Resources, Location, and Subsidiary Structure on
the Performance of Emerging Market Firms: Case
of Poland
Master Thesis International Economics and Business
University of Groningen
Faculty of Economics and Business
Coco de Braconier
s2404427
c.de.braconier@student.rug.nl / cdebrac@gmail.com Supervisor: dr. Raquel Ortega Argilés
2
Abstract
3
Table of Contents
1. Introduction ... 4
2. Literature review and hypotheses ... 8
2.1 International diversification of firms from emerging markets ... 8
2.2 Resources and firm performance ... 9
2.3 Location and firm performance ... 13
2.4 Subsidiary structure and firm performance ... 14
3. Methodology ... 16
3.1 Sample and data collection ... 16
3.2 Dependent variable ... 16 3.3 Independent variable ... 17 3.4 Control variables ... 18 3.5 Model ... 19 3.6 Assumptions ... 21 3.6.1 Normality ... 21 3.6.2 Multi-collinearity ... 22 3.6.3 Homoscedasticity ... 23 3.6.4 Autocorrelation ... 23 4. Results ... 24 4.1 Descriptive statistics ... 24 4.2. Regression results ... 25 4.3 Robustness checks ... 31 5. Discussion ... 33 6. Conclusion ... 35
6.1 Limitations and future research ... 35
7. References ... 36
Appendices ... 46
Appendix 1: unbalanced dataset ... 46
Appendix 2: industry descriptions ... 47
Appendix 3: List of host countries ... 48
Appendix 4: Normality ... 49
Appendix 5: multi-collinearity ... 50
Appendix 6: Homoscedasticity ... 51
Appendix 7: Autocorrelation ... 53
4
1. Introduction
Over the last three decades, economic liberalization and globalization have led emerging and developing market firms to diversify across nations (Sahaym & Nam, 2012). As a result, the role of those markets in the world economy has been increasing, and in this way, their share on the world’s output, trade, and capital flows (Ghosh, 1996; Gubbi et al., 2010; Khanna & Palepu, 2010; Hoskisson et al., 2013). Sahaym and Nam (2013) theorize that the drive behind this phenomenon is related to two issues. Firstly, increased support of institutions enabled firms from emerging markets to pursue international diversification (Peng et al., 2008). Secondly, there was a major shift towards exploiting resources and capabilities by firms from emerging markets (Sahaym & Nam, 2013). A well-known example of a firm that prospered as a result of the phenomenon’s described by Sahaym and Nam (2012) is Samsung. Where government supported the Korean labour force by providing education to some of its workforce and Samsung gained competitive advantages by combining the cheap but highly skilled educational labour force (Magaziner & Patinkin, 1989).
Increasing growth and firm performance are motives for firms to pursue an international diversification strategy (Capar & Kotabe, 2003). The Resource-Based-view theory argues that in order to successfully pursue an international diversification strategy firms need resources superior to their competitors. However, firms from emerging markets lack resources compared to firms from developed markets (Khanna & Palepu, 2006). Previous studies have examined the relationship between international diversification and firm performance, most of these studies analyze the phenomenon in developed markets (e.g. Tallman & Li, 1996, Capar & Kotabe, 2003, and Lu & Beamish, 2004). Therefore, the objective of this study is to measure the performance of firms pursuing an international diversification strategy in the case of emerging market firms.
Although research has been done on emerging market firms (e.g. Kumar, 1985 on Korean firms; Child & Rodrigues, 2005 on Chinese Firms; Contractor et al., 2007 on Indian firms; Elango & Pattnaik, 2007 on Indian firms; and Lin & Wu, 2014 on Taiwanese firms), to my best knowledge, upon till now there is not yet a study which uses a sample of firms from Eastern Europe. There is a growing number of Eastern European countries joining the European Union1. These countries were regarded as “troubled economies” where unemployment increased and economic growth declined. As a result, Western European
1
5
policymakers worried that the poor economic conditions would spillover to their markets (The Economist, 2011). However, these Eastern European countries have rapidly improved growth rates compared to those in Western European. This is illustrated in graph 1 as it shows that growth of GDP per capita in Poland, Czech Republic and Hungary, which are Eastern European countries, is higher than the average growth in the Euro area. During that same period, Eastern European countries also increased their exports, which is illustrated in graph 2.
Graph 1.1: growing GDP per capita in Poland, Czech Republic, Hungary, and Euro area. Source: World Bank Database.
6
Graph 1.2: growing exports of goods and services in Poland, Czech Republic, Hungary, and Euro area. Source: World Bank Database.
This research chooses Poland, as recent globalization and transition of the economy of Poland make it interesting to investigate. In 2004, Poland joined the European Union. Joining the European single market benefits Polish firms by, more cross-border trade, better borrowing, planning and investment, better access to capital and more international trade (European Commission, 2013). Where most European countries were hit by the global economic crisis, Poland was able to keep the damage to a minimum (Velculescu, 2009). In addition, Poland is ahead in terms of GDP and FDI compared to other Eastern European countries (World Bank). Not surprisingly, since Poland is labeled as an emerging market by the International Monetary Fund (IMF). Thus, this research concludes that firms from Poland are very important and interesting to investigate.
Considering, Poland as an emerging market, this research will compare Polish international diversified firms to the ones that are exclusively active in Poland. Thus, the specific research question is formulated as follows: What drives the performance of emerging
7 by the location of the foreign markets where emerging market firms diversify? (3) Does the ownership structure of subsidiaries, located in foreign markets, influence the performance of emerging market firms?
Due to the growing importance of emerging markets and emerging market firms in the global economy it is important to study those firms in order to understand what drives their success in terms of firm performance. Therefore, results of this research can contribute to the growing importance of explaining the relationship of international diversification and firm performance. Since, as mentioned previously, emerging market firms differ from those of developed markets.
8
2. Literature review and hypotheses
2.1 International diversification of firms from emerging markets
International diversification2 creates opportunities for firms (Capar & Kotabe, 2003; Elango & Pattnaik, 2007). However why some firms enter foreign markets, despite having disadvantages regarding the market knowledge compared to local firms, has been a topic since early research on the theory of international diversification (Qian & Rugman, 2013). As a result in disadvantage in market knowledge, cost arise from cultural, political, and economic differences. In addition to the cost of coordination firm’s activities across geographic distance also known as the liability of foreignness (Zaheer, 1995).
Dunning (1980) theorizes that ownership, location, and internalization advantages are the three determinants needed by firms to successfully diversify into foreign markets. In the presence of unique assets, superior to that of competitors, and with the ability of firms to coordinate and control such assets in foreign markets, they possess ownership advantages over foreign firms (Dunning, 1980; Dunning & Rugman, 1985; Rugman & Verbeke, 1992).
Location advantages are related to the firms’ ability to use the benefits associated with
particular geographical locations (Rugman & Verbeke, 1992). While the benefits associated with entry modes of international diversification decisions made by firms are referred to as
internalization advantages (Rugman 1981; Rugman & Verbeke, 1992). In other words, the
likelihood of firms to be successful in international diversification depends on their ability to transfer unique assets to foreign locations through subsidiaries that can benefit from those unique assets. If firms possess such abilities, they can overcome the liability of foreignness.
The characteristics of firms in emerging markets show dissimilarities with those from developed markets (Contractor et al. 2007; Stucchi, 2012). As discussed by Contractor et al. (2007), emerging market firms are typically smaller, have relative less resources and are more distant from major markets compared to their competitors of developed markets. Moreover, whereas developed market firms expand abroad by exploiting resources, emerging market firms diversify in foreign market to access the resources they lack (Mathews, 2006). Thus,
2 As introduced by Capar & Kotabe (2003) international diversification can be defined as a firm’s expansion
beyond the borders of its home country across different countries and geographical regions. This research will
9
scholars have doubted whether traditional theory, such as the OLI3 framework, can still be used to explain the competitive advantages of multinationals (Dunning, 2006).
The view that to diversify successfully in international market firms need to possess unique assets (i.e. ownership advantages) is primarily based on the Resource-Based View (RBV) (Kraaijenbrink et al., 2010). In addition, recent researchers have used the RBV for investigating the relationship between ownership advantages and firm performance (e.g. Capar & Kotabe, 2003; Nath et al., 2010; and Lin & Wu, 2014). Therefore, following the recent trend this research will continue using RBV theory for explaining the relationship between resources and firm performance in section 2.2. Furthermore section 2.2 will also take in consideration aspects such as: the experience gains in exploiting resources and the ability to cope with foreign environments. Similar to the available resources, the selection of markets to diversify in is likely to affect firm performance (Jean et al., 2011). Thus, in section 2.3 the effect of diversifying in attractive markets on firm performance is elaborated. Lastly, this research will also take in consideration the appropriate ownership structure (i.e. wholly-owned or partly wholly-owned subsidiary) for operating in foreign markets. Hence, in section 2.4 the effect of selecting joint venture or wholly-owned subsidiary as a preferred subsidiary structure on the performance of the parent firm is investigated.
2.2 Resources and firm performance
Similar to the ownership advantages of the OLI framework, the RBV is concerned with the question how firm-specific ownership advantages can create and maintain competitive advantage. In this view, the firm-specific ownership advantages are the resources that make firms heterogeneous (Peteraf, 1993). Resources can be defined as tangible and intangible assets and define the strengths and weaknesses of firms (Wernerfelt, 1984). Examples of intangible assets are: brand names, patents, copyrights, goodwill, or firm incorporated knowledge. While on the other hand tangible assets are physically present in a firm such as machinery, buildings, natural resources, or labor. As acknowledged by Dunning (1981), both the tangible and intangible assets are the ownership advantages of a firm, and these advantages make firms competitive in foreign markets. In other words, ownership advantages results from the available resources which create competitive advantage over competitors.
10
However it is important to note that only those resources which are considered unique and hard to copy by other firms have the potential to create competitive advantage (Porter, 1986). The RBV suggest that firms that are able to identify and use those unique resources are more likely to operate successfully in foreign markets (Kraaijenbrink et al., 2010; Peng, 2011). Since unique resources create competitive advantage and competitive advantage increase firm performance (Newbert, 2008). However, for a firm to be successful according to the RBV, it first needs to overcome a threshold of productivity in exploiting their knowledge and resources in domestic market before pursuing any international ambitions (Khanna & Palepu, 2006). Thus, when a firm has developed strong core competences at their home market it can apply these in foreign markets (Bartlett & Ghoshal, 1999). As a result, firms investing in foreign markets are generally bigger, more profitable, and more productive than those not active in foreign markets (Mayer & Ottaviano, 2008). Previous research examined the relationship between international diversification and firm performance. For example, Heyder et al. (2011) found, using a sample of 21 leading European cooperatives in the dairy and meat sector, a significant and positive effect between international diversification and performance. Whereas, Mayar and Ottaviano (2008), who investigated European firms, established that firms that invest in foreign markets outperform those that do not. Nevertheless, there also exists research that could not confirm this significant positive relationship (e.g. Kumar, 1985; and Michel & Shaked, 1986). Due to the combination of theory and empirical results, the following hypothesis is conducted:
H1: International diversification has a positive effect on firm performance.
Once firms are able to perform profitably in foreign markets, they can exploit their resources by expanding operations. To which degree firms are expanding operations in foreign markets, in terms of investing in assets and/or controlling activities, is referred to as multinationality4 (Teece, 1980). Experience in managing subsidiaries generates general knowledge and capabilities which results in lower costs related to diversifying (Tallman & Li, 1996; Wiersema & Bowen, 2011). Thus, by expanding operations at lower costs, firms are provided with the opportunity to exploit the benefits of their resources, and in this way, increase their performance (Kim et al., 1993). In other words, by increasing multinationality firms are able to benefit from economies of scale. Previous studies have found contradicting
4 Multinationality is mostly measured as foreign sales to total sales in previous studies (e.g. Al-Obaidan &
11
results in examining the relationship of multinationality and firm performance. Where researchers have examined linear relationships (e.g. Han & Suk, 1998; and Kirca et al., 2011) as well as non-linear relationships5 (e.g. Tallman & Li, 1996; Hitt et al., 1997; Capar & Kotabe, 2003 and Endo & Ozaki, 2011). Moreover, the results range from a positive effect (e.g. Ramírez-Alesón & Antonio Espitia-Escuer, 2001; and Kirca et al., 2011) to a negative effect (e.g. Al-Obaidan & Scully, 1995, and Riahi-Belkaoui & Alnajjar, 2002), along with no effect at all (e.g. Buhner, 1987). Nonetheless, this research argues that by increasing multinationality a firm lowers the cost associated with the liability of foreignness, i.e. more experience with exploiting resources in a foreign market in order to gain competitive advantage (Hitt et al., 2006). Hence, this research hypothesize that:
H2: The degree of multinationality has a positive effect on firm performance.
As discussed in the literature leading to hypotheses 1 and 2, the emphasis is on the resources that a firm has to its disposal. Firms from developed markets possess specific assets, such as advanced technology, which are not available for most emerging market firms (Khanna & Palepu, 2006; Miller et al., 2008; Denk et al., 2012). However, firms can overcome this disadvantage by developing home-grown resources which are valuable, rare, inimitable, and non-substitutable (Denk et al., 2012). Moreover, intangible assets can reduce costs from liability of foreignness and are used to outperform local competitors (Gardberg & Fombrun, 2006; Yiu & Bruton, 2007). Thus, it is theorized that emerging market firms should possess home-grown intangible assets. Previous empirical literature has recognized intangible assets possessed by emerging market firms that diversify internationally, however, mainly focus on the intangible benefits of strategic alliances (e.g. Hitt et al., and 2000; Hoskisson et al., 2005; and Elango & Pattnaik, 2007). However, emerging market firms should also able to benefit from particular firm specific intangible assets such as research expenses, goodwill, and development expenses. Therefore, this research hypothesize that:
H3: Intangible assets has a positive effect on firm performance
Similar to the increase in a firm’ capabilities through using their resources, as discussed in the literature leading to hypothesis 2, firms gain experience in their ability to deal with unfamiliar environments (Tallman & Li, 1996). By increasing the geographical scope of operations firms are able to benefit through accumulating knowledge from diversifying and applying this knowledge in foreign markets (Qian et al., 2010). So next to multinationality,
12
which is related to the size of foreign operations, firms can increase international diversification by expanding their geographical scope (Porter, 1986). However, it is important to recognize the difference, with increase of multinationality it is expected that firm’s experience with exploiting resources increases, while with increase in geographical scope firms gain in their ability to cope with foreign markets (Kogut, 1985; Tallman & Li, 1996). In other words, multinationality is associated with the experience of exploiting resources at bigger scale whereas geographical scope is related to the ability of using resources in multiple environments. Empirical studies mainly found two different results in examining the relationship of geographical scope and firm performance. On the one hand, researchers found a positive and linear relationship between geographical scope and firm performance (Tallman & Li, 1996; Delios & Beamish, 2001). On the other hand, researchers found a nonlinear relationship where geographical scope enhances firm performance to a certain threshold and then the effect becomes negative (Hitt et al., 1997; Qian et al., 2008). Nonetheless, following the theory discussed earlier, this research argues that there should be a positive and linear effect as firms gain experience in entering foreign environments. As for example, Delios and Beamish (2001), using Japanese manufacturing firms, demonstrated that there is a positive relation between geographical scope and firm performance. In addition, they found that the relation had an even stronger effect in the presence of unique resources. In similar vein, Lu and Beamish (2004) showed that intangible assets are likely to be the essential unique resources fostering the geographical scope of foreign operations and firm performance relationship. The positive relation between geographical scope and unique resources comes from firms their ability in leveraging intangible assets in foreign markets (Wiersema & Bowen, 2011). Thus, firms expanding the amount of foreign markets to operate in should perform better in the presence of intangible assets. Hence, the hypotheses:
H4a: The degree of geographical scope has a positive effect on firm performance. H4b: The interaction of geographical scope and intangible assets has a positive effect on firm
performance.
13
located in each country, even though there is little interaction between them (Lovelock, 1999). However, this research theorizes that increasing both the geographical scope and multinationality should positively influence firm performance. Thus, the hypothesis:
H5: The interaction of multinationality and geographical scope has a positive effect on firm performance.
2.3 Location and firm performance
Similar to the available resources, the selected foreign markets of firms to diversify in, is expected to influence their competitive advantage. As explained previously, by entering a foreign market, firms are confronted with the liability of foreignness. A source of liability of foreignness are the cost resulting from the lack of legitimacy in the foreign environment (Zaheer, 1995). Thus, legitimacy directly influences the internationally diversification decisions of firms (Xu and Shenkar, 2002).
Nowadays there is a growing collection of literature with a focus on the institutional-based view (IBV) (Yiu et al., 2007; Peng et al., 2008; Peng et al., 2009). Where it previously was assumed that institutions did not influence the competitive nature of firms, institutional theory now is insightful for understanding why some firms have competitive advantage over others (Oliver, 1997; North; 2003). The IBV takes in consideration formal and informal institutions, also known under scholars as the rules of the game (North, 1990). According to Scott (1995: 33) institutions are “regulative, normative and cognitive structures and activities that provide stability and meaning to social behavior”. Formal institutions exists of “constitutions, statues, common low, and other governmental regulations” (Pejovich, 2012). While informal institutions are known as the “traditions customs, moral values, religious beliefs, and all other norms of behavior that have passed the test of time” (Pejovich, 2012). As a consequence, firms who know how to “play” the rules of game in domestic and foreign markets are able to gain competitive advantage (Peng et al., 2008). Therefore, the IBV is consistent with the concept that a MNE would be more likely to invest in markets where it is more familiar, thus has more legitimacy, and in this way, reduce the liability of foreignness.
14
confronted with similarities in culture, traditions, customs, and behaviors (i.e. rules of the game). In a study done by Lee and Beamish (1995), the results show that Korean firms investing in developed markets where outperformed by those in developing markets as a result of dissimilarities in environments. Therefore, following the IBV literature, which argues that familiarity within a host environment reduces disadvantages associated with the liability of foreignness, it is more attractive for firms from emerging markets to invest in emerging or developing markets. Thus, this research hypothesis that:
H6: Diversifying in emerging or developing markets (i.e. attractive markets) has a positive effect on firm performance.
2.4 Subsidiary structure and firm performance
After having decided the location of a foreign market to operate in, firms have to select the appropriate ownership structure to enter the market (Erramilli & Rao, 1993; Nakos, & Brouthers, 2002; Nielsen & Nielsen, 2011). Non-equity contractual mode (e.g. licensing), partly-owned subsidiary (also known as joint venture), and wholly owned subsidiary are among the available ownership structures which are commonly used for competing in foreign markets (Hill et al,. 1990). As recognized by previous literature, selecting the appropriate ownership structure for competing in a foreign market is a crucial decision, since performance levels significantly differ among levels of ownership structures in foreign markets (Root, 1994; Woodcock & Beamish, 1994; Brouthers et al., 2003; Brouthers et al., 2008). Due to data restrictions, this research focuses on Joint Venture (JV) and Wholly Owned Subsidiary (WOS) as ownership structures. Furthermore, previous empirical studies mainly investigated the performance of the subsidiaries in foreign markets (e.g. Woodcock & Beamish, 1994; Nitsch, et al., 1996; and Brouthers, 2013). However, as acknowledged by Woodcock and Beamish (1994), it is difficult to obtain valid and reliable data on the performance of subsidiary. Therefore, this research investigates the influence of choosing a particular ownership structure for foreign subsidiaries on parent firm performance.
15
recognized that there are differences in strategies of firms from emerging and developed markets (Filatotchev et al., 2007).
Previous research tend to rely on Transaction Cost Theory (TCT)6 for explaining the appropriate entry mode, which predicts that firms are most likely to select the entry mode from an economical point of view (Brouthers et al, 2003; Boeh & Peamish, 2012; Brouthers, 2013). In other words, a firm will compare the costs and benefits (e.g. investments risk, access to resources, subsidiary control) of alternative ownership structures. In contrary, IBV argues that firms seek legitimacy within a market (Xu and Shenkar, 2002). By engaging in JVs with local firms legitimacy is gained (Rao et al., 2008), and in this way, reduce the liability of foreignness. Although both perspectives are useful in explaining the optimal ownership structure, it is important to recognize them both (Chan & Makino, 2007). To optimize the performance, both costs and legitimacy should be taken into account (Kostovo & Roth, 2002).
Since emerging market firms are more likely to lack resources and knowledge, compared to developed market firms, it is theorized that they have a competitive disadvantage in foreign markets (Contractor et al., 2007). By choosing for a cooperative strategy firms can reduce their competitive disadvantage (Hennart, 1988; Kogut, 1988), since JVs need less investment and firms can benefit from pooling complementary bits of knowledge and resources (Hennart, 1991). In similar vein, Gaur and Kumar (2009) identify the benefits of business group affiliations7 in foreign markets for many emerging market firms, in which group members are committing to collective goals. Thus, choosing JV from a TCT perspective is more likely to benefit emerging market firms. While, from an IBV perspective, JVs are expected to compensate for disadvantages in the liability of foreignness since cooperating with a local partner benefits seeking legitimacy in a foreign market (Makino & Delios, 1996). Therefore, this research theorizes that if emerging markets firms were to diversify into foreign markets, opting for a JV strategy would be more likely to be beneficial for performance. This leads to the following hypothesis:
H7: Choosing JV as ownership structure for subsidiaries has a positive effect on parent firm performance.
6 Transaction cost theory relates to all the costs of participating in a market, such as search and information costs,
bargaining costs, and policing and enforcement costs.
16
3. Methodology
3.1 Sample and data collection
For this research, a sample of Polish firms was obtained from the ORBIS database8, as presented by Bureau van Dijk. The database contains data of 120 million private companies worldwide. This research is interested in the time period between 2003 and 2012. As it is explained in the introduction, during that time period Poland outperformed the average Euro country in terms of GDP and growth, making het an interesting period to investigate. Unfortunately, not all Polish firms could be included since many observation are missing between the time period 2003 and 2012, which would result in an unbalanced sample9. Therefore, this research made a selection of firms for which data was available. Information such as revenue, total assets, intangible assets, industry, number subsidiaries, and firm age are collected on firm-level. On subsidiary-level information such as location and the percentage of shares held by parent-firm in the subsidiary.
In the end, this research obtained a final data set containing 230 Polish firms and a total of 2300 observations during the time period 2003 to 2012 are included. Of the 230 firms, 79 are internationally diversified10.
3.2 Dependent variable
Different approaches have been adopted in previous literature to measure the performance of firms (Capar & Kotabe, 2003, Hult et al., 2008). In a study by Hult et al., 96 articles who make use of performance measurements were examined. Results of the study indicated that, for measuring performance on firm level, most studies used a sales-based measure (44%) or the measurement return on assets (40%). As noted by Capar and Kotable (2003), an asset-based measurement is less favorable when estimating the effect of intangible assets. This research predicts that intangible assets are an important driver behind the performance of emerging market firms. In addition, it is expected that there are considerably large differences in the amount of intangible assets between firms (for example, service vs manufacturing firms). As a result, a sales-based measurement is more likely to be reliable, since it is not influenced by the large difference in assets (Habib & Victor 1991). Thus, this research adopts Revenue, measured as the natural logarithm of a firms’ total revenue in a particular year, as the indicator for firm performance.
8 Link to Orbis database: https://orbis.bvdinfo.com/ 9
See Appendix 1 for the number of observations per year.
17
3.3 Independent variable
According to the hypotheses, firm performance is influenced by six explanatory variables:
International Diversified, Multinationality, Geographical Scope, Intangible Assets, Attractive Markets, and Subsidiary Structure.
The first independent variable is the dummy variable explaining if a firm is internationally diversified or not. International Diversification was measured based on a firm having subsidiaries located outside their home market. The variable was coded as 1 for firms owning one or more foreign subsidiaries in a particular year, and 0 otherwise.
The second independent variable, Multinationality, is measured as the number of foreign subsidiaries in a particular year. Prior research commonly used sales-based measures for Multinationality (e.g., Al-Obaidan & Scully, 1995 and Tallman and Li, 1996). As discussed by Qian (1996), a sales-based measurement does not truly reflect international diversification since it fails to measure the multiplicity of a firm’ ability to operate in a foreign market. Other researchers use other measurements, such as total number of foreign subsidiaries (Lu & Beamish, 2004), ratio foreign to total firm employees (Kim et al, 1993), ratio foreign to total subsidiaries, and ratio foreign to total sales (Contractor et al., 2003). Since the interest of this research is regarded to firms their experience in exploiting resources through subsidiaries, a measurement based on subsidiaries would be adequate. Hence, this research will measure Multinationality as the total number of foreign subsidiaries in a particular year. As this research theorizes that the total number of foreign subsidiaries represents to what extent that firm is able to operate in foreign markets.11
No previous research, similar to this one, has been found using the total amount
Intangible Assets. Previous studies, examining international diversification of emerging
market firms, mainly focus on a firms’ expenditure on research and development (R&D) (e.g. Yiu et al., 2007; and Filatotchev et al., 2007). However, due to data limitations on R&D expenditures, for firms included in the sample, this research uses the total amount of
Intangible Assets owned by a firm in a particular year. Orbis defines Intangible Assets as
“assets such as formation expenses, research expenses, goodwill, development expenses and all other expenses with a long term effect”12
. As such, using Intangible Assets, as defined by
11
This research essentially tried to use the ratio of foreign subsidiaries to total subsidiaries in a particular year as a measurement of multinationality. However, due to high correlations with the variable international
diversification the assumption of multi-collinearity in section 3.6.2 was violated.
18
Orbis, is justified since it includes R&D expenses and more. Furthermore, the independent variable of Intangible Assets is calculated as the natural logarithm of total intangible assets to prevent outliers13.
The fourth independent variable measured the Geographical Scope. Similar to previous research (e.g. Tallman & Li, 1996; and Delios & Beamish, 2001), Geographical
Scope is measured as a continuous variable to proxy for the number of foreign locations
where firms own subsidiaries.
As discussed previously in the literature review leading to the fifth hypothesis, investing in emerging or developing markets is expected to benefit emerging market firms. Emerging and developing markets are from now on referred to as Attractive Markets. Thus, the variable Attractive Markets is related to markets where firms own foreign subsidiaries, and in this way, affect their performance. Operating in markets with similar characteristics increases firm performance as it lowers the liability of foreignness (Garg & Delios, 2007). Hence, operating in developing or emerging markets should benefit the sample of Polish firms, and thus, are considered as attractive. The measure for economic development of a market is identified using the OECD standard. Previous research commonly use the OECD standard as a measurement for the level of economic development (e.g. Buckley et al., and Gubii et al., 2010). Where countries which are considered as developed according to the OECD standard are not attractive for firms included in this research. This research uses a continuous variable to proxy for the ratio of foreign subsidiaries in attractive markets compared to total foreign subsidiaries.
Lastly, this research includes an independent variable to proxy for Subsidiary Strategy. As discussed in the theory leading to hypothesis 7, firms opting for a JV Subsidiary Strategy are predicted to perform superior to those using WOS. As such, a continuous variable computed as the ratio of JV to total subsidiaries is included. This current study distinguishes a JV subsidiary in which parent firms own less than 95% of the shares.
3.4 Control variables
In addition to the independent variables, this research includes control variables to confirm the relationship between the dependent variable and independent variables. The first control variable relates to the size of the firm since bigger firms are more likely to possess
13
19
resources to exploit, and in this way, improve performance (Capar & Kotabe, 2003). As frequently done in research, Firm Size is measured as the natural logarithm of total assets of a firm in a particular year (Chitoor et al, 2009).
Second, this research will include a control variable relating to firm’ Experience. The general Experience of a firm matters as it influences the firms specific knowledge incorporated in the firm (Gubbi et al. 2010; Chittoor et al., 2009). Similar to the theory of multinationality, more experienced firms are able to seize opportunities to exploit resources, and in this way, increasing performance (Kin, Hwang & Burger, 1993). However, whereas multinationality is measured based on a firm’ subsidiaries, firm Experience is measured by the number of years since incorporation in a particular year14.
Third, this research will include a series of dummy variables to control for Industry effects (Contractor et al., 2002, Gaur & Kumar, 2009). Firms are grouped into ten industry sections according to the NACE Rev. 2 classifications of Eurostat (see Appendix 2 for detailed information on the 10 industry sections).
Fourth, a dummy variable to control for the host countries of all the foreign subsidiaries is included. To do so, this research divided all the host markets of the firms their subsidiaries into three categories: (1) Western, (2) Eastern, and (3) Others (see Appendix 3 for which countries are located in the different categories). The dummy variable was coded as 1 in the category in which a firm owns its largest share of foreign subsidiaries. By doing so, this research hopes to control for the effect of mainly diversifying in a particular set of markets.
Lastly, a series of dummy variables are included to control for yearly period effects. Each year included in this research (2003 to 2012) has its own Year dummy variable, coded as 1 in that particular year and 0 otherwise.
3.5 Model
As explained in the introduction and sample collection, data ranges from 2003 to 2012. Therefore, this research uses panel data as it combines both cross-sectional and time-series. Previous research that investigated the diversification performance relationship mostly
14 The variable Experience contains outliers. However, using the winsorized variable Experience did not change
any of the results in any of the models. Therefore, this research continued using the non-winsorized variable
20
included samples of multiple years (e.g. Tallman & Li, 1996, Capar & Kotabe, 2003; Gozzi et al., 2008; and Endo & Ozaki, 2011), thus, panel data analysis appears the obvious choice.
Next, this research has to decide on the appropriate technique to analyze panel data; i.e., Fixed Effects (FE) or Random Effects (RE). The FE model is beneficial when interested in tracking variables that vary over time. The advantage of the model is that it can explore the unique attributes of individual units that are time invariant (Clarke et al., 2010). Since, FE controls for the time invariant variables that are correlated with the error term, and in this way, individual differences are captured by differences in the intercept parameter (Wooldridge, 2002; Hill, et al., 2012). In other words, variables, relating to firm characteristics, in the dataset may not differ over time. By omitting those variables which do not differ in time, the model captures the real effects on firm performance. Yet, using FE also has its disadvantages. First of all, as already discussed, it does not allow the estimation of time-invariant variables. A second disadvantage is that FE is inefficient in estimating the effect of variables that have little variance within (Plumper & Troeger, 2007).
In contrary to the FE model, the RE model does not control for the time invariant variables. This model assumes that variation between individual units are random and uncorrelated with the variables. Thus, in the expectation that random individual differences may influence the dependent variable, random effects model is preferred. In addition, the random effects model has a greater precision in its estimations of the independent variables, thus, is a more attractive model to use. However, endogeneity is a potential problem when using the random effects model (Hill, et al., 2008).
21
results can be interesting as it allows us to observe if results are due to unobserved heterogeneity.
By using Ordinary Least Square (OLS) regressions, this research will investigate the relationship between the firm performance and the independent variables leading to the hypotheses that have been discussed. As a result of the dependent, independent and control variables, the following formula is constructed:
( ) ( ) 3.6 Assumptions
Before being able to run quantitative models this research has to justify the use of regression analysis in order to be certain the results are trustworthy (Osborne & Waters, 2002). This research will discuss the assumptions of: (1) normality, (2) multi-collinearity, (3) homoscedasticity, and (4) autocorrelation.
3.6.1 Normality
22
Lumley (2002) state that with large data sets, such as used in this research, the residuals average around the mean, and thus, are normally distributed.
3.6.2 Multi-collinearity
In the situation where several independent variables are correlated, the problem of multi-collinearity arises. Correlation between independent variables indicates that one can be linearly predicted from others. As a result, one of the assumption for the linear regression is violated since it creates instability in the coefficients, and in this way, estimations of the model are less accurate (Baltagi, 2008). Some degree of correlation is acceptable, however, when variables are perfectly correlated than one needs to be excluded from the regression. This research uses the variance inflation factor (VIF), displayed in table 3.3, to diagnose the degree of multi-collinearity. There exists some disagreement among academics about the maximum VIF value a variable is allowed to take. Most commonly, the rule of thumb is that values exceeding 10 need further investigation (O’brien, 2007). However, some others get concerned with values exceeding 2.5 (Allison, 2012). Inspection of the VIF values of the independent variables show that the variable of International Diversification has a value of 2.93 while the value of Geographical Scope is 2.56, which could be indicated as troublesome. However, since International Diversification is a dummy variable a high VIF value acceptable (Allison, 2012).
Table 3.1: Variance Inflation Factor
Variable VIF 1/VIF
International diversification Geographical scope Attractive Markets Multinationality Intangible assets Subsidiary strategy 2.93 2.56 2.19 1.98 1.29 1.06 0.341598 0.390175 0.456922 0.506326 0.774151 0.939844 Mean VIF 2.00
23
3.6.3 Homoscedasticity
The assumption of homoscedasticity is whether or not the variance of the residuals show similarities. A violation of this assumption implies that a model is not well-fitted, and in this way, the variance of the residuals are not constant; meaning that they are heteroscedastic (Hill, et al., 2012). The null hypothesis of no heteroscedasticity is tested applying the robvar for testing group wise heteroscedasticity. Since this research deals with both cross-sectional and time series, it is beneficial to test for groupwise heteroscedasticity (Baum, 2006). The end results of the test, shown in table 3.4, clearly rejects the null hypothesis of no heteroscedasticity. In order to deal with the problem of heteroscedasticity, this research will run the models using robust standard errors.
Table 3.4: Robvar test for groupwise heteroskedasticity model 1
Note: results for the other models are shown in Appendix 6.
3.6.4 Autocorrelation
As discussed previously, this research uses panel data as it combines both cross-sectional and time-series. Therefore, the problem of autocorrelation can arise when an observation in one period depends on what happened in the former period (Hill, et al., 2008). In other words, a firm in the sample which performed well in 2003 is also more likely to perform well in the 2004. This research will use the Wooldridge test for autocorrelation. The testdetermines whether we can confirm the null hypothesis of no autocorrelation. By inspecting the results shown in table 3.5, this research can reject the null hypothesis at a 5% significance level. Hence, robust standard errors will be used in order to deal with the autocorrelation.
Table 3.5: Wooldridge test for autocorrelation model 1
Wooldridge test for autocorrelation in panel data H0: no first-order autocorrelation
F( 1, 228) = 29.183 Prob > F = 0.0000
Note: results for the other models are shown Appendix 7.
24
4. Results
4.1 Descriptive statistics
In this section the results of the models will be presented and explained. Before the hypotheses are tested, the descriptive statistics will be discussed. Table 4.1 shows the descriptive statistics of the dependent and independent, and control variables.
Table 4.1: Descriptive Statistics
Obs. Mean Std. Dev. Min Max
Revenue International diversification Multinationality Intangible assets Geographical scope Attractive Markets Subsidiary strategy Firm size Experience 2300 2300 2300 2267 2300 2300 2300 2278 2290 10.26286 .3434783 9.2 4.880627 1.117391 .1872182 .3126039 10.46044 24.62926 2.713295 .4749727 22.06684 4.020529 2.537094 .3504008 .3412763 3.003498 26.86637 -.3753494 0 0 -2.040221 0 0 0 3.350474 1 17.47861 1 219 14.88439 15 1 1 17.94941 209
25
are active in more than one market. Moreover, the firm RAWLPLUG SA is active in the most foreign markets, with a Geographical Scope of 15. With an average of 19%, and a standard deviation of 35%, for the variable Attractive Markets, most points are located in the range of 0% to 54%, and in this way, most subsidiaries are located in attractive markets (i.e. emerging or developing markets). In similar vein, based on the average and standard deviation of
Subsidiary Strategy this research concludes that the largest part of the subsidiaries are wholly
owned. Besides the dependent variable and independent variables, the descriptive statistics also include the control variables. As has been mentioned previously, Firm Size is measured as the natural logarithm of total assets. In absolute value, the size of the firms varies between 2.851.624 euros for PSE Inwestycje SA and 62.420.828 euros for PKO BB SA. The variable
Experience reveals that, on average, during the analyses firms are 25 years old, while the
oldest firm has the respectable age of 209. Furthermore, the set of dummy variables controlling for industries, host markets of subsidiaries, and yearly effect, were not included to keep it detailed. However, to briefly discus them, most firms are participating in an industry which is manufacturing related, while most foreign subsidiaries are located in Eastern markets. To summarize, the descriptive statistics demonstrates that the firms which diversify internationally mostly invest in WOS located in developing or emerging markets.
4.2. Regression results
Table 4.2 displays the results of the regression analysis. First the control variables are discussed. Firm Size and Experience are positively associated with Revenue in all models of both POLS and FE. Where Firm Size is highly significant in all models (1%), the significance of Experience varies between 1%, 5%, and 10%. As discussed previously, sets of dummy variables are included to control for Industry, Time, and Host Countries effects. Overall, the results of the individual dummy variables showed limited levels of significance. Hence, a Wald test was performed to test for a joint significance among the sets of variables. As can be seen, the dummy variables for Host Countries and Industry15 show a joint significance in all
models. Meaning that at least some of them are significant, and therefore, add value to the model (Hill, et al.,2012). For the dummy variables Year the results are, as suspected, slightly different. The dummy variables of Year did show a joint significance in the FE models, however, not in most of the POLS models. Nonetheless, this was expected as POLS does not account for yearly individual heterogeneity among firms which might lead to different coefficients (Hill, et al.,2012). To summarize, results of the Wald test indicate that the dummy
15
26
variables for Host countries, Industry and Year add value to the models for which it was expected.
Turning to the dependent variables, Hypothesis 1 predicts that international diversified firms are likely to perform better than non-diversified firms. On the basis of the FE regression, the hypothesis is supported. As can be seen, the coefficient in model 1 (b = 0.268 significant at 5%) and model 8 (b = 0.280 significant at 5%) is positive and significant. This confirms that firms which are internationally diversified perform better than those which are not internationally diversified. Moreover, results are consistent with previous empirical studies (e.g. Mayar & Ottaviano 2008; and Heyder et al. 2011). Hence, emerging market firms seem to have developed competences at their home market and are able to apply these in foreign markets (Bartlett & Ghoshal, 1999). Unfortunately, no significant results were captured by the POLS regression. As can be seen, the FE regression omits the time-invariant control variables associated to the industries of firms. This suggest that the industry-specific, time-invariant, omitted variables generate an effect that is not significant in the POLS. In other words, if we do not account for individual differences in firms their industry specification, but instead pool them together, than it cannot be confirmed if international diversification leads to higher performance.
Table 4.2: Regressions results
Notes: * significant at 10%, ** significant at 5%; *** significant at 1%, dependent variable: Revenue, estimation method: OLS, time period: 2003-2012.
Model 1 Model 2 Model 3 Model 4a Model 4b
POLS FE POLS FE POLS FE POLS FE POLS FE
International Diversification 0.074 (0.98) 0.268** (2.01) Multinationality 0.003*** (3.50) 0.203*** (3.55) Intangible Assets 0.074*** (6.82) 0.069*** (2.65) 0.083*** (7.12) 0.071** (2.53) Geographical Scope 0.003 (0.46) 0.017 (0.09) 0.051*** (4.88) 0.046 (0.25) Attractive Markets Subsidiary Strategy
Geographical Scope*Intangible Assets -0.006***
(4.42) -0.002 (0.43) Multinationality*Geographical Scope Firm Size 0.829*** (60.26) 0.484*** (3.87) 0.828*** (60.72) 0.468*** (3.55) 0.780*** (42.56) 0.423*** (3.57) 0.830*** (60.36) 0.468*** (3.75) 0.776*** (41.70) 0.424*** (3.56) Experience 0.004*** (5.25) 0.033* (1.82) 0.003*** (4.66) 0.039** (2.16) 0.005*** (6.42) 0.031** (1.97) 0.004*** (5.33) 0.039** (2.17) 0.004*** (5.56) 0.031** (2.01) Constant 0.162 (0.80) 4.178*** (4.64) 0.154 (0.76) 2.352 (2.18)** 0.164 (0.81) 4.545 (5.28)*** 0.150 (0.74) 4.213*** (4.83) 0.169 (0.84) 4.491*** (5.27) Wald industry-dummies joint significance
test
76.08*** 73.44*** 61.35*** 75.97*** 62.35***
Wald host country-dummies joint significance test
Table 4.2: Regressions results (continuing)
Notes: * significant at 10%, ** significant at 5%; *** significant at 1%, dependent variable: Revenue, estimation method: OLS, time period: 2003-2012.
Model 5 Model 6 Model 7 Model 8
POLS FE POLS FE POLS FE POLS FE
International Diversification 0.113 (1.20) 0.280** (1.92) Multinationality 0.026*** (6.93) 0.221*** (3.33) 0.018*** (5.09) 0.243** (1.92) Intangible Assets 0.075*** (6.18) 0.073** (2.55) Geographical Scope 0.036*** (3.80) 0.367* (1.82) 0.103*** (3.39) 0.496* (1.96) Attractive Markets 0.046 (0.57) 3.694*** (2.64) -0.085 (0.93) 3.125** (2.10) Subsidiary Strategy 0.012 (0.17) -0.332 (0.76) -0.010 (0.14) 0.307 (0.63)
Geographical Scope*Intangible Assets -0.008**
(2.87) -0.003 (0.63) Multinationality*Geographical Scope -0.002*** (6.75) -0.003** (2.33) -0.001*** (4.67) -0.003** (2.22) Firm Size 0.798*** (50.02) 0.468*** (3.74) 0.829*** (60.22) 0.467*** (3.74) 0.830*** (60.69) 0.468*** (3.75) 0.760*** (39.01) 0.435*** (3.63) Experience 0.005*** (6.10) 0.039** (2.16) 0.004*** (5.31) 0.038** (2.15) 0.004*** (5.30) 0.039** (2.17) 0.004*** (6.15) 0.031** (2.00) Constant 0.294 (1.21) 1.914** (2.04) 0.157 (0.77) 3.564*** (4.65) 0.147 (0.74) 4.336*** (5.30) 0.258 (1.28) 1.457 (1.05) Wald industry-dummies joint significance test 75.86*** 76.38*** 67.53*** 53.71***
Hypothesis 3 states that emerging market firms are more likely to perform better when utilizing intangible assets. This hypothesis is supported, as results show that Intangible Assets is positively associated with Revenue in both the POLS and FE regressions. The significant coefficients in model 2 (b = 0.084 significant at 1% for POLS, and b = 0.069 significant at 1% for FE) and model 8 (b = 0.075 significant at 1% for POLS, and b = 0.073 significant at 5% for FE), means that firms that invest more in formation expenses, research expenses, goodwill, development expenses, and all other expenses with a long term effect are beneficial for performance. Hence, this results confirms the theory that emerging market firms perform better in the presence of home-grown intangible resources (Denk et al., 2012). Moreover, previous researchers emphasized on the importance of intangible assets which are based on business group affiliations (e.g. Hitt et al., and 2000; Hoskisson et al., 2005; and Elango & Pattnaik, 2007). Unfortunately, due data limitations, this research could not find such information for the firms included in this sample.
Hypothesis 4a states that the geographical scope of firm’s operations positively influences performance, whereas hypothesis 4b states that intangible assets are likely to strengthen the positive relationship of hypothesis 4a. Unfortunately, the results are not consistent across the models and both hypothesis 4a and 4b are not supported. As can be seen in model 4a, results lack the significance for supporting hypothesis 4a. However, when the variable geographical scope is regressed simultaneously with other variables in model 4b and model 8, it does appear to produce positive and significant coefficients. Therefore, this research cannot, keeping all other variables constant, confirm that increasing geographical scope positively influences performance, however, a robustness check will be performed in the following section to further investigate the relationship between the variables
Geographical Scope and Revenue. In addition, it could not be confirmed that intangible assets
30
FE model. Indicating that the industry dummy variables generated the significant and positive effects.
Hypothesis 5 states that the interaction of geographical scope and multinationality should positively influence firm performance. This hypothesis is not supported, since the coefficients of the interaction variable (i.e. Multinationality * Geographical Scope) in model 5 (b = -0.003 significant at 1% for POLS, and b = -0.003 significant at 1% for FE) and model 8 (b = -0.001 significant at 1% for POLS, and b = -0.001 significant at 1% for FE) are negative and significant for both POLS and FE. Indicating that if emerging market firms would simultaneously increase their multinationality (i.e. economies of scale) and geographical scope (i.e. economies of scope) their performance diminishes. The current study theorized that, by increasing the size of foreign operations plus the experience in leveraging this in multiple environments should yield a positive interaction effects. However, as we know now, by interacting the variables, the positive effects of geographical scope and multinationality gets negatively moderated. As mentioned previously, this could be evidence of a multi-domestic strategy, where multiple subsidiaries located in each country and are not collaborating, and in this way, are not benefiting from economies of scale (Lovelock, 1999).
Hypothesis 6 states that diversifying in emerging or developing markets should positively influence firm performance for emerging market firms. Based on the FE regression this hypothesis is supported. From model 6 (b = 3.694 significant at 1%, and b = 3.125 significant at 1%) it is evident that there is a positive and significant effect of diversifying in attractive markets. Indicating that emerging market firms which invest more in similar markets (i.e. emerging or developing markets) in relative to other markets (i.e. developed markets) turn out to be more profitable. Thus, confirming the IBV literature, which argues that familiarity within host environments reduces the disadvantages associated with liability of foreignness, and accordingly, improves performance. No significant results are found by the POLS in model 6 and 8. Meaning that pooling the industry control dummies into the regression leads to non-significant results. However, the overall F-test reported in table 4.2 shows that there are significant differences between at least some firms. Hence, it is justified to confirm Hypothesis 6 based on the results of FE.
31
IBV, emerging market firms should, in theory, benefit from adopting a cooperative strategy. If anything, the negative coefficients in model 7 (b = -0.332 not significant) would indicate, keeping other variables constant, that choosing JV as a ownership structure of subsidiary is not beneficial for firm performance. An explanation for a negative coefficient is found in the results from an empirical study done by Makino and Delios (1996). Their results indicate that in the absence of international experience or local knowledge JVs are more profitable than WOS, however, the need for a cooperative strategy declines as experience in foreign markets increase. Another explanation could be that firms are likely to opt for a subsidiary structure in which they are familiar. As is seen from the descriptive statistics analysis, firms included in this research mainly choose of WOS as ownership structure of their subsidiaries. Hence, it could be theorized that firms are familiar with opting for WOS, although theory predicts that JV is more appropriate for emerging market firms, their experience lets them opt for a well-known ownership structure (Yiu & Makino, 2002). Nonetheless, results are not significant and therefore lack any statistical power.
4.3 Robustness checks
The first sets of robustness checks involve investigating the potential non-linear relationship of the variables Multinationality and Geographical Scope with Revenue. To begin with the relationship between Multinationality and Revenue. Results in table 8.1 model 9 (Appendix 8) are conclusive: increasing multinationality is beneficial for firm performance. However, as discussed previously, there have been empirical studies finding a non-linear relationship (e.g. Hitt et al., 1997; Capar & Kotabe, 2003 and Endo & Ozaki, 2011). The results from POLS model demonstrate that the variable Multinationality starts positive (b = 0.002 significant at 1%) before it turns negative (b = -0.491 significant at 5%). Meaning that there is an inverted U-shaped non-linear relationship between the variable Multinationality and Revenue, where expending the number of subsidiaries is beneficial for performance, however, expansion beyond an optimal level decreases performance. In other words, the cost of increasing multinationality will eventually exceed the benefits (Hitt et al., 1997; Endo & Ozaki, 2011). Unfortunately, the non-linear relationship between multinationality and performance is not confirmed by FE.
32
regression, the variable Geographical Scope turned out to have a non-linear relationship with the variable Revenue. As can be seen based on the coefficients in table 8.1 model 10 (Appendix 8), Geographical Scope starts out negative (b = -6.978 significant at 5%), however, than turns positive (b = 0.332 significant at 5%). In other words, between
Geographical Scope and Revenue there exists an U-shaped relationship, which suggest an
initially negative effect of expending operations across multiple markets on firm performance, before the positive returns of expending are realized. Meaning that, when firms begin expending operations across multiple markets, they are faced with costs related to liability of foreignness. However, firms expanding their operations over a particular threshold do seem to obtain the ability for exploiting their resources efficiently (Qian et al., 2010).
The last robustness check involves the interaction effect of the variables
Multinationality and Geographical Scope. As discussed in the previous section, results
33
5. Discussion
For the discussion of the research results, lets return to the main research question first: What drives the performance of emerging market firms? The results clearly show that the performance of emerging market firms is resource driven. As explained by using the RBV, firms that diversify internationally perform better as result of the development of competences in exploiting knowledge and resources in a more efficient way than others. Additionally, as firms increase their presence in foreign markets (i.e. increase multinationality), higher returns are realized. This is consistent with both theory and previous empirical studies (e.g. Ramírez-Alesón & Antonio Espitia-Escuer, 2001; and Kirca et al., 2011). Through expanding operations, the opportunity to exploit the advantages of their resources improves (i.e. economies of scale). As a robustness check, this research further investigated the relationship between multinationality and performance. Based on the POLS results, this research concludes that there is an inverted U-shaped non-linear effect, meaning that firms benefit from expending their foreign presence. However, expansion beyond an optimal level decreases performance. In other words, the cost of increasing multinationality will eventually exceed the benefits (Hitt et al., 1997; Endo & Ozaki, 2011). In the end, the results indicate diminishing returns on economies of scale.
34
increasing the scope of operations turns out to be beneficial. An explanation for the non-linear relationship can be found in the costs related to the liability of foreignness. When firms begin expending their geographical scope, their ability to exploit resources does not exceed the costs. However, firms that have expended their geographical scope behind an optimal level, seem to obtain the ability of not exceeding the costs when exploiting resources (Qian et al., 2010). Moreover, as the results of the robustness check confirm, only experienced firms seem to be able to benefit from increasing their multinationality and geographical scope all at once. In other words, only experienced firms possess the knowledge, which enables them to benefit from economies of scale simultaneously in multiple foreign markets.
After examining the influence of the resources, within this research also the choices of foreign locations for investments by emerging market firms are tested. The results show that the choice of the location significantly influences the performance of the firms. According to the IBV, emerging market firms benefit from diversifying in markets with similarities in culture, traditions, customs, and behaviors (i.e. rules of the game) (Lee and Beamish, 1995). As a result, firms than are less confronted with costs related to the liability of foreignness. Another aspect is the scope of investing in similar markets on firm performance. Where this research has showed several positive effects on firm performance, none of these has a more significant impact on performance as choosing to invest in a similar market (see table 4.2).
35
6. Conclusion
This research contributes to the growing importance of understanding emerging market firms. While there is a large body of research already, both theoretical and empirical, dedicated to understanding the drivers behind firm performance, most researches focus on firms from developed markets. Therefore, this research examined the drivers behind the performance of emerging market firms in today’s global market, by focusing on its resources, location choices in foreign markets and ownership structures of subsidiaries. Since the role of emerging markets and emerging market firms in the world economy is increasing, the main goal of this research is to better understand what drives their performance. Where previous empirical studies use samples of firms from emerging markets, for example China or India, this research used a sample of firms from Poland. As a result, this research contributes to the international business literature. The results show that emerging market firms benefit from expending their operations in foreign markets until a particular optimal level. Nonetheless, only the experienced firms should try expending operations simultaneously in multiple host markets. Additionally, emerging market firms tend to benefit, in terms of performance, from unique intangible resources that are not transferable to foreign markets. Notable is also that emerging market firms seem to benefit from investing in markets with similar characteristics.
6.1 Limitations and future research
36
7. References
Al-Obaidan, A. M., & Scully, G. W. (1995). The theory and measurement of the net benefits of multinationality: the case of the international petroleum industry. Applied
Economics, 27(2), 231-238.
Allison, P. (2012) When Can You Safely Ignore Multicollinearity? Retrieved from http://www.statisticalhorizons.com/multicollinearity on 21-05-2014
Andersen, O. (1997). Internationalization and market entry mode: A review of theories and conceptual frameworks. MIR: Management International Review, 27-42.
Baltagi, B. (2008). Econometric analysis of panel data. John Wiley & Sons.
Bartlett, C. A., & Ghoshal, S. (1999). Managing across borders: The transnational
solution (Vol. 2). Harvard Business School Press.
Baum, C. F. (2006). Stata tip 38: Testing for group wise heteroskedasticity. Stata
Journal, 6(4), 590-592.
Boeh, K. K., & Beamish, P. W. (2012). Travel time and the liability of distance in foreign direct investment: Location choice and entry mode. Journal of International Business
Studies, 43(5), 525-535.
Brouthers, K. D., Brouthers, L. E., & Werner, S. (2003). Transaction cost enhanced entry mode choices and firm performance. Strategic Management Journal, 24(12), 1239-1248.
Brouthers, K. D., Brouthers, L. E., & Werner, S. (2008). Real options, international entry mode choice and performance. Journal of Management Studies, 45(5), 936-960.
Brouthers, K. D. (2013). Institutional, cultural and transaction cost influences on entry mode choice and performance. Journal of International Business Studies, 44(1), 1-13.
Buckley, P. J., Clegg, L. J., Cross, A. R., Liu, X., Voss, H., & Zheng, P. (2007). The determinants of Chinese outward foreign direct investment. Journal of international business
studies, 38(4), 499-518.
Bühner, R. (1987). Assessing international diversification of West German
