% 2,07% 0,697

Public Administration 0 0,00% 0 0,00% 0,00% n.a.

unknown 0 0,00% 0 0,00% 0,00% n.a.

(18)

3.4. Event study

The event study estimates the abnormal returns that are caused by the announcement of an acquisition. These abnormal returns can be estimated, because in an efficient capital market, the stock prices react immediately on the news of an acquisition and include the expected change in value of the combined company. This event study thus only has a short-term focus and does not examine any long-term operational success.

The event day is either the announcement day itself or the first following trading day in the case the announcement day itself is not a trading day. This day is known as day 0, around which the event window is placed. It is customary to include the days surrounding the announcement day to examine if they also, besides the announcement day itself, have abnormal returns. It could be that prior to the announcement day the market acquired information about the acquisition and after the announcement day it captured price effects that occurred after the stock market closed on the announcement day (MacKinlay, 1997). The event window starts 20 days before the event day and closes 20 days after it.

To estimate the abnormal returns caused by the announcement, the normal return or benchmark return should be determined. To find the normal return an estimation window is used. The estimation window is the period before the event window in which the market model parameters are estimated. These parameters are then used in the event window, to calculate what would have been the normal return is this period. The estimation window is 200 trading days, from day -220 to day -20 before the announcement day. The event window directly follows the estimations window.

I assume, like most of the event-studies, that the pre-merger strategy of the bidder (and the target company) continues, so the normal return would have been the return if the event never took place. This assumption allows us to use the market model, which assumes a stable linear relation between the market return and the security return. The market model normal returns are measured in the estimation window and given by;

(19)

Rjt is the rate of return of security j at time t. Rmt is the actual market return on day t. αj is the

intercept term for the regression equation of security j, βj is the regression coefficient that

captures the risk of the security with respect to the actual market. The index used as a proxy for the market is the S&P Europe 350. 5_{This index can be obtained from DataStream. Ε}

jt is the error

term for security j at time t.

I do not always use an Ordinary Least Squares (OLS) regression to calculate the normal returns, because using OLS implies that the data meets some specific requirements. In the context of financial time series this is sometimes unlikely. It could be that the variance of error terms may change over time. In these cases it is better to use a model that does not assume that the variance is constant, but one that assumes the error terms are heteroskedastic; therefore the generalized autoregressive conditional heteroskedasticity (GARCH) model is used. (Brooks, 2002, p. 452). This model uses the same regression function as mentioned above, only the error terms are treated differently; jt jt jt

v

σ

ε

=

(2) For every event I test whether the data is heteroskedastic or not. This is done with the ARCH-LM test. If an event has error terms that are heteroskedastic the GARCH (1.1) model is used, otherwise I use the OLS regression to estimate the α and β. The GARCH model was used for 42 European-Chinese events and for 34 European-American events.

The α and the β parameters are used to calculate the abnormal returns. The daily abnormal returns are computed by subtracting the predicted normal return from the actual return for each day in the event window;

)

(

j j mt jt jt

R

AR

=

−

α

+

β

(3)

The abnormal returns are examined over five event windows within the original event window; ( 41 days (-20 to +20), 21 days (-10 to +10), 11 days (-5 to +5), 5 days (-2 to +2) and 3 days (-1 to +1)). For the different intervals the abnormal returns are accumulated over time and over the

5

S&P Europe 350 is the master index of the S&P European index series. It provides broad regional exposure to 17 leading markets in Europe — the 12 members of the Euro zone plus Denmark, Norway, Sweden, Switzerland and theUnited Kingdom.

(20)

securities. These cumulative average abnormal returns (CAARs) are calculated for N securities as follows;

∑

∑∑

= = = = = = N j N j t t jt jt N AR CAR N CAAR 1 1 2 1 / 1 / 1 τ τ τ (4)

The τ indicates the event window (t1 and t2), for which t1≥ -20 and t2 ≤ +20. The sign and size of the CAARs show whether the investors reaction towards the announcements is positive or negative.

The cumulative abnormal returns will be tested on their statistical significance to see if the acquisition announcement really had an influence on the stock returns. The null hypothesis indicates that the CAAR, during a given event window, is equal to zero. I will use the standard parametric t-test, discussed by Brown and Warner (1985) and Martynova and Renneboog (2006), to test this hypothesis. The statistic follows a Student-t distribution as is calculated as follows;

)

(

ˆ

τ τ

σ

CAAR

t

p

=

(5)

Where the standard deviation of the CAARs (

σ

ˆ

CAAR) can be calculated as followed;

∑∑

= = =

=

N j t t j

N

CAAR

1 2 2 2 1

ˆ

1 )

(

ˆ

τ τ τ

σ

(6)

And

σ

ˆ

j is an estimator of the standard deviation of the ARs for security j calculated over the

estimation window as follows:

∑

= = − − − = T j oj T mt j j t i j j R R

L

1 2 , ) ( 2 1

ˆ

α

β

σ

(7)

Where Lj is the number of observations in the estimation window for security i.

3.5. Univariate analysis

(21)

legal status of the target (private or public), the method of payment and the corporate strategy (focus or diversification). The theory why some of these characteristics could influence the bidder returns in explained in section 2.2 In this section I also mention relative size and hostile or friendly bid as characteristics that is influences the bidder returns. Because of a lack of available data on deal values, enterprise values and type of deals in my dataset I will not investigate the influence of these characteristics on the bidder returns. I do look at some other characteristics, like form of the acquisition, bid completion status and industry, because I think they might have an influence on the bidder returns in this sample.

3.6. Cross-sectional analysis

Next I will perform a cross-sectional analysis. In the cross-sectional analysis I will look at the influence of gaining a majority control over the target, the influence of the method of payment, the period in which the takeover was announced, the success or failure of the deal (completed, announced or pending), the legal status of the target (private or public).

For the cross-sectional regression the sample is divided in three different time periods. In the first period, from 1999 till 2001, there were not many acquisitions with Chinese targets, this because of the trading restrictions. In the second period (2002 till 2003), the number of acquisitions with Chinese targets is growing as a consequence of China’s entry to the WTO and number of deals with American decreases after the internet bubble in 2001. In the third period (2004 till the first half of 2006) the number of acquisitions with Chinese companies is still rapidly growing, while the number of takeover with the USA does not reaches the same level as before the internet bubble.

The CAR is then used to run the following regression model;

ε

β

δ

α

+

=

)

(

)

(

)

(

)

(

)

(

)

3 (

)

2 (

)

1 (

)

(

)

arg

(

)

(

)

(

)

(

)

01 (

13 12 11 10 9 8 7 6 5 4 3 2 1

IP

mtheUK

Dbidderfro

Dservice

ring

Dmanufactu

Dfinance

Dperiod

Dcash

etispublic

Dt

Dfocus

Dcomplete

Dmajority

D

CAR

(8)

(22)

Dcomplete = dummy variable for the completion status of the deal (completed = 1,

not completed =0 )

Dfocus = dummy variable for the corporate strategy of the bidder (focus =1, diversification=0)

The dummy variable for the corporate strategy is 1 when the four digit SIC code for the bidder is the same as the target.

Dtargetispublic = dummy variable for the legal status of the target (public = 1, private= 0) Dcash = dummy variable for the method of payment (cash = 1, other = 0)

Dperiod1, Dperiod2, Dperiod3 = dummy variables for three different announcement periods.

(Period 1; 1=1999 till 2001, 0= other)(Period 2; 1=2002 or 2003, 0= other) (Period 3; 1=2004 till 2006, 0 = other).

Dmanufacturing, Dfinance, Dservice = dummy variables for the industry (Manufacturing = 1

other = 0) (Finance, Insurance, and Real Estate, = 1, other = 0) (Services =1, other =0)

DbidderfromtheUK = dummy variable for the home country of the bidder (UK = 1, Continental

Europe= 0)

IP = is the only variable that is not a dummy variable. The score on the investor protection index

of the bidder minus the score on the investor protection index of the target.

The parameters of the dummy variables measure the expected difference in the cumulative abnormal returns between the two groups (1 or 0) within one category. For instance the parameter δ6 measures the expected difference in CAR due to a payment with Cash.

Some of the qualitative factors that are used in this cross-sectional analysis have more than two categories. For each category a separate binary dummy variable is created. In the case of our dummy variables for the year of the announcement the dummy categories are exhaustive, which means that the sum of the dummy would be one, when dummies are for all three periods. This creates a model in which exact collinearity exists (Hill, Griffith and Judge, 2001 p206). To avoid this, one of the dummy variables (Dperiod1) is omitted, this dummy variable is defined as the reference group. The coefficients of the other dummy variables for the year of the announcement measure the expected differences relative to this intercept coefficient.

(23)

3.6.1. Investor Protection variable

To see whether the differences between the corporate governance systems of the bidder and the target have an influence on the abnormal returns around a takeover announcement an investor protection variable is added. The World Bank gives grades (from zero till ten) to 175 countries which measure of how good this country protects the minority shareholders against misuse of corporate assets by directors for their personal gain (www.doningbusiness.org). The investor protection index is composed out of three other indexes; the disclosure index, the director liability index and the shareholder suits index, for every bidder and target country in our sample the scores are shown in table 6. China gets a 5 in this investor protection index while the USA scored an 8.3. It is remarkable to see that some of the continental European countries (The Netherlands, Austria and Greece) score lower than China in this index and that for other continental European countries the differences with China are not very big. It could be that for bidders from a country with a low score on investor protection it is more difficult to undertake a successful acquisition with an emerging country, because in that case the target can not take advantage of the better investor protection to be provided by the new owner of the company. The investor protection variable in this cross-sectional regression is the score on the investor protection index of the bidder minus the score on the investor protection index for the target.

3.6.2. Interaction variables

I also add interaction variables to the cross-sectional regression; these variables are a combination of the UK vs. Continental Europe variable and the private vs. public variable. This variable is inserted because the output of the univariate analysis gave some remarkable results for these variables and it could be that they together could have a significant impact in the cross-sectional analysis. The reference group for these interaction variables is the group of deals which contains a bidder from the United Kingdom and a target that is public. The coefficients of the other interaction variables measure the expected differences relative to this reference group.

3.6.3. Multicollinearity

(24)

which each other. Following Hill, Griffiths and Judge (2001) pairwise correlation higher than 0.8 will mean that one of the variables can not be used in the regression analysis.

Table 10 shows the correlation matrix. A high correlation between the interaction variables and the two variables of which the interaction variable is constructed is found, but these variables are not used at the same time in a regression model. The interaction variables are used in a regression function where the variable UK vs. Continental Europe and the variable private vs. public are eliminated. For the other variables, I do not find correlations above the 0.7 so there is no indication for multicollinearity.

3.6.4. Heteroskedasticity

(25)

4. Results

In this part the results of the event study, the univariate analysis and the cross-sectional analysis will be discussed.

4.1. Results of the event study

Table 3 reports the cumulative average abnormal returns for the various estimation windows. For both groups, the European takeovers of Chinese targets and the European takeovers of American targets, an announcement does not lead to significant abnormal returns around the event date. The difference between the CAARs of European-Chinese takeovers and European-American takeovers is also insignificant.

Table 3: Descriptive statistics for various event windows of the cumulative abnormal returns for the bidder

CAAR CAAR CAAR CAAR CAAR

41 days 21 days 11 days 5 days 3 days

Europe-China (N=114) Mean -0,79% 0,91% 0,48% 0,53% 0,39% Maximum 27,18% 52,06% 28,77% 17,59% 12,02% Minimum -36,08% -17,78% -13,84% -8,34% -7,16% Standard deviation of the mean 1,80% 1,29% 0,93% 0,63% 0,12% t-statistic -0,44 0,71 0,51 0,85 0,81 Europe-USA (N=116) Mean 1,75% -0,03% -0,89% -0,22% -0,33% Maximum 83,76% 52,58% 34,47% 38,52% 45,23% Minimum -70,16% -79,45% -73,31% -74,79% -75,33% Standard deviation of the mean 1,99% 1,43% 1,03% 0,70% 0,54% t-statistic 0,88 -0,02 -0,87 -0,32 -0,61 Difference (China-USA) -2,54% 0,94% 1,37% 0,76% 0,72% T-statistic -1,40 0,61 1,05 0,70 0,96

(26)

main hypothesis, that the bidder returns for takeovers with Chinese target are higher than for takeovers with American targets, is rejected.

Figure 1: The cumulative average abnormal returns for European –Chinese and European-American takeover

-1,00% -0,50% 0,00% 0,50% 1,00% 1,50% 2,00% 2,50% 3,00% 3,50% -20 -16 -12 -8 -4 0 4 8 12 16 20 event days pe rc en ta ge CAAR usa CAAR china

4.2. Results of the univariate analysis

The complete results of the univariate analysis can be found in appendix 3. With this analysis I analyze whether there are subsamples within our original sample, which based on a specific characteristic, do have significant abnormal returns around the event date. The figures in appendix 3 show the movement of the CAAR during the whole event window, the tables show only the CAARs for the event windows of 41 days. There are some significant results from this analysis.

The only subsample that is significantly different from zero is shown in table 19 of appendix 3. The acquisition of a public target in the USA, this leads to a negative return for the bidder of -8.74%, at a significance level of 5%.

(27)

that are only announced. The deals that are only announced have a CAAR that is higher at a 10% significance level. Deals that are completed in this data-set have reached this status before October the sixth 2006 according to the Zephyr database. Apparently it is not possible for investors to anticipate at the time of the announcement whether the deal will be completed or not. It even seems that investors value deals that are not completed yet better than deals that are completed.

Secondly, as shown in table 16, there is a difference of 4.31% in the CAARs between Continental Europe-Chinese takeovers and Continental Europe-USA takeovers. The takeovers of the American targets are valued positively, while the takeovers of the Chinese targets are valued negatively. Martynova and Renneboog (2006) found that Continental European bidders value a cross-border takeover higher than UK bidders, but this was only for takeovers within Europe. When looking at some of the Continental European countries separately in table 14, there are no significant findings. The difference in the CAARs for the Netherlands is nevertheless notable; takeovers with Chinese targets have a large negative CAAR, while takeovers with an American target have a large positive CAAR, but the difference is not significant.

(28)

There is finally, a significant difference between the CAARs of both groups for the ‘unknown payment method’ category, as shown in table 20 of appendix 3. The difference between the two samples for this category is 5.73 % at a 5 % significance level. Also within the USA there is a significant difference between takeovers paid with cash and the ‘unknown payment method’ takeovers. Deals of which the payment method is unknown have a higher CAAR. Because it is the unknown group, the explanation behind this difference is hard to find out.

Figure 2 and figure 3 of appendix 3 show the development of the CAARs by the form of the deal. Chari, Quimet and Tesar (2004) showed that acquiring a majority control of a target from an emerging country led to significant higher abnormal returns for the bidder, than acquiring a minority control. Our data do not support this finding; for both groups the abnormal return are not significantly different from zero and figure 2 and 3 show that the minority control CAARs are even larger than the majority control CAARs.

4.3.

Results of the cross-sectional analysis

Table 4, on the next page, shows the results of the cross-sectional analysis of the whole sample, for the event window of 41 days and for the event window of 3 days. For both event windows the dummy variable for the target country (China or the USA) does not show that there is a significance difference between takeovers of targets from China and targets from the USA. In other words, investors, on average do not value a European takeover with a Chinese target differently from a European takeover with an American target.

For the 41-days event window two dummy variables influence the CAAR. When the target is a public company this has a negative influence on the CAR at a 10% significance level and when the bidder is manufacturer this also has a negative influence on the CAR. The other control variables do not have coefficients that are significantly different from zero. The analysis for the 3-days event window does not give any significant coefficients.

(29)

For targets from China, the results of the 41-days event period shows that takeovers in period 2 (the years 2002 and 2003) had significantly negative consequences. Table 21 also shows that the legal status of the target for the Chinese 3-days event window is significant; if the Chinese target is public this has a positive influence on the CAR at a 5 % significance level.

Table 4 Results cross section analysis whole sample

This table shows the results of the cross-sectional regression, where the dependent variable is CAAR(41 days) or CAAR (3days). All variables except the ‘difference in investor protection’ variable are dummy variables. For dummy variable ‘Target is from China’ 1=China and 0= the USA. For dummy variable ‘Majority’ 1= after the takeover the bidders has a majority control over the target and 0=minority control or joint venture. For dummy variable ‘Completed’ 1= the deal is completed and 0=the deal is only announced or pending. For dummy variable ‘Focus’ 1= the bidder and target have the same four-digit sic code and 0=other. For dummy variable ‘Target is public’ 1= the target is a public company and 0= the target is a private company. For dummy variable ‘Cash’ 1= the deal is paid with cash and 0= paid otherwise or payment method is unknown. Dummy variable ‘Period 2’ denotes the period 2002-2003 and dummy variable ‘Period 3’denoted the period 2004 till the first half of 2006. For dummy variable ‘Finance’ 1= the bidder is in the finance sector and 0=other. Manufacturing; 1= the bidder is in the manufacturing sector and 0= other. Service; 1=the bidder is in the service sector and 0= other. The ‘bidder from the UK’ dummy gives 1= for bidders from the UK and 0= other. The last variable ‘Diff in inv. Protection’ is the score on the investor protection index of the bidder minus the score of the target. Statistical significance is based on White’s heteroscedasticity-adjusted standard errors. Statistical significance at 10%, 5% and 1% level is denoted with *,** and ***.

Total sample CAARs 41 days CAARs 3 days

Coefficient Prob. Coefficient Prob.

Intercept 0.006 0.608 0.000 0.937

Target is from China -0.045 0.313 0.013 0.504

Majority -0.017 0.547 0.005 0.668 Completed -0.024 0.291 -0.011 0.314 Focus 0.018 0.446 0.009 0.403 Target is public -0.073 *0.057 0.006 0.78 Cash -0.024 0.299 -0.014 0.211 Period 2 -0.041 0.29 -0.015 0.344 Period 3 -0.031 0.387 -0.010 0.478 Finance -0.022 0.536 0.015 0.461 Manufacturing -0.045 *0.087 0.000 0.979 Service -0.035 0.434 0.003 0.852

Bidder from the UK -0.025 0.542 -0.008 0.637

Diff in inv. Protection 0.003 0.789 -0.002 0.634

Observations N=230 N=230

Adj. R-squared -0.003 -0.024

F-statistic 0.948 0.591

(30)

Table 5 Results cross section analysis for the USA separately

For explanation of the variables in this cross-sectional regression see table 4. Statistical significance at 10%, 5% and 1% level is denoted with *,** and ***.

USA CAARs 41 days CAARs 3 days

Intercept 0.02 0.704 -0.012 0.579 Majority -0.007 0.888 0.012 0.616 Completed -0.054 0.226 -0.026 0.223 Focus 0.023 0.599 0.02 0.358 Target is public -0.102 **0.030 -0.008 0.809 Cash -0.074 *0.093 -0.019 0.4 Period 2 -0.025 0.709 -0.028 0.363 Period 3 -0.036 0.583 -0.016 0.547 Finance -0.078 0.11 0.009 0.818 Manufacturing -0.101 **0.043 -0.005 0.854 Service -0.079 0.34 -0.012 0.727

Bidder from the UK -0.025 0.767 -0.012 0.739

Diff in inv.Protection -0.003 0.917 -0.007 0.53

Adj. R-squared 0.003 -0.046

Prob. (F-statistic) 0.426 0.855

Table 6 Results cross section analysis for China separately

For explanation of the variables in this cross-sectional regression see table 4. Statistical significance at 10%, 5% and 1% level is denoted with *,** and ***.

China CAARs 41 days CAARs 3 days

Intercept -0.013 0.39 0.005 0.143 Majority -0.012 0.587 0.002 0.697 Completed -0.001 0.964 -0.001 0.912 Focus 0.001 0.976 0.000 0.995 Target is public 0.042 0.484 0.050 ***0.000 Cash 0.02 0.27 -0.007 0.163 Period 2 -0.063 *0.068 -0.006 0.400 Period 3 -0.03 0.318 -0.002 0.74 Finance 0.012 0.798 0.011 0.245 Manufacturing 0.009 0.623 0.008 0.217 Service 0.011 0.678 0.011 0.162

Bidder from the UK -0.006 0.861 0.003 0.705

Diff in inv.Protection 0.004 0.625 0.001 0.789

Adj. R-squared -0.022 0.106

(31)

4.3.1. The results of the cross-sectional regression with interaction variables

The cross-sectional regression is also conducted using an interaction variable which combines the variables ‘legal status of the target’ and ‘UK vs. Continental European bidders’. I also included an extra dummy variable for bidders from the Netherlands, because of the notable results from the univariate analysis. The rest of dummy variables stay the same. The results of these cross-sectional regressions are in table 22, table 23 and table 24 of Appendix 4.

Table 22 shows the results for the whole sample. Only when the bidder is a manufacturer CAR is significantly negative. For both event windows I find no any other variables that are significant. Compared to the prior results of the regression analysis without interaction variables, this means that the interaction variables have less influence on the CAR than the variable legal status alone. All three interaction variables (Cont. Europe-Private, Cont. Europe-Public and UK-Private) have a positive coefficient, this indicates that there are expected differences between these CARs that are positive in relation to the reference group (UK- public). Oaxaca and Ransom (1999) mention that it is not possible to identify the separate contributions of the dummy variables, but it is possible to estimate the relative effects of the dummy variables.

Table 23 provides the results for the takeovers of American targets. The results are similar to the results of the regression without the interaction variables; paying with cash and being a manufacturer negatively influence the CAR of the bidder. In this regression the three interaction variables also have positive coefficients, but none of them is significant.

The results of the regression with interaction variables of Chinese targets can be found in table 24. For the 41-days event window all the interaction variables and the period 2002-2003 have significant negative differences to their reference group. Also being a bidder from the Netherlands seems to have a negative influence on the returns for the bidder.

In the results for the 3-days event window the Continental Europe-Public variable and the variable for period 2 (2002 till 2003) is no longer significantly different from the reference group. The other two interaction variables are still significantly different from the reference group. Being a bidder from the Netherlands still has a negative influence, moreover being a bidder from the service sector has a positive influence on the CAR.

(32)

protection variable influences the CAR. This is regretful, because it may have explained why there are differences between the European countries and the different results they gain from takeovers with American targets or with Chinese targets.

4.3.2. The results of White’s Heteroskedasticity test

The results of White’s heteroskedasticity can be found in

Table 25 in Appendix 4. There is a strong indication that the error terms are heteroskedastic in some of the cross-sectional analyses. Therefore the regression is estimated with heteroskedasticity robust standard errors; this however does not provide much improvement

.

6

6_{One of the variables that seem to cause this heteroskedasticity is the dummy variable ‘target is from China’. Apparently there is a}

(33)

5. Conclusion

This thesis studied the differences between two groups of takeovers; European bidders taking over Chinese targets and European bidders taking over targets from the USA. I studied the period between 1999 and the first half of 2006 and investigated 114 deals with China and 116 with the USA.

Prior research on cross-border acquisitions showed that it is hard for the bidder to get a positive stock return around the announcement date of the deal (Campa and Hernando, 2004) (Moeller and Schlingemann, 2004). But Chari, Ouimet and Tesar (2004) showed that bidders from a developed country did get a positive return when they invested in emerging countries. An important factor in the success of the deal for the bidder was to own, after the takeover, a majority share of the target company.

However, the results of the event study in this thesis do not show significant abnormal returns in the bidder’s stock returns around the announcement date. This finding is the same for both takeovers with Chinese targets as well as for takeovers with target from the USA. This contradicts the hypothesis that acquisitions of Chinese targets would lead to more value creation for the bidder than acquisitions of targets from the USA. It even seems to be the other way around; deals with targets from the USA gave higher positive returns than deals with Chinese targets, though this difference is not significant either.

In the univariate analysis and the cross-sectional analysis I tried to find deal characteristics that influenced the bidder returns. For many of the characteristics, that according to the literature could have an effect on the value of a takeover for the bidder, I do not find significant results. This is the case for acquiring a majority control and following a focus strategy.

(34)

emerging countries as China, deals with public companies are valued higher than deals with a private target, for which it is hard to find the fundamental firm value. In developed countries these information problems are smaller and there are several reasons why private targets provide higher returns for the bidder, like a less competitive takeover market.

Another important significant finding, from the univariate analysis, is that bidders from Continental Europe get higher returns from deals with the USA, than from deals with China. For bidders from the United Kingdom there is no significant different between both target countries. A possible explanation could be the difference in corporate governance systems between the UK and Continental Europe, but I do not find further evidence for this. A notable finding from the cross-sectional analysis is that deals that involved a bidder that is a manufacturer also had a negative influence on the value creation process compared to bidders from other industries.

The interaction variables only led to some significant results for acquisitions with China; compared to bidders from the UK with public targets, all other three groups have significant negative coefficients.

In summary, the hypothesis that investing in China is a great opportunity for European bidders compared to the USA is rejected. In other words acquisitions of Chinese targets are fully accepted in the world of cross-border mergers and acquisitions and it not possible to gain better returns with these acquisitions.

5.1. Recommendations for future research

For future research there are some recommendations I would like to give. This paper is a contribution to the existing literature on cross-border acquisitions from Europe with targets from China and the USA, but there are some issues that deserve special attention.

Firstly, in this thesis I do not include any target firm characteristics, because these characteristics are not widely available. If this were not the case, maybe some important target characteristics could be found to have an important influence on the bidder returns.

(35)

included, while being aware that it could be difficult to recognize the abnormal returns that the different announcements create in the bidder returns

Thirdly, this paper only has a short-term focus, to observe whether there are differences between undertaking an acquisition with China and the USA. Other research could also take the long term success also could be taken into account. It could be that there is a difference between the short-term reaction of the shareholders of the bidder and the long short-term operational success. In that case not only stock returns should be used as a measurement of success, but also accounting data could be used.

(36)

References

Aw M.S.B. and Chatterjee R.A. (2004), “The performance of UK firms acquiring large cross-border and domestic takeover targets”, Applied Financial Economics, 14, p.337-349.

Aminian N. and Campari S. (2005), “Macroeconomic determinants of Cross-border mergers and acquisitions: European and Asian evidence”

http://www.univ-lehavre.fr/actu/itlcsge/aminian.pdf

Berkovitch E. and Narayanan M.P. (1993), “Motives for Takeovers: An Empirical Investigation”,

The Journal of Financial and Quantitative Analysis, 28(3), p.347-p362.

Brooks C. (2002), “Introductory Econometrics for Finance”, University Press, Cambridge

Brown S.J. and Warner J.B. (1985), “Using daily stock returns”, Journal of Financial Economics, 14, p.3-31.

Campa J.M. and Hernando I. (2004), “ Shareholder Value Creation in European M&As”,

European Financial Management, 10(1), p.47-81.

Campa, J.M. and Kedia S. (2002), “Explaining the Diversification Discount”, The Journal of

Finance, 57(4), p.1731-1762.

Chari A., Ouimet P.P. and Tesar L.L (2004), “Acquiring Control in Emerging Markets: Evidence from the Stock Market”, http://emlab.berkeley.edu/users/webfac/gourinchas/e281_sp05/e281-chari.pdf

Cheng L.T.W, Fung J.K.W. and Lam K. (1998), “An examination of the determinants of stock price effects of US-Chinese joint venture announcements”, International Business Review, 7, p.151-161.

Conn R.L., Cosh A, Guest P.M. and Hughes A. (2005), “The Impact on UK Acquirers of Domestic, Cross-border, Public and Private Acquisitions”, Journal of Business Finance &

Accounting,32, p.815-870.

Dennis, D. J., Diane K. D. and Keven Y. (2002), “Global Diversification, Industrial Diversification, and Firm Value”, Journal of Financ ,57(5), p.1951-1979.

Eckbo, B. E. (1992) Acquisitions, in The New Palgrave Dictionary of Money & Finance (Eds) J. Eatwell, M. Milgate and P. Newman, Macmillan, London.

Froot, K. and Stein, J. (1991), “Exchange rates and foreign direct investments: an imperfect capital markets approach”, Quarterly Journal of Economics, 106, p.1191–1271.

Gaughan, P.A. (2002), Mergers, Acquisitions, and Corporate Restructuring, (New York: Wiley)

(37)

Healy P., Palepu, K. and Ruback, R. (1997), “Which takeovers are profitable? Strategic or financial?” , Sloan Management Review, 38, 45-57.

Hill R.C., Griffiths W.E. and George G.J. (2001) “Undergraduate Econometrics”, John Wiley & Sons, Inc

La Porta R., López-de-Silanes F., Shleifer A. and Vishny R. (1998), “Law and Finance”, Journal

of Political Economy, 106(6), p.1113-1155.

Lay H. T. (2003), “Mergers and acquisitions in China” SSRN: http://ssrn.com/abstract= 527442 Khanna, T. and Palepu K. (1997), “Why Focused Strategies May Be Wrong for

Emerging Markets?” Harvard Business Review, 75 (4), p.41-49.

MacKinlay A.C. (1997), “Event studies in Economics and Finance”, Journal of Economic

Literature, 35, P.13-39.

Martynova M. and Renneboog L. (2006), “Mergers and acquisitions in Europe”, Discussion

paper, ISSN 0924-7815

Mody A. and Negishi S. (2002), “The role of cross-border mergers and acquisitions in Asian restructuring” FIAS/PREM seminar series, ifc.org

Moeller S.B, Schlingemann F.P. and Stulz, R. (2004a), “Firm size and the gains from acquisitions”, Journal of Financial Economics, 73, p.201–228.

Moeller S.B., Schlingemann F.P. (2004), “Are cross-border acquisitions different from domestic acquisitions? Evidence on stock and operating performance for U.S. acquirers”

Forthcoming in the Journal of Banking and Finance

Moeller S., Schlingemann F. and Stulz R. (2005), “Wealth Destruction on a Massive Scale? A Study of Acquiring-Firm Returns in the Recent Merger Wave”, Journal of Finance 60, p.757-782. Norton P. M. and Chao, H. (2001), “Mergers and Acquisitions in China” China Business Review, 28 (5), p.46-53.

Oaxaca R.L. and Ransom M.R. (1999) “Identification in detailed wage decompositions”, The

Review of Economics and Statistic, 81(1), p.154-157.

Peng M. W. (2006), “Making M&A Fly in China.”, Harvard Business Review, Vol. 84 Issue 3, p.26-27.

Rau R. P. and Vermaelen T. (1998), “Glamour, value and the post-acquisition performance of acquiring firms”, Journal of Financial Economics, 49, p.223–53.

Servaes H. and Zenner M. (1994), “Taxes and the returns to foreign acquisitions in the United States”, Financial Management, 23, p.42-56.

Seth A. (1990), “Sources of value creation in acquisitions: An empirical investigation”, Strategic

(38)

Shaked I., Michel A. and McClain D. (1991), “The foreign acquirer bonanza: myth or reality?”,

Journal of Business Finance & Accounting, 18, p.431-447.

Tan, L.H. (2003), “Mergers and Acquisitions in China”, SSRN: http://ssrn.com/abstract=527442

websites

www.evd.nl

(39)

Appendix 1 Takeover Characteristics

Table 7: Takeover characteristics (source Zephyr)

This table describes the total sample of announcement of cross-border deals of bidders from the EU-15 countries with targets from China and the USA between 1999 and the first half of 2006 according the Zephyr database. The bidder must be listed on a European stock market. Statistical significance at 1%, 5% and 10% level is denoted with a, b and c.

China USA

No. % No. % Difference p-value

(40)

Table 8: Takeover Characteristics (source Zephyr)

This table describes the total sample of announcement of cross-border deals of bidders from the EU-15 countries with targets from China and the USA between 1999 and the first half of 2006 according the Zephyr database. The bidder must be listed on a European stock market. Statistical significance at 1%, 5% and 10% level is denoted with a, b and c.

China USA

No. % No. % Difference P-value

Method of payment Cash 122 _40,9% 986 _{45,6% -4,69% 0,124} Converted Debt _{0 0,0% 1} _0,0% _0,317 Debt 2 132 _{6,1% -5,44% 0,000 a} Deferred payment 2 _0,7% 28 _1,3% _0,242 Earn out 6 95 _{4,4% -2,38% 0,010 b} Loan notes 0 _0,0% 17 _0,8% _{0,000 a} Shares 21 348 _{16,1% -9,06% 0,000 a} Other 2 _0,7% 17 _0,8% _0,826 Unknown method of payment 143 986 45,6% 2,36% 0,447 Total 298 _100,0% 2161 _100,0% deal type Acquisitions 152 54,5% 1807 83,9% 0,000 a Joint Ventures 48 28 1,3% 15,90% 0,000 a Minority Shares 78 28,0% 320 14,8% 0,000 a IPO 1 0 0,0% 0,36% 0,317 Merger _{0 0,0% 1 0,0%} -0,05% 0,7% -0,62% 2,0% -0,79% 7,0% -0,12% 48,0% -29,37% 17,2% 13,11% 0,4% -0,05% 0,317 Institutional buy-out _{0 0,0%} 0,4% _{-0,37% 0,005} Total ₂₇₉ _{2155 100,0%} Deal status 111 _{39,78% 407} _{20,95% 0,000} Completed 134 _{1659 76,77%}_-28,74% _a Pending (regulating approval) _{8,24% 22 1,02%} _0,000_a 3 1,08% 11 8 _b 100,0% Announced _18,83% _a 48,03% 0,000 23 7,23% Pending (shareholder approval) _0,51%_{0,57% 0,430}

(41)

Appendix 2 Overview Corporate Governance indices and correlation matrix

Table 9: Overview of corporate governance indices (source: the World Bank)

DO EUROPEAN SHAREHOLDERS BELIEVE THAT INVESTING IN CHINA IS MORE VALUABLE?

THAT INVESTING IN CHINA IS MORE

VALUABLE?

An event study about the differences in the announcement effect

between acquisitions of Chinese and American companies.

Hélène F. M. Wiegerinck

University of Groningen

Faculty of Economics, Finance department