“Do cross-border acquisition announcements to emerging markets create shareholder value; a study considering the effect of legal and financial institutions”

(1)

“Do cross-border acquisition announcements to emerging markets create

shareholder value; a study considering the effect of legal and financial

institutions”

Jan Willem Stoffer (s1906771)

Master’s Thesis MSc IFM Date: 20-06-2014

Abstract

While emerging countries liberalize their economies and rapid globalization of business increases the pressure on developed market firms to expand abroad, there is only little empirical evidence available on shareholder wealth creation from cross-border acquisitions to emerging markets. This study conducts an event-study on a sample of 2149 cross-border acquisitions announcements from 24 different developed countries to 54 different emerging countries. Over a two day event-window a significant cumulative abnormal return of, on average, 0.7% is found for developed market acquirers. Furthermore a cross-sectional regression analysis is conducted on the sample of cumulative abnormal returns to explain the variance in returns. A significant positive effect on acquirer abnormal returns is found for cross-border takeovers where there is a pronounced disparity between home and host country legal and financial institutions. First, the disparity in the in quality of legal institutions indicates that a developed market acquirer can create value by sharing better legal institutions, corporate governance and accounting standards. Second, the disparity in quality of financial institutions enables the developed market acquirer to create value from providing the target firm with better options to finance growth opportunities which they otherwise had to forgo.

Supervisor: Nanne Brunia

(2)

Introduction

During the last few decades many emerging countries liberalised their economies and opened up their closed business environments. A result of this shift toward a more liberal economy is that both foreign and local firms actively begun undertaking acquisitions. There is a massive increase in the number of acquisitions in emerging countries following their liberalizations compared to the pre-liberalization period. During the period of 1991 to 2003 cross-border acquisitions accounted for 48% of foreign direct investment (FDI) in East Asia and 61% of FDI in Latin America, this compared to respectively 4% and 10 % during the 1980’s (Chari et al, 2010). The opportunities and pressures to engage in cross-border acquisitions have become even greater due to the rapid pace of technological developments and globalization of

business (Hitt, 2000).

Previous research on cross-border acquisitions has focussed on several different topics such as; the mode of entry (Kogut and Singh, 1988; Hennart and Reddy, 1997; Andersen, 1997; Barkema and Pennings, 1996; Brouthers and Brouthers, 2000), organizational learning and transfer of knowledge (Bresman et al., 1999; Vermeulen and Barkema, 2001; Bhagat et al., 2002), target integration processes (Olie, 1994; Weber et al., 1996; Lubatkin et al., 1998; Inkpen et al., 2000; Child et al., 2001) and creation of shareholder wealth (Kang, 1993; Markides and Ittner, 1994; Datta and Puia, 1995; Morck and Yeung, 1992; Francis et al., 2008; Chari et al 2010.). However, only a small part of the literature about cross-border acquisitions empirically studied the effect of host country institutions on shareholder wealth creation (Chari et al. 2010). This lack of coverage is strange when you realize that a weak institutional environment is one of the key differences between cross-border and domestic acquisitions (Khanna et al., 2005)

(4)

2

Economic Freedom Index). In general it is assumed that this makes doing business in these countries more complicated (Khanna et al., 2005).

However if developed-market acquirer firms are able to circumvent these weak institutions, by either creating an internal market for valuable assets or extending their excellent institutions abroad, this should be reflected in value gains for the acquirer firm shareholders. Therefore the research question of this paper is: ‘Do weak legal and financial institution, present in emerging countries, offer an opportunity to create value for a developed country acquirer? To answer this question this paper investigate how different strengths in target country legal and financial institutions affect acquirer abnormal returns. This makes this research related to the earlier work of Francis et al., (2008) and Chari et al., (2010). For example, Francis et al., (2008) found higher abnormal returns for US acquirers when splitting up their sample in financial integrated and financial segmented markets, the effect was even stronger for the acquirers with the lowest cost of capital. Chari et al., (2010) found higher abnormal returns when a developed market firm acquires a firm from an emerging country with weak contractual institutions. They used a distance score between the strength of the acquirer and target country’s contracting institutions, based on the index provided by La Porta (1998).

This study differs from Francis et al. (2008) and Chari et al. (2010) since it makes use of the Economic Freedom Index (EFI) to assert the quality of the legal and financial

institutions of the target countries. This index is an annual guide published by The Heritage Foundation and The Wall Street Journal. Every year they rank 183 countries on a subset of relevant economic variables, which can be used as indicator for several institutions. In contrast the index of La Porta (1998) is compiled based on information from 1982 to 1998. Since institutions develop over time this might be a less reliable measure for more recent periods. Second, whether the previous studies use a sample from which the latest is until 2006, this paper extends the findings by making use of a sample from 1997 to 2013. It is especially valid to see if the results still hold for an extended period, because the development of institutions is a dynamic process. Third, where most of the studies on cross-border

acquisitions take into account only US acquirers and mostly developed market target firms, this study accounts for 2149 cross-border acquisitions from 24 different developed countries to 54 different emerging countries. The aim of this paper is to contribute to the scarce

(5)

3

The remainder of this paper is structured as follows: in section 1 the related literature is reviewed and the hypothesis are developed. The data collection and the method of analyses are explained in section 2. Than the results of both the event study and cross-section analyses are presented and discussed in section 3. Finally section 4 concludes on the findings, present limitations and provide suggestions for future research.

1. Literature

1.1 Shareholder wealth creation from acquisitions

Gains in shareholder value are market based measures of expected returns from a company’s investment. Immediately after the acquisition announcement acquirer firm shareholders revise their expectations based on the expected net present value (NPV) of the acquisition

(Panayides and Gong, 2002). For the net present value of the investment they take into account the price paid for the target firm, the discounted future cash-flows and the possible synergies from the investment (Brooks, 2008). Synergy between two firms exist when the value of the combined firm is higher than the sum of the values of the individual companies, this can be either due to cost reduction or revenue enhancement (Seth, 1990; Bradley et al., 1988). If the acquirer firm shareholders believes that the NPV of the investment is positive the acquisition is seen as adding value to shareholder equity and so result in a positive abnormal return following the announcement. When the NPV is expected to be negative, the acquisition is seen as destroying shareholder value and so result in a negative abnormal return for the buyer firm.

1.2 Determinants influencing shareholder wealth creation from acquisitions

The average value gain in studies about both domestic and cross-border acquisitions depends crucially on the composition of the sample. First it is discovered that acquirer firm

shareholders gain more from acquisitions of privately held firms than from acquisitions of market listed firms (Chang, 1998; Fuller et al. 2002, Moeller et al. 2004; Faccio et al. 2006; Officer et al. 2009). This can be due to a liquidity discount for private firms. A private target often can’t be bought and sold as easily as a listed firm, because they don’t have an

(6)

4

required rate of return is used to discount future cash flows. Second the method of payment is seen as an important characteristic. All acquisitions paid with cash resulted in higher bidder returns than acquisitions paid with stock (Franks et al. 1988; Huang and Walking, 1987; Franks and Harris, 1989;Yook, 2003). Payment with stock can signals to investors that management believe that their stock is overvalued (Goergen & Renneboog, 2004). Third, unrelated acquisitions by means of industrial diversification are found to be value destroying (Maquiera et al., 1998; Berger and Ofek, 1995; Doukas et al. 2001 Moeller and

Schlingemann, 2005; Denis et al. 2002). This confirms the general agreement that no value can be created from industrial diversification, because investors can diversify on their own by creating a portfolio of stocks from different industries. Fourth, studies of Gorton et al. (2009), Alexandridis (2013) and Alexandridis et al. (2010) found that lower premiums are paid for large targets, this because of the limited competition for these large firms. Although at first sight this seems positive, they found substantial value destruction at the time of the cross-border acquisition announcement of large targets. They argue that this is due to the fact that investors might be highly concerned about the realisation of synergies from large and complex deals. Fifth, Fuller et al. (2002) empirically found that the larger a target firm is compared to the acquirer the lower the acquirer abnormal returns will be. In line with this, Higgins and Beckman (2006) argue that acquirer firms should have as certain size and financial power to cope with the often large acquisition and integration cost. Sixth the Q hypothesis claims that a high ratio of market to book value is a good indicator of the ability of a firm to use their assets well and create value from invested resources. The empirical

literature found a positive effect on abnormal returns for acquirers with high Q ratios (Jovanovic and Rousseau, 2002; Servaes, 1991; Lang et al. 1989). Finally, Maloney et al. (1993) and Kang et al. (2000) found that highly leveraged acquirers gain higher abnormal returns when announcing an acquisition compared to low leverage acquirers. They see this as support for the theory that the level of debt can reduce agency costs of the acquirer firm. The high level of debt serves as an extra control on how management spent their cash and restrict the potential danger of management acting in their own interest. It is important to control for these determinants when studying the effect of host country legal and financial institutions on acquirer abnormal returns. This in order to ensure the effect isn’t influenced by factors

(7)

5

1.3 Value creation from cross-border acquisitions to emerging markets

Business environments in emerging markets differ in a number of ways from developed markets. Where competition in developed markets is saturated and institutions to facilitate business are of relatively high quality, emerging markets are characterised by high market growth and low quality institutions to facilitate business. The different characteristics between developed and emerging markets can offer (Khanna & Palepu, 1997)several potential

synergies for developed market firms when acquiring an emerging market target firm. Buying an already existing firm in emerging markets allows developed country acquirers to obtain access to new growth markets and gain local knowledge about doing business in these institutional underdeveloped environments (Shimizu 2004). Furthermore the internalization theory asserts that firms can create an internal market on which they can transfer technological know-how, marketing skills, management skills and all other kinds of intangible assets, which become more valuable in direct proportion to a firm’s market size (Pantzalis, 2001; Shimizu et al., 2004; Morck & Yeung 1992). In emerging markets with both high transaction cost of doing business and large growth potential, this might be an especially valuable option (Khanna & Palepu, 1997). Second, the multinationality hypothesis states that companies operating in many foreign countries can make use of a transnational network of operations which give them an advantage over domestic companies (Pantzalis, 2001). They can for example arbitrage across market segmentations by deciding where to source inputs such as raw materials and labour or where to allocate profit and losses. If the number of international strategic options increase, the incremental value of these options should be reflected in the value of the firm (Doukas & Travlos, 1988; Errunza & Senbet, 1981, 1984) In line with this theory Doukas and travlos (1988) found that abnormal returns for cross-border acquisitions where higher when US firms expanded to countries which were less developed than the US economy and in which they weren’t present before. Lastly Chang (1998) found that when competition for ownership of the target firm is low, the acquirer firm receives a limited competition discount. This can be particularly true for target firms located in emerging markets with less extensive financing opportunities or which restrict foreign investment flows. The lack of competition enables the acquirer to pay a lower price for the target and capture a larger proportion of the synergy.

(8)

6

their business (Argarwal and Ramaswami, 1992; Anderson and Gatignon, 1986; Beckman and Haunschild, 2002). Previous international experience is an indication of already existing processes and systems in place to manage cross-border activities and capitalize on the synergies (Shimizu et al., 2004). Based on the above the following is hypothesized

Hypothesis 1: Cross-border acquisitions of emerging market target firms by developed market acquirers generate positive abnormal returns for acquiring firm’s shareholders.

1.4 Host country legal institutions and acquirer abnormal return

This paper define legal institutions as the rules and laws in place to protect property rights, prevent corruption and enforce contracts (Khanna & Palepu, 1997). In emerging countries where official discretion instead of legal institutions protect property such as technological know-how, marketing skills and patents it can be extremely costly for foreign firms to protect these assets by arm-length contracts (La Porta, et al. 1997; Coase 1937; Alchian et al. 1978; Williamson, 1979; Grossman and Hart 1986). By internalizing the market for these valuable assets, the acquirer firm can circumvent the difficulties and cost associated with protecting these assets. In emerging markets with weak protection of property rights and low contract enforcement this might be a value creating strategy (Seth et al. 2002). Especially because these assets become more valuable in direct proportion to a firm’s market size. In line with this theory Morck & Yeung (1992) found higher acquirer abnormal returns for US foreign acquisitions when they possess knowledge-based assets.

Furthermore in emerging markets with weak legal institutions, the level of corruption is often high (La Porta et al. 1997;Weitzel et al. 2006; Hoskisson et al. 2000). Weitzel et al. (2006) shows that in these environments the bribery of government officials and red tape result in a higher discount on expected cash flows. The higher required rate of return, used as a compensation for the higher risk, reduces the NPV of the takeover and result in a lower price paid for the target firm (Weitzel et al. 2006). This allows the acquirer to capture more of the synergy.

lastly a country’s legal infrastructure also provide the fundamentals for effective corporate governance to protect property rights of minority shareholders (La Porta et al. 1997; Hoskisson et al. 2000). When this is absent, monitoring of management and large

shareholders will be a costly and time consuming task. This paves the way for both

(9)

7

the expected cash flows of the firm (La Porta et al. 1997). The lower expected cash flows of the firm will also result in a lower valuation of the firm. By gaining substantial control the acquirer can share legal, corporate governance and accounting standards to control for the use of a firm’s resources (Chari et al., 2010). Therefore, the ability to transfer better corporate governance and accounting standards to target firms located in emerging-markets can create potential value gains for buyer firm shareholders (Chari et al., 2010). Chari et al. (2010) argue that the effect might be even bigger when the disparity between home and host country’s legal, corporate governance and accounting standards becomes greater. This because it leaves more room for improvement.

To summarize, the above discussion argues that developed market acquirers have less opportunities to gain from cross-border acquisitions when the legal institutions in the target country are of high quality and more value creation opportunities will exists when the legal institutions are of low quality. Based on this the following two hypotheses are developed.

Hypothesis 2a: Higher quality of host country legal institutions will negatively influence

acquirer abnormal return.

Hypothesis 2b: Greater disparity in legal institutions, between acquirer home and target host

country, will have a positive effect on acquirer abnormal return.

1.5 Host country Financial institutions and acquirer abnormal returns.

(10)

8

Although during the early 1990s many emerging countries liberalized their financial markets and allowed both domestic and foreign investors to compete and move their resources across a country’s borders, integration into the world capital markets is still relatively little in comparison to developed countries (Francis et al. 2008). Errunza and Miller (2000) found that due to less developed and internationally integrated capital markets, firms from emerging countries face a higher cost of capital. In these financial segmented markets, providers of finance require a higher return on investment. This to compensate for the less liquid financial market and because they have fewer possibilities to diversify their portfolio internationally (Stulz, 1999).

This paper argues that if a firm from a financially developed market acquirers a target from a financially underdeveloped emerging market, both a more efficient internal and foreign capital market can be used to raise cheaper capital for financing growth opportunities (Stulz, 1999; Baker et al. 2009 Francis et al. 2008). This cheaper capital can be used to invest in positive NPV projects of the emerging market target firm which they would otherwise have to pass up. It is expected that this potential value creation, caused by a reduction in the cost of capital for the target firm, will result in a positive effect on acquirer abnormal return. In line with this Francis et al. (2008) found positive abnormal returns for US acquirer firms when acquiring a target firm located in a financial segmented market. They even found that the effect was more pronounced for acquirers with the lowest cost of capital. Therefore it is expected that larger disparity, between developed and emerging market financial institutions, offer greater opportunities to reduce the cost of capital and create value for the developed market acquirer.

To summarize, the above discussion suggest that developed market acquirers have less opportunities to gain from a cross-border acquisitions when the financial institutions in the target country are of high quality and more value creation opportunities when the financial institutions are of low quality. Based on this the following two hypotheses are developed.

Hypothesis 3a: Higher quality of host country financial institutions will negatively influence

acquirer abnormal return.

Hypothesis 3b: Greater disparity in financial institutions, between acquirer home and target

(11)

9

2. Data and Methodology

The Zephyr databank on M&As is used to identify the cross-border acquisitions of interest. All cross-border acquisition announcements over the time period 1997-2013 are included in the sample, if the acquirer is form a developed country and the target is located in an

emerging country. The developed and emerging countries included in the sample are derived from the MSCI index and are listed in appendix 1. Furthermore information such as deal value, industry SIC codes and method of payment are also derived from the Zephyr databank on M&As. Acquirer stock returns and returns on the market portfolio are derived from Datastream International. Stock returns must be available for 255 days before and 30 days after the acquisition announcement. This data is used to test hypothesis 1 with event-study methodology. Additionally accounting data such as market value of equity, book value of debt, book value of assets, intangible assets, foreign sales and total sales also come from Datastream International. This accounting data is used to compose several control variables which are included in the cross-sectional analyses.

The Economic Freedom Index is used to collect data for the quality of a countries legal and financial institutions. To recap, this index is an annual guide published by The Heritage Foundation and The Wall Street Journal. Every year they rank 183 countries on a subset of economic relevant variables, which can be used as indicator of several institutions. For both developed and emerging countries information is collected for the years 1997 to 2013. This data is needed to test hypothesis 2a, 2b, 3a and 3b. More detailed information about the independent variables will be discussed in section 2.2 which is about the cross-sectional analyses of the cumulative abnormal returns

Screening on the availability of data, to conduct both event-study and cross-sectional regression analyses, a workable sample of 2149 cross-border acquisition announcements is compiled. The distribution of takeovers over the years 1997 till 2013 can be seen in figure 1. One could notice a gradually growing number of cross-border acquisitions from 1997 till 2006, with a small drop in 2002. However it must be mentioned that the stark growth might be due to the fact that the Zephyr databank on M&As does not include all acquisitions in the early years of the sample. After the peak of 190 cross-border acquisitions in 2006 the number of acquisitions decline till 2009 and then reach its peak again in 2011, with 203 acquisitions.

(12)

10

manufacturing (1018 firms), finance, insurance & real estate (318 firms) and mining (197 firms). The three main target firm industries are: Manufacturing (897 firms), finance, insurances & real estate (287 firms) and services (259 firms).

Furthermore the average deal value of the sample is €167 million with a minimum value of €1 million and an maximum of €14,596 million. 386 acquisitions have a public target whereas 1763 have a private target. 1502 out of the 2149 acquisitions are related to the acquirer industry and 647 are unrelated to the acquirer industry. Lastly, considering the method of payment; 875 out of 2149 are cash acquisition, 55 are financed solely with stock and 1219 are a mixture of cash, stock and other methods of payment.

Figure 1:Yearly distribution of Acquisitions 1997-2013

Table 1: Number of acquirers and targets per industry

Industry Acquirer Target

Agriculture, Forestry, & Fishing 72 152

Mining 197 129

Construction 94 38

Manufacturing 1018 897

Transportation & Public Utilities 179 240

Wholesale Trade 53 63

Retail Trade 62 83

Finance, Insurance, & Real Estate 318 287

(13)

11

2.1 Event study methodology

Hypothesis 1 is tested by making use of event-study methodology. According to Mackinlay (1997) event study methodology is appropriate in assessing the effect of a cross-border acquisition announcement on abnormal returns. Abnormal return is the difference between the actual return on a security and the expected return of the security (Mackinlay , 1997). Furthermore in strategic management and finance this method is extensively used for

studying mergers and acquisitions (e.g. Chari et al. 2010, Markides & ittner, 1994 and Doukas & Travlos, 1988).

Two models commonly used in event-studies are the constant-mean return model and the market-model. Compared to the constant-mean return model the market-model reduces the variance of the abnormal return by taking away the part of the return that is caused by the variation in the return of the market portfolio (Mackinlay, 1997). This could lead to an increased possibility of detecting abnormal returns. For robustness of the results this paper makes use of both models.

To calculate abnormal returns with the ‘constant-mean return model’ the following formula is used

̅

(1)

Where

(2)

Here ARit is the abnormal return for security i at day t , Rit is the actual return, ̅i is the mean

return for security i over the estimation window [-250,-30] and ln(TRit) is the natural

(14)

12

The more sophisticated model used in this paper is the market-model. This model relates the return of each individual security to the return on the market portfolio. The formula to calculate the expected return based on the market-model is noted as

(3)

Where

(4)

Here E(Rit )is the expected return,

α

i is the risk free, βi is the firm’s sensitivity to the market

portfolio, Rmt is the return of the market portfolio and ln(TRmt) is the natural logarithm of the

total return index of the market portfolio. By running an OLS regression over the estimation window [-250,-30], with Rit and Rmt, the parameters ̂ and ̂ are estimated. The abnormal

returns, based on the market-model, are calculated with the following formula

̂

(5)

The average abnormal return (AAR) is calculated to obtain a general impression of the abnormal return observations for the sample of N firms. The formula to calculate the average abnormal return is noted as

∑

(6)

Since the effect of an event might also influence the returns for days close to the event it is common to use a multiple day event window (Mackinlay , 1997). This paper makes use of a two day event-window including day 0 (the announcement day) and day +1 (day after the announcement day). The cumulative abnormal return (CAR) and the average cumulative abnormal return (ACAR) over a chosen event-window [t1,t2] are calculated by using the following formulas

∑

(15)

13 And

∑

(8)

Because some stocks in the sample have a higher total risk than others, Mackinlay (1997) recommend to work with the standardized abnormal return (SAR) when determining the significance of the returns. The standardised abnormal return is calculated by the formula

(9)

Here is the standard deviation of firm i calculated over the estimation window[-250,-30].

By replacing ARit for SARit in formula 6 and 8, the ‘average standardized abnormal

return’ (ASAR) and the ‘average standardized cumulative abnormal return’ (ASCAR) can be calculated. By making use of ASAR and ASCAR the significance of the abnormal returns are tested. However since the coefficients of the standardized returns can’t be interpreted economically, the coefficients of the AAR and ACAR are presented in the result section. Based on Mackinlay (1997) the next two parametric t-tests are conducted

√ (10) And √ √ (11)

(16)

14

abnormal return in the event-window is in line with the expected percentage of positive abnormal returns based on the estimation-window[-250,-30]. This test statistic is calculated by the formula

̂

√ ̂ ̂

(12)

Here w is the number of abnormal returns bigger than 0 within the event-window, n is the number of cross-border observations in the sample and ̂ is the fraction of abnormal returns bigger than 0 within the estimation window[-250,-30].

2.2 Cross-sectional analyses of cumulative abnormal returns

To test hypothesis 2a, 2b, 3a and 3b the cumulative abnormal returns derived from the event-study are regressed on both the independent and control variables. This to explain the

variation in cumulative abnormal returns over the sample of 2149 cross-border acquisitions announcements. Descriptive statistics of all variables can be found in table 2.

2.2.1 Dependent variable

This paper uses the cumulative abnormal returns from the market-model, over day 0 and day +1, as the dependent variable in the regression models. The calculation of the cumulative abnormal returns over the window[0,+1] can be found in formula 7. The choice for this two-day window is based on the significance of the AARs presented in table 4 of the result section.

2.2.2 Independent variables

Legal institutions: Data for determining the strength of an acquirer and target country’s legal

(17)

15

This variable is composed by subtracting the score of the target country from the score of the acquirer country. The aim is to establish if greater disparity between acquirer and target country legal institutions result in higher acquirer abnormal returns (hypothesis 2b). Financial institutions: Also the data for this variable is derived from the Economic Freedom Index. They rank each country on a score between 0 to 100 according to the level of capital market development, financial services in place and openness to foreign competition from both foreign capital markets and individual investors. Also with this score two different variables are composed for the cross-sectional regression. The first variable is called

‘financial institutions unadjusted’. This variable is composed by using the unadjusted score from the economic freedom index in the regression model to determine if better financial institutions in the target country result in lower acquirer abnormal returns (hypothesis 3a). The second variable is ‘financial institutions distance’. This variable is composed by subtracting the score of the target country from the score of the acquirer. The aim is to establish if greater disparity between acquirer and target firm financial institutions result in higher acquirer abnormal returns (hypothesis 3b).

2.2.3 Control variables

To make sure that the results of the independent variables of interest aren’t influenced by other variables external to the regression model, several control variables are included. All controls used are already addressed in the literature section therefore only the data collection and the composition of the control variables is described next.

International experience: To control for the acquirer international experience this

paper use the percentage of foreign sales to total sales as a proxy for international experience. Both foreign and total sales are collected from Datastream International one year prior to the year of the announcement (year -1). In line with Agarwal and Ramaswami, (1992), Anderson and Gatignon, (1986) and Beckman and Haunschild, (2002) a positive coefficient is expected for international experience.

Leverage: This variable measures the degree of leverage of the acquirer firm in the

year prior to the year of the announcement (year-1) and is compiled by dividing total debt by total shareholder equity. In line with Maloney et al. (1993) and Kang et al (2000) a positive coefficient is expected with regard to leverage of the acquirer.

Intangible assets: This variable control for the amount of intangible assets of the

(18)

16

announcement (year -1). In line with Morck and Yeung (1992) a positive coefficient is expected for intangible assets intensity.

Relative purchase price: This variable controls for the price paid for the target firm

relative to the size of the acquirer. For the calculation of this variable the transaction price is scaled by the book value of assets. Here the purchase price is derived from Zephyr databank on M&As and the book value of assets is from DataStream International in the year prior to the year of the announcement (year-1). In line with Fuller et al. (2002) the coefficient is expected to be negative when the relative purchase price increase

Tobin’s Q: The Tobin’s Q ratio is included in the model as a control variable and is

computed by the market value of equity plus the book value of debt divided by the book value of assets. The data to compose this variable is derived from DataStream International one year prior to the year of announcement (year -1). In line with Jovanovic et al. (2002), Servaes (1991) and Lang et al. (1989) a negative coefficient for Tobin’s Q is expected.

Target size: This control variable is computed by taking the log of the transaction

price. The deal value is derived from Zephyr databank on M&As. In line with Gorton et al. (2009), Alexandris (2013) and Alexanridis et al. (2010) the target size control is expected to show a negative coefficient.

GDP growth rate: Yearly growth rates between 1997 and 2013 are derived from The

World Bank. For each target country the annual GDP growth rate is used to control for the growth of a country’s economy. A positive relation is expected between annual GDP growth rate and acquirer abnormal returns.

Cash dummy: The model control for payment by cash. The dummy variable on this

control will take a value of 1 when the acquisition is paid with cash and 0 otherwise. The data for the method of payment is derived from the Zephyr databank on M&As. In line with Franks et al. (1988), Huang and Walking (1987), Franks and Harris (1989) and Yook (2003) a positive coefficient for cash payment is expected over the other forms of payment

Private dummy: Several studies on domestic acquisitions found positive abnormal

returns for the private-target acquisitions and no or negative returns for public target

acquisitions (Chang, 1998; Fuller et al. 2002, Moeller et al. 2004; Faccio et al. 2006; Officer et al. 2009) The listing status of the target is derived from the Zephyr databank on M&As. A dummy variable is composed which takes a value of 0 when the target is listed and 1 when the target is private. A positive coefficient is expected for the private dummy.

Related dummy: To control for the effect of industrial diversification, the industry SIC

(19)

17

as the SIC code of the acquirer, the dummy takes a value of 1 and otherwise 0. In line with (Maquiera et al., (1998), Berger and Ofek, (1995) an Doukas et al. (2001) a positive coefficient is expected for related acquisition.

Table 2: Descriptive statistics

Variable Observations Mean Std. Dev. JB (p-value)

Dependent variable:

CAR [0,+1] 2149 0.007 0.053 (0.00)

Independent variable:

Legal institutions unadjusted 2149 46.90 15.52 (0.00)

Legal institutions distance 2149 37.67 17.77 (0.00)

Financial institutions unadjusted 2149 48.61 13.24 0.00)

Financial institutions distance 2149 23.99 17.96 (0.00)

Control variable:

International experience 2149 52.11 34.20 (0.00)

Leverage 2149 2.05 28.81 (0.00)

Intangible assets 2149 0.17 0.19 (0.00)

Relative purchase price 2149 0.55 17.22 (0.00)

Tobins' Q 2149 2.24 43.02 (0.00) Target Size 2149 17.13 1.84 (0.00) GDP growth rate 2149 4.87 2.41 (0.00) Cash dummy 875 0.41 0.49 (0.00) Private dummy 1763 0.82 0.38 (0.00) Related dummy 1502 0.70 0.46 (0.00)

Here JB is the Jarque-Bera test of normality, where the null hypothesis of normality is rejected when the p-value is significant at a 5% level.

2.2.4 Regression model

Both univariate and multivariate regressions are run (Table 6 and Table 7), this to see whether the sign or size of the individual variables remains constant when they are combined in one regression model. Furthermore the univariate regressions already give useful information about the explanatory power of each individual variable in the model (Brooks, 2008).

Based on the correlations between the variables (table 3) and the results from the univariate regressions (Table 6) the control variables intangible assets, relative purchase price and cash dummy are excluded from the multivariate regression models. The variables

(20)

18

will become significant in the multivariate regressions since they already have non

explanatory power on their own. Including too many variables that are insignificant can also have an adverse effect on the significance of the other variables (Brooks, 2008). Lastly the control variable relative purchase price is excluded from the model because of a correlation coefficient of 0.96 with Tobins’Q and the lower level of significane in the univariate regression (table 3).

With a sufficient large sample of 2149 observations no concerns should have to be made about non-normality (Brooks, 2008). Furthermore there is no need to check for autocorrelation in the error terms since the regression is cross-sectional over firms and 24 different countries. This makes it unlikely that returns influence each other (Brooks, 2008). Finally it is important to control for heteroscedasticity in the regression models. Since daily stock data is used for the dependent variable CAR, there is no reason to assume that the standard errors are homoscedastic (Mackinlay, 1997). The ‘white heteroskedasticity-consistent’ least square regression is used to control for heteroscedasticity (Brooks, 2008).

Together with the remaining control variables regression models 1 to 4 (Table 7) one by one include the four variables of interest: legal institutions unadjusted, legal institutions distance, financial institutions unadjusted, financial institutions distance. This because of the potential danger of multicollinearity when the two unadjusted or distance variables are jointly included in one model (see Table 3).

Table 3: Correlation diagram

Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14

1. Legal institutions unadjusted 1.00

(21)

19

3. Results

3.1 Event study

This paper analysed stock price movements of 2149 firms in the period surrounding a cross-border acquisition announcement. Both the ‘mean-return model’ and the market-model mentioned in Mackinlay (1997) are used to test for the presence of abnormal returns. The results of the mean-return model are similar to the results of the market-model. This is in line with expectations based on Brown and Warner (1980,1985) who found that the constant-mean return model often find similar results as the more sophisticated models such as the market- model. Although the results of the mean-return model are not discussed in the result section, the results can be found in appendix 3 and 4. Furthermore the abnormal returns from 29 days before the announcement of the takeover till 30 days after the announcement are analysed. Since from the market-model no significant average abnormal returns can be related to the period outside the interval [-5,+5], table 4 only present the average abnormal returns over the interval [-5,+5]. The results for the entire event-window can be found in appendix 2.

From table 4 a significant positive average abnormal return (AAR) can be noticed at the day of the announcement (day 0) and one day after the announcement (day +1) . The AAR on day 0 and day +1 are respectively 0.5% (p=0.00) and 0.2% (p=0.02). Also the non-parametric test shows that, at the day of the takeover announcement, 54% of the abnormal returns is above zero (p = 0.03). At the day after the announcement 51% is above zero

,although the result is not significant (p=0.35). Thus also from the non-parametric test, at least for the day of the announcement, this percentage is significantly higher than what can be expected from the percentage positive abnormal returns in the estimation-window[-250,-30]

Based on the significance of the AAR the average cumulative abnormal returns (ACAR) over three event-windows are composed and tested. The results are presented in

table 5 and show a significant positive ACAR of 0.7% (p=0.00) over the two day event-

(22)

20

The results confirm hypothesis 1, which state that cross-border acquisitions of emerging market target firms generate positive abnormal returns for developed market acquirer firms. So in general investors perceive cross-border acquisitions of an emerging market target firm as a value creating strategic initiative. The result is in line with positive returns found in related studies that focus on cross-border acquisitions (Doukas & Travlos, 1988;Markides and Ittner, 1994; Morck and Yeung, 1992; Francis et al., 2008; Chari et al. 2010). However it must be mentioned that most of these studies, except the one by Chari et al. (2010), takes only into account cross-border acquisitions by US firms to more or less developed markets.

Therefore, in the end, the positive returns found in this paper are most in line with earlier findings of Chari et al. (2010) who found a 1.16% acquirer abnormal return between 1986 and 2006 over a three day event window and use a similar sample of developed market acquirer firms and emerging market target firms.

The positive abnormal return also contrast with several studies that found non-significant or negative returns for acquirer firm shareholders (Datta and Puia, 1995; Moeller and Schlingemann, 2005; Denis et al. 2001). A possible explanation for the contrasting results found in this paper might be the examination of different time periods. Where for example Moeller and Schlingeman (1995) and Denis et al. (2001) used a time frame from respectively 1985 to 1995 and 1984 to 1997, this paper examines the period 1997-2013. The latter period is characterised by increased globalization of business and rapid technological development. This increases the pressure on developed market firms to expand abroad and capitalize on new growth markets to stay profitable (Hitt, 2000). A second explanation might be the difference in the geographical focus of the studies. The studies that show a negative return take only into account US cross-border acquisitions to developed markets with relatively similar characteristics as the home country (Datta and Puia, 1995; Moeller and Schlingemann, 2005; Denis et al. 2001). Emerging markets possess several different

characteristics such as high growth potential and mostly underdeveloped legal and financial institutions (Khanna & Palepu, 1997). This paper hypothesized several value creation

opportunities for developed market firms which can arise from these unique characteristics of emerging market countries. Based on the cross-sectional regression analyses in the next section the results for these hypotheses will be discussed in some more detail.

(23)

21

Table 4: Average abnormal returns ‘market model’ day -5 to day +5

Parametric Non-parametric Parametric Non-parametric

Day AARt T1 Art > 0 T3 Day AARt T1

Art > 0% T3 -5 -0.001 (0.08)* 48% (0.03)** +1 0.002 (0.02)** 51% (0.35) -4 0.000 (0.28) 47% (0.02)** +2 0.000 (0.10)* 48% (0.04)** -3 0.000 (0.12) 49% (0.05)** +3 0.001 (0.20) 48% (0.04)** -2 0.000 (0.12) 47% (0.01)*** +4 0.000 (0.35) 48% (0.03)** -1 0.000 (0.35) 48% (0.02)** +5 0.000 (0.21) 49% (0.10)* 0 0.005 (0.00)*** 54% (0.03)**

The parametric test is

√ and the non-parametric test is

̂

√ ̂ ̂ . The p-values are presented.

*Significant at 10% level, **Significant at 5% level and ***Significant at 1% level.

Table 5: Average cumulative abnormal returns ‘market model’

Parametric Non-Parametric

ACAR[t1,t2] Std. Dev. t-statistic P-value CARt > 0 P-value

Window[-29,-1] -0.0015 0.12 -1.35 (0.13) 49% (0.08)*

Window[0,+1] 0.007 0.04 7.68 (0.00)*** 54% (0.03)**

Window[+2,+30] -0.008 0.11 -1.94 (0.07)* 48% (0.03)**

√ √

and the non-parametric test is ̂

√ ̂ ̂ .

3.2 Cross-sectional analysis of cumulative abnormal returns

To test hypothesis 2a, 2b, 3a and 3b, both univariate and several multivariate regressions are performed. The CARs over the two-day window[0,+1], derived from the event-study, are regressed on both the explanatory variables of interest and control variables. The ‘white heteroskedasticity-consistent’ least square regression is used to control for heteroscedasticity (Brooks, 2008). The results of both the univariate and multivariate regressions are presented in table 6 and table 7 respectively.

(24)

22

result is just insignificant (p=0.11). Furthermore the F-statistic for model 1 is significant at a 1% level and the adjusted-R2 is 0.009. The F-statistic test the overall significance of the regression model and indicate if the regression model has enough validity in fitting the data. In other words when the F-statistic is significant it can be assumed that the independent variables are not purely random with respect to the dependent variable. The adjusted-R2 indicates how much of the variance is explained by the regression model.

Hypothesis 2b expect that greater disparity in legal institutions between acquirer home and target host country have a positive effect on acquirer abnormal returns. Both the

univariate regression (Table 6) and model 2 of the multivariate regression confirm this hypothesis. For both regressions a positive coefficient of 0.0002 is found and the results are significant at 1% level (p=0.01). The F-statistic of model 2 is significant at a 1% level and the adjusted-R2 is 0.0105. Compared to the adjusted-R2 of model 1, this model explains more of the variance in abnormal returns.

So with regard to the quality of a host country’s legal institutions significant support can only be found for hypotheses 2b. The result shows that the opportunity to gain from the acquisition increases when the disparity between home and host country legal institutions becomes bigger. This is in line with expectations based Chari et. al (2010) who found that under control of a developed market acquirer better legal institutions, corporate governance and accounting standards can be shared with the target firm. When the target country lack the legal infrastructure to provide effective corporate governance and property rights protection, the improvement in corporate governance and accounting standards can drive value gains for acquirer firm shareholders. The result also confirms to the multinationality hypothesis, which is built on the believe that firms can arbitrage market segmentations between home and host countries included in their international network (Pantzalis, 2001; Shimizu et al., 2004)

Hypothesis 3a expect that higher quality of host country financial institutions will negatively influence acquirer abnormal returns. Both the univariate regression analyses and

model 3 from the multivariate regression analyses show that this hypothesis can’t be

confirmed. The negative sign is as expected but the result is highly insignificant. The F-statistic of the model is significant at a 1% level and the adjusted R2 has a value of 0.0081.

(25)

23

model 4 is also significant at a 1% level and the adjusted-R2 is 0.0105. Comparing the adjusted-R2 of model 3 and model 4, also here the model including the distance measure explain more of the variance in abnormal returns.

Considering the quality of financial institutions this paper found support for the theory that cross-border acquisitions of targets located in countries with weak financial institutions can create value for developed market acquirer firms. However based on the fact that only hypothesis 3b can be confirmed, the disparity between developed market and emerging market financial institutions seems to be an important prerequisite for the positive effect on abnormal returns. If companies from emerging markets face higher cost of capital, because of less developed and international integrated capital markets (Errunza and Miller, 2000; Stulz 1999), developed market acquirer firms can improve the financing conditions of the target firm by creating an internal capital market or giving access to more developed foreign capital markets. This is in line with Francis et al. (2008) who found that if US firms acquirer foreign targets from segmented financial markets they experience higher abnormal returns. Furthermore the result is also in line with the work of Baker et al. (2009) who found that low cost of capital in the home country can be an important driver for cross-border acquisitions, especially when the target country impose capital account restrictions.

Looking at the control variables one can see that from the univariate regression analyses international experience (p=0.10), leverage (0.01), relative purchase price (p=0.03), tobins’Q (p=0.00), private dummy (p=0.04) and related dummy (p=0.01) are significant. However in the multivariate regression models only leverage, tobins’Q and related dummy stay significant at a 1% level. The sign of leverage is as expected and is in line with Maloney et al. (1993) and Kang et al (2000). They found that highly leveraged acquirers gain higher abnormal returns when announcing an acquisition compared to low leverage acquirers. The high level of debt can serve as an extra control on how management spent their cash. Next, the sign of tobins Q is negative and is opposed from what was expected based on the Q hypothesis. The Q

(26)

24

possible explanation for the positive effect of unrelated cross-border acquisitions to emerging markets can be found in Khanna and Palepu (1997). They state that business groups in

emerging markets can be attractive to invest in for firms from developed countries. This because they possess great market knowledge and are seen as reliable by local consumers. The latter is especially important since consumer protection is often very low. Khanna and Palepu (1997) state that these firms often start businesses in complete different industries to leverage their strong local knowledge and reputation. If developed market firms acquirer a stake in such a unrelated business group to use their knowledge and reputation this might be perceived as positive by acquirer firm shareholders.

Table 6: Univariate regression analyses

Variable Sign Coefficient T-statistic p-value R-square

Legal institutions unadjusted (H2a) - -0.0002 -2.19 (0.03)** 0.0026 Legal institutions distance (H2b) + 0.0002 2.81 (0.01)*** 0.0046 Financial institutions unadjusted (H3a) - -0.0001 -1.36 (0.17) 0.0008 Financial institutions distance (H3b) + 0.0002 2.69 (0.01)*** 0.0043

International experience + 0.0001 1.63 (0.10)* 0.0017

Leverage + 0.0001 2.56 (0.01)*** 0.0025

Intangible assets + 0.0007 0.13 (0.90) 0.00001

Relative purchase price - -0.0001 -2.11 (0.03)** 0.0003

(27)

25

Table 7: Multivariate regression analyses

Variable Sign Model 1 Model 2 Model 3 Model 4

Intercept 0.0264 0.0151 0.0189 0.0185

(0.09)* (0.26) (0.25) (0.18) Legal institutions unadjusted (H2a) - -0.0001

(0.11)

Legal institutions distance (H2b) + 0.0002

(0.01)***

Financial institutions unadjusted (H3a) - 0.0000

(0.89)

Financial institutions distance (H3b) + 0.0002

(28)

26

4. Conclusion

In this study it is investigated whether shareholders of a developed market firm perceive a cross-border acquisition of an emerging market firm as a value creating strategic option. Event-study methodology is used to study a sample of 2149 cross-border acquisitions announcements from 1997 to 2013. This to see whether in general stock prices react positively or negatively in the days surrounding a cross-border acquisition announcement. The sample of 2149 observations comprise a variety of acquisitions from 24 developed countries to 54 emerging countries (see Appendix 1).

The findings from the event-study show that on average the stock markets react positively to cross-border acquisitions of emerging market target firms by developed market acquirer firms. The average cumulative abnormal return over an event-window including day 0 (announcement day) and day + 1(day after the announcement) is 0.7% (see Table 5).

Although the effect for a small individual investor might be not be sufficiently large, for a big institutional investor this can be a substantial increase in the value of their equity holdings in the firm. The result from the event-study also contribute to the inconclusive findings about shareholder wealth creation from cross-border acquisitions and add insights to the discussion about the proper valuation of holding foreign assets (see, e.g. Moeller and Schlingemann, 2005; Stulz, 1999).

The sectional regression analyses conducted over the sample of 2149 cross-border acquisition announcements shows that increased disparity in quality between home and host country legal and financial institutions have a positive effect on acquirer abnormal returns. First, the positive effect on abnormal returns resulting from increased disparity in quality of legal institutions can be explained by the opportunity of the developed market firm to create value by sharing better legal institutions, corporate governance and accounting standards. Second, the disparity in quality of financial institutions enables the developed market acquirer to create value from providing the target firm with better options to finance growth opportunities which they otherwise had to forgo. It should be emphasized that the disparity between the quality of home and host country institutions seems to be an important requirement for the developed market acquirer to be able to create value form the institutional weak environment in the host country. Only taking into account the quality of the host

(29)

27

Of course this study also have a number of limitations. First, the effect of positive abnormal returns is measured over a large sample of cross-border acquisition announcements for which the acquirer can be from 24 different developed countries and the target can be located in 54 different emerging countries. Although this makes the result generalizable over a number of developed market cross-border acquisition between 1997 till 2013, the results can’t be generalised to cross-border acquisition of target firms located in developed countries. Second, the results might be biased towards the countries in the sample with the largest number of observations. This makes that one should be cautious when generalizing the results to the countries with fewer observations in the sample (see Appendix 1). Lastly , the event-study methodology used in this paper assume that investors reaction to the cross-border acquisition announcement is immediately, unbiased and complete, based on the semi-strong form of the efficient market hypothesis. Hence, cautious interpretation of the value creation from cross-border acquisition announcements is desirable. This because abnormal returns are the market based assessments of often complex takeovers. Investors may not fully understand the strategy implications of the takeover and therefore heuristic biases might exist. The latter limitation offers a direct opportunity for future research.

Future research can use performance measures which focus on the long term performance. These studies can check for the robustness of the results found in this paper. Second, Khanna et al. (2005) state that firms often use country rankings as source of

information for their globalization strategy. Apart from the economic freedom index and the index created by La Porta (1998) they mention the ‘world economic forum’, ‘world bank’, ‘transparency international corruption index’ and ‘international country risk guide’ as often used indices to assert the institutional environment. Future research can use these indices to check their usefulness and see if similar results can be found. Finally, during the past decade companies from emerging markets become active players in the global takeover markets (Luo and Tung, 2007). They aggressively acquirer companies from developed markets to obtain sophisticated technology and manufacturing know-how, but also try to compensate for weak legal property rights protection, low contract enforcement and underdeveloped factor markets (Luo and Tung, 2007). Future research can focus on emerging market acquirer firms

(30)

28

Appendix

Appendix 1: acquirer and target countries from MSCI index

Acquirer country Numer of Acquirers Emerging market Target country (1) Number of targets Emerging market target country (2) Number of targets

Australia (AU) 75 Czech republic (CZ) 68 Bulgaria (BG) 26

Austria (AT) 43 Egypt (EG) 27 Croatia (HR) 15

Belgium (BE ) 43 Chile (CL) 59 Estonia (EE) 23

Canada (CA) 59 Colombia (CO) 30 Lithuania (LT) 24

Denmark (DK) 25 Greece (GR) 29 Kazakhstan (KZ) 10

Finland (FI) 55 Hungary (HU) 29 Romenia (RO) 43

France (FR) 189 Latvia (LV) 11 Serbia (RS) 19

Germany (DE) 86 Indonesia (ID) 58 Slovenia (SI) 18

Hong Kong (HK) 72 Malaysia (MY) 70 Ukraine (UA) 24

Ireland (IE) 15 Mexico (MX) 97 Botswana (BW) 2

Israel (IL) 17 Poland (PL) 113 Ghana (GH) 4

Norway (NO) 36 Peru (PE) 23 Kenya (KE) 4

Portugal (PT) 16 Philippines (PH) 21 Mauritius (MU) 3

Italy (IT) 75 Slovak republic (SK) 23 Morrocco (MA) 16

Japan (JP) 143 Taiwan (TW) 49 Nigeria (NG) 10

South Korea (KR) 39 Thailand (TH) 49 Tunesia (TN) 6

The Netherlands

(NL) 58 Turkey (TR) 67 Zimbabwe (ZW) 2

Singapore (SG) 156 Venezuela (VE) 6 Jordan (JO) 4

Spain (ES) 99 Brazil (BR) 213 Kuwait (KW) 2

Sweden (SE) 131 Russia (RU) 113 Oman (OM) 1

Switserland (CH) 57 India (IN) 202 Qatar (QA) 1

United Kingdom

(GB) 298 China (CN) 328 Saudi Arabia (SA) 5

United states (US) 362 South – Africa (ZA) 88

Unit. Arab Emirat.

(AE) 15

Argentina (AR) 63 Bangladesch (BD) 3

Jamaica (JM) 2 Pakistan (PK) 3

Trinidad & Tobago

(TT) 3 Sri Lanka (LK) 4

Bosnia Herzegovina

(BA) 7 Vietnam (VN) 14

(31)

29

Appendix 2: Average abnormal returns ‘market model’ day -29 to day +30

Day AARt T1 Art >0 T3 Day AARt T1 Art > 0 T3 -29 0.002 (0.35) 48% (0.03)** +1 0.002 (0.02)** 51% (0.35) -28 0.001 (0.20) 47% (0.01)*** +2 0.000 (0.10)* 48% (0.04)** -27 0.000 (0.28) 48% (0.03)** +3 0.001 (0.20) 48% (0.04)** -26 0.001 (0.10)* 47% (0.01)*** +4 0.000 (0.35) 48% (0.03)** -25 0.001 (0.05)** 51% (0.33) +5 0.000 (0.21) 49% (0.10) -24 0.001 (0.15) 50% (0.13) +6 0.000 (0.26) 48% (0.02)** -23 0.000 (0.13) 48% (0.02)** +7 0.000 (0.35) 46% (0.01)*** -22 0.000 (0.35) 49% (0.09)* +8 0.000 (0.06)* 51% (0.35) -21 0.000 (0.21) 48% (0.04)** +9 0.001 (0.11) 47% (0.01)*** -20 0.001 (0.32) 48% (0.03)** +10 0.000 (0.35) 48% (0.04)** -19 0.001 (0.35) 49% (0.06)* +11 0.000 (0.25) 49% (0.05)** -18 0.000 (0.17) 47% (0.02)** +12 0.000 (0.14) 48% (0.04)** -17 0.000 (0.32) 48% (0.02)** +13 0.000 (0.33) 48% (0.03)** -16 0.001 (0.07)* 49% (0.05)** +14 0.001 (0.12) 46% (0.01)*** -15 0.000 (0.31) 46% (0.01)*** +15 0.000 (0.16) 49% (0.05)** -14 0.000 (0.08)* 47% (0.02)** +16 0.000 (0.25) 49% (0.07)* -13 0.001 (0.12) 47% (0.01)*** +17 0.000 (0.13) 47% (0.01)*** -12 0.000 (0.14) 49% (0.06)* +18 0.001 (0.05)** 47% (0.02)** -11 0.000 (0.34) 48% (0.04)** +19 0.001 (0.16) 49% (0.05)** -10 0.000 (0.09)* 49% (0.06)* +20 0.002 (0.04)** 47% (0.02)** -9 0.000 (0.32) 48% (0.02)** +21 0.000 (0.28) 47% (0.01)*** -8 0.000 (0.33) 47% (0.02)** +22 0.000 (0.33) 49% (0.05)** -7 0.001 (0.26) 48% (0.04)** +23 0.000 (0.19) 48% (0.02)** -6 0.000 (0.14) 47% (0.01)*** +24 0.000 (0.34) 47% (0.02)** -5 0.001 (0.08)* 48% (0.03)** +25 0.000 (0.33) 48% (0.04)** -4 0.000 (0.28) 47% (0.02)** +26 0.001 (0.21) 49% (0.05)** -3 0.000 (0.12) 49% (0.05)** +27 0.001 (0.03)** 47% (0.01)*** -2 0.000 (0.12) 47% (0.01)*** +28 0.001 (0.15) 48% (0.03)** -1 0.000 (0.35) 48% (0.02)** +29 0.001 (0.12) 47% (0.01)*** 0 0.005 (0.00)*** 54% (0.03)** +30 0.000 (0.30) 48% (0.02)**

̂

√ ̂ ̂

.

The p-values are presented.

(32)

30

Appendix 3: Average abnormal returns ‘mean-return model’ day -29 to day +30

Day AARt T1 Art > 0 T3 Day AARt T1 Art > 0% T3 -29 0.001 (0.28) 49% (0.21) +1 0.002 (0.04)** 51% (0.04)** -28 0.001 (0.13) 48% (0.35) +2 0.000 (0.13) 49% (0.15) -27 0.000 (0.35) 48% (0.33) +3 0.001 (0.27) 49% (0.24) -26 0.001 (0.09)* 46% (0.11) +4 0.000 (0.35) 49% (0.21) -25 0.001 (0.17) 50% (0.10) +5 0.000 (0.35) 49% (0.15) -24 0.001 (0.15) 50% (0.09)* +6 0.001 (0.10)* 47% (0.21) -23 0.001 (0.08)* 46% (0.13) +7 0.000 (0.35) 49% (0.20) -22 0.000 (0.30) 49% (0.19) +8 0.001 (0.08)* 49% (0.24) -21 0.000 (0.16) 49% (0.28) +9 0.001 (0.18) 48% (0.35) -20 0.001 (0.35) 49% (0.24) +10 0.000 (0.34) 48% (0.32) -19 0.001 (0.10)* 47% (0.21) +11 0.000 (0.27) 48% (0.29) -18 0.000 (0.20) 48% (0.35) +12 0.000 (0.21) 49% (0.20) -17 0.000 (0.33) 49% (0.19) +13 0.000 (0.15) 48% (0.32) -16 0.001 (0.20) 47% (0.29) +14 0.001 (0.07)* 47% (0.18) -15 0.000 (0.20) 46% (0.14) +15 0.000 (0.18) 49% (0.13) -14 0.001 (0.05)** 47% (0.23) +16 0.000 (0.20) 50% (0.11) -13 0.001 (0.15) 48% (0.35) +17 0.001 (0.05)** 45% (0.04)** -12 0.000 (0.31) 49% (0.26) +18 0.001 (0.15) 46% (0.13) -11 0.000 (0.25) 48% (0.35) +19 0.001 (0.05)** 47% (0.27) -10 0.000 (0.18) 48% (0.34) +20 0.001 (0.11) 48% (0.31) -9 0.001 (0.29) 48% (0.33) +21 0.000 (0.26) 48% (0.32) -8 0.000 (0.34) 49% (0.28) +22 0.000 (0.33) 49% (0.25) -7 0.000 (0.31) 48% (0.30) +23 0.001 (0.12) 48% (0.31) -6 0.000 (0.21) 48% (0.35) +24 0.000 (0.33) 50% (0.10)* -5 0.001 (0.13) 49% (0.27) +25 0.000 (0.33) 49% (0.25) -4 0.001 (0.35) 48% (0.31) +26 0.000 (0.24) 48% (0.35) -3 0.000 (0.14) 48% (0.32) +27 0.001 (0.06)* 47% (0.27) -2 0.001 (0.05)* 47% (0.20) +28 0.001 (0.16) 49% (0.16) -1 0.001 (0.21) 49% (0.12) +29 0.001 (0.04)** 46% (0.12) 0 0.006 (0.00)*** 55% (0.00)*** +30 0.000 (0.32) 47% (0.31)

̂

√ ̂ ̂

.

The p-values are presented.

Appendix 4: Average cumulative abnormal returns ‘mean-return model’

Parametric Non-Parametric

ACAR[t1,t2] Std. Dev. t-statistic P-value CARt > 0 P-value

Window[-29,-1] -0.0002 0.12 -2.50 (0.04)** 48% (0.34)

Window[0,+1] 0.007 0.04 7.15 (0.00)*** 56% (0.00)***

Window[+2,+30] -0.009 0.11 -3.43 (0.02)** 49% (0.24)

√ √ and the non-parametric test is

̂ √ ̂ ̂ .

(33)

31

References

Agarwal, S., & Ramaswami, S. N. 1992. Choice of foreign market entry mode: Impact of ownership, location and internalization factors. Journal of International business studies, 1-27.

Andersen, O. (1997). Internationalization and market entry mode: A review of theories and conceptual

frameworks.MIR: Management International Review, 27-42.

Alchian, A., R. Crawford, and B. Klein. 1978. Vertical Integration, Appropriable Rents, and the Competitive Contracting Process. Journal of Law and Economics 21(2):297–326.

Alexandridis, G., Fuller, K. P., Terhaar, L., & Travlos, N. G. 2013. Deal size, acquisition premia and shareholder gains. Journal of Corporate Finance, 20, 1-13.

Alexandridis, G., Petmezas, D., & Travlos, N. G. 2010. Gains from mergers and acquisitions around the world: New evidence. Financial Management, 39(4), 1671-1695.

Anderson, E., & Gatignon, H. 1986. Modes of foreign entry: A transaction cost analysis and propositions. Journal of international business studies, 17(3), 1-26.

Baker, M., Foley, C. F., & Wurgler, J. (2009). Multinationals as arbitrageurs: The effect of stock market

valuations on foreign direct investment.Review of Financial Studies,22(1), 337-369.

Barkema, H. G., Bell, H. J. H., & Pennings, J. M. 1996. Foreign entry, cultural barriers, and learning. Strategic Management Journal, 17(2): 151–166

Beckman, C. M., & Haunschild, P. R. 2002. Network learning: The effects of partners' heterogeneity of experience on corporate acquisitions. Administrative science quarterly, 47(1), 92-124.

Berger, P. G., & Ofek, E. 1995. Diversification's effect on firm value. Journal of financial economics, 37(1), 39-65.

Bhagat, R. S., Kedia, B. L., Harveston, P. D., & Triandis, H. C. (2002). Cultural variations in the cross-border

transfer of organizational knowledge: An integrative framework.Academy of management review,27(2),

204-221.

Bradley, M., Desai, A., & Kim, E. H. (1988). Synergistic gains from corporate acquisitions and their division

(34)

32

Bresman, H., Birkinshaw, J., & Nobel, R. (1999). Knowledge transfer in international acquisitions. Journal of

international business studies, 439-462.

Brooks, C. 2008, Introductory econometrics for finance, 2nd edition, Cambridge University Press, Cambridge, UK

Brouthers, K. D., & Brouthers, L. E. (2000). Acquisition or greenfield start‐ up? Institutional, cultural and transaction cost influences.Strategic Management Journal,21(1), 89-97.

Brown, S. J., & Warner, J. B. 1980. Measuring security price performance.Journal of financial Economics, 8(3), 205-258.

Brown, S. J., & Warner, J. B. 1985. Using daily stock returns: The case of event studies. Journal of financial economics, 14(1), 3-31.

Chang, S., 1998. Takeovers of privately held targets, method of payment and acquirer returns. Journal of Finance 53, 773–784

Chari, Anusha, Paige P. Ouimet, and Linda L. Tesar. 2010. “The Value of Control in Emerging Markets.” Review of Financial Studies 23:4, 1741–1770.

Child, J., Faulkner, D., & Pitkethly, R. 2001.The management of international acquisitions. Oxford University

Press.

Coase, R. 1937. The Nature of the Firm. Economica 4(16):386–405.

Datta, D. K., & Puia, G. 1995. Cross-border acquisitions: An examination of the influence of relatedness and cultural fit on shareholder value creation in US acquiring firms. MIR: Management International Review, 337-359.

Denis, D. J., Denis, D. K., & Yost, K. (2002). Global diversification, industrial diversification, and firm

value.The Journal of Finance,57(5), 1951-1979.

Doukas, J., & Travlos, N. 1988. The effect of corporate multinationalism on shareholders' wealth: Evidence from international acquisitions. Journal of Finance, 43(5): 1161–1175.

“Do cross-border acquisition announcements to emerging markets create shareholder value; a study considering the effect of legal and financial institutions”