M&A foreign investment decisions

(1)

RIJKSUNIVERSITEIT GRONINGEN

FACULTY OF ECONOMICS AND

BUSINESS

UPPSALA UNIVERSITETET

FACULTY OF SOCIAL SCIENCES

M&A foreign investment decisions

-A study of cultural influence in foreign bias effects-

Master Thesis

MSc IB&M – International Financial Management

Programme Coordinator and Thesis Supervisor: Dr. W. Westerman

MSc Student: Lucian Pătulea

(2)

1

ABSTRACT

Limited research has been granted so far to foreign bias behavioural phenomena in financial resource distribution and international strategy building, and even less to focusing on the cultural reasoning behind such managerial decision making processes. Moreover, in the last years M&As have become the bulk of FDI cross-national relations and one particular area where deals have developed in intensity and frequency is the European context. Based on the previous stated, the paper focuses on M&A deals that have taken place within the last decade among a European set of countries with a distinct array of cultural backgrounds. Furthermore, the deals under study are formed between companies from the financial sector, as being the most dynamic in cross-national interactions. Thus, an effort will be made in trying to find patterns of foreign investment deals and foreign bias trends, linked to the cultural dimension variables of individualism and uncertainty avoidance. Results overall show a negative impact of individualism on M&A foreign distributions, and moreover, a negative impact as well of both individualism and uncertainty avoidance on foreign bias effects (namely irrational foreign funds allocation that hinder investment diversification gains).

Key words: behavioural finance, foreign bias, home bias, cross-national M&As, cultural

dimensions, individualism, uncertainty avoidance, Europe, EU.

Acknowledgements

(3)

2

CHAPTER 1 RESEARCH OUTLINE

1.1 INTRODUCTION

Decision-making process is overall a rather difficult and non-sequential series of a random amount of steps that lead to a more or less appropriate solution for a given situation. Economic science has rigorously made the effort to step in and help managerial functions, by attempting through a number of models to predict and standardise the above referred process. It might be easier to integrate and stick to the rules in theory than in practice though. Practicality of the real world brings to view lack of available information and bounded or limited rational behaviour, which breaks apart from the basis of economic thought up to this point in time (March, 1994). Moreover, it involves the study of a parallel field, with a certainty of fewer rules to follow, psychology. It brings to view the aspects of cognitive biases and intuition, thoughts that come to mind in a rush and with little basis of reflection and apparent motivation (Kahneman, 2002), but that represents an important weight in the process of taking decisions.

(5)

4

future research in these directions, it has been a rather neglected cocktail of factors studied so far (Beugelsdijk and Frijns, 2010).

More specifically, the current research is focusing on cultural impact of investment behaviour of financial institutions, in an international context. It studies rationality of such decision making processes, influenced by home bias (disproportionate allocation of funds to the domestic market) and more specifically foreign bias (under or overweighting foreign markets in asset allocation to the detriment of others, based on different than rational cognitive processes) (Chan, Covrig and Ng, 2005). Additionally the impact of several more traditional, but secondary factors (like GDP growth and geographical proximity) will undergo the analysis. Cultural proximity, in sense of familiarity, is expected to have a more demanding appeal than the apparent loss of rational behaviour, the lack of diversification, and consequent opportunity costs that are accumulated at organizational and national level. The research promises relevance in both the business area, and in the academic one. Thus, relevantly from a managerial point of view, the acknowledgement of cultural effects on current and future strategies may well improve rationality in decision making and minimise costs of suboptimal planning, when focusing on expansion and investment home and abroad. Subjectivity and biases may be minimised and results of future returns and strategic plans, on the contrary, maximised.

(6)

5

foreign bias in investment decisions, but to complete the object of the study, it analyzes specifically the weight of the cultural impact and how it is affected and moderated by unification efforts at EU level. The detailed explanation of choices and analysis of each element and every method are discussed later on in the current paper.

1.2 RESEARCH QUESTION

Given the so far stated, I can formulate the main research question that the study will attempt to investigate, among other secondary effects.

Do cultural dimensions (individualism and uncertainty avoidance) influence choices concerning M&A deals, leading to distinct patterns of deviation from rational foreign investment behaviour (foreign bias)?

1.3 BRINGING A FRESH PERSPECTIVE

Set apart from previous behavioural studies, the current research takes the aim of focusing mostly on the cultural impact and effects that are developed at the psychological level so as to deter and offset managerial thinking patterns away from the traditional rational steps. It analyses the choices, at country level, of when and where to focus financial efforts so as to get best possible returns on investment. The cultural distance is a factor that will be taken in account but at a more basic level than the one that previous studies preferred to attempt, like calculating a distance index. Distance indexes are proven to be difficult in estimation and understanding, resulting in misleading and debatable values with a rather questionable explanatory power (Shenkar, 2001). In such circumstances I prefer to choose a relatively broader system of measurement and calculate distance through the appurtenance, or not, of a target country to the same cultural cluster, as determined by the GLOBE project. Moreover this enables the opportunity of combining two cultural measurement systems (the one by Hofstede and GLOBE’s). Although being related, the two frameworks combined improve the potential accuracy of results and associated confidence.

(7)

6

though the level of specificity of the study is expected to bring more pertinent and easily applicable results to the professional and academic setting, it does unfortunately contribute to a more acute degree of limitations that are to be explained more thoroughly towards the final part of the paper.

From an economic, geographical and regional point of view, the perspective rests at European level, where a great part of the country clusters are met (obtaining a desired level of cultural variation), but also where a high and intensive degree of M&A and general FDI activities are present. Buelens (2008) indicates in his presentation for the European Commission these recent trends and the increase in importance of the European market for M&As. As an additional factor that explains the choice of the world focus, is the location of the European Union, as the most developed and multitasked regional economic and politic block, with persistent and fierce efforts of integration. Such efforts of integration and globalization effects will be accounted for as a secondary aspect of study, to observe results of overcoming a strong cultural diversity. Finally all phenomena described will be under a recent period of study (last decade) and variations within this period will be observed for determining weakening or stronger trends in investment patterns and cultural impacts. Although a small period per se, minimal temporal variations may prove a good predictor of the recent future.

1.4 PAPER STRUCTURE

(8)

7

interest to the present topic to pursue, perfect, sustain or constructively contradict its findings (in Chapter 5).

CHAPTER 2 THEORETICAL CONSIDERATIONS

The following pages will introduce the reader to basic elements and manifestations of behavioural finance and its main causes. Further on, trends and relevant results of investment will be outlined and more specifically, M&As’ strategies. Thus, here the theory will build towards the primer effect under observation, foreign bias, and shall additionally be giving details about recent M&As trends in the European context. The following subsection will continue with the cause-focus of the current paper, cultural effects, and build an explanation of the relevance of culture and its measurement techniques. Lastly, the final remarks of the chapter are reached, towards developing the research hypotheses.

2.1 BEHAVIOURAL CORPORATE FINANCE

Thinking on the basis of traditional finance, the main concepts on which these theories are constructed consist of economic agents’ rational behaviour, market efficiency and the rules of the capital asset pricing model (Shefrin, 2001). These elementary parts determine the process and range of outcomes from a managerial decision-taking perspective. However, a more recent stream of financial studies shows that choices are not always the outcome of such rigorous methods. Cognitive resources (like time, memory, and attention) are limited and most often developed under the impact of feelings or emotions. Consequently, decision-making processes’ “thoroughness” is affected by imperfection or by the so-called heuristics (Solnik, 2006). Overconfidence, overreaction, loss aversion, and narrow framing, or the difficult to explain phenomena like the “January effect” or “day of the week effect”, all come under the study of behavioural sciences. These concepts confirm the psychological imperfection of the human decision making. Some choices are made under consistent social pressure and need for conformity, others come as mere intuition (thoughts that arise hastily and without reflection) (Kahneman, 2002), and sometimes even under the pressure of a feeling, an affect that guides judgement without too much consciousness (Slovic, Finucane, Peters and MacGregor, 2002).

(9)

8

decision, the economic agent is put under the real pressure of uncertainty of the given alternatives, constant assumptions modification and subjective personal perception (risk averse or risk taker; enthusiast or sceptic). Thus, he or she is subjected to limited or bounded rationality, the grindstone of behavioural finance. Decision making is not solely processed under the sign of objective risk, but once more the situations are perceived to be more complex when taking in account the subjective risk (Statman, Fisher and Anginer, 2008). Diving just for a moment in the consumer behaviour area of study, Zajonc describes a realistic hypothetical situation as such: “We buy the cars we ‘like,’ choose the jobs [...] we find ‘attractive,’ ”; “We do not just see ‘a house’: we see ‘a handsome house’, ‘an ugly house’ ” (Zajonc, 1980). The importance of affect (integral or even incidental to the situation per se) and personal experiences are further underlined in other behavioural studies. For example, Welch in his doctoral dissertation in 1999, performs a study where he concludes that after being exposed to watching an intense emotional movie, subjects’ risk aversion was substantially increased, although the phenomenon was nothing more that an emotion carried after the initial context it was associated with (Statman, Fisher and Anginer, 2008).

At corporate level, psychological forces and behavioural off-sets from traditional and rational decision-making processes, lead to local behavioural costs (at the level of the corporation itself) and other drawbacks, from behavioural errors made by analysts and investors (external to the corporation) (Shefrin, 2001). The focus of the current research will be at the managerial level of internationally oriented investors, within corporations. Their decision making processes and outcomes will be analysed and behavioural constraints are to be identified and measured according to their impact.

2.2 INTERNATIONAL ENVIRONMENT AND INVESTMENT TRENDS (THE FOREIGN BIAS)

(10)

9

financial institutional investments are made from: pension funds (employees and workers’ contributions for post-retirement cash flows), insurance companies (with focus on life insurance policies) and general investment companies (mutual funds, investment trusts, private equity funds). The increase in international flows should theoretically encourage improved risk management and spreading of investor risk among its active base.

The behavioural effects are commonly noticed at international asset allocation and asset management level. It narrows down to two strongly related phenomena identified as home or domestic bias (disproportionate allocation of funds domestically) and foreign bias (irrational foreign funds allocation so as not to fully beneficiate of potential diversification investment gains) (Chan, Covrig and Ng, 2005). In this paper foreign bias will be measured as the proportional difference between optimal investment and actual investment that a country performs in another one, at a given moment in time. Familiarity, economic development, capital control and withholding from tax variations are just some of the main reasons to determine the over or underweighting of foreign markets in investment patterns (as manifestations of foreign bias). Standard, more traditional financial rationale (biases based on restriction of capital flows, non-tradable instruments, institutional barriers and transaction costs), holds equal ground and even gives way to subjective and personal preferences (overconfidence, personality compatibilities, language barriers, geographic proximity and cultural distance) (Beugelsdijk and Frijns, 2010).

(11)

10

connection to investment behaviour, through the principle of familiarity, are to be further explained in the following section.

Studies on optimal portfolio theory have proven that foreign assets contain a vital role in the diversification gains that investors can obtain, although a bigger percentage of domestic stakes may be a defining element as well (Flavin and Wickens, 2001). Empirical proof and theoretical thought consistently keep the domestic bias myth still very much alive though (Appendix I: Table A1). The domestic stakes in the five largest stock markets regarding corporate equity prove to be larger than 80%, which is connected with a considerably small amount of diversification (French and Poterba, 1991). Consequently the diversification level is well under the contemporary international constraints. Foreign and home biases are believed to have roots in instincts, compensating for hard to predict future returns and perceptions of risk that replace thorough historical data based risk evaluation.

Although it is generally agreed upon the fact that home bias is an intense and perpetual phenomenon with not so impressive weakening trends, foreign asset allocation is nevertheless subjected to investors’ preference as well. The bias originates on the basis of different considered reasons, varying from financial ones (market development and tax incentives) to more socially rooted familiarity (language, geographical distance, culture) (Beugelsdijk and Frijns, 2010). The latter represents the major focus of the current research. Such behavioural patterns are expected to manifest in a stronger manner in non-professional investors than institutional ones. The first may rely more on country specific factors as they are subject to even more limited information (Giofré, 2008). Furthermore, studies with a Finnish background prove that less sophisticated investors (small companies and households) are more prone to be subject to distance, culture and language differences (Grinblatt and Keloharju, 2001).

(12)

11

after 1992, when companies were facing heightened competition and free borders, nowadays the M&As activity has somewhat stabilised and gained a more national character once more. Moreover, companies from external areas to the EU, after craving for a foothold in the newly formed market have now reached a satisfaction point (Ietto-Gillies, Meschi and Simonetti, 2005). EU market’s M&As consequently faced internally a shortfall from 2000-2003 onwards, and have refocused on external markets (mostly distant ones like the Asian) (Buelens, 2008).

Difficulty of integration of M&As has been a constant detrimental factor. Nevertheless, the desire to reach synergies, to gain know-how and increase managerial ability has pushed internally such activities, focused on a regional scale though. The regional concentration of these international expansions has indeed a negative impact on diversification (García-Herrero and Vásquez, 2007). Clustering among European countries is an apparently persistent phenomenon among other local European characteristics, such as the primacy of the United Kingdom in M&As and the divergent countries’ goals regarding integration and globalization (Ietto-Gillies, Meschi and Simonetti, 2005). Being such a heterogeneous mix, EU countries may face a longer path to total integration and still currently experience the forming of internal blocks and clusters. The attractive destination and major originating point of FDI in general remains the UK, since it has the most liberal and earliest policies, concerning attitudes towards trans-national companies.

Researchers have been split among the two mainstream trends across time periods: to encourage the resolution of foreign bias diversification issues, or rather encourage home bias altogether, as far as international M&As are concerned. Aybar and Ficici (2009) point rather to destruction of company value once engaging in cross-border acquisitions. The main reason that may be brought to view is that diversification benefits may be offset by institutional, national and organizational culture differences. These bring into discussion the real challenge of working in a foreign environment and with a foreign partner, apart from basic issues arising by the simple act of interacting with a partner per se (Lee, Shenkar and Li, 2005).

2.3 THE CULTURAL IMPACT

(13)

12

directly connected with the way one takes decisions. Although claimed not to have yet been receiving the necessary attention from the finance research community, culture, through its influence on the probabilistic thinking style of investors, differentiates among risk-taking attitudes, negotiations techniques for investment, mergers and acquisitions strategies and, sometimes even in bonds’ and stocks’ returns at local markets’ level (Hens and Wang, 2007). As Hofstede describes culture as a “collective programming of the mind” (Hofstede, 2001), the alteration of the system of values determines culture to have more than just a residual role, but actually have an important impact on investment decisions through the support for the familiarity principle at least (Beugelsdijk and Frijns, 2010). The impact of culture on the world of behavioural finance, which both are intricately and deterministically connected, has been measured empirically as well. For example some studies connect overconfidence bias with collectivist cultures, since in these cultures the accent lies on the necessity of conformity, and different opinions are not specifically encouraged (Yates, Lee and Shinotsuka, 1996). At a managerial level, culture acts as deterministic factor for compatibility. Uppsala University’s research has shown that culture contributes to development of “psychic distance” that on its turn determines the level of uncertainty that companies will manifest towards foreign markets of choice (Kogut and Singh, 1988). Culture thus influences perceptions of organisational costs and the degree of uncertainty among available alternatives of action. Perspectives and domestic cultural traits are that deterministic in decision-taking processes as such, that, as studies have shown, it depends on direction and focus and is not reciprocal. More specifically said, culture “distance” is not identical in reverse when two focal points (partners or potential partners in two distinct national cultures) are chosen (Lee, Shenkar and Li, 2008).

(14)

13

partnerships, empirical work has also proved that international diversification does indeed enhance the risk-return performance of a company, so as to compensate and balance the pros and cons on engaging in cross-border M&As (Kim, Hwang and Burgers, 1993).

Success rate of M&As and moreover, the probability of a company to engage in such activities may not all be correlated with cultural distance per se. Researchers have shown that the importance of the cultural characteristics of the originating country is rather more significant in these types of decisions, than the target’s (Fisher and Ranasinghe, 2000), hence the motivation to rather focus on cultural characteristics of these country-positions further on. On the one hand, uncertainty avoidance scores of the acquirer’s nationality, in direct correlation with risk averseness may be a potential powerful factor in explaining target country choices, taking the socio-cultural differences of the business environment into account. For example, a connection may be projected between high uncertainty avoidance values and the diminishing trade in similar countries, with possible connections to both foreign and home bias. One the other hand, high power distance discourages foreign investment, while individualism and masculinity on the contrary, overall encourage seeking such actions (Aggarwal, Kearney and Lucey, 2009). Nevertheless, a firm’s difficulty in foreign investment is far from being as simple and unidirectional as above. In order to be able to refer to a cultural distance, the home country is only one piece of the puzzle. Empirical results have reached conclusions regarding the choice of the host country, due to its cultural traits in particular. For example when choosing a target for engaging in foreign investment, countries with low uncertainty avoidance scores and high level of trust are better potential partners since an easier co-integration and acceptance is expected (Bhardwaj, Dietz and Beamish, 2007). For this purpose cultural distance is chosen as integrating factor for targets’ cultural influence in the following methodological approach.

(15)

14

and leads to great stress and acculturation issues at the organizational level or even representing an un-crossable line of failure, but this matter aside, distance may also lead to improved cultural tolerance or even a multi-cultural attitude (Cartweight and Price, 2003). Culture per se is certainly a difficult phenomenon to understand, it is complex and indeed extremely difficult to measure, not to mention the task of defining it (Carpenter and Wyman, 2009). Most efforts to measure cultural effects and more exactly cultural distance do rely on Hofstede’s work up to date. On the other hand culture’s impact, or its supposed impact to be more exact, on international M&As is particularly clear. Scientists so far have provided a sort of definition of the negative consequences of culture, connecting them with decreased productivity, which leads to lower revenues and lower eventual income, and overall result in forming an entity worth less than initially expected (Sherman, 2006). Consensus among researchers though has not yet been reached, if ever to be expected. Some still claim that M&As in cultural distinct contexts, or among culturally distant countries (including language and religion in the studies), do perform relatively better, but in the long-run (Chakrabarti, Jayaraman and Mukherjee, 2009). Most researchers stick to the mainstream line and acknowledge managerial conflicts and early dissolution vulnerability of international alliances. The latter add to a potential poor communication that would lead to obstructing the channel of management practices and technology transfer. Moreover some recognize the effects of “double-layered acculturation” when differences are striking, not only at national but also at the organizational cultural level (Barkema, Bell and Pennings, 1996). Contradictions, as mentioned before, do exist and some will go so far as to prove that cultural distance has no such powerful effect on international organizations, and it is not a deterministic factor of dissolutions (Park and Ungson, 2008).

(16)

15

the initial four (individualism vs. collectivism, uncertainty avoidance, masculinity vs. femininity, power distance) and a later fifth addition, Confucian dynamism or short-term vs. long-term orientation, which has been studied at a smaller level. In the current paper solely the first two are to be chosen and studied for the impact on international M&As, due to their most often association in previous research with investment behaviour. Individualism is a keen association with the ruthless gain driven investor, while uncertainty avoidance is connected most likely with the protectionist and risk averse behaviour of a common rational economic agent. The other cultural dimensions seemed to have a more moderating than deterministic impact, but so as to correlate for all additional cultural effects, a distance measure will be taken in account.

Hofstede’s work is not without a healthy and steady stream of theoretical critique. His “legendary” research has been attracting a lot of questions concerning first of all the plausibility of the national culture which Hofstede considers as implicit, systematically causal, territorially unique, shared and carried by all individuals at least as a central tendency (McSweeney, 2002). He assumes a sort of uniformity which is rather difficult to take as such. His nations have been compared to Andersons’ “imagined communities”, to strike the distance from realism, and many have been putting under the sign of doubt his measuring system, as Smelser definition of “a misguided attempt to measure the unmeasurable”. Lastly, his study is strongly limited to a sole organizational context and culture, being conducted within the IBM organization with a questionnaire of which many doubt its power of responsiveness (McSweeney, 2002).

(17)

16

2.4 HYPOTHESES CONSTRUCTION

2.4.1 Individualism and investment behaviour

According to Hofstede’s model, individualism/collectivism is a bipolar cultural dimension proposed first in 1984, and highly utilised ever since. It describes the emotional independence and the autonomy of a person or group of persons, and the strict concern and care for oneself and immediate relations, rather than for the large group one does belong to (Gouveia and Ros, 2000). At a national level, individualism may be part of the prevailing ideological mix, which is described as the capability of a nation to compete in an economic context within the international setting. Individualism, forming a dominant atomistic view, and closely related to attitudes of trust and equality, will indeed stir the need and desire for intense competitiveness, but might not always get favourable results due precisely to lack in flexibility and adaptability at group level (Lodge, 2009), and thus leading to a smaller expected rate of M&A deals to reach completion (the particular object of the study). The ideological clash between M&As’ partners may be sufficient to predict the unfavourable future resilience of such a project, and solely cultural understanding and efforts in this direction may lead to another fate (Ross, 1999). The main cultural effect on firms direct interacting in similar situations is mainly focused on the ability to combine efforts so as to lower costs and increase revenues, which is not always so easily done as said (Livermore, 2010).

(18)

17

(Livermore, 2010). One main supporting argument to the final statement might be the correlation between the amount of information needed and the one obtained when considering M&As, which is in a higher intentional percentage within sharing and trustworthy cultures, the collectivistic ones (Doney, Cannon and Mullen, 1998).

Individualistic cultures are more focused on attitudes, personal needs, rights and contracts so as to guide their social and business behaviour. They are highly independent, acting on own beliefs and personal experience, but most importantly develop an intensively competitive business background from the emphasis put on rational analysis and by attentively measuring advantages and disadvantages (Huff and Kelley, 2010). Rational investor behaviour is expected to diminish potential foreign bias and to take in account the potential of diversification only when benefits surpass costs, including cultural driven costs. However, M&A activities would be encouraged by intention, since this form of FDI per se is very “individualistic” in cultural orientation (Cartwright and Price, 2003). Researchers strongly correlated so far individualism, alongside other cultural dimensions (in the case of the original four, masculinity supports further the correlation), with a positive outlook in cross-border M&As strategies (Aggarwal, Kearney and Lucey, 2009), making a possible connection to a decreased investment home bias as well, solely if beneficial. Moreover, culture is proven to have a powerful effect upon probabilistic thinking styles. Individualism specifically, through competitive and assertive rationale behaviour, has thus a strong softening impact on risk averseness and an overall moderating effect on risk perception (Hens and Wang, 2007). Individualist decision making may thus have a more realistic perceiving of risk. However, at the same time it may create a more biased perceiving on return since individualism has indeed been correlated with another behavioural phenomenon and thinking pattern, overconfidence (Chui, Titman and Wei, 2010). Thus, although aggressive in initiating cross-border M&As, the completion level of such deals is expected to be lower, especially in culturally distant contexts. However, a more rational attitude economically-wise is adjacently correlated and consequently foreign bias effects might be expected to diminish as well.

Given the arguments constructed so far I build the following hypotheses.

(19)

18

Though, given an enhanced rational investment attitude:

H3: Individualism is negatively correlated with foreign bias effects;

2.4.2 Uncertainty avoidance and risk assessment

(20)

19

foreign M&A deals altogether. Desire for kinship is correlated with a smaller rate of investment distribution abroad and higher home biases as a dominant behaviour. The downsizing effect of investment per se is expected to surpass in consequence the level of completion of M&A deals based on trustworthiness.

Within the European context, uncertainty avoidance is more intensively manifested towards the Eastern cultures, with the exceptions of Belgium and Portugal, which have quite high levels as well (Appendix I: Fig. A2). Overall, uncertainty avoidant countries are empirically proven to trade disproportionally less than other when it comes to cultural distance (Gupta, Hanges and Dorfman, 2002), and, on their turn, do not attract as much foreign assets as others do (Tripodo and Dazzi, 1995). This could be a direct consequence of perception distortion. Alternatives in decision-making processes are more intensively perceived as uncertain, intensifying sentiments of risk aversion. Furthermore, modifications of assumptions and existing knowledge, due to continuous information flow, induce constant change in preferences and scepticism (March, 1994).

Uncertainty avoidance, in relation to investment patterns and home and foreign behavioural biases, is mainly correlated in research with risk aversion and avoidance of non-compatible and profitable situations (Beugelsdijk and Frijns, 2010; Hens and Wang, 2007), although some experts may disagree, including Hofstede himself (Hofstede, 2001). Empirical support mostly holds with the mainstream correlation though, and sometimes it might come from associations as creative as the one to high levels of insurance “consumption” (Chui and Kwok, 2008). In consequence M&A deals involving highly uncertain avoidant countries are considerably less but may have a stronger survival expectation (which is not a main focus of the current study though).

Taking in account the previous theoretical construction, I hypothesise:

H2: Uncertainty avoidance is negatively correlated with the overall rate of cross-border M&A deals (completed or non-completed);

Moreover,

(21)

20

In sum, previous studies have shown the existence of behavioural finance phenomena in

managerial decisions concerning international M&As investments. Furthermore, these manifestations lead to creation of potentially significant business costs. Major effects have been estimated in the case of persistent preference for domestic markets (home bias), and foreign under or over-diversification in international M&As (foreign bias). Cultural influence is one of the major familiarity determinants of such phenomena and in consequence gains primer focus in the current research, rather than a residual position as in most other studies. Finally, it is hypothesised that individualism and uncertainty avoidance will manifest negative correlations not only with foreign distribution of M&A completed deals in the European context, but with foreign bias diversification matrixes as well (Appendix IV: Table C1).

CHAPTER 3 METHODOLOGY

The methodological chapter will begin with a first section that outlines the sample under study. The sample is based mostly on secondary data from electronic databases, manipulated into a desirable grouping and format. Measure units and time dates are correlated so as to have statistical sense. After offering a thorough description of data and an explanation of the variables by choice and construct, the second and last section will develop the model construct, culminating with the description of the equations on M&A foreign distribution and foreign bias effects, which are going to be used in the present research.

3.1 DATA SAMPLE AND STATISTICAL APPROACH

As previously stated with several occasions, the focus of this country study is situated along the cultural influences manifested in the decision making process of M&As, within the European context. The time range, a variable that most studies so far exploring phenomena related to home and foreign bias tend to ignore, is spanning the period of the last decade, from 2000 to mid 2010. The first two quarters of year 2010 have been introduced to the analysis so as not only to take in account the most recent period possible for up to date meaningful expected results, but also given the fact that in the last period of the decade the volatile economical environment predicts a significant amount of contextual changes in a short periods of time. Along this time axis, numbers are gathered concerning the deal value of completed mergers and acquisitions between the countries of reference, collected on a

quarterly interval, from 1st January 2000 till 1st of July 2010 (the beginning of the 3rd quarter

(22)

21

non-available values were further filtered and subject to elimination so as not to bias expected results. The range of countries comprise of 19 entities, all from European region (inside and outside of the EU), covering an expansive array of cultural diversity. The companies chosen to be taken in account are part of the large industry group of financial services, expecting higher relevance and occurrence within this area of performed international and national M&A deals. The industry choice was based on the statistical classification of economic activities NACE Rev 2, the latest system provided by Eurostat, the statistical office of the European Communities (Eurostat, Nace Rev 2, European Commission). The chosen groups that were taken for analysis were: 64 (Financial service activities, except insurance and pension funding), 65 (Insurance, reinsurance and pension funding, except compulsory social security) and 66 (Activities auxiliary to financial services and insurance services).

The focal point was choosing the acquirer’s country of origin, and identifying all the M&A deals, of the given time period, concluded with the array of the 19 countries of the sample of choice. An important remark, made priory in hypothesis construction, is that the data base consists of completed deals, so not the intent but rather the result of engaging in M&As is under research. The sample of 19 countries has been built for covering the intended range of diversity from a cultural point of view. The decision has been made according to the classification of countries among cultural clusters, a project started by the GLOBE initiative (performed on 62 cultures) and based in a major part on Hofstede’s research up to that point (House, Hanges, Javidan, Dorfman and Gupta, 2004). Extensive research has been conducted to verify the validity of the 10 general country clusters and the results have proven that the categorization is a quite solid and reliable method (Gupta, Hanges and Dorfman, 2002). The following table shows the results that the GLOBE initiative has obtained concerning the cultural clusters, adapted to the current context (Table 1).

Anglo (European) Latin Europe Nordic Europe Germanic Europe Eastern Europe

England France Denmark Austria Poland

Ireland Spain Finland Germany Russia

Italy Sweden Netherlands Greece

Portugal Hungary

Table 1 European Cultural Clusters by the GLOBE project

(23)

22

The clusters from this study all belong to the European context. Belgium has been considered additionally to the Germanic cluster due to the half existing Dutch speaking region that economically is more developed and dominant in comparison to the French one (at this point in time). Moreover preliminary studies have associated similar behaviour in M&A trends between Belgium and the Germanic cluster. Romania is additional to the Eastern Europe cluster, due to the historical and cultural appurtenance (except linguistic) and is chosen due to the important amount of FDI received during the last decade (UNCTAD). Norway has been added to the Nordic Cluster based on the belonging to the historic cultural region of Scandinavia. Thus the 19 chosen countries are distributed culturally as follows: Anglo cluster (Great Britain - GB, Ireland - IE), Latin European cluster (France - FR, Italy - IT, Portugal - PT, Spain - ES), Nordic European (Denmark - DK, Finland - FI, Sweden -SE, and Norway - NO), Germanic European (Austria - AT, Germany - DE, Netherlands - NL, and Belgium - BE) and Eastern European (Greece - GR, Hungary - HU, Poland - PL, Russia – RU, and Romania - RO). The cluster separation can be seen as useful in indentifying a measurement of cultural distance between the chosen countries under study and assures needed cultural variation.

3.1.1 Dependent variables

(24)

23

resource distribution for M&A deals, of that particular situation. The variable is calculated in consequence as a weighted average and it formulates the first dependent variable.

C ijt

= INV

ijt

/ ∑

j=1 19

INV

ijt

The rational investor theories mentioned above push towards an optimal allocation of resources and optimal weights that a country should manifest towards another. These weights can be defined according to the market value of a country j in comparison to all other markets of the chosen sample, at time t. Being a closed system, solely the 19 countries under study will form the pallet of choice that one particular state has for investing. The market values in this case are defined as gross domestic product values (forwardly mentioned as GDP). Thus:

C̽

jt

= GDP

jt

/ ∑

j=1 19

GDP

jt

The difference between the actual and the optimal distribution of resources in M&A deals, at a given moment, represents thus the foreign bias of a country manifested towards another. This gives birth to the second dependent variable under study, and will be computed as the following equation illustrates:

FB ijt

= log

10

(C

ijt

/ C̽

jt

)

The variable is constructed based on the discrepancy between actual investment weights (Cijt)

and the optimal theoretical weights that a country should allocate for another, at time t (

C̽

_jt).

(25)

24

3.1.2 Independent variables

The independent variables are the two variables connected directly with the hypothesis development, respectively individualism (IND) and uncertainty avoidance (UA). Quantifying and measuring cultural related phenomena, as previously stated, is quite a challenging and relative task per se, and researchers usually agree to disagree on such subjective matters. In the effort to keep on track with the most often and efficient used techniques so far (Chakrabarti, Jayaraman and Mukherjee, 2004), I choose to adopt Hofstede’s measurements for these particular cultural dimensions and develop consequently the independent variables to be used in the study (Appendix I: Table A2).

From the values shown in the table, the most individualistic countries are United Kingdom (89) and The Netherlands (80), with an average leading Anglo or Anglo-Saxon cluster (79.5), and the lowest values are in Russia (39), Greece (35) and Romania (30), part of the Eastern European cluster (that has an average of 48.8). The highest uncertainty avoidance scores can be identified in Greece (112), Belgium (94), Romania (90) and Poland (93), with a leading Eastern European Group (94.4 on average), and lowest values in Denmark (23), Sweden (29) and the United Kingdom (35), with lowest average in the Anglo-Saxon cluster (35).

3.1.3 Control variables

(26)

25

Cultural Distance: Clustering effects One of the most commonly chosen methods of

computing cultural effects is to calculate a form of cultural distance, which encompasses all differentials at national cultural level between the home and host country (in this case to control for phenomena unrelated to individualism and uncertainty avoidance). Cultural distance is mostly perceived as to enhance value destruction and discourage M&A activities (Aybar and Ficici, 2009; Majidi, 2007). Indexes of cultural measures (one of the most used being the cultural distance index of Kogut and Singh based on Hofstede’s cultural dimensions) lead often to inconclusive results and tend to annul effects of individual cultural factors when computed in one singular variable. Furthermore, some recent opinions incline to believe that cultural distance is not axially symmetric and does differ if the focal point and destination switch among themselves (Lee, Shenkar and Li, 2008). In such a given situation, I incline to formulate cultural distance as having only a controlling effect (apart from the independent variables), and will use solely the appurtenance or not to clusters developed by the GLOBE project, described above. Clusters represent a group of countries with a common cultural basis. I develop thus the dummy variable Cl which will indicate with a value of 1 if the acquirer targets a country in the same cluster as its own, or 0 otherwise. This variable will control for the preference of countries to orientate towards proximal cultures when developing economically abroad. Consequently it is expected to have a negative impact on M&A foreign distribution and a positive impact on foreign bias indices. The effects are expected to be rather small given the recent effort of the European countries in the last decades to converge, economically at least.

Home bias behavioural developments Home bias is a persistent and intensive behavioural

(27)

26

has on foreign bias. Optimal foreign asset allocation is virtually impossible if the resources are in a great majority restricted to local investment. In the case of foreign bias, funds allocated for external M&As consequently overall decrease, and by this correlation it is expected that foreign bias will actually decrease alike (Beugelsdijk and Frijns, 2010). In such circumstance a control variable for domestic bias is a natural and very much necessary inclusion to the model. I choose thus the dummy Home, which takes values of 1 if the M&A at time t is performed with a domestic company and 0 if the deal is performed abroad.

Market attractiveness: GDP growth The push of globalization, a phenomenon increasingly

intensified during the last years’ economical evolution, has urged companies to search more and further for perfect opportunities that would assure a high level of qualitative service, a much needed flexibility for an increasingly competitive international environment, the leverage to keep the so much needed one step in advance, and a continuous desire of lowering costs and maximising profits, among many other considerations. From a strictly international financial perspective, when looking for opportunities abroad, managerial decisions do need to take in account a rigorous evidence of a country’s development, its stage in the economic cycle and the overall economic status, and such information is commonly associated with the growth rate of the GDP (Tripodo and Dazzi, 1995). In these circumstances M&A deals’ choice of orientation may be influenced by the host country’s GDP growth within a time period, and this variance is described by GDP growth control variable from the data sample. It is expected that the foreign bias effect would be intensified in favour of countries with a higher GDP growth and accordingly investment spread diminished by decrease in available funds. The variable is computed quarterly based on monthly evolutions taken from the International Monetary Fund (IMF) Statistical Data. Missing data was completed by assuming previous or future values for the last or first month when such numbers were indeed available.

Regional economic blocks: the European Union impact International, as well as domestic

(28)

27

M&A from countries outside its borders, wanting to gain access to this vast pool of potential opportunities. Taking in consideration these “friendship” effects (Beugelsdijk and Frijns, 2010), the resources allocated for foreign countries are expected to increase, but the overall foreign bias is also likely to gain value since the efforts are expected to be concentrated solely at EU level, rather than in other European countries as well. The control variable EU will take value 1 if the host country is part of the European Union and 0 if not so. The variable’s values are updated within the sample in pace with the official expansion of the EU during the last decade. For example, although Poland is not part of the EU in 2002, this changes in the first quarter of 2004, when it was accepted in the regional economic and political block.

Familiarity: Geographical Distance So as to control for all aspects of familiarity, prior

research suggests geographical distance and language similarities (Beugelsdijk and Frijns, 2010). Language similarities do overlap mostly cluster grouping (with some exceptions where Romanian language has a grater appurtenance to the Latin European cluster than to the East European, or the isolation of the Greek language), and are reduced in importance in case of M&As, since corporate language in most companies in Europe has formally been declared as English. Geographical distance, although not very significant within Europe’s geographical context, can be though expected to hinder the intent of engaging in M&A activities by enhancing the perception of other differential variables, such as culture (increase biases). In consequence I choose GEO dummy variable to take the following values: 0 if the country of investment coincides with the target country or it is within the same cluster (most of times clusters define an area of geographic proximity: Appendix I - Fig. A3), 1 if the country is a neighbour but nor part of the same cluster, and 2 if the countries are not geographically or culturally adjacent.

Time changing factors and trends As a new development to the row of studies that focus on

behavioural manifestations in asset allocations and the cultural effect, time will be a variable taken in account. The data is gathered in groups according to year and quarter of deal completion, so as to predict minimal possible trends in foreign bias and distribution patterns throughout the last decade. The surge in M&A deals at the beginning of the decade in developed countries and towards the end in developing countries, followed by the global financial downturn are expected to influence M&A activities at European level. The variables

YEAR and QUARTER are proposed. In early and preliminary tests the impact of quarter

(29)

28

particular year there is a favourable quarter for performing M&A deals. Although countries individually may manifest such “traditions”, in the overall sample there is shown no relevant significance and the variable QUARTER is dropped. The remaining time variable, YEAR, is expected thus to show a decrease in foreign bias and a more intensified resource allocation as it progresses further.

3.2 DEVELOPING THE EQUATIONS

Predicting the hypothesis development there are two equations that will interpret by OLS regression the pattern of distribution of asset allocation in the performing of M&A deals across the given sample and the foreign bias effects, all under the direct influence of individualism and uncertainty avoidance. In the following, IND and UA are the independent variables from Hofstede’s study, i is the country of origin and j is the target country, while t signifies the moment in time.

Equation 1 (M&A foreign distribution):

C ijt = α + β * IND i + γ * UA i + δ * Cl j + ζ * Home j + η * GDP growth jt + θ * EU jt + ι *

GEO j + κ * YEAR + ε ijt

Equation 2 (foreign bias effects):

FB ijt = α + β * IND i + γ * UA i + δ * Cl j + ζ * Home j + η * GDP growth jt + θ * EU jt + ι

* GEO j + κ * YEAR + ε ijt

As a conclusion, the coefficients are to predict the impact of each independent and control

(30)

29

CHAPTER 4 FINDINGS

The current chapter begins with a first section which presents the results of the previously explained model, expressed as output of the two regressions ran. In the second part a robustness analysis of the model and alternative approaches to data interpretations are taken under study, alongside statistical limitations and additional details that will be explained. The latter part of the chapter will be dedicated to the interpretation of the results discovered and will try to further give a sustainable explanation for the specific details discovered based on potential solutions and existing literature.

4.1 RESULTS

In this subsection the results of running the two regressions are going to be presented and a brief interpretation will be given. The independent and control variables will be introduced all at once as in the equations described above and not in a sequential manner (multivariate analysis). This is due to the fact that all secondary effects of the variables under study will accordingly be controlled for and results are expected to be statistically superior in an aggregate format. Univariate analysis will only be shown in the case when additional explanatory power might be needed for the already obtained results. The statistical program used for performing the two regressions is E-views 7.

4.1.1 International Distribution of M&A dedicated resources

(31)

30 Type of variable Variable name Predicted Sign Actual Sign t-statistics Coefficient

and significance level

Independent IND i - - -6,856264 -0,377964* Independent UA i - - -1,307025 -0,040717 Control Cl - - -0,844209 -0,089071 Control HOME - - -18,02267 -0,299309* Control GEO - + 2,867957 0,139974* Control GDP growth j - - -1,411967 -0,152876 Control EU + + 4,051231 0,111586* Control YEAR + - -1,449335 -0,062678

R-squared: 0,486136 Mean dependent var: 1,442172

Adjusted R-squared: 0,482956 S.D. dependent var: 0,346746 S.E. of regression: 0,249330 Durbin-Watson: 2,243834 F-statistic: 152,9035

Prob(F-statistic): 0,000000

Table 2 Results for Equation 11;

Individualism, as the first of Hofstede’s couple of variables to be considered as an independent factor (here IND i), has a statistically significant negative influence on the dependent variable, thus validating hypothesis number 1 (H 1). On the other hand, the second cultural variable, uncertainty avoidance (UA i), although presenting a correspondent sign with the one predicted by hypothesis number 2, does not show sufficient statistical significance. Univariate analysis of uncertainty avoidance on resource distribution of M&A shows a significant and surprisingly positive impact. Nevertheless, at aggregate level of the regression performed on equations 1, statistical significance could not be proven and consequently it can only be concluded that H 2 is not sustained by the constructed model.

Focusing on adjacent control variables, it can be observed that the clustering or cultural distance effect in this case (Cl) does not seem to have a significant impact on the distribution of M&A resources within European companies. On the other hand, the HOME variable,

1

(32)

31

controlling for domestic bias manifestations has a very strong negative impact in resource allocation, being one of the strongest significant variables from the equation analysis. GEO or geographical distance does contribute unexpectedly significantly positive on the dependent variable. On the other hand, so does in this case the EU factor (or EU membership). The GDP growth of the host country (GDP growth j) and the time distribution (YEAR), although rather close to show an influence, have no apparent impact on the dependent variable in the chosen context of statistical significance.

4.1.2 Foreign Bias effects in M&A deals

In a similar manner as to equation 1, the output of equation 2 is presented in Table 3, concerning the impact on the foreign bias and corresponding manifestation in M&A deals, performed within the contextual European setting, in the last decade. After minor data adjustment procedures on the sample of 1302 observations (that will be further explained in the following section), the equation’s output was, as previously, checked for all levels of statistical significance and meaningful basic interpretations are made.

The interpretation of the data goes as following. Individualism presents once more a negative significant expected impact on the dependent variable, as well as uncertainty avoidance has a similar and very much so significant influence (validating Hypotheses 3 and 4). The single control variable with no apparent proven significance in the whole methodological model is the timing factor, expressed through YEAR. However, in the univariate analysis, even this particular variable manifests moderate statistical significance for the second equation, with an expected negative impact on the foreign bias element. Furthermore, in the case of the second equation, GEO (geographical distance) has no significant impact as well, although as the time variable before, in univariate statistical approach it manifests a strong significant positive influence.

(33)

32 Type of variable Variable name Predicted Sign Actual Sign t-statistics Coefficient

and significance level

Independent IND i - - -11,87790 -0,304927* Independent UA i - - -2,945082 -0,042725* Control Cl + + 2,814483 0,138286* Control HOME - - -11,51096 -0,089024* Control GEO + - -1,026147 -0,023323 Control GDP growth j + + 2,417729 0,121903** Control EU + + 7,386793 0, 094748* Control YEAR - - -1,354273 -0,027274

R-squared: 0,435578 Mean dependent var: 0,403626

Adjusted R-squared: 0,432085 S.D. dependent var: 0,154073 S.E. of regression: 0,116110 Durbin-Watson: 2,044607 F-statistic: 124,7297

Prob(F-statistic): 0,000000

Table 3 Results for Equation 22;

4.2 ROBUSTNESS ANALYSIS

Before structurally analysing more in debt the model presented in the previous chapter, the focus will shift to a brief description of the basic characteristics of the chosen sample. The table below (Table 4) will summarise the descriptive statistics of each variable apart.

_Cijt _Fbijt _IND _UA _Cl _EU _{GDP gr.} _GEO _HOME _YEAR

Mean 53% 0.99 68.17 65.19 0.64 0.82 2.18 0.66 0.48 2,005 Median 55% 1.18 71.00 65.00 1.00 1.00 2.44 0.00 0.00 2,005 Maximum 100% 2.88 89.00 112.00 1.00 1.00 19.68 2.00 1.00 2,010 Minimum 0% -4.40 27.00 23.00 0.00 0.00 -10.94 0.00 0.00 2,000 St. Dev. 42% 1.14 15.52 24.80 0.48 0.38 3.31 0.88 0.50 2.90 Table 4 Descriptive statistics variables’ analysis;

2

(34)

33

Even though having no inference strength for the whole population, descriptive statistics may present some relevant traits of the sample used and important interpretations may be given. The table above shows some basic central tendency and dispersion characteristics. For example, in the case of resource distribution for M&A deals, one may observe a mean and median above the average value between the maximum and minimum (above 50%). This makes reference to the fact that foreign distributions may have begun the process of receiving more attention. The conclusion is sustained by a home bias variable with a mean slightly below 0.5 (mid value between its maximum and minimum). From another point of view the standard deviation of Cijt is quite high (42%), and this is probably due to the dominant positions of no investment in a particular country (0%) or the other way around, complete investment in a moment in time in a specific location (100%). The central tendencies of the independent variables (individualism and uncertainty avoidance) show that overall within the sample there are quite high scores in both cases, which may prove a smaller than expected variance between the European cultural clusters, at least from these points of view. The cluster and EU membership values show as predicted a general preference towards investing in proximal cultural backgrounds and in the same time the dominance of EU appurtenance. During the time frame chosen, the EU has enlarged and ended up containing a great part of the sample (in 2010, all except Norway and Russia). As a final mention, the geographical distance mean differs significantly from 0, making reference to a grater than expected number of deals performed regardless of geographical distance.

(35)

34

cultural distance), there is a multicoliniarity effect of 0.948658. This is above the 0.8 average value proposed by Newbold’s studies of coliniarity (Guerard Jr. and Clemen, 1989), and thus it has been addressed accordingly. However, the happening is rather naturally expected since both variables measure a similar type of distance. If a country chooses to invest in its own cultural cluster, it usually means it is investing quite close by, geographically wise, since cultural clusters (as an observation stated above in theoretical background) tend to form naturally within a proximal area. Regressions for both the cases, when Cl or GEO variables were sequentially excluded, were performed, and no significant modification in results has been observed (although there is no case where both variables were proven statistically significant simultaneously). Thus I tend to safely exclude the problem of multicoliniarity between the two variables. The coefficient of determination (R square: Table 2 – 0.48 and Table 3 – 0.43) shows in both equations an impact of the independent and control variables on the dependent variables, being above 0.3. Nevertheless, having values below 0.75, it makes reference to the fact that additional factors and different combinations may prove even a stronger influence than the system described so far. Although the scatter diagrams of both equations do not show strong evidence of heteroskedasticity, White tests have been performed in both cases and confirmed existing traces (Appendix II). Efforts like data restructuring and eliminations so to eliminate such effects have not succeeded in completely eliminating the phenomenon.

(36)

35

primarily constructed for estimating a quarterly analysis, although this particular variable has been dropped due to little relevance and no significant effects in the preliminary tests. Given this particular format and the focus on inter-country relations only when such relations did exist, there has been no direct potential harming effect on the analysis due to non-available data or persistent “0” values.

4.3 DISCUSSION AND INTERPRETATIONS

The above mentioned results of the study validate three hypotheses and reject the remaining fourth. Although theory usually inclines to support uncertainty avoidance as having a more important impact culturally-wise, only one of the related hypotheses is backed up by the current statistical model, while the individualism-related ones both hold. More exactly individualism is shown as having a negative impact on both foreign bias and M&A foreign distribution (Hypothesis 1 and 3) and uncertainty avoidance with a similar negative effect on foreign bias (Hypothesis 4). However, uncertainty avoidance is not supported as being negatively correlated to M&A foreign distribution (Hypothesis 2). An array of control variables have been shown to have a significant influence alike, and, in the following paragraphs, relevant connections will be drawn between the results that the sample imply and the theoretical implications that these could have.

(37)

36

correlated to the stabilising number of cross-border M&As that recently occurs as a trend in Europe (Ietto-Gillies, Meschi and Simonetti, 2005). The second hypothesis taken in account so far that connects individualism with lower levels of foreign bias, can be correlated furthermore with Great Britain’s leading level in Europe’s market for M&As, having better diversified portfolios and more important levels of synergetic effects, since UK has the highest level of individualism from the country sample (Appendix I: Table A2).

In the case of Hypothesis 4, uncertainty avoidance does appear to have a strong negative impact on foreign bias, due in great probability to the lack of foreign M&A operations based on weak risk-taking behaviour. Making a connection of the finding with the existing data sample of countries, Easters European cluster countries do indeed record considerably less foreign M&A activities, whilst having in the same time the highest levels of uncertainty avoidance (Table 2). Unfortunately, this finding cannot be related with a better diversification of M&A activities per se, since domestic bias is still a very much existing strong behaviour in the current case.

(38)

37

Another adjacent perspective can be formed when paying attention to secondary variables of the study that may demonstrate connections worth taken in account. A very strong expected influence within the control variables is given by the home bias factor. It seems that indeed companies from a particular state do manifest a clear preference for the economic and cultural environment in which they were born and developed so far. This does lead to a statistically significant perturbation in foreign mergers and acquisitions (a decrease in calculated foreign bias), and in a correspondent decrease of foreign investment and M&A deals per se. The persistent existence of home bias is associated by some researchers with the development and balance between two parallel phenomena. On the one hand worldwide financial openness increases and that is a downsizing bias contributor but on the other hand, the necessity and capacity of information increases exponentially alike, which although engaging companies in investment opportunities constantly, represents an additional constant advantage of the home market (Mondria and Wu, 2006). The decrease in foreign bias effects, as expected, is most likely a signalling of decreased available funds for external use. As Merkel suggested in his article “In defence of home bias” (2010), it can also be interpreted as being indeed economically beneficial for companies to invest proportionally more in their own countries. This can thus be a probable factor that led to the survival of home bias for such a long period of time, with little evidence of significant decrease.

M&A foreign investment decisions

RIJKSUNIVERSITEIT GRONINGEN

FACULTY OF ECONOMICS AND

BUSINESS

UPPSALA UNIVERSITETET

FACULTY OF SOCIAL SCIENCES

M&A foreign investment decisions

-A study of cultural influence in foreign bias effects-

Master Thesis

MSc IB&M – International Financial Management

Programme Coordinator and Thesis Supervisor: Dr. W. Westerman

MSc Student: Lucian Pătulea

ABSTRACT

TABLE OF CONTENTS

CHAPTER 1 RESEARCH OUTLINE

CHAPTER 2 THEORETICAL CONSIDERATIONS

CHAPTER 3 METHODOLOGY

C ijt

= INV

/ ∑

INV

C̽

= GDP

/ ∑

GDP

FB ijt

= log

(C

/ C̽

)

C̽

CHAPTER 4 FINDINGS