Master Thesis
Acquisitions or Joint Ventures? The influence of deal type on post-deal firm performance in the motor
manufacturing industry and the moderating effects of strategic similarity and cross-national distance
George Taylor – S3596184/B7061184
MSc Advanced International Business Management Dual Award – Newcastle University & University of Groningen
Supervisors – Jiyoung Shin & Stefanie Reissner
Abstract
Exogenous growth is a well-documented way for firms to facilitate access to new markets, knowledge, and resources. More recently literature has started to discuss the potential benefits that joint ventures (JV) can bring when compared to the traditional methods of expansion. This debate is particularly strong in the automotive industry where a rapid rate of acquisitions has created a hyper-competitive and saturated market. However, little is actually known about under which circumstances each of the available deal types should be used. The objective of this master thesis is to contribute to the debate by exploring the effects of choosing between a JV and an acquisition on the performance of the deal. It is a proposition of the research that the choice of JV as a deal type will lead to increased post deal performance in the automotive industry. Furthermore, the moderating effects of cross-national distance and strategic similarity are assessed in this context. Analysis conducted through both OLS and robust linear regression shows no evidence that choice of deal type has any effect on performance outcomes, and contrary to expectations cross-national distance has no effect on the relationship. The results for strategic similarity however are more significant, with support for a direct positive relationship between strategic similarity and performance, and partial support for an interaction between strategic similarity and deal type. The findings of the study challenge some existing theories and methods that have been taken for granted and hence provide important theoretical and practical implications on the debate of what constitutes a successful deal.
Keywords: Joint ventures, acquisitions, consolidation, cooperation, motor manufacturing, cross-national distance, strategic similarity, deal performance
Acknowledgements
Firstly, I would like to express my appreciation for my supervisors Stefanie Reissner and Jiyoung Shin. Their invaluable constructive feedback, support, and guidance enabled a challenging but rewarding experience in writing my thesis. Further, I would like to thank Prof.
Sjoerd Beugelsdijk for his help in making data on cross-national distance available to students for their master theses. Lastly, I would like to thank all my friends and colleagues who I met as part of the AIBM course, both in Newcastle University, and the University of Groningen, for making the last 18 months an amazing experience.
Table of Contents
Abstract ... 2
Acknowledgements ... 3
List of Figures ... 6
List of Tables ... 6
List of Abbreviations ... 6
Chapter 1: Introduction ... 7
1.1 Research Context ... 7
1.2 Theoretical Foundations ... 8
1.3 Summary of Methods ... 10
Chapter 2: Literature Review ... 11
2.1 Deal Type ... 11
2.2 Deal Performance ... 12
2.3 Joint Ventures vs Acquisitions ... 13
2.4 Cross-National Distance ... 15
2.4.1 Current research on Cross-National Distance ... 15
2.4.2 Cross National Distance as a Moderator ... 16
2.5 Strategic Similarity ... 17
2.5.1 Current Research on Strategy ... 17
2.5.2 Strategic Similarity as a Moderator ... 19
2.6 Conceptual Model ... 20
2.7 Literature Review Conclusion ... 21
Chapter 3: Research Methodology ... 22
3.1 Research Philosophy ... 22
3.2 Data ... 23
3.3 Measures ... 24
3.3.1 Dependent Variable: Post-Deal Firm Performance ... 24
3.3.2 Independent Variable: Deal Type ... 25
3.3.3 Moderating Variables: Cross-National Distance ... 25
3.3.4 Moderating Variables: Firm Strategy ... 26
3.3.5 Control Variables ... 27
Chapter 4: Results ... 29
4.1 Preliminary Tests ... 29
4.2 Descriptive Statistics ... 30
4.3 Correlations ... 32
4.4 Regression Results ... 34
4.5 Robust Regression ... 37
Chapter 5: Discussion ... 38
Chapter 6: Conclusion ... 42
6.1 Theoretical Implications ... 42
6.2 Practical Implications ... 43
6.3 Limitations and Future Research ... 44
References ... 48
Appendices ... 58
Appendix A: Preliminary Tests ... 58
Appendix B: Individual Moderator Regression Analysis ... 59
Appendix C: Robust Regression Analysis ... 60
Appendix D: Deals ... 61
List of Figures
Figure 1: Conceptual Model……….20
List of Tables
Table 1: Descriptive Statistics…….……….30 Table 2: Correlation Matrix……….….33 Table 3: Regression Analysis……….……….………….36
List of Abbreviations IB – International Business JV – Joint Venture
M&A – Mergers and Acquisitions OLS – Order of Least Squares PRC – Peoples Republic of China RBV – Resource Based View ROA – Return on Assets
SGA – Selling, General and Administrative
Chapter 1: Introduction
“Coming together is a beginning, staying together is progress, and working together is success”- Henry Ford
1.1 Research Context
It is a well-documented phenomenon that as industries grow and mature, they often undergo a process of consolidation and cooperation in which the number of firms reduces, and those firms that remain increasingly integrate processes to achieve competitive advantage. The automotive industry is no exception to this. Whilst case level analysis of unsuccessful deals such as DaimlerChrysler, and successful deals such as TATA/Ford, have been well documented in business literature, little attention has been paid to the choice of deal type that firms elect when expanding operations, and to what extent this choice determines the success factor of increasing firm performance. In this research this issue is directly analysed, specifically focusing on identifying which factors ensure a successful deal between motor manufacturers.
The automotive industry has often been referred to as the ‘industry of industries’ (Drucker, 1946). There are few other industries that carry the same amount of deep-rooted national pride in their successes. Motor manufacturers have often been the first to adopt and deploy revolutionary business and production strategies, for example the conception of Fordism by Henry Ford enabled the first mass production assembly line of standardised goods. Later the motor industry gave rise to the first multi-divisional firm in General Motors, in order to combat the multi-functional empire of Ford (Chandler, 1962). Another form of production that was pioneered by the industry is that of lean production, which involves maximising efficiencies across all areas of the business; from a constant process of checking for defects on the production line, minimising space, and minimising the amount of materials/parts required for production; to lean IT systems, and lean relationships with suppliers (Womack, Jones & Roos, 2007). As well as this motor manufacturers were some of the first to adopt ‘just-in-time’
delivery (JIT). JIT delivery is a system of production organisation that minimises buffer stocks (Sadler, 1994) by ensuring materials and parts are delivered only at the exact time required.
Thus, we can see how firms in the motor industry are under constant pressure to stay ahead of the competition.
Consolidation in the automotive industry has been prevalent for almost as long as the industry has existed. In America General Motors was formed by the bundling together of 200 small firms, in Europe almost the entire British car industry was rolled into British Leyland, and Citroen was swallowed by Peugeot (The Economist, 2005). More recently the failed merger of Daimler and Chrysler and the blossoming Nissan Renault alliance are two of the most notable examples, however what pressures do motor manufacturers face that pushes them to pool resources? The manufacture of automobiles has become an increasingly risky venture (Orsato
& Wells, 2007). Orsato & Wells identify a central characteristic of the automotive industry, that there are slim margins between the profit and loss faced by most manufacturers due to the sizeable economies of scale that firms have imposed upon themselves. A combination of slim margins and high break-even points has meant that auto-makers are susceptible to market fluctuations (Golding, 1999). The answer to this volatility is therefore size in order to achieve the economies of scale required to compete. Feast (2000) predicted that by 2020 only six vehicle manufacturers would remain in the industry, and although this prediction looks unlikely to come true, it gives a clear indication to the rapid pace of consolidation.
Recently the focus of manufacturers has expanded to areas where cooperation through shared platforms can reduce excess capacity and concentrate supply. With each manufacturer building its own engines, transmissions and chassis, the amount of duplication is significant. Factors such as this rarely play an important role in the end customers purchase decision with their focus primarily falling upon style, quality and reliability (Pwc, 2017). The mechanism through which this cooperation can be achieved is therefore a JV. For example, the Global Engine Alliance between Chrysler, Mitsubishi, and Hyundai was a JV that involved the development of shared engines. This type of cooperation can result in savings in R&D and supply chains.
These trends therefore feed into this research in analysing the choice of deal available to motor manufacturers and the performance outcomes associated.
1.2 Theoretical Foundations
When firms engage in expanding operations, they can do this by either internal, or external growth strategies. Internal strategies involve organic practices such as the launch of new products, hiring more staff, and increased marketing investments (Klette & Griliches, 2000).
On the other hand, external strategies imply the pooling of resources with another firm (Lechner & Dowling, 2003). This thesis focuses on external strategies of growth as a
determinant of firm performance. When pursuing external growth there are various governance structures available to firms, of which here two are analysed. Firstly, acquisitions in which one company buys a majority stake in another company in order to assume control. Secondly, a JV, in which two companies set up a separate legally distinct business owned by two or more parent firms (Beamish & Lupton, 2009). Beamish & Lupton (2009) take the view that JVs enable firms to access each other’s complementary resources and capabilities in order to achieve economies of scope and/or scale, and to develop new products faster, more reliably, and more cheaply than could be done by either firm acting alone or through acquisition. The growing traction that JVs are gaining in the literature influences this thesis’ focus on the type of deal and its direct effect on performance outcomes.
The international business (IB) literature around external expansion has paid much attention to the idea of cross-national ‘distance’. One framework that is prominent in the literature is Ghemawat’s (2001) CAGE framework (Boschma, 2005; Aaker & McLoughlin, 2010) that assesses differences between countries on the grounds of cultural, administrative (or institutional), geographic, and economic distance. The consensus here is that the greater the distance between home and host countries, the more difficult it is to operate effectively in the host environment (Ahern, Daminelli, & Fracassi, 2015; Peng & Pleggenkuhle-Miles, 2009).
However, little research has been conducted into whether the type of deal is affected by cross national distance and thus whether it moderates the relationship with performance. By including this variable, it is hoped to bring indication as to whether certain deal types, such as JVs, are more successful in limiting the negative effects of distance.
Another important aspect of external growth is the strategy of each firm prior to the deal. It has been previously argued that firms which exhibit similar strategic characteristics prior to the deal will experience increased performance outcomes (Ramaswamy, 19997). For example, taking Porter’s (1980) generic strategies, it is unlikely that a ‘cost leader’ could benefit from merging with a ‘focused differentiator’ as their strategies would be out of line, whereas two
‘cost leaders’ pooling resources would only strengthen their position in this segment. The same argument can also stand when comparing strategies on the basis of the resource-based view (RBV) (Wernerfelt, 1984). Despite this, evidence from previous studies is conflicting and contradictory, therefore this research aims to further understanding in this area by analysing the interaction effect of strategic similarity alongside the type of deal and their contribution to
Combining all the factors discussed above, the research questions are formulated: Does choice of deal type influence post-deal performance? If so, is this relationship moderated by strategic similarity and cross-national distance?
1.3 Summary of Methods
Based on the theoretical underpinnings and research context surrounding the motor manufacturing industry, this thesis attempts to analyse the effect of JV as a deal type on deal performance when compared to acquisitions. Data was collected from primarily 4 sources:
Zephyr; Orbis; Datastream: and Reuters ThomsonOne. This data was then categorised into 11 variables measuring deal type, deal performance, and the moderating variables of cross- national distance and strategic similarity. The data was then analysed using OLS as well as robust multiple linear regression analysis. The findings of this research have led to conclusions that can be applied by managers of automobile manufacturers to future decision making. By identifying the importance of strategic similarity in the deal process, managers will be able to take clear and informed decisions on whether to proceed with such actions. If the decision to proceed is taken, then this research also provides new evidence on the importance of R&D differences between firms when choosing which deal-type to pursue.
This thesis will follow a traditional structure of academic work. Firstly, a comprehensive literature review will be presented providing critical insight to the subject and building from this the hypotheses will be developed. Secondly, the hypotheses will be presented in a conceptual model to be tested. Thirdly, the research methodology is detailed along with clear justifications for the statistical tests used and an understanding of the limitations of the analysis.
Fourthly, the results of the main regression analysis are disclosed and subsequently discussed resulting in an advancement in understanding of the key issues. Finally, results are summarised in a conclusion that brings both theoretical and practical implications to light, as well as further discussion of the study’s limitations and potential avenues for further research.
Chapter 2: Literature Review
In undertaking a literature review of the debate on the effects of deal type, cross-national distance, and strategic similarity on deal performance, this thesis first attempts to highlight the possible deal types available to firms pursuing exogeneous growth and present the trade-off between acquisitions and JVs. Then the theory on how deal performance can be measured is compared and contrasted. Hypotheses are then developed from existing works to explain the relationship between deal type and performance as well as the moderating effects of cross- national distance and strategic similarity. Finally, the derived hypotheses are presented in a conceptual model.
2.1 Deal Type
When choosing exogenous growth there are a plethora of deal structures available to firms.
From acquisitions to alliances, wholly owned subsidiaries, mergers, or strategic partnerships the amount of governance structures is vast and sometimes confusing. Before diving into how a deal can be considered successful, it is first important to understand the differing definitions of deals available and the exposure they have gained in literature.
By far the most studied forms of deal activity are mergers and acquisitions (M&A). Shleifer &
Vishny (2001) break down the decades of M&A activity into distinct waves, with a
‘conglomerate’ wave in the 60s, moving to a ‘hostile’ wave in the 80s, and a stock price driven wave in the 1990s. Despite this, many papers fail to provide a clear definition of the deal type they analyse. One of the most cited works on mergers defines a merger as a majority purchase of shares that is approved by shareholder majority vote (Manne, 1965). On the other hand, Iannotta (2010) defines acquisitions as a bidder firm purchasing all or part of the targets stock or assets. There would appear to be very little difference between the two definitions, therefore this paper clarifies any confusion by taking an explicit definition of acquisitions. An acquisition is defined as a situation in which one company buys a majority stake in another company thus assuming control. There are two reasons why this thesis does not focus on mergers in the motor manufacturing industry. Primarily, as described the definition of what constitutes a merger is sometimes unclear in the literature and studies often include M&A together as one. Secondly, that there have been a number of high-profile failed mergers in the motor manufacturing industry e.g. DaimlerChrysler, Ford & Volvo (Badrtalei & Bates, 2007; Salama et al., 2003).
This has led to a change in appetite away from traditional merger activities and a look to more nuanced deal methods such as JVs (Ghosn, 2011).
A JV can be defined as a situation in which two companies set up a separate legally distinct business owned by two or more parent firms (Beamish & Lupton, 2009). This common legal organisation results in the pooling of the parent firms resources (Kogut, 1988). The number of JVs has been increasing, and thus the amount of attention that this deal type is getting in IB literature is also increasing (Lane, Salk & Lyles, 2001; Yan & Luo, 2016). Yan & Luo 2016 state that JVs offer an important alternative to traditional methods such as acquisitions by allowing firm cooperation without the loss of corporate identity. It is due to this apparent decision that emphasises this research’s decision to compare acquisitions with JVs.
2.2 Deal Performance
In measuring the performance of a deal there are conflicting points of view within the literature.
The elusive equation that has plagued boardrooms and management for decades is how to make 1+1=3 (Marks & Mirvis, 2010). Whilst this equation is simple and highlights the ultimate aim of any deal should be value creation, this means different things for different stakeholders.
Value creation for shareholders involves rising share prices and dividend payments. Value creation for customers on the other hand means better quality products at lower prices. Even management and employees may see value creation as lavish offices and superior benefits.
Lubatkin (1988) sees shareholder value as the key component in analysing success as a high valuation enables a firm to satisfy the financial claims of employees, customers, suppliers and stockholders, whilst also mitigating battles for corporate control and allowing access to cheap equity capital. Fridolfsson & Stennek (2005) argue that where the stock market understands the dynamics of a deal, the change in firms’ stock market values reflects the change in their true values. The authors argue that a puzzle of some deals reducing profit whilst raising share prices can be explained when being an insider (consolidated firm) is better than being an outsider (non-consolidated firm), thus pre-empting their partner merging with a rival. On the other hand, stock market valuations are not always based on fundamentals and are often subject to speculation. Shelton (1988) found that in the three days surrounding a deal announcement selling firms experienced stock gains of 11% on average. This type of volatility questions whether share prices can be seen as a reliable measure of deal performance. Some scholars therefore choose to focus deal performance around accounting measures of profitability as
indicators of post-merger efficiency, such as Meeks & Meeks (1981) who look at the: profit- sales ratio; rate of return on equity; and rate of return on total net assets as potential measures for post-merger efficiency gains to determine success. These types of measures allow for a greater number of deals to be analysed as firms need not be exchange-listed, whilst also avoiding the speculation that is associated with share prices. Such measures are regularly used in more recent studies measuring deal performance (Zollo & Singh, 2004; Porrini, 2004). There is however, an argument in the literature that financial measures are sometimes not applicable given the objectives of a JV (Beamish & Lupton, 2009). Anderson (1990) argued that subjective measures of JV performance are preferred. If the goal of a JV is not financial e.g.
access to local market information or technology, then the argument is made that the best judge of performance is likely to be made by those who closely manage it. Despite this, previous evidence shows that subjective and objective measures tend to be somewhat related (Geringer
& Hebert, 1991; Glaister & Buckley, 1998). Beamish & Lupton (2009) conclude that purposive sampling may be required depending on the performance measure of interest in a particular study. In the motor manufacturing industry JVs are often manufacturing and sales partnerships between a global manufacturer and a local partner e.g. the JV between General Motors and AvtoVAZ in Russia. For such relationships where a legally separate company is engaged in a full production and sales process, the reporting and analysis of financial statements is undoubtedly a key indicator of performance.
2.3 Joint Ventures vs Acquisitions
In this research the type of deal analysed is split into two discrete categories of acquisitions and JVs. Acquisitions involve a company buying a majority stake in another company thus assuming control, whereas JVs involve two parent companies setting up a separate legally distinct business in order to cooperate and pool resources. The distinction that is drawn here is that acquisitions represent a conquest, whereas JVs represent cooperation. Wang and Zajac (2007) state that JVs and acquisitions are seen to be opposing choices for firms interested in combining their resources. During the 1990s M&A became well known instruments through which firms could increase their market power, enter new markets, and/or enhance their capabilities (Hagedoorn & Duysters, 2002). Recently, the number of JVs has grown. A search on the database Zephyr for JVs in the industry showed a 50 % increase from the decade between 1997 and 2007, to the following decade 2007 to 2017. Cooperation has grown around a number of business functions such as R&D (Agostini & Caviggioli, 2015), and marketing (Andersson
& Nyberg, 1998). JVs allow access to local knowledge concerning markets and business practices while allowing retention of operational and strategic control in the parent companies (Beamish & Lupton, 2009). This allows parent firms to reduce the risk of their investments by partnering with firms that have more local knowledge and international experience (Lu &
Beamish, 2001). Woodcock, Beamish & Makino (1994) found that Japanese-American JVs established in North America outperformed subsidiaries established by acquisition.
There are various factors through which JVs achieve superior performance. The transfer of knowledge is a key aspect of JVs that distinguishes it from other methods of expansion. JVs are often used to access new knowledge, or to profit from existing knowledge (Inkpen &
Crossan, 1995; Shenkar & Li, 1999). Knowledge tends to flow more freely, and capabilities are developed more easily in JVs than in wholly owned subsidiaries (Luo, 2002). Unlike acquisitions, JVs includes capital investment from both parent organisations. This signals commitment on both sides of the venture, instead of the one-sided nature of an acquisition. Partner commitment, which is especially important when firms are competitors, enhances the probability of success (Beamish & Lupton, 2009).
Hennart & Reddy (1997) break down the reasons why JVs might outperform acquisitions into 4 categories. Firstly, management costs. The integration of management styles, routines, and employees can be a costly business. The JV safeguards the incentives that both firms employees have to maximise profits and avoids the management costs associated with acquisitions. Secondly, there may be difficulties in assessing the actual value a target firm.
When engaging in an acquisition a price must be paid for the target firm. When the value of this firms assets is not known this may lead to the bidding firm overpaying. The JV allows access to complimentary assets at a relatively low cost compared to acquisitions. Thirdly, related to the second point, is that of indivisibilities. This occurs when the desired assets are hard to distinguish from the undesired ones. For example, a motor manufacturer may wish to engage in a deal to start producing light commercial vehicles. It could be possible to buy a manufacturer that already produces light commercial vehicles. The target firm may be predominantly a light commercial vehicle manufacturer, but also combine these operations with the manufacture of minibuses. In this case the indivisibility of the desired assets would pose a problem in an acquisition. Instead a JV with the target firm may be preferred in which the firm gets access to only the resources it needs. Finally, Hennart & Reddy (1997) identified
occurs when some countries ban acquisitions by foreign companies e.g. China/Russia or make it difficult for them e.g. Japanese cross-shareholdings.
One reason why the emphasis on cooperation in the automotive industry is growing could be the importance of the brand. Very few industries carry the same amount of national pride as the automotive industry (Heneric, Licht & Sofka, 2005). Thus, by retaining autonomy through JVs, firms can retain their identity whilst also avoiding the pitfalls associated with acquisitions such as cross-national distance, alongside limiting cooperation to only areas where true synergies can be achieved. Upon the announcement of the Nissan-Renault alliance, the joint CEO of the two companies Carlos Ghosn highlighted the mood of favourability towards cooperation, stating that a merger would backfire due to cultural differences and a desire to remain independent (Ghosn, 2011). Despite this, relatively little is known about the circumstances in which a JV should be pursued over an acquisition (Wang & Zajac, 2007;
Frehse, 2012). In order to better understand the effect of deal type on performance outcomes the following hypothesis is developed:
H1: Joint ventures in the motor-manufacturing industry lead to greater deal performance than acquisitions.
2.4 Cross-National Distance
2.4.1 Current research on Cross-National Distance
Despite the trend of consolidation and cooperation not all deals are successful. Opportunistic takeovers are a risky business (Rhys, 1999). Bigger corporations do not necessarily stand a better chance of surviving in competitive environments (Wells & Nieuwenhuis, 2000). There is a long list of potential pitfalls a deal could fall into. The DaimlerChrysler case of a cultural mismatch has been well documented in business literature, with the apparent ‘merger of equals’
ultimately leading to a ‘clash of culture’ (Hollman et al., 2010). However, culture is not the only form of cross-national distance faced by firms operating internationally. The IB literature devotes a substantial amount of interest to cross-national distance, with a prominent model on the subject being Johanson & Vahlne’s (1977) Uppsala internationalisation model. The model suggests that as host countries are seen as more distant from an investor's home country, internationalisation through FDI is likely to be preceded by market entry modes with less commitment e.g. exporting/sales subsidiaries. Furthermore, where FDI does occur it is likely
to occur in countries more ‘similar’ to that of the home country. The overarching theme is such that the greater the distance between home and host country, the harder it becomes for an MNE to operate in that country. One prominent framework that is often referenced when discussing cross-national distance is CAGE (Ghemawat, 2001). The framework sets out cultural, administrative (or institutional), geographical, and economic distance as the four most important aspects of cross-national distance.
Drawing on Hofstede’s (1980) definition of culture, cultural distance can be seen as the differences between the collective programming of the mind distinguishing the members of one group or category of people from others. In this study the group or category is measured at the country level. Ahern et al. (2015) find that the volume of cross-border mergers decreases as cultural distance increases and show that greater cultural distance can lead to lower combined announcement returns. Administrative distance is defined as the differences in legal system, political system, and financial institutions between nations (Ghemawat, 2001). Peng
& Pleggenkuhle-Miles (2009) state that differences in host countries compared to MNE’s home countries make it more difficult to operate in these foreign environments, and a key difference lies in the quality of nations institutions. Geographic distance in general deals with the idea that the farther away you are from a country the harder it will be to conduct business in that country, due to factors such as transportation costs (Ghemawat, 2001). Economic distance deals with the level of GDP per capita, and research has shown that rich countries engage with relatively more cross-border economic activity than poorer countries, and this activity is mostly with other rich countries (Ghemawat, 2001). It is therefore logical to include the effect of cross-national distance in the model by including Ghemawat’s CAGE framework.
2.4.2 Cross National Distance as a Moderator
In socio-culturally distant markets firms choose JVs as they provide better management of operations by local partners who presumably have a greater understanding of the local market (Argawal, 1994). Scholars such as Barkema, Bell, and Pennings (1996), found that cultural differences could affect the long-term survival of JVs, with a positive relationship being found between firms’ prior experience in a country and the longevity of a JV. Park and Ungson (1997) found that U.S.-Japanese JVs actually survived longer than domestic U.S. JVs. This was argued to be down to the compatibility of Japanese culture in sustaining alliances along the grounds of trust, learning, and long-term horizons. Using Hofstede’s cultural dimension of
individualism, Li, Lam & Qian (2001) showed that greater differences in individualism may actually improve JV profitability and productivity as more conflict is likely between individualistic partners. Therefore, we can see how JV may be the preferred choice when cultural distance is present. When administrative distance is present, local partners in a JV better understand the local polity and legal system. Thus, they are able to better manage relationships with governments (Root, 1987). Argawal (1994) highlights the importance of economic distance in a JV, stating that the importance of a local partner is great as it facilitates better supervision of employee relations. A firm coming from a developed country may not have the knowledge of a less developed labour market, and a local partner can help better understand and resolve problems. Finally, a JV can better alleviate the problems of geographic distance by utilising a partners knowledge of how to manage relationships with suppliers and buyers (Root, 1987). This brings substantiation to the debate that when cross-national distance is present JVs may actually deal with the problems that this presents better than acquisitions, and thus positively moderate the relationship between deal type and deal performance:
H2A: Cultural distance between firms’ home countries positively moderates the positive relationship between choice of JV as a deal-type and post-deal firm performance.
H2B: Administrative distance between firms’ home countries positively moderates the positive relationship between choice of JV as a deal-type and post-deal firm performance.
H2C: Geographic distance between firms’ home countries positively moderates the positive relationship between choice of JV as a deal-type and post-deal firm performance.
H2D: Economic distance between firms’ home countries positively moderates the positive relationship between choice of JV as a deal-type and post-deal firm performance.
2.5 Strategic Similarity
2.5.1 Current Research on Strategy
Strategy has been defined in many ways in business literature. In its simplest form strategy can be ‘a coordinated set of actions’ (Ahlstrom & Bruton, 2009), however this definition can also be expanded to encompass a long-time frame and the business environment (MacCrimmon, 1993). Rumelt (1991) found that effects specific to individual businesses explain the largest
portion of the variance of business-level profitability. It is therefore apparent how business strategy is of upmost importance to profit-maximising firms.
Many attempts have been made to classify strategy, however because of strategy’s broad definition there is a diverse range of classifications in the literature. The most seminal work on strategy comes from Michael E. Porter (1980). Porter attempted to simplify firms’ strategic actions into 4 ‘generic’ strategies. Firms were distinguished on 2 levels, being source of competitive advantage (cost or differentiation) and scope of market (narrow or broad), then allocated to the category of either cost focus, cost leadership, differentiation focus, or differentiation. Despite the vast number of citations Porter’s work continues to garner amongst academic work, some argue that the idea of mutually exclusive ‘generic’ strategies is a flawed assumption in the modern business environment (Hill, 1988; Miller, 1992). This has given rise to the view that the application of mixed strategies is a possible avenue for firms to pursue.
Porter’s argument is that this would lead to firms being “stuck in the middle”, failing to achieve the benefits of either low cost production or differentiation, and leaving other firms in a better position to compete in every segment. On the other hand, Hill (1988) argues that in a mature oligopolistic industry it is possible to pursue both a low-cost position along with differentiation when that low-cost position gained by the firm is not unique within the industry. A prime example of which would be the automotive industry.
One contrasting theory of corporate strategy to that of Porter is the RBV (Wernerfelt, 1984;
Barney, 1991). The RBV explains performance differences between competing firms by focusing on the resources as the tangible and intangible assets that a firm possesses, and how these resources contribute to its competences and strategic capabilities (Grant, 1991). The RBV also emphasises the rooting of capabilities in the culture, or evolution, of the organisation (Mintzberg, 2008). Examples of resources that would contribute to the strategic direction and profitability of motor manufacturers would include: research and development assets such as patents, process technology, and product technology; as well as marketing, distribution, and service capabilities alongside the power of the brand. Grant (1991) emphasises the importance of harmonizing the exploitation of existing resources with the development of the resources and capabilities for competitive advantage in the future. The key to this is investment in the firm’s resource base, and therefore this thesis takes a resource-based view of strategy as being signalled by these investments.
2.5.2 Strategic Similarity as a Moderator
When assessing what impact strategic similarity between firms will have on deal performance, two opposing schools of thought are prominent. There are those who argue that strategic similarity is beneficial to the deal process as it allows for gains in economies of scale and increased market penetration (Datta, Grant & Rajagopalan, 1991; Ramaswamy, 1997).
Conversely there are those that argue in order to best achieve value creation and exploit synergies it is best to have an element of dissimilarity between firms (Harrison, Hitt, Hoskisson, & Ireland, 1991), as this allows for the maximum potential of learning effects from what the opposing firm does well, and vice versa. Harrison et al. (1991) demonstrate that dissimilarities between firms on strategic variables such as R&D, administrative, capital and debt intensity are positively related to post-deal performance. This brings evidence to a theory that differences may in fact amplify the synergies that can be achieved on uniquely valuable differences.
On the other hand, Ramaswamy (1997) uses the concept of strategic similarity to explain performance differences following bank takeovers and shows that deals between firms exhibiting similar strategic characteristics are more successful than those that are strategically dissimilar. Datta, Grant & Rajagopalan (1991) analysed management styles of acquirer and target firms and found that inconsistency between management teams along the lines of decision-making approach, risk-taking propensity, and time orientation was negatively related to post-deal performance. Incorporating the RBV of firm strategy (Wernerfelt, 1984) suggests that firms combining similar resource allocations to critical areas could be expected to enjoy greater performance than firms combining disparate resource allocations (Harrison, Hitt, Hoskisson, & Ireland, 1991). In the motor manufacturing industry where margins are slim and the importance of economies of scale is great, an argument can be made that deals between firms exhibiting strategic similarity will positively moderate the relationship between deal type and deal performance.
Whilst existing literature on the moderating relationship of strategic similarity in this context is limited, Fey and Beamish (2001) found that JVs with similar organizational cultures had a higher probability of success. Whilst these differences would also affect a merger or acquisition, they likely affect a JV more because the two parent firms retain their separate management (Beamish & Lupton, 2009). In order to incorporate the RBV into a measure of
capabilities as being an integral part of strategy. Previous studies quantifying firm strategy have used market coverage and marketing activity as strategic variables (Ramaswamy, 1997).
Therefore, this thesis uses 3 strategic variables of: firm internationalisation; selling, general, and administrative expenditure (encompassing marketing expenditure); and R&D expenditure to measure strategy. The following hypotheses are therefore developed:
H3A: Strategic similarity in firms internationalisation strategies positively moderates the positive relationship between choice of JV as a deal-type and post-deal firm performance
H3B: Strategic similarity in firms selling, general, and administrative expenditure positively moderates the positive relationship between choice of JV as a deal-type and post-deal firm performance
H3C: Strategic similarity in firms research and development expenditures positively moderates the positive relationship between choice of JV as a deal-type and post-deal firm performance
2.6 Conceptual Model
Based on the previous theory and to summarise the established hypotheses, the conceptual model (Figure 1) shows the proposed positive relationship between choice of JV as deal type and post-deal performance (H1). Further, the positive moderating effect of cross-national distance (H2A-D), and the positive moderating effect of strategic similarity (H3A-C), on the positive influence of JV choice on deal performance are integrated into the model.
Figure 1: Conceptual Model
2.7 Literature Review Conclusion
To conclude the literature review, a brief summary of the main points to be understood is presented here. Firstly, evidence on current research around the choices of deal types available to firms undertaking exogenous growth was presented, and an argument was made for the opposing choices that acquisitions and JVs represent. Secondly, the literature around deal performance was critically analysed. It appears that accounting-based financial measures of performance, although not perfect, are frequently used in judging ‘success’, of which ROA is the most common used measure. The moderating effects were then brought into the review with a discussion on how JVs can better cope with the challenges that cross-national distance presents to international expansion, as well as the importance of strategic similarity in a JV due to the retention of two separate parent company strategies that may clash if not aligned. Finally, the conceptual model was developed from the hypotheses and presented.
Chapter 3: Research Methodology
In order to determine the effect of deal choice on post-deal performance quantitative analysis techniques are employed. Hypotheses are evaluated using IBM SPSS statistics (version 24).
Due to the conceptual model containing more than one independent variable explaining one continuous dependant variable, linear regression analysis based on order of least squares (OLS) is used. Other studies using similar variables employ a technique of hierarchical regression analysis (Ramaswamy, 1997; Zollo & Singh, 2004), thus a similar method is followed in this study by adding variables to the model step-by-step.
3.1 Research Philosophy
Johnson & Duberly (2000) state that crucial issues such as the way in which a research question is phrased, and the reasons behind the selection of a particular methodology, can be traced to a researcher’s epistemological beliefs, irrespective of whether these beliefs have been consciously examined or discussed. It is therefore becoming increasingly important for management researchers to reflect on their own epistemological beliefs and the implications for their conducted research.
This research has been conducted in the realms of positivism in which it is assumed there exists a reality that is independent from the observer (Easterby-Smith, Thorpe, & Jackson, 2012).
This approach allows the application of data to verify or falsify predetermined hypotheses. To allow for this data is expressed in quantitative form and a large sample of organisations and deals is collected. As Easterby-Smith et al. (2012) point out, this will allow the uncovering of patterns and regularity, however they also concede that it is only a matter of probability that the views collected will provide an accurate indication of the underlying situation.
The main criticism of a positivistic approach is its failure to appreciate the objective nature of data and methods (Kim & Donaldson, 2016). Although it is true that these methods are not perfect in understanding underlying processes and significance that people attach to actions, they are useful in understanding a wide coverage of the situation as well as being of considerable relevance to policy considerations (Easterby-Smith et al. 2012). In this study a quantitative approach allows a fast and easily understandable analysis of the data. Although the research has been conducted with a neutral approach, it is likely that the researcher’s own background and beliefs have influenced the interpretation of the data to some extent. It is
possible that other researchers conducting similar analyses may conclude different interpretations, however it is hoped that insights gained here can be incorporated into future research alongside helping managers gain a better understanding of the mechanics behind a successful deal.
3.2 Data
In collating a sample for the research, data was obtained on deals in the industry through Zephyr. Zephyr is a comprehensive M&A database with integrated detailed company information, making it easy to find and classify deals. Firms were categorised by North American Industry Classification System (NAICS 2017) code 3361 relating to motor vehicle manufacturing. A preliminary sample acquired through Zephyr for all completed acquisitions and JVs between 01/01/2010 and 01/01/2015 consisted of 278 deals. After data collection and application of exclusion criteria, the final sample size consisted of 110 deals. The reduction in sample size primarily was due to a lack of data for the smaller firms in the sample, for example some smaller firms from countries such as Russia and China did not report data allowing the calculation of ROA, R&D etc.
The 5-year time period 2010-2015 was chosen due to the fact that whilst recent enough to ensure the managerial relevance of the implications hoped to be drawn from the study, the period avoids the turbulence of the 2008 financial crisis. Prior studies have used time periods of similar length for example Datta et al. (1991) take a 4-year period, Singh & Montgomery (1987) a 5-year period, and Ramaswamy (1997) uses a 7-year period. The 3 studies also carry out analysis on as little as 46 deals, to as high as 703. Adopting the rule of Harrel (2001) that a sample should include a minimum of 10 observations per variable, 110 should be the minimum sample size to account for 2 independent variables, 7 moderating variables, and 2 control variables. Therefore, it is clear that the final sample of 110 is sufficient to obtain reliable results.
To be included in the sample, a number of exclusion criteria were applied. Firstly, deals must be confirmed to have occurred. Whilst Zephyr classifies confirmed deals this was checked against news articles from the time to authenticate the deal. Secondly, sufficient data for all variables must be available from the data services used (Orbis/DataStream/ThomsonOne/annual reports). Thirdly, firms should not have been
involved in further deal activity during the 3-year measurement period, as this allowed for analysis of a single deal in isolation, without any external influence (for JVs this applies to the JV only and not parent companies). Only when the criteria were met were the deals included in the final sample. After data collection a final sample of 110 was deals was used to test the hypotheses.
3.3 Measures
3.3.1 Dependent Variable: Post-Deal Firm Performance
In determining a measure for deal performance, the question arises as to what determines success? Should deal performance be measured in terms of share price? Or are there more appropriate measures of success available? Whilst a number of studies focus on share price, some such as Datta et al. (1991) have even used a more objective based measure of performance, by asking top managers to fill in a questionnaire on their assessment of the impact of a merger or acquisition. For the purpose of this research performance is measured by the return on assets (ROA) of the post-deal firm. As argued in chapter 2.2 financial measures of performance avoid the problems associated with speculative and objective measures and are widely used in similar studies (Zollo & Singh, 2004; Porrini, 2004). For acquisitions the target firm is used and for JVs the JV firm itself is used. The ROA for each firm is collected from Orbis, a database with information on over 300 million companies worldwide. Meeks & Meeks (1981) argue that of accounting measures of profitability, the ROA is least affected by the estimation bias that can occur due to changes in leverage or bargaining power that can occur with a merger. The question then becomes how long after the merger should data be collected.
There is no exact science to this, as value creation can be spread out over a number of years.
Ramaswamy (1997) uses a 3-year period post-deal (and a 3-year pre-deal period used as a control variable), whilst discounting the merger year as it is difficult to pinpoint when both sets of accounts were combined. Post-deal performance was therefore measured as the ROA of the post-deal firm for a 3-year period, an increase over this period compared to the pre-deal ROA of the firm would suggest improved performance. The dependent variable ∆ROA is therefore calculated as the change in ROA between the year prior to the deal and the 3-year post-deal measurement.
3.3.2 Independent Variable: Deal Type
The independent variable in the model is the type of deal, here divided into acquisitions and JVs. In categorising the deal, a methodology similar to Hennart & Reddy (1997) was employed.
Hennart & Reddy used a dummy variable to capture mode of entry into the US by taking the form 0 if a Japanese parent made an acquisition and 1 if they established a JV with an American firm. Akin to this, using secondary data it was possible to categorise each deal into one of the categories. A dichotomous variable for deal type was created that takes the value 1 when a deal is a JV or 0 when it is an acquisition.
3.3.3 Moderating Variables: Cross-National Distance
The data used for cross-national distance in this thesis was made available to students of the University of Groningen in the module Comparative Environmental Analysis taught by Prof.
Sjoerd Beugelsdijk. The data consists of scores across all 4 dimensions of distance (cultural, administrative, economic, and geographic) for over 200 countries.
Cultural distance is measured through the application of Hofstede’s (1980) cultural dimensions framework and calculated through a Mahalanobis index. The Mahalanobis index is scale- invariant and takes into consideration the variance–covariance matrix, a feature that facilitates approaching distance as a construct made of multiple, partially overlapping dimensions (Berry, Guillén, & Zhou, 2010). Berry et al. (2010) showed that for multidimensional frameworks with correlated dimensions (such as Hofstede’s) a Mahalanobis correction is required. This shows the dyadic cultural distance between home countries of two firms. Although the work of Hofstede has been the subject of much criticism in the literature (McSweeney, 2002;
Baskerville, 2003), Drogendijk & Slangen (2006) argue that Hofstede-based measures of cultural distance have significant explanatory power and are reliable measures of cultural distance in the form of a Mahalanobis index.
In calculating the administrative distance between two countries the Mahalanobis distance of quality of governance from the World Bank Data on World Development Indicators (WDI) was used. This data collates governance indicators on the grounds of: voice and accountability;
political stability and absence of violence; government effectiveness; regulatory quality; rule of law; and control of corruption. The aggregate indicators are based on several hundred individual underlying variables, taken from a wide variety of existing data sources. The data
reflect the views on governance of survey respondents and public, private, and NGO sector experts worldwide (Kaufmann, Kraay & Mastruzzi, 2011). This therefore provides a comprehensive analysis of the administrative environment in each country.
Geographic distance was calculated by using the total greater circle distance between two countries, measured from the geographic centre of the home country of one firm to the geographic centre of the home country of the other firm akin to the method employed by Bailey
& Li (2015). Exact coordinates can be obtained using the CIA factbook and calculated by Berry et al. (2010).
Economic distance was calculated by taking the GDP per capita of a country (in USD) and comparing this to another country at the same point in time. Similar measurements have been made previously by Berry et al. (2010). Data was collected from the WDI. This gives a consistent measure on which economic development can be compared across borders.
3.3.4 Moderating Variables: Firm Strategy
Data on firm strategy was collected primarily from Thomson Reuters DataStream. DataStream is the world’s largest historical financial database, with data available for companies worldwide. Where data could not be ascertained from DataStream, Thomson One was used, and where both databases failed to provide adequate information, the annual reports of the companies themselves were used.
Despite firm strategy being a complex and normative concept, many attempts have been made to include it in empirical research. Strategy may be stated by top management (MacCrimmon, 1993). Therefore, one potential course of action would be to analyse, categorise, and code, management statements on strategy prior to consolidation taking place, and compare these with the counterpart of the other firm. On the other hand, a more categorical assessment of firm strategy could be implemented by using Porter’s Generic Strategy framework (1980) and placing firms in one of the four strategic options available. This would give a clear indication of firms strategic similarity, however there is an argument that motor manufacturers now adopt approaches of ‘mixed strategies’ as argued in chapter 2.5.1, thus it may prove difficult to categorise these firms into a discrete category.
Previous studies on strategic similarity have often focused on resource allocations as a key indicator of strategy. Ramaswamy (1997) uses five strategic variables designed for the banking industry (market coverage, operational efficiency, level of marketing activity, client mix, and risk propensity) to calculate the similarity between firms. A similar approach to Ramaswamy was followed in this thesis, however adapting the indicators to the motor manufacturing sector.
Firstly, market coverage was determined as a construct of the internationalisation of a firm.
Inspiration here was drawn from Rugman & Verbeke (2004) who measure global strategies of firms through the percentage of their sales in different regions. For this measurement a simplification was made that any firm for which greater than 50% of total revenue came from sales outside their home nation was considered to be ‘international’. If more than 50% of total revenue came from sales inside their home nation, then the firm was considered to be ‘national’.
Greater than half of a firm’s revenue coming from outside its home nation, would likely translate into its strategy focusing on international operations. This therefore represents a distinct approach in order to realise competitive advantage. A dummy variable was produced on this basis taking value 1 for an ‘international’ firm and 0 for a ‘national’ firm. Secondly, expenditure on advertising, selling, and delivering a product is a fundamental indicator of strategic differences between firms. In financial statements firms report these costs as selling general and administrative (SGA) expenses. Whilst some motor manufacturers spend significant sums of money on pursuing such costs, others restrain expenditure to focus on other aspects of the business. By taking a ratio of SGA expenditure to revenues it is possible to compare this level of strategy. Finally, akin to SGA expenditure, R&D is a key variable of strategy in motor manufacturing, as firms are under constant pressure to stay ahead of the competition. A ratio of R&D expenditure to revenues gives a clear indicator of whether firms are pursuing a strategy of innovation. The ability to take ratios of expenditures to total revenues means that the resulting figures are easily compared across firms irrespective of size. This results in the formation of 3 variables measuring the strategic differences between pre-deal firms that are ∆Internationalisation, ∆SGA Expense, and ∆R&D Expense. All measured as the difference in strategies of the two firms in the accounting year prior to the deal.
3.3.5 Control Variables
Two control variables are proposed for the model. These are the weighted average pre-deal ROA of the two firms and the relative size of the acquirer and target firms to each other.
Relative size is to be used as a control variable as prior studies have shown that it has a
significant impact on post-deal performance (Kitching, 1967; Kusewitt, 1985). A significantly larger firm buying a smaller firm could impose immediate economies of scale benefits on the target firm that do not represent an increase in underlying efficiencies. The use of pre-deal performance as a control variable accounts for the ‘floor/ceiling’ effects that can be observed when looking at magnitudes of change. This occurs as manufacturers that were already performing well may be unable to improve their performance as much as those that were low performers (Ramaswamy, 1997).
Chapter 4: Results
The following section outlines first the preliminary tests for the assumptions of linear regression to ensure the validity of the results obtained. Secondly, descriptive statistics and correlations are analysed. Then the results of the main regression analysis are presented. As moderation effects are included in this study the continuous moderators for Cross-national Distance, as well as the moderators of ∆SGA Expenses and ∆R&D Expenses are mean centred for the entire analysis.
4.1 Preliminary Tests
Normal distribution was first tested for alongside skewness and kurtosis through a Shapiro- Wilk test across both levels of the independent variable deal type. The results show that skewness and kurtosis appear to be present in the data and that the data appears not to be normally distributed. Despite this, normality of the raw data is not an essential requirement for linear regression (Lumley, Diehr, Emerson, & Chen, 2002). To ensure that this skewness &
kurtosis does not create non-parametric, non-linear effects the Pearson and Spearman correlation coefficients were compared. On the whole these delivered similar results thus this does not seem to disturb the data. A key assumption of linear regression however, is that the residuals follow a normal distribution. With the variables presented in their raw data form, this was not the case. Therefore, a transformation was required to ensure that this assumption was met.
In order to transform the data, the 2-step process advocated by Templeton (2011) was used.
This created a new dependent variable Deal Performance out of the raw data on ∆ROA. The first step of this process involves transforming the variable into a percentile rank, resulting in uniformly distributed probabilities. The second step then involves applying the inverse-normal transformation of the previous step’s results to form a new variable consisting of a normally distributed Z score. This method allows the retention of the original series standard deviation with a similar mean, and once completed the new variable exhibited normally distributed residuals, with all standardized residuals falling within the acceptable -3 to +3 range, alongside no outliers being present when analysed with SPSS’s casewise diagnostics. Due to the control variable Pre-Deal Performance also being measured on the same scale as the dependent variable, the 2-step process was also applied here to ensure continuity in the results.
Next multicollinearity was tested for by using the variance inflation factor (VIF). The VIF and tolerance for most variables was in the acceptable range of not higher than 10 and not lower than .2 respectively (Field, 2009). The only variables for which this was not the case were Cultural Distance, Administrative Distance and Economic Distance. This multicollinearity is not surprising as these three measures of cross-national distance are likely correlated due to the fact that those nations that are richer (Economic Distance) compared to their counterparts, are likely to have differing institutions (Administrative Distance). As well as those nations that are culturally distant (Cultural Distance) being more likely to have different economic, political, and legal systems. The decision was made to leave these variables in the model and run separate regressions testing the impact of these moderator variables both individually and then together.
Autocorrelation between the error terms was tested for using a Durbin-Watson test to evaluate the independence of the errors. The tests for the regressions produced scores within the range 1.752 to 2.038 which are all within the acceptable range of between 1.5 and 2.5 (Field, 2009).
Therefore, it can be concluded that autocorrelation is not present. Finally, in order to test for heteroscedasticity, both a Breusch-Pagan & Koenker test were performed on all regression models used in the main analysis. For all models heteroscedasticity is not present, therefore it is concluded that it does not disturb the results.
4.2 Descriptive Statistics
The descriptive statistics for the data used in the analysis are presented in Table 1. Of the 107 deals sampled 45 were cross-border deals, with the remaining 65 being conducted within the borders of a nation. Firms involved in the deals hailed from a total of 25 countries, ranging from the US, Mexico and China, to Kazakhstan and The Philippines.
Table 1: Descriptive Statistics
Variable Min Max Mean SD
∆ROA (raw data) -60.00 46.76 -1.42 18.21
Deal Performance -39.53 41.58 -1.02 17.96
Deal Type 0 1 .24 .43
Cultural Distance 0 4.15 .91 1.25
Administrative Distance 0 2.61 .64 .94
Economic Distance 0 47001.22 12328.55 17697.01
Geographic Distance 0 18549.61 2296.26 3821.02
∆Internationalisation 0 1 .35 .48
∆SGA Expenses 0 45.64 4.92 7.18
∆R&D Expenses 0 8.22 1.70 1.93
Pre-Deal ROA (raw data) -90.15 46.45 -1.05 16.55
Pre-Deal Performance (transformed) -40.15 38.05 -.69 16.37
∆Size 0 2.08E+11 2.36E+10 4.01E+10
Note: N=110 deals, Variables are presented in original form and not mean centered
The raw data for ∆ROA shows that firm performance post-deal varies greatly ranging from -60 to +46.76. It is interesting that the average value of the data is negative (mean=-1.42). This reflects the skew of the data towards the lower-value range, however it shows that in this sample on average deals in the industry lead to a reduction in ROA in the firms analysed. The transformed variable Deal Performance addresses the issue of skewness in the data set while keeping a similar mean and standard deviation (mean=-1.02, SD=17.96).
Out of the 110 deals in the industry between 2010 and 2015 only 26 were JVs. The remaining 84 were acquisitions. This shows that despite the seeming growing favourability towards JVs, the vast majority of deals are still acquisitions. Cultural Distance ranges from a value of 0 for those deals that were not international, to 4.15 for the acquisition of Volvo (Sweden) by Zhejiang Geely (China). Again, the number of international deals being conducted at 45 represents the fact that in the sample the majority of deals were conducted within borders.
Administrative Distance ranged from 0 to 2.61 in the most extreme case in the acquisition of Russian AVTOVAZ by Dutch based Alliance Rostec Auto BV. Economic Distance also varied from 0 to a maximum of 47001.22 in the Zhejiang Geely takeover of Volvo. The greatest Geographic Distance in the sample came from the formation of Mazda’s JV in Brazil with a distance of 18549.61 between the countries. Of the firms involved in the deals in the sample 155 were judged to be ‘national’ in the year prior to the deal and 65 were judged to be
‘internationalised’. This shows that despite the most well-known brands having a global presence such as Nissan, Renault, Mitsubishi, FiatChrysler etc. There are still a significant number of manufacturers that choose to focus efforts in their home nation. From the sample this seemed to predominantly stem from Chinese manufacturers, for which all bar 5 focused sales efforts at the significant PRC market. Descriptive statistics for differences in SGA and R&D expenses show that in the sample there seems to be greater differences in ∆SGA (mean=4.92) than ∆R&D (mean=1.70). This is to be expected as firms often spend significantly larger amounts on SGA expenses that R&D expenses, so greater variation is not unusual. The control variables show that the average firm in the sample experienced a negative ROA prior to involvement in a deal (mean=-1.05) and that the average size difference between firms was substantial (mean=$23.6bn), showing that in the majority of cases a larger firm is involved in a deal with a smaller firm.
4.3 Correlations
The correlation matrix is presented in table 2 below. In the data set the only variables that correlate significantly with Deal Performance are ∆SGA Expense, ∆R&D Expense, and Pre- Deal Performance. Cultural, Administrative, Economic, and Geographic Distance significantly correlate together. This is to be expected and fits with prior theory as they are all measures for the construct of Cross-National Distance. ∆Internationalisation seems to correlate significantly with all aspects of Cross-National Distance as well. Again, this is not surprising as logic would suggest that an internationalised firm engaged in a deal with a national firm would be doing so across borders. Some aspects of Strategic Similarity seem to be correlated, namely ∆SGA Expense and ∆Internationalisation, and ∆SGA Expense and
∆R&D. As these are measurements of the same construct it is not unusual that there would be correlations between the variables. In order to prevent this changing results in the regression, the moderators for Cross-national Distance as well as Strategic Similarity were first tested separately, then simultaneously in the analysis.
