Radboud University Nijmegen Nijmegen School of Management
Master Business Administration, International Business
Name: Koen Andel
Student number: S4337425
Supervisor: prof. dr. N.A. Dentchev
Second examiner: dr. ir. G.W. Ziggers
Date: 20-05-2020
The impact of CSR similarity on
cross-border M&A success
2
Abstract
This thesis looks at the impact of corporate social responsibility (CSR) on cross -border merger and acquisition (CBM&A) success. Using a sample of 196 cross-border deals from 2002 onwards, the effects of CSR similarity on CBM&A deal completion, deal rapidness and post-deal integration are being examined. Additionally, the moderating effect of firms’ overall CSR scores on this relationship is investigated. The hypotheses are tested using logistic regressions, cox-hazard models and ordinary least squares regressions. The main results show that CSR similarity has no statistically significant impact on CBM&A success and that firms’ overall CSR scores have no significant moderating effect. However, when looking at the individual components of CSR this study finds similarity on emissions to negatively impact deal rapidness, and similarity on product responsibility to positively impact the post-deal integration process. Furthermore, environmental innovation and workforce are moderated by firms’ overall CSR score.
3
Table of Content
1. Introduction ... 5
2. Literature review ... 8
2.1 M&A research ... 8
2.2 Corporate Social Responsibility ... 9
2.3 Linking Corporate Social Responsibility to M&A ...12
3. Methodology...16
3.1 Data and sample description ...16
3.1.1 CSR data ...16 3.1.2 CBM&A data ...16 3.2. Dependent variables ...18 3.3 Independent variable ...19 3.4. Moderator variables ...19 3.5 Control variables...20 3.6 Analysis ...21 4. Results ...23 4.1 Descriptive statistics ...23 4.2 Correlation matrix ...24 4.3 Test of hypotheses ...25 4.4 Robustness checks ...28
5. Discussion and conclusion ...30
References ...34
Appendix A ...40
Appendix B ...41
Appendix C ...44
4
List of Figures
Figure 1 Conceptual model 15
Figure 2 CSR scores 16
List of Tables
Table 1 Search results 17
Table 2 Descriptive statistics 23
Table 3 Correlation matrix 24
Table 4 Effect of CSR Similarity on deal completion and deal rapidness 25
Table 5 Effect of CSR similarity on post-deal integration 27
Table 6 Classification table logistic regression model 1 40
Table 7 Classification table logistic regression model 2 40
Table 8 Classification table logistic regression model 3 40
Table 9 Effect of CSR similarity (individual components) on deal completion 41 Table 10 Effect of CSR similarity (individual components) on deal rapidness 42 Table 11 Effect of CSR similarity (individual components) on post deal integration 43 Table 12 Effect of CSR similarity (Euclidian) on deal completion and deal rapidness 44
Table 13 Effect of CSR similarity (Euclidian) on post-deal integration 45
Table 14 Effect of CSR similarity on post-deal integration (Net profit income) 46
List of abbreviations
CAR Cumulative abnormal return M&A Merger and acquisition
CBM&A Cross-border merger and acquisition MNE Multinational enterprise
CD Cultural distance OLS Ordinary least squares
CSR Corporate social responsibility ROA Return on assets
ESG Environmental, social and governance VIF Variance inflation factor
EUC Euclidian SRI socially responsible investments
5
1. Introduction
Economic globalization is an irreversible trend which continues to flatten the world and increase interdependence of world economies. Swift scientific and technological advancements greatly reduced and continue to reduce transportation and communication costs. At the same time, international trade agreements under the WTO like the GATT and later the Marrakesh Agreement further lowered international trading costs by cutting tariffs, enabling more and more global trade (Shangquan, 2000). Along with an increasingly globalized world came an explosive rise in the number of cross -border mergers and acquisitions (CBM&As). In 1998, about 23% of total merger volume was caused by CBM&As and by 2007 this was about 45%, creating a multi-trillion dollar segment (Erel, Liao & Weisbach, 2012). It is therefore not surprising that CBM&As and M&As in general have gained a lot of academic attention (Hitt et al., 2012).
The reason to engage in mergers and acquisitions (M&As) is to increase a company’s competitive advantage (Hitt et al., 2012). Creating synergies is a huge opportunity to reinforce one’s competitive position. This can be done in a variety of ways, but integrating the two businesses is always part of it (Hitt et al., 2012). Research shows that failed integration is the number one reason for failed M&As (Hay Group, 2007). Examples are the failed mergers of Daimler-Chrysler (1998) or the new vs old media clash of AOL-Warner (2000). In almost all failed M&As, cultural clashes were blamed for the negative results (Teerikangas & Véry, 2012). In this sense, culture has become sort of a black box. Managers did not know exactly how it was responsible for the lack of synergy creation, but it certainly was culpable for it. Research was quick to follow and found copious effects of culture on M&A success, often incongruent and conflicting with one another (Teerikangas & Very, 2008; Weber et al., 2009). An important contribution to this debate is offered by Stahl and Voight (2008), who reconciled the conflicting perspectives and state that the effect of cultural differences on M&A performance depends on contingencies, including the degree of relatedness and the dimensions of cultural distance. This cultural distance is an important contributor to integration success or failure as well as pre-deal variables, making it a significant determinant in M&A practices (Alexandridis et al., 2015). Theory states that the more alike two cultures are, the smoother the two organizations can integrate. Corporate culture is described by O’Reilly and Chatman (1996) as “a system of shared values defining what is important, and norms, defining appropriate attitudes and behaviors, that guide members’ attitudes and behaviors”.
An increasingly important topic within a company’s culture is corporate social responsibility (CSR). CSR is often used by companies to communicate about their vision, mission and values (PricewaterhouseCoopers, 2010). It is also driven by stakeholders’ preferences (Benabou and Tirole,
6 2010) and therefore reflects the shared beliefs and values within an organization, making it an important part of corporate culture (Hoi, Wu & Zhang, 2013). CSR has been researched a lot over the last decades and its effect on social topics and legitimacy are well known (Werther Jr. & Chandler, 2010). Research also links CSR to financial performance and stock-market returns (Carroll & Shabana 2010; Orlitzky et al., 2003; Margolis & Walsh, 2003). Even though there is no consensus on these effects, there are signs that companies engaging in CSR are performing better than the ones that lack CSR. Given the strategic importance of M&As and its huge impact on competitive advantage and performance, the next step is to look for the effects of CSR on M&As. Despite the benefits, CSR dimensions in M&A decisions have been more or less lacking in finance literature (Aktas et al., 2011). However, anecdotal evidence indicates that acquiring companies are actively looking at target companies’ CSR performance before engaging in the actual take-over. General Electric had a special division of auditors look into target companies’ environmental performance as part of their due diligence (Berchicci, Dowel & King, 2012), Unilever took over Ben and Jerry’s to learn from their CSR activities and L’Oréal took over the Body Shop for their environmental protection performance and its animal-free experiments (Austin & Leonard, 2008). It seems that companies are actively looking for and buying CSR, hinting at the importance of CSR in strategy formulation. The few Business studies that do link CSR to M&A confirm this stance, finding numerous (positive) effects of higher CSR performance in the acquiring company during an M&A (Arouri, et al., 2019; Deng, et al., 2013) and in the targeted company (Gomes, 2019; Gomes & Marsat, 2018; Qiao & Wu, 2019; Aktas, et al., 2011). Even though there is abundant research linking similarities in (corporate) culture to (CB)M&A success (e.g. Datta, 1991; Chatterjee, 1992; Weber & Camerer, 2003), the effects of similarities in CSR practices on M&As success are hardly investigated. Research of this similarity on CBM&As seems to be completely absent. Because CSR deals with core beliefs, norms and values of a company and its employees and is rooted in, and perhaps is a great part of, their corporate culture, it can potentially have a significant impact on M&A success. Therefore, this master thesis aims to fill this gap by answering the following research problem:
“What is the impact of CSR similarity on cross-border M&A success?”
Of course, success can be measured in a variety of ways. Therefore, a few important variables inherent to success are researched. In order to adequately assess the research problem, two research questions are composed. Because of the immense time and costs associated with M&As and the impact this can have on both the acquiring and the targeted company, it is important to successfully complete the process, and, preferably, in as little as time as possible. As previous research shows, up to 25% of M&As are withdrawn after they were announced (Holl & Kyriazis, 1996), resulting in huge financial losses of
7 up to 6% of the transaction value (Rosenkranz & Weitze, 2005) and serious damages to a company’s reputation (Luo, 2005). Completing the deal faster does not only reduce time spend, costs and uncertainty, it is also an important component of post-deal integration (Bereskin et al. 2018). Knowing the factors that could increase the rapidness and chances of completing the deal can therefore prevent time and money wasting ventures. The first research question therefore is:
“What is the impact of CSR similarity on cross-border deal completion?”
However, completing the deal is only one of two mayor obstacles in achieving a successful merger. When a deal is completed the acquiring company is left with another huge challenge. As already mentioned, a huge part of failed M&As were blamed on failed integration. Successfully integrating the two businesses comes with an abundance of difficulties, not the least being cultural differences. Research shows similar cultures can ease the integration process, but can also lead to less synergies created because of the lack of a learning effect which two different cultures can deliver. It is therefore important to know the effects of CSR on the integration process. The second research question sounds:
“What is the impact of CSR similarity on post-deal integration”
To assess the research problems and its attached research questions, the remainder of this master thesis is organized in four chapters. First, a literature review will be given in chapter 2 which presents in further detail the literature on M&A, CSR and cultural distance. Next, the methodology is formed in chapter 3, in which details of data and sample, variables and analyses are explained. Chapter 4 will present the results of the analyses, including descriptive statistics and robustness checks. Lastly, chapter 5 contains the discussion and conclusion. It is used to discuss the results, reflect on this thesis and to give recommendations for future research.
8
2. Literature review
In this chapter, relevant literature and important constructs will be discussed. Furthermore, the link between the main subjects in previous research will be given and the hypotheses will be constructed. First, M&As are reviewed, afterwards relevant literature on CSR is discussed and then the combination of these two fields will be evaluated.
2.1 M&A research
M&As are an expensive and time-consuming occupation. The reason for the widespread application of the practice is the potential competitive advantage that can be gained from a successful merger or acquisition via synergy creation, for example by acquiring complementary capabilities or the removal of redundant activities (Hitt et al., 2012). However, many of the announced M&As fail to deliver the expected results (Hitt et al., 2012). Damages can range from gaining slightly less revenues than anticipated to such vigorous blows to a company’s financial or intangible resources that it goes bankrupt. Some studies even find 90% of European M&As failing to reach the set objectives (Hay Group, 2007). Considering the substantial associated risks and rewards, it is only logical that practitioners want to know as much as possible about the factors influencing M&A success. It therefore would not raise too many eyebrows to find that M&As are one of the most studied topics in International Business research. Hitt et al. (2012) give a comprehensive overview of the most common independent variables in M&A research, with the effects of diversifying into related or unrelated businesses, firm size and acquisition experience topping the list. Even though there is widespread belief that integration issues are the main reason for M&As’ failure to create value (Olie, 1990), cultural effects are less frequently researched. It are exactly these cultural effects that play a huge role in the post-deal integration process (Stahl & Voigt, 2008).
Cultural distance literature aims to find the influence of cultural differences and similarities on M&As. Two opposing theories set the stage for investigating the impact of cultural distance on M&A performance. It is the culture clashes versus culture synergy debate (Alexandridis et al., 2015). The culture clashes theory states that cultural differences will lead to friction and create issues in the integration process and therefore, companies which have a similar cultures will work well together. The culture synergy theory sees cultural distance as an opportunity to learn from each other and create value, and argues that firms with different cultures will perform better because of the potential learning effect (Alexandridis et al., 2015).
9 The two main kinds of culture that influence CBM&As are national culture and corporate culture. Studies of cultural dimensions in M&As started in the 1980s with research on domestic deals in the United States. Therefore, the impact of the clash of corporate cultures during a merger on firms (financial) performance was the main topic under investigation, as national cultural differences were not at play yet (Teerikangas & Véry, 2012). A lot of positive effects of cultural fit are found, but contradictions do exist. Datta (1991), Chatterjee et al. (1992) and Weber & Camerer (2003) all find that cultural differences negatively impact firm performance. Weber (1996) did not find a direct relationship of cultural differences on firm performance, but he did find these differences negatively influencing the integration process and commitment. Another research found a positive relation to exist, but the functional background of the top management team was used to proxy for culture and might therefore not adequately measure cultural differences so much as management’s overall experience in different fields (Krishnan et al., 1997). Larsson and Risberg (1998) do find a genuine positive relation between corporate cultural distance and synergy creation. In the 1990s, the rise of European cross-border M&As caused national culture to become an important dimension in the study of culture in CBM&As (Teerikangas & Véry, 2012). Olie (1990) was the first to include the impact of national culture next to corporate culture in CBM&As, with national culture influencing a firm’s corporate culture.
Not only the post deal integration process is influenced by culture, also pre-deal processes can be impacted by it. During the process of negotiation conflicts and incomprehension can arise due to cultural differences (Tse, Francis & Wall, 1994). These conflicts can lead to harder renegotiations and negatively impact the likelihood of deal completion. The awareness of cultural differences and their potential difficulties can also impact selection criteria in choosing the target. A growing number of research suggests that corporate culture is more important in M&As than previously believed (Vezér & Morrow, 2017).
2.2 Corporate Social Responsibility
The concept of CSR has been around for centuries, with formal writing on CSR beginning mainly in the 1950s (Carroll, 1999). Since then, CSR has gained more and more importance in both the academic and corporate world and the general public. It has become an almost unavoidable pressure for companies around the world to implement CSR policies (Porter & Kramer, 2006). This pressure comes from a variety of stakeholders, including customers, suppliers, employees, governments, media, activists, communities and shareholders, whom all have different, sometimes conflicting goals (Dahlsrud, 2008).
10 Despite the widespread relevance of CSR, a general definition remains absent (Dahlsrud, 2008). There are ample definitions available but they all seem to be biased towards specific interests, preventing consensus of the concept (Van Marrewijk, 2003). Furthermore, the idea of what constitutes as (corporate) responsibility changes over time, dragging the meaning of CSR along with it (Carvalho, Jonker & Dentchev, 2014). However, a few topics seem to have rooted in the core notion of contemporary CSR; social, environmental, economic, stakeholder and voluntariness, with a 97% probability that at least three of these dimensions are used in a random definition (Dahlsrud, 2008). In 2001, the European Commission defined CSR as “a concept whereby companies integrate social and environmental concerns in their business operations and in their interaction with their stakeholders on a voluntary basis” and simplified this to “the responsibility of enterprises for their impacts on society” in 20171. Useful for this master thesis is to see CSR as an umbrella term for the reflection of the social
consequences of business success (Matten and Moon, 2008).
Partly due to globalization and climate change, people’s interests and concerns about global environmental and social issues kept increasing. A shared belief that CSR benefits both society and business became more widespread and the idea of ‘doing well by doing good’ was introduced (Falck & Heblich, 2007). Along with research that positively links CSR to consumer behaviour, more and more companies engaged in CSR practices, resulting in 80% of Fortune-500 companies addressing CSR on their websites by the end of the 20th century (Sen & Bhattacharya 2001). As Werther Jr. and Chandler
(2010) state, CSR has become an essential component of strategy and is not merely a necessary expense in order to comply to regulation and social pressures. It can offer strategic and economic benefits, maximize competencies and create competitive advantages as well as help evade criticism and other forms of potentially value decreasing attacks (Werther Jr. & Chandler, 2010).
Research found positive relations between CSR and a wide range of topics such as risk reduction, legitimacy and reputation, competitive advantage, synergy creation and stakeholder relationships (Carroll & Shabana, 2010; Dentchev, 2004). For practitioners, however, perhaps the most interesting and important influence any factor could have is its influence on a firm’s financial performance. Not surprisingly, literature linking CSR to corporate financial performance is rich and contentious. Due to the (in)direct influences of a multiplicity of factors, neutral and negative effects have been found, but a lot of positive effects are concluded as well. Arguinis and Glavas (2012) did a meta-study of close to 700 articles and books about CSR and found a small but positive relationship between CSR and financial performance to exist. Likewise, Qian, Junsheng and Shenghua (2016) had similar results in their
11 analysis concluding an enhancing effect of CSR on financial performance, especially in the developed world.
As Aguinis and Glavas (2012) assert, little is understood about the underlying mechanisms and factors that influence the relationship of CSR on financial performance. A salient detail is brought into the debate by Brammer and Millington (2008) who challenged the idea of a linear relationship between CSR and financial performance and found that the relationship might exist in an U shape. This curvilinear form was earlier found in the relationship between socially responsible investments and financial returns by Barnett and Salomon (2002). In short, they argue that firms that have really low or really high CSR performance achieve the best results. In contrast, much earlier, Ullmann (1985) found an inverted U-shaped correlation between social and economic performance, suggesting that there is an optimal level of CSR and corresponding resource allocation.
As CSR is often used by companies to communicate about their vision, mission and values (PricewaterhouseCoopers, 2010) and are driven by stakeholders’ preferences (Benabou and Tirole, 2010), it reflects the shared beliefs, norms and values within an organization (Hoi, Wu & Zhang, 2013). It also covers a lot of important dimensions of corporate culture such as governance, employees, product and customers (Alexandridis et al., 2015) and can therefore be defined as being (an important) part of corporate culture. By doing so, cultural distance literature offers insights on the potential effects of CSR on M&As.
CSR similarity is defined in this master thesis as two firms having more or less the same score on the individual components of CSR and can be compared to a low degree in cultural distance. Stahl and Voigt (2008) tried to reconcile the conflicting perspectives in the cultural distance literature and state that the effect of cultural differences on M&A performance depends on contingencies, including the degree of relatedness and the dimensions of cultural differences. It is therefore important to note that comparability or similarity of CSR practices in this sense does not mean that two companies which have the same overall CSR scores are similar. In order for them to be similar, they must score more or less the same on the individual components of CSR. To illustrate, imagine two companies which both score high on their overall CSR performance. The first company focusses on employee satisfaction and governance while the other is mainly concerned with environmental responsible practices. Even though both have high performing overall CSR scores, a merger between these two firms is not seen as a merger between two companies with similar CSR practices.
12
2.3 Linking Corporate Social Responsibility to M&A
After the effects of CSR on organizational legitimacy, reputation, competitive advantage and performance became more well known, the inevitable question of its impact on M&A success arose. Insights could be found in earlier research of socially responsible investments (SRI). For example, Aktas et al. (2011) found acquirer gains to increase when it targeted a company which was more socially and environmentally responsible. Likewise, Derwall et al. (2005) found SRI to improve portfolio performance. However, even though social and environmental performance are important aspects of CSR, it does not quite cover the whole spectrum of the concept. Deng et al. (2013) were the first to actually look at an acquiring company’s CSR performance and link it to M&A performance. They found that high CSR acquirers, compared to low CSR acquirers, led to higher announcement stock returns, larger increases in long-term operating performance and stock returns, and a higher likelihood and shorter duration of deal completion. Qiao et al. (2018) also concluded a positive relation between higher CSR performance of the acquirer and long-term M&A performance. Their sample consists of M&As conducted in China, suggesting that CSR might also be rewarded in emerging economies. Later, Arouri et al. (2019) conducted the first study examining the effect of CSR on CBM&A completion uncertainty, moving the field to an international context. They found better CSR performance of acquiring companies to be associated with lower uncertainty. Positive effects are also found when looking not at the acquirer’s CSR performance, but at the target company’s CSR. Gomes (2019) found target companies with higher CSR scores are more likely to be targeted. Moreover, Gomes and Marsat (2018) and Qiao and Wu (2019) concluded that such companies receive higher bid premiums due to their intangible strategic asset.
There is very little research that looks at CSR similarity and links it to M&A dimensions. Bereskin et al. (2018) use the similarity in firms’ CSR characteristics to proxy for cultural similarity and find that similar firms are more likely to merge, receive greater synergies, have superior long-run operating performance and fewer write-offs of goodwill. There sample however, only consists of domestic deals. Another research looks at environmental, social and governance (ESG) compatibility and finds that deals between companies with more comparable ESG scores outperform companies with distant stances on ESG (Vezér & Morrow, 2017). Alexandridis et al. (2015) also look at ESG scores, finding numerous positive results of similarity, including a more rapid deal completion and an increased likelihood of completing an announced deal in domestic and CBM&As.
Even though the importance of cultural fit is well-known among one stream of research, the other stream highly values the learning effects generated by cultural distance. They argue that companies should try to target culturally distant firms to gain more benefits. The only way to end this debate is by opening the black box and look at the individual components of culture and see how they influence
13 merger success. Therefore, this thesis aims to do exactly that. The effects of CSR similarity on M&A success are only investigated in a handful of studies, creating a huge gap in our knowledge in this field. This is even more so for the influence of CSR similarity on CBM&As. The reasons to look at CSR are plentiful. First, it is an excellent proxy for corporate culture as it looks at dimensions of governance, employees, products and customers. It also encompasses the shared beliefs and values of the organization, offering a revealing insight in the company’s norms. Second, measuring corporate culture in a quantitative way is almost impossible. CSR data however, quantifies a company’s corporate culture based on more objective criteria, enabling relevant statistical analysis (Alexandridis et al., 2015; Bereskin et al., 2018).
According to cultural distance literature, the main positive effects from cultural distant companies stem from its learning effects and the main positive effects from cultural similar companies come from mutual understanding and its smooth integration process. Even though this master thesis looks at CSR from a corporate culture perspective, there are some important distinctions to be made that possibly influence the outcomes. This thesis hypothesizes in favor of the cultural clashes theory, arguing that CSR similarity between two merging firms will yield more positive results than CSR distant firms. Apart from the positive effects of similarity as mentioned in literature the following reasonings support this stance. First of all, the learning effects that the merging of two cultural different companies offer, as theorized by the cultural synergy school, will be less profound when it comes to CSR. Perhaps more than corporate culture itself, CSR can be seen as a strategic asset (Qiao & Wu, 2019). When a target company is known for its high CSR performance, and the acquiring company actually bought the target for this reason, it will be very careful not to contaminate the target with their own culture, limiting the (positive) learning effects. Unilever for example, was very careful not to ruin Ben & Jerry’s unique culture with their own (Austin & Leonard, 2008). If it does integrate fully, you could say the target also learns the (bad) habits from the acquirer. In the meantime, a M&A like this will have experienced the negative effects of their cultural distance, mainly less smooth negotiations and post-deal integration. If the acquiring company does not care for CSR performance, then the practice will be ignored and likely die-out altogether. Second, CSR is such a complex and multifaceted concept that even companies with very high CSR similarity still have much to learn each other, becoming even better at achieving their socially responsibly goals. Think of two companies which both care for the environment. One might invest a lot in symptomatic treatment, such as carbon dioxide compensation and supporting environmental charities while the other is busy producing less waste, installing solar panels etc. Even though they would both care a great deal about the environment, they could still learn each other a lot. In other words, there are a multiple ways to achieve the same CSR goals. Another reason for the existing learning potential between CSR similar firms is that when looking at corporate culture as a
14 whole, there often is not a preferred direction in which the merged cultures should evolve. In the case of CSR however, better CSR performance is often favored and is a clear and logical objective to try to achieve. Both companies will likely be open towards new ways to achieve their CSR goals, particularly if they both have high overall CSR scores. In short this thesis hypothesizes that the learning effects of CSR distant firms are limited (or also go the other way around, leveling the effects), and the learning effects of CSR similar firms are far from zero, minimizing the positive effects of the cultural synergy school and enlarging the positive effects of the cultural clashes school.
An important factor contributing to failed M&As is a failed integration. Because of mutual understanding due to cultural similarity CSR similar firms are more likely to complete a deal after its announcement. An additional factor is the international context. CSR similarity might function as a stabilizing factor in the national cultural differences and other corporate cultural elements. CSR similar firms probably understand each other better, increasing the chances of deal completion and making them more rapid. Or, as stated differently by Alexandridis et al. (2015), different business behavior and ethics might create frictions resulting in less chances of completing the deal. The following hypotheses are:
Hypothesis 1: CSR similarity increases cross-border deal completion chances
Hypothesis 2: CSR similarity increases cross-border deal completion rapidness
In line with literature as described in the previous paragraphs, cultural similarity makes the integration process go smoother. The same rationale goes for CSR similarity and therefore the second hypothesis is:
Hypothesis 3: CSR similarity increases the cross-border post deal integration process
Research did not only find positive effects for cultural similarity, but also for high overall CSR performance in either the acquiring or the target company. These studies did not look at the CSR similarity. However, it is possible that apart from the positive effects gained by merging with a similar firm, they also benefit from having high overall CSR performance as found by earlier research. It is therefore possible that merging firms which both have a high overall CSR score and have similar CSR practices experience better results than two firms which have a low overall CSR score and similar CSR practices. I must be careful not to mistake the total effect of similarity for similarity itself. For example, the total effect of CSR similar firms might be X, but when zoomed in, the effect of a high-high couple is X+10 and a low-low couple X-10. It would be wrong to say that CSR similarity has an effect of X, and that companies should therefore try to aim for CSR similar firms (if X is positive and <10), while in reality only if they too have a high CSR overall score it will create beneficial results. Because a high CSR
15 performance offers a lot of benefits both during and outside of M&As, it is likely that the positive effects of CSR similarity are higher in high-high couples. In order to make sure this thesis actually measures similarity and not just the effects of either a high-high couple or low-low couple and also to gain deeper understanding of the underlying structures that influence the effects, the final hypothesis is as follows:
Hypothesis 4: The effects of CSR similarity on CBM&A success are stronger for cross-border deals in which both companies have high overall CSR scores than for deals in which both companies have low
overall CSR scores.
16
3. Methodology
3.1 Data and sample description
3.1.1 CSR data
To find CSR data for the companies involved in the deals, the Thomson Reuters ESG Scores database (hereafter: Thomson) is used. It is an enhancement and replacement of the ASSET4 ratings. Companies are scored based on three subjects; environmental, governance and social. Each subject consists of 3 or 4 components (10 in total), which in turn contain a multitude of sub-components, totaling 178 different scores among the three themes. See figure 2 for a visualization. For this master thesis, the components of each subject are used to measure similarity. The database consists of 8,044 companies scored on their ESG practices going back to 2002. Since an acquired target is often not listed anymore, there is insufficient CSR data available in the year the M&A took place. Therefore, following related research (e.g. Bereskin et al. (2018) Alexandridis et al. (2015)), CSR data of both the acquirer and the target firm is taken from the year previous to the year in which the deal was announced.
Figure 2. Visualisation of ESG data.
Source: Thomsons Reuters ESG Scores, May 2018. http://zeerovery.nl/blogfiles/esg-scores-methodology.pdf
3.1.2 CBM&A data
The data on mergers and acquisitions are derived from Zephyr. This database collects data on M&As worldwide from various sources and contains detailed information about the deals. Zephyr offers information on the date of the announcement, deal status, individual deal details, and target and
17 acquirers profiles. This database consists of close to 2 million deals. Random cross-checks of the data between Zephyr and company annual reports are done to assess the accuracy of the information. Table 1 shows the search criteria and the retrieved amount of deals from Zephyr and refinements.
Table 1. Search results
Search criteria Results
Acquirer and target ISIN numbers existent in Thomson 56,632
Deal type: acquisition or merger 2,661
Initial stake max. 49.9% and final stake min. 50.1% 2,099
Cross border deal 1,012
Deal value min. 1 million euros 894
Include only completed, completed-assumed and withdrawn deals 316 Manual refinement
Delete missing data 196
- Of which completed 127
- Of which withdrawn 69
Zephyr offered a sample of 316 deals. Following previous literature (e.g. Bereskin et al. (2018), Deng et al. (2013), Alexandridis et al. (2015)), the minimum deal value is at least 1 million euros and the initial stake is less than 50%. As a clear change of control is needed for this research, the final stake needs to enable a majority position. Therefore the final stake needs to be at least 50.1%. Additional data is derived from Thomson Reuters Datastream to measure integration via synergy creation. Deals lacking an announcement date and deals that were completed-assumed were manually searched and added using the Thomson Reuters Eikon database. Out of the 31 deals that lacked an announcement date, 18 were found and added. Since most cancelled deals were withdrawn in the same year as they were announced (69 out of 88), the withdrawn date was used to extract CSR data for the remaining 13 deals. This has no influence on the outcome of hypothesis 1b since all deals lacking an announcement date were withdrawn deals. Furthermore, out of 9 deals that were marked as completed assumed, 4 were changed to completed and 1 to withdrawn. The remaining 4 deals were deleted. 2 deals were removed which were marked by Zephyr as being cross border but were actually
18 domestic. Lastly, if CSR data was not available for the selected year of either the acquiring or targeted company, the deal is deleted.2 196 deals remain.
3.2. Dependent variables
The first and second dependent variables are Deal Completion, and Deal Rapidness. Following Deng et al. (2013) and Bereskin et al. (2018) for both dependent variables, deal completion is measured using a dummy variable which equals 1 if the announced deal is completed and 0 otherwise. Deal rapidness is measured by looking at the time passed in days between the announcement date and the completion date.
The third dependent variable is Integration. It is measured by looking at post-deal synergy realization. Without successful integration, synergies are less likely to be materialized as post-deal integration is the single most important determinant of synergy realization (Barkema & Schijven, 2008). Synergies created means the integration process has worked and, to a certain extent, more synergies created means the integration process has worked better. Previous research of culture on M&A performance takes two to five years accounting performance post-deal as a time frame (e.g. Weber (1996), Datta (1991)). Even though synergies can take up to twelve years to fully realize (Teerikangas & Joseph, 2012), due to data restrictions this research takes the lower limit and follows Stahl & Voigt (2008) in their control point of two years, measuring integration via synergy realization two years after deal completion.3 Although there are ample ways to measure synergies, the return on assets ratio
(Hereafter: ROA) is chosen as it is influenced less than other ratios by potential biases (Meeks & Meeks, 1981) and it is the most used ratio in M&A literature (Thanos & Papadakis, 2012). It is measured by dividing the acquirer’s net profit income by their total assets at a certain time. The ROA of the acquirer is taken two years after the year in which the deal was completed and compared to the ROA of the acquirer two years before the announcement date. The formula is:
Change in ROA = (ROAT+2 – ROAT-2)
Two years has been chosen to give the integration process and synergy creation time to materialize. The net profit income is described as the “income after all operating and non-operating income and expense, reserves, income taxes, minority interest and extraordinary items.” The benefit of this
2 In this sample, missing data on one component of CSR means data is missing for every component. Therefore,
the whole deal can be deleted since no data is available for that year. In a few cases, only the overall ESG score was available, but this information alone is unusable for the purpose of this research.
3 Stahl & voigt (2008) control for time of measurement <2 or ≥2 years after the announcement date. This
19 measure is that synergies usually have two ways of creating value, either via cost reduction or via efficiency gains, and the return on assets allows both types to be expressed. As Stahl and Voigt (2008) mention “…accounting measures of postacquisition performance are often preferred to stock market based measures, because they represent actual economic benefits generated by M&A rather than by market expectations”. An advantage of this method over calculating the cumulative abnormal returns (CAR) like most closely related research does (e.g. Bereskin et al. (2018), Deng et al. (2013), Alexandridis et al. (2015)), is that stock prices and therefor the CAR includes a lot of sentiment and speculation. It measures the expected synergetic gains and how the market values them. The ROA offers better insight in the actual synergy gains and with it, an assessment of whether the integration was successful. Also, said research looks at a time frame of just days after the completion date, vastly ignoring the time synergies need to occur and prosper.
3.3 Independent variable
The independent variable is CSR similarity. It is measured by looking at the scores of each of the 10 components of the acquirer and target companies and comparing them. Since each company receives a score for each component ranging from 0-100, the maximum total score for all components combined is 1000. Similarity is then measured by looking at the square root of the square difference (i.e. the absolute difference) in scores in each component and subtracting the cumulative difference from 1000. By dividing the outcome by ten, deals will be scored on their CSR similarity ranging from 0 to 100. The companies in deals scoring closer to 100 are more similar and companies in deals scoring closer to 0 are more distant in their CSR practices. In formula:
CSR Similarity =
1000 − ∑ √(SA,i−ST,i)2 10
i=1 10
In which SA,1 is the score of the acquirer on component 1 and ST,1 is the score of the target on
component 1.
3.4. Moderator variables
To test the third hypothesis and look at the effects of CSR similarity on M&A success in relation to the firms overall CSR score, the moderator variable Acquirer overall CSR score is added. This term is given
20 by Thomson and it is measured as a weighted average of all the components of CSR combined. As mentioned in chapter 2, acquirer and target companies with higher overall CSR scores often experience a wide range of benefits when engaging in M&As. To further test the determinants of the results the variable Acquirer overall ESG score is added so that the impact of overall CSR on the effect of CSR similarity on CBM&A success can be assessed.
3.5 Control variables
To control for various factors that can potentially influence CBM&A success. multiple control variables are included in the model.
Industry relatedness has multiple influences on CBM&A success. As Rhodes-Kropf & Robinson (2008) find, firms in related industries tend to perform better. Following Alexandridis et al. (2015) industry relatedness is measured as a dummy. When the acquirer and target company have the same first two numbers of their four-digit SIC code it receives a score of 1 and a 0 in all other cases.
Deal size potentially destroys more value as it gets bigger due to the additional complexity caused by large targets (Alexandridis et al., 2013). It is therefore harder to receive the expected economic benefits. Deal size as a control variable is included using the deal value in thousands of euros as measured by Zephyr.
The method of payment may have an influence on M&A performance Hitt et al. (2012) and Alexandridis et al. (2015) find that payments made in cash increase the likelihood of deal completion. Therefore a dummy is included which equals 1 if the deal is predominantly financed by cash(reserves) and 0 otherwise.4
Firm size of the acquirer can influence M&A success in a few ways. It can affect deal premiums through greater bargaining power during negotiations and larger firms usually have more financial resources (Qiao & Wu, 2019). Moeller et al. (2004) find that smaller acquirer firms perform better than larger firms, possibly due to better alignment of interests between managers and shareholders. The acquirer firm size is therefore included and it is measured as the year-end total assets the year prior to the announcement.
The relative size might also influence M&A success. As Hitt et al. (2012) state, not only independent variables of firm size are important, also the relative size of the acquirer and target are of relevance. Alexandridis et al. (2015) for example, find a larger target relative to the acquirer to bring more
21 opportunity for resource sharing and synergy value creation, while Bereskin et al. (2018) find integration issues for relatively larger targets to be more important. The relative size is used as a control variable and measured by dividing the target size by the size of the bidder (e.g. Aktas et al. (2011), Datta et al. (1991)). Following Chatterjee et al. (1992), this research controls for relative size by dividing the target total assets by the acquirer total assets.
National cultural distance is often equally important as corporate cultural distance and therefore a national cultural distance composite is created and added as a control variable. The first four Hofstede (1980) dimensions are used to calculate national cultural distance. They are individualism, power distance, uncertainty avoidance and masculinity. Cultural distance is then calculated using the Kogut and Singh Index, resulting in the following formula:
NationalCD = √∑ (SA,i−ST,i) 2 4 i=1 4
Wherein SA,i is the score for each acquiring company’s home country on dimension i, and ST,i is the score
for each target company’s home country on dimension i.
3.6 Analysis
The first dependent variable in this research is the probability of deal completion. Since this is a dummy variable and therefore dichotomous, a logistic regression model is used to measure the effect of CSR similarity on deal completion. Another method could be the discriminant analysis, but as Hair (2014) states, a logistic regression is the appropriate statistical technique when the dependent variable is a categorical (nominal or nonmetric) variable and the independent variables are metric or nonmetric variables. It does have the advantage over discriminant analyses, however, of easily incorporating nonmetric variables as independent variables. Also, when normality of variables is not met, logistic regression is less affected than discriminant analysis. Therefore, in line with similar research (e.g. Alexandridis et al (2015), Bereskin et al (2018)) the logistic analysis is used.
One of the assumptions for logistic regression is that the sample must be large enough. Hair (2014) states the size of the total sample should be at least 400. Also, every group should have at least 10 observations for each estimated parameter. The sample of this thesis consists of 196 observations in total and 127 and 69 observations for both groups and therefore does not meet the minimum total sample size (and does not meet the group sample size by 1 observation). However, there is much support stating that smaller sample sizes are justified when using logistic regression, as long as a
22 minimum of ten events per variable is met (Peduzzi et al., 1996). Therefore, in line with for example Alexandridis et al. (2015) who also use a sample of just 220 observations, a logistic regression is the appropriate method. The base model looks as follows:
Logit (Deal completion) = β0 + β1 Industry relatedness + β2 Deal size + β3 Method of payment +
β4 Firm size + β5 Relative size + β6 National CD + ε
Wherein β0 is the constant and ε is the error term. In model 1 CSR_similarity is added to test the first
hypothesis. Model 2 includes Overall CSR score to assess the individual effect of overall CSR score on the likelihood of deal completion. Lastly, model 3 then adds the interaction effect CSR similarity and the overall CSR score as to measure the moderating effect of firms overall CSR score on the probability of deal completion. In order to prevent multicollinearity problems, the two independent variables are first standardized using the Z scores before they are multiplied.
To measure the effect on the second dependent variable, deal completion time, a logistic regression model cannot be used since this dependent variable is not dichotomous. A Kaplan Meier analysis is insufficient as well because more than one explanatory variables are included and one of them is quantitative. Instead, similar to Bereskin et al. (2018), a Cox proportional hazard model is used to assess the influence of CSR similarity on the number of days it takes to complete a deal. A Cox model will not only indicate whether there is an effect or not, but it will also indicate the magnitude of that effect. The basic model is depicted below.
Deal rapidness(t) = ho(t) * exp(β1 Industry relatedness + β2 Deal size + β3 Method of payment +
β4 Firm size + β5 Relative size + β6 National CD
In which t is the survival time, h0 is the value of the hazard when all variables are equal to zero and exp
is the hazard ratio. Similar to the logistic regression, the independent variables are added in model 1, 2 and 3.
Lastly, to look at the effect of CSR similarity on the post deal integration process, an OLS regression is conducted. Integration is measured via synergy by looking at the return on assets, which is an interval variable. Therefore, a logistic regression is not suitable to test the second hypothesis. The base model looks very similar to the base model of the logistic regression, with the same variables added in models 1-3.
Integration = β0 + β1 Industry relatedness + β2 Deal size + β3 Method of payment + β4 Firm size
23
4. Results
In this chapter the outcome of the various analyses are discussed, after showing the descriptive statistics and the correlation between the variables. Then, the regressions are performed and the hypotheses are tested. Robustness checks are done to test whether the results are robust to different measures of CSR similarity.
4.1 Descriptive statistics
Table 2 presents the descriptive statistics for all variables. Days to Complete and Change in ROA have lower observations because only completed deals are included in these variables. The mean of CSR similarity is 70,68, meaning that on average companies differ around 29 points on a 0-100 scale in each of the 10 CSR components. Likewise, the deals between the most and least similar companies differed about 6 and 57 points respectively on average on each component. In comparison with National CD, CSR similarity is more concentrated around the mean since its standard deviation is relatively lower.
Table 2. Descriptive statistics
Variable Obs. Min Max Mean Std. Deviation
Deal Status 196 0 1 0,65 0,479
Days to complete 125 0 631 167,80 118,60194
Change in ROA 85 -0,27081 0,08359 -,0042033 0,03693113
CSR similarity 195 42,64600 94,20500 70,6786769 10,50161056
Acquirer overall ESG 195 13,69 90,93 61,9867 16,25627
Related Industry 196 0 1 0,70 0,460 Deal Value 196 22925,0 140550893,7 11616223,63 20391901,48 Method of payment 186 0 1 0,55 0,498 Firm size 196 154608,00 1476355248 84956439,79 225631625,2 Relative size 196 0 2,33 0,4205 0,46221 NationalCD 177 1,39194 25,45093 9,7858315 5,78066205 Euc. CSR similarity 196 0 15,3952 7,655372 3,3425626
Table 2 also shows that the mean of the change in ROA is negative, indicating that a lot of acquiring companies had a decrease in their performance two years after the CBM&A compared to two years before it. This is consistent with previous research (Papadakis & Thanos, 2010). In the sample, 40.0% of acquirers had a negative change in their ROA.
24
4.2 Correlation matrix
The correlation matrix is depicted in table 3 and shows the correlation of all independent variables. Significant correlations are marked with an asterisk. There is some correlation between the explanatory variables, but as most are below or just above 0.30, these are considered weak or moderate (Hair, 2014). One interesting correlation is that of the acquirer overall ESG and CSR similarity, which is negative (r= -0,271) and highly significant (P<0,01), indicating that the higher an acquirer’s overall ESG, the least similar their CBM&A targets are in terms of their CSR practices. In other words, when an acquiring company’s overall ESG score is higher, they are less likely to target another company which has similar CSR practices. Deals in which companies have more CSR similarity have higher deal value, as shown by the positive correlation between CSR similarity and deal value (r=0.164). Relative size also correlates moderately high with CSR similarity (r=0,362), indicating that acquirers and targets with more CSR similarity engage in deals where the target’s size is relatively bigger. As indicated by the negative correlation of relative size and method of payment (r=-0,381), acquirers in deals with relatively bigger targets are less prone to use cash as a means of financing the deal. The high correlation between the Euclidian CSR similarity and CSR similarity (r=0,974) is to be expected as the ways of measuring are very much alike, which will be explained in chapter 4.4. Based on the correlation matrix multicollinearity problems are not suspected and to further verify this, the variance inflation factor (VIF) of each independent variable is calculated. All VIFs are between 1.027 and 1.386. Therefore, multicollinearity is not an issue since they are far below both the critical value of 10 (Kutner et al., 2004) and the more stringent cut off value of 5 (Sheather, 2009).
Table 3. Correlation matrix
1 2 3 4 5 6 7 8 9
1. CSR similarity 1
2. Acquirer overall ESG -,271** 1
3. Related Industry ,057 -,034 1 4. Deal Value ,164* ,120 ,118 1 5. Method of payment -,252** ,009 -0,094 -,174* 1 6. Firm size -,025 ,277** ,058 ,186** ,141 1 7. Relative size ,362** -,128 ,132 ,244** -,381** -,137 1 8. NationalCD -,001 ,152* -,040 -,092 ,129 ,091 -,082 1 9. Euc. CSR similarity ,974** -,229** ,063 ,179* -,232** -,012 ,363** ,021 1 * Correlation is significant at the 0,05 level (2-tailed)
25
4.3 Test of hypotheses
A logistic regression and a cox proportional hazard regression analysis are performed and the results are presented in table 4. The dependent variable of models 1 through 3 is deal completion, which is binary and takes the value of 1 if the deal is completed and 0 if it was withdrawn. The dependent variable in models 4 through 6 is deal rapidness, which includes only completed deals and is measured in days between the announcement date and the completion date. Similar to Bereskin et al. (2018), the models include the coefficients, standard errors and Pseudo R2. First in model 1, without the
variables included, 64,5% is guessed correctly, meaning that 64,5% of the deals are completed and the model always guessed 1. Then, the model includes the independent variables, causing it to predict the outcomes and make a correct guess in 68.1% of the cases (3,6% increase). The explained variation in the dependent variable deal completion based on our model ranges from 12.8% to 17,6%. Appendix A shows the classification tables before and after including the variables.
Table 4. Effect of CSR Similarity on deal completion and deal rapidness
Deal Completion Deal Rapidness
1 2 3 4 5 6 Variables Related industry 0.352 (.381) 0.426 (.385) 0.378 (.393) -0.326 (0.227) -0,401* (0.233) -0.376 (0.234) Deal value 0.000** (.000) 0.000** * (.000) 0.000** (.000) 0.000** (0.000) 0.000** (0.000) 0.000** (0.000) Deal method of payment -0.395 (.401) -0.277 (.407) -0.268 (.409) 0.239 (0.229) 0.196 (0.229) 0.183 (0.231) Firm size 0.000 (.000) 0.000 (.000) 0.000 (.000) 0.000 (0.000) 0.000 (0.000) 0.000 (0.000) Relative size -0.241 (.433) -0.216 (.433) -0.177 (.438) -0.060 (0.269) 0.044 (0.274) 0.018 (0.280) National CD 0.058* (.032) 0.048 (.033) 0.049 (.033) -0.011 (0.019) -0.008 (0.019) -0.010 (0.019) CSR similarity -0.014 (.019) -0.005 (.020) -0.002 (.021) -0.006 (0.011) -0.017 (0.013) -0.019 (0.013) Acquirer overall ESG score 0.015 (.012) 0.017 (.012) -0.013* (0.008) -0.015* (0.008) CSR similarity*Acquirer overall ESG score
-0.119 (.176) 0.080 (0.099) Constant 1.610 (1.372) 0.017 (1.776) -0.338 (1.872) Pseudo R2 0.176 0.191 0.195
*, ** and *** indicates significance at the 10%, 5% and 1% level respectively. Standard errors are given in parentheses.
26 Hypothesis 1 predicts a positive relation between CSR similarity and deal completion chances. As can be seen in table 4 almost no variable adds significantly to the prediction. Deal value is significant, but has an extremely small odds ratio. When deal value is divided by 100,000 to find the effect, Exp(B) is 0,996. This means that an increase of 10 million euros5 decreases the chances of completing the deal
by 1/0,996= 1,004 times. National CD is positive and significant (β = 0.058, P < 0.10), indicating that more national cultural differences between acquirer and target countries increase the likelihood of deal completion. More importantly, contrary to prior research, the independent variable CSR similarity has a negative and non-significant relation with deal completion (e.g. Bereskin et al. 2018; Alexandridis et al. 2015). Model 2 adds the moderator variable Acquirer Overall ESG score. As can be seen in Appendix A table 7, the model guesses 70.9% correct, which is an increase of 6.1% compared to guessing that the deal was completed in all cases. The explained variation is between 13.9% and 19.1%. The acquirer overall ESG score is not significant and the impact of national CD is not significant at the 10% level anymore. Model 3 adds the interaction effect of CSR similarity and overall ESG score. As mentioned, the scores are first standardized before they are multiplied in order to prevent multicollinearity problems. The model guesses 71,5% correctly and has an explained variation of 14,2% to 19,5%. The independent variable, the moderator variable and the interaction term are all insignificant. Therefore, hypothesis 1 and hypothesis 4 (for the part about deal completion) are rejected.
Similarly, table 4 shows the results of the cox regression in models 4 through 6. CSR similarity and the interaction term with the moderator do not have significant effects on deal rapidness. However, the moderator variable does have a significant direct effect. As can be seen in model 2, acquirer overall ESG score has a significant negative effect (β = -0.013, P < 0.10), indicating that the higher the acquirer overall ESG score is, the smaller the hazard ratio gets, also meaning longer time to complete the deal (Klein et al., 2014). Adding the interaction term in model 6 makes this effect slightly stronger, but does not make any significant changes to the prediction. Therefore, CSR similarity has no significant effect on deal rapidness and hypotheses 2 and 4 (for the part about deal rapidness) cannot be accepted.
To test hypothesis 3 and analyze if CSR similarity increases post-deal integration, an OLS regression is conducted to assess synergy creation. Table 5 shows the results of this regression. Models 1 through 3 show that CSR similarity, acquirer overall ESG score and the interaction term have no significant effect on the change in ROA. M&As in related industries (β = 0.015, P < 0.05) and with a higher deal
27 value (β = 9.045E-10, P < 0.01) experience more return on assets two year after the deal is completed, when all other variables stay the same. In line with literature, bigger acquirers gain less return on assets (β = -9.414E-11, P < 0.01). In order to check if the different levels of acquirer overall ESG are not suppressing each other and leveling out the total effect of CSR similarity on synergy creation, three groups of acquirer overall ESG score are created based on the quartiles, similar to Bereskin et al. (2018). The low group contains the acquirers with the lowest 25% of overall ESG, the moderate contains 25-75% and the high overall ESG group contains acquirers with >25-75% of overall ESG. The results can be seen in models 4 through 6 in table 5.
Table 5. Effect of CSR Similarity on post-deal integration
Change in ROA Change in ROA
1 2 3 4 Low ESG 5 Mod. ESG 6 High ESG Variables Related industry 0.015** (0.007) 0.015** (0.007) 0.014* (0.007) 0.027* (0.012) -0.001* (0.007) 0.027 (0.020)
Deal value
9.054E-10*** (0.000) 8.626E-10*** (0.000) 9.489E-10*** (0.000) 1.168E-9** (0.000) 6.466E-10** (0.000) -1.243E-9 (0.000) Deal method of payment 0.008 (0.008) 0.007 (0.008) 0.008 (0.008) 0.037** (0.013) -0.003 (0.007) -0.008 (0.023)
Firm size
-9.414E-11*** (0.000) -9.030E-11*** (0.000) -9.119E-11*** (0.000) -2.288E-11 (0.000) 4.629E-11 (0.000) -8.374E-11*** (0.000) Relative size -0.007 (0.009) -0.007 (0.009) -0.006 (0.009) -0.007 (0.010) -0.011 (0.012) 0.008 (0.031) National CD -0.001 (0.001) -0.001 (0.001) -0.001 (0.001) -0.002** (0.001) 0.001 (0.001) -0.004 (0.002) CSR similarity 0.000 (0.000) 0.000 (0.000) 0.000 (0.000) 0.001 (0.001) -0.000 (0.000) 0.000 (0.001) Acquirer overall ESG score 0.000 (0.000) 0.000 (0.000) CSR similarity*Acquirer overall ESG score
-0.002 (0.003) Constant 0.007 (0.024) 0.033 (0.034) 0.030 (0.035) -0.079 (0.056) 0.010 (0.028) -0.006 (0.059) Adjusted R2 0.433 0.434 0.430 0.649 0.084 0.435
*, ** and *** indicates significance at the 10%, 5% and 1% level respectively. Standard errors are given in parentheses.
As shown by model 4, acquirers with low overall ESG scores will gain less return on assets as national cultural distance increases (β = -0.002, P < 0.05). Acquiring a target in a related industry seems to have a negative effect for acquirers with moderate ESG scores (β = -0.001, P < 0.10) and has no significant
28 effect for acquirers with high ESG scores. Furthermore, model 4, 5 and 6 show no significant effect of CSR similarity on the dependent variable. This means that acquirers with low, moderate and high ESG scores experience no significant influence of CSR similarity on their return on assets. Therefore, CSR similarity has no significant effect on cross-border post-deal integration and the overall ESG score does not moderate this effect. Hence, hypothesis 3 and 4 must be rejected.
In order to gain more insight in the effects of the individual components of CSR similarity, each of the ten components is assessed separately. This way, potential opposing influences of the components that could result in a leveled effect of CSR similarity as a whole can be identified. Similarity of a component is measured by taking the absolute difference of the acquirer’s and target’s score on that component and then subtracting it from 100 to transpose the data, meaning a higher score means more similarity. The same analyses are used as were used to study the aggregate of CSR similarity. The results can be found in appendix B. Model 3 of table 9 shows that the interaction term is significant (β = -0.404, P < 0.10). The effect of environmental innovation on CBM&A deal completion is therefore moderated by the acquirer’s overall ESG score. Since this is negative, it means the higher the ESG score of the acquirer is, the weaker the effect becomes. In table 10, model 2 shows the effect of emission score similarity on deal rapidness. Higher similarity in emission scores decrease deal rapidness, as can be seen by the negative coefficient in the cox hazard model (β = -0.009, P < 0.05). The acquirer overall ESG score also has a negative effect (β = -0.013, P < 0.10). Model 4 shows that the acquirer overall ESG score positively moderates the relation between workforce similarity and deal rapidness (β = 0.325, P < 0.01), making the effect of workforce similarity on deal rapidness stronger when the acquirer’s overall ESG score gets higher. Model 7 in table 11 shows that similarity in product responsibility has a positive effect on the return on assets (β = 0.0003, P < 0.05). Model 3 shows that the interaction term is negative and significant (β = -0.009, P < 0.05), indicating a moderating effect of acquirer ESG score on the effect of environmental innovation on post-deal integration.
4.4 Robustness checks
To check the robustness of the results additional tests are done. First of all, the models are re-estimated by measuring the independent variable CSR similarity in a different way. Following Alexandridis et al. (2015), the Euclidian distance will be used, which differs slightly from the original measure. The Euclidian distance is calculated as follows:
CSR similarity.Euc =
√∑10 (SA,i−ST,i)2
i=1
29 The value of the highest outcome subtracted by the score is taken to transpose the data as to a higher score meaning more CSR similarity. The same statistical analyses are used. The results can be found in appendix C. Table 12 shows similar results for the effect of CSR similarity on deal completion and deal rapidness. The effect of the overall ESG score of the acquirer remains the same (β = -0.013, P < 0.10). The models show no significant changes to the results in the main analysis. Likewise, table 13 shows no relevant changes for the effect of CSR similarity on post-deal integration.
Second, in order to check the robustness of the effect of CSR similarity on integration via synergy creation, a different measure of this dependent variable is used. The difference in net profit income of the acquirers and targets two years before and two years after the deal are calculated. Table 14 in appendix D shows the results, which are very similar to the main analysis. No significant effect of CSR similarity, the moderator and the interaction term can be found. Therefore, the results of the main analysis are robust to different measures of post-deal integration.
30
5. Discussion and conclusion
In this master thesis, we have examined the effect of CSR similarity on CBM&A success. Based on the theoretical perspective of (corporate) culture, four hypotheses were developed. Hypothesis 1 argues that CSR similarity will increase cross-border deal completion chances. Hypothesis 2 expects that CSR similarity will increase cross-border deal rapidness and hypothesis 3 expects that CSR similarity will increase cross-border post deal integration. Hypothesis 4 expects that the overall ESG score will have a moderating effect on the effect of CSR similarity on CBM&A success, making the effect stronger when the overall ESG score is higher. Logistic regressions, cox regressions and OLS regressions are performed to test these hypotheses. The results of the main analysis show that none of the hypotheses can be supported. CSR similarity has no significant effect on the likelihood of deal completion, deal rapidness and post-deal integration of CBM&As. There is also no significant moderation effect of different levels of overall ESG score on the effect of CSR similarity on CBM&A success. This means that there is no statistically significant difference in the effect of CSR similarity in deals between high, moderate and low overall CSR firms. These results are robust for different measures of CSR similarity and post-deal integration.
Looking closer at CSR similarity however, we have tested the impact of similarity of individual CSR components on CBM&A performance. This thesis finds similarity on emissions to negatively impact deal rapidness, and similarity on product responsibility to positively impact their ROA. The product responsibility score is defined by Thomson as “a company’s capacity to produce quality goods and services integrating the customer’s health and safety, integrity and data privacy.” This indicates that having the same stance on what the quality of the goods and services must look like, be that both high or low, results in better post-deal integration. Perhaps this is the result of not losing time over disagreements and modifications to adjust to the new quality standards. Furthermore, the acquirer overall ESG score negatively moderates the relation between environmental innovation and deal completion and positively moderates the relation between workforce and deal rapidness, but no direct effect has been found. This could mean there is a cross-over interaction, meaning the effect of similarity of environmental innovation score on deal completion chances can be both positive and negative, depending on how high the acquirer overall ESG score is (Loftus, 1978). Finding significant results on the component level is quite interesting as it seems to indicate that CSR similarity as a whole does not significantly predict CBM&A success, but some individual components do. Perhaps the aggregate of CSR similarity is too broad a subject to adequately predict CBM&A success and the individual components are better suited to explain the impact of CSR similarity on CBM&A success. As mentioned before, literature about CSR is rich and contentious. Research found both positive and negative effects of CSR on firms’ financial performance, M&As and stock returns (e.g. Carroll &
