Master’s Thesis IB&M

Master’s Thesis IB&M

EMMNE internationalization: Do full or partial

acquisitions lead to better performance? The

moderating impact of absorptive capacity,

institutional distance, and cultural distance.

Jeremy Eefting – S2876914

University of Groningen

Faculty of economics and business

Word count: 14.010

Master’s Thesis IB&M

EMMNE internationalization: Do full or partial

acquisitions lead to better performance? The

moderating impact of absorptive capacity,

institutional distance, and cultural distance.

ABSTRACT

Purpose: Using a multi home, multi host country approach, this thesis aims to determine under which

circumstances (cultural/institutional), emerging market multinational enterprises (EMMNE) acquisitions mode choice (full vs. partial) lead to better firm performance. The results are compared to traditional IB theories to determine if there are any differences in acquisition strategies between EMMNEs and developed market MNEs (DMMNE). Additionally, this thesis examines whether absorptive capacity moderates the relationship between full acquisitions and firm performance.

Design/Methodology/Approach: Using panel data of cross-border acquisitions by EMMNEs in

developed markets in the years 2012-2018, this thesis tests whether the independent and moderator variables have the hypothesized effect on firm performance. A random effects panel data regression model was used to perform the analysis.

Findings: The results of the thesis show that full acquisitions do not lead to better firm performance

compared to partial acquisitions per se. Furthermore, no evidence was found that there is an interaction between absorptive capacity and full acquisitions on firm performance. EMMNEs that engaged in full acquisitions in institutionally distant countries performed better than those that engaged in partial acquisitions. This defies classic IB theories based on DMMNEs, but is consistent with past research in the context of EMMNEs. EMMNEs that engaged in partial acquisitions in culturally distant countries performed better than those that engaged in full acquisitions, which lends support to past research on entry modes and firm performance.

Originality/Value: This thesis enhances the understanding of whether full- partial acquisitions

undertaken by EMMNEs in developed markets lead to different outcomes in terms of performance. It provides an extra addition to the debate regarding EMMNEs in internationalization theory. Moreover, it identifies under which circumstances (institutional/cultural distance) full acquisitions improve EMMNE performance, which is relevant for EMMNE managers.

KEYWORDS: Emerging markets, EMMNE, CBA, absorptive capacity, institutional distance, cultural

Table of contents

1. Introduction 4

2. Literature review 7

2.1 Internationalization 7

2.2 EMMNE characteristics 8

2.3 EMMNE International Process 9

2.4 Cross-Border Acquisitions and firm performance 10

3. Methodology 18

3.1 Data selection and sample 18

3.2 Dependent variable (ROA) 20

3.3 Independent variable 21

3.3.1 Entry mode (full vs. partial acquisition) 21

3.4 Moderating variables 21 3.4.1 Absorptive capacity 21 3.4.2 Institutional Distance 22 3.4.3 Cultural Distance 22 3.5 Control variables 23 3.6 Analysis 25 4. Results 27 4.1 Sample description 27

4.2 Description and assumptions 28

4.3 Regression results 31

4.4 Robustness 35

5. Discussion 38

6. Conclusion 42

6.1 Implications for theory 42

6.2 Implications for practice 43

6.3 Limitations and future research 44

7. Literature list 47

1. INTRODUCTION

Emerging market multinational enterprises (EMMNEs) have become more dominant in the global market during the last 15 years. Back in 1995, there were only 20 EMMNEs on the Global Fortune 500 list. Over the last two decades, this number grew to 164 in 2017 (Figure 1). The vast majority of

these EMMNEs originate from the BRIC (Brazil, Russia, India, China) and VISTA (Vietnam, Indonesia, South Africa, Turkey, Argentina) countries (Luo & Tung, 2018). As the importance of EMMNEs on the world stage grew, so did the amount of research on their expansion behavior (e.g. Luo & Tung, 2007; Cuervo-Cazurra, 2012; Mathews, 2006). This has expanded our knowledge about their motives and strategies.

Figure 1. Number of EMMNEs in Fortune Global 500 list

Source: Luo & Tung, 2018

Deciding the “right” entry mode is of great strategic importance. Antecedents, processes, and outcomes of different entries differ per entry mode (Cheng, 2008; Chen & Hennart, 2004). Cross-border acquisitions (CBAs), the seemingly preferred choice of EMMNEs (Luo & Tung, 2007), provide the acquiring firm with access to tacit foreign resources that cannot be acquired through the market, which can be recombined with proprietary resources and improve the firm’s performance (Tuch & O’Sullivan, 2007). On the other hand, acquisitions are riskier than lower stake entry modes, due to high initial investments and transaction costs (Johanson & Vahlne, 1977; Anderson & Gatignon, 1986). These costs are amplified by cultural and institutional differences between the home- and host countries (Kostova & Zaheer, 1999). Research that focussed on DMMNEs finds that CBAs usually decrease firm performance (Campa & Hernando, 2004; Gupta & Misra, 2004), while others find positive (Dickerson et al., 1997), or insignificant results (Walker, 2000).

The impact of full versus partial CBAs on firm performance for EMMNEs is under researched compared to DMMNEs (Yang, 2015). Research on CBAs undertaken by EMMNEs mainly finds a positive impact on firm performance (Boateng et al., 2008; Gubbi et al., 2010). The difference could be explained by the argument that EMMNEs venture abroad to acquire strategic resources and improve their legitimacy (Luo & Tung, 2007). Differences in culture and institutional quality can be recognized to improve current operations (Yang, 2015), rather than obstacles that could be reduced by lower equity entry modes (Johanson & Vahlne, 1977). The ability to internalize these foreign resources and recombine them to improve current operations depends on the firm’s absorptive capacity (Cohen & Levinthal, 1990).

Research Question: Do EMMNEs that engage in full acquisitions in developed markets show better

financial performance than those that engage in partial acquisitions? And do cultural distance,

institutional distance, and absorptive capacity moderate this relationship?

This study contributes to the literature by extending our knowledge of the relationship between firm performance and entry strategies, which is prominently studied in the context of DMMNEs. This may also shed light on the usefulness of existing IB theories about internationalization strategies and firm performance, and whether the current theories are of generic characteristics that can apply in both developed and emerging markets

This thesis uses acquisition data from EMMNEs from the BRIC (Brazil, Russia, India, China) and VISTA (Vietnam, Indonesia, South Africa, Turkey, Argentina) in the years 2012-2018 to determine the impact of full acquisitions on EMMNE performance compared to partial acquisitions. The thesis further examines whether the relationship between the type of acquisition and firm performance is moderated by the firm’s absorptive capacity, cultural-, and institutional distance. The results show that the decision to fully acquire a target firm in a developed market does not improve performance per se. This thesis does not find any evidence that absorptive capacity has a moderating impact on the relationship between full acquisitions and firm performance. EMMNEs engaging in full acquisitions performed marginally worse in culturally distant countries, which is in line with traditional IB literature. EMMNEs that engaged in full acquisitions in institutionally distant countries performed marginally better, which contradicts traditional IB literature, but lends support to recent research on EMMNE entry modes and performance.

2. LITERATURE REVIEW & HYPOTHESES

2.1 Internationalization

According to Hennart (2012), a Multinational Enterprise (MNE) can be defined as “a firm that produces goods and services in a foreign country with their own employees, as opposed to firms that export to these countries or that license or franchise producers located there”. The behavior of MNEs has been central to the field of IB. Johanson and Vahlne (1977) define internationalization as a process in which firms gradually increase their international involvement. Being present in a foreign country, the MNE experiences extra costs that local firms do not encounter. These disadvantages are labeled as the “liability of foreignness” (Zaheer, 1995). These costs stem from (1) costs associated with spatial distance, such as travel, transportation, and communication costs; (2) firm-specific costs due to unfamiliarity with and a lack of roots in the host country; (3) costs resulting from the host country environment, such as a lack of legitimacy of foreign firms and economic nationalism; and (4) costs of the home country environment, such as export and import restrictions (Zaheer, 1995). Since there are so many extra costs involved in the process of internationalization, there must be advantages that offset these costs. Otherwise, the firm would be better off by staying at home.

2.2 EMMNE characteristics

The abovementioned frameworks were introduced at a time when the focus of IB was mainly on DMMNEs, as these firms predominantly engaged in internationalization. Over the past two decades, MNEs from emerging markets established a greater presence in the global market. Emerging markets are characterized by rapidly growing national economies, dramatic structural industrial changes, and promising markets despite volatility and weak legal systems (Hoskisson et al., 2000; Marquis & Raynard, 2015; Luo & Tung, 2007). Emerging markets are not entirely homogenous, but enterprises in these countries face some similar constraints, share similar motives, and have common experiences in international business (Luo & Tung, 2007).

These different findings on the ability of EMMNEs to internationalize have caused a debate in the IB field, also known as the “goldilocks debate” (Cuervo-Cazurra, 2012). One stream of the literature argues that EMMNEs are different from DMMNEs and thus require other frameworks than the “traditional frameworks” (Luo & Tung, 2007; Mathews, 2002). Other researchers argue that traditional frameworks are also applicable to EMMNEs (Verbeke & Kano, 2015). A third stream of the literature states that traditional frameworks can partly explain the internationalization behavior of EMMNEs, but that it needs some adjustment (Ramamurti, 2009).

2.3 EMMNE internationalization process

As mentioned earlier, the literature on the internationalization process suggests that MNEs are more likely to invest in countries that are geographically, culturally, and institutionally proximate to the home country through entry modes that require relatively low initial commitment (Johanson & Vahlne, 1977; Benito et al., 2009). This pattern, which is primarily observed in DMMNEs, hardly applies to the aggressive internationalization strategies of EMMNEs in developed markets (Gaffney et al., 2016).

Rui, 2009). In this thesis, the focus will be only on the entry into developed markets, as this seems to be the dominant strategy (Deng, 2004).

A second aspect of the IP model that EMMNEs seem to defy, is the speed at which they internationalize (Madhok & Keyhani, 2012). More specifically, Johanson and Vahlne (1977) argue that firms incrementally increase their entry mode strategy in a foreign country once they have gained experience in the host country. This implies that MNEs first opt for an exporting or licensing oriented strategy and gradually increase resources and commitment if operations in the host country go well. As mentioned before, internationalization is a costly and hardly irreversible strategy, surrounded by uncertainty and risk stemming from the liability of foreignness (Zaheer, 1995), which explains why firms often test the waters first before entering a new market through an acquisition. EMMNEs do not seem to follow this path as they invest aggressively in acquisitions in distant developed markets (Deng, 2004; Luo & Tung, 2007; Rui & Yip, 2008).

2.4 Cross-border acquisitions and firm performance

Research has shown that choosing the “right” entry mode has a significant impact on firm performance (e.g. Contractor et al., 2014; Brouthers, 2013; Brouthers & Brouthers, 2003). Most of these papers take the transaction costs of entry modes in foreign countries(Anderson & Gatignon, 1986; Brouthers, 2013), or the impact of the resource-based-view in determining which entry mode (Morosini et al., 1998; Vermeulen & Barkema, 2001) into consideration. Since EMMNEs seem to defy traditional entry mode patterns, it is interesting to see whether the chosen entry strategies improve their performance. After all, some EMMNEs have become very successful in their industry using this strategy (Luo & Tung, 2007; Ramamurti, 2012). The growing involvement of EMMNEs in CBAs is partly driven by globalization (Shimizu et al., 2004), but also by deregulatory policies imposed by emerging market governments (Aybar & Ficici, 2009).

& O’Sullivan, 2007). MNEs acquire foreign firms to quickly overcome the entry barriers of potentially attractive foreign markets (Vermeulen & Barkema, 2001), acquire new skills, knowledge, and resources that are not available at the home market (Ahuja & Katila, 2001), and gain access to knowledge derived from foreign location advantages (Makino & Delios, 1996). As a result, the firm develops internal learning capabilities that allow for integration and recombination of acquired resources with proprietary resources (Zollo & Singh, 2004; Zahra & George, 2002). These recombined resources can be utilized to develop or strengthen competitive advantages, which may improve firm performance (Zahra & George, 2002).

However, compared to the acquisition of a firm in the home country, CBAs are more challenging, especially in the post-acquisition phase (Child, 2001). Considering the ambiguity of CBAs, it is not surprising that the results in terms of its impact on performance are mixed, at least for DMMNEs (Tuch & O’Sullivan, 2007). Several papers based on short-term shareholder value report significant negative returns (Campa & Hernando, 2004; Gupta & Misra, 2004; Sudarsanam & Mahate; 2003), while others found insignificant results (Walker 2000; Song & Walking, 2004). Researchers who considered accounting performance as a measurement of firm performance also found varying results. Lu (2004) found that performance significantly decreased after the acquisition. Dickerson et al. (1997) found that DMMNEs that engaged in CBAs outperformed those that didn’t.

When acquiring a foreign target firm, high-level managers need to decide on an appropriate ownership level; a correctly assessed ownership level can result in significant benefits through post-acquisition integration and assets synergies (Contractor & Lahiri, 2014). A wrongly assessed ownership level may lead to a mismatch between resource commitment and risk, suboptimal integration of the target firm, and less-than-desired rent appropriation from the acquisition (Contractor & Lahiri, 2014). Prior IB literature distinguishes between full acquisitions, where the acquirer buys more than 95% of the stakes in the target firm, and partial acquisitions where less than 95% of the stakes are bought (Brouthers & Nakos, 2004; Hollender et al., 2017). The decision to engage in a full acquisition or a partial acquisition involves different strategic considerations, as each of these entry strategies has its advantages and disadvantages (Chen & Hennart, 2004; Lopez-Duarte & Garcia-Canal, 2004).

Full acquisitions are the highest form of commitment in terms of (financial) resources when entering a foreign country (Johanson & Vahlne, 1977). The purchase of a target firm as a whole provides the acquirer complete control over its resources, assets, and returns (Yang, 2015). Furthermore, it gives the MNE complete operational freedom to implement its own strategies in the target firm (Contractor & Lahiri, 2014). Additionally, full acquisitions allow for quicker access to desired resources than partial acquisitions, which speeds up the development of synergetic value (Deng, 2004).

from the acquisition, the risk of opportunistic behavior from the target, and a limited ability to fully integrate its operations (Yang, 2015; Anderson & Gatignon, 1986; Hennart, 1991).

When deciding the level of ownership in a foreign target, the EMMNE needs to carefully weigh the costs of the investment (including risks) against the potential benefits and determine which acquisition strategy is best. Ample literature has shown that among EMMNEs strategic asset seeking is the most common reason to acquire firms in developed countries (UNCTAD, 2006; Makino et al., 2002; Teece et al., 1997). According to Wesson (2004), Deng (2009), and Rui and Yip (2008), EMMNEs use their aggressive acquisition strategy to acquire the knowledge and resources that reduce their latecomer status and overcome their lack of sophisticated resources. Furthermore, Gubbi et al. (2010) found that MNEs from India achieved significantly better performance after they internalized tangible and intangible assets from developed markets that are difficult to access through market transactions. A full acquisition provides the EMMNE with quicker access to resources and more strategic control over the target firm’s assets (Yang, 2015; Contractor et al., 2014) compared to a partial acquisition, allowing for greater potential benefits. Based on the abovementioned literature, the following hypothesis is derived:

Hypothesis 1: EMMNEs that engage in a full acquisition in a developed market perform better than

those that engage in a partial acquisition in a developed market, ceteris paribus.

However, EMMNEs differ in the extent to which they can assimilate knowledge from developed markets and use it to create synergetic value M&As (Zollo & Singh, 2004; Deng, 2009). Simply being exposed to tacit knowledge that is hard to extract is not enough to generate synergistic value; the firm needs absorptive capacity (Amit & Schoemaker, 1993; Guo et al., 2017)

diversity of the firms, increasing the knowledge available for problem-solving and improves the ability to cope with environmental change (Grimpe & Sofka, 2009). Second, for a given quantity of identified external knowledge, the degree to which firms can extract benefits from external knowledge depends on their absorptive capacity (Guo et al., 2017). This explains why varying levels of absorptive capacity can influence achieved levels or returns (Volberda et al., 2010). In other words, firms that possess greater absorptive capacity will generate more value from absorbed knowledge (Tsai, 2001).

It is argued that EMMNEs face extra barriers to develop absorptive capacity compared to DMMNEs (Cuervo-Cazurra & Rui, 2017; Awate et al., 2012, 2015). A lack of home country information intermediaries, poor implementation of resources, and home government protectionism hinder the development of absorptive capacity (Cuervo-Cazurra & Rui, 2017; Khanna & Palepu, 2010; Djankov et al., 2002). Despite these barriers, Deng (2009) and Bilgili et al. (2016) found that those EMMNEs that successfully grew their business after acquisitions in developed markets did so by investing in R&D and strong combinative skills, which are pre-requisites for absorptive capacity. Considering EMMNEs expand to developed markets to acquire strategic assets and resources, those EMMNEs that have greater absorptive capacity are more likely to extract and recombine these resources, amplifying the positive effect of full acquisitions on firm performance.

Hypothesis 2: Absorptive capacity will positively moderate the relationship between full acquisitions

and firm performance.

Massoud, 2008), and weaker investor protection procedures (Pagano et al., 2002), compared to developed economies.

As a result, the transaction costs of operating in an institutionally distant country increases due to unfamiliarity with local businesses, regulations, institutions, and a heightened complexity to monitor and maintain business relationships with suppliers and local customers (Dikova et al., 2010; Peng et al., 2008). To mitigate these risks, MNEs often opt for lower equity entry modes, where the local partner could assist them in establishing legitimacy to overcome institutional uncertainty (Eden & Miller, 2004). IB research that traditionally focuses on the relationships between DMMNE entry mode strategy and firm performance supports this notion. Dikova et al. (2010) empirically show that the likelihood of an MNE completing a CBA decreases when the institutional distance increases. Furthermore, Xu et al. (2004) found that a high institutional distance correlates with a low ownership acquisition. Shirodkar and Konara (2017) empirically showed that DMMNEs that entered markets characterized by a high institutional distance experienced a decrease in firm performance, but this reduction was reduced when they entered through partial ownership. The outcome of these researches provides support for Johanson and Vahlne’s (1977) internationalization process theory.

such a correlation and showed that these firms perform better than those that engaged partial acquisitions. Therefore, I hypothesize the following:

Hypothesis 3: When institutional distance increases, EMMNEs that engage in full acquisitions perform

better than those that engage in partial acquisitions; institutional distance positively moderates the

relationship between full acquisitions and firm performance.

Another complicating factor for CBAs is the difference in national culture between acquiring and acquired firms. The main argument is that cultural distance, which is the extent to which cultural values between the home- and the host market are different (Shenkar, 2001), increases the costs of operating in the host market (Anderson & Gatignon, 1986). Firms entering a foreign market face additional difficulties and challenges due to a lack of knowledge and understanding of the host country (Beugelsdijk et al., 2018). A lack of familiarity with local values and customs, as well as diverse social routines and implicit assumptions, require adjustment and adaptation (Chakrabarti, 2009; Hofstede, 2001; Morosini et al., 1998). Moreover, the high cultural distance increases the challenges to establish, monitor, and maintain relationships with local management and various stakeholders, which increases the transaction costs of the acquisition (Contractor et al., 2014).

To overcome the costs of cultural distance, firms traditionally opt for lower-equity entry modes to lower the exposure to the risk and ambiguity surrounding internationalization (Johanson & Vahlne, 1977). Empirical studies show that firms (both EMMNEs and DMMNEs) are more likely to engage in partial acquisitions than full acquisitions in culturally distant countries (Barkema & Vermeulen 1998; Tihanyi et al., 2005; Contractor et al., 2014). The main advantage of a full acquisition is the complete strategic control over the target’s resources, but unfamiliarity with the target’s national culture complicates the integration and internalization process (Randoy & Dibrell, 2002). A partial acquisition, on the other hand, allows the firm more time to familiarize itself with the host market and requires less cost ex-ante, which reduces the risk of international expansion (Contractor et al., 2014).

policies and strategies by an EMMNE than vice versa (Yildiz & Fey, 2016). All in all, cultural distance reduces the ability of the EMMNE to quickly extract strategic resources from the foreign target (Slangen, 2006), which diminishes the main advantage of a full acquisition. Therefore:

Hypothesis 4: When cultural distance increases, EMMNEs that engage in full acquisitions perform

worse than those that engage in partial acquisitions; cultural distance negatively moderates the

relationship between full acquisitions and firm performance.

While the primary objective of this thesis is to distinguish the different impacts of full versus partial acquisitions on firm performance, the overall performance after the acquisition will be controlled for. Du and Boateng (2012) found that the majority of studies on EMMNE acquisitions finds a positive impact on firm performance. Therefore, it is expected that firm performance improves in the years after the acquisition.

Figure 2. Conceptual model

This conceptual model is developed based on this thesis’ hypotheses. It states that it is expected that EMMNEs that engage in a full acquisition attain a higher return on assets than those that engage in a partial acquisition. This relationship is positively moderated (strengthened) when the firm has a higher level of absorptive capacity and when institutional distance increases. The relationship between entry mode and return on assets is negatively moderated by cultural distance, indicating that when the cultural distance increases, the advantages of a full acquisition are weakened compared to a partial acquisition.

3. METHODOLOGY 3.1 Data selection and sample

For this thesis firm- and country-level data is used. While there are different classifications for emerging countries (Hoskisson et al., 2013), which often includes a wide variety of countries, the focus of this thesis lies on MNEs of countries that are most active in acquisitions in developed markets. Therefore, the BRIC (Brazil, Russia, India, China) and VISTA (Vietnam, Indonesia, South Africa, Turkey, Argentina) countries are selected for this analysis (Hennart, 2012; Winterhalter, 2017). Developed markets were selected based on the World Economic Situation and Prospects report (2014) of the United Nations (see Figure 3).

Nicholson, 2016). Using the firm identifier (ISIN) of the acquiring firm, firm-level data was extracted from the Orbis database, which also integrates multiple databases by Bureau van Dijk. The Orbis database has data on around 200 million firms worldwide. This data consists mainly of firm-specific data, ranging from firm descriptive data to financial data and ratios. Only firms that had data on R&D intensity, Return On Assets (ROA), and total revenue at the year of the acquisition and the two years prior and after the acquisition were included in the sample. Data on the cultural distance between the home- and host country is retrieved from the Hofstede-Insights website, which offers a tool to extract the country scores on the six Hofstede dimensions. The institutional distance scores are retrieved from the World Governance Indicators website, where the scores of each country on the WGI indicators are listed. Lastly, the GDP per capita scores of the emerging markets are extracted from the World Bank website.

Figure 3. UN classification of developed countries

Europe European Union New EU member states Other Europe Other countries Major developed economies G7

EU-15 Bulgaria Iceland Australia Canada

Austria Croatia Norway Canada Japan

Belgium Cyprus Switzerland Japan Germany

Denmark Czech Republic

New Zealand

Italy

Finland Estonia United

States

United Kingdom

France Hungary United States

Germany Latvia Greece Lithuania Ireland Malta Italy Poland Luxembourg Romania Netherlands Slovakia Portugal Slovenia Spain Sweden United Kingdom

3.2 Dependent variable (Firm performance (ROA))

The dependent variable for this research is firm performance. To measure firm performance, researchers have used a variety of metrics. For this thesis, the accounting metric Return On Assets (ROA) is used, which is calculated by dividing the firm’s Earnings Before Interest and Taxes (EBIT) by the firm’s total assets. For the analysis in this thesis, following Krishnan et al. (1997), ROA is measured in the year of the acquisition and the two years after the acquisitions. The underlying rationale is that an acquisition does not immediately increase firm performance, as it will take some time before the firm can internalize the partner’s resources and improve its operations (Krishnan et al., 1997; Tuch & O’Sullivan, 2007). ROA is widely used in the international business, and mergers and acquisitions literature (e.g. Haar, 1989; Ramaswamy, 1995; Gomes & Ramaswamy, 1999; Kostopoulos et al., 2001; Contractor et al., 2007; Lu, 2004). The underlying rationale is that the synergistic benefits from the acquisition will eventually appear in the firm’s accounting records (Tuch & O’Sullivan, 2007).

3.3. Independent variable

3.3.2 Entry mode (full vs. partial acquisitions).

The independent variable (entry mode) reflects the percentage of stakes bought by the EMMNE. The percentages range from 10-100. To make the distinction between partial- and full acquisition, a dummy variable is created. This method has been used in IB literature before, for example by Hollender et al. (2017) and Brouthers and Nakos (2004). Partial acquisitions (≤ 95%) will be labeled "0" and full acquisitions (> 95%) will be labeled "1". Some researchers argue that this method is too broad since the level of commitment between a 20% acquisition and an 80% acquisition differ greatly (Contractor et al., 2014). While this argument is understandable, the focus of this thesis is to distinguish between firms that have complete control over the resources of the target and those that do not. Different acquisition stakes will be considered in the robustness check. The data is extracted from the Zephyr database. It is expected that EMMNEs that engage in a full acquisition attain better financial performance than those who engage in partial acquisitions, ceteris paribus.

3.4 Moderating variables

3.4.1 Absorptive capacity

absorptive capacity, it is well-grounded in the literature (Veugelers, 1997; Stock et al., 2001; Tsai, 2001). Therefore, it is expected that absorptive capacity positively moderates the relationship between full acquisitions and EMMNE performance.

3.4.2 Institutional distance

To measure the institutional differences between countries, the World Governance Indicator (WGI) dimensions are used. The index has been used before to determine cross-country institutional differences (Van Hoorn & Maseland, 2016; Law et al., 2013) and has proven to be useful to make cross-country comparisons (Kraay et al., 2010). The following dimensions make up the WGI: Voice and Accountability, Political Stability and Absence of Violence/Terrorism, Government Effectiveness,

Regulatory Quality, Rule of Law, and Control of Corruption. To determine the institutional distance,

the following model is used:

𝐼𝑛𝑠𝑡𝑖𝑡𝑢𝑡𝑖𝑜𝑛𝑎𝑙 𝑑𝑖𝑠𝑡𝑎𝑛𝑐𝑒 = ∑{𝐼𝑖𝑗− 𝐼𝑖ℎ𝑜𝑚𝑒) 2 /𝑉𝑖}/6 𝑖 𝑖=1 Where: - 𝐼𝑖𝑗 = 𝑇ℎ𝑒 𝑖𝑛𝑑𝑒𝑥 𝑓𝑜𝑟 𝑡ℎ𝑒 𝑖𝑡ℎ 𝑑𝑖𝑚𝑒𝑛𝑠𝑖𝑜𝑛 𝑎𝑛𝑑 𝑡ℎ𝑒 𝑗𝑡ℎ 𝑐𝑜𝑢𝑛𝑡𝑟𝑦 - 𝑣𝑖 = 𝑇ℎ𝑒 𝑣𝑎𝑟𝑖𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 𝑡ℎ𝑒 𝑖𝑡ℎ 𝑑𝑖𝑚𝑒𝑛𝑠𝑖𝑜𝑛 (variance based on all country scores in

dimension i)

It is expected that institutional distance positively moderates the relationship between full acquisitions and financial performance.

3.4.3 Cultural distance

To measure the cultural differences between countries, the Hofstede Dimensions are used. The framework consists of the following dimensions: power distance, individualism, masculinity, uncertainty avoidance, long term orientation, and indulgence (Hofstede, 1984; Hofstede & Minkov,

𝐶𝑢𝑙𝑡𝑢𝑟𝑎𝑙 𝑑𝑖𝑠𝑡𝑎𝑛𝑐𝑒 = ∑{𝐼𝑖𝑗− 𝐼𝑖ℎ𝑜𝑚𝑒) 2 /𝑉𝑖}/6 𝑖 𝑖=1 Where: - 𝐼𝑖𝑗 = 𝑇ℎ𝑒 𝑖𝑛𝑑𝑒𝑥 𝑓𝑜𝑟 𝑡ℎ𝑒 𝑖𝑡ℎ 𝑑𝑖𝑚𝑒𝑛𝑠𝑖𝑜𝑛 𝑎𝑛𝑑 𝑡ℎ𝑒 𝑗𝑡ℎ 𝑐𝑜𝑢𝑛𝑡𝑟𝑦 - 𝑣𝑖 = 𝑇ℎ𝑒 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑖𝑡ℎ 𝑑𝑖𝑚𝑒𝑛𝑠𝑖𝑜𝑛 (variance based on all country scores in

dimension i)

The use of this Euclidian distance to calculate cultural distance has been subject to criticism (Shenkar, 2001). Firstly, it assumes that the distance between the two countries is symmetrical, meaning the distance between country A and B is the same as between country B and A. Second, it assumes that the distance remains stable over time, not accounting for cultural change. Maseland et al. (2018) found that cultures do change, but in parallel, meaning that the relative distance remains unchanged. Third, it assumes that all cultural dimensions have the same weight. For acquisitions, it might be that some dimensions are more disruptive than others (Contractor et al., 2014). Despite the limitations, the cultural distance concept still holds value in entry mode and firm performance research (Maseland et al., 2018). It is expected that cultural distance negatively moderates the relationship between full acquisitions and financial performance.

3.5 Control variables

Post-acquisition period

Firm size

The firm’s size will be controlled for, as larger firms possess more (financial) resources and are therefore more likely to engage in acquisitions, as opposed to lower-stake entry modes (Inkpen & Beamish, 1997). Larger firms can better exploit economies of scale and scope and have higher bargaining power (Mansfield, 1962). On the contrary, smaller firms can implement changes more easily, as they suffer less from organizational inertia (Yang & Cheng, 2009). As a measure of firm size, total revenue is used (Dang et al., 2018). It is expected that larger firms are more able to derive higher ROA.

Industry

Furthermore, this thesis controls for industry factors, dividing the firms into service and manufacturing firms. To make this distinction, a dummy variable was created. Manufacturing firms are labeled “0” and service firms were labeled “1”. It is argued that firms in different industries have a different preference for acquisition stakes (Brouthers & Brouthers, 2003), and firms in the service sector attain higher ROA (Rumelt, 1991). It is expected that service firms report better financial performance than manufacturing firms.

Age

Moreover, a firm's age might impact cross-national operations and performance (Zahra & Hayton, 2008). It is argued that experience comes with age and older firms are therefore more capable of handling an acquisition. Age is measured by retrieving the year of establishment from the Orbis database (Brouthers, 2002). It is expected that older firms are more experienced and are therefore more likely to achieve better financial performance.

GDP per capita

Multiple acquisitions

When firms engage in multiple acquisitions, they develop learning and recombination abilities that make them more efficient when they engage in an additional acquisition (Buckley et al., 2014). Although the costs and risk of buying multiple target firms increase as well, it is expected that those firms that engage in multiple acquisitions will attain a higher ROA.

3.6 Analysis

To test the abovementioned hypotheses, a panel data regression is used. A panel data analysis consists of at least two dimensions: a cross-sectional dimension indicated by the subscript i, and a time series dimension indicated by the subscript t (Hsiao, 2007). The acquisitions happened in different years across the 2012-2018 time period. Hence, observations based on firm data are not all present in the same years. As a result, the panel data is likely to be unbalanced. To determine whether a fixed-effects or a random-effects model is appropriate for this thesis, a Hausman test is performed.

Models:

1. 𝑅𝑂𝐴𝑖𝑡 = 𝛽0 + 𝛽1 ∙ 𝐸𝑁𝑇𝑅𝑌𝑖𝑡+∈

Where ROA stands for return on assets, and ENTRY is a dummy variable that takes “1” for a full acquisition and “0” for a partial acquisition. i stands for firm and t stands for year. In the subsequent models we add more variables, as below. ∈ is the error term.

2. 𝑅𝑂𝐴𝑖𝑡 = 𝛽0 + 𝛽1 ∙ 𝐸𝑁𝑇𝑅𝑌𝑖𝑡+ 𝛽2 ∙ 𝐴𝐶𝐴𝑃𝑖𝑡−2+ ∈

Where ACAP is the firm’s R&D intensity, to measure absorptive capacity, measured at t-2, where t0 is the year of acquisition.

3. 𝑅𝑂𝐴𝑖𝑡 = 𝛽0 + 𝛽1 ∙ 𝐸𝑁𝑇𝑅𝑌𝑖𝑡+ 𝛽2 ∙ 𝐴𝐶𝐴𝑃𝑖𝑡−2+ 𝛽3 ∙ 𝐸𝑁𝑇𝑅𝑌𝑖𝑡∗ 𝐴𝐶𝐴𝑃𝑖𝑡−2+∈

Where ENTRY * ACAP is the interaction between absorptive capacity and entry mode. 4. 𝑅𝑂𝐴𝑖𝑡 = 𝛽0 + 𝛽1 ∙ 𝐸𝑁𝑇𝑅𝑌𝑖𝑡+ 𝛽2 ∙ 𝐴𝐶𝐴𝑃𝑖𝑡−2+ 𝛽3 ∙ 𝐸𝑁𝑇𝑅𝑌 ∗ 𝐴𝐶𝐴𝑃𝑖𝑡−2+ 𝛽4 ∙ 𝐼𝑁𝑆𝑇𝑖𝑡+

𝛽5 ∙ 𝐸𝑁𝑇𝑅𝑌 ∗ 𝐼𝑁𝑆𝑇𝑖𝑡+ 𝛽6 ∙ 𝐶𝑈𝐿𝑇𝑈𝑅𝐸𝑖𝑡+ 𝛽7 ∙ 𝐸𝑁𝑇𝑅𝑌 ∗ 𝐶𝑈𝐿𝑇𝑈𝑅𝐸𝑖𝑡+∈

5. Where INST is the institutional distance between the home and host country. ENTRY*INST is the interaction between institutional distance and entry mode. CULTURE is the cultural distance between the home and host country. ENTRY*CULTURE is the interaction between cultural distance and entry mode.

6. 𝑅𝑂𝐴𝑖𝑡 = 𝛽0 + 𝛽1 ∙ 𝐸𝑁𝑇𝑅𝑌𝑖𝑡+ 𝛽2 ∙ 𝐴𝐶𝐴𝑃𝑖𝑡−2+ 𝛽3 ∙ 𝐸𝑁𝑇𝑅𝑌 ∗ 𝐴𝐶𝐴𝑃𝑖𝑡−2+ 𝛽4 ∙ 𝐼𝑁𝑆𝑇𝑖𝑡+ 𝛽5 ∙ 𝐸𝑁𝑇𝑅𝑌 ∗ 𝐼𝑁𝑆𝑇𝑖𝑡+ 𝛽6 ∙ 𝐶𝑈𝐿𝑇𝑈𝑅𝐸𝑖𝑡+ 𝛽7 ∙ 𝐸𝑁𝑇𝑅𝑌 ∗ 𝐶𝑈𝐿𝑇𝑈𝑅𝐸𝑖𝑡+ 𝛽8 ∙ 𝑃𝐴𝐶𝑄 + 𝛽9 ∙ 𝑆𝐼𝑍𝐸𝑖𝑡+ 𝛽10 ∙ 𝐴𝐺𝐸𝑖𝑡+ 𝛽11 ∙ 𝐺𝐷𝑃𝑖𝑡+ 𝛽12 ∙ 𝑆𝐸𝑅𝑉𝑖𝑡 + 𝛽13 ∙ 𝑀𝐴𝐶𝑄+ ∈

Where PACQ refers to the years after the acquisition, SIZE is the firm size measured by total operating revenue, AGE is the firm age measured at time t, derived from the year of establishment. GPD is the home country’s GDP per capita, SERV is a dummy variable that takes “1” when the firm is operating in the service industry and “0” if it does not. Lastly, MACQ is a dummy variable that takes “1” if the firm acquired multiple firms in the observed time period and “0” if it did not. 7. 𝑅𝑂𝐴𝑖𝑡 = 𝛽0 + 𝛽1 ∙ 𝐸𝑁𝑇𝑅𝑌𝑖𝑡+ 𝛽2 ∙ 𝐴𝐶𝐴𝑃𝑖𝑡−2+ 𝛽4 ∙ 𝐼𝑁𝑆𝑇𝑖𝑡+ 𝛽5 ∙ 𝐸𝑁𝑇𝑅𝑌 ∗ 𝐼𝑁𝑆𝑇𝑖𝑡+ 𝛽6 ∙

Where the interaction between ENTRY and ACAP is excluded.

Although coefficients of common variables are expressed by the same letter in each equation, their estimated variables are likely to differ, which also holds for the error term.

4. RESULTS

4.1 Sample description

After withdrawing the M&A data from Zephyr and the firm-level data from Orbis, the final sample included 223 merger deals where the EMMNE engaged in a full or partial acquisition in a developed market. 45 of deals were classified as partial acquisitions (20.15%) and 178 were classified as full acquisitions (79,85%) (Table 1). Since the difference between the two groups is rather large, the generalizability of the results has to be taken with caution, and a robustness test was performed to test the sensitivity of the results. The vast majority of these EMMNEs originate from China (51.49%) followed by India with 26.27%. Target firms stemming from the US were most popular (28.81%), followed by the UK (15.82%) and Germany (15.97%). Table 2 provides the complete overview of the EMMNEs' home countries that were included and Table 3 provides the overview of the targeted host countries.

Table 1. Summary of acquisitions Table 2. Summary of EMMNE home countries

Freq. Percent Cum. Partial acquisition 135 20.15 20.15 Full acquisition 535 79.85 100.00 Total 670 100.00 EMMNE Home country

Freq. Percent Cum.

Table 3. Summary of target countries

Target Country

Freq.

Percent

Cum.

Austria

6

0.90

0.90

Australia

33

4.93

5.83

Belgium

6

0.90

6.72

Canada

6

0.90

7.61

Germany

107

15.97

23.58

Denmark

9

1.34

24.93

Spain

27

4.03

28.96

Finland

12

1.79

30.75

France

18

2.69

33.43

Great Britain

107

15.82

49.40

Ireland

6 0.90

50.30

Italy

48

7.16

57.46

Japan

3

0.45

57.91

Croatia

3

0.45

58.36

Luxemburg

12

1.79

60.15

Malta

3

0.45

60.60

Netherlands

27

4.03

64.63

Norway

6

0.90

65.52

New Zealand

6

0.90

66.42

Poland

6

0.90

67.31

Portugal

3

0.45

67.76

Sweden

9

1.34

69.10

Singapore

14

2.09

71.19

United States

193

28.81

100.00

Total

670

100.00

4.2 Description and assumptions

forms. When two or more predictors are included in a regression model, it is important that a strong correlation between predictors, also known as multicollinearity, is absent (Field, 2013). In general, multicollinearity was not an issue in the sample, as the correlation between independent variables is often low. As shown in Table 5, an exception is a correlation between the interaction variable full

acquisition*absorptive capacity and the moderator variable absorptive capacity (0.937). To further

Table 4. Descriptive Statistics

Variables Obs Mean Std. Dev. Min Max

ROA 668 7.828085 11.0027 -79.004 63.57 Full acquisition 670 .7985075 0.4014147 0 1 Absorptive capacity (ACAP) 660 2.532492 4.503939 0 26.01 Full acquisition * ACAP 660 2.116882 4.418331 0 26.01 Institutional distance (ID) 670 1.80e-08 7.811082 -18.08115 19.91041 Full acquisition * ID 670 .0038513 6.850261 -14.39215 19.91041 Cultural distance (CD) 670 -1.96e-08 7.355826 -11.52692 18.30219 Full acquisition * CD Post-acquisition period 670 670 0.0310358 .9985075 6.623481 .8174101 -11.52692 0 18.30219 2 Firm size 670 6377500 2.85e+07 16.81842 3.60e+08

Age 670 25.61045 17.14535 2 98 GDP per capita 670 6743.74 3333.123 1449.606 15974.65 Services 670 .3358209 0.4726295 0 1 Multiple acquisitions 670 .3119403 0.4636317 0 1

Table 5. Pairwise correlations

Table 6. Variable Inflation Factors

Table 7. Variable Inflation Factors without “full acquisition * absorptive capacity*

4.3 Regression results

After setting up the data, it became clear that the panel data was unbalanced, meaning that not all variables have information for the same years. As a result, the model contains an additional error term “e”, which stems from the random effect term and might lead to less optimal estimates compared to balanced panel data (Baltagi, 2005). To determine whether a fixed or random-effects model is appropriate for this thesis, a Hausman-Test was performed for every model. “The difference between fixed and random effects is whether the unobserved individual effect embodies elements that are correlated with the regressors in the model, not whether these effects are stochastic or not” (Greene, 2008, p.183). The Hausman-Tests indicated that a random-effects model was appropriate for this thesis

Variable VIF 1/VIF

Full acquisition * ACAP 13.63 0.073352

Absorptive capacity (ACAP) 13.45 0.074373

Cultural distance (CD) 7.08 0.141161

Full acquisition * CD 6.98 0.143340

Full acquisition * ID Institutional distance (ID) Full acquisition GDP per capita Firm age Post-acquisition period 5.96 5.78 4.89 4.31 3.45 2.43 0.167887 0.173114 0.204429 0.231892 0.290228 0.411600 Multiple acquisitions 1.72 0.582538 Services 1.69 0.592647 Firm size 1.07 0.934346 Mean VIF 5.57

Variable VIF 1/VIF

Cultural Distance (CD) 7.01 0.142592

Full acquisition * CD 6.86 0.145874

Full acquisition * ID 5.93 0.168703

Institutional distance (ID) 5.74 0.174223

Full acquisition GDP per capita Firm age Post-acquisition period 3.82 3.73 3.43 2.42 0.261600 0.267927 0.292443 0.413011 Multiple acquisitions 1.72 0.582665 Services 1.67 0.597481

Absorptive capacity (ACAP) 1.33 0.752673

Firm size 1.07 0.935590

regression and an OLS regression, the Breusch-Pagan Lagrange Multiplier (LM) test was performed. The p-value of this test was lower than 0.01, which indicates that there is a panel effect and that the random-effects model should be used. The results of the Hausman and LM tests for each model can be found in the appendix.

The regression analysis was performed hierarchically, meaning that each set of summary statistics is repeated for each stage in the hierarchy (Field, 2013). The first Model only includes the independent dummy variable full acquisition vs. partial acquisition. The second Model also includes absorptive

capacity. The third Model adds the interaction effect between absorptive capacity and full acquisition

on firm performance. The fourth Model adds institutional distance, cultural distance, and the interaction effects of institutional distance and cultural distance on the main effects of a full or partial CBA. In the fifth model, the control variables are added to the regression. Lastly, Model 5 was re-run with the interaction effect of absorptive capacity * full acquisition. The regression results are reported in Table 7.

As shown in Table 7, there was a positive but insignificant impact of full acquisitions on firm performance, compared to partial acquisitions. The estimated coefficient ranges from 0.523 to 2.114 across the six different Models. Although the coefficient is positive in all Models, no significant correlation was found in any of the Models. Without further context, a full acquisition by an EMMNE in a developed market does not lead to significantly better firm performance, on average, compared to a partial acquisition. Therefore, there is no evidence to support hypothesis 1.

additional regression was run without the interaction term of absorptive capacity and full acquisition. Again, the main effect of absorptive capacity on ROA was significant (p < 0.01). An increase in R&D intensity ratio by one results in an increase in the ROA ratio by 0.255. It can be concluded that absorptive capacity has a positive impact on ROA but based on this thesis’ data, it cannot be concluded that there is a moderating effect of absorptive capacity and full acquisitions on firm performance. Therefore, there is no evidence to support hypothesis 2.

Table 7. Random Effects Model: Dependent variable is ROA

Institutional distance as a predictor value has a negative but insignificant effect on firm performance in Model 4, but turns moderately significant in Models 5 and 6, with coefficients of -0.358 and -0.367, respectively. This might indicate that expanding to institutionally distant countries can lower firm performance. The outcome of the interaction, however is more interesting for this thesis. In Model 4 the interaction effect of institutional was insignificant, providing no support for hypothesis 3. However, when control variables were included in Models 5 and 6, the interaction between full acquisitions and institutional distance was positive and significant at 5 percent level. The coefficients were 0.504 and 0.513 respectively which might suggest that although the average effect of full acquisitions on firm performance is zero, a full acquisition in institutionally distant countries may improve firm performance. When institutional distance increases, ROA is reduced by 0.367. When a firm engages in a full acquisition when institutional distance increases, ROA improves by 0.513, which overturns the direct impact of institutional distance. Therefore, ROA improves slightly when an EMMNE engages in a full acquisition when institutional distance is high. However, since there is no significant relationship between full acquisitions and firm performance, the effects are marginal. The coefficient of institutional distance as a moderate did not remain constant across models, but was strongly significant in Model 5 and 6, which included all variables that were deemed relevant. Thus, there is evidence to support hypothesis 3.

relationship between full acquisitions and firm performance, there is a negative marginal effect of full acquisitions in culturally distant countries on firm performance, which provides support for hypothesis 4.

Model 5 also included several control variables that might have had an impact on the EMMNEs’ ROA, regardless of the acquisition choice. Overall, the firms’ performance seemed to decrease in the years after the merger (β = -0.910, p < 0.05). A possible explanation could be that the firms experience extra costs of integrating the foreign target were higher than the benefits derived from the acquisition (Child, 2001). This might also explain why GDP per capita was significant (p < 0.01) but negative (-0.00104). While the performance of the firms in the sample generally decreased, the GDP per capita in the home countries, in general, steadily (Worldbank, 2020). The ROA of firms that operate in the service industry was on average 3.345 higher than for firms that operate in manufacturing industries (p < 0.05). The variables firm size, firm age, and multiple acquisitions all had insignificant coefficients. It must be noted that the explanatory power of the Models is quite limited. The main Models (5 & 6), had the highest R-square values of 0.2127 and 0.2108 respectively, just above the threshold of 0.2, which Hair et al. (2010) deem as an ‘acceptable fit’.

4.4 Robustness

that this method led to an increase in the full acquisition to partial acquisition ratio, which might have impacted the results.

Table 8. Robustness tests: Dependent variable is ROA

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

acquired and institutional distance was positive but insignificant, meaning that an increase of percentage stake by one in a culturally distant target country does not lead to significantly better performance. Furthermore, the interaction effect of cultural distance and acquired stake on firm performance reduced the significance (p < 0.1 in Model 9) compared to the main model (p < 0.05). However, the results here might be impacted because the vast majority of the EMMNEs acquired 100% of the target’s stake. Lower acquisition stakes had only a few individual observations, so the impact of the lower percentages is not as strong as those of 100% acquisitions. The impact of absorptive capacity and the control variables remained unchanged.

Lastly, as can be seen in Table 2, EMMNEs from China and to a lesser extent India, dominate the sample. To determine whether the results would hold without these two countries, the main Model was re-run without China (Model 10) and China and India (Model 11). When China was taken out of the model, the interaction effect between institutional distance and full acquisitions on firm performance became more significant (β = 0.982, p < 0.01). The interaction effect of cultural distance, on the other hand, became insignificant. The absence of Chinese firms in the analysis somewhat altered the results. Absorptive capacity and the interaction between institutional distance and full acquisitions on firm performance remained significant, which also holds for GDP per capita and the impact of service firms. The performance after the acquisition did not significantly worsen for non-Chinese firms.

1.312, p < 0.01). The interaction effect between cultural distance and full acquisitions remained negative and become more significant compared to the main model (β = -2.186, p < 0.01). The results of Model 11 should be considered with caution since taking both China and India out of the equation reduced the number of acquisitions to 50, which reduces the robustness of the model (Riley et al., 2020).

Overall the interaction effects of the direct impact of absorptive capacity on firm performance were relatively robust. The direct impact of full acquisitions on firm performance was influenced by the country of origin and the sample size.

5. DISCUSSION

This thesis aims to extend the knowledge of entry mode strategies and firm performance for EMMNEs. More specifically, under which circumstances does a full- or partial acquisition of a firm in a developed market by an MNE from an emerging market lead to better financial performance. Since EMMNEs seem to follow different internationalization strategies compared to DMMNEs (Luo & Tung, 2018; Ramamurti, 2012), there is a debate within the IB field whether the current internationalization theories can be applied to EMMNEs. On the one hand, some researchers argue that EMMNEs are so different from DMMNEs that traditional internationalization theories are not applicable and new internationalization theories are needed (Luo & Tung, 2007; Mathews, 2006). On the other hand, researchers argue that EMMNEs and DMMNEs are not so different and that traditional internationalization theory is just fine (Narula, 2006; Rugman, 2010). The third stream of researchers argues that traditional theories apply to EMMNEs as well, but these need some refinement and adaptation since EMMNEs have different features (Ramamurti, 2012; Cuervo-Cazurra, 2012; Verbeke & Kano, 2015). The goal of this thesis is to add additional insights to this debate by looking at which entry mode (full or partial acquisitions) results in better financial performance when investing in culturally and institutionally distant countries.

2014, Yang, 2015). Considering the recommended strategies by the literature, a panel data regression is performed to test whether the proposed choice for a full or partial acquisition actually leads to better performance. The outcome of the regression is then compared to the dominant literature on the impact of cultural and institutional distance on DMMNE performance (e.g. Contractor et al., 2014, Brouthers, 2013; Gaur & Lu; 2007; Kogut & Singh, 1988). Additionally, the moderating impact of absorptive capacity on EMMNE performance is discussed, as there is debate about the capabilities of EMMNEs regarding absorptive capacity (Deng, 2009; Awate et al., 2012).

The results of this thesis show that firm performance of EMMNEs generally decreased after the acquisition. This contradicts the findings of Du and Boateng (2012), who argue that most research finds that CBAs undertaken by EMMNEs improves firm performance. These results do lend support to Aybar and Fitici’s (2009) paper, where they find that CBAs by EMMNEs reduce firm performance in more than half of the cases. While this is an interesting result, the main focus of this is to determine whether full acquisitions undertaken by EMMNES leads to better performance compared to partial acquisition. The results indicate that there is no significant unconditional difference between full or partial acquisitions on firm performance by EMMNEs in developed markets. From the data sample, it is clear that most of the acquisitions by EMMNEs in the 2012-2018 period were full acquisitions. This confirms the arguments of Deng (2009) and Luo and Tung (2007) that aggressive high stake acquisitions are the preferred entry mode in developed economies. However, full acquisitions do not lead to performance per se. The acquisition strategy depends on conditional factors that determine the entry stake.

absorptive capacity and its predictor variable. The thesis did find that firms that invested in R&D and thus improved their absorptive capacity, attained a higher ROA. Deng (2009) found similar results when comparing the foreign acquisitions of Chinese firms. The firms that had a higher level of absorptive capacity were better at integrating the externally acquired knowledge and resources and combining those with their own resources, leading to better financial performance (Deng, 2009).

Some researchers argue that the absorptive capacity of EMMNEs is limited due to knowledge deficiencies (Madhok & Keyhani, 2012). Furthermore, Cuervo-Cazzura and Rui (2007) found that EMMNEs face extra barriers to attain absorptive capacity due to characteristics of EMMNEs and home country institutional factors. The results of this thesis confirm that EMMNEs that invested in their absorptive capacity performed better, but it did not confirm that these effects are stronger for full acquisitions than for partial acquisitions.

Furthermore, this thesis tested the interaction effect between full acquisitions and institutional distance on firm performance. While there was no significant direct impact between full acquisitions and firm performance, it did find that those EMMNEs that engaged in full acquisitions performed marginally better when institutional distance increases. This is a particularly interesting finding since it deviates from what traditional IB theories find (Brouthers & Hennart, 2007; Xu et al., 2004). Traditionally, low-equity strategies are recommended when firms enter institutionally distant countries, due to high transaction costs, uncertainty, and a lack of legitimacy in the host country (Kostova et al., 1997). The main argument for this distinction between EMMNEs and DMMNEs is that EMMNEs invest in developed markets to improve their resources and legitimacy, while DMMNEs engage in partial acquisitions to reduce the risk of dealing with lower-quality institutions and poorer regulatory systems (Contractor et al., 2014; Lahiri et al., 2014).

Contractor et al. (2014) and Lahiri et al. (2014). This result is also supported by Yang (2015), who also found that EMMNEs that engaged in full acquisitions in institutionally distant countries performed better than those that engaged in a partial acquisition. Since there seems to be a difference between EMMNEs and DMMNEs regarding their views towards the institutional distance, there is some support for the view that EMMNEs need at least an adjustment of internationalization theory. More empirical research will be necessary to strengthen that view.

Lastly, this thesis finds a moderate impact of the interaction between full acquisitions and cultural distance on firm performance. While full acquisitions did not result in significantly better results, those EMMNEs that engaged in full acquisitions in culturally distant host countries seemed to perform marginally worse than those that engage in partial acquisitions. The results of this thesis are in line with those from traditional IB views, in which unfamiliarity with the host country complicates the process of extracting resources from the foreign target firm (Johanson & Vahlne, 1977; Quer et al., 2007).

Contractor et al. (2014) proposed that EMMNEs tend to prefer partial acquisitions in culturally distant countries, while Yang (2015) found that these indeed perform better than those that engaged in full acquisitions. Based on these studies and the results of this thesis one could argue that EMMNEs and DMMNEs do not differ significantly when it comes to acquisitions in culturally distant countries. On the other hand, other researchers find a positive result when firms engage in CBAs in culturally distant countries (Morosini et al., 1998; Chakrabarti et al., 2009). Aybar and Ficici (2009) found that acquisitions by EMMNEs in developed markets destroy value altogether. These studies did not make a distinction between full and partial acquisitions.

6. CONCLUSIONS

6.1 Implications for theory

The main goal of this thesis is to empirically test whether full or partial acquisitions undertaken by EMMNEs in developed markets had different impacts on firm performance. Most literature on EMMNEs focussed on the motives (Luo & Tung, 2007), the characteristics of EMMNEs (Ramamurti, 2012), and whether cultural and institutional distance leads to higher or lower ownership participation (Contractor et al., 2014; Lahiri et al., 2014). Compared to entry mode research on DMMNEs, little is known whether these different entry mode strategies result in better performance (Yang, 2015).

Furthermore, the findings of this thesis extend the debate about whether different theories are needed to explain EMMNEs internationalization. The results show that there are differences and similarities between EMMNEs and DMMNEs internationalization performance. Building on the recommendations by Contractor et al. (2014) and Lahiri et al. (2014), this thesis lends support to the notion that EMMNEs may perform better when a full acquisition is undertaken in an institutionally distant country, compared to a partial acquisition. This contradicts the traditional IB view that firms should incrementally increase their commitment and investment in less distant countries and then expand to more distant countries (Johanson & Vahlne, 1977).

On the other hand, when EMMNEs venture to culturally distant countries, they may perform better when entering through a partial acquisition. Traditional IB literature also recommends lower-risk entry modes to overcome the liability of foreignness (Zaheer, 1995). Considering these outcomes, a completely new theory to explain EMMNE expansion behavior seems unnecessary. However, there do appear to be differences between EMMNEs and DMMNEs. Therefore, the findings of this thesis seem to support the earlier views that some adjustment of traditional IB theory to the features of EMMNEs might help to gain a further understanding of the topic of EMMNE internationalization (Verbeke & Kano, 2015; Ramamurti, 2012).

could not be concluded based on the data. This adds to the literature that even though EMMNEs suffer from knowledge and resources deficiencies (Madhok & Keyhani, 2012) and face extra barriers to attain absorptive capacity (Cuervo-Cazurra & Rui, 2007), they can still benefit from investing in absorptive capacity, confirming Deng’s (2009) findings.

6.2 Implications for practice

This thesis also has practical implications. First, it shows that full acquisitions do not lead to more positive spill-overs than partial acquisitions per se, but that the decision to acquire a foreign firm is a conditional strategic decision. Furthermore, the results of this thesis imply that investing in absorptive capacity might lead to better financial performance, but do not indicate that these effects are stronger in full acquisitions. EMMNE managers who want to acquire firms in developed markets to attain strategic assets, integrate them into the focal firm, and combine them with their own resources and knowledge might consider investing in R&D to develop the capabilities that improve this process.

Moreover, the results of this thesis indicate that there are both differences and similarities between the acquisition strategies of DMMNEs and EMMNEs. Managers of EMMNEs might consider full acquisitions in institutionally distant countries when the goal is to improve legitimacy, accounting standards, and improve the protection of property rights. In other words, the positive spill-overs from institutionally developed countries can contribute to the EMMNE’s performance. This strategy differs from the recommended entry modes for DMMNEs in institutionally distant markets.

6.3 Limitations and future research

Like most empirical research, this thesis has some limitations. First, this thesis only considers cross-border acquisitions, while EMMNEs also enter developed markets through joint-ventures, greenfield acquisitions, or strategic alliances. The results of this thesis might not apply to those entry mode strategies. Furthermore, there is a domination of full acquisition in this thesis. While it is explained by the literature that full acquisitions are the preferred entry mode strategy (Luo & Tung, 2007), a sample that contains more partial acquisitions might improve the reliability and generalizability of the results. Moreover, using R&D intensity might not capture the firm’s absorptive capacity in detail (Vega-Jurado et al., 2008). Although the metric is grounded in the literature (Veugelers, 1997; Stock et al., 2001; Tsai, 2001), a more combination of multiple metrics might provide greater insight into the target firm’s ability to extract and recombine the resources of the acquired target (Vega-Jurado, 2008). Other determinants of potential and realized absorptive capacity are employee’s level of education, formalization of processes, and the presence of social integration mechanisms (Vega-Jurado et al., 2008; Lenox & King, 2004; Zahra & George, 2002). Based on the data in this thesis, there was multicollinearity between the interaction term of absorptive capacity and its predictor variable. This is mainly attributed to the relatively small portion of partial acquisitions in the sample and a larger dataset with more partial acquisitions might provide a more reliable outcome.

abnormal returns as a result of the acquisition (Papadakis & Thanos, 2010). This method is based on the principle that the primary goal of the firm is to generate value for the shareholder (McGee et al., 2005), which is better captured using market-based performance measures. However, the short term impact of an acquisition on stock value does not capture the firm’s ability to extract value from the acquisition and generate synergistic gains, as this is a medium- to long term process. Therefore, the use of both long- and short term event windows are recommended (Du & Boateng, 2012). Considering the complexity of acquisitions and their impact on performance, it would be even better to combine market-based measures with accounting measures and organization metrics (Papadakis & Thanos, 2010; Meglio & Risberg, 2011).

Furthermore, despite being heavily used in the IB literature, the concept of cultural distance is subject to criticism (Beugelsdijk et al., 2018). Firstly, McSweeney (2002) argues that clear national cultures might not exist, but it consists of multiple subcultures that are not limited to national country borders. Another criticism of Hofstede’s Dimension is that the questionnaire was designed for IBM employees across the world, which might be more representative of corporate rather than national culture (McSweeney, 2002). The use of the Kogut and Singh Index also has its limitations. The Kogut and Singh Index has been accused of assuming that the distances between countries are symmetrical (Shenkar, 2001). When applied to M&As, some of the cultural dimensions might be disruptive than others (Contractor et al., 2014). The Kogut and Singh Index does not make a distinction between dimensions, as each dimension has the same weight. Whether the moderating impact of cultural distance holds, still needs more research.

correlation between individual dimensions (Beugelsdijk et al., 2018). This method can also be used for the concept of institutional distance.

The classification of emerging markets is rather broad and the results of this thesis might not apply to all emerging countries. Hoskisson et al. (2013) divide 60 emerging countries into (1) traditional emerging economies, (2) newly developed economies, and (3) mid-range emerging economies. The BRIC and VISTA countries all belong to the mid-range emerging economies that have already experienced some degree of economic and institutional growth (Biligili et al., 2016; Hoskisson et al., 2013). The results of this thesis might not apply to traditional emerging economies and newly developed counties, due to significant differences between these groups of emerging markets. Even within the mid-range emerging economies, there is some heterogeneity (Cuervo-Cazurra & Rui, 2007). Moreover, this thesis is based on the assumption that EMMNEs enter developed markets to acquire assets and resources that are not available at the home market and recombine those with firm assets to generate synergetic gains (Luo & Tung, 2007). More in-depth research is needed for EMMNEs that have different motives to venture abroad, as these also require different metrics for performance measurement.