Local, regional, or global? Geographic orientation and relative financial performance of emerging market multinational enterprises

European Management Journal (2009) 27, 344– 355

journal homepage: www.elsevier.com/locate/emj

Local, regional, or global? Geographic orientation and relative financial performance of emerging market multinational enterprises Elitsa R. Banalieva

a,*

, Michael D. Santoro

b,1

a

International Business and Strategy Group, College of Business Administration, Northeastern University, 360 Huntington Avenue, 313 Hayden Hall, Boston, MA 02115-5000, United States b College of Business and Economics, Lehigh University, 621 Taylor Street, Bethlehem, PA 18015, United States

KEYWORDS Geographic orientations; Regional/global strategies; Performance; Emerging markets; Multinationals

We extend the regional/global strategies literature by analyzing the relative performance of emerging market (EM) multinational enterprises (MNEs) based on their geographic orientation. We develop a framework showing that firms adopt three geographic orientations—local, regional, and global—and test our framework with the market penetration strategies (sales) of 701 MNEs from 28 EMs during 2000–2006. Our analysis shows that distinguishing among these three geographic segments is important, as not all of these geographic segments enhance firmsÕ financial performance relative to their industry: a combination of local and global orientations enhances while regional orientation reduces the relative financial performance of EM MNEs. ª 2009 Elsevier Ltd. All rights reserved.

Summary

Introduction Are emerging market multinational enterprises (EM MNEs) mostly locally, regionally, or globally oriented? How does their geographic orientation, i.e., local, regional, or global orientation, affect financial performance relative to their industry rivals? These two research questions are at the heart of our study. * Corresponding author. Tel.: +1 617 373 4756; fax: +1 617 373 8628. E-mail addresses: [email protected] (E.R. Banalieva), [email protected] (M.D. Santoro). 1 Tel.: +1 610 758 6414; fax: +1 610 758 6941.

Emerging markets represent an interesting and relevant context when analyzing the recently advanced regionalization hypothesis (Rugman and Verbeke, 2004, 2007, 2008) for several key reasons. First, EM MNEs face many challenges in the global arena. One notable challenge is that EM MNEs can sometimes be late movers with respect to multinationality as compared to developed market MNEs (Luo and Tung, 2007). Yet, an advantage for EM MNEs is that they are often first movers in gaining market opportunities in their often-times underdeveloped home countries. Second, EMs, defined as ‘‘low-income, rapid-growth countries using economic liberalization as their primary engine for growth’’ (Hoskisson et al., 2000, p. 249), comprise more than half of the worldÕs population and account for 45% of the worldÕs

0263-2373/$ - see front matter ª 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.emj.2009.04.001

Local, regional, or global? Geographic orientation and relative financial performance of emerging market exports (Economist, 2007). Third, despite the notable importance of EMs to the global economy and the challenges EM MNEs face in their internationalization, no study in the regional/global strategies literature to date has fully examined the link between regionalization and relative financial performance in the context of EM MNEs. So far, the literatureÕs focus has been largely on Triad MNEs (e.g., Delios and Beamish, 2005; Li, 2005; Rugman and Verbeke, 2004, 2007, 2008). We aim to fill this gap with the present study. Specifically, we build upon previous literature (Delios and Beamish, 2005; Li, 2005; Rugman and Verbeke, 2008) that suggests that an MNEÕs home region includes two distinct geographic segments—local, i.e., activities within the home country; and regional, i.e., activities outside the home country but within the proximate home regional confines. We also expand on an even greater multinationality literature (e.g., Li, 2005; Hennart, 2007; Wan and Hoskisson, 2003) by incorporating a third important component; i.e., the global segment, since some MNEs also operate globally (i.e., beyond their home country and home region). Since learning the competitive conditions of doing business and creating a competitive advantage in each of these three segments is costly for MNE managers, MNEs do not typically focus equally on the three geographic segments. Instead, firms, particularly EM MNEs, orient to various extents toward one of three main geographic segments: local, regional, or global. For instance, in 2006, Naspers, a SouthAfrican MNE focused on electronic and print media, extracted roughly 76% of its total sales from Africa and only 11.7% from the rest of the African continent; while Acer, a Taiwanese electronics MNE, extracted 63% of its total sales from the global segment outside of Asia, and 20% from Taiwan. A key question from these brief examples relates to how the differences in the firmsÕ geographic orientations affect their financial performance relative to their industry rivals. Focusing on the firmsÕ financial performance compared to that of their industry allows us to draw implications for the relative market position of the firm; i.e., whether the EM MNEs following each of these geographic orientations perform better or worse than their industry rivals. From a strategy perspective, performing better than the industry ensures the longer-term survival of the company, especially in an emerging market context where country-specific advantages such as lower labor costs enable firms to cut costs and compete more efficiently against each other, as it shows that the outperforming firm is capable of guarding its know-how and market share well against its industry rivals. Thus, analyzing how the geographic orientations of local, regional, and global relate to firmsÕ relative financial performance represents an important and interesting research question, which we address in this article. With this study, we address this issue and, in doing so, attempt to make both theoretical and empirical contributions to the growing regional/global strategies literature. First, we propose a finer-grained classification of firmsÕ geographic orientations that takes into account both firmsÕ intra regional (i.e., between home country and rest of home region) and global (i.e., outside the home region) expansions2 of EM MNEs. Second, by using a comprehensive dataset of 2

By expansion, we mean the scope of operations of the MNEs.

345

701 MNEs based in 28 EMs from Africa, the Americas, Asia, and Europe during 2000–2006, we analyze whether EM MNEs are mostly locally, regionally, or globally oriented and examine the relationships between these three geographic orientations and relative financial performance. Third, by investigating the geographic orientations of EM MNEs and relating them to the firmsÕ relative financial performance, we test our theory and add to previous research by examining this phenomenon with an understudied sample that is of growing importance in an increasingly interdependent global economy. In pursuing this line of inquiry we acknowledge that as Hennart (2007) notes, the internationalization-performance literature has produced inconclusive results, since in just the past decade alone, more than one hundred empirical articles have been written. However, they have failed to offer robust results regarding the relationship between multinationality and relative financial performance (Hennart, 2007). Our intent is to advance the theory through theoretical parsimony in the context of EM MNEs. To accomplish this, we follow HennartÕs (2007) call to examine the internationalization-performance relationship at lower levels of analysis, which allow the investigation and control for likely crucial variables at several levels of analysis. We, therefore, include firm-, industry-, country-, and region-level constructs in our analysis to address HennartÕs (2007) recommendations. This study is organized as follows. The next section proposes our conceptual framework of MNEsÕ geographic orientations and their relationship to the firmsÕ relative financial performance. We then present our research design, followed by a discussion of our research results and our contributions to theory, future research, and management practice. We conclude by reviewing the caveats and main findings of this study.

Framework of MNEsÕ geographic orientations We build upon the two dimensions of intra-regional and global orientations as the foundation for our proposed framework. Figure 1 shows our Organizing Framework of

Local Intra Regional

Regional

Global

Figure 1 An organizing framework of multinationalsÕ geographic orientations. Based on Rugman and VerbekeÕs (2008) study, where Regional corresponds to their ‘‘rest of home region’’ and Global corresponds to their ‘‘rest of world’’ segments.

346 MultinationalsÕ Geographic Orientations and illustrates the three geographic segments at the heart of our conceptual framework—local, regional, and global—with the combination of local and regional segments forming the intra-regional dimension (Rugman and Verbeke, 2004, 2007, 2008). We illustrate our framework in Figure 1 with respect to MNEsÕ market penetration strategies, i.e., sales, to maintain consistency with a large number of previous studies on regional/global strategies of MNEs (e.g., Li, 2005; Rugman and Verbeke, 2004, 2007, 2008)3.

Intra-regional orientation In Figure 1, we distinguish between an MNEÕs local and regional geographic segments, both of which comprise the intraregional orientation of an MNE. Distinguishing between the local and regional segments is important. An MNE, for example, could have high home regional sales possibly due to a high local and low regional orientation, or vice versa. Referring to a multinational as ‘‘home-regionally oriented,’’ as is sometimes seen in the literature, is not sufficient because such a label may be disguising either a local orientation, i.e., when the MNE is focused more on its home country segment rather than regional segment, or a regional orientation, i.e., when the MNE is focused more on its regional segment rather than home country segment. Thus, our framework shows that it should be specified whether the firm is specifically locally or regionally oriented within its home region. We also suggest an objective way to determine the MNEÕs geographic orientation not based on random cut-off percentage points (e.g., Collinson and Rugman, 2008; Delios and Beamish, 2005; Osegowitch and Sammartino, 2008; Rugman et al., 2008; Rugman and Verbeke, 2004, 2008), but, instead, on overall comparison of the three geographic segments. As seen in Figure 1, a firm can be said to  L have a local orientation if its ratio of local-to-total R sales T exceeds the ratios sales T and global-to-total sales G of regional-to-total L R L G > and > . Locally-oriented MNEs pursue ; i.e., T T T T T home markets rather than regional and global markets because they are unable to overcome the liability of regional foreignness (LORF) (Rugman and Verbeke, 2008), and thus prefer to keep most of their activities within their home base. For instance, companies like Naspers (a South African media firm) or PKN Orlen (a Polish oil refinery and petrochemical retailer) have a local orientation, as they received 76.36% and 53.79% of their total sales from their home country markets, respectively, in 2006. Naspers further emphasizes the importance of its local market by noting that, ‘‘the group derives revenues from television platform services, print media activities, internet services, technology products and services, book publishing and private education. The activities in the Republic of South Africa are the most significant in this segment’’ (Naspers, 2006). Likewise, a firm can be said to have a regional orientation, or to be ‘‘regionalizing,’’ if the regional-to-total sales ratio is the highest; i.e., TR > TL and TR > GT. Unlike the local orientation, the regional orientation posits firms focus more on their regional markets relative to their local and more 3

However, our framework can easily be adapted to other measures of international involvement such as assets, employees, or subsidiaries, for instance.

E.R. Banalieva, M.D. Santoro distant global markets because they are unable to overcome the liability of global foreignness (LOGF) (Rugman and Verbeke, 2008). For instance, companies like Viohalco (a Greek metals processing company) and Alco Hellas (another Greek metals processing company) obtained 58% and 70%, respectively, of their total sales from the rest of their European home region in 2006 and would be considered to have a regional orientation. Rugman and Verbeke (2008) note that LOGF is generally expected to be higher than LORF, which may be one important reason why so many MNEs are often unable to reach a meaningful global scope of operations. In an EM context, the home region of many EM MNEs is frequently a collection of other developing EMs. For instance, South AfricaÕs African home region comprises many EM neighbors, characterized by unstable political and economic systems. Likewise, Asia is home to many EMs, such as Malaysia, Indonesia, Vietnam, China, India, etc., where market liberalization is advancing; yet, the region also is home to Japan, a much more developed market. These neighboring countries, part of an EM MNEÕs home region, are often similar to one another in the levels of economic, political, financial, and overall business development. Customers in these EM neighboring countries often lack the high purchasing power that their developed market (e.g., the US, Germany, UK, or Japan) counterparts have to buy goods and services. Thus, when EM MNEs try to internationalize to the surrounding nations within their home region that tend to be EMs, we would expect performance to suffer because the combination of developed and emerging markets within a home region represents unique challenges to the multinational manager. From a transactions costs economics perspective, a volatile surrounding home region environment also increases the risk and costs of doing business within the home region, as opposed to within oneÕs home country. Moreover, LORF may be too high for many EM MNEs to overcome, even though it is lower than LOGF, because EM MNEs often lack technology advantages. Conversely, focusing on the domestic market, with the difficulties and challenges the EM MNE already is familiar with and knows how to manage, can be a winning strategy for the EM MNE. The home country market often involves less risk and fewer exploration resources, which are sometimes scant in an EM MNEsÕ internationalization arsenal. Moreover, the home market often represents a more familiar landscape in terms of geography, language, customs, and consumer tastes and preferences, which EM MNEs know how to manage best and, consequently, translate into less uncertainty for the firms (e.g., Hejazi, 2007). Additionally, customers in the home country may be more familiar with, trust more, and consequently be more willing or compelled to buy their domestic producersÕ goods than lesser known brands produced from surrounding developing nations, particularly when political/legal pressures to buy from domestic producers are intensified in a home market. All of these home country benefits and neighboring country disadvantages translate into greater success for the EM MNE. Thus, we expect that the home country of an EM MNE provides a ‘‘safe heaven’’ in which the firm has developed and accumulated knowledge, which causes the firm to gain more performance advantages via a more local orientation. Accordingly, we propose that intra-regionally:

Local, regional, or global? Geographic orientation and relative financial performance of emerging market H1: The higher the local orientation, the higher the relative financial performance.

Global orientation Besides MNEs expanding intra-regionally, they also can ‘‘leapfrog’’ from their local home country markets to more distant global markets. As seen in Figure 1, a firm has a global orientation if its global-to-total sales ratio is the highest; i.e., GT > TL and GT > TR. Here, MNEs pay most of their attention and focus on distant global markets and give the least attention to local competitive conditions. For instance, companies like Acer (a Taiwanese electronics firm) or Cemex (a Mexican manufacturer and distributor of cement) obtain 63% and 53%, respectively, from their total sales from their distant global markets, and would be considered to have a global orientation. A growing body of literature suggests a large number of MNEs tend to be home-regionally oriented instead of globally dispersed because the variance of home regional conditions that includes critical factors such as geography, economics, and politics tends to be smaller than the variance found in more global conditions (Li, 2005; Rugman and Verbeke, 2004, 2007, 2008). To this point, the regional and global strategies literature emphasizes that doing business in countries that are less proximate and more dissimilar than the firmÕs home country in terms of geography, economics, and politics often entail higher risk, more complexity, and greater uncertainty. The reason is that in managing globally-dispersed lines of businesses, managers must obtain and update different types of knowledge about many more diverse global environments resulting in information overload and added costs. Conversely, MNEs operating regionally deal with less uncertainty, less information overload, and less complexity due to economic, political, and geographic conditions that are more proximate and similar to that of the firmÕs home country, often resulting in reduced costs and greater organizational success (Rugman and Verbeke, 1993, 2008). Thus, in general, LOGF is larger than LORF for MNEs, which suggests that expanding regionally rather than globally reduces transaction costs as an MNE considers the factor conditions not only within its domestic market but also within another neighboring home-region country (Rugman and Verbeke, 1993). When EM MNEs do not have a particular preference for either the local or regional segment, but attribute equal importance to each (i.e., when R/T = L/T), having a lower global orientation decreases the firmÕs exposure to LOGF and, consequently, increases the relative financial performance of firms, such that: H2: The lower the global orientation, the higher the relative financial performance.

Methodology Data We started with Hoskisson et al.Õs (2000) list of 51 EMs worldwide. The sample of publicly-listed EM MNEs based in

347

these 51 countries was drawn from OSIRIS, a financial database commercially provided by Bureau van Djik. We excluded firms in the financial sector and firms, which were not independent4. To ensure that the firms were multinational, we excluded firms with less than 10% foreign-to-total sales and we included in the final sample only those MNEs that had reported and separated home country sales, rest of home region sales, and global sales. Since not all firms had data on all the variables necessary for our regression analysis, our final sample was unbalanced and included 701 MNEs, or 2105 firm-year observations for 2000–2006, based in 28 EMs. The 28 EMs were located in four geographic regions: Europe (Croatia, Estonia, Greece, Hungary, Latvia, Lithuania, Poland, Portugal, Russia, Slovenia, and Turkey)— 5% of the sample; Asia (China, India, Indonesia, Israel, Malaysia, Philippines, Saudi Arabia, Sri Lanka, Taiwan, and Thailand)—89% of the sample5; the Americas (Mexico, Peru, and Venezuela)—3% of the sample; and Africa (Kenya, Nigeria, Mauritius, and South Africa)—3% of the sample. Even though the sample turned out to be primarily oriented toward Asian MNEs, we did not want to exclude those firms in Europe, the Americas and Africa since we wanted a comprehensive and representative sample of EM MNEs based in as many regions as possible to increase the generalizability of our findings.

Measures Dependent variable As in a number of previous studies, we measured financial performance using return on assets (ROA) (e.g., Hitt et al., 1997; Kotabe et al., 2002; Wan and Hoskisson, 2003). ROA, calculated as net income-to-total assets, reflects how successful a company is at generating income from its invested capital, therefore capturing both the prof4 In OSIRIS, independent are firms that have no ‘‘ultimate parent owners,’’ i.e., firms in which no other firm held more than 25% ownership (this threshold is provided by OSIRIS). The latter step ensures that the firms were not subsidiaries of another firm in the sample, which would result in double-counting. 5 Collinson and Rugman (2008) also found that Asian firms in the broad Triad, roughly 1/3 of the 500 largest firms, are dominated by Japanese firms reporting regional sales data. Thus, we too note that the prevalence of Asian firms in our sample indicates Asian firms have more consistent data reporting than do firms in other regions. We thank one of our reviewerÕs for pointing this out. For instance, most of the data we lost was due to the fact that many non-Asian MNEs reported their domestic sales together with their rest of home region sales into one combined number, therefore making it impossible for us to separate the Local and Regional segments, as operationalized in our conceptual framework. Additionally, we opted against dropping firms in Africa, Europe, and Americas since we wanted to have all possible EM MNEs of the world included in our sample to increase our studyÕs generalizability. Dropping Africa, Europe, and Americas would have meant that our results would be only applicable to Asian-based EM MNEs. We wanted our study to have a broader view as we believe it is important to include some of the lesser-known emerging markets of the world such as Africa and the Americas, because some of the countries in these continents have been severely understudied in the international business literature and our study highlights these areas for the IB research agenda, as viable markets for future research.

348 itability and efficiency of assets employed. Studies in international business have sometimes used return on sales (ROS) as the dependent variable but we opted against ROS for several reasons, despite its high correlation with ROA (54.52%). First, ROA has been the most widely used measure of a companyÕs financial performance in the internationalizationperformance literature, so using it allowed us to maintain maximum consistency with prior studies. Second, given that ROS and several of our independent and control variables are functions of total firm sales, using ROS would have artificially increased the correlations between these salesrelated variables. Indeed, prior research has cautioned that ‘‘regression equations with ROS as the dependent variable might reflect mathematical artifacts as well as true relations’’ (Hitt et al., 1997, p. 778). Furthermore, since our theoretical development pertains to the financial performance of EM MNEs relative to their industry rivals, we matched our theoretical arguments with our empirical operationalization by subtracting the average ROA for each firmÕs respective industry (SIC2) from the firmÕs ROA value (Cho and Cohen, 1997). This performance adjustment allowed us to analyze how the firms were performing financially relative to their industry. Positive (negative) values of this variable mean that the focal firm is performing better (worse) than the industry.

Independent variables Following our proposed conceptual framework, we measured the Intra-Regional Orientation of firms as the difference of two ratios: Sales/Total Sales minus Local  Regional  Sales/Total Sales TR  TL . Previous research also uses such ratios to capture the geographic orientation of firms (e.g., Elango, 2004; Rugman and Verbeke, 2008), but does not take the difference of these ratios as we do in this study in order to capture the Intra Regional Orientation of the firm. This difference can be greater or less than 0. This natural zero cutoff point facilitates determining the geographic orientation of the firm. The TR  TL difference reflects MNEsÕ intra-regional expansion within the home region; when it is <0, the MNE has a local geographic orientation of TR < TL , and when TR  TL is >0, the MNE has a regional geographic orientation of TR  TL . Thus, with one variable, we were able to capture two types of geographic orientations—regional or local—that firms undertake within their home regions. Because a firmÕs global segment is so distinct from its home region segment due to the expected higher LOGF and higher diversity globally, we measured Global Orientation as a stand-alone ratio, as opposed to using the difference in ratios as we did for the Intra-Regional Orientation. Following previous research (e.g., Elango, 2004; Rugman and Verbeke, 2008), we too measured Global   Orientation by the ratio of Global Sales/Total Sales GT , which ranges from 0 (the firm has no global sales and all of its total sales come from the home region segment) to 1 (the firm has no home region sales and all of its total sales come from the global segment). Thus, higher values of GT designate a firm focused more on its global segment than its home regional segment. However, because the three geographic ratios that we use are all out of total sales, they allow us to capture the shift across the three geographic

E.R. Banalieva, M.D. Santoro segments toward the prevailing geographic segment. For instance, if a firmÕs L/T = 10% and R/T = 60%, then its G/ T = 30% since G/T = 100  (L/T + R/T) and this firm will have a regional orientation. Conversely, if a firmÕs L/T = 30% and R/T = 10%, then its G/T = 60% since G/T = 100  (L/T + R/T) and this firm will have a global orientation.

Control variables At the firm level, we controlled for Firm Size (the natural log of total sales in thousandths of USD), to account for economies of scale; Firm Leverage ((current liabilities + non-current liabilities)/shareholdersÕ equity), to control for the differences in capital structures across firms; Firm Age (the natural log of the number of years since incorporation), to account for learning effects Pthat accrue with age; and Industrial Diversification ð1  ðsi Þ2 where si is the share of sales from business segment i (Berry, 1971)). At the industry level, we controlled for whether the industry the firm operates in is High Tech or not. We did this by assigning 1 to firms based in industries belonging to the high tech sectors as defined by the Advancing the Business of Technology agency6, and 0 otherwise. At the home country level, we controlled for the Home Country Market Size with the natural log of the GDP of the firmÕs home country measured in current USD (Hennart, 2007; World Bank, 2008). At the regional level, we followed prior scholarship (e.g., Hejazi, 2007) and controlled for the Intra Regional Attractiveness of each of the four regions in our sample by the difference between (regional GDP-to-world GDP) minus (local GDP-to-world GDP) (World Bank, 2008), keeping it consistent with our operationalization of the Intra Regional Orientation variable. We also controlled for possible Region of Origin effects by including three 0–1 dummy variables for Africa, Europe, and the Americas, and performing the analysis with respect to Asia. Finally, since this is a longitudinal dataset, we controlled for year effects by including six dummy terms (coded 0 or 1) for years 2001 through 2006 and performing the analysis with respect to year 2000 to neutralize potential effects common across all firms in a given year (Li, 2005).

Statistical analysis The Hausman test suggested that fixed effects are better than random effects. However, most of the independent and control variables in our study are either quasi-time invariant (Global Orientation, Firm Age, Industrial Diversification, Firm Size, Home Country Market Size, Intra Regional Attractiveness), i.e., slowly changing over time, or strictly time-invariant (High Tech Sector and Region of Origin dummies). The presence of such explanatory variables makes standard fixed effects (FE) estimation infeasible and inefficient (Plu ¨mper and Troeger, 2007). First, the FE estimator cannot estimate the theoretically interesting but time invariant variables in our sample. Second, the FE estimator relies only on the within-variance of variables, thus using less information than is available to estimate the model. Third, Plu ¨mper and Troeger (2007) add that 6

http://www.aeanet.org/Publications/IDMK_definition.asp.

Local, regional, or global? Geographic orientation and relative financial performance of emerging market the inefficiency of the FE model in the presence of slowly changing independent variables will not only lead to higher standard errors but also to unreliable point estimates. Plu ¨mper and Troeger (2007) suggest employing the fixed effects vector decomposition (FEVD) estimator, developed particularly for FE estimation of panel data with such time invariant and/or quasi-time invariant explanatory variables. The FEVD estimator is superior to FE when estimating variables with little year-to-year variance. Following this rationale, we too employed the FEVD estimator in our regression analysis.

Descriptive statistics Figure 2 illustrates the characteristics of the EM MNEs in our sample frame in terms of Region of Origin (Europe, Asia, Americas, and Africa) vs. Geographic Orientation (Local, Regional, and Global). As a first general observation, Figure 2 illustrates that region of origin represents an interesting variation in our data, as some firms based in some regions are more local than other firms based in other regions. Although the patterns seem similar for firms in the Americas, Europe, and Asia where the full range of local, regional and global strategies is pursued, they occur with varying concentrations. For instance, most African EM MNEs are local, whereas most European EM MNEs are regional. These findings suggest that it is possible the Intra Regional Attractiveness is larger for the European firms due to the higher integration benefits within the region, whereas the home countries seem to be more attractive for African-based MNEs perhaps due to their larger size. Second, as the data are weighted by Firm Size, Figure 2 also punctuates that Firm Size is an important consideration and a necessary control in our regression analyses, as larger firms seem to group around the regional (in particular European MNEs) and global (in particular Americas MNEs) orientations. We also

349

found that while the larger African firms seem to prefer the local orientation, the smaller African-based firms followed mostly the regional and global orientations, which may be a result of searching for new opportunities on which their larger counterparts have not yet focused their attention. It is also interesting to note that it is the relatively larger EM MNEs that adopt a local orientation (except for the European MNEs), but it is the relatively smaller EM MNEs (except the AmericasÕ MNEs) that are able to achieve a global orientation. Table 1 reveals the descriptive statistics and correlations of the main constructs in our study. The average EM MNE in our sample had nearly $5 mill in sales (computed by exponentiating the ln-value of 11.46), 15% of its total sales came from the global segment, and the average Firm Age was 16 years (computed by exponentiating the ln-value of 2.78). The average market size of the economies in the sample was $218 billion (computed by exponentiating the ln-value of 26.11). It is interesting to note that the average EM MNE in the sample was also more locally than regionally oriented, as evident by the negative value of 0.22 for the intra-regional orientation variable. Lastly, on average, the Intra Regional Attractiveness variable was positive (0.24), which means that the international home regions were more attractive for business than were the home countries. A review of the correlations among the variables reveals that multicollinearity is not a problem since the highest correlation that we observed among any two variables was 34% between Firm Size and Firm Age, which is well below the threshold of 80% used as a rule of thumb for collinear relationships (Judge et al., 1980, p. 459). Table 2 synthesizes the main characteristics of the EM MNEs in each geographic orientation. It shows that the Local Orientation is the most prevalent for the EM MNEs—64% of the sample, followed by the Regional Orientation with 23%, and the Global Orientation with the remaining 13%.

Americas

1

Africa

0

REGION OF ORIGIN

REGION

Asia

2

3

Europe

1 Local

1.5

2 2.5 Regional GEOGRAPHIC ORIENTATION

3 Global

GEOGRAPHIC ORIENTATION

Figure 2 The EM MNEsÕ geographic orientations by region of origin and firm size. Note: Circles represent firms. Larger circles designate bigger firms. Bolder circles mean that more firms are concentrated following that strategy.

350

E.R. Banalieva, M.D. Santoro

Table 1

1 2 3 4 5 6 7 8 9 10

Descriptive statistics and correlations.

Relative financial performance Intra regional orientation Global orientation Firm leverage Firm age Industrial diversification Firm size Intra regional attractiveness Home country size High tech sector

Mean

SD

1

2

3

4

5

6

7

8

9

0.00 0.22 0.15 1.20 2.78 0.16 11.46 0.24 26.11 0.39

0.11 0.41 0.24 2.87 0.83 0.23 1.64 0.05 0.79 0.49

1.00 0.05 0.12 0.08 0.03 0.07 0.13 0.06 0.10 0.03

1.00 0.19 0.02 0.04 0.09 0.03 0.13 0.02 0.11

1.00 0.01 0.07 0.03 0.04 0.02 0.09 0.14

1.00 0.05 0.05 0.09 0.03 0.01 0.08

1.00 0.08 0.34 0.11 0.02 0.14

1.00 0.08 0.13 0.20 0.26

1.00 0.08 0.24 0.11

1.00 0.15 0.02

1.00 0.18

Note: Bold indicates significance at 5%. Firm Age and Firm Size are represented in natural logs. N = 2, 105.

Table 2

EM MNEs characteristics by geographic orientation. Local orientation

Regional orientation

Global orientation

Relative financial performance Firm leverage Firm age Industrial diversification Firm size

0.6% 1.14 2.80 0.17 11.54

0.5% 1.25 2.81 0.12 11.49

3.8% 1.38 2.62 0.15 11.03

Firm-year obs.:

1348

481

276

Note: Firm Age and Firm Size are in natural logs. Presented are the averages for each variable in each strategy.

Our finding that Global Orientation is the least popular geographic orientation for EM MNEs to attain is consistent with prior research on Triad-based Fortune 500 Global MNEs (Rugman and Verbeke, 2004). Thus, our study, along with other prior works, reinforces that a Global Orientation poses notable difficulties to not only MNEs in developed economies, but also to those in emerging economies (Li, 2005; Rugman and Verbeke, 2004, 2007, 2008). Additionally, Table 2 reveals that the EM MNEs which were globally oriented had the lowest relative financial performance (3.80%) and the EM MNEs which were locally oriented had the highest relative financial performance (0.6%), providing preliminary evidence for the negative effect of global orientation on firm performance. This can, in part, be due to the fact that the local and regional EM MNEs were slightly older and larger than their globally oriented counterparts, and hence were more established, successful companies. Across the three geographic orientations, the EM MNEs were very little diversified industrially, which is not surprising, given the emerging economy nature of our sample. Table 3 presents the longitudinal trends of the geographic orientations by region of origin, and the subsequent change of observations per geographic orientation between 2000 and 2006. For the locally-oriented firms, the Americas MNEs were the only ones that decreased their local orientation between 2000 and 2006, and did so drastically—with a 900% decrease over time. Asian and European MNEs increased their local orientations the most—85% and 70%, respectively—for the same period. African MNEs also became

more locally oriented over time, but not by much—they increased their local orientation by only about 29%. For the regionally-oriented firms, European and Asian MNEs increased their regional orientation the most, with 100% and 89.26%, respectively, between 2000 and 2006. There was no change in the regional orientation of Americas MNEs while African MNEs slightly decreased theirs for the same period. For the globally-oriented firms, Americas and European MNEs were not following a global orientation in 2000; however, in 2006, Americas MNEs had already managed to move in the global direction. Additionally, Asian MNEs increased their global orientation by 73.21% between 2000 and 2006. Finally, for the entire sample across all regions, it was the regional orientation that experienced the largest increase between 2000 and 2006—close to 90%, followed by the local orientation at 81.31% and the global orientation, which increased by 74.58%. These results confirm the regionalization hypothesis since we found that EM MNEs stick to their home regions rather than operate on a global scale during the period, consistent with recent findings on Triad-based MNEs (Delios and Beamish, 2005; Rugman, 2005; Rugman and Verbeke, 2004; 2008). On the other hand, our results in Table 3 also extend the regionalization hypothesis advanced to date because they show that while the home region is important for EM MNEs, the entire region is not equally important, but it is mostly the local segment of the EM MNE that is key, as the local orientation was the most popular among the EM MNEs in our study. This finding is also supported by Hejazi

Local, regional, or global? Geographic orientation and relative financial performance of emerging market Table 3

351

Longitudinal trends of geographic orientation by region of origina. 2000

2001

2002

2003

2004

2005

2006

Change (%)

Local orientation Africa Americas Asia Europe

5 10 45 3

8 8 52 6

9 9 95 6

9 11 121 8

7 1 234 12

8 1 329 14

7 1 319 10

28.57 900.00 85.89 70.00

Total

63

74

119

149

254

352

337

81.31

Regional orientation Africa 1 Americas Asia 16 Europe

1 2 17

1 2 28 4

1 3 37 3

1 67 6

1 1 116 10

149 14

0.00 89.26 100.00

17

20

35

44

74

128

163

89.57

1 56 2

0.00 100.00 73.21 100.00

59

74.58

Total

Global orientation Africa Americas Asia 15 Europe Total *

15

1

2

1 25

33

38

49 2

1 48 2

26

34

38

53

51

Numbers in cells designate firm-year observations. Change is calculated between years 2006 and 2000.

(2007, p. 5) who also found that, in a US context, most home regional activities of US-based MNEs are ‘‘driven by the national dimension.’’ As Hejazi (2007) further notes, this importance of the local geographic orientation is consistent with economic theory, as when firms tend to expand domestically, they do so in response to meet local consumer needs and preferences, which are what they know best. This notion also makes sense in an EM context, because EM MNEs are typically late-comers to the internationalization stage (Luo and Tung, 2007).

Results Our regression results are presented in Table 4. Models 1 through 3 are biased specifications as they did not include all of the key independent variables. Model 4 includes all independent and control variables used to test our hypotheses, while Model 5 was included to show the interaction term of Intra-Regional Orientation · Global Orientation. Although we did not offer a formal hypothesis for this interaction term, we added this interaction term to this model in order to examine if the interplay of these two geographic orientations had a significant effect on relative financial performance. Since our results were consistent across all model specifications, we tested our hypotheses using Model 5 because it had the highest R squared, and because it included the statistically significant interaction term, suggesting it has important implications for firmsÕ relative financial performance. In Model 5, the negative and significant (p < 0.001) coefficient of intra-regional orientation denotes that the more regionally-oriented the EM MNE is and conditional on the firm having no global sales (i.e., G/T = 0), the lower its financial performance relative to its industry. In other

words, when EM MNEs have a higher local orientation and expand less outside of their home countries but within their home regions, they perform better relative to their industries. Thus, Hypothesis 1 is supported. Additionally, the negative and significant (p < 0.01) coefficient of global orientation suggests that the more globally oriented an EM MNE becomes, conditional on the firm having no preference for either local or regional orientation (i.e., R/T = L/T), the lower its financial performance relative to its industry. This finding supports Hypothesis 2 as well. Finally, we included and examined the interaction term of Intra-Regional Orientation · Global Orientation and found a positive and significant (p < 0.05) interaction coefficient, which suggests that when the firm is more locally than regionally oriented, its relative financial performance is higher with higher global orientation. To further examine this significant interaction effect, we graphed the predicted values from the relative financial performance in Model 5 on the vertical axis against the intra-regional orientation on the horizontal axis in Figure 3 at high and low levels of global orientation. More negative (i.e., to the left of 0) values of intra-regional orientation designate a more local orientation. Conversely, more positive (i.e., to the right of 0) values designate a more regional orientation. Figure 3 indicates that when the EM MNEs maintain a combination of higher local and global orientations, they experience superior performance gains relative to their industry rivals that stay concentrated within their home regions. Interestingly, when the firm has a low global orientation (the solid line in Figure 3), meaning most of its total sales come from the home region, its relative financial performance exceeds that of its industry as long as the EM MNEÕs local orientation does not exceed its regional orientation by more than 0.5.

352 Table 4

E.R. Banalieva, M.D. Santoro Regression results.

Control variables Firm leverage Firm age Industrial diversification Firm size Intra regional attractiveness Home country market size High tech sector Africa Americas Europe Eta Constant

Model 1

Model 2

Model 3

Model 4

Model 5

0.000 (0.735) 0.014*** (5.986) 0.034*** (4.142) 0.011*** (8.339) 0.511** (3.201) 0.014*** (3.587) 0.018*** (4.802) 0.154*** (3.806) 0.112*** (4.831) 0.045** (2.981) 1.000*** (49.417) 0.585*** (4.297)

0.000 (0.582) 0.014*** (6.139) 0.038*** (4.713) 0.011*** (8.515) 0.525*** (3.305) 0.012** (3.003) 0.015*** (3.837) 0.145*** (3.596) 0.110*** (4.771) 0.041** (2.711) 1.000*** (49.736) 0.546*** (4.031)

0.000 (0.675) 0.014*** (6.194) 0.032*** (3.956) 0.011*** (8.464) 0.437** (2.730) 0.011** (2.854) 0.014*** (3.770) 0.136*** (3.346) 0.104*** (4.473) 0.041** (2.682) 1.000*** (49.012) 0.488*** (3.559)

0.000 (0.529) 0.014*** (6.267) 0.037*** (4.593) 0.011*** (8.591) 0.479** (3.002) 0.010* (2.540) 0.012** (3.188) 0.134*** (3.302) 0.105*** (4.543) 0.038* (2.522) 1.000*** (49.377) 0.486*** (3.558)

0.000 (0.518) 0.015*** (6.408) 0.035*** (4.326) 0.011*** (8.696) 0.485** (3.041) 0.011** (2.699) 0.013*** (3.410) 0.140*** (3.460) 0.107*** (4.614) 0.037* (2.466) 1.000*** (49.364) 0.506*** (3.698)

0.042*** (5.666)

0.046*** (10.402) 0.026*** (3.496)

0.051*** (10.092) 0.024** (3.168)

Independent variables Intra regional orientation

0.046*** (10.649)

Global orientation Intra regional orientation · global orientation

0.081* (2.566) R squared N

0.660*** 2105.000

0.663*** 2105.000

0.660*** 2105.000

0.663*** 2105.000

0.664*** 2105.000

Note: t-Statistics in parentheses. All regressions control for year effects (not shown to preserve space). Eta is a coefficient automatically added and estimated by the Stata software as part of the FEVD estimation procedure. * p < 0.05. ** p < 0.01. *** p < 0.001.

Among the control variables in Model 5, larger firms had higher relative financial performance than smaller ones, as evident by the positive and significant (p < 0.001) coefficient of Firm Size. However, the EM MNEs that were older (p < 0.001), more industrially diversified (p < 0.001), and were operating in the High Tech sector (p < 0.001) experienced inferior performance relative to their industry. A larger home country market size also enabled firms to perform better, as evident by the positive and significant (p < 0.01) coefficient of home country market size variable. Additionally, when the regional market was more attractive than the home country market, firms experienced superior performance (p < 0.01). Among the region of origin dummies, African EM MNEs performed significantly (p < 0.001) better than

their Asian counterparts, while American (p < 0.001) and European (p < 0.05) EM MNEs performed significantly worse than their Asian counterparts. Lastly, Firm Leverage did not significantly affect the relative financial performance of the EM MNEs examined here.

Discussion Contributions to theory and future research In this study, we offered a framework of firmsÕ geographic orientations and linked these orientations to relative financial performance within the context of EM MNEs. In

Local, regional, or global? Geographic orientation and relative financial performance of emerging market

353

0.03

Relative Financial Performance

0.02 0.01 0 -1-0.500.51 -0.01 -0.02 -0.03 -0.04 -0.05 Intra Regional Orientation High Global Orientation

Figure 3

Low Global Orientation

The interaction effect. Note: Negative (positive) values on the horizontal axis designate a local (regional) orientation.

our proposed framework, we presented a finer-grained classification of firmsÕ geographic orientations that encompass local, regional, and global orientations. A first major theoretical and research implication from this study is that geographic orientation is important, since we found a large percentage of the worldÕs EM MNEs that tend to be locally oriented and that they experience superior performance by staying close to home. This finding is in line with HennartÕs (2007) notion that attaining optimal scale economies does not automatically require a firm to venture internationally. Indeed, we found that MNEs from smaller-sized EMs, similar to their larger market counterparts, were also able to stay local and attain superior performance relative to their industry rivals. We also found that firms that leapfrog to a high global orientation, while maintaining their local orientation within their home region, experience superior relative financial performance relative to firms that are internationalizing both regionally and globally simultaneously. This suggests that EM MNEs still need to keep their home country bases as strongholds of their competitive advantages, even though they internationalize globally. Our finding within the context of EM MNEs contrasts with Delios and BeamishÕs (2005) work on Japanese MNEs, where they found that Japanese MNEs with at least 50 foreign direct investments pursing the global strategy had the highest performance, and with LiÕs (2005) research of US MNEs which shows that a regional orientation increases firm performance. Instead, we found that in an EM context, a regional orientation of the EM MNEs increases their relative performance only when global orientation is low. Our findings, however, do support the regionalization hypothesis, which posits that a global orientation is hard to achieve for Triad-based MNEs since we saw that very few of our sampled EM MNEs had a global orientation. Based on these findings, we suggest future research explore in further detail how and why certain firms are able to overcome LOGF and others are not, especially in the context of EM MNEs.

Second, with respect to our proposed notion of leapfrogging, we surmise firms prefer to keep strong local bases while internationalizing in more distant markets because locallyoriented firms are most familiar with their home country markets because they may have already concentrated most of their operations there. Conversely, when the company is already regionally oriented, having a higher global orientation decreases performance because the EM MNE is exposed to two different liabilities of foreignness simultaneously: LORF and LOGF. In other words, being highly regional and global at the same time distracts the EM MNE manager who may be somewhat inexperienced in the global scene, and poses too much uncertainty and risk for the MNE. When the firm is regionally oriented, it has focused most of its resources to master and control LORF and has learned how to operate in other risky home region neighbors, which tend to share more similarities than do non-home region countries. Thus, with the EM MNEÕs time and resources dedicated to its home region operations, the firm may suffer financially relative to its rivals, since it has stretched itself too thin by orienting itself both globally and regionally. Third, a number of scholars have grappled for some time with the relationship between internationalization and performance and have found inconsistent results (e.g., Geringer et al., 2000; Hitt et al., 1997; Kotabe et al., 2002; Li, 2005; Wan and Hoskisson, 2003). Our research shows that one possible reason for these mixed findings may be that previous studies have focused too heavily on the process of internationalization and overlooked the geographic orientation of the MNEsÕ segments and their relationships to performance. The international label has largely blended two different segments—regional and global—that we have parsed out, examined, and shown to have different implications for relative financial performance. Based on the results we found here, we surmise that because the regional and global orientations have different effects on relative financial performance, previous studies show mixed results perhaps because they ignored the specific prevailing

354 geographic orientations of the MNEs. Future studies can consider explaining the three geographic orientations proposed in this study—local, regional, and global—as dependent variables to determine the firm, industry, country, and region factors that influence each. Fourth, our findings provide a springboard for some potentially interesting future research. For example, future studies can consider how the intra-regional and global competitions among MNEs and other domestic firms affect relative financial performance and the competitiveness of the firmsÕ respective home regions. Along this line, some intriguing possible emerging research paths and future research questions include: Are EM MNEs prevailing geographic orientations due to mutual forbearance (i.e., EM MNEs may not act aggressively in some geographic segments to avoid possible retaliation from their competitors in those segments) or escalating competition (i.e., EM MNEs take aggressive action to undermine their competitors in the same geographic segment)? What compels EM MNEs to ‘‘divide and conquer’’ their geographic segments into zones of influence and pre-empt certain geographic segments and not others—insufficient strength of the firm-specific advantages, or unfavorable institutional environment conditions, or a combination of the two?

Contributions to management practice Our study brings important implications for managers of EM MNEs as well. First, managers of EM MNEs may want to consider that internationalization should not always be an automatic response to the prevailing notion of expansion through globalization of world markets since achieving higher profits by merely internationalizing within and beyond the home region is not an automatic outcome. Rather, the role of effective management in this process is paramount, since greater financial success, as it relates to multinationality, is often due to managementÕs developing and pursuing the proper strategies, along with employing the appropriate organizational structures that can facilitate these strategies (Hennart, 2007). Second, careful consideration as to where an MNE can expand geographically is required in order to attain greater financial performance over industry rivals. A local orientation is linked to greater performance, which may be due to the lower risk and uncertainty within the home country market, as compared to the regional and global markets, since the firm is doing business in more familiar surroundings in the way of geography, language, customs, and consumer preferences. We also suspect the firmÕs existing knowledge of their home market dynamics and the firmÕs ability to leverage this knowledge along with exploiting any political/legal advantages in their home market play a crucial role in facilitating greater performance in the local market (Hejazi, 2007). Third, counter to conventional thinking regarding possible expansion strategies, a continued emphasis in local markets along with leapfrogging, i.e., a local orientation combined with a high global orientation, may be a viable alternative. Following the notion of regionalization, the home region markets are often the first step for many firms

E.R. Banalieva, M.D. Santoro as they move beyond their domestic market to become ‘‘international.’’ Leapfrogging appears to be a beneficial strategy because firms that leapfrog bypass their overcrowded regional markets where perhaps many of their competitors have already established a first or second mover advantage. By leapfrogging, firms can be successful by venturing beyond their domestic and home region markets to reap first or second mover advantages in untapped global markets. Our findings also highlight that ‘‘regionalizing,’’ i.e., internationalizing within the home region, leads to performance gains only when done in combination with a low global orientation, and leads to performance decreases when done in combination with a high global orientation due to the increased exposure to LORF and LOGF simultaneously. This combination of global and local orientations is also suggested by prior research, which mentions the possibility that ‘‘[a]s the company moves into a globalization phase, it may still rely more heavily on its local market, which is more in line with the demands the company is best able to satisfy’’ (Hejazi, 2007, p. 7). In combination, we, therefore, offer EM MNE managers with a contingency perspective that underscores the notion that one size does not fit all when it comes to expanding a firmÕs scope of operations, but rather there are alternatives and conditions that warrant thoughtful consideration. As with any other study, our study has some limitations. First, our final sample frame was unexpectedly over-represented by Asian-based EM MNEs. Although we started with a large number of EM countries, our final sample was reduced due to the fact that not all firms reported the complete data for all variables needed for our analyses. Although beyond the scope of this current study, perhaps future research could examine in greater detail why Asianbased EM MNEs report more of their financial, geographic, and business segment data more consistently than their European, Americas, and African counterparts. Second, we only explored the market penetration strategies (i.e., sales) of EM MNEs. We think it might be interesting if researchers consider applying our proposed framework to other measures of MNE geographic involvement, such as total assets, number of employees, or patent output. This will provide an even broader view of the geographic orientation of EM MNEs. Third, our focus in this study was on the linkage between geographic orientation and financial performance, captured by ROA. While ROA as our dependent variable is a widely used measure to reflect firm performance, other aspects come to bear when considering a firmÕs full range of internationalization strategic intent. That is, even though our findings show a firmÕs focus on certain geographic locations over others are more positively or negatively related to ROA, firms pursue various geographic locations for a variety of reasons. For instance, a firmÕs focus on local, regional, or global markets could be a result of and prove beneficial as they consider other key objectives, such as learning, obtaining crucial technologies and resources, and/or gaining economies of scale, deemed necessary for improving the overall value of the firm. These additional aspects and their linkage to geographic orientation could be concentrated on and further teased out in future studies. Fourth and finally, future research may extend our framework to examine the linkage between developed market MNEsÕ international diversification and performance for a

Local, regional, or global? Geographic orientation and relative financial performance of emerging market longer time frame than the 7-year period employed in this study to increase the generalizability of our results.

Conclusion To conclude, we developed and examined a framework to determine MNEsÕ prevailing geographic orientation by taking into account the intra-regional and global orientations of EM MNEs. We presented our conceptual framework that expanded existing theory by including a global segment and by further breaking down the broad home region into local and regional segments. Our results show that in the context of EM MNEsÕ market penetration strategies, the local orientation is the most prevalent, followed by the regional and finally global orientations. We also confirmed that when EM MNEs leapfrog and maintain both a local and a global orientation simultaneously by keeping their regional orientation low, they gain superior financial performance.

Acknowledgements We have benefited greatly from useful comments by Alan Rugman, K. Sivakumar, Christian G. Asmussen, Ravi Sarathy, Paul Gooderham, Peter Buckley, Ron Rivas and other participants of the 2007 JIBS/AIB PDW in Indianapolis, IN. We also would like to thank the three anonymous reviewers of this Special Issue whose comments helped us significantly improve our work. The usual disclaimer applies.

References Berry, C. (1971) Corporate growth and diversification. Journal of Law and Economics 14(2), 371–384. Cho, M.-H. and Cohen, M. (1997) The economic causes and consequences of corporate divestiture. Managerial and Decision Economics 18, 367–374. Collinson, S. and Rugman, A. M. (2008) The regional nature of Japanese multinational business. Journal of International Business Studies 39(2), 215–230. Delios, A. and Beamish, P. (2005) Regional and global strategies of Japanese firms. Management International Review, 19–36, Special Issue. Economist. (2007) Dizzy in boomtown. http://www.economist. com/finance/displaystory. cfm?story_id=10136509, Accessed Dec. 1, 2007. Elango, B. (2004) Geographic scope of operation by multinational companies: An exploratory study of regional and global strategies. European Management Journal 22(4), 431–441. Geringer, J. M., Tallman, S. and Olsen, D. M. (2000) Product and international diversification among Japanese multinational firms. Strategic Management Journal 21(1), 51–80. Hejazi, W. (2007) Reconsidering the concentration of US MNE activity: Is it global, regional, or national? Management International Review 47(1), 5–27. Hennart, J.-F. (2007) The theoretical rationale for a multi-nationality-performance relationship. Management International Review 47(3), 423–452. Hitt, M. A., Hoskisson, R. E. and Kim, H. (1997) International diversification: effects of innovation and firm performance. Academy of Management Journal 40(4), 767–798.

355

Hoskisson, R. E., Eden, L., Lau, C. M. and Wright, M. (2000) Strategy in emerging economies. Academy of Management Journal 43(3), 249–267. Judge, G., Griffith, W., Hill, C. and Lee, T. (1980) The theory and practice of econometrics. Wiley, NY. Kotabe, M., Srinivasan, S. S. and Aulakh, P. S. (2002) Multinationality and firm performance: The moderating role of R&D and marketing capabilities. Journal of International Business Studies 33(1), 79–97. Li, L. (2005) Is regional strategy more effective than global strategy in the U.S. service industries? Management International Review 45, 37–57. Luo, J. and Tung, R. (2007) International expansion of emerging market enterprises: A springboard perspective. Journal of International Business Studies 38(4), 481–498. Naspers. (2006) Annual Report. Osegowitsch, T. and Sammartino, A. (2008) Reassessing (home-) regionalization. Journal of International Business Studies 39, 184–196. Plu ¨mper, T. and Troeger, V. E. (2007) Efficient estimation of timeinvariant and rarely changing variables in finite sample panel analyses with unit fixed effects. Political Analysis, 15. Rugman, A. M. and Verbeke, A. (1993) Foreign subsidiaries and multinational strategic management: An extension and correction of PorterÕs single diamond framework. Management International Review 2, 71–84, Special Issue. Rugman, A. M. and Verbeke, A. (2004) A perspective on regional and global strategies of multinational enterprises. Journal of International Business Studies 35(1), 1–15. Rugman, A. and Verbeke, A. (2007) Liabilities of regional foreignness and the use of firm-level versus country-level data: A response to Dunning et al. (2007). Journal of International Business Studies, 1–6. Rugman, A. M. and Verbeke, A. (2008) The theory and practice of regional strategy: a response to Osegowitsch and Sammartino. Journal of International Business Studies 39, 326–332. Rugman, A. M., Yip, G. and Jayaratne, S. (2008) A note on return on foreign assets and foreign presence for UK multinationals. British Journal of Management 19(2), 162–170. Wan, W. P. and Hoskisson, R. E. (2003) Home country environments, corporate diversity strategies, and firm performance. Academy of Management Journal 46(1), 27–45. World Bank. (2008). ELITSA R. BANALIEVAÕs research focuses on the regional/global strategies, internationalization-performance, and strategic alliances and divestitures of MNEs.

MICHAEL SANTOROÕs principal research interests are in the areas of strategic alliances and the external sourcing of knowledge and technological innovation.