سرمایه گذاری مستقیم خارجی ظاهری در چین: انتخاب محل سکونت و مالکیت شرکت

This article evaluates the international location decisions made by public listed Chinese firms during the period 2006–2008, using a Poisson count data regression model. Further, we categorize the firms into state-controlled and privately owned according to majority ownership. We find that the determinants of internationalization differ based on ownership. State-controlled firms are attracted to countries with large sources of natural resources and risky political environments. Private firms are more market seekers. Although all firms have strategic intent, the attraction is commercially viable technology rather than core research content. Our findings also show that existing theories can sufficiently explain the actions of private Chinese firms, but adjustments are needed to understand the behavior of state-controlled multinationals.

Outward FDI from emerging economies has been considered to be one of the “big questions” in the 21st century International Business research agenda (Mathews, 2006). Two reasons motivate this agenda. First, FDI outflow from developing countries (e.g. China, India, and Malaysia) has increased dramatically in recent years. They accounted for 16% of global outward FDI in 2008, up 13% from 2007 (UNCTAD, 2009). Second, the ability of multinationals (MNCs) from the developing world to invest abroad seems to defy the fundamentals theories of internationalization (Child & Rodrigues, 2005). How could firms without obvious ownership advantages succeed to become among the world's largest firms in their respective industries? This predicament has mooted several researchers to develop new theories. At the very least an overhaul of existing theories is warranted (Buckley et al., 2007, Liu et al., 2005 and Mathews, 2006). No emerging economy has received as much attention as China—whether by researchers or popular media. Three reasons explain the limelight that China gets. First, until recently, China has been known as a destination of global investment. However, since 2003, investment abroad by Chinese firms has increased substantially (see Fig. 1). In 2008, outward FDI from China surged to USD 52 billion, up 132% from 2007, making it the 13th largest source of capital in the world and third among developing countries (UNCTAD, 2009). However, the size of China's OFDI has to be seen in perspective. Averaging the amount between 2006 and 2008, China's OFDI is a little more than a tenth of FDI from the US, about a fifth of the UK and Germany, and a little more than a half of Hong Kong. Despite the relatively small volume of investment, a great deal of publicity has been generated by the actions of Chinese companies, particularly through M&As of several high profile targets.1 Second, Chinese firms seem to be investing into countries that do not fit the standard profile of host locations. Since 1991, there have been some dramatic changes in the geographical distribution of China's OFDI. In the 1990s, Canada, the U.S. and Australia hosted about 40% of Chinese OFDI, but by 2005, the proportion had reduced to 10%. In contrast, developing countries, particularly in Asia and South America, accounted for nearly 90% in 2005. In 2008, Asia continues to dominate (mainly due to flows to Hong Kong) while countries in the African continent accounted for nearly 10% of investment flows (see Table 1). A cursory look at the top destinations of China's OFDI reveals a rather strange set. As shown in Table 2, OFDI tends to flow to tax havens like Cayman Islands, neighboring territories like Hong Kong as well as untypical destinations like Laos, Nigeria and Mali. Third, the change in the industrial distribution of OFDI in recent years is also equally dramatic. Manufacturing accounted for nearly 60% of OFDI in the 1990s (Cheung & Qian, 2009), but had dropped to a mere 3% by 2008. Despite sensational media reports, the mining sector accounted for only about 10% of OFDI in 2008 (see Table 3). Sectors that seem to be gaining momentum are in the services sector, particularly business services, finance and retail. The objective of this paper is to consider one aspect of China's OFDI, namely its locational determinants. Despite a few attempts by others (Buckley et al., 2007, Cheng and Ma, 2008, Cheung and Qian, 2009 and Kolstad and Wiig, 2009) to consider this issue systematically, results are inconclusive. The scarcity of such studies is due to limited time series in Chinese OFDI on the one hand,2 and the type of outflow data used on the other.3 Perhaps it is for these reasons that other researchers have resorted to case based analysis to unravel the motivations behind the internationalization process of Chinese firms (Deng, 2007, Deng, 2009 and Rui and Yip, 2008). In this paper, we attempt to merge these two aspects – where and why – of Chinese OFDI by using a unique dataset of listed Chinese companies. Our dataset, which comprises the location of individual firm's FDI, allows us to segment internationalizing Chinese firms into state-owned and private. Thus, the determinants of location decisions based on ownership can be examined. In addition, by utilizing count data, i.e. the number of investment projects in country i, we avoid the tax haven problem that has plagued other similar studies ( Cheng and Ma, 2008, Kolstad and Wiig, 2009 and Morck et al., 2008). Our findings suggest that there is indeed a need to reconsider the existing theories of international location choice, particularly when OFDI of firms from developing country like China is brought into the equation. More specifically, we find that some determinants of location choice among state-owned Chinese firms are inconsistent with existing theories. Our results have a direct implication to host countries intending to attract greater Chinese FDI. Given the differences between state-owned and private firms, targeting the right type of firms based on their locational determinants becomes imperative.

We considered both the Poisson and negative binomial models, and the results are quite similar. Our final interpretation is based on the Poisson model (see Table 6) because the likelihood-ratio tests for over-dispersions (H0: α = 0) for the binomial models were insignificant, implying that the Poisson counterparts may be more appropriate.8 The estimated coefficients for the various models reveal that the criteria for choosing investment locations among firms with different ownerships are not the same.9 When the complete set of data, i.e. countries where Chinese firms irrespective of ownership type, are considered, model ALL(1) of Table 6 shows that eight variables are significant at least at the 10% level. Ownership of natural resources, the importance of the host country to China's exports and its proximity to China are significant at the 1% level. The host country average level of income, its level of technology exports and a sizeable Chinese population are significant at the 5% level. Although the number of patents registered in the host country is significant at the 5% level, it carries a negative signs. On the other hand, the alternative proxy we used to capture technology has the correct sign and is significant. Finally, OFDI is attracted to countries that are politically risky since the co-efficient is negative. Our overall results are quite consistent with those of Buckley et al. (2007). Our empirical results confirm H1 that Chinese firms are indeed attracted to natural resource rich countries. The speed of economic growth that China has experienced over the last three decades and the projected growth in the future means continued demand for natural resources (Davies, 2009). For instance, since 2009 China has become the largest market for passenger cars. This means an increased demand for oil in addition to the energy required to power its industries. Where does China source its natural resources? Our results also accept H2. Chinese investments are attracted to countries that are politically risky. We confirm the results obtained by others (Buckley et al., 2007 and Kolstad and Wiig, 2009) that countries (e.g. Sudan, Mali, Lao PDR) which are less than desirable by other (western) investments have become important destinations for resource seeking Chinese companies. The trend of heading to these risky countries that seem to have started after 1992 (Buckley et al., 2007) seem to have continued into this century. The interaction variable in model ALL(2) in Table 6, i.e. political risk and natural resource endowment turns out to be insignificant. This would mean we have to reject H3, as we are unable to find evidence to show that Chinese investments are attracted to risky countries to tap on their natural resources. Our results also accept H4 that Chinese firms are motivated to improve their competitive disadvantage in innovation and technology. Chinese firms are attracted to countries that show superiority in these areas. However, our findings show that Chinese investment may not be attracted to countries that excel in core research as our proxy, number of registered patents, shows a significant negative coefficient. Instead, Chinese companies are attracted to countries (e.g. Malaysia and Thailand) that are able to convert core research into commercially viable products and services since our alternate proxy (EXTECH) is significant and positive. Chinese firms seem quite pragmatic in the sense that bringing back core research home need not necessarily increase their core competencies if the human capital and other capabilities in China are unable to add value to this core technology. On the other hand, applied research could find more usage, and given the home advantage in the form of low cost production that these firms already have, a synergy can be created. However, the interacting variable in ALL(3) in Table 6 (i.e. political risk × exports of technology) turns out insignificant, and thus H5 is rejected. We find no significant evidence that the attraction to politically stable countries is in search of strategic assets. Turning now to the locational decisions made by the various types of Chinese firms, our results confirm H6a. Countries with large natural resource reserves (e.g. Indonesia, Ukraine and Tajikistan) are equally attractive to private firms. On closer scrutiny, it should be noted that none of the private Chinese firms with investment abroad in our analysis are mining firms. Instead, those that do invest in resource rich countries are heavy equipment companies like Sany Group Ltd. (for example with its investments in Ukraine) and Tebian Electric Apparatus Stock Co. Ltd. (or TBEA with its investments in Tajikistan). This suggest that private companies follow their state-owned counterparts by investing in natural resource rich countries and provide related products and services to the deals already made by their respective governments. The appetite for risk of private firms is not as large as state-owned ones. We accept H6b. Our results clearly show that both SASAC and SOELG controlled firms have a higher tendency to invest in countries with weak political systems. As mentioned earlier, private firms are relatively more risk averse and our results corroborate this view (H6f). Unlike their state-owned counterparts, private firms are more attracted to countries that are closer to home (e.g. Hong Kong), and weak political systems are not significantly attractive. The familiarity of Chinese SOEs to a highly regulated business environment at home provides a competitive advantage over other multinationals in politically risky countries. For instance, our results show the significance of the Chinese population in host countries (e.g. Indonesia) in location decisions among state-owned firms. This may imply that the use of international guanxi among the Chinese Diaspora seems prevalent. It should be noted that SOEs view political risks differently than private firms. IB literature tends to view undemocratic countries negatively because of the lack of institutions that can provide the legitimacy and rights required for long term investments (Globerman & Shapiro, 2002). However, SOEs rely more on the government-to-government (G2G) relationship as the basis of their decisions. Uncertainties like nationalization and contracts failures maybe less likely when the investment is based on a G2G foundation. Although the interaction variables were not significant when all firms were considered, some interesting results emerge when firms are segregated according to ownership. For SOELGs, the interaction variable in model SO(2) in Table 6 is negative and significant at the 5% level, while in all other cases no significant results are found. In other words, we can accept H6c partially as only SOELGs are attracted to political risky countries for natural resources. In like manner, H6d can be partially accepted because the interaction variable is only significant for SASAC controlled firms (see Model SASAC (3) in Table 6), indicating that they are attracted to politically stable countries for strategic asset seeking motives. These findings show a clear difference even between state-owned firms. To make a general statement that Chinese OFDI exploits countries with poor institutions and large natural resources, such as Sudan (Kolstad & Wiig, 2009) maybe premature as there is a need to consider the type of firm. The consequence of a large appetite for risks among SOELGs however can be seen from the proportion of local government controlled firms that incurred losses in 2009. Beijing Review reported that 28% of the SOELGs were in the red compared to 15% among central government managed enterprises.10 Weak corporate governance structures and risky management practices were among the reasons cited for this loss. Private firms are clearly market seekers (H6e). Our results show a significant positive coefficient for the size of the market (GDP). State-owned firms are also market seekers but there seems to be an attraction to relatively richer countries. In fact, SOELGs are more attracted to smaller rich economies. The significant outflow of investments from state-owned financial institutions like the Construction Bank of China, Industrial and Commercial Bank of China and Ping An Insurance to locations like Hong Kong, Luxembourg and the Netherlands are good examples of such investments. The high level of significance for the export variable further implies that Chinese firms, irrespective of ownership, are keen to invest in countries that are already a significant market for Chinese goods. In this sense, Chinese investment tends to follow traditional IB theories like Vernon's product life cycle. On the other hand, the intention to move upstream and control sources of imports seem to be significant only among private Chinese firms. Our findings also clarifies the findings of previous research (Child and Rodrigues, 2005 and Morck et al., 2008) in that the attraction of Chinese OFDI to emerging markets/developing economies due to their firm specific advantages is limited to private and SOELG firms. The larger SASAC controlled firms are attracted to richer developed markets. Finally, the arguments made by other studies (Deng, 2007 and Rui and Yip, 2008) that Chinese investments are made with a strategic intention of acquiring technology, brands, marketing, management and other know-how finds more support among state-owned firms. The intention of such acquisition is not only to increase the capacity to compete in the global markets, but also to maintain market shares at home. A large number of industries in China (e.g. utilities and mining) function in oligopoly market conditions with companies owned by the central and provincial governments vying against each other to attract the domestic business. To withstand the onslaught of other foreign competitors who are already in China or waiting at its doorstep, state-owned corporations are attracted to those countries (e.g. South Korea and Hong Kong) that are able to provide them with the competencies necessary for survival in China. Interestingly, technical superiority does not attract private firms. It is likely that at this stage of the internationalization process, private firms are more interested in market expansion using home grown competencies.