Syndication and the performance of venture capital investments

Transcription

1 Syndication and the performance of venture capital investments Abstract This paper examines syndicates of local venture capital (VC) firms with foreign VC partners in an Asian setting. We find that post-syndication, local VCs have a higher likelihood of a successful exit from their investee companies. We interpret this as an outcome of the syndication in which local VCs undergo organisational learning by acquiring knowledge and skills from their foreign partners. Our empirical results also suggest that local VC firms are more likely to invest in high-tech industries after their syndication with foreign VCs. JEL G24, G32, G33, G34, G1 Key words: VC exit, cross-border deals, syndication, experience 1

2 1. Introduction Cross-border venture capital (VC) investment has become a trend in recent years (Aizenman and Kendall, 2008; Dai, et al. 2012). VC firms are actively expanding their operations internationally, particularly in Asian countries (Deloitte, 2006, 2007). Prior studies have examined various aspects of cross-border VC investment from foreign investors view point (Dai et al., These studies conclude that syndication or joint venture with local VC firms is an effective way for foreign VCs to alleviate information asymmetry and enhance their investment performance. However, the issue of whether the collaboration experience with foreign VC firms is also beneficial for the local VC firms remains unexplored. Foreign VC firms and their partnership with local VCs may add value to the expertise of local VC firms. The potential benefits might include adopting standard contracts, better knowledge on advising and nurturing entrepreneurs, and better knowledge on constructing deals and taking those firms to success exits (Dai et al., 2012). This paper aims to fill gap in the literature by examining whether syndicate experience with foreign VC firms benefit local VC firms in terms of investment selection process and performance. We use a sample of 3,966 investment rounds made by 2,094 local VC firms in 3,309 portfolio companies from 1996 to We investigated the performance of VC investments in portfolio companies that are located in Asia. These portfolio companies are mainly based in China, Japan, India, South Korea, Singapore, Hong Kong, Taiwan, Vietnam, Malaysia, Thailand, Indonesia, Philippines and Pakistan. Our empirical results show that local VC firms having syndicated with foreign VCs are more likely to invest in high-tech industries, in particular, information technology and telecommunication industries. Provided the fact that these local VC firms have been investing in non-high tech sector prior to their syndication with foreign VC firms; change in their investments preference from non-tech to high-tech might be due to the syndication experience. In terms of investment performance, in both deal and company level analysis, we find that syndication with foreign (mainly Europe and North America) VC firms 2

3 increases the likelihood of a successful exit for local VC firms. Similarly, the likelihood of successful exits increases post syndication with foreign VC firms as compared to pre-syndication. This suggests that local VC firms learn from their foreign VC partner and enhance the performance of their investments in terms of successful exits. These findings also provide evidence that local VC firms benefit from foreign VC syndication and the benefits are mutual between local and foreign VC firms. The rest of the paper is organised as follows. Section 2 highlights related studies and raises a few testable hypotheses; section 3 outlines and data and methodology; section 4 analyses the empirical results; and section 5 concludes the paper. 2. Literature and hypotheses 2.1 Cross-border VC investment Cross-border venture capital (VC) investment has become a trend in recent years (Aizenman and Kendall, 2008; Dai, et al. 2012). The international VC activity has been increasing manifold in terms of number of transactions and capital involved. VC firms are actively expanding their operations internationally, particularly in Asian countries (Deloitte, 2006, 2007). There is a large body of literature that examine various aspects of cross-border VC investment (Wang and Wang, 2011; Li et al., 2014). For instance, Dai et al. (2012) s multi-country analysis on Asian countries find that foreign VCs could alleviate information asymmetry caused by cultural and geographical distance by forming partnerships with local VCs. Similarly, by focusing solely on China, Jenner and Suchard (2013) conclude that foreign VCs could increase the likelihood of a successful exit through joint ventures with local VC firms. These studies have concluded that syndication or joint venture with local VC firms is an effective way for foreign VCs to enhance their investment performance. However, the formation of such partnership might be based on mutual interests. In other words, if foreign VCs benefit from such collaboration (e.g. lower level of information asymmetry), there should also be accrued benefits to local VC firms. Hence, a related research question is whether the collaboration experience with foreign VC firms also to some extent benefit the local VC firms? The presence of foreign VC firms and their 3

4 partnership with local VCs may help professionalise the local VC firms. The potential benefits to local VC firm might include adopting standard contracts, better knowledge on advising and nurturing entrepreneurs, and better knowledge on constructing deals and taking those firms to success exits (Dai et al., 2012). This study investigates whether syndicate experience with foreign VC firms influence local VC firms in terms of investment selection and successful exits. 2.2 Theoretical framework VC syndication has always been one of the typical and enduring characteristics of the VC industry (Tykvová and Schertler, 2011; Meuleman and Wright, 2011). Two or more venture capital firms co-invest in an investee firm and share a joint pay-off. In essence, syndicates are a form of inter-firm alliance (Wright and Lockett, 2003). There is a substantial amount of studies on strategic alliances in management literature. One of the issues that is related to our study is the objectives of alliances (i.e. why do firms form alliances with each other?). A variety of objectives have been suggested to explain firm s motives for forming alliances, including risk-reduction, economies of scale, access to markets, and the search for legitimacy (Contractor and Lorange, 1988; Harrigan, 1986; Hennart, 1988; Kogut, 1988). Similarly, in the venture capital literature, it is suggested that syndicates are formed due to the fact that VC firms seek to share risk (the risk-sharing perspective) or to access valuable resources (resource-based perspective). The risk-sharing perspective view syndication as a mean of diversifying risk without sacrificing the return (Lockett and Wright, 1999). Under this perspective, information asymmetries and adverse selection problems could be mitigated through syndication since joint decision enhances the accuracy of the assessment and provide increased deal flows at the pre-investment stage (Lerner, 1994). On the other hand, the resource-based perspective views syndication as a way of sharing resources among the VC firms (Hopp and Rieder, 2010). In such case, different information and heterogeneous skills brought by various VC could enhance the post-investment performance of the ventures (Ferrary, 2010). Apart from the incentives mentioned above, many scholars (Grant, 1996; Hamel, 1991; Khanna et al., 1998; Kogut, 1998) have suggested another important incentive for forming alliances. It is suggested that firms form alliances in order to form a 4

5 platform for organisational learning, which provides access to knowledge of their partner. Through the experience of mutual interdependence and problem solving, and observation of alliance activities and outcomes, participating firms are able to learn from their partners (Inkpen, 2000). Inkpen (1998) finds that the formation of an alliance is an acknowledgement that an alliance partner has useful knowledge, which can be used by the alliance partner to enhance its own strategy and operations. This type of knowledge is valuable to the partner firm even outside the specific terms of the alliance agreement. Alliances between firms provide a better platform for organisational learning than other contexts because it can reduce the risk that the knowledge will dissipate quickly (Powell, 1987). A number of empirical studies (Dodgson, 1993; Hamel, 1991; Inkpen and Crossan, 1995; Lane and Lubatkin, 1998; Mowery et al., 1996; Powell et al., 1996; Simonin, 1997, 1999) have also addressed the issue of learning through alliances. Within the context of cross-border venture capital investments, the formation of alliances/partnerships/syndicates between local and foreign VC firms may also be attributed to the learning motive. From the perspective of foreign VC firms, the inclusion of a local partner, who is geographically close to the investee company and has superior knowledge on the local market, technology, legal environments, linguistic skills and valuable contacts may help to reduce the level of information asymmetry caused by cultural disparity and geographical distance (Mäkelä and Maula 2006, 2008). In other words, foreign VC firms learn from the knowledge and experience provided by local VC firms and achieve lower level of information asymmetry and better investment performance (Jenner and Suchard, 2013; Dai et al., 2012). On the other hand, from the perspective of local VC firms, forming partnerships with foreign VC firms may allow them to access heterogeneous knowledge and skills and provide them opportunities to learn from their foreign partners. The learning opportunities for the local VC firms lie in the fact that the venture capital industry in Asia, in general, is rather young and underdeveloped (Dai et al., 2012). The number of local VC firms is small and they often operate on a small scale and not able to provide value-adding advice to their portfolio companies due to their limited experience (Bruton and Ahlstrom, 2003). Foreign VC firms, especially VC firms from North America and Europe have relative rich experience developed in their home 5

6 countries and are able to provide a larger amount of capital and better networks (Dai et al., 2012). Through the collaboration, local VC firms are able to access foreign VCs rich experience and expertise and acquire better knowledge on how to advise and nurture entrepreneurial firms and exit their portfolio successfully. 2.3 Syndicate experience and investment selection As discussed above, local VC firms may learn from their foreign partners during their collaboration and apply those acquired knowledge in their following-up investments. In other words, the syndicate experience may influence on local VCs investment behaviour. The first aspect of investment behaviour is investment selection. Foreign VCs are in general more experienced in investing in high-tech industries, especially in information technology firms. According to Cumming and Dai (2010), 64.3% of VC investments in the U.S. between 1980 and 2009 are in information technology sector (with the rest 18% in Medical and 18% in others). Asian VC firms, on the other hand, tend to invest more in traditional industries. As Dai et al. (2012) documented, about 30% of VC investments in Asia between 1996 and 2006 are in non-technology industries. That is to say, Asian VCs are less experienced and less likely to invest in high-tech industries. By working side-by-side with foreign VC firms who have extensive experience and knowledge of the high-tech industries, local VC firms could learn from their foreign partners and acquire necessary skills in selecting promising deals. Ultimately, in their following-up investments after the collaboration, they would be able to use those acquired knowledge to make more appropriate investment decisions and feel more confident in making investments in high-tech industries. Therefore, we hypothesise that local VC firms that have syndicate experience with foreign VC firms are more likely to invest in high-tech: Hypothesis 1a: Local VC firms with broad foreign syndicate experience are more likely to invest in high-technology ventures. In the previous hypothesis, we classify a VC firm as foreign and local. For example, if a Chinese VC firm syndicated with a Singapore VC, we will consider the Chinese VC as having a foreign syndicate experience. Despite that the lower cultural and institutional distance between Asian countries may lead to better collaboration 6

7 during the process, the learning capacity in this type of syndication might be limited. Because in general the Asian VC firms are less experienced and less developed than those from western economies (Dai, et al., 2012), the level of knowledge and expertise brought by another Asian VC firm into the syndicate might not be as extensive as those by western VC firms. For example, a Chinese VC firm may learn more by syndicating with an American VC firm than with a Taiwanese VC firm. Therefore, in order to more accurately measure the influence of foreign syndicate experience, in this case we only consider a syndicate experience with a western VC firm (i.e. VC firms from North America or Europe) as foreign. And we hypothesise that local VC firms with syndicate experience with western VC firm are more likely to invest in high-tech ventures: Hypothesis 1b: Local VC firms with western foreign syndicate experience are more likely to invest in high-technology ventures. 2.4 Syndicate experience and investment performance In the previous section, we posit that the syndicate experience with foreign VC firms would influence investment selection of local VC firms. Here in this section, we turn our focus onto the final stage of VC investment, which is exit. The issue of VC exits has been examined extensively in the developed market. Various studies including Cumming and Johan (2007), Giot and Schwienbacher (2007), Isaksson (2007), examine VC exits within different contexts. A large body of literature concludes that exits are influenced by various factors. Some studies (Cumming and Johan 2007 and Elisabete, Cesaltina and Mohamed 2008) report that the characteristics of VC firms and investee companies affect the exits. Others (Cumming, Fleming and Schwienbacher 2006; and Cumming and MacIntosh 2003) find that better economic condition and legal environment increase the likelihood of exits. In addition, VC syndication (Megginson and Weiss, 1991; Lerner, 1994; Giot and Schwienbacher, 2007) geographical distance and cultural disparity (Cumming and Dai, 2010; Buchner, Khurshed and Mohamed, 2012) also influence the VC exit within the context of cross-border VC investments. Recent studies on the Asian VC markets (Dai et al. 2012; Wang and Wang, 2012; Jenner and Suchard, 2013) find supportive evidence that a joint venture or a partnership with local VC lead to better investment 7

8 performance. However, previous studies examined the influence of syndicate on foreign investors but ignored the effect it may have on local VC firms. As discussed above, the knowledge and skills acquired by local VC firms may be used in their following-up investments. And the actual influence and effectiveness of those acquired knowledge will be ultimately reflected in the investment performance. If local VC firms have gained experience by syndicating with foreign partners, their investment performance might be improve. Therefore, we hypothesise that local VC firms syndicate experience with foreign (or western) VCs will increase the likelihood of a successful exit: Hypothesis 2a: Local VCs syndicate experience with foreign VCs will increase the likelihood of a successful exit. Hypothesis 2b: Local VCs syndicate experience with Western VCs will increase the likelihood of a successful exit. 3. Data and methodology 3.1 Data and sample This study focuses on how syndicate experience with foreign VC firms influence local VC firms investment selection and performance. We collected venture capital investment and exit data from Asia Venture Capital Journal (AVCJ) database, which has been used by reputable institutions (INSEAD, 2012) and researchers (Brander, Du, and Hellmann, 2010) due to its better coverage for Asian deals as compared to VentureXpert. We constructed our sample as follows. First, we started with all available venture capital investments in the AVCJ database, which includes 11,748 VC investments made by both foreign and local VC firms from 1990 to We included investments that received initial funding between 1996 and 2009 because globalization of venture capital has only gathered pace since the mid-1990s (Iriyama 8

9 et al., 2010; Kenney et al., 2008). We then tracked the outcome of each investment until the end of 2012 and allowed for at least three years to observe an exit for an investment made as late as in 2009 ( Gompers and Lerner, 1998; Hochberg et al., 2007; Nahata, 2008). We only included ventures that received local VC firm funding in all rounds in order to avoid potential foreign VC firm influence on investment performance. Our final sample consists of 3,966 rounds of investments made by 2,094 local VC firms in 3,309 portfolio companies from 1996 to 2009 in the following countries: China, Japan, India, South Korea, Singapore, Hong Kong, Taiwan, Vietnam, Malaysia, Thailand, Indonesia, Philippines and Pakistan. As shown in table 2, the majority of VC investments in Asia during the 13-year period are made in China, India, Japan and South Korea. China has the most active VC market among all developing countries as suggested by its largest number of rounds (927, 23.9%) and number of ventures (822, 24.8%). Japan s VC market remains the most active among developed countries. In terms of total capital received, China accounts for more than 1/3 ($920.3 million, 34.4%) of all capital received in the Asian VC market. And it has the largest investment size with an average of $1.12 million. We further breakdown the VC investments into two different categories based on whether a particular round or venture has foreign syndicate experience (broad or western). As shown in the table, almost 1/3 of all investments in Asia are subject to the potential influence of foreign syndicate experience (either broad or western) brought by local VC firms. Notably, China and India have the highest proportion of investments that are subject to potential foreign syndicate experience influence among developing countries, while Japan and South Korea have the highest proportion among developed countries. [INSERT TABLE 2 HERE] 3.2 Dependent variables The dependent variable in the company level analysis is a dummy that takes the value of one if the investee company ultimately went public or was acquired by the end of In the deal level analysis, the dependent variable is the time to exit computed as the time difference (in years) between initial funding date and exit date. 9

10 In the case of not-yet-exited investments, the time to exit is the time difference between initial funding date and the last day of We do not consider the not-yet-exited investments as unsuccessful but treat them as right-censored. We consider an exit in the form of IPO or M&A as successful since VCs generate returns primarily from successful exits through ether IPO or M&A (NVCA, 2008; Triantis, 2001). This measure of venture capital investment success has been adopted by a large number of recent studies (e.g., Bottazzi et al., 2008; Cumming and Dai, 2010; Dai et al., 2012; Nahata, 2008; Zarutskie, 2010). 3.3 Explanatory variables At least one broad foreign syndicate experience This is a dummy variable that takes the value of one if the VC firm has had at least one syndicate experience with a non-local VC firm in the past five years and zero otherwise. We consider any syndicate experience with a non-local VC firm as a foreign syndicate experience. At least one Western foreign syndicate experience It is measured as a dummy variable that takes the value of one if the VC firm has had at least one syndicate experience with a Western VC firm in the past five years and zero otherwise. We only consider a syndicate experience with Western VC firms as a foreign syndicate experience. Broad foreign syndicate experience It is measured as the amount of broad foreign syndicate experience in the past five years normalized by the total amount of investments the VC has made in the past five years. We consider any syndicate experience with a non-local VC firm as a broad foreign syndicate experience. This measure is time varying and VC specific. It ranges from zero (least experienced) to one (most experienced) Western foreign syndicate experience This is measured as the amount of foreign syndicate experience with Western VC firms in the past five years normalised by the total number of investments the VC made in the past five years. We only consider syndicate experience with a Western VC 10

11 firm as a foreign syndicate experience. It ranges from zero to one and is time varying and VC specific. 3.4 Control variables The selection and performance is likely to be influenced by other factors. We included several variables to control for the VC characteristics, investee companies, and country of the VC firms. Unobservable VC firm-specific characteristics might influence investment performance, which could be captured by VC past success. It is measured as the number of ventures funded by the VC firm that has gone public or been acquired in the past five years. This variable captures VC firm s ability to take its portfolio companies to successful exits. In a recent study on cross-border VC investment by Li et al., (2013), VC past success has been proven to be a significant factor influencing investment performance. VC syndication is important and systematically reduces the level of uncertainty. Megginson and Weiss (1991) and Lerner (1994) find that VC syndication is positively related to the likelihood of IPO exits. Giot and Schwienbacher (2007) find that the larger the size of VC syndication the shorter the time to exit a portfolio company. Similarly, recent studies on the Asian and Chinese market (Dai, et al. 2012; Wang and Wang 2012; Jenner and Suchard, 2013) provide evidence on the VC syndication for VC investment performance. Therefore, we included a dummy variable that takes the value of one if the deal involves more than one VC firm. Furthermore, we also included VC type dummies to indicate difference types of VC firms. We control for venture-related characteristics. We included dummies variables that indicate venture s stage when it received its first round of financing. Specifically, we included early-stage, expansion-stage, and later stage dummies in our analysis. We include these variables because previous studies suggest that early-stage ventures are risky and have a high failure risk (Cochrane, 2005). The level of information asymmetry and uncertainty are high at early stage than the later stage (Dai et al., 2012). In addition, we also account for venture industry-specific fixed effects by including industry dummies in our estimations. 11

12 Time varying variable, related to country of the VC firms are included to capture the effect of selection and investment performance. Countries with higher GDP (Venture nation GDP) might have more entrepreneurial potential. On the other hand, VC firms may rush to such bigger markets and create competition in these countries, which might dampen the probabilities of success. Stock market development measures the level of stock market development in the venture s nation. It is measured as the average of stock market capitalization scaled by GDP. Previous studies in the U.S. such as Black and Gilson (1998) show that a well-developed market is extremely important to the development of VC industry as it provides a viable exit mechanism for both investors and entrepreneurs. The importance of the stock market has also been proved within the context of cross-border investments, which show that the level of stock market development matters for VC investment (Hazarika et al., 2010; Jeng and Wells, 2000). In addition, we included venture-nation dummies and year dummies to control for venture country fixed effects and unobservable temporal effects. 3.2 Estimation models a. Logit model We use logit model to estimate selection and investment performance (company level). Since the dependent variables in both analysis are binary in nature, we apply logit model (Greene, 2002). The basic function of the non-linear model is described as: (1) In table 6 (model 1 and 2), is the probability of having at least one BFSE or WFSE for the ith investment, equals 1 if the VC firm has at least BFSE or WFSE and equals 0 otherwise In table 7, is the estimated probability of a successful exit for the ith investment, equals 1 if the deal is successfully exited by the end of 2012 and equals 0 otherwise.. u is the normal linear regression model, which is: 12

13 (2) Where is the constant and to are coefficients of independent variables to. The log transformation of the logistic model is given by: (3) The parameters are estimated through maximum likelihood method. To test the statistical significance of the predictor variable, Wald test is used. Pseudo is used to measure the goodness fit of the model. Pseudo is similar to in the OLS, the larger the pseudo, the better the goodness of fit. b. Cox proportional hazard model We examine the time-to-exit /exit rate using Cox proportional hazard model. The model is used in our deal level analysis. The dependent variable is the hazard rate, which is the probability of exiting an investment given that the exits have not occurred. The following is the hazard model: (5) where, is the proportional hazard rate, is the baseline hazard rate at time t. j is the index for individual firm, is a vector of independent variables, which includes VC firm-related factors, portfolio company-related factors, and venture nation-related factors. are coefficients to be estimated through maximum likelihood. The Cox model makes no assumption about the distribution of the hazard rate and can take any shape (i.e. could be increasing or decreasing functions). In our analysis of exit performance, we use the computed time to exit as the dependent variables. The survival time in years is either the time between the first investment date the exit date or difference between investment date and 31/03/2012. We do not consider the not-yet-exited deals as unsuccessful but treat them as right-censored. 13

14 4. Results and Analysis We conduct our empirical analysis as follows: we first undertake a general analysis of the companies in which local VC firms invest to examine whether foreign syndicate experience influence local VC firm s investment selection. We then examine the impact of foreign syndicate experience on venture performance. 4.1 Investment selection by local VC firms We analyse the investment decisions of local VC firms at portfolio company level. Here, we examine a cross-sectional data set where the unit-of-analysis is at portfolio company level (i.e. there is one observation for each portfolio company). The univariate analysis is in Table 4. The first column provides analysis of the full sample (i.e. 3,309 portfolio companies). As shown in the table, VC firms are mostly independent VCs and more likely to make sole investment rather than syndicated investments. In terms of portfolio company-related factors, the majority investments are in expansion stage (65.2%) and non-technology industries (50.5%). Among all the investments, 20.4% have been exited by the end of 2012 and it takes around 6 years for a company to go public or been acquired. The next four columns provide a comparison between ventures that have at least one foreign syndicate experience (broad) and ventures with no experience. As can be seen, VC firms with broad foreign syndicate experience are more experienced in taking portfolio companies to successful exits and are more likely to participate in syndicated investments. There is no significant difference between two groups in terms of investment stage but there are differences in industry preferences. Specifically, VC firms with foreign syndicate experience (broad) invest more in Information Technology and Telecommunication industries but less in non-technology industries. The last four columns provide a comparison between ventures that have at least one foreign syndicate experience (western) and ventures that have no such experience. In this case, we only consider a syndicate experience with North American or European VC firms as foreign. Results are similar to the previous analysis in terms of VC firm-related factors and investment industries (i.e. more 14

15 experienced in taking portfolio companies to successful exits and invest more in IT and Telecom industries). VC firms with foreign syndicate experience (western) in this case invest more in early stage and less in expansion stage. This may suggest that local VC firms learn from their previous foreign partners and become more experienced and knowledgeable in advising and nurturing early stage ventures. Finally, in terms of investment performance, broad foreign syndicate experience shows no significant impact on the probability of successful exits but western syndicate experience does. And both broad and western foreign syndicate experience shortens the time to exit. [INSERT TABLE 4 HERE] The company level regressions are in Table 5. The models are logit models that predict investment by VC firms with foreign syndicate experience (broad or western) and tobit models that analyse the proportion of foreign syndicate experience (broad or western). We only include independent variables that would be exogenous to the decision by a local VC firm to invest in the company. The main results are as follows: First, the foreign syndicate experience, either broad or western, shows no significant impact on investment stage preference of local VC firms. Second, there are some strong industry effects. Specifically, VC firms with foreign syndicate experience (broad or western) are more likely to invest in Information Technology and Telecommunication industries. This may suggest that they learn from their previous foreign partners, who are mostly experienced in high-tech industries. In addition, a well-developed stock market provides local VC firms more opportunities to learn from foreign VC firms as suggested the significantly positive relationship between all four dependent variables and stock market development. [INSERT TABLE 5 HERE] 4.2 Investment performance by local VC firms We examine the impact of foreign syndicate experience on investment performance by using two models. We implement Cox hazard model in the deal level analysis and logit model in the company level analysis. The dependent variable in the 15

16 deal level analysis is the time to exit, which is the time difference in years between the date of first round financing and exit date (or 31/12/2012 in the case of not-yet-exited investments). We do not consider the not-yet-exited investments as unsuccessful but as right-censored. In the company level analysis (i.e. one observation per portfolio company), the dependent variable is a dummy variable which takes the value of one if the portfolio company ultimately went public or been acquired by the end of The independent variables are based on the characteristics of the first round financing because VCs make important strategic decisions in the first round and their interactions with investee companies in the initial round lays the foundation for subsequent exchange and is critical for investment s ultimate success (De Clercq et al., 2006; Fitza et al., 2009; Sahlman, 1990). The results are presented in Table 6. Model 1-4 are deal level analysis while Model 5-8 are company level analysis. In model 1, the explanatory variable is a dummy variable which takes the value of one if there is at least one foreign syndicate experience (broad) in a particular deal and zero otherwise. The results of model 1 indicate that the participation of VC firms with broad foreign syndicate experience has no significant impact on the investment performance. In model 2, we use a different proxy for foreign syndicate experience. In this case, we only consider a previous syndicate experience with North American or European VC firms as foreign. As shown in model 2, the coefficient of the explanatory variable is positively significant at 1% level, suggesting that the participation of VC firm with western foreign syndicate experience increases the likelihood of successful exits. In model 3 and model 4, we do not use dummy variables as proxies for syndicate experience but use actual experiences. For example, broad foreign experience is measured as the amount of broad foreign syndicate experience in the past five years normalised by the total amount of investments the VC has made in the past five years. This measure is time varying and specific to VC firms. It ranges from zero (least experienced) to one (most experienced). By using this proxy, we are able to more accurately measure the magnitude and the impact of foreign syndicate experience on investment performance. The results in model 3 and model 4 indicate that broad foreign syndicate experience has no significant impact on investment performance but western foreign syndicate experience is positively significantly related to the likelihood of successful exits. 16

17 Model 5-8 are company level analysis. The explanatory variables are the same as used in model 1-4. As can be seen from the results, only western foreign syndicate experience is positively significantly related to the likelihood of successful exits but other explanatory variables do not have significant impact on investment s ultimate success. This suggests that local VC firms only benefit from their previous collaboration experience with western VC firms. In other words, the partnership with western VC firms helped professionalize the local VC firms. Local VC firms learnt from western VC firms and acquired better knowledge on advising and nurturing entrepreneurs and taking those firms to successful exits. Several empirical findings about the control variables in Table 6 are noteworthy. Both VC firm s past success and VC syndicate help improve the success rate of the investment project. At deal level analysis, we find that Corporate VC firms perform better than Independent VC firms but company level analysis does not show such results. As expected, early and expansion stage investments are less likely to be exited compared to later stage investments. High-tech industries such Computer Related and Information Technology investments in general are less likely to be exited than non-technology investments. We also find that total amount invested is positively related to investment performance at company level analysis but not at deal level analysis. In terms of VC country-related factors, we find that GDP is negatively related to investment success suggesting that VC firms rush to such bigger markets and increase competition in these countries, which dampen the probabilities of success. In addition, we find that good stock market development in the country of a portfolio company improve the likelihood of success exits, which suggests that a viable exit mechanism is very important to both investors and entrepreneurs. Overall, we find that broad foreign syndicate experience of local VC firms are not directly related to investment success but syndicate experience with western VC firms improve the likelihood of successful exits. This suggests that local VC firms learned from their previous collaboration experience with western VC firms and acquired better experience on advising and nurturing entrepreneurs and taking these portfolios to successful exits. [INSERT TABLE 6 HERE] 17

18 4.2 Robustness check In this section, we check the robustness of our results. Since we observe the exit success of a local VC investment conditional on deal that have only local VC firms, we use a Heckman selection model to ameliorate endogeneity concerns. In the first stage, we use probit model in the selection equation to estimate the likelihood of an investment having only local investors. The control group is also obtained from AVCJ database, which contains 3,950 investments made by only foreign VC firms during In the selection model, we also include control variables used in the main regression and control for industry fixed effects, country fixed effects, and year fixed effects. The results are similar to our previous findings suggesting that our results are robust after controlling for potential selection bias. 6. CONCLUSION There has been a large number of studies on cross-border VC activities in recent years (Dai et al., 2012, Jenner and Suchard, 2013, Li et al., 2014). There is a consensus that forming partnership or syndicate with local VCs is an effective way of alleviate information asymmetry caused by cultural and geographical distance. However these studies are mainly from the perspective of foreign investors. Our study extents their contribution by examining the syndicates influence on local VC firms. Specifically, we examine whether the syndicate experience with foreign VCs influence local VCs following-up investment behaviour in terms of selection and performance. We first examine whether syndicate experience with foreign VC firms influence local VC firms investment behaviour in terms of selection. We use two different proxies for syndicate experience: syndicate experience with non-local VCs including VC firms in other Asian countries. Our results indicate that both types of syndicate experience have significant impact on local VCs selection. Specifically, local VCs are more likely to invest in high-tech industries such as information technology and telecommunications after their syndicates. We then examine whether the syndicate experience s impact on investment performance in terms of exit success. The results 18

19 show that only syndicate experience with western VC firms improve local VCs investment performance. This suggests that local VCs learn from their foreign partners during the syndicate. But the knowledge acquired from western VC firms is more abundant and effective. This study provides several practical implications for market practitioners. For instance, local VC firms who have no syndicate experience with foreign VCs should consider such collaboration as a platform for organisational learning, where they could acquire knowledge and skills to improve investment performance. Local VCs with the intention to syndicate may prefer VC firms from North America or Europe over VCs from other Asian countries because learning capacity in such syndicates is higher. 19

20 REFERENCE Aizenman, J., Kendal, J. (2008), The internationalization of VC and private equity, working paper. UCSC and the NBER; The World Bank Black, B., Gilson, R., Venture capital and the structure of capital markets: banks versus stock markets. J. Financ. Econ. 47 (3), Bottazzi, L., Da Rin, M., Hellmann, T., 2008.Who are the active investors? Evidence from venture capital. J. Financ. Econ. 89 (3), Brander, J., Du, Q., and Hellmann, T. (2010), The effects of government-sponsored VC: international evidence, Working Paper. Bruton, G. D., Fried, V., and Manigart, S. Institutional influences on the worldwide expansion of VC. Entrepreneurship Theory and Practice, 2005.,29(6): Buckley, P. J. and Casson M. (1988). A theory of cooperation in international business. In Contractor, F. and Lorange, P. (Eds), Cooperative Strategies in International Business. Lexington, MA: Lexington Books, pp Cochrane, J.H., The risk and return of venture capital. J. Financ. Econ. 75, Cumming D. and Johan, S. (2008) Pre-planned exit strategies in VC. European Economic Review. 52: Cumming, D. Fleming, G. and Schwienbacher, A. (2006). Legality and VC exits. Journal of Corporate Finance. 12: Cumming, D. J. and MacIntosh, J.G.(2003). VC exits in Canada and the United States. University of Toronto Law Journal, 53: Cumming, D., Dai, N., Local bias in VC investments. Journal of Empirical Finance 17, Dai, N., Jo, H., and Kassicieh, S. (2012), Cross border VC investments in Asia: selection and exit performance, Journal of Business Venturing, page 19 De Clercq, D., Fried, V.H., Lehtonen, O., Sapienza, H.J., 2006.An entrepreneur's guide to the venture capital galaxy. Acad. Manag. Perspect. 20 (3), Deloitte, Global Trends in VC Survey. Deloitte, Global Trends in VC Survey. Dodgson M. (1993). Learning, trust, and technological collaboration. Human Relations, 46, Elisabete G., Cesaltina P. and Mohamed, A. (2008). The Exit Decision in the European VC Market. Working Paper Ferrary M. Syndication of VC investment: The art of resource pooling. Entrepreneurship Theory and Practice, 2010, 34(5): Fitza, M., Matusik, S.F., Mosakowski, E., 2009.Do VCs matter? The importance of owners on performance variance in start-up firms. Strateg. Manag. J. 30 (4), Giot, P., and Schwienbacher, A. (2007), IPOs, Trade Sales and Liquiditions: Modelling VC exits using survival analysis, Journal of Banking and Finance, 31(3), Gompers, P., Lerner, J., 1998.The determinants of corporate venture capital successes: 20

21 organizational structure, incentives, and complementarities. Working paper NBER. Grant R. M. (1996). Prospering in dynamically-competitive environments: organizational capability as knowledge integration. Organization Science, 7, Hamel G. (1991). Competition for competence and inter-partner learning within international strategic alliances.strategic Management Journal, 12, Harrigan K. R. (1986). Managing for JV Success. Lexington, MA: Lexington Books. Hazarika, R., Nahata, R., Tandon, K., Success in Global VC Investing: Do Institutional and Cultural Differences Matter? Working Paper. Baruch College, City University of New York. Hennart J. F. (1988). A transactions costs theory of equity JVs. Strategic Management Journal, 9, Hochberg, Y.V., Ljungqvist, A., Lu, Y., 2007.Whom you know matters: venture capital networks and investment performance. J. Financ. 62 (1), Hopp, C and Rieder F., What drives VC syndication? Applied Economics, 2010, In press. Inkpen A. C. and Crossan M. M. (1995). Believing is seeing: joint ventures and organization learning. Journal of Management Studies, 32, Inkpen, A. (1998) Learning, knowledge acquisition, and strategic alliances. European Management Journal 16 (2), Iriyama, A., Li, Y., Madhavan, R., Spiky globalization of venture capital investments: the influence of prior human networks. Strateg. Entrep. J. 4 (2), Isaksson, A. (2007). Exit strategy and the intensity of exit-directed activities among VC-backed entrepreneurs in Sweden. Phd dissertation, Umeå School of Business Jeng, L., Wells, P., The determinants of VC funding: evidence across countries. Journal of Corporate Finance 6 (3), Jenner, M., H., Suchard, J., A. (2013), Foreign VC and venture success: Evidence from China, Journal of Corporate Finance, 21, Kenney, M., Haemmig, M., Goe, W.R., 2008.The globalization of the venture capital industry. In: Macher, J.T., Mowery, D.C. (Eds.), Innovation in Global Industries: U.S. Firms Competing in a New World. National Academies Press, Washington, D.C., pp Khanna T., Gulati R. and Nohria N. (1998). The dynamics of learning alliances: competition, cooperation, and relative scope.strategic Management Journal, 19, Kogut B. (1998). Joint ventures: theoretical and empirical perspectives. Strategic Management Journal, 9, Kogut B. and Zander U. (1992). Knowledge of the firm, combinative capabilities, and the replication of technology. Organization Science, 3, Lane P. J. and Lubatkin M. (1998). Relative absorptive capacity and interorganizational learning. Strategic Management Journal, 19, Lerner J. The syndication of VC investments. Financial Management, 1994, vol. 23, 3:

22 Lerner, J., (1994).VC exists and the decision to go public. Journal of Financial economics 35, Li, Y., et al., National distances, international experience, and venture capital investment performance J. Bus. Venturing (2013) Lockett, A., Wright, M., (2001), The syndication of VC investments, Omega 29, Mäkelä, M. M. and M. Maula, (2008) Attracting cross-border VC: the role of a local investor. Entrepreneurship and Regional Development, 20(3): Mäkelä, M. M. and M. Maula. (2006) Interorganizational commitment in syndicated cross-border VC investments. Entrepreneurship Theory and Practice, 30(2): Megginson,W.,Weiss,K.,(1991).VC certification in initial public offerings. Journal of Finance 46, Meuleman M. Wright M. Cross-border private equity syndication: Institutional context and learning Journal of Business Venturing, 2010, in press. Mowery D. C., Oxley J. E. and Silverman B. S. (1996). Strategic alliances and interfirm knowledge transfer. Strategic Management Journal, 17, Special Issue, Winter, Nahata, R., 2008.Venture capital reputation and investment performance. J. Financ. Econ. 90 (2), NVCA, The venture capital industry an overview. (Visited Jan 15, 2014). Powell W. W. (1987). Hybrid organizational forms. California Management Review, 30, 1, Powell W.W.,Koput K. W. and Smith-Doerr, L. (1996). Interorganizational collaboration and the locus of innovation: networks of learning in biotechnology. Administrative Science Quarterly, 41, Sahlman, W.A., 1990.The structure and governance of venture-capital organizations. J. Financ. Econ. 27 (2), Schertler A. and Tykvova T. VC and Internationalization, International Business Review, 2010, in press. Simonin B. L. (1997). The importance of collaborative know-how: an empirical test of the learning organization. Academy of Management Journal, 40, strategic alliances. Strategic Management Journal, 20, Triantis, G.G., 2001.Financial contract design in the world of venture capital. Univ. Chic. Law Rev. 68 (1), Wang, L. and Wang, S. (2012), Cross-border VC performance: Evidence from China, Pan-Basin Finance Journal, 19:71-79 Wright, M. and Lockett, A. (2003), The structure and management of alliances@ syndication in the venture capital industry, Journal of Management Studies 40:8 Zarutskie, R., 2010.The role of top management team human capital in venture capital markets: evidence from first-time funds. J. Bus. Ventur. 25 (1),

23 Table 1: Definition of variables Variable name Broad foreign syndicate experience Broad foreign syndicate experience (I) Western foreign syndicate experience Western foreign syndicate experience (I) VC firm-related factors VC past success VC syndication Independent VC indicator Corporate VC indicator Venture-related factors Definition of variable The amount of broad foreign syndicate experience in the past five years normalized by the total amount of investments the VC has made in the past five years. A dummy variable which take the value of 1 if the deal involves VC firms that have at least one broad foreign syndicate experience in the past five years, and 0 otherwise The amount of western foreign syndicate experience in the past five years normalized by the total amount of investments the VC has made in the past five years. A dummy variable which take the value of 1 if the deal involves VC firms that have at least one western foreign syndicate experience in the past five years, and 0 otherwise The number of ventures funded by the VC firm that has gone public or been acquired in the past five years in a particular round (or in the first round in company level analysis) A dummy variable which takes the value of 1 if the deal involves more than one VC firm, and 0 otherwise A dummy variable which takes the value of 1 if the VC firm is an independent VC firm, and 0 otherwise A dummy variable which takes the value of 1 if the VC firm is an independent VC firm, and 0 otherwise 23

24 Early-stage ventures Expansion stage ventures Computer related Information Technology Medical/Health care Telecommunications Total amount invested Venture nation-related factors Venture nation GDP Stock market development A dummy variable which takes the value of 1 if the venture is in early-stage when it receive its initial founding, and 0 otherwise. A dummy variable which takes the value of 1 if the venture is in expansion-stage when it receive its initial founding, and 0 otherwise. A dummy variable which takes the value of 1 if the venture is in computer related industry A dummy variable which takes the value of 1 if the venture is in information technology industry A dummy variable which takes the value of 1 if the venture is in medical/health care industry A dummy variable which takes the value of 1 if the venture is in telecommunication industry In deal level analysis, total amount invested is the amount invested by all VC firms in a particular round. In company level analysis, it is the total amount invested in all rounds by all VC firms. The GDP per capita measured in USD in the venture's nation in a particular year, obtained from The World Bank. It is measured as the total market capitalisation of the venture nation's stock market in a particular year over the GDP. 24

25 Table 2: Local venture capital investments in Asia, China (PRC) Japan India South Korea Singapore Hong Kong Taiwan Other Total N % N % N % N % N % N % N % N % N % Number of rounds % % % % % % % % % Broad foreign syndicate exerience (I) % % % % % % % % % Western foreign syndicate exerience (I) % % % % % 5 7.9% % % % Number of ventures % % % % % % % % % Broad foreign syndicate exerience (I) % % % % % % % % % Western foreign syndicate exerience (I) % % % % % 5 9.3% % % % Total capital ($M) % % % % % % % % % Mean round size ($M)