What Community Really Means
Different Understandings of Community and Post-Merger and Acquisition
Performance
Master Thesis
M.Sc. International Business & Management (Groningen) M.Sc. Managing Multinationals (Uppsala)
Jord Kalle Geerts
j.k.geerts@student.rug.nl Student Number: 2968142
University of Groningen, The Netherlands Uppsala University, Sweden Faculties of Economics and Business
Duisenberg Building 9700 AB Groningen
Supervisor: Dr. B. J. W. Pennink Co-Assessor: Dr. O. Lindahl
Date of Submission: 20th of January 2020
2
Abstract
Cross-country comparisons have been at the heart of the field of international business, both in a theoretical and practical sense. Nevertheless, these kind of comparisons are ever difficult because the interpretation and understanding of concepts and dimensions used for comparison can be different across different countries. In order to prevent comparing the figurative apples with pears, the understanding of some concepts has to be examined in more detail. The first part of this paper addresses this issue by outlining the understanding of community across different societies. Four countries are evaluated, China and South Korea are found to have a similar in- versus out-group understanding of community whereas Sweden and South Africa have a more open idea of community. The second part focuses on the more practical aspect of the field of international business. The development of a new community, through social integration processes, after a merger or acquisition is important for the performance of the said merger or acquisition. Therefore, it is examined how the different understanding of community could have an impact on the performance of mergers or acquisitions. However, no statistical evidence is found for the expectation that performance differs across the acquirers from different nations. Thus, the effect of having a different understanding on the performance of mergers and acquisitions does not become clear in this paper.
Keywords: Community, Post-merger Performance, Sweden, China, South Korea, South
3
Table of Contents
1.
Introduction ... 5
2.
Literature review ... 7
2.1 Community ... 7
2.2 Hofstede and collectivism ... 9
2.3 Beyond Hofstede ... 10
2.4 Community and M&A’s ... 11
3.
The understanding of community in different nations ... 14
3.1 The case of China and South Korea ... 14
3.3 The case of South Africa and Sweden ... 15
4.
Comparing the understandings of community ... 18
4.1 How the understanding differs across nations ... 18
5.
Community and performance ... 21
5.1 Social integration in M&As ... 21
5.1.1 Chinese and South Korean perspective on mergers ... 22
5.1.2 Swedish and South African perspective on mergers ... 23
6.
Hypotheses ... 24
7.
Research method ... 25
7.1 Data sources and sample selection ... 25
7.2 Data collection and analysis ... 26
7.2.1 Dependent variable ... 26
7.2.2 Independent variable ... 28
7.2.3 Control variables ... 29
7.3 Conceptual Model ... 30
8.
Results ... 32
8.1 Results for cumulative abnormal returns ... 32
8.1.1 Empirical assumptions ... 32
4
8.1.3 Results For Positive And Negative Cumulative Abnormal Returns ... 36
8.2 Results for abnormal returns in the 12th month ... 39
8.2.1 Results for positive and negative abnormal returns in the 12th month ... 42
9.
Conclusion ... 44
9.1 Discussion ... 44
9.2 Limitations ... 45
9.3 Recommendations for future research ... 46
9.4 Conclusion ... 47
10.
References... 48
11.
Appendix... 60
11.1 Appendix A: Results related to all the cumulative abnormal returns (CAR B) ... 60
5
1. Introduction
“[….], if the cultures vary considerably in the norms associated with a particular dimension, cross-cultural comparisons will be confounded (Heine, Lehman, Peng, K. & Greenholtz, 2002: 905). This statement accurately reflects the current issues related to cross-cultural research: How to compare, for instance, the attractiveness of people from country A with those of country B, when the norm and understanding of “attractive” is different in both countries? If blonde hair is an attractive trait in country A but country B considers brown hair to be attractive, the common understanding of attractiveness is lacking. In other words: when the understanding of a certain dimension is different for two countries, comparing country specific scores on that dimension would be like comparing the figurative apples with pears.
This article contributes to the theoretical field by arguing that the proposed dimensions as they exist today (e.g. Hofstede, 1980, Schwartz, & Boehnke, 2004 and Trompenaars, 1994) are not all encompassing. Since the understanding of the concepts are not always similar, cross-country comparisons are sometimes based on false premises. Consequently, the results of comparison can be wrongly interpreted and even become irrelevant or useless. Thus, this article fits in with authors that have criticized the so widely accepted measurements (e.g. Heine, Lehman, Peng, K. & Greenholtz, 2002, Shenkar, 2001 and Schneider, 1988), while simultaneously offering a deeper understanding on how one concept can be understood differently in different countries. Community was chosen to be the concept under investigation since the a seemingly clear definition of McMillan and Chavis (1986) actually leaves room for a certain degree of variation in the interpretation and understanding across different countries. In addition, previous literature indicates that a feeling of community is likely to develop in collectivistic societies, and not in those that are individualistic (Hornik & Tupchiv, 2006), a statement that can be challenged when the understanding of community is found to differ between individualistic and collectivistic countries. The first research question, relevant for the theoretical contribution of this paper, is stated as follows:
1. What is the understanding of community in different nations?
6 of a merger or acquisition has been neglected, but could be of great importance. Therefore, it was decided to investigate the relationship between the understanding of community and the performance of cross-border M&As. Several cross-border M&As are evaluated, with South African companies representing the targets and Swedish, Chinese and South Korean companies serving as acquirers. The research question associated with the practical relevance of this paper is given below.
2. Does the understanding of community have an influence on the integration process and the subsequent performance of cross-border M&As.
Sweden and South Africa are chosen because of their apparent individualistic character (Hofstede, 1980) and their intensive business relationship as shown by the high level of Foreign Direct Investment (FDI) from Sweden to South Africa (UN Comtrade, 2018). In a practical sense, both countries could benefit from gaining a deeper understanding of how the society in the other country is organized. After all, both already committed a significant amount of resources into this relationship and it is important to keep on developing the relationships and reap the benefits from trade and investment. China and South Korea, on the other hand, are used as comparison countries because of the apparent collectivistic ideas that are present within the society (Hofstede, 1980).
7
2. Literature review
In this section, the definition of community will be given, and the relationship with the individualism-collectivism dimension of Hofstede (1980) will be explained. On first sight, the concepts of collectivism and community overlap to a great extent. However, in the following section the differences are outlined with regards to the interpretation and the understanding of community as it exists within different societies. Later in this section, the connection will be drawn to the business environment and the effect of culture on mergers and acquisitions will be taken into consideration. After all, in order for learning to take place and the M&A to be a success, all members should share a feeling of belonging to the newly developed community (Lave & Wenger, 1991 and Gates & Very, 2003).
2.1 Community
One could state that the first research question, concerning the understanding of community, is relatively easy to answer and there is no need for a comprehensive literature review on the topic. After all, McMillan and Chavis (1986: pp. 9) provide a crystal clear definition of the concept: [community is] a feeling that members have of belonging, a feeling that members matter to one another and to the group, and a shared faith that members’ needs will be met through their commitment to be together”. However, the interpretation of this definition could lead to several questions, such as: How open is the community to new members? Who are the members, do they come from the same social background or are high and low status individuals mixed together? The first part of this paper sheds light on the interpretation and understanding of community for Sweden, South Africa, China and South Korea.
8 developing a sense of community as it develops connectedness to the fellow group, or soon to be community, members (Gunawardena, 1995).
Nevertheless, the fact that there is interaction among individuals does not necessarily indicate that we can speak of a community. Already in 1982, before McMillan and Chavis formed their definition of community, McAdam stated that being part of a community is much more than just being connected to each other or being recognized by the group members. The feeling of belonging to a community, which is incorporated in the element of membership, is even more important, along with the central thought that one’s own objectives and goals can be achieved by actively engaging in the community (McAdam, 1982). In his later work, McMillan (1996) renames the four elements as; spirit, trust, trade and art respectively. In the words of McMillan (1996: 315): I view Sense of Community as a spirit of belonging together, a feeling that there is an authority structure that can be trusted, and awareness that trade, and mutual benefit come from being together and a spirit that comes from shared experiences that are preserved as art.
9
2.2 Hofstede and collectivism
Within existing literature in the field of international business, academic researchers often refer to the work of Hofstede (1980). In fact, his research is among the most cited and respected papers in the field (Bond, 2002). Two of the six dimensions are of specific importance for this paper; individualism versus collectivism and long- versus short-term orientation. The latter was not included in Hofstede’s original research, but was added after it became clear that there was a Western oriented reflection of cultural dimensions (Hofstede & Bond, 1988).
With individualism on the one side, and collectivism on the other, this dimension indicates the extent to which people are integrated in groups, or communities. In a highly individualistic society people are “I”-conscious, which indicates that every individual only takes care of him or herself, and their immediate family members. With this explanation in mind, one would not expect a high regard for community, in any sense of the word, in an individualistic society. The person is seen as an individual, as opposed to being group-dependent (Rudvin, 2007) which is the case in collectivistic societies. When a country scores relatively low on the individualism dimension, it is considered to be collectivistic, people take care of their in-group and harmony among group members is crucial. Loyalty is often a cornerstone of society, and the group depends on its members to protect each other and the group as a whole (Hofstede, 2001). This resembles the elements of spirit, trust and even trade, as coined by McMillan (1996). Thus, the concepts of collectivism and community show a high degree of similarity and, in some instances, the words for community and collectivism have become intertwined (e.g. Mamontov, Kozhevnikova & Radyukova, 2014) and are used to indicate one and the same thing. However, in the section “Beyond Hofstede” later in this paper, it is argued that collectivism and community are in fact different constructs and that a feeling of community can exist in a society that scores low on collectivism.
long-10 term oriented society (Hofstede & Minkov, 2010), which could mean that there is better cooperation (Hilbig & Zettler, 2009). Ultimately, displaying honesty also contributes towards building a, stronger, community (Strong, Ringer & Taylor, 2001).
2.3 Beyond Hofstede
Even though collectivism and community are sometimes used as synonyms, and the need for belonging to a community is highly present in a collectivistic society (Ahuvia, 2002), the two concepts should not be confused with each other. Feelings of community can also exists outside of collectivistic societies (Tame, 1998). At this point, this paper forms an addition to the research on cross-country comparisons by demonstrating in more detail that one concept might have different interpretations in different societies. From a theoretical perspective, this is an important realization. Scores on dimensions, such as those of Hofstede (1980) are used by academics to indicate the degree of difference between one country or society and another. However, when the interpretation of the dimension differs between societies, one would be comparing the figurative apples with pears.
Hofstede (1980) describes both South Africa as well as Sweden as relatively individualistic and short-term oriented, especially once they are compared to the scores of China and South Korea on the same dimensions (see Graph 1).
Graph 1. Scores for the individualism and long-term orientation dimensions: South Africa,
Sweden, China and South Korea.
Source: Scores on the dimensions are derived from: Hofstede Insights.
As can be observed, the scores for individualism are noticeably higher in Sweden and South Africa as opposed to China and South Korea. Nevertheless, the community does in fact represent a focal point in Scandinavia and the Swedish companies are often trying to achieve
South Africa Sweden
China South Korea
11 the goals of society as a whole (Grennes, 2013). Moreover, Scandinavian managers often realize they have a responsibility towards the organization and therefore pursue the development of the organization as a whole, rather than pursuing their own individual goals such as their promotion (Head, Larsen, Nielsen & Sorensen Jr, 1993). Lastly, avoiding confrontation and maintaining harmony is often associated with a collectivistic society (Hofstede, 2010) and not with a society that scores high on individualism. Nevertheless, Swedish management is characterized by compromising and consensus (Brewster, Lundmark & Holden, 1993), which are indicators that the objective of harmony should be pursued.
With regards to Ubuntu and South Africa, described as an individualistic society, the task is supposedly more important than the relationship. However, as emphasized by Turaki (2006), a crucial aspect of African cultures and Ubuntu are the community, and the subsequent relationships as they evolved within the community. Ubuntu describes that one should care for the community (Mangaliso, 2001), and an individual can only achieve his own objectives by doing good for the broader community. In other words, the dimension by Hofstede (1980), and the value awarded to it, seems to give a too simplistic view of the real-life situation.
Furthermore, as we further investigate the values that Hofstede (1980) has ascribed to these countries, another value can be highlighted. Represented with a value of 34, the long-term orientation of South Africans seems low, especially relative to the countries that are offered as comparison. This would indicate that they are focused on the short term, whereas Ncube (2010: 78) clearly states that: “the hallmarks of Ubuntu are harmony and continuity”.
On a concluding note, it can be stated that there seem to be different understandings of community across different cultures and societies. Even when, according to Hofstede’s model, there does not seem to be a focus on the collective as the society is given an individualistic rating, there can still be a sense of belonging, which is a key aspect of the concept of community (McMillan & Chavis, 1986), that exists in the society.
2.4 Community and M&A’s
Up until this point, community has been the core concept under inspection. However, the connection to the (international) business environment should not be neglected, which is why this section will form as a reminder why the understanding of community should be further investigated.
12 resembles the unity of the new organization. Failure to develop a sense of unity can result in failure of the M&A, as was the case for Amazon and Whole Foods (Gelfand, Gordon, Li, Choi & Prokopowicz, 2018) as well as for Daimler-Chrysler (Dealbook, 2010). In order to make the merger or acquisition a success, the employees should share a sense of belonging and the feeling of “us versus them” should be avoided (Vaara, Sarala, Stahl & Björkman, 2012). However, this feeling of us versus them is often exactly what is present in societies that are described as collectivistic (Hofstede, 1980). In order to avoid such a stand-off, a social integration process needs to take place that can bring the two communities together (Bresman, Birkinshaw & Nobel, 1999), with the goal to develop a new community that all employees can identify themselves with (Empson, 2004). When there is a feeling of belonging to the new community, learning can take place (Lave & Wenger, 1991), which is beneficial for the knowledge sharing process and can contribute to the success of the merger (Gates & Very, 2003). However, when neither of the two communities is open to the other community, it is unlikely that a shared feeling of belonging will develop. Instead, a feeling of us versus them is present and the merger will most likely fail (Vaara, Sarala, Stahl & Björkman, 2012). This is often the case when there are clear status differences between the groups as groups that are perceived to have a low status often feel vulnerable in case of a merger, whereas the higher status groups have a more positive reaction (Fischer, Greitemeyer, Omay & Frey, 2007). Developing the feeling of belonging to the new community as well as the sense that the employees need each other to grow and develop the organization are, thus, key aspects of the social integration process.
Previous research has since long indicated the necessity of developing a common identity after an M&A (Olie, 1990), and several best practices have been developed in order to improve the social integration that is necessary to develop a common identity (Wolf, 2003). Furthermore, research has even indicated the influence of national culture on the integration process (Vaara, Sarala, Stahl & Björkman, 2012). Below, in Figure 1, the process towards a shared identity is visualized.
Figure 1. Creating a shared identity through social integration.
Community A Community B Social integration after M&A Shared identity in Community C
A variety of factors influencing the social integration (outside
14
3. The understanding of community in different nations
3.1 The case of China and South Korea
The cultures of China and South Korea score relatively low on the individualism scale and are therefore seen as relatively collectivistic societies according to Hofstede (1980). Therefore, in theory, these cultures should reflect a high regard for the group and the community.
15 With regards to the South Korean situation, on the basis of Hofstede’s comparison (see Graph 1), it could be expected that the society as a whole is more collectivistic and long-term oriented as Sweden, South Africa and even China. The existence of the Chaebol in South Korea is a vivid example of the long-term orientation and collectivistic outlook on life. Chaebol firms are large business groups, or conglomerates, that have a dominant presence in the Korean business world (Chang, 1988). Often owned and controlled by a small group of shareholders, in some cases family members, the group members depend on and benefit from each other (Choi & Cowing, 1999). It resembles a structure where one company is favored over the other, the members of the group share resources and internal buying and selling is common (Chang & Hong, 2000). Thus, the existence of a community can be expressed on a business level, where several firms cluster together and represent themselves as one. Rivalry exists between the different Chaebol, and community members try to positively enhance the performance of in-group members whereas they negatively influence the performance of other in-groups. An example is given by Song, Mantecon and Altintig (2012), who discovered that analysts who are incorporated in one Chaebol would positively rate the members of this Chaebol whereas they would give negative recommendations about firms from other Chaebol. The sense of belonging to one community also shows the importance of relationships, belonging to a strong Chaebol would mean that the firm could survive even when the firm itself is unprofitable. Nevertheless, it is not only in-group relationships that are of importance, as Chaebol are involved in unrelated and sometimes unprofitable markets, these Chaebol take on significant amounts of debt from banks (Campbell & Keys, 2002). Consequently, it is crucial for managers and executives to develop relationships with high status bank officials to finance their activities, and these officials often neglect to monitor the Chaebol (Campbell & Keys, 2002). In addition to that, developing ties with government officials, whom are also perceived to have high status, can prove to be of importance as well. Directors who have political connections can cause an increase in performance while simultaneously reducing the risks for the Chaebol overall (Shin, Hyun, Oh & Yang, 2017).
3.3 The case of South Africa and Sweden
16 Ubuntu stems from the African traditional cultures (Lutz, 2009) and is focused on the “we”, rather than increasing its own profits the firm should be focused on striving to achieve an increase in well-being for all the parties that are connected to the firm. Not only in South Africa, but actually in Africa as a whole this focus on the community can be found (Mbigi, 2005). To make it easier to grasp for the Western reader, Ubuntu could be linked to the stakeholder theory as proposed by Freeman in 1984, where the focus extended from shareholders and their profits to the broader network of the organization and the common good. In Ubuntu, the common good is inherently linked to the individual’s goals as well, he or she can attain its own goals through pursuing the common good (Lutz, 2009). Relationships play a vital role (Shutte, 2001), since there would be no common good if there were no relationships to begin with. It is remarkable that it is not until the last couple of years that Western academics have begun to describe the importance of relationships whereas this idea already existed in South Africa. In the article of Fosfuri, Giarratana and Roca (2011, pp: 222), the authors state that their article “introduces the notion of community-focused strategies”, while this idea has existed for many years already in the African communities. To be even more explicit, the responsibility for the community in Ubuntu goes far beyond individual or firm boundaries, rather it is directed at every community that the person or institution is involved in, from village to nation and even to the global community (Nussbaum, 2003). Furthermore, and connected to the importance of relationships, Ubuntu places an emphasis on a non-materialistic life-style. Pursuing material wealth is often subordinate to the goals of developing meaningful relationships with others (Shonhiwa, 2006). Thus, summarizing this philosophy in one sentence, one who lives by the values of Ubuntu would develop relationships with others and simultaneously enhances his own well-being and the well-being of the community by pursuing the (non-materialistic) common good.
17 between different actors that the overall well-being of the greater community is pursued and achieved. Harmony is not only important within the community, but should also be maintained between the different communities, which is why democratic processes and cooperation are inherited in the Scandinavian management style (Lindkvist, 1988). Additionally, maintaining harmony means that conflicts are avoided (Daun, 1986), and reaching a consensus is more important than persuading others (Brewster, Lundmark & Holden, 1993). In practice this means that an authority structure is largely absent and status differences not amplified (Berg, 1986), managing is done through inclusion and consulting rather than directing and ordering. Moreover, it is showcased among Swedish companies that monetary profits are no longer the one and only goal of the organizations. Sweden is a role-model for advocating corporate social responsibility (CSR) (KPMG 2010), again accomplished due to extensive cooperation between several parties of different statuses, such as businesses, labor unions, the government but also non-governmental organizations. Additionally, Swedish companies are engaging in CSR voluntarily, because the managers feel as if they have a responsibility towards the community (Hąbek & Wolniak, 2013). In fact, the relatively new term of ‘stakeholder capitalism’ can be used to describe Scandinavian management, in the words of Tor Grennes (2013: 13): […] the Scandinavian model promotes long-term ties between owners, managers, workers, and society, where the role of the company includes promotion of goals of society at large. In some cases, pursuing non-monetary CSR goals outside of the firm’s boundaries can lead to an increase in profits for the firm as well (Wang, 2014). By striving for the community and societal desirable goals, Swedish firms secure their survival, they achieve individual goals by pursuing community goals.
As indicated above, there are in fact certain areas or values which overlap between Ubuntu philosophy and the Scandinavian society. Both are focused on the community as a whole, and there is a clear realization that individual objectives can be achieved through the pursuance of community objectives. Strong members of society sometimes take care of the weak and, although manifested in slightly different ways, relationships are a key aspect and they should be maintained over time. Lastly, material wealth is not the main achievement, it is more important to keep the community happy and to be a good, corporate, citizen.
18
4. Comparing the understandings of community
4.1 How the understanding differs across nations
With the analysis of the four countries and their respective ideas about the society, groups and community we can now provide a clear overview of the differences that exist in understanding what community means in different societies.
In previous literature, the concept of community seems to correspond with the dimension of collectivism as introduced by Hofstede (1980), and there are cases where the words for community and collectivism have become intertwined and are used to describe the same concept (Mamontov, Kozhevnikova & Radyukova, 2014). When we consider China and South Korea, we can indeed notice that the collectivistic score as given by Hofstede (1980) corresponds with the understanding of community that exists within these societies. In South Korea, the Chaebol form an example where a community is formed that tries to outperform the other communities, and an us versus them feeling is present. In China, the harmony that should exist between members of the same in-group should be maintained, and relationships are a crucial aspect of society. Status plays an important role in China, and it is sometimes difficult to develop a relationship with higher-status individuals, or to be adopted in a community of high-status individuals, which also coincides with the ideas of collectivism that there is an us versus them outlook on life.
19 Notable is the difference between the understanding of community in Sweden and South Africa relative to China and South Korea. In the latter two countries, the feeling of community does indeed correlate with the definition of collectivism, and there are many in- and out-groups to be identified. The high- versus low-status groups in China and the Chaebol structures in South Korea indicate that there is rivalry between different communities, and the primary concern of every member is their own community. In Sweden and South Africa, on the other hand, the understanding of community relates to the broader society as a whole. In Sweden, the employees, employers and the government are described to work together often. In South Africa, the regard for the greater community and society as a whole is best described by the examples of Nussbaum (2003) who describes the understanding of community that is not limited to the workplace, the village or even the nation but could concern the world as a whole. In another paper, Nussbaum (2003: 23) states that: following a war between two tribes in South Africa, war healers from each side would together arrange for a cleansing ceremony involving those who fought on both sides. This is a vivid example that, even though competition is present, the rival party should be treated with respect. The feeling of us versus them seems to play less of an important role.
After the four countries have been analyzed and we have observed a different understanding of community across the societies, an answer to the first research question can be given. In order to recap, the question is given: What is the understanding of community in
different nations?
20 include reaching consensus and maintain harmony between groups. The communities are more open and connections with individuals of various levels of status (i.e. employees, employers and government officials) are based on mutual benefits. Thus, in Sweden and South Africa the relative absence of collectivistic ideas, as described by Hofstede (1980), does not indicate the absence of a feeling of community. Rather, the understanding of community is different. Consequently, both academics and managers can not state that there is no regard for the community because Hofstede (1980) states that Sweden and South Africa are relatively individualistic. Additionally, in previous research the concepts of collectivism and community are considered to have a degree of overlap (Hornik, 2006) and are sometimes used as synonyms (Mamontov, Kozhevnikova & Radyukova, 2014). However, the distinction between a broad and narrow understanding as presented in this piece can not be ignored. Country comparisons would form a one-sided and incomplete representation of the reality because apples would be compared with pears when researchers would continue to use the terms community and collectivism interchangeably.
21
5. Community and performance
5.1 Social integration in M&As
Now that differences in the understanding of community have been provided, the attention will shift to the second research question related to the consequences of different understandings of community on the performance of M&As. However, before an answer can be provided, a bridge has to be built between the fields of sociology (understanding of community) and international business (cross-border M&As). Before the countries are discussed individually, the importance of social integration after a merger or acquisition is addressed. For the country specific perspective on M&As we build on the claim that the Swedish and South African culture have a higher regard for the broader notion of community whereas the understanding of community in China and South Korea is focused on dividing the society into in- and out-groups. Sweden and South Africa are deemed to have a similar understanding, as are China and South Korea.
22 of status, is crucial in this process, since the managers and employees of the acquired firm often feel inferior or unappreciated, which hinders the interaction process (Mirvis & Marks, 1992).
5.1.1 Chinese and South Korean perspective on mergers
Chinese corporations acquiring and merging with other firms might also showcase a high regard for the required socialization efforts that follows such an event. In Guanxi, and Confucianism, there is considerable consideration for acts of Ren (benevolence), which indicates that managers display kindness and try to include everyone in the social network that exists within the corporation (Ling, Chia & Fang, 2000). Nevertheless, acts that are a result of Guanxi are often focused on the individual, rather than the group as a whole (Bedford & Hwang, 2013). Moreover, it has been indicated that, in the case of China, significant differences exist between state-owned enterprises (SOEs) and companies that are privately owned. Managers in the latter are often forgetting about Guanxi, as they prefer to focus on profits and actions that are focused on increasing the efficiency and effectiveness of the workforce (Xang & Liu, 2016). Additionally, in the case of state-owned enterprises, difficulties with respect to post-merger integration can arise as well. SOEs often have evolved into complex structures, with unclear objectives and low transparency, which makes it difficult for the managers and employees of the acquired firm since they do not know what is expected of them (Liu & Woywode, 2013).
23 employees, which is often the case when a Korean firm expands abroad (Yang, 2014). In the absence of informal ties the local employees are undervalued and there is no empowerment (Fodchuk & Sherman, 2008) and trust issues exist between managers and employees (Teng & Newell, 2004).
5.1.2 Swedish and South African perspective on mergers
It is grounded in the literature that Swedish companies, in the case that they are the acquiring firms, will be more successful in bringing the merging process to a successful end. After all, cooperation is highly valued in the Swedish society (Tornblom, Jonssons & Foa, 1985) and reaching a consensus is often a crucial objective when negotiating (Brewster, Lundmark & Holden, 1993). Additionally, the Swedes tend to have a feeling of collective responsibility (Czarniawska-Joerges, 1993), which means that all the parties involved are responsible for achieving the objectives that they collectively decided upon. Providing the employees of the acquired company with equal purpose in the organization, and not stripping away their autonomy will lead to a more successful integration process and, hence, a more successful merger (Weber, Shenkar & Raveh, 1996).
24
6. Hypotheses
Considering that the understanding of community has been outlined for Sweden, South Africa, China and South Korea, and that the perspective on M&As has been outlined, the focus shifts to the second research question: Does the understanding of community have an influence
on the integration process and the subsequent performance of cross-border M&As?
In order to provide a clear structure, the following hypothesis is proposed:
Hypothesis 1a: Swedish acquirers will be more successful in an M&A with a South African target, compared to Chinese and South Korean acquirers, because of their “broad” understanding of community.
Hypothesis 1b: Swedish acquirers will be more successful in an M&A with a South African target, compared to Chinese and South Korean acquirers, because of the similarity in understanding of community between Sweden and South Africa.
With the understanding of community we refer to the narrow or broad understanding of the societies incorporated in this research. Unfortunately, at this moment, the level can only be displayed with its two extremes and will be represented by a dummy variable in the analysis. Since there are only a few countries incorporated in this research, it would be unfair to assign values and to display the dimension as a scale. Additionally, it would be necessary to develop surveys or use another measurement in order to come up with a logical scale on which societies could be ranked before a logical value could be given to any society. This could not be executed because of the time and budget constraints connected to this research.
25
7. Research method
In order to find an answer to the second research question, a quantitative method will be used. This section provides information on the data and sample selection, the description of the variables incorporated and the method of analysis that is used to compare the performance of different M&As.
7.1 Data sources and sample selection
The required information on M&As can be found in Zephyr and Thomas Reuters Eikon, where the acquiring country of origin as well as the location of the target company can be specified. The following criteria are used to define the data set:
1. The target company has to be listed in South Africa
2. The acquiring companies must stem from either Sweden, China or South Korea
3. The event of a merger has to be both announced and completed in the past, but later than the year 2000
4. Financial information of all the acquiring company, with respect to changes in stock prices, must be available in the Thomas Reuters Eikon database
5. The acquirer obtains a majority stake directly (over 50%) or obtains such an amount of stakes that the total number of stakes represent a majority indirectly (share increase) 6. Information with respect to the number of employees working for the acquiring firm has
to be available in either the Orbis or Thomas Reuters databases.
26
7.2 Data collection and analysis
7.2.1 Dependent variable
One of the major problems with research regarding mergers and acquisitions is the measurement of performance of the new firm after the event has occurred. Meglio and Risberg (2011) even go as far as stating that M&A performance should not be seen as one concept, but that academics should make subdivisions related to the various aspects incorporated in the overall notion of M&A performance. Since the understanding of community is something that exists within (a group of) individuals, it would mean we would consider the people on the work floor. The most straightforward and effective measurement for this aspect would be to look at employee satisfaction and social integration (Stahl & Voigt, 2004). However, one would be required to do on-site interviews with the employees involved, which is both costly and time consuming. However, because of the limited time span of this research, and the scarcity of funding, it was decided to not use this approach. In this paper, the dependent variable of performance will be evaluated by exploiting the methods of abnormal returns (AR) and cumulative abnormal returns (CAR). According to Zollo and Meier (2008), this method can indeed reflect the success of the integration process following an M&A event, to a certain extent. Healey, Palepu and Ruback (1992) go even as far as stating that incorporating a short term window, such as a couple of days, can reflect performance in the long run, but this is considered to be a controversial statement and other researchers have not always found the same relationship (Harrison, Oler & Allen, 2005). On the other hand, there are those who state that a two or even three year period is better to evaluate the performance of a merger or acquisition (Jemison & Stikin, 1986). Therefore, it was chosen to focus on the middle long-term with respect to the abnormal returns and calculate them up until 12 months after the event. This approach is also grounded in the more recent literature on the evaluation of abnormal and cumulative abnormal returns (Boehme, Danielsen & Sorescu, 2006 and Dube & Glascock, 2006). In the analysis, the 11th and 12th month after the event are given special consideration since they are at the end of the period. The cumulative abnormal returns, on their turn, reflect the period from 2 months up until and including the 12th month after the event. This method is visualized further on in this section.
27 the day of the event and the closing price on the day itself. However, it could also be used for days or months further away from the event day, in order to evaluate the returns over a longer period of time. The expected return is determined by examining the average daily return of the company during an estimation window, which covers a wide range of days, or months, before the event in order to account for small daily fluctuations of the stock price. Following the example of Datta and Puia (1995) the AR method was deemed to be an appropriate method to evaluate the performance of a cross-border merger or acquisition.
In order to evaluate the long run performance after the completion date, we follow the example of Barber and Lyon (1997), who state that the abnormal returns should be reflected over a month, rather than day by day. The estimation window and event window are therefore based on months before and after the completion of the acquisition (t = 0). Figure 2 is provided as a visualization of the estimation and event windows.
Figure 2. Event Study: Estimation and Event Window.
Thus, for every acquiring company the event(s) of M&A announcements will be evaluated according to this method, in which the abnormal return is calculated according to formula 1:
Formula 1. Abnormal returns.
𝐴𝑅𝑖𝑡 = 𝑅𝑖𝑡− 𝑅̅𝑡
𝐴𝑅𝑖𝑡 = Abnormal return of firm “i” in month “t” 𝑅𝑖𝑡 = Return of firm “i” in month “t”
𝑅̅𝑡 = Average return of the firm “i” in the estimation period 1
𝑡1−𝑡0∑ 𝑡 ∈ |𝑡0,𝑡1|𝑅𝑖𝑡
𝑡0 𝑡1 𝑡2 0 𝑡3
Event Month t = (0)
Estimation window t = (-8; -2) months before announcement
28 Furthermore, now that the abnormal returns can be calculated, we only have to sum up the monthly abnormal returns to derive at the cumulated abnormal returns (Ma, Pagán & Chu, 2009). It was decided to reflect the CARs associated with the announcement event in two different ways, CAR A represents the cumulated ARs of the months leading up to the date of announcement and CAR B reflects the cumulated ARs of the months following the event. The AR in the actual month of the announcement (t = 0) is included in CAR B. Formulas 2 and 3 are presented to visualize both CAR A and CAR B respectively.
Formula 2. Cumulative abnormal returns before t = 0.
𝐶𝐴𝑅 𝐴𝑖 = ∑ 𝐴𝑅𝑖𝑡
𝐶𝐴𝑅 𝐴𝑖 = cumulative abnormal return of firm “i” in the period (−2𝑡; −1𝑡) 𝐴𝑅𝑖𝑡 =Abnormal return of firm “i” in month “t” in the period (−2𝑡; −1𝑡)
Formula 3. Cumulative abnormal returns after t = 0.
𝐶𝐴𝑅 𝐵𝑖 = ∑ 𝐴𝑅𝑖𝑡
𝐶𝐴𝑅 𝐵𝑖 = cumulative abnormal return of firm “i” in the period (0𝑡; +12𝑡) 𝐴𝑅𝑖𝑡 = Abnormal return of firm “i” in the period (0𝑡; +12𝑡)
The value for the cumulative abnormal returns after the event (CAR B) and the values for the abnormal returns in the last month under consideration (AR + 12) are, in this research, considered as the most important indicators of performance in the middle-long run after the M&A. The corresponding values for both the CAR B and AR + 12 are used to compare the success of one merger or acquisition against the other, with higher values indicating more success.
7.2.2 Independent variable
29 research indicate that cross-border M&As between firms from culturally distant countries improves the long-term result of the company (Chakrabarti, Gupta-Mukherjee, & Jayaraman, 2009), whereas others imply that cultural distance leads to destruction of performance and value (Brouthers, 2002). One major difficulty that arises when using this independent variable is that it is a concept, and not a measurable construct which can be indicated with, for example, numbers or values. Therefore it is deemed necessary to approximate the understanding of community as it exists across societies. Based on the findings reported earlier in this research reflecting the differences and similarities of the notion of community between the chosen societies, it was decided to use reflect the narrow understanding by using China and South Korea. The broad understanding is reflected by Sweden and South Africa. With respect to the analysis, a dummy variable is used in which Sweden is represented by the value of 1 and the other cases by a value of 0.
7.2.3 Control variables
Within a scientific research project it is necessary to make sure that the effect of the independent variable on the dependent variable is explained as good as one possibly can. Therefore, it is proposed to use several control variables which could have an effect on post-merger performance as well. There is an array of examples in the literature of variables that could have an effect on post-merger performance, included but not limited to firm size, method of payment, type of merger or acquisition and macro-economic conditions (Ismail, Abdou & Annis, 2011). Nevertheless, it is unrealistic that every variable that has an effect can be controlled for, as it would mean that company specific data has to be accessed that might be secured. Additionally, not all the indicators of acquisition performance are currently known (King, Dalton, Daily & Covin, 2004).
30 Secondly, it is important to consider the type of industry that the firms are operating in. According to Ahern and Harford (2014) mergers are often more successful if the target company is operating in a related industry with regards to the acquiring company. Mergers in unrelated businesses, on the other hand, are often less successful. In order to control for this effect on the post-merger performance, a dummy variable is included in the analysis. The value of 1 will be used when companies operate in a related, or the same, industry. The value of 0, on the other hand, indicates that companies are operating in unrelated businesses. In order to indicate the difference between industries, the SIC standard codes will be evaluated (Datta, 1991). In the data set, the industry relatedness will also be reflected using a dummy variable, where the value of 1 means that the industries in which the target and acquirer operate are related. Logically, the value of 0 indicates that the industries in which they operate are different.
Third of all, it was decided to incorporate GDP per capita as a control variable, this in order to control for the economic difference, and distance, that exist on the country level. More economic distance can be an indication that socio-economic differences exist between the two nations, which can negatively impact the acquisition performance (Chakrabarti, Gupta-Mukherjee, & Jayaraman, 2009). In this case, the level of GDP differs not only per country, but also per year in which the acquisition was completed. Data on the GDP per capita was retrieved from the World Bank (World Bank, 2019), and reported in US dollars in order to make logical comparisons between the different nations. For every case, except two, the GDP of the nation of the acquirer was higher than the GDP in South Africa in the corresponding year. Thus, an increase in GDP is expected to increase the socio-economic distance in all cases but two. In the result section, the regression will be run twice, including and excluding these two cases to observe a possible difference.
7.3 Conceptual Model
In order to visualize the proposed relationships between the independent, dependent as well as the control variables, the following conceptual model is provided below. For the sake of clarity, a short description is given for every concept included.
31 and South Africa. However, because there is no accurate measure or numerical difference for the level of understanding and the degree of similarity, it is impossible to measure exactly what we want to measure (as depicted in Figure 3). Moreover, the development of a new community after the M&A can only adequately be captured by conducting on-site interviews, which falls outside of the time frame and budget that are bound to this paper. Therefore, the conceptual model has to be modified (see Figure 4) in order to perform test with the available data at this point in time.
Figure 3. Conceptual model including independent, dependent and control variables.
Figure 4: Conceptual model which is tested in this paper.
Level of understanding of community (from
narrow to broad).
Development of a new
community after an M&A Performance of M&A. Degree of similarity
in the understanding of community.
Firm size (no. of employees). Industry relatedness. GDP per capita. Economic distance.
Dummy variable: Country from acquirer (South Korea and China represented with the value
of 0,00 and Sweden with the value of 1,00).
Performance of M&A with South African target (measured by abnormal and cumulative abnormal returns).
Firm size (no. of employees).
32
8. Results
8.1 Results for cumulative abnormal returns
Before the analysis is conducted, a short recap is provided with respect to the (cumulative) abnormal return method. The abnormal return is the difference between the return in a certain month minus the normal return of the estimation period. The cumulative abnormal return, consequently, denotes the sum of the abnormal returns over a certain period. The CAR B value, which is used to compare the performance of the acquiring companies, reflects the cumulative abnormal returns from the month of the M&A completion (t = 0) up until 12 months after the event.
8.1.1 Empirical assumptions
Several assumptions are tested before the regressions and tests for differences in performance can be performed. The assumptions of normality, multicollinearity and homoscedasticity are examined for the CAR Bs of all companies included in the data set.
First of all, normality is tested by considering the p-plot of the regression, and the data is considered normal if there are no large deviations from the normal probability line. As can be observed in Appendix A1, there is no perfect correlation with the probability line. However, the deviation is relatively small and not considered large enough to state that the assumption of normality is not violated.
Secondly, the presence of multicollinearity is considered. In order to test for multicollinearity, which illustrates the correlation between the independent variables in a multiple regression equation (Wooldridge, 2013), the variance inflation factors (VIF) are considered. Multicollinearity is considered to be absent if the reported VIF values are below the common threshold of 10 (Penders, 2016). If the VIF values are lower than 10, it is assumed that the predictors in the model do not have a linear relationship with each other. Since the VIF values as given by SPSS are all below 2, and therefore well below the threshold of 10, it is concluded that multicollinearity is not a problem in this data set (see Appendix A2).
33 account for this. Consequently, it was decided to exploit the Weighted Least Squares (WLS) approach (Kaufman, 2013), which alters the weights assigned to the error variances. Smaller variances are given more weight, whereas larger variances are given smaller weights. The WLS were assigned through SPSS, and a newly conducted variable was used for the following regressions.
8.1.2 Results for all cumulative abnormal returns
From Table 1 it can be concluded that the tested model, including the independent variable of country indication and the controls of firm size, industry indication and GDP per capita, only has a small effect on the dependent variable of CAR B. As indicated by R squared, only 11,4 percent of the change in CAR B is predicted by this model. Furthermore, even though the independent variable related to the country of the acquirer reports a relatively large coefficient and thereby is responsible for a relatively large part of the change in CAR B, the relationship is not close to the significance thresholds of either the 1, 5 or 10 percent (p = ,908, see Appendix A4). Running the regression again but excluding the two cases in which the GDP per capita of the nation of the acquirer was lower compared to the GDP in South Africa does not seem to influence these results (see Appendix A5).
Table 1. Model summary of the effect of several predictors on the cumulative abnormal
returns.
R 𝑅2
Adjusted 𝑅2
Std error of
the estimate 𝑅2 Change F
Change df1 df2
Sig. F change
,337 ,114 -,064 21,345 ,114 ,640 4 20 ,640
Predictors: Firm size, industry relatedness, GDP per capita, country
Dependent variable: CAR B (all included) corrected through Weighted Least Squares
34 More important than the regression is the comparison of the performance of the Swedish and Chinese/South Korean acquirers. Here the mean performance, as depicted by the cumulative abnormal returns, is evaluated. Because the sample size for both groups is different, the Mann Whitney-U test was preferred as an alternative to the t-test (Zimmerman, 1987) to compare the mean performance. Additionally, this test using ranking to compare the (cumulative) abnormal returns, a method which automatically controls for the effect of outliers and as a result the robustness is greater compared to the t-test (Coombs, Algina & Oltman, 1996). Therefore, no additional robustness checks are performed. The null hypothesis (𝐻0), and alternative hypothesis (𝐻1), are reported in Formula 4:
Formula 4: Null and alternative hypotheses for the cumulative abnormal returns.
𝐻0: µ1 = µ2 𝐻1: µ1 ≠ µ2
µ1 = The mean of the CAR Bs reported by Swedish companies µ2 = The mean of the CAR Bs reported by all other companies
35
Table 2: Mean rank of all the cumulative abnormal returns per country of acquirer.
Country N Mean rank Sum of ranks CAR B ,00 7 10,29 72,00 1,00 18 14,06 253,00 Total 25
Note: Swedish acquirers are represented by the value of 1. All other cases take the value of 0.
Table 3: Result of the Mann Whitney U test with all cumulative abnormal returns included.
CAR B Mann-whitney u 44,000
Wilcoxon w 72,000
Z -1,150
Asymp. Sig. (2-tailed) ,250 Exact sig. [2*(1-tailed
sig.)]
,270b
Note: P = ,250. Since ,250 > ,10, the difference in performance is not considered to be
significant.
36
Graph 2: Difference in cumulative abnormal returns between Swedish and other acquirers.
Graph 3: Difference in cumulative abnormal returns between Swedish and other acquirers
(outliers removed).
8.1.3 Results For Positive And Negative Cumulative Abnormal Returns
When the performance optically differs to such an extent, it seems appropriate to perform multiple tests with adjusted data sets. Firstly, the companies which had negative cumulative abnormal returns (negative values for CAR B) were excluded from the data set, which means that only companies with a positive accumulative return are included. Following this strategy, however, leads to further reducing the already small sample size. Consequently, the reliability of the outcomes decreases (Morrow Jr & Jackson, 1993), and it becomes difficult to draw definitive and cohesive conclusions from the test. Nevertheless, even a rough
37 impression of the different groups of acquirers can still form an indication on the differences or similarities in performance. Since performing a multivariate regression is not advisable under these circumstances (Nemes, Jonasson, Genell, & Steineck, 2009) it was decided to only run a test for equality of means. As indicated, the number of observations is rather low, which is why the Mann Whitney-U test is preferred as an alternative over the independent sample t-test (Ruxton, 2006) to test for the mean performance.
The mean rank of Swedish acquirers is noticeably higher than the mean rank of the other companies (see Table 4) and this difference is considered to be significant (see Table 5). Thus, the following statement can be made: When acquirers get a positive return, the returns of Swedish acquirers are significantly higher than the returns of Chinese or South Korean acquirers. This in accordance with the visual representation of the returns as shown in Graph 2 and 3. Nevertheless, this relationship can not be proven to the fullest extent since the control variables are not included in the Mann Whitney-U test. Additionally, as previously indicated, the small sample size forms problems for the reliability of the results (Raudys & Jain, 1991).
Table 4. Mean rank for the positive cumulative abnormal returns
Note: Only companies with positive cumulative abnormal returns included. Swedish acquirers
are represented by the value of 1. All other cases take the value of 0.
Table 5. Result of the Mann Whitney U test with only positive cumulative abnormal returns
included.
Note: P = ,021. Since ,021 < ,05 the difference in performance is considered to be significant
on a 5% level.
Country N Mean rank Sum of ranks
0,00 3 2,67 8,00 1,00 12 9,33 112,00 Total 15 CAR B Mann whitney u 2,000 Wilcoxon w 8,000 Z -2,309
Aysmp. Sig. (2-tailed ,021 Exact sig. (2*(1-tailed
sig.))
38 Moreover, the same process is repeated for the companies that have a negative CAR B. The division between Swedish and Chinese/South Korean firms is now rather fair, with 6 and 4 firms respectively (see Table 5). Before the Mann Whitney-U test was performed for the negative cases, the negative sign was removed, which means that the input consisted only of the absolute values of the cumulative abnormal returns. This was done because a negative sign can influence the test statistic whereas the absolute value indicates the same difference between the normal and abnormal returns but forms no threat to the test statistic. Concluding from Table 6, it can be stated that the mean rank of the Swedish firms is lower, which would indicate worse performance. Table 7 indicates that the difference in performance is in fact significant. Based on these results, the following is stated: When both groups of acquirers report negative returns, the returns of Swedish acquirers are significantly lower than the returns of Chinese and South Korean companies. This is again in accordance with the visual representation of the cumulative abnormal returns in Graph 2 and 3. Nevertheless, caution is advised with the real life application of these results because of the small sample size and the absence of control variables in the Mann Whitney-U test.
Table 5. Mean rank for the negative cumulative abnormal returns.
Country N Mean rank Sum of ranks
0,00 4 2,50 10,00
1,00 6 7,50 45,00
Total 10
Note: Only cases with negative cumulative abnormal returns included. Swedish acquirers are
represented by the value of 1. All other cases take the value of 0.
Table 6. Result of the Mann Whitney U test with only positive cumulative abnormal returns
included.
CAR B
Mann whitney u ,000
Wilcoxon w 10,000
Z -2,558
Aysmp. Sig. (2-tailed ,011 Exact sig. (2*(1-tailed sig.)) 0,10
Note: P = ,011. Since ,011 < ,05 the difference in performance is considered to be significant
39
8.2 Results for abnormal returns in the 12
thmonth
In order to provide a more inclusive picture on the performance of Swedish acquirers versus the other group of acquirers (Chinese and South Korean), it was decided to run several tests with respect to the monthly abnormal returns reported for each company. The idea is that after 12 months the performance has somewhat stabilized and the acquisition has been either successful or a failure. By examining the monthly values independently we can exclude the optimism, or pessimism, that might have existed closely after the merger. In other words, the cumulative abnormal returns can be positive because the firm is performing well in the first months after the acquisition. However, after 12 months, it might be the case that some companies reported negative returns because problems arise with the operationalization, whereas this is not shown by examining only the cumulative abnormal returns. For three Swedish companies it was the case that their cumulative abnormal returns were positive, whereas the abnormal return in the 12th month was negative. For 1 Chinese case the opposite was true, the abnormal return in the 12th month was positive but the cumulative abnormal returns were negative.
Just as with the CAR B values, we will show the regression outputs with respect to all the abnormal returns and perform either independent t-test or Mann Whitney-U tests for all the abnormal returns, the positive cases and the negative cases respectively. In order for a case to be included in the sample set of positive returns, the company has to have positive returns for the last two months (t = 0 +11 and t = 0 + 12). This was done to filter the results and exclude the companies that show high fluctuations with respect to the abnormal returns.
40 considered to be significant and shows a very low coefficient (see Appendix B4). This means that it can not be held responsible for a lot of variation in the monthly abnormal returns.
Table 6. Model summary of the effect of several predictors on the abnormal returns in the 12th
month.
R 𝑅2
Adjusted 𝑅2
Std error of
the estimate 𝑅2 Change F
Change df1 df2
Sig. F change
,890 ,792 ,749 1,029 ,792 18,138 4 19 ,000
Predictors: Firm size, industry relatedness, GDP per capita, country
Dependent variable: AR + 12 (all included) corrected through Weighted Least Squares
Correlation forms, just as with the CAR B results, no problem (see Appendix B5). The highest correlation occurs, again, between the GDP per capita and country variable, but as GDP per capita can be seen as an aspect of the country, this correlation is disregarded. Another Mann Whitney-U test was conducted, with the same null and alternative hypothesis as stated in Formula 4. From Table 7 it becomes clear that the mean rank of Chinese and South Korean companies is slightly higher as compared to the Swedish mean rank, but the difference is not considered to be significant (see Table 8).
Table 7. Mean rank for the all the abnormal returns in the 12th month. Country N Mean rank Sum of ranks
0,00 7 13,43 94,00
1,00 17 12,12 206,00
Total 34
41
Table 8. Result of the Mann Whitney U test with all the abnormal returns in the 12th month included.
AR + 12
Mann-whitney u 53,000
Wilcoxon w 206,000
Z -,413
Asymp. Sig. (2-tailed) ,680 Exact sig. [2*(1-tailed sig.)] ,710b
Note: P = ,680. Since ,680 > ,10 the difference in performance is not considered to be
significant on a 10% level.
Nevertheless, a similar image to the cumulative abnormal returns can be observed when we visualize the abnormal returns of the acquirers in the 12th month after the completion of the M&A (see Graph 4). The results from Swedish acquirers are again noticeably higher as well as lower compared to the other cases (see Appendix B6). Interestingly enough, the outlier is again on the Swedish side, but this time on the negative side of the spectrum, whereas it was on the positive side for the cumulative abnormal returns (see Graph 2). This indicates that it was a right decision to evaluate the performance by examining both the CAR B and the AR + 12 values, since incorporating only one of them might have provided an incomplete picture of the situation and the performance of the acquirers.
42 8.2.1 Results for positive and negative abnormal returns in the 12th month
Again, it is optically justified to test the difference in performance in more detail. From Graph 5 it is clear that, when the returns are positive, the Swedish companies get higher abnormal returns, on average, compared to Chinese and South Korean acquirers. The Mann Whitney-U finds similar results since the mean rank of Swedish companies is significantly higher (see Tables 9 and 10), which indicates that the abnormal returns for Swedish acquirers are, on average, higher. Based on these results, the following conclusion can be drawn: When the acquirers report positive abnormal returns in the 12th month after the completion date, the abnormal returns of Swedish companies are higher compared to Chinese or South Korean acquirers. However, caution is advised with interpreting the results because the small sample size decreases the reliability.
Table 9. Mean rank for the positive abnormal returns in the 12th month. Country N Mean rank Sum of ranks
0,00 4 3,25 13,00
1,00 8 8,13 65,00
Total 12
Note: Only cases with positive abnormal returns in the 12th month included. Swedish acquirers are represented by the value of 1. All other cases take the value of 0.
Table 10. Result of the Mann Whitney U test with only positive abnormal returns in the 12th
month included.
AR + 12
Mann whitney u 3,000
Wilcoxon w 13,000
Z -2,208
Aysmp. Sig. (2-tailed ,027 Exact sig. (2*(1-tailed sig.)) ,028
Note: P = ,027. Since ,027 < ,05 the difference in performance is considered to be significant
on a 5% level.
43 negative abnormal returns in the 12th month after the acquisition, the abnormal returns of Swedish acquirers are significantly lower compared to the Chinese and South Korean acquirers. Again, just as with the other results, the sample size is rather small which means that caution is advised with regards to the reliability.
Table 11. Mean rank for the negative abnormal returns in the 12th month. Country N Mean rank Sum of ranks
0,00 3 2,00 6,00
1,00 9 8,00 72,00
Total 12
Note: Only cases with negative abnormal returns in the 12th month included. Swedish acquirers are represented by the value of 1. All other cases take the value of 0.
Table 12. Result of the Mann Whitney U test with only negative abnormal returns in the 12th
month included.
AR + 12
Mann whitney u ,000
Wilcoxon w 6,000
Z -2,496
Aysmp. Sig. (2-tailed ,013 Exact sig. (2*(1-tailed sig.)) ,009
Note: P = ,013. Since ,013 < ,05 the difference in performance is considered to be significant
