Japanese subsidiaries in the European Union: Entry modes and performance

Abstract Japanese Foreign Direct Investment (FDI) in the European Union and its performance were analysed in this work. Three different FDI or entry modes used by Japanese companies to enter the European market were compared, and the presence of a relationship between the selected entry mode and the performance of the subsidiary was investigated. We found that more than half of the Japanese investments in Europe took the form of new ventures, approximately 40% were joint ventures and less than 6% were acquisitions. We found that no specific entry mode performed better than another.


introduction
In this work, we study Japanese Foreign Direct Investment (FDI) in Europe and its performance. Japan is the world third's largest national economy. In July 2012, the European Commission (EC) asked the European Union (EU) Member States for their agreement to open negotiations for a free trade agreement with Japan. As a highly developed economy and major global trader and investor, Japan is an important partner for the EU (EC, 2012

RESEARCh GRoUp ovERvIEW
Jean-Louis Schaan is currently the Donald F. hunter professor of International Business at the Ivey Business School at Western University, Canada. he holds a phD in Strategic Management from the Ivey Business School and a MBA from the Telfer School of Management at the University of ottawa. he has extensive teaching experience in both degree and non-degree executive education. his research focuses on strategic alliances and international management, as well as advising executives internationally. David Tanganelli is associate professor of Applied Economics and International Management at the Universitat Internacional de Catalunya, Spain. his research issues include strategic alliances, international management and commercial diplomacy. he holds a phD in Economic and Business Sciences from the Universitat de Barcelona. he has been a visiting professor at Ivey Business School.

pUBLIC InTEREST STATEMEnT
This work compares and contrasts three different entry modes used by Japanese firms to compete in European markets and investigates the relationship between the selected entry mode and the performance of European subsidiaries. Over the period studied, more than half of the Japanese investments in Europe took the form of new ventures. Approximately, 40% were joint ventures and less than 6% were acquisitions. We found that no specific entry mode performed better than another.
We analyse different entry modes used by Japanese multinationals (MnEs) to enter the European market and investigate whether there is a relationship between the entry mode and the Japanese subsidiary performance. The choice of entry mode into foreign markets has received significant attention from international business researchers in recent decades (harzing, 2002) because selection of an adequate mode of entry in foreign markets is a critical and strategic decision (Arregle, hébert, & Beamish, 2006;Lu, 2002;Reus & Ritchie III, 2004).
Some studies have analysed factors that influence the choice of entry modes (harzing, 2002 for a revision) mainly focusing on three alternatives: licensing, joint ventures and wholly owned subsidiaries. If a wholly owned subsidiary is chosen, it could be a through full acquisition or through the establishment of a new facility (greenfield investment). Several researchers have analysed main factors that influence the selection of equity-based entry mode (Brouthers & Brouthers, 2000;hennart, Larimo, & Chen, 1995;hennart & park, 1993) usually underpinned by transaction cost theory.
This paper aims to contribute to the existing literature by providing additional information about the association between the selected entry mode and the performance of Japanese direct investment in Europe. We compare and contrast the relative performance of ownership-based foreign entry modes. Specifically, we update a previous analysis of the Japanese subsidiaries in Europe (nitsch, Beamish, & Makino, 1995, 1996 with new data, to examine the links between entry mode and financial performance. We considered three ownership-based entry modes (following hennart & Reddy, 1997; Kogut & Singh, 1988;Woodcock, Beamish, & Makino, 1994): greenfield investments (FDI where a parent company starts a new venture in a foreign country by constructing new operational facilities from the ground up), acquisitions of already existing companies and joint ventures with a European partner. A joint venture is a firm created and run by two or more parent firms with not only joint ownership from different firms but also joint control by parent firms (Leung & Cheung, 2007). The joint venture can be the result of the parent companies investing equity in a newly created joint company or one of the parents taking an equity position in its partner firm. The first two cases-greenfield and acquisition-are wholly owned modes, whereas the third is a shared ownership mode. Subsequent sections of the paper are structured as follows. First, we review previous research on entry mode and performance. Then, we describe data and our research methodology. Finally, we report our main findings and discussion.

entry Mode Performance: Previous Research
Among the many different theoretical lenses used in studies about entry mode, internalization and transaction cost theories are the most commonly found (Anderson & Gatignon, 1986;Chen, 2010;Erramilli & Rao, 1993;Gatignon & Anderson, 1988;hennart, 1988). The joint venture mode is selected to internalize a failing market of intermediate inputs when indivisibilities make full ownership of the relevant assets inefficient or when the assets are firm-specific public goods and the acquisition of the firm holding them add significant management costs. By contrast, the resource-based theory emphasizes the organizational capabilities of the firm and distinguishes between cost and value aspects in the analysis of decisions related to firm boundaries (Madhok, 1997). Third, the eclectic paradigm (Dunning, 1988) is a multi-theoretical approach for studying the choice of entry mode and permits researchers to use three determinants to predict the selected entry mode: ownership, location advantages and market internalization. Finally, institutional theory (Scott, 2001;Yiu & Makino, 2002) emphasizes the importance of the influence of institutional forces embedded in national environments in the choice of a foreign entry mode.
Several works have studied the relationship between entry mode and performance (Delios & Beamish, 2004 include a review). Using a sample from a single host country, Woodcock et al. (1994) found a strong statistical relationship between entry mode and performance when comparing pairs of entry modes. The authors suggested that greenfield investments generally perform better than joint ventures, which perform better than acquisitions. Using a smaller sample from 13 European countries, nitsch et al. (1996) supported the same entry mode hierarchy but with weak statistical significance. Li and Guisinger (1991) examined the performance of new ventures, joint ventures and acquisitions using failure rates as a performance measure. They found that acquisitions had a significantly higher failure rate than new ventures. however, the failure rate of joint ventures was not significantly different from the other two entry modes. Altogether, these studies suggest that multinational companies entering foreign markets through acquisitions were less likely to be successful compared with greenfield investments and joint ventures and there was no clear difference between the performance of greenfield investments and joint ventures. however, Makino and Beamish (1998a) showed that the relationships between entry mode and both financial performance and survival likelihood may not be as straightforward as has been suggested previously. Their results indicate that there was no consistent association between entry mode and financial performance across host countries in Asia. Further, Delios and Beamish (2004) reported that joint ventures had financial performance levels comparable to wholly owned subsidiaries.
Because we update the previous analysis of Japanese subsidiaries in Europe conducted by nitsch et al. (1995,1996) with new data, we too adopt the eclectic paradigm of international production (Dunning, 1988) to analyse the impact of ownership, location and internalization advantages on MnEs' decisions around FDI. First, among ownership advantages that contribute to enhance the focal firm's competitive advantage are tangible and intangible assets such as: skills, capabilities, property rights, technology, experience, brand, reputation, innovatory capacity, governance skills, specialized know-how about production and marketing expertise. Such advantages help MnEs exploit accumulated resources to gain scale economies, synergies, complementary resources and innovation capabilities as they select how to enter new foreign markets (Dunning & Lundan, 2008). Specifically, links have been established between these resource-based concepts and the choice of wholly owned modes (Kogut & Singh, 1988;Kogut & Zander, 1993), or the choice between wholly owned modes and shared ownership modes such as joint ventures (nitsch et al., 1996).
Second, location-specific advantages arising from the acquisition of natural resources, transport and communication infrastructures, strong market demand, economies of scale, low-cost factors and the abundance of innovative individuals, firms and universities influence the selection of potential and actual target countries (peng, 2009). To a certain extent, this study controls for location by considering only Japanese subsidiaries in Europe.
Finally, internalization explains the cost advantage of internal hierarchies over market transaction costs for intermediate products. When external costs are higher than the internal costs of establishing a hierarchical system of transactions controlled by the firm, MnEs achieve internalization advantages by entering new markets through FDI (Dunning & Lundan, 2008). In this paper, internalization relates to the control costs borne by a firm in a given entry mode.
This study analyses three ownership-based entry modes. of these, greenfield and acquisition are wholly owned modes, and the third, joint ventures, are characterized by shared ownership between Japanese and host-country parents. By comparing pairs of entry modes and using the internalization/control cost and ownership/resource requirements concepts from the Eclectic Theory, we propose a hierarchy of performance based on earlier entry mode research: greenfield entry will perform better than entry through joint venture which in turn will perform better than entry through acquisition (nitsch et al., 1996;Woodcock et al., 1994).
The argument supporting the first articulation of the hierarchy (that the greenfield entry mode is more successful than acquisitions) is that acquisitions require MnEs to engage more resources than greenfield operations throughout all phases in the life of the investment, from initial search for targets and due diligence to execution of the deal and ongoing monitoring and control post-deal. Information asymmetry between buyer and seller in a transaction makes it difficult for the buyer, no matter how thorough the due diligence, to ascertain the earning potential of the assets being acquired and thus to determine a fair price, the risks of over valuation and over paying as a result are significant. The transfer of an MnE's identity is easier and less costly in the case of a greenfield investment than it is in the case of the acquisition of a local company since the MnE will not have to incur the costs of integrating the corporate culture and management infrastructure, including dealing with legacy systems and processes which both contribute to make complex (and costly) the exercise of control. Further in most acquisitions managers on the buy side fail to capture potential revenue and cost synergies, thus increasing the financial burden associated with the financing of the deal. For these reasons, we hypothesize: hypothesis 1 (a): greenfield entry mode will perform better than acquisitions.
The second level of the performance hierarchy among foreign market entry mode compares greenfield investments with joint ventures. We argue that joint ventures will face higher ownership and managerial costs than greenfield entry mode. Creating a joint venture entails significant costs from the search for a partner, through the screening of potential partners, negotiations, due diligence, designing and structuring the joint venture to monitoring and controlling the joint venture. All of these activities are much simpler and efficient when there is no need to negotiate and compromise with one or more partners as it is the case for greenfield investments. Thus, we hypothesize: hypothesis 1 (b): greenfield entry mode will perform better than joint ventures.
The third level of the hierarchy entails comparing acquisitions with joint ventures. We argue that joint ventures will entail lower costs of resource acquisition than acquisitions. Information and overall transaction costs will be higher in the case of acquisitions with the MnE bearing these costs alone instead of sharing in the investment and in the control post closing. If the joint venture has been designed and structure in a way to meet mutual needs and expectations of the partners increases the likelihood that partners will trust each other and thus will need to allocate fewer resources to control the investment (Beamish & Banks, 1987). For all these reasons, we hypothesize: hypothesis 2: Joint ventures will perform better than acquisitions.

Data and Methodology
The data was collected from a survey of Japanese subsidiaries at the end of 2007 that appeared in Kaigai Shinshutsu Kigyou Souran (Japanese overseas Investments), an annual publication of Toyo Keizai Inc. that provided subsidiary-level information on the overseas activities of Japanese MnEs from 1985 to 2007. More than 90 articles in refereed academic journals over the past 20 years have drawn on this data-set (Arregle, hébert, & Beamish, 2009). This study used only information about subsidiaries in the EU whose parent companies were listed on the major Japanese stock exchanges.
Additionally, the sample selected was based on the availability of performance data. Financial performance data was not available but the database contained a subjective measure of performance. The performance of Japanese subsidiaries was measured by asking the top manager in each subsidiary to rate his/her subsidiary's performance by using a three-point scale of loss, breakeven and profitable. This performance measure is then comparable across host countries, firms and industries since respondents were all from Japan and were at similar levels in their organizations. The use of subjective, perceptual measures of performance is well supported in the literature. This particular measure has been used as a proxy for performance in a number of studies about the conduct of Japanese subsidiaries (Cieślik & Ryan, 2009;Kasuga, 2008;Kumarasinghe & hoshino, 2009;Makino & Beamish, 1998b;nitsch et al., 1996 among others).
our original sample considered the Japanese subsidiaries operating in the EU from 1985 to 2005. There were several motives for choosing this period. First, the Japanese subsidiaries in Europe from 1968 to 1994 were analysed in previous works (nitsch et al., 1995, 1996), and our analysis updates these works. Second, the Japanese FDI in the EU was studied at a time where the Union included 15 members (EU-15): Spain and portugal joined the EU in 1985, Finland joined in 1995 and Austria and Sweden followed, bringing the total number of countries to 15. In 2004, 10 more countries joined the EU. however, consistent with our criterion that a subsidiary must be at least two years old for its performance to stabilize, as several studies show (nitsch et al., 1996;Woodcock et al., 1994), we did not consider these subsidiaries in our sample. Subsidiaries were included in our sample if they were operating at the time of data collection, a single shareholder owned more than 5% of their equity (Beamish, Delios, & Lecraw, 1997) and reported clear and complete data and the subsidiaries were at least two years old. The final sample consisted of 146 Japanese subsidiaries operating in different industries that were classified as per our definition of joint venture, acquisition and greenfield, and for all cases we had performance data.
In this work, a descriptive analysis is used to show the entry modes selected by Japanese companies to operate in Europe, the performance of these subsidiaries and the target country of these investments. In addition, two tests were applied to determine if there was a statistically significant relationship between entry mode and performance. The pearson chi-squared test was used to see if there was a relationship between entry mode and performance. The pearson chi-squared is usually applied to categorical variables, and allows to test the differences between predicted and observed occurrences. The Wilcoxon rank sum test compared the performance of pairs of entry modes against one another. This non-parametric test has a Mann-Whitney U distribution and a power efficiency of 95.5%, and it can be used instead of other parametric tests, such as the unpaired T-test.

empirical Findings
The overall distribution of entry modes is summarized in Table 1 according to country, entry mode and performance. over the period covered by the study, the main form of ownership-based entry mode that Japanese subsidiaries used was greenfield investment with 79 subsidiaries, which represents 54.1% of the sample (see Figure 1). Joint venture was the second most frequently used mode, with 59 subsidiaries (approximately 40.4%). Acquisition was used in only eight cases (5.5%). Table 2 summarizes the main results of the entry mode and performance. While 70% of subsidiaries set up as a greenfield investment or joint venture reported a positive performance, only 50% of subsidiaries formed through acquisitions reported positive performance. The percentage of unprofitable subsidiaries was the lowest in joint venture mode.
In our study, the results of the pearson chi-squared test (p-value = .683) showed non-statistically significant differences between the different entry modes and performance for our sample (see Table 3). We applied Wilcoxon rank sum tests to the three pairs of entry modes corresponding to our hypotheses in order to contrast the greenfield > joint venture > acquisition performance hierarchy found in previous works. our data did not support the hypotheses related to the hierarchy: the p-values of these tests were greenfield/acquisition p = .266, greenfield/joint venture p = .725 and joint venture/acquisition p = .193. Thus, no significant difference appeared to exist between any pair of entry modes for our sample. hypotheses 1(a), 1(b) and 2 were not supported with our data. These results are partially similar to those of Delios and Beamish (2004), who found that joint ventures exhibited perceived financial performance levels comparable to wholly owned subsidiaries.

concluding Remarks
Do greenfield investments perform better than acquisitions? Do greenfield investments outperform joint ventures? Do we expect acquisitions to be more profitable than acquisitions? We found nonstatistically significant differences between any of these entry modes and performance. our sample did not support the hierarchy greenfield > joint venture > acquisition performance reported in nitsch et al. (1996). practitioners should take into account that the literature offers no clear theoretical predictions. however, if the conclusions from our study were validated elsewhere they would suggest that to understand subsidiary performance one needs to consider factors other than the entry mode. It may be a matter of execution, and we would find greater insights about the relative performance of alternative entry mode by comparing and contrasting successful vs. less successful  experiences in each mode type. our conclusions about no differences between the three entry modes analysed and performance disagreed with past studies except Delios and Beamish (2004), who reported similar results by comparing joint ventures with wholly owned subsidiaries.
Several explanations for this discrepancy can be offered. First, the samples used in Woodcock et al. (1994) andnitsch et al. (1996) only referred to manufacturing subsidiaries. our sample covered a broader diversity of industrial sectors, including wholesale, services, retail, construction, finance, transportation, electricity, gas and sanitary services. The aggregate analysis may have blurred differences between sectors. The influence of industrial factors on many aspects of FDI performance may have contributed to the difference in results. Second, the study by Woodcock et al. (1994) used a larger sample (321 Japanese subsidiaries) that focused on a single host country, whereas our sample included 146 subsidiaries from 14 countries. Therefore, locational factors may have influenced the results. Third, cases in our study were not as evenly distributed across categories as in the samples from Woodcock et al. (1994) andnitsch et al. (1996): our data only had eight cases in the acquisition category.
Generalizations should be stated with caution. Firstly, because of the characteristics of the sample noted above. Secondly, due to our performance measure, that is subjective or perceptual. Thirdly, because all findings are based on bivariate tests rather than a multivariate model as such tests do not control for other factors influencing performance.