Cross-border M&As : an analysis on shareholder returns for emerging and non-emerging market mergers

(1)

1

Cross-border M&As: An analysis on shareholder returns for

emerging and non-emerging market mergers.

Amsterdam Business School

Name Mike Trott

Student number 10264620

Program Business Economics

Specialization Finance Number of ECTS 15

Supervisor dr. I. Naaborg Target completion 7/7/2016

Abstract

This paper examines if mergers and acquisitions (M&As) in emerging markets create more shareholder value than non-emerging market cross-border M&As in the period of 2002-2016. Several cross-border specific determinants that can influence shareholder value are also examined in this paper by looking at factors such as the level of corporate governance, institutional infrastructure, investment size, level of control, public status, cultural distance, top mediation and debt-to-asset ratio. The analysis suggests that cultural distance and institutional infrastructure have a negative effect on abnormal returns.

Explanatory variables such as investment size and level of control proved to have no significant effect on abnormal returns. The results also show a positive effect on shareholder value for corporate governance and firms operating in the same industry. Finally the analysis shows that emerging market M&As lead to value destruction which is in contrast to findings of previous literature. However, limitations of this paper are that it only adopts a bidder side view and does not look at combined wealth effects. Another limitation is that investors do not always fully understand mergers which can bias the results.

(2)

2

1. Introduction

During the recent economic crisis Merger & Acquisition (M&A) transactions declined. In 2015 cross-border M&A in which the target is located in an emerging market increased to approximately 22%, in 2014 this was around 16% (Grave et al., 2012). During the crisis, investor confidence was low and managers were looking for investment opportunities to diversify their portfolios. The low prices and other M&A benefits such as market expansion resulted in emerging markets, such as Asia and South America, becoming more popular as acquisition targets (Aybar & Ficici, 2009).

In general M&As will only occur when the M&A increases a firms’ wealth. There are several motives for a firm to consider a domestic merger. These include synergy gains in which the combining of the two firms offers an improvement in efficiency of production, tax benefits or even increased market power. When considering cross-border M&As there are also other, cross-border specific, factors to take into account that can either add or destroy value. For example, a lower level of corporate governance can hinder a merger as the lack of effective monitoring systems can create incentives for managers to increase value at the expense of shareholders. This is especially important in emerging markets due to their underdeveloped nature and therefore lower level of corporate governance compared to non-emerging markets. (Erel, Liao, & Weisbach, 2012). For example, emerged market firms are more likely to already have experienced a large growth. Therefore, it is also more likely that these firms have adapted to the increased risk of managerial misbehavior by increasing the quality of the corporate governance. This does not hold for all emerging market targets, as they are still in the developing phase. Another influential factor is institutional infrastructure which captures the level of market regulation in a country such as property right protection, environmental regulation and trade barriers. The institutional

infrastructure is most likely to be worse in emerging markets than emerged markets as emerging markets are by definition less developed that emerged markets (Grave et al., 2012). This leads to higher levels of operation and investment risk due to inefficient legislation and corrupt legal infrastructure. The increase in investment and operation risk leads to more frictions and added costs which decreases the overall shareholder gain for emerging market mergers (Aybar & Ficici, 2009). For example, when a country has a low level of institutional infrastructure it might be harder for an acquiring firm get or enforce a patent in the target country. This could increase costs and thus frictions due to differences in institutional infrastructure and thus reduce shareholder gains.

The aim of this paper is to evaluate the extent to which emerged market cross-border M&As differ in their effect on shareholder gain, from a U.S. investor perspective, as compared to cross-border

(4)

4 M&As in emerging markets. Currently there are several papers of which most compare cross-border M&As to domestic M&As. For example, Aybar & Ficici (2009) studied the effect of domestic and cross-border M&A on the shareholder value of the acquirer. Their analysis shows that cross-cross-border mergers do not create shareholder value, they destroy shareholder value. Lowinski et al. (2004) conducted an event study on the effect of domestic and cross-border acquisitions on shareholder wealth. Their results show that there is no significant difference in wealth creation for shareholders either domestic or cross-border M&As but that the overall effect on shareholder wealth is positive. However, the lack of

difference in wealth effect could be due to the fact that most cross-border M&As were in the E.U. and due to the higher financial integration of the European capital market the wealth gains could be offset. Doukas & Travlos (1988) find that cross-border mergers in general have a more positive effect on shareholder value for the acquirer as compared to domestic mergers. Especially when the acquirer has no previous presence in the target country. Rossi & Volpin (2004) also found similar results in their sample. Overall the literature suggest that cross-border mergers have a positive effect on shareholder wealth or perform at least similar to domestic M&As. However, what can be noticed is that the focus of the current literature is more on the difference between cross-border and domestic mergers. There is less focus on the differences between cross-border mergers in emerging markets and emerged markets. Therefore, the contribution of this thesis is to examine cross-border mergers with the interest to see if there is a difference between a cross-border merger in an emerged country and an emerging market in terms of shareholder value.

This ultimately leads to the following research question; Do cross-border M&As in emerging markets differ in their effect on shareholder value compared to non-emerging market M&As?

To answer the research question the paper will start by performing an event study in the period of 2002-2016 to measure the overall effect of a cross-border in developed countries and emerging market countries on shareholder value. The shareholder value for the event study will be measured by the Cumulative Average Abnormal Returns (CAARs) which is frequently used in studies as a measure of shareholder wealth. To provide more insight this thesis will use a cross-sectional analysis to explain the variation in Cumulative Abnormal Returns (CARs) by adopting a similar method as Aybar & Ficici (2009). Emerging market countries will be selected by country using the MSCI Emerging Market Index. Currently there are 23 emerging markets of which Brazil, Chile, Colombia, Mexico, Peru, Czech Republic, Egypt, Greece, Hungary, Poland, Qatar, Russia, South Africa, Turkey, United Arab Emirates, China, India, Indonesia, Korea, Malaysia, Philippines, Taiwan and Thailand.

(5)

5 the influential factors. This will provide the basis for establishing the hypothesis. Chapter 3 will explain the methodology in more detail by elaborating the multi-factor model as well as the calculation for the CAAR values. Chapter 4 will provide the data as well as the accompanying descriptive statistics. Chapter 5 will show the results of the event study, the results and interpretation from the multi factor model as well as the robustness checks and lastly chapter 6 will contain the conclusion of this paper.

2. Literature review

In general, a firm will only consider an M&A if it will create value. Therefore, the literature review will start by examining whether or not cross-border M&As actually do create value. When looking at cross-border M&As in general, several studies have shown that overall cross-border M&As do in fact create value. Based on the outcome of their analysis on the wealth effect of cross-border and domestic M&A, Doukas & Travlos (1988) find that cross-border M&As do have a significiant positive effect on shareholder value. Lowinski, Schiereck, & Thomas (2004) find that on average cross-border M&As perform at least the same as domestic M&A. However, due to the fact that the sample of Lowinski, Schiereck, & Thomas (2004) only contained European firms they state that this result is most likely due to the high capital market integration in Europe as when international capital markets are highly integrated and information is readily available then no diversification gains should be earned by cross-border M&As. On contrary, there are also papers that find less positive effects of cross-border M&A such as Camp & Hernando (2004) who show that domestic M&As outperform cross-border M&As. However, they also conclude that there are still cases in which cross-border M&As outperform domestic M&As such as when there is a lower level of regululation in emerging markets. Therefore, these papers show that cross-border M&As perform at least as good as domestic mergers.

To see which factors could influence shareholder value Aybar & Ficici (2009) studied the effect of cross-border M&As in emerging markets compared to domestic M&As. According to Aybar & Ficici (2009), there are several specific factors that can influence the Cumulative Abnormal Returns (CAR), such as; cultural distance, investment size, level of control, company status, corporate governance and institutional geography. These factors also correspond with the results from Erel, Liao, & Weisbach (2012) who find that the most important factors for a cross-border merger are cultural distance,

differences in corporate governance as well as the level of institutional infrastructure. To further explain the factors this chapter will make a distinction between influential factors and explain how they

(6)

6

2.1 Corporate Governance

The level of corporate governance can be an explaining factor in CAAR deviations, especially for emerging market targets. For example, emerged market firms already experienced company and economic growth in their emerging phase, therefore they have already adapted to the increase in managerial discretion by increasing their effective monitoring systems, as opposed to emerging market firms. Therefore, emerging market firms still create relatively more incentives for managers to increase value at the expense of the shareholders. Aybar & Ficici (2009) studied the effect of corporate

governance in the period of 1991-2004. The cross-sectional OLS results of Aybar & Ficici (2009) show that the lower level of corporate governance has a negative effect on shareholder wealth. Harford, Humphery-Jenner, & Powell (2012) also conducted a cross-sectional analysis on the effect of corporate governance on abnormal returns. Their results show that bad corporate governance in cross-border targets ultimately leads to value destruction due to managerial misbehavior. Their results also show that bad coporate governance is more likely to be the case in the case of private targets as the disceplenary role of shareholders is removed. Another addition was by Starks & Kelsey (2013) who investigated the effect of corporate goverance on CARs using five corporate governance proxies such as whether or not the target is an G7 country, the countries legal origin, a shareholder rights measure, the level of accounting disclosure and a shareholder protection measure. Their results suggest that the acquirers’ shareholders require compenstaion for low levels of corporate governance in cross-border M&As. Their results also show that bad corporate governance can be partially solved by a majority of control. However, it should also be noted that a U.S. acquiring firm is overall more developed than an emerging market firm, therefore it has a more developed corporate governance system. When the U.S. firm acquires the emerging market firm, the emerging market firm can also learn from the more developed corporate governance systems of the acquirer or vice versa. This could lead to an increase shareholder value. When comparing emerged-cross border to non-emerged cross-border M&As these results suggest that, due to the relatively low levels of corporate governance in most emerging market countries, the negative effect on shareholder gain would be greater implying a better performance of emerged cross-border M&As.

2.2 Institutional Infrastructure

Another factor that can benefit emerged market M&As is institutional infrastructure which captures the level of market regulation in a target country such as property right protection,

(7)

7 emerging markets than non-emerging markets due to their underdeveloped nature as emerging markets are just adapting to the changes arising from the economic development. This leads to higher levels of operation and investment risk due to inefficient legislation arising from corrupt legal infrastructure, which in turn creates more frictions and added costs that decrease the overall shareholder gain (Aybar & Ficici, 2009). Chari, Ouimet, & Tesar (2010) conducted an cross-sectional OLS regression in the period of 1986 - 2006 on the effect of insitutional infrastracture on abnormal return for mergers in Latin America and East-Asia. Their findings show that better institutional infrastructure addresses poor property rights and contractual problems which leads to better facilities for the acquirer. This implies that the acquirer can utilize the targets resources better thereby raising expected future cash flows for investors and thus shareholder value. For example, consider a target firm that is also a tech firm. The acquirer knows that the target has valuable new technology which they are reluctant to release due to bad patent enforcement. Therefore, the acquirer would not benefit from the new technology and thus may lose any extra cash flows and thus shareholder value. Rossi & Volpin (2004) also looked at the effect of institutional infrastructure on shareholder value between 1990 and 1999 by using an investment index from La Porta (1998). The La Porta Index is based on the quality of the accounting standards, quality of law enforcement and a common law dummy and can therefore be seen as an approximation for the institutional infrastructure. The results show that a higher level of institutional infrastructure leads to a higher abnormal returns, which leads to a higher volume of cross-border M&A in general. However, Coffee (1999) finds that even though institutional geography might seem

important, the more takeovers by developed firms in an emerging country the higher the quality of the institutional infrastructure. According to Coffee (1999) the developed firms will influence the current institutional infrastructure by increasing pressure on the government to improve their institutional infrastructure, for instance by lobbying. Ultimately, these analyses suggest that emerging market firms would perform worse than emerged market firms, in terms of their effect on shareholder gains.

2.3 Investment Size

The next determinant is investment size which is related to the fact that a merger can lead to economies of scale in production, marketing and distribution. Investment size is defined as the ratio of acquired stake to acquirers’ market value. According to Lamacchia (1997) cross-border mergers may have significant benefits from the more efficient use of fixed capital and market presence. The results show that investment size has a positive influence on shareholder value for the acquirer as the expected future cash flows would be higher. This would especially affect emerging market firms. Due to their lack

(8)

8 of expertise, the allocation of capital would be less efficient in emerging markets, therefore the benefit from the knowledge U.S. firms bring would lead to more efficient use of capital. These efficiency gains raise expected future cash flows ultimately leading to higher shareholder value. As emerged market firms are generally more developed the gain in efficiency would be lower in case of an M&A. Another addition to the literature is by Moeller & Schlingemann (2005) who performed a cross-sectional analysis on the effect of several M&A determinants on abnormal returns, such as investment size, in the period of 1985-1995. The results of Moeller & Schlingemann show weak evidence for a positive influence of investment size on abnormal returns. The results of Eduardo (2009) also show that investment size also leads to higher shareholder value in cross-border M&A. The analysis was based on a dataset containing 513 cross-border M&As in Latin America. More interestingly, the countries contained in the dataset are all classified as emerging market countries. Therefore, it is expected that emerging market mergers would see higher shareholder wealth as compared to emerged market mergers (Aybar & Ficici, 2009).

2.4 Public Status of target

The public status of a target firm can also have an impact on shareholder value. For a public firm information is readily available for valuation. Therefore, the bidding firm can make a better

assessment of the value of the target, as opposed to a private target. In the case of a private target, the acquirers will have less information available for a proper value assessment. This will give acquirers more bargaining power over the private target as there is more room to negotiate about the value of the company. This in turn can lead to larger value creation for the acquirer in the case of a private target (Aybar & Ficici, 2009). Fuller, Netter, & Stegemoller (2002) conducted a cross-sectional OLS regression on the effect of public versus private firms on CARs in the period of 1990-2000. They estimate the effect of public status for different methods of payment and find similar results as Aybar & Ficici (2009), they find negative returns for acquirers when buying public firms. Moeller & Schlingemann (2005) also looked at the effect of a private target on abnormal returns in their cross-sectional analysis in the period of 1985-1995. Moeller & Schlingemann (2005) find that on average private targets yield around 1.8% higher returns compared to public targets. According to Moeller & Schlingemann (2005) this result is due to the fact that in most privately held firms takeovers of these firms have the tendency to create large blockholders. These blockholders are in turn effective monitors of company and managerial performance, which can increase expected future performance. Another addition is by Harford et al. (2012) who also investigated the effect of several determinants of cross-border M&As on the abnormal returns, such as the effect of target company status. The results of their OLS cross-sectional analysis

(9)

9 shows a highly significant positve influence of a private target on abnormal returns also explained by the monitoring effect found by Moeller & Schlingemann (2005). In most of the emerging markets the financial system is less developed as the economy is still in an emerging phase, therefore countries are just starting to adapt to the growing economy for example by improving the financial regulation. Since the financial system is less developed in most cases going public does not really benefit the firm as the markets can still be more inefficent compared to more developed financial markts (Aybar & Ficici, 2009). This would imply that there would be relatively more private firms. The opposite is true for emerged market as their financial system is relatively more developed and effcient, therefore there will be more public firms. This would mean that it would be expected that emerging market M&As could result in higher shareholder value compared to emerged market targets as there are relatively less public firms.

2.5 Level of Control in target

The final factor that can influence shareholder value is the level of control in the target by the acquirer after the merger. The higher level of control the more the acquirer can use its expertise to steer the target in a better direction. Therefore, emerging market target firms can benefit from the better management which could lead to higher future cash flows for the target thereby raising shareholder value (Aybar & Ficici, 2009). This result is complemented by the analysis of Chari, Ouimet, & Tesar (2010) who conducted a cross-sectional analysis on the effect of the level of control on CARs for emerging market firms between 1986 and 2006. They find that, on average, CARs in emerging market mergers are approximately 1% higher than domestic mergers. Francis, Hasan, & Sun (2008) also conclude that in the period of 1990-2002 a majority of control earns significant positive results for acquirers’ shareholder wealth since emerging market firms can benefit from the supervision from the acquirer. Especially if the acquirer has a relatively big percentage of control to actually steer the target firm. Another addition to the literature is by Bhaumik & Selarka (2012) who looked at the effect of ownership and other M&A charateristics on firm performance in India, an emerging market. The results show that higher levels of control lead to better target firm performance. Thus raising current and possibly future cash flows, which in turn could lead to higher abnormal returns for the acquirer. Sudarsanam, Holl & Salami (1996) also looked at the effect of ownership on abnormal returns. The cross-sectional model is based on a sample containing 429 cross-border mergers in the period of 1980-1990. Their results also show that increased level of control leads to higher abnormal returns, but not due to better performance by improved supervision but due to the decrease of the conflicts of interest between shareholders and management. This would lead to less value destruction by management and could thus raise

(10)

10 shareholder value for the acquirer. Overall it is expected that an emerging market M&A would lead to a higher shareholder wealth as compared to emerged market M&A due to the effect of better

management.

2.6 Cultural Distance

As explained earlier geography and cultural distance measures the absolute and cultural differences between countries. The lower the cultural and geographical distance between them the more likely that they will have similarities thus reducing frictions caused by communication or habits. Thus possibly increasing shareholder gains. However, this paper will only focus on the cultural distance between two countries as the geographical distance has less explanatory power and would interfere with the results (Aybar & Ficici, 2009). For example, the U.S. and Australia are geographically far from each other however culturally they can be virtually the same. Therefore, geographical distance does not say much about differences between them. Brouthers & Brouthers (2001) who analyze the effect of cultural distance on abnormal returns show that a lower cultural distance increases shareholder abnormal returns. Their dataset is based on five Eastern European emerging countries such as Poland, Czech Republic, Russia and Romania. Their analysis shows that the larger the differences between countries, the less likely it is for a developed firm to enter due to higher investment risk. Morosini, Shane & Singh (1998) also investigated the effect cultural distance on CARs between 1987 and 1992 using an OLS model. A similarity with this paper is that they also use Hofstede’ dimensions as a proxy for cultural distance. They find that a lower cultural distance enhances cross-border acquisition

performance due to the fact that an acquirer will have easier access to different routines from the target, related to innovation and entrepreneurship which can increase the merged firms’ performance over time. The same is concluded by Shenkar (2012) who conducted an empirical study on the effect of cultural distance on merger activity. Shenkar (2012) finds that a low cultural distance can raise merger activity as it helps improve corporate governance by increasing similarities and thus returns. Overall it can be stated that the higher the cultural distance, the lower the overall shareholder gains.

2.7 Mediation

Another factor that can influence both the emerging market M&As and emerged market M&As is whether or not the deal is made by a top intermediary investment bank. According to Lowinski et al, (2004) a top intermediary investment bank can negotiate better deals or can facilitate complex M&A process in emerging countries which results in higher abnormal returns for investors. Bowers & Miller

(11)

11 (1990) show that there is a positive relation between a top 25 investment bank acting as an advisor and shareholder wealth creation. Bowers & Miller (1990) conduct an OLS regression on the effect of top investment bank mediation on CARs in the period of 1981-1986. They use deal characteristics to divide the data based on whether or not the bidder, target or both are assisted by a top mediation investment bank. Their results show a positive effect on the acquirers’ abnormal returns when a top intermediary investment bank is used. Servaes & Zenner (1996) also studied the effect of top intermediary investment bank on abnormal returns. Their results show that top investment bank mediation raises abnormal returns only in complex transactions, this is due to asymmetric information for which investment banks are better equipped. However, Srinivasan & Saunders (2001) find that there is no significant effect on announcement returns in their cross-sectional OLS regression. In fact, the results show no direct relationship between the acquiring firm and the investment bank. However, they do see a signaling effect in which the top 25 investment bank mediation signals a higher probability of a successful merger. The same is reported by Rau (2000) who also conducted an OLS regression on the effect of top

mediation on CARs for M&As in the period of 1959-2000. The results show no significant effect of top intermediary on shareholder wealth for the acquiring firm. There are also papers that top mediation can have a negative influence on shareholder value, for example McLaughlin (1992) shows that mediation by a top intermediary investment bank can lead to conflicts of interest between advisors and clients in M&As due to contract incentive features and thus possibly lowering shareholder value. Therefore, overall the literature suggests that mediation by a top intermediary investment bank will have a positive influence on shareholder value for both emerging and emerged markets.

2.8 Leverage of target

The final determinant is the amount of leverage of a target firm. A high leverage ratio of a target could indicate that the firm is in financial distress. Therefore, it will most likely sell at discount as the firm could me more eager to sell and the fact that the quality of the firm is worse which would decrease the value of the firm. If the high leverage ratio, indicating financial distress, is due to bad management the acquiring firm can improve the performance of the company and therefore shareholder gain. However, if it is not due to bad management but for instance due to market changes, it would be expected to see negative effects on shareholder wealth due to higher risk. Another effect of a high debt-to-assets ratio is that the high ratio will make investors more hesitant to invest thereby causing negative abnormal returns. The total effect therefore depends on which effect is stronger. Hardford et al. (2012) also looked at the effect of leverage on abnormal returns of the acquirer in the period of 1990-2005 and

(12)

12 finds that a higher leverage leads to lower abnormal returns as it decreases the overall profitability of a target as well as the increased risk. Arikawa & Miyajima (2012) investigated the drivers behind the M&A boom in Japan and their effect on abnormal returns. They find that high leverage plays a significant role in the determination of a target as it is in indicator for future performance as a higher leverage can indicate distress and lower returns. According to Jensen (1993) this is the result of the managers’ fear of defaulting thereby leading to lower abnormal returns. These results thus imply that a higher debt ratio leads to lower returns in both emerging and non-emerging markets.

All the literature above shows that there is not enough literature available on the effect of cross-border M&A in emerging and non-emerging countries. However, given the literature above it is seen that there are multiple explanatory factors that can explain the difference in shareholder value between emerging and non-emerging country M&As. To summarize the literature suggests that cultural distance would have a negative effect on shareholder value due to increasing differences between countries. The literature suggests that in terms of investment size emerging markets M&As will earn higher shareholder gains as compared to emerged market M&As as due to higher efficiency gains of the target, the expected future cash flows are higher.The public status of the target is expected to benefit emerging market M&As more than emerged market M&As due to the less developed financial system and more private firms in emerging markets. This will lead to a better bargaining position for the acquirer and thus have a more more positive effect on shareholder wealth. The quality of corporate governance is another important factor as the higher the level of corporate governance, the less incentives there are for managers to persue value decreasing activities. Compared to emerged market cross-border M&As, emerging market cross-border M&As will have a lower qualility of corporate governance thus implying that emerging market M&As will have a more positive effect on shareholder value. Furthermore, Institutional geography also plays a significant role as the better the institutional geograhpy, the better the legal protection, the less risk the acquirer faces which raises shareholder value. In emerging market M&As this risk is higher, therefore the it has a relatively more negative effect on shareholer value compared to an emerged market M&A. Finally the level of control in the target after the merger can also be an inlfuential factor as the higher the level of control, the more the acquirer can influence decisions to steer the target and earn higher future cash flows. The emerging market firm has more potential as they are generally less developed than emerged market firms, leading to higher expected future cash flows and thus shareholder value compared to emerged market firms.

This suggests the following hypothesis: Emerging market M&As will lead to higher shareholder value compared to non-emerging market M&As.

(13)

13

3. Methodology

3.1 Event Study

In the finance literature one of the most widely used approach to evaluate stock price reactions in case of a specific event is the event study methodology. It gives researches the ability to conclude whether the event in question had a positive of negative effect on shareholder wealth. Using this methodology, this thesis therefore aims to see if there is a difference between shareholder value creation for emerging market M&As and emerged market M&As. This analysis will utilize the market adjusted approach in which it is assumed that the return of a security depends on the return of the market portfolio. This can be written as follows:

𝑅𝑖𝑡 =∝𝑖+ 𝛽𝑖𝑅𝑚𝑘𝑡+ 𝜀𝑖𝑡 (1)

In which the subscript t indicates the time, subscript i indicates certain security and subscript mkt indicates the market. Therefore, 𝑅𝑖𝑡 is the return of security i at time t, ∝𝑖 is the intercept estimated by

the model, 𝛽𝑖 measures a firm specific coefficient which is also estimated by the model, 𝑅𝑚𝑘𝑡measures

the return on the market portfolio and 𝜀𝑖𝑡 measures a random error term or abnormal return for a

specific security i at time t.

To calculate the abnormal returns we need to subtract the predicted returns from the observed returns:

𝐴𝑅𝑖𝑡 = 𝑅𝑖𝑡− (∝𝑖+ 𝛽𝑖𝑅𝑚𝑘𝑡) (2)

To make the analysis more intuitive this thesis will use the Cumulative Average Abnormal Returns (CAAR) for the event study. To calculate the CAAR we need to start with calculating the Cumulative Abnormal Returns, which is the sum of all the abnormal returns:

𝐶𝐴𝑅𝑖 = ∑𝑁𝑖=1𝐴𝑅𝑖𝑡 (3)

From this the CAAR can be calculated by taking the average CAR for each transaction:

𝐶𝐴𝐴𝑅 = 1

𝑁∑ 𝐶𝐴𝑅𝑖𝑡 (4)

𝑁

𝑖=1

This would imply that a nonzero CAAR would indicate an abnormal performance. For instance, a positive CAAR would imply positive abnormal returns. To test if the CAAR value is statistically different from zero, a t-test will be used.

(14)

14 Several event windows will be used to calculate the abnormal return deviations around the announcement date. The event windows chosen are [-5, -1], [-1, +1], [+1, +5] and [-5, +5]. This means that the expected returns for Equation (1) for each day in the event window as well as the observed return for each day in the event window. The event windows are chosen as per Aybar & Ficici (2004), who state that moving further than the current windows does not yield more significant results.

3.2 Cross sectional Analysis

As stated in the literature review, the value created by an M&A depends on a range of different factors. In order to explain the variation in the CARs this thesis will adopt a modified model based on Aybar & Ficici (2009). In the model the dependent variable is the CAR and the independent variables are; log of the bidder firms’ total assets, similar industry, company status dummy, percentage of control in the target after the merger, investment size, the level of development of institutional infrastructure, cultural distance, the level of corporate governance and a region dummy which is 1 if the target is in an emerging market and 0 otherwise. The additions are a debt-to asset ratio and a top mediation bank dummy. This yields the following model:

𝐶𝐴𝑅𝑖𝑡 = 𝛽0+ 𝛽1𝑆𝑖𝑧𝑒𝑖𝑡+ 𝛽2𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦𝑖+ 𝛽3𝐶𝑆𝑡𝑎𝑡𝑢𝑠𝑖𝑡+ 𝛽4𝑝𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑖𝑡+ 𝛽5𝐼𝑛𝑣𝑒𝑠𝑡𝑆𝑖𝑧𝑒𝑖𝑡+

𝛽6𝐼𝑛𝑠𝑡𝑖𝑡𝑢𝑡𝑖𝑜𝑛𝑖𝑡+ 𝛽7𝐺𝑜𝑣𝑒𝑟𝑛𝑎𝑛𝑐𝑒𝑖𝑡+ 𝛽8𝐸𝑚𝑒𝑟𝑔𝑖𝑛𝑔𝑖+ 𝛽9𝑇𝑜𝑝 + 𝛽10𝐷𝐴 + 𝛽11𝐶𝐷𝐼 + 𝜀

In which;

Size = the log of the bidder firms’ total assets

Industry = Industry similarity dummy. This dummy measures if both firms operate in the same industry, measured by the first two number of the SIC code of both the target and acquirer. The variable is equal to 1 if they share the same industry and 0 otherwise.

CStatus = company status dummy. This measures the status of the target firm, either public or private. The dummy equals 1 if the firm is public and 0 otherwise.

pControl = the percentage of ownership in the target after the merger.

InvestSize = defined as investment over bidder market value. It measures efficiency in which the targets’ resources can be used for achieving the companies’ goals.

Institution = a measure of the level of institutional infrastructure. This is measured by the Economic World Freedom index by the Heritage Foundation.

(15)

15 Governance = the level of corporate governance dummy measured by the Entrancement Index by Lucian Bebchuk.

Emerging = the Emerging market dummy which is the main variable of interest, it equals 1 if the target is located in an emerging market and 0 if the target is in a non-emerging market. If positive this would imply better performance of emerging market M&As compared to non-emerging market M&As. As this paper is interested to see whether or not emerging market M&As outperform non-emerging market M&As, this variable is one that has to be included in the regression.

Top = a top intermediary bank dummy which measures if the intermediary bank of the acquirer was in the top 10 financial advisor list from the Merger Market M&A report of 2015.

DA = a continuous variable that measures the debt to assets ratio of the target firm. The debt to asset ratio is defined as Total Debt over Total Assets. A higher level of Debt-to-Asset ratio could indicate CDI = the Cultural Distance Index between a target an acquirer. This is a measure proposed by Aybar & Ficici (2009) and is based on Hofstede’s five dimensions such as power distance, individuality,

masculinity, uncertainty aversion and long-term orientation. However, just as Aybar & Ficici (2009) this thesis will not consider the long-term orientation due to data constraints. The CDI index takes values between 1 and 0 with a value near 1 implying a high cultural difference on contrary a value near 0 implies a low cultural difference. The calculation of the CDI index will be explained in the Appendix.

To check for the robustness of the model, a binary logistics regression will be used. The dependent variable will be a dichotomous version of the CAR values for each time window. The dichotomous variable will have a value of one when the CAR is positive and a value of zero otherwise. This implies the following function:

𝐷𝐶𝐴𝑅𝑖𝑡 = 𝛽0+ 𝛽1𝑆𝑖𝑧𝑒𝑖𝑡+ 𝛽2𝐼𝑛𝑑𝑢𝑠𝑡𝑟𝑦𝑖+ 𝛽3𝐶𝑆𝑡𝑎𝑡𝑢𝑠𝑖𝑡+ 𝛽4𝑝𝐶𝑜𝑛𝑡𝑟𝑜𝑙𝑖𝑡+ 𝛽5𝐼𝑛𝑣𝑒𝑠𝑡𝑆𝑖𝑧𝑒𝑖𝑡+

𝛽6𝐼𝑛𝑠𝑡𝑖𝑡𝑢𝑡𝑖𝑜𝑛𝑖𝑡+ 𝛽7𝐺𝑜𝑣𝑒𝑟𝑛𝑎𝑛𝑐𝑒𝑖𝑡+ 𝛽8𝐸𝑚𝑒𝑟𝑔𝑖𝑛𝑔𝑖+ 𝛽9𝑇𝑜𝑝 + 𝛽10𝐷𝐴 + 𝛽11𝐶𝐷𝐼 + 𝜀

The logistic model results will then be compared with the OLS regression to check for the robustness of the results.

(16)

16

4 Data

The study will focus on acquisitions undertaken by publicly listed firms in the U.S. firms in the period of 2002-2015. To make a distinction between emerging and emerged market firms the most current MSCI Emerging market index will be used. Currently there are 23 emerging markets of which Brazil, Chile, Colombia, Mexico, Peru, Czech Republic, Egypt, Greece, Hungary, Poland, Qatar, Russia, South Africa, Turkey, United Arab Emirates, China, India, Indonesia, Korea, Malaysia, Philippines, Taiwan and Thailand.

To limit the sample size firms will be selected on the deal transaction value which must be higher than 10 million dollars and the control after transaction must be higher than 50% as used by Aybar & Ficici (2009). This will ensure that the sample will only contain mergers from a reasonable magnitude and acquiring firms can influence the target firm to some degree.

Data on M&A size, value, and type of merger will be retrieved from the Thompson One and ZEPHYR database. Furthermore, data on the target size, book values as well as market values will be gathered from DataStream. The top intermediary list will be collected from the Merger Market M&A report of 2015. To measure the level of development of the institutional infrastructure, the World Economic Freedom index will be used. This index is based on four pillars namely, rule of law,

governmental freedom, regulatory efficiency and open markets and can therefore be seen as a proxy for the institutional infrastructure. The data on the World Economic Freedom Index is readily available from the Heritage Foundation. The measure for cultural distance is based on the CDI index from Aybar & Ficici (2009) which is based on Hofstede’s cultural dimensions. The data on these dimensions are available from Hofstede’s online database. The Hofstede dimensions are a proxy for the national culture and are based on estimations on power distance, individualism, masculinity, uncertainty avoidance, long term orientation and Indulgence. Finally, the data for the event study will be gathered with the use of Eventus.

After gathering the required data and applying the selection criteria the sample contained around 1,598 complete observations in the period of 2002-2016 of which 1,203 targets were in non-emerging countries and 395 in non-emerging countries. Further statistics are provided in table 1 and 2. Looking at table 2 it can be seen that South Korea has the most emerging market transactions and accounts for 42% of the total emerging market mergers. South Korea is then followed by India, China and Brazil.

(17)

17

Table 1 Descriptive statistics of the sample used

(1) (2) (3) (4) (5)

Variables N Mean St.Dev. Min Max

Deal Value (mil $) 1,541 406.5 917.0 10 9,583

Institutional Infrastructure 1,408 7.558 0.554 3.090 8.870

Firm Size (mil $) 1,598 8.706 1.696 4.607 14.58

Top Mediation 1,598 0.281 0.450 0 1

Governance Index 1,598 2.038 1.213 0 5

Cultural Distance Index 1,446 0.141 0.143 0.00289 0.844

Company Status 1,598 0.357 0.479 0 1

Debt-to-Assets Target 1,317 0.159 0.263 0 1.113

Investment Size (mil $) 1,541 0.0466 0.0568 0.00118 0.208

Table 2 Number of transactions per emerging market country

No. of Transactions Percentage of total emerging market mergers (%) Country Taiwan 13 3.2 Brazil 39 9.9 Chili 14 3.5 Colombia 2 0.51 Mexico 20 5.1 Peru 2 0.51 Czech Republic 4 1.1 Egypt Greece Hungary Poland Qatar Russian Federation South Africa Turkey UAE China India South Korea Malaysia Philippines Thailand 0 2 2 9 1 12 4 5 1 40 45 167 3 3 2 0 0.51 0.51 2.3 0.25 3.0 1.1 1.3 0.25 10.1 11.4 42.3 0.76 0.76 0.51 Total 395 100

(18)

18 Table 3 Correlation matrix regression variables

Size Top Med D/A

Invest Size Comp. Stat. Corp. Gov. Cult. Dist. Sim. Ind. Inst. Infra. Emer. Mkt. Size 1.00 Top Med. 0.28*** 1.00 Debt-to-Assets -0.16*** -0.18*** 1.00 Investment Size -0.45*** 0.09*** 0.09** 1.00 Company Status 0.23*** 0.24*** -0.21*** -0.06*** 1.00 Corporate Gov. 0.35*** -0.22*** 0.15*** 0.20*** -0.20*** 1.00 Cultural Distance 0.05** 0.04 -0.01* -0.12*** -0.10*** -0.02 1.00 Sim. Industry -0.10*** 0.26*** -0.07*** 0.09*** -0.21*** 0.01 0.15*** 1.00 Institutional Infra -0.02 -0.07*** 0.07 0.12*** 0.07*** -0.02 -0.68*** -0.17*** 1.00 Emerging Market 0.09*** 0.36*** -0.25*** 0.21*** 0.28*** -0.17*** 0.59*** 0.05*** -0.68*** 1.00 *** p<0.01, ** p<0.05, * p<0.1

Table 3 shows the correlation matrix between the main variables of the regression. The correlation matrix shows mostly significant correlations between the variables and does not show any perfect or near perfect relationships, however there are still some relationships that are noteworthy such as the relationship between size and investment size. Table 3 shows a correlation of -0.45 which indicates a semi-strong negative relationship between the two factors. In other words, the bigger a company the lower the investment size. Another noteworthy relationship is between cultural distance and an emerging market country which has a significant correlation of 0.59. This correlation shows that being in an emerging market country increases the cultural distance, therefore it is in line with the hypothesis of Aybar & Ficici (2009) who claim state the cultural distance between an emerging and emerged market will most likely be higher due differences cultural dimensions such as power distance, masculinity and long term orientation. The relationship between cultural distance and institutional infrastructure also proves to be strong with a negative correlation coefficient of -0.68 which implies that the lower the cultural distance between two countries the better the institutional infrastructure. Another insight is the correlation coefficient of an emerging market and the institutional infrastructure which is also -0.68 thus indicating that an emerging market country has a lower institutional

(19)

19 Table 4 Number of mergers and average transaction value for emerging and non-emerging M&As

Year Emerging Non-Emerging Total

Avg. Trans. Value

Market M&As Market M&As M&As

EMM* (in Mil $) NEMM** (in Mil $) 2002 10 96 106 85,34 235,09 2003 10 108 118 137 1334,53 2004 168 126 294 183,49 258,27 2005 45 105 150 200,75 278,14 2006 28 89 117 281,9 172,85 2007 14 94 108 129,99 345,79 2008 24 68 92 468,04 370,4 2009 18 113 131 177,04 286,44 2010 21 96 117 457,21 344,14 2011 16 74 90 59,52 648,2 2012 11 74 85 585,52 464,44 2013 12 38 50 483,6 388,94 2014 7 74 81 1266,56 832,97 2015 8 51 59 374,54 485,66

*EMM stands for Emerging Market Mergers, ** Non-Emerging Market Mergers

Table 4 shows the number of Emerging market M&As, non-emerging market M&As, total M&As and the average transaction value per year for emerging and non-emerging markets. When looking at the Total M&As it is seen that there is a peak in total M&As in 2004. When looking at the distribution between emerging market M&As and non-emerging market M&As in 2004. Even though emerging market M&As dominated in terms of total mergers table 4 also reveals that the average transaction value was higher for non-emerging market M&As in 2004. However, from 2010 it is noticed that emerging market M&As mostly dominated non-emerging market M&As in terms of deal value. Table 4 also shows that prior to the crisis cross-border M&As were more total mergers compared to post crisis. This result is in contrast with the findings of Grave et al. (2012) who finds that during the crisis many firms diversified by investing in foreign firms, implying that there should be an increase in total M&A. However, this is effect is not shown in table 4.

5. Results

5.1 Event study results

The first result to be discussed are the event study results. The results of table 5 and 6, show that announcements of both types of cross-border mergers are, on average, associated with positive abnormal returns. The results also show that most announcement period CAARs are significant at 1% level except for the non-emerging [-5, -1] event window. The negative abnormal returns could indicate

(20)

20 that investors do not perceive the merger as value increasing prior to the announcement (Aybar & Ficici, 2009), however since the result of the [-5, -1] window is not significantly different from zero no

conclusions can be drawn. Furthermore, tables 5 and 6 show that in the window [+1, +5] non-emerging market mergers earn higher abnormal returns compared to emerging market mergers. The difference in means for the [+1, +5] time window is tested and showed that non-emerging markets obtain higher CAARs, significant at 5% level. Thus implying that on average emerging market mergers earn higher abnormal returns.

The positive and higher CAARs for emerging market M&As is in line with similar event study results from Morck & Yeung (1992) and Doukas & Travlos (1988), who state that emerging market firms earn higher CAARs values overall. These results are also in line with Lowinski, Schiereck, & Thomas (2004) who find that on average cross-border M&As perform at least the same as domestic M&A. However, due to the fact that the sample of Lowinski, Schiereck, & Thomas (2004) only contained European firms they state that this effect is likely to be underestimated due to high capital market integration in Europe. Furthermore, these results are not in line with Kissin & Herrera (1990) and Click & Harrisonwho show that domestic M&As outperform cross-border mergers.

Even though the emerging markets on average earn higher abnormal returns for most event windows than non-emerging markets, the results could still indicate that the distribution between positive and negative results differ. For example, on average the return is higher for emerging firms is higher but it is possible that these differences arise due to differences in firm specifics, target market institutional infrastructure and other deal specific factors. These are possibilities that can affect investor reactions. The next section will explore theses option is by using the cross-sectional regression results.

Table 5 Announcement period CAAR for non-emerging markets

Event Window CAAR (%) Positive (%) Negative (%) t-value p-value

[+1, +5] 0.712*** 77 23 5.8741 0.0000

[-1, +1] 0.494*** 75 25 4.1416 0.0000

[-5, -1] -0.039 73 27 -0.3313 0.7405

[-5, +5] 0.817*** 76 24 4.1190 0.0000

Note: *** indicates that the coefficient is statistically different from 0 at 1% level

Table 6 Announcement period CAAR for emerging markets

Event Window CAAR (%) Positive (%) Negative (%) t-value p-value

[+1, +5] 0.55*** 72 28 3.3564 0.0009

[-1, +1] 0.51*** 68 32 4.0804 0.0001

[-5, -1] 0.75*** 66 34 4.3813 0.0000

[-5, +5] 1.65*** 70 30 6.3319 0.0000

(21)

21

5.2 Cross-sectional regression results

To explain cross-sectional variation in CAARs this paper will consider several factors such as firm size, debt-to-assets ratios, investment size and intuitional infrastructure. The results in table 7 show that firm size, defined as the total assets of the acquiring firm, has as positive effect on the CARs thus suggesting that the larger the acquiring firm the higher the abnormal returns. Furthermore, these results are only significant at 1% for the event windows [-1, 1], [-5, -1] and [-5, +5]. However, this result is in contrast with the literature. Moeller et al. (2003) show that the size of the acquirer is negatively related to the CARs as large firms tend to overpay in M&A transactions due to empire building or overconfidence. These results are also in contrast with Delios & Beamish (1999) who state that small firms are more cautious than big firms in terms of M&As therefore taking potential shareholder gains more seriously. This ultimately leads to higher shareholder value. However, the OLS regression results in table 7 results show the opposite.

In terms of mediation by a top 10 investment bank the results show a weak positive effect on shareholder wealth as the only significant result at 10 % level is for the [+1, +5] event window. This is in line with the results of Lowinski, Schiereck, & Thomas (2004) and Brouthers & Brouthers (2001) who conclude that top mediation has a positive effect on shareholder wealth as they can facilitate more complicated deals or negotiate better deals for the acquiring firms. However, these results are in contrast with the results form Srinivasan & Saunders (2001) and Rau (2000) who state that top mediation by an investmentbank has no significant on abnormal returns. This is also in contrast with McLaughlin (1992) who claims that top mediation has a negative effect on abnormal returns due to conflicts of interest between shareholders and mediators arising from contract incentive features. These results suggest that mediation by a top 10 ranked mediation bank offers benefits in terms of

shareholder gain after the announcement, however these results are limited to the defined event window [+1, +5] and not significant for other event windows.

The results from the debt-to-assets ratio of the target show that in the event windows [-5, -1] and [-5, +5], a higher debt-to-asset ratio leads to lower abnormal returns. This is in line with the hypothesis that a higher debt-to-asset ratio can imply a higher level of financial distress and thus a higher perceived risk. Thereby reducing abnormal returns. This result also implies that the hypothesis that a financially distress firm would lead to a discount and thereby increasing the abnormal returns, is incorrect as we find a negative coefficient. However, this could also mean that the discount effect could be superseded by the increased financial risk effect on abnormal returns. The results are also in line with

(22)

22 the results from Harford, Humphery-Jenner, & Powell (2012) who conclude that a higher leverage increases the financial risk for investors as well as decreasing the expected profitabilitiy and therefore lowering the abnormal returns.

Investment Size is related to the idea that firms can achieve economies of scale and scope by efficient use of the investment. Therefore, a higher investment size is expected to have a positive influence on shareholder value for the acquirer as the expected future cash flows would be higher (Lamacchia, 1997). The results shown in table 7 are not in line with the results from Aybar & Ficici (2009) & Lamacchia (1997) who find mostly significant positive results for company status. However, although the effect of Investment Size is positive the results are not significantly different from zero in all time windows. Therefore, no conclusions can be drawn from these results. This result is more in line with the results from Moeller & Schlingemann (2005) who find very weak evidence for a positive effect on shareholder value. The same holds for the Company Status of a firm. The results show a negative influence on abnormal returns which is consistent with the literature just as Aybar & Ficici (2009), Moeller & Schlingemann (2005) and Harford, Humphery-Jenner, & Powell (2012). However, none of the results are significant for all event windows. Thus implying that whether or not a firm is public or private does not have a significant effect on the abnormal returns.

The results of the Corporate governance show contradicting results as the effect on abnormal returns is negative in the [+1, +5] window but positive in the [-1, -5], significant at 5% and 10% level. The [-1, -5] result is in line with Aybar & Ficici (2009) who find that the higher the level of corporate

governance, the less incentives there are for managers to persue value decreasing activities. Thus implying a positive effect on abnormal returns. However, with these results it can only be concluded that prior to announcement date a higher corporate governance level seems to have a negative effect on abnormal returns but after announcement date the effect is positive.

When considering the Cultural Distance Index (CDI), which captures the cultural distance

between countries, the results show highly significant negative effects forthe [+1, +5] and [-5, +5] event windows thus implying that the higher the cultural distance the lower the abnormal returns for

shareholders. This is in line with the results of Brouthers & Brouthers (2001) and Morosini, Shane & Singh (1998) who show that the larger the cultural distance, the higher the level of friction leading to lower abnormal returns. Thus contradicting the results from Aybar & Ficici (2009) who find that there is no significant effect on abnormal returns.

Another possible determinant examined in this study is institutional infrastructure. The results show that institutional infrastructure has a negative effect on the abnormal returns for all time

(23)

23 windows. However, the results are only significant at 1% for the event windows [-1, -5] and [-5, +5]. This suggests that the more developed a countries’ institutional infrastructure, the lower the abnormal returns implying that a less developed institutional infrastructure with bad patent protection etc. is expected to earn higher abnormal returns. This is an unexpected result as these results do not

complement the results of Aybar & Ficici (2009) on the effect of the level of institutional development as in an institutionally underdeveloped country the acquirer faces a tradeoff between gaining from the M&A and the increased costs of uncertainty and bureaucracy. Implying that the higher the institutional development, the lower the higher the abnormal returns should be.

The similar industry dummy results, used as a control variable, shows positive results for the event windows [-1, +1] and [-5, +5] significant at 1% and 5% respectively. This is in line with the hypothesis that when firms operate in the same industry the acquirer can share its expertise with the target, possibly increasing expected future cash flow which could lead to higher abnormal returns for acquirers’ shareholders.

The final variable to examine is the Emerging market dummy, which equals 1 if a country is an emerging market and 0 otherwise. The results show negative effects for all event windows, which are significant at 1% level for [-1, 1], [-5, -1] and [-5, +5]. Implying that an emerging market M&A leads to value destruction compared to an emerged market M&A. These results speak against the results from the announcement period CAARS from table 5 and 6, in which the results show higher CAARs overall for emerging markets except for the [+1, +5] window. However, these results are in line with the results from Aybar & Ficici (2009), as their results also show negative returns for emerging market in Asia and South America. The results of table 5 and 6 can explain the surge to emerging markets in 2011, as the CAARs of emerging markets were on average higher than the non-emerging market M&As. However, the results of table 7 contradict this by showing that an emerging market M&A leads to lower abnormal returns compared to non-emerging M&As. This thus implies that the difference between emerging and non-emerging markets must lie in other factors.

(24)

24 Table 7 Cross sectional OLS regression results.

(1) (2) (3) (4)

Event Window Event Window Event Window Event Window Independent variables [+1, +5] [-1,1] [-5, -1] [-5, +5] Firm Size 0.000321 0.00743*** 0.00722*** 0.00760*** (0.00149) (0.00180) (0.00220) (0.00281) Top Mediation 0.00780* 0.00630 0.00616 0.00464 (0.00444) (0.00590) (0.00600) (0.00797) Debt-to-Assets -0.00630 0.00357 -0.0319*** -0.0322** (0.00652) (0.00801) (0.00999) (0.0151) Percentage of Shares owned -1.17e-06 7.01e-05 -6.28e-05 1.32e-05 (6.05e-05) (8.08e-05) (6.74e-05) (0.000119)

Investment Size 0.0427 0.0400 0.0438 0.0190 (0.0447) (0.0516) (0.0454) (0.0734) Company Status -0.00166 -0.00384 -0.00163 -0.00794 (0.00472) (0.00561) (0.00807) (0.00945) Governance Index -0.00415** 0.000740 0.00432* 0.000439 (0.00180) (0.00232) (0.00226) (0.00305) CDI -0.0762*** -0.0244 -0.0119 -0.0866*** (0.0170) (0.0182) (0.0195) (0.0255) Institutional Infrastructure -0.00749 -0.00866 -0.0282*** -0.0377*** (0.00482) (0.00854) (0.00642) (0.0111) Similar Industry -0.00202 0.0154*** 0.00293 0.0134* (0.00423) (0.00543) (0.00514) (0.00708) Emerging 0.00718 -0.0257*** -0.0608*** -0.0611*** (0.00635) (0.00808) (0.00806) (0.0105) Constant 0.0843** -0.0111 0.154*** 0.244*** (0.0372) (0.0609) (0.0542) (0.0864) Observations 419 419 419 419 R-squared 0.096 0.158 0.249 0.179

Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

The dependent variable in the regression are the daily Cumulative Abnormal Returns of the acquiring firm. Firm size is defined as the log of the acquirers’ Total assets, Debt-to-Assets as the debt over total assets from target. Investment size as Deal value over acquirers’ total assets. CDI is based on Hofstede’s cultural dimensions and Institutional Infrastructure is based on the World Economic Freedom index from the Fraser Institute.

5.3. Robustness Checks

To check for the robustness of the OLS regression results a binary logistic model was used in which the dependent variable is a dichotomous variable. This variable equals 1 if the CAR is positive and 0 otherwise. In this section the results from the OLS regression results from section 5.2 will be compared to the results of our robustness model.

The results of the cross-sectional OLS regression of table 7 show that firm size has a positive effect on shareholder value significant at 1% for the event windows [-1, +1], [-5, -1] and [-5, +5]. The binary logistic regression results in table 8 show similar results with significant results for the event

(25)

25 windows [+1, +5], [-5, -1] and [-5, +5]. For the event window [-1, +1] the results show weak evidence for a positive effect of firm size on CARs. Therefore, the results indicate that the bigger a firm is in terms of total assets, the bigger the effect on abnormal returns. This would also imply that multinational

companies would earn higher abnormal returns than a small national firm given the same deal, as stated in section 5.2.

The analysis of the OLS results on the effect of top mediation shows that there is weak evidence for a positive effect of top mediation on CARs, only in the [+1, +5] time window. Implying that top mediation leads to value creation. Just as the OLS regression, the logistic regression results show that top mediation has a positive effect on CARs, significant at 5% level. Thus confirming our table 7 results. Furthermore, this result is in line with the hypothesis that top mediation leads to higher abnormal returns as a top mediation bank can negotiate better deal or facilitate difficult mergers. A downfall is that this result is only significant for the event window [-1, +1].

When considering the effect of debt-to-asset ratio the results in table 7 show that a higher debt to asset ratio leads to lower abnormal returns for the event windows [-5, -1] and [-5, +5] at 1% and 5% level. Looking at the binary logistic regression results in table 8, the output shows similar results for the [-1, +1], [-5, -1] and [-5, +5] window significant at 1%, 5% and 10% respectively. The event window [-5, +5] shows a positive effect on the abnormal returns however this result is not significantly different from zero. The results of the binary logistic regression are thus in line with the expectation that a high debt to asset ratio, indicating financial distress of a firm, leads to a lower abnormal return due to the higher perceived risk.

Another factor examined is the percentage of ownership of the acquirer after the merger. The OLS regression results from table 7 do not show any significant effect of percentage of owner ship on abnormal returns. The opposite is true for the binary logistic regression results, these show highly significant results for the event window [-1, +1], [-5, -1] and [-5, +5]. However, the results show mixed evidence as the result for the [-1, +1] window is positive as opposed to the other windows. This would indicate that around announcement date a higher percentage of shares in the target is preferred. The positive effect is also what is expected as the higher the percentage of shares, the more control an acquirer has to steer the target in a better direction. However, with these mixed results it can only be stated that in the [-1. +1] window the effect is positive and in the [-5, -1] and [-5, +5] windows not. Therefore, this remains inconclusive.

Investment size is another factor examined in table 7, the results show that there are no significant results whatsoever. The binary logistics regression shows no significant results as well, thus

(26)

26 confirming our table 7 results that investment size does not explain variation in abnormal returns. This result is in contrast with our expectation that that investment size has a positive influence on abnormal returns.

The OLS results of company status show no significant results in all time windows indicating that whether or not a target firm is private or public does not have any effect on the abnormal returns of the acquirer. Even though in table 7 the results are insignificant, they do show the correct sign. When looking at the binary logistic regression results there is positive effect on abnormal returns, significant at 5%. The result is in contrast with the literature as it is expected that a public target firm would render a negative effect on abnormal returns as the acquirer loses its bargaining position and thereby overpaying compared to a private target firm.

The table 8 results of the effect of the level of corporate governance show no significant effects on the abnormal returns. When comparing this result to the table 7 results, the OLS estimates show he mixed results a positive effect for the [-5, -1] event window and a negative effect for the [+1, +5] event window significant at 5% and 10% respectively. The direction of the results match, however the table 8 returns are not significant for all event windows. Therefore, the results of table 8 binary logistic

regression results cannot give concluding evidence in explaining the mixed results of table 7 OLS regression results.

Another explanatory variable is the CDI index which is a measure for the cultural distance. The expectation is that the larger the index, implying a larger cultural distance, the more frictions a target will experience thereby reducing the overall shareholder gain. The table 7 results confirm the negative effect for the [+1, +5] and [-5, +5] event windows by rendering negative coefficients. The table 7 results are complemented by the binary logistic regression results as the results show corresponding signs and significance for the [+1, +5] and [-5, +5] event windows. Thus confirming that a higher cultural distance does have a negative influence on abnormal returns.

The effect of institutional infrastructure is expected to be positive as the higher the level of institutional infrastructure the less frictions the acquirer will face in the target country. The table 7 results show no signs of a positive effect as the results show significant negative coefficients in the [-5, +5] and [-5, -1] event windows. The binary logistic regression results confirm these results with negative for all event windows except the [-5, -1] event window. Therefore, the conclusion can be drawn that a higher level of institutional infrastructure does have a negative influence on abnormal returns. The table 7 results for the similar industry variable show a significant positive effect on abnormal returns when both the target and acquirer operate in the same industry. The results of the

(27)

27 logistic regression find weak evidence for a positive effect on abnormal returns only in the [-1, +1] event window. Thereby, weakly complementing the OLS regression results from table 7.

The final explanatory variable is the emerging market dummy which is the main variable of interest. The OLS regression results show highly significant negative effects on the abnormal returns in all event windows except [+1, +5]. The binary logistic regression results in table 8 also show highly significant and negative coefficients implying that emerging market mergers reduce abnormal returns and thus destroy value. Therefore, the conclusion can be drawn that an emerging market merger leads to value destruction, thus contradicting the hypothesis stated in section 2.

Table 8 Binary logistic regression results.

(1) (2) (3) (4)

Independent variables DCAR [+1, +5] DCAR [-1, +1] DCAR [-5, -1] DCAR [-5, +5] Firm Size 0.320*** 0.215* 0.280** 0.411*** (0.114) (0.111) (0.125) (0.108) Top Mediation -0.0640 0.872** 0.405 0.0555 (0.354) (0.346) (0.362) (0.327) Debt-to-Assets -0.982* -1.846*** -0.984** 0.00986 (0.540) (0.558) (0.442) (0.422)

Percentage of Shares owned 0.00965 0.0296*** -0.0246*** -0.0156*** (0.00641) (0.00538) (0.00597) (0.00572) Investment Size 0.296 -0.884 -0.881 2.971 (2.453) (2.515) (2.496) (2.428) Company Status 1.006** -0.573 0.0570 0.0853 (0.481) (0.619) (0.491) (0.497) Governance Index -0.212 -0.0567 0.213 0.161 (0.129) (0.150) (0.163) (0.131) CDI -9.658*** 0.920 0.532 -5.378** (1.701) (1.427) (1.735) (2.285) Institutional Infrastructure -1.479*** -1.342** -0.810 -1.423* (0.523) (0.533) (0.535) (0.791) Similar Industry -0.386 0.585* -0.263 0.0605 (0.312) (0.317) (0.328) (0.300) Emerging 0.992 -2.542*** -3.446*** -2.471*** (0.631) (0.594) (0.683) (0.813) Constant 9.353** 6.142 5.946 8.971 (3.777) (4.053) (4.418) (6.370)

Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

The dependent variable in the binary logistic regression is a dichotomous version of the Cumulative Abnormal Returns of the acquiring firm in which a positive CAR implies a value of 1 and 0 otherwise. Firm size is defined as the log of the acquirers’ Total assets, Debt-to-Assets as the debt over total assets from target. Investment size as Deal value over acquirers’ total assets. CDI is based on Hofstede’s cultural dimensions and Institutional Infrastructure is based on the World Economic Freedom index from the Fraser Institute.

(28)

28

6. Conclusion

In this paper the difference in effect on shareholder value of a cross-border merger in an

emerging market and a non-emerging market is examined during the 2002-2016 time frame. The sample contains 1,598 acquisition announcements of global mergers from which 395 are emerging market targets. The emerging market countries are selected on the basis of the MSCI Emerging market index and contains 23 emerging market countries.

The results of the event study on CAARs show that on average emerging market mergers result in higher CAARs in most time windows. The average CAAR across event window is 0.496 for emerged market and 0.865 for emerging markets. Furthermore, a higher announcement CAAR thus implies that on average investors believe that emerging market mergers lead to a higher value creation compared to non-emerging market mergers.

To investigate the abnormal returns and look for possible determinants, this study implements a cross-sectional analysis based on Aybar & Ficici (2009). The results show that firm size of the acquirer has a positive effect on shareholder value which is not in line with the results from Moeller,

Schlingemann, & Stulz (2003) who state that large firms tend to overpay in M&A transactions due to empire building and thereby decreasing value.

The results also show weak evidence for a positive effect on shareholder value from mediation by a top investment bank. This result complements the results from Lowinski, Schiereck, & Thomas (2004) who find that top intermediary investment banks can facilitate and negotiate better deals for acquiring firms leading to higher abnormal returns for the acquiring firm. However, this result is not in line with the results of (Srinivasan & Saunders, 2001) and Rau (2000) who find no effect on shareholder wealth as they find no direct relationship between acquiring firms and investment banks. Therefore, it can be stated that intermediation by a top investment bank increases shareholder returns.

When considering the cultural distance, Brouthers & Brouthers (2001), Morosini, Shane, & Singh, 1998 and Shenkar (2012) find that a higher cultural and geographical distance has a negative influence on shareholder wealth as there would be less similarities which can increase costs and frictions. Furthermore, the results of this paper show similar results with negative effects on abonormal returns. Implying that a larger cultural distance leads to lower abnormal returns for shareholders. In terms of institutional infrastructure the results show a negative effect on shareholder abnormal returns. Implying that a lower level of institutional infrastructure leads to higher abnormal returns. This result does not complement the results of Chari, Ouimet, & Tesar (2010) who find that a

Cross-border M&As : an analysis on shareholder returns for emerging and non-emerging market mergers