Cross-border M&As

(1)

Cross-Border M&As

Abstract

In this study I investigate whether firms from emerging countries are undervalued compared to firms from developed countries by analyzing the acquisition premium for domestic and cross-border mergers and acquisitions. I explore the theoretical motivations for a different acquisition premium and test whether this potential difference is statistically significant for the period 2000 to 2008. My results indicate that acquirers from a developed country pay significantly lower acquisition premiums for targets from emerging countries compared to targets from developed countries. Furthermore, I find that deals between firms that both reside in developed countries require significantly higher acquisition premiums than deals amongst firms from emerging countries.

Bachelor Thesis 6013B0326 BSc ECB Tim Bökenkamp 11018682 January 31st_{, 2018} Finance and Organization Faculty of Economics and Business

(2)

Originality Statement

I hereby declare that I am the sole author of this thesis and that the intellectual content belongs to my own research. I certify, to my best knowledge, that this thesis does not fringe upon fraud and inherently does not violate any copyrights or proprietary rights. All sources have been properly acknowledged and are contained in the bibliography at the end of the thesis. No information, beyond my own knowledge, has been used that is not stated in the reference list which uses reference and quotation techniques comprised by APA.

Furthermore, I acknowledge that this is a true copy of my thesis which has been submitted through the University of Amsterdam. No other institution preserves the rights to use or copy my research other than the University of Amsterdam, including proper acknowledgments.

(3)

1. Introduction

An important issue in the study of mergers and acquisitions is to explain the drivers of the size of the acquisition premium (Walking & Edmister, 1985; Varaiya, 1987). The success of an M&A deal does not only depend upon the ability of the acquiring entity to successfully integrate the target into their operations, but also on the ability of the acquirer to effectuate the deal at a price that does not cover all the synergy value. Ideally, an acquirer must offer a price that is below their reservation price and above the price at which the target is willing to accept the offer (Flanagan & O’Shaughnessy, 2001). Prior research has indicated that foreign acquirers offer a higher price for targets than domestic acquirers (Moeller & Schlingemann, 2004). Rossi & Volpin (2004) demonstrate that the acquisition premium for cross-border mergers and acquisitions is higher than the acquisition premium for domestic mergers and acquisitions. Yet, as opposed to Moeller and Schlingemann (2004) and Rossi & Volpin (2004), Francis, Hasan and Sun (2007) focus on the differential impact of a distinction between emerging and developed countries rather than foreign and domestic acquirers. They contend that value is created by combining firms with a different financial integration status, in which cheaper external funds are provided to the firm with the highest cost of capital. Focusing on the extent to which deals between firms from emerging and developed countries can be value enhancing, I examine the extent to which these country characteristics affect the size of the acquisition premium

Emerging countries are stipulated by the level of integration into the global capital, labor and product markets. A typical emerging country is still undergoing industrialization, has a higher economic growth rate than developed countries and has financial markets that induce greater risk (Bodie, Kane & Marcus, 2014). On the contrary, developed countries are integrated and established nations that drive international business. Alone, the developed countries accounted for 68% of world GDP and 85% of world market capitalization in 2010 (Bodie, et al., 2014). Furthermore, more than 80% of all global M&A activity is accounted for by firms from developed countries. Yet, since the early 2000s, globalization and integration of financial markets increased the number of firms from emerging countries that are acquired or are merged with (Francis et al., 2007). Acquirers from developed countries are able to utilize the market imperfections by providing capital at a lower cost, which allows targets from emerging markets to pursue positive NPV-projects. Nevertheless, firms from emerging countries have restricted bargaining power as opposed to firms from developed countries, because emerging countries have less developed capital markets and accounting standards to rectify the value of targets and demand a higher premium. Moreover, the low domestic cost of capital that emerges in

(5)

developed countries reinforces the bargaining position for acquirers, because targets from emerging countries are designated to this cheaper source of funding to abandon their financing constraints on positive NPV-projects (Francis et al., 2007). As a result, I examine whether firms from emerging countries are undervalued in M&A transactions compared to firms from developed countries, by analyzing the acquisition premiums.

In this paper I provide support for the contention that targets from emerging countries receive on average significantly lower acquisition premiums than targets from developed countries if they are acquired by a company from a developed country. I find that the presence of targets from developed countries increase the acquisition premium that is paid to finalize the deal, because firms from developed countries have a better bargaining position to demand a higher premium. Furthermore, shareholders receive greater protection in nations that are well integrated into the global financial market, which positively affects the magnitude of the acquisition premium (Rossi & Volpin, 2004). Moreover, I find that deals amongst firms that both reside in developed countries pay a 27.8% higher premium compared to deals amongst firms in emerging countries. Also, consistent with preliminary research, cross-border M&As are expected to increase the acquisition premium compared to domestic M&As.

Given the relatively recent development of M&A activity in emerging countries, this study contributes to existing M&A literature that primarily distinguishes between cross-border and domestic mergers and acquisitions. Few research papers have explored the difference between cross-border M&As and emerging and developed countries. My paper makes several contributions to the M&A literature. First, it enhances the understanding of the different characteristics that coincide between developed and emerging countries. Second, it amplifies that a distinction between emerging and developed countries is significant to determine that firms from emerging countries receive, on average, a lower acquisition premium. Consequently, shareholders of the acquirer distribute less wealth to the shareholders of the target compared to deals with a target from a developed country. Finally, it contributes to extent that worldwide mergers and acquisitions are dominated by firms from developed countries and that the purported benefits of integration by emerging countries have not yet occurred.

This paper is presented in six parts. The first part provides background information on mergers and acquisitions and discusses general trends as well as the acquisition premium. The second part is devoted to define my hypotheses and motivates why I expect targets from emerging countries to be undervalued. Next, the third part presents the methodology that is used to test my hypotheses followed by the fourth part that describes my sample. Finally, the fifth part discusses the outcome of my regression analysis, which is concluded in the sixth section.

(6)

2. Literature Review

2.1 Background Information

Mergers and acquisitions (M&As) play an important role in corporate finance and investments, because M&As significantly change the structure and organization of the combined entities and are a form of external investment (Copeland, Weston & Shastri, 2015). In M&As, either another corporation or a group of individuals acquires the target firm or the target firm mergers with the acquirer. Both cases represent a change in ownership and control where the acquirer must purchase stock or existing assets of the target by offering either cash or something of equivalent value – such as an exchange of shares (Berk & DeMarzo, 2014).

Since the 1980s and 1990s, M&As have been an active mechanism that has shaped industries and economic growth (Liu & Qiu, 2013). Acquirers engage in M&A transactions, because they believe that they might be able to add economic value that cannot be added by both entities separately. Therefore, the most common justification to acquire a target is the possibility to create synergies (Copeland et al., 2015). There is extensive literature on the rationales for M&A activity and ultimately on the drivers of synergistic value. Typically, synergies translate into cost reductions and revenue enhancement, which subsequently enhances the value of a combined corporation. Synergy value is created by a variety of mechanisms including economies of scale, expertise, vertical integration and market expansion (Flanagan & O’Shaughnessy, 2001). Firstly, economies of scale are used to justify a corporate takeover, because a larger company can enjoy savings from producing in high volume by reducing the costs per unit (Copeland et al, 2014). Secondly, M&As contribute to enhance value from areas where the acquiring entity lacks expertise to compete. By acquiring these resources as an already functioning unit, the firm will be able to reinforce expertise and gain skills to empower market share and ability to compete more effectively (Berk & DeMarzo, 2014). Thirdly, vertical integration refers to a merger or acquisition of two companies that compete in the same industry but make products that are required at different stages of the production cycle. Flanagan and O’Shaughnessy (2001) emphasize in their research that core-related mergers and acquisitions have greater potential to create value through synergies than a combination of firms that are totally unrelated or do not share the area of commonality between the firm’s primary lines of business. As a result, M&As are often justified by vertical integration, because an acquirer can enhance the value of its operations if it has direct control of the inputs for the production process. Finally, M&As provide the possibility to expand in new markets by acquiring a target that has local market knowledge and operations in the market (Danbolt & Maciver, 2012). Deals amongst these firms are likely to create synergistic value, because the

(7)

operational channels of the target provide access to new capital and product markets which inherently accommodates an expansion of product demand and increases in revenue. Altogether, M&As are justified by the ability to create value from combining operations which improves the competitive position of the combined firm and thereby increases long-term financial performance.

Nevertheless, M&A activity is highly cyclical and depends whether the economy is in an expansion or in a recession (Berk & DeMarzo, 2014). This behavior is characterized by merger waves, which entails peaks of heavy M&A activity followed by periods with few transactions. The extent to which an acquirer is able to create synergistic value from a deal is, therefore, triggered by market shocks, which is correlated among countries (Madura, Ngo & Viale, 2011). As a result, they find the general trend that takeover activity is clustered in waves and evolves with environmental changes as well as industrial changes. The early waves are characterized by firms that seek to alter their competitive positions within their country, while contemporary M&A activity is influenced by globalization and integration of product, labor and capital markets among countries (Francis, Hasan & Sun, 2007).

2.2 Acquisition Premium

When one firm acquirers or mergers with another firm, the acquiring entity typically offers to purchase the target’s stock at a premium, which is defined as the offer price in excess of the market price (Nathan & O’Keefe, 1989). One reason is that target’s shareholders will never sell their shares below the market price and only accept an offer that is at least equal to the target’s pre-bid market capitalization plus a premium (Berk & DeMarzo, 2014)). Another reason comes from Roll (1986) who studies the Hubris hypothesis, which holds the overbearing presumption that the acquiring entity believes that their valuation is superior to that of the market and therefore pays too much for the target. The Hubris hypothesis maintains the assumption that financial, product and labor markets are efficient and asserts that all available information about a firm is incorporated in the market price. Consequently, overconfident CEOs pursue deals that have a low probability of creating value, but are completed because they believe that the full economic potential of a combined firm is greater than predictions by the market. Support for this explanation comes from Hayward and Hambrick (1997) who find that CEO hubris increases the acquisition premium, because they are overconfident about their abilities to create value by combining operations. Yet, markets do not recapitulate all relevant information about a potential M&A deal, because of the existence of asymmetric information (Copeland et al., 2015). This contradicts with the hubris assumption that financial, product and labor markets are

(8)

efficient. Therefore, another reason why firms often pay a premium is because they believe that they can create synergy value. Acquisition premiums represent the expected benefits or synergies that the acquiring entity expects to obtain from undertaking the deal (Sonenshine & Reynolds, 2013). As a result, premiums are thereby a proxy for synergies, which is why synergies are the most common justification to pay an acquisition premium and proceed with a takeover (Berk & DeMarzo, 2014). As a result, the deal is only a positive NPV-project if the synergies exceed the premium that is paid for the target firm, where the valuation of the target incorporates these expected future synergy advantages and reimburses a premium that, based on these expectations, creates a positive return.

Acquisition premiums in mergers and acquisitions receive a considered amount of attention in corporate finance and investments since the 1970s, partially because average deal premiums rose significantly in the 1970s and 1980s. (Nathan & O’Keefe, 1989). This fueled an ongoing interest to determine the drivers of this acquisition premium. Walking and Edmister (1985) were one of the first to study these drivers, who hypothesized and showed that the distribution of the acquisition premium depends on a negative function of acquirer’s bargaining power and a positive function of acquisition-related benefits. They also present evidence that firms with lower leverage and a lower market-to-book ratio demand a higher premium. Along this evidence, drivers such as the size of the target and cross-border deals also have a significant effect on the size of the premium (Rossi & Volpin, 2004). Evidence displays the existence of the “border effect”, which entails significantly lower returns for firms that acquire cross-border targets rather than domestic targets (Moeller & Schlingemann, 2004). Rossi & Volpin (2004) show that these lower bidder returns for cross-border M&As are the result of higher acquisition premiums to convince equity holders to give up their shares. Acquirers choose to engage in cross-border M&As, because they perceive improved market access, cost advantages (lower labor costs and economies of scale) and differences in cost of capital (Sonenshine & Reynolds, 2013). As a result, acquirers would be willing to pay a higher acquisition premium to acquire corporate control of target’s assets as opposed to domestic M&As. This strong relationship between the acquisition premium and the cross-border exemplification has increased research on acquisition premiums, because cross-border M&As have grown rapidly over the last decade.

During the 1990s and 2000s the volume of cross-border mergers and acquisitions increased significantly. There are several explanations why the number of cross-border M&As has gone up. On the one hand firms have gone away from traditional Greenway investments (Francis, Hasan & Sun, 2007). On the other hand, integration between international financial markets

(9)

enforced more absolute and relative increases in both domestic and cross-border mergers and acquisitions (Moeller & Schlingemann, 2004). Globalization has resulted in more integrated capital, product and labor markets which disposed uncertainties and created a new opportunity set to increase the likelihood of realizing synergies and efficiency gains (Sonenshine & Reynolds, 2013). A new feature of this increase in M&As has been the significant number of transactions involving firms from emerging markets (Francis et al., 2007). The research by Moeller and Schlingemann (2004) is based on a time that imposed many restrictions on firms in emerging countries that abandoned them from engaging in international business. In the late 1980s and early 1990s many emerging countries reformed and liberalized their financial markets. Integration by these countries in the global economy allowed foreign firms to acquire domestic firms and realize a positive return. Despite liberalization, these emerging markets still faced a higher cost of capital compared to developed countries, which induced more risk and financial constraints on the targets firms (Errunza & Miller, 2000). Another reason for the increase in volume of cross-border M&As is therefore to provide funding to these financially constrained firms, because the higher costs of capital restrained them from investment (Francis et al., 2007). An important benefit of financial integration is the cheaper access to this external capital via the M&A mechanism. Target firms then gain the possibility to extract capital from these integrated-markets, which enable them overcome financial constraints and allow them to pursue positive NPV projects. This effect is stipulated as the wealth effect, which is extensively described in research by Francis, Hasan and Sun (2007) and Erel, Liao and Weisbach (2012). The wealth effect described by Moeller and Schlingemann (2004), therefore, changed after the liberalization of firms from emerging countries, in the sense that cross-border M&As are value enhancing for acquirers, with most of this gain coming from the emerging economies.

2.3 Cross-border M&As Versus Domestic M&As

Conceptually, cross-border mergers and acquisitions involve the same justification as domestic ones – namely the synergistic value that concurs a combined enterprise (Erel, Liao & Weisbach, 2012). Nevertheless, a possible reason to go cross-border is the benefit from international diversification and the access to new markets (Danbolt & Maciver, 2012). In addition, foreign countries may have the advantage of lower labor costs and taxes, or beneficial corporate and regulatory jurisdiction (Sonenshine & Reynolds, 2013). As a result, acquirers are on average willing to offer a higher acquisition premium for a foreign target, because corporate control of the target’s assets flourishes additional benefits beyond the scope of domestic M&As.

(10)

However, national borders add extra dimensions to the calculus that can obstruct or accommodate a M&A deal.

In the research by Erel, Liao and Weisbach (2012), they find evidence for international factors that affect cross-border M&As that are not necessarily present in domestic M&As, such as cultural and geographical differences. Both factors contribute to the likelihood of a cross-border deal and influence the extent to which two firms decide to merge. Consequently, the shorter the distance between two countries the greater the odds of a successful acquisition relative to targets that are far away. In addition, countries that trade more commonly with one another are more likely to share cultural similarities, which increases the likelihood of synergy gains and subsequently the likelihood that an acquisition or merger occurs. Also, the likelihood increases the better the accounting standards and accounting quality in the acquiring country. These findings are consistent with the results from Rossi and Volpin (2004), who show that the volume of M&A activity is higher the better the accounting standards. Also, they demonstrate that the acquisition premium is positively related to the accounting quality. Differences in country development, therefore, is a main driver to the extent it impedes or facilitates cross-border mergers and acquisitions (Liu & Qiu, 2013).

Compared to domestic deals, cross-border mergers and acquisitions illustrate smaller deal values, higher market-to-book values and higher likelihood of all cash payments (Moeller & Schlingemann, 2004). The premium that is offered to acquire a target is conditional on these deal-specific factors (Dimopoulos & Saccetto, 2011). Furthermore, it is argued that acquirers pay more for cross-border targets as opposed to domestic targets, because there is a large benefit to acquire control over another firm’s assets, especially when it is more difficult to write and enforce contracts (Sonenshine & Reynolds, 2013). Cross-border mergers and acquisitions are regarded more complex to write and enforce contracts, due to cultural differences that impede the integration process of the target into the acquiring entity, but more importantly due to currency movements and relative stock market performance (Danbolt & Maciver, 2012). Currency appreciation and relative stock market performance between two countries appear to increase the attractiveness of cross-border mergers and acquisitions compared to domestic ones (Erel et al., 2012). These country-factor effects on merger propensity are indicative for the fact that higher-valued firms tend to purchase lower-valued firms. Higher-valued firms have a tendency to acquire firms from poorer countries, because of the appreciation of the currency and a wealth effect due to the lower cost of capital (Erel et al, 2012). Both effects contribute to the extent that acquirers pursue cross-border deals and pay higher acquisition premiums for them. However, much research has been focused on the additional benefits from cross-border

(11)

M&As, and whether cross-border mergers and acquisitions change the size of the acquisition premium. Despite, few research has been devoted distinguishing between the effect of developed countries and emerging countries on the acquisition premium.

As already indicated by Francis, Hasan and Sun (2007), barriers to international investment have fallen for many emerging countries that allowed firms to conduct business across national borders. Yet, mentioned by Errunza and Miller (2000) and Erel, Liao and Weisback (2012), liberalization into the global financial market is a slow process which initially does not improve the cost of capital. Foreign firms are able to benefit from these market imperfections by utilizing their lower cost of capital. Deals that entail targets from emerging countries can indeed be more value enhancing compared to deals from developed countries, because developed acquirers are able to provide external capital at a lower costs which allows targets to pursue positive NPV-projects and create synergistic value for the M&A deal. (Francis et al., 2007). However, these alluring investment opportunities do not necessarily generalize into higher acquisition premiums, which is done for cross-border deals. This could have a variety of reasons, including accounting standards, bargaining power, volatile exchange rates, and competition. First, emerging countries are less integrated, have less developed capital markets, less developed accounting systems and more volatile exchange rates, which obstruct the valuation of the target and increase the risk that the acquiring entity cannot attain synergy

value (Danbolt & Maciver, 2012). Second, integration is aligned with an increase in

competition, which increases the level of bargaining power to bargain for a higher premium. To the extent that emerging countries are by definition not well integrated, they have severely less bargaining power than developed countries (Moeller & Schlingemann, 2004). Finally, diffused ownership reinforces the free-rider problem by forcing acquirers to pay a higher acquisition than otherwise. Since firms from emerging countries are less likely to have diffused ownership, due to a less developed stock market, their shareholder are less likely to be able to bargain for a higher premium (Rossi & Volpin, 2004). Consequently, by distinguishing between developed and emerging countries, more can be said about the size of the premium and whether this premium is lower for targets in emerging countries as opposed to developed countries.

3 Hypothesis

Due to increased globalization and better regulation, cross-border M&As have substantially increased throughout the world (Danbolt & Maciver, 2012). Among the increase in volume of cross-border M&As has been the increasing amount of activity that occurs in emerging markets (Sonenshine & Reynolds, 2013). Despite the increased integration by many

(12)

emerging countries, countries that lack this development are hit by monetary constraints that discourages firms to invest (Francis et al, 2007). Francis, Hasan and Sun show that targets from emerging markets have a higher beta compared to targets from developed countries. Since a firm’s cost of capital is captured by its beta, these firms do indeed have greater financial constraints relative to firms from developed countries (Copeland, et al., 2014). Ultimately, the M&A mechanism is a powerful tool to diminish these market imperfections by allowing targets from emerging countries to take advantage of the integrated capital markets and pursue positive NPV-projects (Francis et al., 2007). As a result, acquirers from developed countries benefit from these market imperfections by a wealth effect, because a M&A deal with a target from an emerging country is able to create significant value due to the lower cost of capital (Erel et al, 2012). Liu and Qiu (2013) provide evidence that the level of development in a country is indeed significant to the extent that cross-border M&As create more value compared to domestic M&As.

In cross-border deals that comprise acquirers and targets from developed and emerging markets, the acquirer is most likely to be from a developed country and the target from an emerging country. The acquirers are most likely to be from developed countries, because their country development allows them to provide capital at lower cost and thereby finance internal projects that drive synergy value (Erel et al, 2012). Considering that access to emerging markets is highly valuable for an acquirer from a developed country, the target is able to extract a higher acquisition premium compared to deals with an acquirer from an emerging country (Doukas & Travlos, 1988). Deals comprised between an acquirer from a developed country and a target from an emerging country are therefore expected to be characterized by higher acquisition premiums than deals amongst firms from emerging countries. This could have a number of explanations.

Rossi and Volpin (2004) illustrate that the degree of shareholder protection positively relates to an increase in the premium. Shareholder protection reduces the cost of capital, resulting from a decrease in risk, which increases the attractiveness of an investment and ultimately the premium that is paid (Copeland et al., 2014). Rossi and Volpin (2004) provide the notion that the level of shareholder protection increases with the development level in a country, implying that less developed countries have lower standards for shareholder protection and thus that firms from emerging countries suffice with lower acquisition premiums. Furthermore, emerging countries have the disadvantage of less diffused ownership and less competition, which are factors that are proven to be positively related to the size of the acquisition premium (Walking and Edmister, 1985; Rossi & Volpin, 2004). Moreover, their

(13)

capital markets are dictated by more volatile exchange rates, which increases the risk to successfully accommodate the M&A deal (Erel et al., 2012). Consequently, due to these differences in capital markets and accounting standards, firms from emerging markets are more difficult to value, which allows the acquirers of these firms to make more aggressive bids than domestic rival bidders – lower acquisition premiums (Erel et al., 2012). Though, targets from emerging countries are able to demand a higher premium if they are acquired by an acquirer from a developed country compared to an emerging country, because shareholders demand compensation for the synergy gain from the takeover, due to the lower cost of capital (Sonenshine & Reynolds, 2013). Deals amongst firms from emerging countries have a lower probability to create value, because the acquirer cannot provide the target with cheap external funding. Firms from developed countries, in fact, are able to mitigate financial constraints as opposed to firms from emerging countries, which increases the potential that deals amongst these firms create more synergy value (Doukas & Travlos, 1988). Furthermore, acquirers from developed countries may overestimate the synergy value, which could lead to an increase in hubris and agency problems, which increases the acquisition premium as a result of overconfidence (Roll, 1986). By taking into account all these factors, I created the following hypotheses:

Hypothesis 1: Deals between targets from emerging countries and acquirers from

developed countries pay a higher acquisition premium than deals with both a target and an acquirer from emerging countries

Developed countries are highly integrated because their financial, product and labor markets are set up in a way that induces firms to pursue valuable projects, protect the rights and obligations of debt- and shareholders and increase the size/value of the stock market (Moeller & Schlingemann, 2004). This high degree of integration reduces the cost of capital, which subsequently makes developed countries gain from less developed countries due to the wealth effect (Erel et al, 2012). Despite, deals between firms from developed countries are expected to be characterized by higher acquisition premiums compared to deals amongst firms from emerging countries for the following reasons. Firstly, participants from highly integrated markets have a strong bargaining positions that amplifies the amount of the premium (Moeller & Schlingemann). Secondly, firms in developed countries are more likely to have diffused ownership due to a more developed stock market. Diffused ownership reinforces the free-rider problem by forcing acquirers to pay a higher acquisition premium than otherwise (Rossi &

(14)

Volpin, 2004). Thirdly, integrated capital markets have the disadvantage of increasing the level of competition for a corporate takeover, that induces the eventual acquirer to pay a higher premium (Moeller & Schlingemann, 2004). Fourthly, shareholders from firms in developed countries are protected by jurisdiction that requires when existing shareholders of a target firm are forced to sell their shares, they receive a fair value for their shares. As a result, targets from developed countries are able to bargain for a higher acquisition premium in contrast to targets from emerging countries that lack the jurisdiction to protect shareholders (Danbolt & Maciver, 2012). Finally, premiums rise as a result of agency problems and hubris (Hayward & Hambrick, 1997). Diffused ownership imposes few restrictions on the CEO to acquire another company. Managers may prefer to run a larger company due to the extra personal benefits, because they only bears a small fraction of the costs if the merger is unsuccessful, due to their relative small ownership share (Berk & DeMarzo, 2014). Moreover, as a result of overconfidence, acquirers may pay too much for a target, because they truly believe that they can manage to enhance much value (Roll, 1986). Consequently, I expect based on these motivations that deals amongst firms from developed countries have higher acquisition premiums compared to deals with firms from emerging countries. Furthermore, focusing on the extent that emerging markets aren’t

fully integrated and have severely less bargaining power than developed countries, I expect to

find that deals between an acquirer from a developed country and a target from an emerging country have significantly lower acquisition premiums than deals amongst firms from developed countries. This let me to create the following hypothesis:

Hypothesis 2: Deals amongst targets and acquirers that both reside in developed countries

pay a higher acquisition premium than deals with both a target and an acquirer from emerging countries

Hypothesis 3: The acquisition premiums for targets from emerging countries paid by

acquirers from developed countries is significantly lower than the acquisition paid for a target from a developed country

4 Methodology

4.1 Econometric Method

To test my hypotheses, I run the following cross-sectional regression analysis on a sample of acquisition premiums using Ordinary Least Squares (OLS):

(15)

𝐴𝐶𝑄𝑃𝑅 = 𝛽)+ 𝛽+𝐶𝑅𝑂𝑆𝑆𝐵𝑂𝑅𝐷𝐸𝑅 + 𝛽1(𝐷𝐸𝑉 → 𝐸𝑀𝐸) + 𝛽7(𝐸𝑀𝐸 → 𝐷𝐸𝑉) + 𝛽8(𝐷𝐸𝑉 → 𝐷𝐸𝑉)

+ 𝛽9𝐶𝑅𝑂𝑆𝑆𝐵𝑂𝑅𝐷𝐸𝑅 ∗ (𝐷𝐸𝑉 → 𝐷𝐸𝑉) + 𝛽;𝑆𝐼𝑍𝐸 + 𝛽>𝑀𝐴𝑅𝐾𝐸𝑇𝐵𝑂𝑂𝐾

+ 𝛽A𝐷𝑈𝑅𝐴𝑇𝐼𝑂𝑁 + 𝛽D𝐻𝑂𝑆𝑇𝐼𝐿𝐸 + 𝛽+)𝑅𝑂𝐴 + 𝛽++𝐷𝐸𝐴𝐿𝑉𝐴𝐿 + 𝛽+1 𝐶𝐴𝑆𝐻

+ 𝛽+7 𝐿𝐸𝑉𝐸𝑅𝐴𝐺𝐸 + 𝜀

I use the method of Ordinary Least Squares to estimate my regression coefficients, because many studies on the acquisition premium performed a multiple linear regression analysis (Walking & Edmister, 1985; Flanagan & O’Shaughnessy, 2001; Dimopoulos & Sacchetto, 2011; Madura et al., 2011; Sonenshine & Reynolds, 2013).

The dependent variable is the acquisition premium, ACQPR, which is defined as the percentage difference by which the deal value exceeds the pre-bid market capitalization (Flanagan & O’Shaughnessy, 2001). The premium, therefore, incorporates the market capitalization of the target firm prior to the M&A announcement. It is important to determine an appropriate time span to calculate the market capitalization to avoid run-up effects associated with M&A rumors. The run-up effect comprises that rumors inflate pre-announcement stock prices, which adjusts the target market capitalization to a degree in which it diverges from its true value. To limit effects of pre-announcement rumors, I calculate the deal premium with the target’s market capitalization 28 days prior to the announcement (ACQPR28) and 42 days prior to the announcement (ACQPR42) (Dimopoulos & Sacchetto, 2011; Sonenshine & Reynolds, 2013). As a result, the following mathematical definition of the acquisition premiums applies to my study:

𝐴𝐶𝑄𝑃𝑅28 =𝐷𝐸𝐴𝐿𝑉𝐴𝐿 − 𝑆𝐼𝑍𝐸_{𝑆𝐼𝑍𝐸} 1A LMNO

1A LMNO

where ACQPR28 represent the acquisition premium, DEALVAL the price that is paid to acquire the target and SIZE the pre-bid market capitalization of the target calculated on the basis of a 28 days’ time span.

The independent variables are a combination of control variables and research specific variables. The control variables are included because they are highlighted in previous literature about the acquisition premium. This study includes two types of control variables, namely target-specific variables (SIZE, MARKETBOOK, ROA and LEVERAGE) and deal-specific variables (DURATION, HOSTILE, DEALVAL and CASH). Both set of variables have been acknowledged by the majority of studies to the extent that target-specific characteristics (Walking & Edmister, 1985; Flanagan & O’Shaughnessy, 2001) and deal-specific characteristics (Hsieh & Walking, 2005; Dimopoulos & Sacchetto, 2011) affect the size of the acquisition premium. Along with these control variables, the research specific variables are included to find evidence for my hypotheses.

(16)

4.1.1 Control Variables

SIZE = The size of the target firm, which is measured as the natural logarithm of the target’s market capitalization 42 days prior to the announcement (Dimopoulos & Sacchetto, 2011). The natural logarithm transposes the distribution of the market capitalization to a normal distribution, and reinforces that extreme values do not influence the sample.

MARKETBOOK = The target’s market-to-book ratio, which is defined as the market value of equity divided by the book value of equity (Walking & Edmister, 1985). For the market-to-book ratio, data is taken 42 days prior to the announcement, which is in line with the calculation of SIZE and supported by Dimopoulos & Sacchetto (2011).

ROA = The target’s return on assets and is calculated as the fraction of net income over total assets computed with data from the fiscal year prior to the announcement of the deal (Flanagan & O’Shaughnessy, 2001; Madura et al., 2011). I use net income one year prior to the announcement, instead of 42 days, because a fiscal year’s net income is contained in a firm’s annual report and can therefore not be obtained exactly 42 days prior to the announcement

LEVERAGE = The fraction of the target’s net debt over total assets, computed with data from the fiscal year prior to the announcement of the deal. By taking a fraction, I make sure that the variable is scaled and restricted from extreme values.

DURATION = The duration of the M&A deal, which is controlled for in my regression model as the difference between the completion date and the announcement date, measured in days (Hsieh & Walking, 2005)

HOSTILE = A dummy variable that equals 1 if the deals is hostile and 0 otherwise (Rossi & Volpin, 2004)

DEALVAL = The natural logarithm of the value of the deal, measured in thousands of US dollars. Just like SIZE, deal value is transposed via a natural logarithm to avoid outliers and reinforce a normal distribution (Sonenshine & Reynolds, 2013). The amount that an acquirer pays to attain a target in countries that use other currencies than US dollars are converted to the US currency by using daily exchange rates.

CASH = A dummy variable that equals 1 if the acquirer pays for the target solely with cash and 0 if the acquirer pays with stock or a mixture between cash and stock (Sullivan et al., 1994; Madura et al., 2011).

I Include SIZE, because evidence displays that the size of the target seems to have a negative relationship with the size of the acquisition premium, implying that acquirers typically pay a higher acquisition premium for “smaller” firms compared to “larger” firms (Dimopoulos & Sacchetto, 2011). Sonenshine and Reynolds (2013) confirm this negative relationship in their

(17)

research by explaining that the smaller the size of the target, the greater the difference in size between the acquirer and the target, which attenuates the possibility of asset duplication

MARKETBOOK is included, because larger acquisition premiums are associated with lower market-to-book ratios (Walking & Edmister, 1985). Support for this comes from Dimopoulos and Sacchetto (2011) and Berk and DeMarzo (2014) who explain that firms with lower market-to-book ratios have market value of equity relatively close to their book value, and thus a market value that is closely aligned with the current value of the firm. On the contrary, high market-to-book ratios are based on anticipated growth by the market which does not exactly match the current value of the firm. As a result, acquirers pay a higher premium for firms with a low market-to-book ratio, because the target transmits less risk to the acquirer.

As an indicator for the financial performance of the target, the return on assets (ROA) is included. ROA seems to be positively related to the size of the acquisition premium, because it signals the target’s ability to generate profits i.e., the ability to generate cash flows (Flanagan & O’Shaughnessy, 2001). As stated by Walking and Edmister (1985), most of the premium can be tied into the performance of the target, which rectifies the positive relationship. Yet, it could also be argued that it is negatively related to the magnitude of the acquisition premium (Dimopoulos & Sacchetto, 2011). They find a significant negative coefficient for ROA and explain that this could be due that the acquirer will be able to extract synergies by an alternative deployment of assets, i.e., utilize the assets differently post-merger than pre-merger.

Next, LEVERAGE is accounted for in the regression model, because Walking and Edmister (1985) contend that acquirers pay on average greater premiums with declining levels of leverage. Acquirers are willing to pay more for targets with lower levels of debt, because the target’s debt liabilities transfer to their own balance sheet upon completion of the deal (Madura, Ngo & Viale, 2011). However, Flanagan and O’Shaughnessy (2001) find in their research a positive coefficient for leverage along all their regressions. This could be because greater levels of debt provide, at certain levels, the advantage of a tax shield that decreases the amount of corporate taxes for the acquirer (Erel et al, 2012). If acquirers value the advantage of the tax shield, this could explain the positive relation between leverage and the acquisition premium.

The deal-specific variable DEALVAL is included, because evidence displays that deal value negatively affects the size of the acquisition premium, implying that when the offer price increases, the premium decreases (Hsieh & Walking, 2005). Sonenshine and Reynolds (2013), support this in their research and find that the acquisition premium falls with the amount of the deal; for every 1% increase in the transaction size, the deal premium falls by 0.02%. They

(18)

motivate this negative relationship, because smaller deals attenuate the possibility of asset duplication.

I include DURATION, because the magnitude of the acquisition premium is positively related with the duration of the deal (Hsieh & Walking, 2005). The longer the time span, the more time the market has to perceive the deal and determine whether the deal will be a success or not (Flanagan & O’Shaughnessy, 2001). Furthermore, acquirers that perceive value creation from a combined enterprise, would like to effectuate the deal as soon as possible. As a result, the longer the time until a deal is effectuated, the greater the size of the premium.

Numerous studies have incorporated the method of payment in their analysis (Varaiya, 1987; Sullivan, Jensen & Hudson, 1994). I include CASH, because preliminary research has found different, but significant results for the method of payment. On the one hand, deals that are entirely financed with cash are expected to have higher acquisition premiums, because target shareholders demand compensation for the direct taxation of their gains (Sullivan, Jensen & Hudson, 1994). On the other hand, one might argue that there persists a negative relation between a cash transaction and the acquisition premium. The negative relationship is supported by the signaling theory which states that target shareholders will demand a larger premium to all stock transactions compared to cash bids (Eckbo, 2009). The signaling theory hypothesizes that the acquirer will choose to use his own shares as a means of payment when they believe that their shares are overvalued. Target shareholders will then require a higher premium to compensate them for the expected loss once the stock prices return back to their fair value.

Finally, the deal-specific variable HOSTILE is accounted for in my regression model. Despite, results for the effect of hostile offers are mixed. First, evidence displays a positive relationship between the size of the acquisition premium and hostile deals. The target’s board of directors fight a hostile takeover attempt, because they perceive that the acquirer is overvalued and want to protect their shareholders from losing value in this deal (Berk & DeMarzo, 2014). However, this could induce the acquirer to pay more for the target to illustrate that they have the right intentions and want to compensate the target’s shareholder for their fear that they will lose value. Inherent, this would imply a positive coefficient for hostile takeovers (Rossi & Volpin, 2004). Second, there could persist a negative relationship between the size of the acquisition premium and hostile deals. Hostile deals could reduce the acquisition premium, because target’s shareholders directly tender their shares to the acquirer for an offer that is above the current market capitalization, but below an offer desired by the target’s board of directors (Flanagan & O’Shaughnessy, 2001).

(19)

4.1.2 Research Variables

The purpose of my thesis is to find evidence that supports my hypotheses and answers the general question in my research whether firms from emerging countries are undervalued with respect to firms developed countries. Coherent, I created independent variables to test whether my expectations are correct. First, I created a dummy variable CROSSBORDER, that is equal to 1 if a specific deal is cross-border and 0 if it is a domestic deal. This variable will determine whether cross-border deals are inherent to higher acquisition premiums compared to domestic M&As, because cross-border deals have shown to be more value enhancing (Francis et al., 2007; Liu & Qiu, 2013). Second, I created four dummy variables to represent all possible deals, namely deals with an acquirer from a developed country and a target from an emerging country (DevelopedàEmerging), deals with an acquirer from an emerging country and a target from a developed country (EmergingàDeveloped), deals amongst firms from developed countries (DevelopedàDeveloped) and deals amongst firms from emerging countries (EmergingàEmerging). Regression output will therefore contribute to the extent whether these country characteristics affect the size of the acquisition premium and whether there is a significant difference in the premium that acquirers pay based on the country exemplification. In other words, based on whether the size of the premium depends whether a target is from an emerging country or developed country.

Finally, I also included an interaction term between CROSSBORDER and DevelopedàDeveloped to see whether cross-border deals between two firms from developed countries affect the size of the acquisition premium other than both variables separately. This interaction term is of course a dummy variable as both CROSSBORDER and Developedà Developed are dummy variables. Significance of this interaction term infers that beyond the two terms separately, that there is a combined effect on the acquisition premium.

4.2 Econometric Assumptions

I estimate the coefficients in my regression model by using Ordinary Least Squares. By doing so, I assume that the Gauss-Markov assumptions hold. The first assumption concerns that the conditional expectation of the error term on the included regressors is zero, which can be violated if one of the regressors is correlated with the error term. When this first assumption is violated, then the OLS estimates may be biased and inconsistent and cannot be used to make valid inferences. I mitigate this concern, by incorporating the right control variables to diminish omitted variable bias and avoid any correlation between the explanatory variables and the error term – problem of endogeneity. The second assumption implies that both the acquisition

(20)

premium and the control variables should be independently and identically distributed. This condition holds, because the data for my study is collected through simply random sampling. The third assumption states that large outliers are unlikely. I exclude the potential of large outliers in my sample by deleting data points that are clearly composed with an error and by winsorizing variables that seem to have an incongruity in the tales. The fourth assumption regards that no regressors are said to exhibit perfect multicollinearity. As you can see from my regression model, I excluded the deals amongst firms from emerging countries (EmergingàEmerging) to avoid perfect multicollinearity. Any significant coefficient of any of the other deals will therefore stipulate whether they differ in the premium that they pay compared to deals amongst firms from emerging countries. Perfect multicollinearity affects the regression coefficient to the extent that inaccurate inferences will be made from interpreting the results. Also, in the results section, I will therefore first examine whether the independent variables are highly correlated with other variables and remove the variables that exhibit forms of multicollinearity. Finally, the last OLS assumption is that the error term should have a heteroscedastic variance, which implies that the error terms are normally distributed with a mean of zero. When this assumption is violated, the OLS estimates may then be biased and inconsistent and again cannot be used to make valid inferences. This is called the problem of heteroscedasticity, which is one of the most common problems in cross-sectional data analysis (Stock & Watson, 2015). I correct for this by performing a regression analysis using heteroscedastic robust standard errors in Stata. This implication is important, because homoscedasticity cannot be assumed beforehand and would otherwise lead to false conclusions.

Under these Gauss-Markov assumptions the OLS-estimators are unbiased, consistent and asymptotically normally distributed. By correcting for errors in my data set, excluding variables that exhibit forms of multicollinearity and correcting for heteroscedastic standard errors, I will obtain regression output from which I can make accurate inferences. As a result, I conduct an analysis to test my hypotheses in line with table 1.

Table 1 Summary of coefficients used to test hypothesis

Hypotheses Coefficients Testing procedure

Hypothesis 1: Deals between targets from emerging

countries and acquirers from developed countries pay a higher acquisition premium than deals with both a target and an acquirer from emerging countries

𝛽1 H0: 𝛽1= 0

(21)

Hypothesis 2: Deals amongst targets and acquirers that

both reside in developed countries pay a higher

acquisition premium than deals with both a target and an acquirer from emerging countries

𝛽8 H0: 𝛽8= 0

H1: 𝛽8> 0

Hypothesis 3: The acquisition premiums for targets from

emerging countries paid by acquirers from developed countries is significantly lower than the acquisition paid for a target from a developed country

𝛽1+ 𝛽8 H0: 𝛽8= 𝛽1

H1: 𝛽8> 𝛽1

5 Data Collection and Methods

5.1 Sample Selection and Refinements

The data on M&A deals that is analyzed for this study is collected through the Zephyr database of Bureau van Dijk and DataStream. My sample is comprised of completed domestic and cross-border M&A deals over the period 2000 to 2008. I have chosen this period, because M&A activity is highly cyclical and contingent upon market shocks (Madura et al., 2011). Any inferences made with regards to the effect of emerging countries on the acquisition period, would therefore be misled by a period that includes the global financial crisis. Consistent with prior research, I only obtained data where the acquirer acquired 100% of the target’s stock (Hsieh & Walking, 2005). Partial acquisitions would make the M&A deals more complex as a result of mixed control rights, which decreases the acquirer’s estimated synergies from the inability to replace target’s management. In addition, I selected the deals that entail a cash component for the method of payment, because all cash transactions are found to be profoundly different from transactions with stocks or other methods of payment with respect to the acquisition premium (Sullivan et al., 1994). Next, I obtained this data for both domestic as well as cross-border deals to include a potential cross-border effect on the acquisition premium (Rossi & Volpin, 2004; Moeller & Schlingemann, 2004). After applying these selection criteria, I obtained an initial sample of 1345 transactions including 301 cross-border deals.

Next, I applied several filters to ensure consistency of the data. I deleted the observations from which I cannot obtain the required data from DataStream on share prices and shares outstanding. Target’s share prices and shares outstanding are used to calculate the market capitalization, and subsequently the acquisition premium. This market value of equity is composed with data 28 days prior to the announcement and 42 days prior to the announcement of the deal. Finally, after examining the calculated acquisition premiums, I decided to delete deals that are comprised with errors in the share prices or shares outstanding, for example deals with an acquisition premium well below -90% and above 1000%. Therefore, my final sample contains 1082 M&A deals with 243 cross-border deals. This share of cross-border deals in the

(22)

sample is not uncommon, since cross-border deals typically account between the 20 and 40 percent of global mergers and acquisitions (Danbolt & Maciver, 2012).

Table 1 shows the distribution of these cross-border M&As across developed and emerging countries. As already argued in the introduction, most cross-border mergers and acquisitions take place between developed countries, which accounts for 89.7% in this sample. In fact, 48.2% of all cross-border M&As took place amongst firms that reside in just three countries; the United States, the United Kingdom and Canada, as presented in table 2. Similarly, table 3 shows that that the United States, the United Kingdom and Canada accounted for more than 80% of all M&A activity during my sample period. Yet, a much smaller number of deals takes place with firms from emerging countries. Indicated by table 1, only 4.1% of all cross-border M&As take place with an acquirer from a developed country and a target from an emerging country. South Africa, Singapore and China account for 50% of the emerging countries that are acquired. Even a smaller number of cross-border takeovers takes place with two firms that both reside in emerging countries, 0.8%. However, domestic M&A activity for emerging countries is 3.9% for the total duration of my sample period.

Table 1 Distribution of Cross-border M&As Country of Target Firm Developed Emerging Country of Acquiring Firm

Developed 218 (89.7%) 10 (4.1%)

Emerging 13 (5.3%) 2 (0.8%)

Table 2 Cross-border M&As by country Table 3 All M&As by country

Country of Acquirer Firm Country of Target firm Country of Acquirer Firm Country of Target firm

Country Share Total Country Share Total Country Share Total Country Share Total

US 25.5% US 32.5% US 53.3% US 55.2%

UK 16.5% UK 15.6% UK 14.9% UK 14.7%

Canada 14.4% Canada 18.5% Canada 11.6% Canada 12.6%

France 6.2% Australia 6.6% France 2.1% Australia 2.8%

Germany 5.8% Netherlands 5.3% Australia 1.8% Sweden 1.7%

Netherlands 2.9% Sweden 3.7% Germany 1.5% Netherlands 1.6%

Denmark 2.9% France 2.9% Sweden 1.4% France 1.4%

Switzerland 2.9% Denmark 2.1% Japan 1.2% Japan 1.3%

Sweden 2.5% Norway 1.6% Netherlands 1.0% Denmark 1.0%

(23)

5.2 Sample Description

Table 4 presents the descriptive statistics for the sample of 1082 mergers and acquisitions. While not all acquisition premiums are positive, the distribution is averaged around 67.5 for ACQPR28 and 71.5 for ACQPR42, which are not statistically different from each other (t(1082)=-0.05, p>0.05, two-sided). The acquisition premiums show large standard deviations and high economic extremes, which indicates that the distribution is slightly skewed to the right. The occurrence of dispersed values for the acquisition premium is rather normal given the economic nature of premiums (Sonenshine & Reynolds, 2013). Among the research specific variables, DEVàEME and DEVàDEV have mean (standard deviation) values of 0.009 (0.096) and 0.940 (0.238), respectively. As a result, DEVàEME is not statistically different from zero (t(1082)=0.09, p>0.05, two-sided), while DEVàDEV is statistically significant (t(1082)=3.95, p<0.05, two-sided).

Among the control variables, ROA and LEVERAGE are negative, on average -4.964 and -0.007 respectively. The negative average for ROA is consistent with Dimopoulos and Sacchetto (2014) and the negative coefficient for leverage implies that relatively a large fraction of targets in my sample are able to foresee their debt liabilities using only cash or liquid asset. In addition, the market-to-book ratio (MARKETBOOK) is on average positive and greater than one, which infers that the targets have on average a greater market value of equity than book value. Finally, the natural logarithm that is applied to DEALVAL and SIZE induces that the distribution of these variables is indeed normal with mean (standard deviation) values of 12.179 (1.902) and 11.733 (1.952), respectively.

Table 4 Summary Statistics

Variable #Obs Mean Std. Dev Min Max

ACQPR28 1082 67.518 81.537 -42.740 439.773 ACQPR42 ₁₀₈₂ _71.465 _86.100 _-42.783 _474.108 CROSSBORDER 1082 0.225 0.418 0.000 1.000 DEVàEME 1082 0.009 0.096 0.000 1.000 EMEàDEV 1082 0.012 0.109 0.000 1.000 DEVàDEV 1082 0.940 0.238 0.000 1.000 CROSSBORDER * DEVàDEV 1082 0.201 0.401 0.000 1.000 DEALVAL 1082 12.179 1.902 6.409 18.315 DURATION 1082 111.094 94.49 0.000 1506 CASH ₁₀₈₂ _0.417 _0.493 _0.000 _1.000 HOSTILE 1023 0.072 0.259 0.000 1.000 MARKETBOOK 951 2.899 4.452 0.1800 32.700 ROA 984 -4.964 23.627 -122.228 30.553 SIZE 1082 11.733 1.952 5.256 18.121 LEVERAGE 982 -0.007 0.398 -0.9994 1.279

(24)

6 Results

6.1 Correlation between Model Variables

Table 5 presents the correlation matrix for the proposed model on a sample of 1082 M&A transactions, discussed in the methodology.

By analyzing the correlation coefficients, there appears to be a negative correlation between the acquisition premium (ACQPR28 and ACQPR42) and completed M&A deals that have a target from an emerging country. Consistent with my expectations, both DEVàEME and EMEàEME have a negative correlation coefficient with the acquisition premium, -0.0209 and -0.0969 respectively. Further, deals with a target from a developed country appear to be positively correlated with the acquisition premium. Both results align with my expectations that country characteristics are associated with a respective decrease and increase of the acquisition premium. I expect deal premiums to be lower when targets are acquired from emerging countries rather than developed countries, because firms from emerging countries have significantly less bargaining power and are subsequent to more volatile stock prices, which enforces the risk of an international M&A transaction.

The correlations between the independent research specific variables and the proposed control variables in the model are at tolerated levels and do not represent any form of multicollinearity. However, the control variables DEALVAL and SIZE are highly correlated, 0.9783, which could make it difficult to distinguish the effect of one independent variable from the other, especially when the correlation is almost linear. The high correlation can be expected between these two control variables since the value of the deal is a function of the size of the target. When acquirers ought to acquire a target, the firm must establish an offer that is at least equal to the pre-M&A market capitalization. Therefore, both the deal value and the size of the target incorporate the target’s market capitalization, which explains the extreme correlation between the two control variables. To address this potential problem of multicollinearity, I have decided to remove DEALVAL from my regression in order to obtain unbiased and consistent OLS-estimators. I remove DEALVAL instead of SIZE, because SIZE has more consistently proven to significantly affect the magnitude of the acquisition premium compared to the value of the deal (Flanagan & O’Shaughnessy, 2001). Nevertheless, I will incorporate both in the first regression to illustrate that their coefficients affect the outcome of the regression.

(25)

Ta bl e 5 Co rr el at io n Ma tr ix fo r D ep en de nt a nd In dep en den t V ari ab les A C Q PR 28 A C Q PR 42 CRO SS B~ R DE V à EM E EM E à DE V DE V à DE V EM E à EM E CBD E V à DE V DE AL VAL DUR AT ION CA SH HOS T IL E SIZ E MAR KE T B OOK RO A LEV ER A G E A C Q PR 28 1. 0000 A C Q PR 42 0. 7952 1. 0000 CRO SS BO RD E R 0. 0535 0. 0789 1. 0000 DE V à EM E -0. 0209 -0. 0269 0. 1654 1. 0000 EM E à DE V 0. 0259 0. 0212 0. 1981 -0. 0095 1. 0000 DE V à DE V 0. 0761 0. 0665 -0. 0760 -0. 3612 -0. 4324 1. 0000 EM E à EM E -0. 0969 -0. 0799 -0. 0930 -0. 0177 -0. 0212 -0. 8084 1. 0000 CBD E V à DE V 0. 0550 0. 0837 0. 9418 -0. 0450 -0. 0539 0. 1247 -0. 1008 1. 0000 DE AL VAL -0. 0179 -0. 0002 0. 1249 0. 0188 0. 0099 -0. 0345 0. 0277 0. 1210 1. 0000 DUR AT ION -0. 0203 -0. 0316 -0. 0217 0. 0588 0. 0604 -0. 2439 0. 2354 -0. 0507 0. 2121 1. 0000 CA SH 0. 1236 0. 1159 0. 2386 0. 0501 0. 0354 0. 0602 -0. 1156 0. 2291 -0. 0984 -0. 2346 1. 0000 HOS T IL E -0. 0254 -0. 0155 0. 0010 0. 0257 0. 0546 -0. 0846 0. 0607 -0. 0181 0. 0014 0. 0448 0. 0003 1. 0000 SIZ E -0. 1984 -0. 1847 0. 1052 0. 0247 0. 0100 -0. 0537 0. 0482 0. 0988 0. 9783 0. 2113 -0. 1245 0. 0052 1. 0000 MAR KE T B OOK -0. 0736 -0. 0729 0. 0777 -0. 0007 -0. 0134 0. 0524 -0. 0558 0. 0845 0. 2142 0. 0284 -0. 0783 -0. 0389 0. 2240 1. 0000 RO A -0. 0389 -0. 0141 0. 0186 0. 0390 -0. 0128 -0. 0748 0. 0797 0. 0130 0. 2971 0. 0630 -0. 0078 0. 0966 0. 2921 -0. 0669 1. 0000 LEV ER A G E 0. 0966 0. 0972 -0. 0142 -0. 0289 0. 0360 -0. 0107 0. 0065 -0. 0196 0. 2586 0. 1005 -0. 0360 0. 0053 0. 2355 -0. 0258 0. 2801 1. 0000 Bo ld te xt in dic ate s th at co rre la tio n ar e sta tis tic ally s ig ni fic an t a t a 5 % s ig nif ic an ce le ve l Pe ar so n co rre la tio ns a re re po rte d ab ov e in th e ta bl e

(26)

6.2 Regression Results

This section displays the results of my regression analysis in which I account for alternative methods in calculating the acquisition premium. Table 6 reports my regressions results, in which the acquisition premium is based upon the target’s market capitalization 28 days prior to the announcement of the deal. Whereas, table 7 marks the regression output with the target’s market capitalization 42 days prior to the announcement. The method that is used to conduct both regression analyses is fundamentally the same, apart from the dependent variable. Both alternate measures of the acquisition premium contribute to the extent whether inferences about the effect of emerging countries on the acquisition premium are correct and not contingent upon false conclusions.

6.2.1 Acquisition Premium on the basis of 28 days

The first regression model includes 8 control and 5 independent variables including year dummies. The model, however, still includes deal value as a control variable to illustrate the effect of multicollinearity. The coefficients for both deal value and size are significantly different from zero at a 1% significance level (t(892)=16.59, p<0.01, two-sided; t(892)=-16.74, p<0.01, two-sided). Yet, the effect of each variable on the acquisition premium is excluded by the other variable as the coefficients move in the opposite direction and are not significantly different from each other (t(892)=0.074, p>0.1, two-sided). Furthermore, variables such as CASH and MARKETBOOK that are proven to significantly affect the size of the acquisition premium, are insignificant. Therefore, I exclude deal value in regression number two, three and four and focus my analysis along these three regression models.

In model two and four I find a positive but insignificant coefficient for CROSSBORDER (p>0.05, two-sided). This finding is consistent with Rossi and Volpin (2004) and Sonenshine and Reynolds (2013), who show that acquirers are willing to pay a higher premium for cross-border targets, because acquirers are expected to accrue a great amount of synergies from cross-border transactions. However, model three does appear to have a negative coefficient for CROSSBORDER, which is also insignificant (t(892=-0.43, p>0.05, two-sided). This finding is inconsistent with preliminary research, but could be the result from omitted variable bias by excluding the year dummies. Nevertheless, both insignificant regression coefficients imply that cross-border deals and domestic deals do not differ from each other in the premium that acquirers pay for targets.

In each model, I find a positive coefficient for the dummy variable that equals one if both the acquirer and the target are from developed countries (DevelopedàDeveloped), which

(27)

is significantly greater than zero (p<0.01, one-sided). This finding suggests that deals with firms from developed countries pay significantly higher acquisition premiums than deals with firms from emerging countries, which is consistent with hypothesis 2. Consequently, deals amongst firms from developed countries increase the acquisition premium by 27.83% compared to deals amongst firms from emerging countries. Yet, the other deals, DevelopedàEmerging and Emerging à Developed, are not significantly different from zero, which implies that the acquisition premium for these respective deals do not differ from deals amongst firms from emerging countries. However, DevelopedàEmerging and EmergingàDeveloped do appear to positively affect the size of the acquisition premium given their positive coefficients. These findings are consistent with hypothesis 1 and 3, because deals between firms from developed and emerging countries could potentially extract greater synergies from a combined enterprise due to the lower cost of capital. As a result, the target has more bargaining power, which increases the premium (Doukas & Travlos, 1988).

I also test along all four regressions whether there is a significant interaction between cross-border deals and deals with an acquirer and target both from developed countries (C x (DàD)). Both terms appear to be insignificant, but are estimated in regression two and three to be positive (p>0.05). The interaction term could have exemplified the effect of cross-border mergers and acquisition between two firms from developed countries. The significant result for deals comprised with firms from developed countries is already indicative for the fact that these deals coincide with higher acquisition premiums, but the interaction term would have amplified this effect by multiplying this variable with the cross-border dummy. Nevertheless, the interaction is insignificant in all regressions, and does therefore not contribute to an extra effect on the acquisition premium.

Furthermore, a number of control variables that I included, which should affect the size of the acquisition premium, are merely all statistically significant and of the predicted sign. First, the control variable DURATION is positive in each model, and significantly greater than zero in model two and three (p<0.05, one-sided). Hsieh and Walking (2005) confirm this positive relation in their research, which implies that announced deals that take longer to complete coincide with larger acquisition premiums. Second, similar to prior research analyzing the effect of method of payment on the acquisition premium (Sullivan et al., 1994), I find significant positive coefficients among all regressions for CASH (p<0.01, one-sided). This finding suggests that target’s shareholders demand a higher acquisition premium for M&A deals that are entirely financed with cash, to compensate them for the direction taxation of their gains. Third, Rossi and Volpin (2004) find a positive coefficient for hostile takeovers, where I