The influence of syndication characteristics on investment performance of corporate venture capital

(1)

Master thesis

The Influence of Syndication

Characteristics on Investment Performance

of Corporate Venture Capital

Student: Geert Beks

Student number: 11096608 Date submitted: 27-01-2017

MSc. in Business Administration – International Management Track Institution: University of Amsterdam

(2)

STATEMENT OF ORIGINALITY

This document is written by Geert Beks who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

ABSTRACT

Corporate venture capital is an increasingly popular vehicle in which the focus lies more on learning from the portfolio company than on direct financial gain, which characterizes independent venture capital. The extent to which these investments are successful have before been shown to be influenced by, among other factors, the quality of the syndicate in terms of, firm-level factors such as industry type, experience and diversity. 7900 private equity transactions, gathered from the ThomsonOne database, regarding the success of syndication are evaluated, whereby success is measured in terms of exit mode of the portfolio company and analyzed using binary logistic regression. This paper is an addition to the existing literature in that it explores the influence of the strength intellectual property regimes across three dimensions. It is shown that as the strength of the intellectual property regime increases, the positive effects of (1) the amount of partners within a syndicate, (2) the amount of investor types, and (3) the total experience present within the syndicate on the outcome of corporate venture capital investments get stronger. These findings have the implication for academia that some determinants of corporate venture capital performance may not have the expected effect in isolation, but do so when considered in conjunction with the strength of intellectual property regimes. This in turn, has the managerial implication that corporate venture capital firms should consider more specific portfolio company characteristics in combination with more general national characteristics such as intellectual property regime when evaluating investment opportunities.

___________________________________________________________________________ Keywords: corporate venture capital, performance, syndication, intellectual property regime, experience, diversity, binary logistic regression

(5)

1. INTRODUCTION

Corporate venture capital (CVC) is an important topic in international business as there have been several uprisings of CVC in the past, leading to shifts in allocation of investment capital (Basu _{et al.}

, 2016). CVCs invest in portfolio companies with goals that are

not merely financial, but which are also strategic which consequently has implications for the CVC’s own innovations. Any success in such activities could be influenced by different determinants, such as with whom the CVC works together in the investment, that is, with whom the CVC syndicates, that is. Moreover, the strength of protection of intellectual property (IP) in terms of how much firms benefit from their own innovations, in different countries also plays a significant role as it influences the extent to which gains from investments, in the form of patents, are protected (Dushnitsky and Shaver, 2009).

1.1 Three uprisings

Over the years, there has been a varying amount of interest in investment through CVC, due to geopolitical, financial, and other reasons. These reasons have in the past influenced the extent to which the CVC investor stands to gain from investing in a portfolio company. The first uprising can be identified starting in the 1960s, which was enabled by an excess in cash flow among firms, and the success of portfolio companies (Fast, 1978; Gompers and Lerner, 1998). This led to increased room for expansion on the one side, and increased attractiveness of possible targets on the other. The main reason that ended this uprising was the 1973 oil crisis, leading to a decrease in excess cash flow of potential investors (Dushnitsky and Lenox, 2006). Moreover, increasingly, there were frictions between portfolio companies and parent companies, which decreased the performance of CVC at this time (Dushnitsky and Lenox, 2006).

The second uprising was mainly due to legislative changes and growth opportunities. These opportunities led firms from the chemical and technological industries to engage in CVC again, which have historically been two industries that have attracted relatively high amounts of CVC investment. However, this uprising came to an end during the 1987 market crash (Dushnitsky and Lenox, 2006).

The third wave of CVC investments took place during the late 1990s, at a time when the dot-com bubble was inflating. It was during this uprising that the relative importance of

(6)

CVC investments as compared to other forms of venture capital increased notably. Whereas in 1996 CVC investments amounted to 0,5 billion dollars, this figure grew to 15 billion dollars over the course of five years (Dushnitsky and Lenox, 2006). And although this uprising was ended by a crisis as well, the level of CVC investments remained well above historical levels (Chesbrough, 2002).

CVC’s current uprising, combined with the increased need for firms to continually improve their innovation strategies makes CVC an essential topic to be studied further, both in the interest of academia, as well as that of practitioners, because there remain further potential synergies to be gained as established firms work together with small firms . The main reason for this is that large firms may have the sufficient capital backing potential investments, but they may lack finesse and agility in terms of innovation (Dushnitsky and Lenox, 2006).

Despite the dominance of CVC in the US, the center of gravity of CVC is currently shifting away from the US toward emerging markets (Jeng, L. A., Wells, P. C., 2000), where industries such as IT and services are attractive for corporate investors. This is primarily facilitated by increased ease of doing business abroad in such industries through internet. There is a lack of research, however, on what factors drive this expansion, what type of companies are likely to engage in such international expansions through CVC, and what difficulties such firms may face.

Firms face complications due to differences in IP regimes in the different countries in which they operate, as investment through CVC may influence the extent to which the firm is able to derive value from IP (Dushnitsky and Shaver, 2009). Moreover, sharing experience with other CVC-investors, or instead, feeling a competitive need to protect any potential gains from an investment may also influence the outcomes of the CVC investment. This is why this paper revolves around the question of the strength of IP regimes impact the influence of syndicate partner characteristics on CVC performance.

A main focus of research on CVC is on the biopharmaceutical industry, as well as the technology industries, which are widely recognized to make extensive use of advantages of external development through CVC (Dushnitsky and Lenox, 2006). Whereas numerous factors influencing propensity to engage in CVC have been researched, such as industry competition, technological development, and IP regimes, more recently, subjects such as clean energy have come to the attention of research (Basu _{et al.}

(7)

1.2 Definition and position in research

Corporate venture capital is “the equity investment by an established corporation in entrepreneurial ventures” (Basu _{et al.}

, 2016), and these investments are made to further

nurture these entrepreneurial ventures, in order to gain from their innovativeness (Dushnitsky, 2013). A CVC unit separate from the parent company often is responsible for the investment, allowing for a large degree of freedom of decision-making regarding investments (Dushnitsky and Shapira, 2010).

Concerning the relative positioning with respect to the current literature, CVC investments in starting firms have been approached from different angles, such as psychology, economics and several different fields (Ireland and Webb, 2007). Compared to the conventional view of venture capital, CVC is distinct regarding the strategic interests that the investing firm has (Chesbrough, 2000). That is, the CVC is looking for a certain amount of fit with the portfolio company in order to achieve results that are not necessarily pecuniary, but that involve goals such as increasing the smoothness of knowledge flows within the firm and increasing innovativeness (Dushnitsky and Lenox, 2006).

Following this line of reasoning, it is specifically this fit between the CVC and portfolio company that leads to a potential gain from complementarities in assets between the two parties (Ivanov and Xie, 2010). This is markedly different from the aims of individual corporate ventures, of which the main objectives are financial and which have a looser relationship with portfolio companies and therefore require a lower level of fit between the investing party and the portfolio company.

1.3 Description of the sector and historical data of CVC

Examples of firms that engage in CVC are Intel, Cisco, Microsoft, Comdisco, and Dell, which are IT-based companies that have made investments over the course of the 1980s and 1990s. Of these firms, Intel is the largest CVC investor with 1,486 million US dollars invested in CVC. Intel illustrates the dominant position that the US has occupied within the field of CVC, in the sense that the lion's share of Intel’s investments have been in US-based firms. However, this focus is starting to shift toward more non-US-based firms (Tyler, 2011). Important non-IT-based CVC investors include Johnson & Johnson and Ford Motor, a pharmaceutical company and a car company respectively, with total CVC investments

(8)

amounting to 196 million US dollars and 146 million US dollars (Dushnitsky and Lenox, 2006).

Although CVC investors from the chemical industry were active during the 1980s, CVC activity from other industries exceeded the amount of investment in chemical industries during the 1990s, most likely due to the increasing importance of the internet at the time and its investment opportunities (Dushnitsky and Lenox, 2006). An important attracting factor for CVC investments is the technological opportunities offered in an industry (as illustrated by the amount of patents awarded) and the strength of IP protection, as this will ensure gains from the cooperation with the portfolio company. The presence of complementary capabilities, as well as the investor’s absorptive capacity matter as well, as a higher absorptive capacity, ceteris paribus, allows the investor to learn more from its portfolio companies (Dushnitsky and Lenox, 2006).

After the three peaks in CVC activity in the 1960s, 1970s, and 1990s discussed previously, we are currently experiencing another peek in CVC activity, which is mainly powered by the increasing need for large companies to engage in constant innovation (Basu _et

al.

, 2016). Such innovations increase economic growth and offers opportunities for economic

rent for firms that engage in these innovations, at the expense of lagging incumbents, which is the process of creative destruction as discussed by Schumpeter (1942). Moreover, CVC activity is currently characterized by the need for innovation to which CVCs respond by syndicating, investing in the same portfolio company with other investors that is, in order to either assess attractiveness, reduce risk, or establish scale (Chesbrough, 2000).

Literature currently has widely addressed the relationships between syndication and CVC performance, as well as how such performance is influenced by IP regimes. However, where literature is lacking is coverage of how these issues may be interlinked, which is what this paper addresses. It does so by reviewing the literature on syndication, CVC, IP regimes and by statistically analyzing past performance of CVCs using the ThomsonOne database corporate venture transactions, as well as other sources for the evaluation of the strength of IP regimes.

The remainder of this paper is organized as follows. First, the literature is reviewed, considering mostly papers that were produced in the last two decades, in order to give an up-to-date representation of the current thinking on CVC and its determinants. Next, the gap in the literature is discussed, leading to the research question, which is followed by a

(9)

conceptual framework, depicting hypotheses derived from the literature. This is followed by a discussion of the research design, along with the relevant variables and sample, which is discussed and from which several conclusions are drawn for scholars, managers and for policy makers.

2. LITERATURE REVIEW

In this section, the existing literature related to CVC is discussed. For example, the different forms that venture capital can take are discussed and the purpose of these forms of venture is contrasted with each other. The position of the portfolio company is discussed, as well as the position of the investing firm within a group of investors, a syndicate. Moreover, the relation of CVCs with regard to IP regimes is discussed in the light of attractiveness of cooperation with portfolio companies.

The main level of analysis is the portfolio company, and central theme is how the relation of CVCs with other investors influences performance. By taking this perspective the emphasis shifts from a broader view of an entire industry towards a more focused view, eventually illustrating how the the link between syndication and investment performance is influenced by IP regimes. This review will lead to a discussion of where the current literature is lacking, which is where the research gap is identified.

2.1 Corporate venture capital, independent venture capital and government

venture capital

Several kinds of venture capital investors can be identified, other than CVC. Examples include independent venture capital and government venture capital. Whereas in the case of CVC the investments have mainly strategic objectives, in the case of independent venture capital, the objective is solely financial returns (Dushnitsky and Lenox, 2006).

Furthermore, government venture capital investors have in the past often subsidized high-tech ventures in countries such as Israel, Singapore, and Taiwan (Lerner, 1999). These investments can be motivated by the fact that returns from investments in R&D may be larger in terms of social benefit, than for private gain for private investors, or the fact that the

(10)

involvement of governments in such investments can eliminate information asymmetries that otherwise might have obstructed such investments (Lerner, 1999).

CVC is different from independent venture capital in the extent of knowledge and support that the corporate investor transfers to the portfolio company. Whereas other forms of alliances imply mutual dependence, in the case of CVC, there is mainly a unidirectional flow of capital, to the portfolio company that is (Dushnitsky and Shapira, 2010). In this sense, CVC is complicated by the need to manage both the relationship with the portfolio company, in order to reap the benefits from cooperation, as well as the relationship with the parent firm, in order to maintain the financing power to continue investing in these portfolio companies (Dushnitsky and Lenox, 2006).

Even though the popularity of CVC investments has been increasing over the last decades, its impact has historically been limited relative to the impact of independent venture capital (Chesbrough, 2000). Still, performance can ben increased by designing CVC investments to take advantage of the agility of independent venture capital, while also maximizing strategic benefits from complementarities between the portfolio company and the CVC firm (Chesbrough, 2000).

2.2 Motivations for CVC and measurement issues

Firms are increasingly seeing the need for innovation outside the boundaries of the firm itself, which is indicated by increased spending in small firms for innovation; a manifestation of CVC, which has risen during the period of 1891 to 2009 (National Science Foundation, 2012). This can be described as a shift from a focus on financial gains, toward investments with strategic goals (Dushnitsky _{et al.}

, 2009). However, a main issue in the

evaluation of effectiveness of CVC is that often financial measures are used to evaluate performance, because of difficulties in measuring the effectiveness of CVC in different ways. That is, gains from CVC are mainly tacit, with little direct financial benefit to be traced back to the CVC investment, but gains which are important nonetheless, as they may boost long-term performance.

The increased popularity of CVC is illustrated by the increased lifetime of portfolio companies, which indicates a greater involvement of corporate investors, having an increasingly more important stake in its investments (Dushnitsky, 2012). Corporate investors

(11)

are motivated to engage in CVC by several factors such as obtaining financial returns, complementing R&D efforts, gaining access to emerging markets, and gaining access to disruptive technology and business models (Bielesch _{et al.}

, 2012).

What is more, by engaging in CVC, firms are able to stay up-to-date with regard to specific market trends, when it comes to promising innovations (Ivanov and Xie, 2010), or in other cases, decreasing production costs of the firm, depending on whether the value CVC investments is indeed larger than ‘regular’ IVC investments (Dushnitsky and Lenox, 2006). Whether this is the case depends on for example the extent to which CVCs can leverage their industry-specific knowledge to aid in the portfolio company’s development, and the extent to which the existence of a cooperation between CVC and portfolio company provides the portfolio company with additional credibility, which could lead to higher valuations of potential future buyers of the portfolio company (Ivanov and Xie, 2010).

In a similar vein, Ernst _{et al. (2005) ascribe the attractiveness of CVC to the fact that} it allows the investor to monitor relevant market developments, that it gives access to experts in certain fields of technology, that it encourages the investor’s internal entrepreneurial spirit and improves the R&D efficiency. These last few aspects can be ascribed to the the investment in portfolio companies bringing in some sort of competition into the investor’s organizational structure.

However, apart from such potential advantages which CVCs might enjoy, there are also disadvantages that could potentially outweigh them. For example, as opposed to managers of IVCs, managers of CVCs often lack sufficient incentives to put in their best efforts to optimally develop portfolio companies, which on its turn leads to a higher turnover of qualified employees of CVCs, relative to those of IVCs (Gompers and Lerner, 2000).

An explanation for the lack of sufficient incentives can be found in the absence of an easily observable relationship between CVC investments and financial performance, as well as the fact that such CVC investments benefit multiple aspects of the firm, once again making it difficult to trace back the source of the benefits. Moreover, efforts building up to certain success may be accredited to one manager, even though the preparation of such a success may have been done by a previous manager (Rind, 1981).

Moreover, any gains from CVC may be rather limited due to unforeseen incompatibility between the investor and the portfolio company. Gains from these investments are most likely when the activities of both parties are somewhat related, but if the

(12)

extent of relatedness is overestimated, gains will be limited (Ernst _{et al.}

, 2005). Besides, the

parent firm of the CVC may not be fully prepared to absorb the knowledge gained from the cooperation with portfolio companies, which is a prerequisite for a successful cooperation (Ernst _{et al.}

, 2005).

Another potential disadvantage of CVCs relative to IVCs, which is brought forward by Ivanov and Xie (2010), is that due to the additional engagement of CVCs with portfolio companies, conflicts of interest may arise. CVCs learn increasingly more about the portfolio companies’ strategies, which could prevent the investor from engaging in potentially rewarding cooperations. This argument may, however, be lacking to the extent that IVCs have even less information available of the portfolio companies and might therefore engage in suboptimal cooperations, while foregoing profitable ones. One could therefore argue that more information should lead to better decision-making, and therefore to more profitable outcomes, at least for the investor. For the portfolio company, however, it is indeed true that development may be constrained due to the CVC investor’s own strategy.

Moreover, corporate investors look for complementarities between their own R&D and the capabilities of the portfolio company. Such complementarities are needed to leverage the benefits gained from CVC investments throughout the investing firm, instead of merely taking over certain activities of portfolio companies (Chesbrough, 2000). On the other hand, portfolio companies are motivated by increased access to capital, but more importantly they are motivated by the access to knowledge resources, reputation, and network advantages (Basu _{et al.}

, 2016). Such resources can be used by the portfolio company to better position

itself with regard to the competition. In the electronics industry, this is often achieved through customer and supplier relations. In the biopharmaceutical industry it is most apparent in marketing and product development and research agreements (Ivanov and Xie, 2010).

Such advantages may, however, be limited as CVCs have less autonomy over their investments compared to IVCs (Ivanov and Xie, 2010). What is more, IVCs often have substantial industry-specific knowledge which allows these IVCs to provide assistance to portfolio companies which is on par with CVCs when it comes to usefulness of these resources (Ivanov and Xie, 2010).

Firms face a choice between investment through CVC or acquisitions, which has been shown to be affected by the extent of uncertainty (Tong and Li, 2011). In this sense, CVC can be seen as real options, as the investor’s stake in a portfolio company can, but does not have

(13)

to result in a complete acquisition of the portfolio company, if the portfolio company develops well. This can be described as a positive relationship between uncertainty and level of CVC investment. The strength of this relationship is increased by investment irreversibility, whereas it is decreased by the level of growth opportunities (Tong and Li, 2011). Another way in which CVC investments can be seen as a real option, is that as technological and market uncertainty decreases, the CVC and portfolio company can engage in a strategic alliance, in which further investments are made to ensure the transfer of technological knowledge, which is linked to the subject of syndication as different investors may engage in the investment. Such formation of a strategic alliance is, however, not required and thus the investment is a real option instead of a direct investment with all its ties.

Despite increasing popularity of CVC, not all large firms engage in it. Dushnitsky and Lenox (2005) describes examples of underlying reasons, such as the strength of the technological base, the importance of complementary distribution capability, the corporate investor’s cash flow, and absorptive capacity, the extent of industry IP protection. Consequently, if industry IP protection influences the popularity of CVC investments, this is a factor that firms should take into account when investing in portfolio companies, in conjunction with other more firm-specific factors.

2.3 Bargaining power of the portfolio company and CVC investor

A main concern for portfolio companies is that they have few means to show their attractiveness to potential corporate investors, as their main value lies in the knowledge they possess, while putting this knowledge on display deteriorates their bargaining position (Maula, M. V., Autio, E., Murray, G. C., 2009). This issue can be addressed in several ways. Firstly, the portfolio company can try to limit the corporate investor’s ownership, which results in more control for the portfolio company, allowing it to maintain secrecy over IP. Secondly, the portfolio company can increase its bargaining power by timing the launch of inventions, as portfolio companies have the flexibility to do so (Maula et al., 2009), while these are tools that are not available to the corporate investors. This ability to time the launch of inventions at the time that suits the portfolio company is powerful, because the portfolio company can pressure the corporate investor by delaying the launch, but it can also simply

(14)

launch the invention when the expected return is highest for the portfolio company itself, without having to take into account other parties.

Besides these mechanisms, portfolio companies also have social defense mechanisms to their availability, as described by Hallen (2014), who shows that CVC investors can help mitigate the vulnerability of their resources. Although these tools do help the portfolio company in maintaining its bargaining power, the application of these tools does reduce the efficiency of the overall cooperation, as being defensive during the cooperation prevents the parties from achieving their full potential due to frictions (Maula et al., 2009). The impact of such defense mechanisms does merit further research, with regard to the long-term impacts of such decreased efficiency (Basu _{et al.}

, 2016).

Portfolio companies are often in need of additional means to engage in the commercialization of their products, or to pursue further development, which requires CVC investments (Dushnitsky and Lenox, 2006). However, such ties may limit the portfolio companies to gather complementary support from the open market, which results in a trade-off between support and flexibility. Park (2012) finds that in such a situation, CVC is beneficial for the portfolio company in case it offers specialized complementary assets, or if the portfolio company operates in an uncertain environment.

2.4 The effects of CVC on performance

There have been several studies researching the effects of CVC on the portfolio company’s performance, which indicate that portfolio companies that have access to a corporate investor’s knowledge have a faster growth rate and have a higher probability of exiting via an initial price offering (IPO), which is preferable to an exit via acquisition (Weber, B., Weber, C., 2007; Gompers and Lerner, 1998). This success does however come at a premium, as only the highest performing investors have sufficient cash flow to engage in CVC, often barely being able to break-even (Allen, S. A., Hevert, K. T. (2007).

CVCs can have substantial effects on the innovativeness of firms, which can complement other innovation activities within these firms (Chesbrough and Tucci, 2004). That is to say, as absorptive capacity of CVCs increases, its capacity to learn from innovative initiatives undertaken by portfolio companies increases, even though one should bear in mind that the actual absorptive capacity can vary from one domain of knowledge to the next

(15)

(Cohen and Levinthal, 1990). Such cooperation can be beneficial to both parties (Chesbrough, 2002), as this cooperation allows the firm to gain insights from a broader field of practitioners, resulting in both parties being more prepared to deal with unforeseen circumstances (Chesbrough, 2012).

CVCs are often run by experienced managers who know the market, who see opportunities, and who therefore possess capabilities that are specifically applicable to the firm, which has as a positive consequence that the CVC is able to respond to opportunities that have a specifically high potential for the CVC. The specificity in the managers’ capabilities makes them valuable. However, since CVC managers are often paid through equity shares in the portfolio company, which often results in lower payments than other forms of payment, a shift from CVC to independent venture capital attractive for these managers, resulting in a lack of qualified managers available to the corporate investor (Chesbrough, 2000).

Another challenge for CVC investors is that if they syndicate, they must be prudent in picking the right partners, as their goals are mainly strategic, so cooperating with direct competitors could increase the risk of losing any of the obtained gains to said competitors.

When engaging in CVC, owners of firms may aim for short-term financial gains. This may conflict with the long-term vision of other leading people within an organization, who aim for strategic benefits through cooperating with portfolio companies. Eventually, all investments are required to be feasible financially, but if these managers are not given the freedom to pursue these objectives, there may be only limited strategic gains to be made from CVC, due to an overemphasis on short-term financial gains, imposed by owners (Ernst_{et al.}

,

2005), resulting in the CVC not realizing its full potential.

2.5 Syndication

Corporate investors often work together in syndicates to find potential portfolio companies for CVC investments, as this could lead to risk reduction, although this finding is subject to different situational factors, such as experience of local partners or international syndicate partners (Dai, N., Nahata, R., 2015). Investors that have experience syndicate primarily with other investors with a similar amount of experience, as this allows the investor to gain insight into the attractiveness of the investment (Lerner, 1994). Potential syndicate

(16)

partners with different amounts of experience, on the other hand, may not be as likely to cooperate as firms with more experience, ceteris paribus, may have less incentive to engage with firms with less experience as they have less to gain from the firm with less experience..

The benefits of syndication are described by amongst others Markku (2013), noting that timely attention to technological discontinuities is achieved by forming heterophilous ties through CVC, that is, by co-investing with other VCs, instead of by forming homophilous ties by forming alliances with industry peers. This line of reasoning resonates in the current paper, considering the impact that diversity within a group of investors has on CVC performance.

Markku also points out that decisions regarding syndication in the early stages of the investment can have long-term consequences. That is, portfolio companies that are being invested in by international investors and domestic investors with strong international syndication networks increases the probability of an international exit. Moreover, from the perspective of the CVC, getting involved in an international syndicate is more challenging than merely staying domestic. However, it does provide the investor with the capabilities allowing for successful international exits, as relations of the investor in terms of international syndication partners decrease negative effects of geographic distances (Markku, 2013).

Moreover, investors can be motivated to engage in syndication as they are able to obtain a certain degree of critical size by cooperating with firms in a syndicate, allowing the CVC investor to enter markets it would have had no access to by itself. Also, engaging in a syndicate may give rise to unique complementarities with the portfolio company’s capabilities, which allows the investing CVC to obtain a structural advantage (Chesbrough, 2000).

With regard to venture capitalists, Lerner (1994) suggests that syndication leads investors to make superior selections for portfolio companies, as the potential investment comes under more fierce scrutiny than would be the case, had there been just one investor, which can be seen as a double check of the potential of the investment.

The level of knowledge that the investing firm can use for further innovation has been shown to follow an inverted U-shape, with respect to the amount of CVC investments made, which is positively moderated by the level of knowledge diversity, which is another motivation for syndication (Lee, 2015). This means that initial positive effects from CVC investments in terms of knowledge sharing, increase as diversity increases. This further supports the argument brought forward in the current paper that diversity is an important

(17)

factor determining CVC performance. The stage of development of the portfolio company at which corporate investors engage in CVC differs from that of other kinds of investors, such as angel investors, which engage in the seeding stage, and independent venture capital.

As Dushnitsky and Lenox (2006) describe, the size of syndicates is influenced by the presence of CVC investors in the syndicate, in which the presence of CVC in the syndicate is related to larger syndicate size. This can be explained by CVC possibly having a facilitating role within the syndicate in the sense that CVCs have a fundamentally different objective when engaging in an investment than does, for example, an independent venture capital firm. Dushnitsky and Lenox (2006) also indicate that ventures that are backed by CVCs perform better, both in absolute and relative terms compared to portfolio companies that are backed only by non-CVC investors. This observation holds true both when the success of the investment in terms of IPOs is considered, as well as when deal values of acquisitions are considered (Ivanov and Xie, 2010).

It remains unclear, however, if this success is due to superior development capabilities or superior selection capabilities. Ivanov and Xie (2010) presents evidence of CVCs being able to select more profitable projects, but besides this effect, they also add value to their portfolio companies, which can be explained by CVCs holding relatively larger stakes in portfolio companies than do other investors, giving them a stronger incentive to get involved with the portfolio company’s daily business in order to improve it.

With regard to the composition of syndicates, Hill et al. (2004) find that cooperation of a CVC with independent venture capital firms leads to higher performance, both in financial terms, as well as in strategic terms, depending on the governance structure. Moreover, as Dushnitsky and Lenox (2006) indicate in their handbook on further CVC research, the influence of IP protection on the success of CVC merits further research. This means that countries have different characteristics that can influence their attractiveness for CVC. One of these characteristics determines the extent to which the CVC investor can be confident that the innovations it develops maintain their value to the firm, namely the strength of a country’s IP regime, which is discussed next.

(18)

2.6 IP regimes

The extent to which a CVC can benefit from the investments it makes, and the risk it takes in cooperating with startups as potential future partners, depends on the strength of the IP regime, especially that of the country in which the portfolio company is based, because this is where the CVC has a disadvantage with regard to experience with the local system. This relationship between strength of the IP regime and the likelihood of investments is influenced by the industry in which the firm and portfolio company operate . That is to say, if the firm and portfolio company are active within the same industry, the likelihood of CVC investments being established is higher when the IP regime is high, and vice versa. This can be explained by the fact that close ties with outside partners in the same industry bring along risk of imitation if the IP is not protected sufficiently (Dushnitsky and Shaver, 2009).

Gaining access to new resources can be one motivation for engaging in a syndicate. This does, however, bring along the risk of knowledge spillover and other shifts in resource allocation, eventually leading to a deterioration of the negotiation position of the focal firm (Katila, 2008). There may be a good potential match between a certain CVC and portfolio company, but cooperation may still be inhibited due to a lack of legal recourse in case of a lack of IP protection. In such cases, it may be attractive for both the CVC, as well as for the portfolio company to have a well-functioning protection system available, as this facilitates further business.

3. RESEARCH GAP, RESEARCH QUESTION AND

HYPOTHESES

3.1 Research gap

There has been extensive research on the effects of IP regimes in relation to the engagement of firms in CVC (Chesbrough, 2012; Dushnitsky and Shaver, 2009; Basu _{et al.}

,

2016). This stream of research mainly finds that it is desirable for CVCs to be active in environments with a relatively strong IP regime, as this allows investors to gain financially from any developments that result from an investment.

With regard to syndication, insightful research has emerged as well (Basu, 2016), indicating the added value of a CVC present in a syndicate of investors in the form of

(19)

improved innovation, as well as by providing different a different point of view within the syndicate. Moreover, syndication sends a positive message to other potential partners if previous partners have a good reputation. This can thus be used as an indication of quality by these partners (Ernst, _{et al.}

, 2005).

However, there has been limited research on these two factors, the strength of IP regimes and the characteristics of the investing syndicate, in conjunction. Extending academic knowledge in this area is important, because what literature tells us, is that IP regimes matter for CVCs, due to the risk of not realizing the full potential benefit of investments that is the results of a weak IP regime. This is a negative effect that might be mitigated by working together in a syndication, as partners can supply the CVC investor with experience on how to deal with said issue. Apart from that, there are other factors that could potentially influence this relationship, such as the syndicate’s diversity, as several different perspectives could lead to better solutions, and the amount of syndication partners, as this could improve the negotiation position of the syndicate with regard to the portfolio company.

3.2 Research question

Considering the limitations of the literature up until now, it can be concluded that there are several avenues for further research. The impact of institutions on the attractiveness to foreign investors, the attractiveness of syndication, the extent to which diversification impacts CVC, how knowledge is actually transferred between the corporate investor and portfolio company, the impact of relations of the corporate investor with its portfolio companies and corporate parent, all are corporate-level subjects that deserve further investigation.

On a lower level, of the portfolio company, there are possibilities for further research on the impact of managerial composition and how this composition influences innovation behavior and bargaining power, as these compositions do vary widely among portfolio companies. The characteristics of syndication of independent venture capital partners in terms of experience is discussed at length by Lerner (1994), but the extent of applicability to CVC investments remains unclear and needs further research.

Not all these issues can be resolved in a single research. However, as the goal of this paper is to combine the different levels of analysis, several elements are isolated, in order to

(20)

give an overview of CVC syndication in relation to IP regimes. This combines the firm level with the portfolio company level.

More specifically, this research revolves around the amount of syndication partners, their diversity, the experience of syndication partners, and IP regimes. This last dimension has a distinct focus on international business, as IP regimes differ among countries, which influences the amount of possible innovation (Dushnitsky and Lenox, 2006).

Therefore, this research focuses on how the syndication of CVC investment efforts impacts performance, and how this relationship is influenced by IP regime characteristics, resulting in the following research question:

How does the strength of an IP regime moderate the influence of syndicate characteristics on CVC performance?

This research question, which is used as a guide throughout this research, consists of three elements, namely the IP regime, CVC performance, and syndicate characteristics, the last of which is divided further into amount of syndicate partners, diversity within the syndicate, and experience present within the syndicate. These relations are visualized in the conceptual model in figure 1, which is discussed next.

3.3 Hypothesis development

In this section, the research question is further specified in hypotheses, from which a conceptual framework is abstracted, which is depicted in figure 1. This conceptual framework contains three syndication characteristics on the left side as independent variables, CVC performance as dependent variable on the right, and the strength of the IP regime as a moderating variable in the middle, influencing all three independent variables.

As the amount of partners in the syndicate increases, there is a larger opportunity to form a strong block in any negotiations, as partners can have each other’s back and thereby improve their bargaining position relative to the portfolio company. As a stronger bargaining position of the syndicate partners relative to the portfolio company leads to a less desirable outcome for this portfolio company, CVC performance as measured by amount of IPOs, may not be in the best interest of the portfolio company itself, but all the more so for the syndicate

(21)

members, as long as the syndicate members do not exploit their negotiation position at the expense of potential future gains from cooperation with the portfolio company. This is related to the reasoning followed later in this section, in the discussion of the hypothesis related to the experience within the syndicate. However, the perspective taken in the discussion of this hypothesis is different in that it is not the sheer amount of years of experience that is taken into account, but also the amount of different parties involved which are expected to improve the outcome of CVC performance. Hence, it is hypothesized that:

H1a: As the amount of syndicate partners increases, CVC performance increases

Having multiple points of view within a syndicate can have positive consequences in the sense that there are different ways in which problems facing a syndicate can be dealt with. However, whereas more syndicate partners being present in a syndicate can be expected to have a positive effect on the syndicate performance, the increase in this amount also brings along increased risks for the focal firm. For example, there are more parties to deal with, which could make it harder to maintain all of the firm’s relationships. One way in which such risk could be dealt with, in theory, is a stronger IP regime (Dushnitsky and Shaver, 2009), facilitating an improved cooperation between a larger number of syndicate partners and their portfolio company.

H1b: As the strength of an IP regime increases, the positive effect of amount of syndicate partners on CVC performance increases

As more partners with different kinds of investment experience engage in a syndication, there is a more diverse spectrum of experience available for the focal firm to make use of. This is a similar point to the one that is discussed in the previous section, but whereas before, the focus lied on the interfirm relationships, here the focus shifts towards the variety of the industries themselves. That is to say, having connections to several different industries allows the entire syndicate to respond more effectively to threats to their development, as individual partners possess the appropriate experience to deal with different situations.

(22)

H2a: As the diversity of syndicate partners increases, CVC performance increases

Individual firms may be hesitant to expose the knowledge and capabilities that give them their competitive advantage, even though it is exactly this kind mechanism that could add value to the syndicate itself. A strong IP regime can allow syndicate partners with a very specific set of knowledge to expose their knowledge in the best interest of the syndicate, as this strong regime secures these investors’ delicate position of exposing themselves to the benefit of the entire syndicate.

H2b: As the strength of an IP regime increases, the positive effect of syndicate partner diversity on CVC performance increases

As partners with more investment experience engage in a syndication, there is also more experience available for the focus firm to make use of within the syndicate. This allows the entire syndicate to respond more effectively to threats, as individual partners possess appropriate experience for different situations. Experience is described by Ernst _{et al. (2005)} as being an important factor that determines the success of CVC investments. Managers of the CVC need to have substantial professional experience, which allows them to assess investments better. Moreover, these managers should have a profound knowledge of the parent firm, in order to be able to transfer any gains from these investments to the parent firm, and to identify potential synergies between the CVC and the portfolio company. One problem, however, is that a combination of both a thorough knowledge of the parent company and entrepreneurial experience is rare (Ernst_{et al.}

, 2005). This does, however, imply a higher

performance of firms that do manage to combine these two factors.

H3a: As experience of syndicate partners increases, CVC performance increases

Combining the evaluation of IP regimes with that of syndicates, one can expect experience of syndicate partners to decrease the need for concrete protection of IP via governmental institutions. The reason for this is that the knowledge of syndicate partners can be used to mitigate any potential risk of loss of this IP, which follows the reasoning of Dushnitsky and Lenox (2005). Moreover, the behavior of firms is influenced by both formal

(23)

and informal constraints set by institutions, such as government protection, for example in the form of IP protection (North, 1990). This would imply that the firm would look for protection of IP through either government or syndicates.

H3b: As the strength of an IP regime increases, the positive effect of syndicate partner experience on CVC performance increases

It should be noted at this point that the unique elements of the hypotheses may have some strong causal relations with each other, considering the main idea in the argument that syndicate partners provide protection through their experience, amount and diversity.

Figure 1: Conceptual framework

All this taken together leads to a network of expected effects working together, influencing the success of the investment performance of CVC as depicted in figure 1. So summarize, it is hypothesized that as the amount of syndication partners, the diversity of the syndicate, or the total amount of syndicate experience within a syndicate increases, so does the investment performance. Moreover, it is hypothesized that these three direct effects get stronger as the strength of the IP regime increases, which is thus expected to have a moderating role within this framework.

(24)

4. METHODOLOGY

4.1 Research type

This research has an empirical character and aims to find a relationship based on data gathered from databases both freely accessible via UvA and publicly accessible. Instead of looking for a causal relationship, by demonstrating covariation in panel data, which would be present in an explanatory study, this project has a more exploratory focus, in which the aim is a view of CVC that is generalizable to the extent that it can be considered an extension of current research on the topic of CVC, and a descriptive focus in that the current situation in CVC with regard to institutions and syndication is represented accurately. Therefore, this study has a quantitative approach in which exploratory research relationships is performed through databases such as, primarily, ThomsonONE and the World Intellectual Property Organization.

4.2 Sample

The used sample is as large as possible, while keeping the focus on larger international corporate investors in order to minimize bias and consists of CVC investments from different industries, using different technologies, in order to connect with the main research objective. Moreover, as the databases that are available are extensive enough, it is possible to obtain a random sample, instead of a convenience sample which would have led to biases. A main limitation of this choice is that subtle, or not so subtle details may get lost in this process, which would have been picked up if a different research method were to be used.

The data is statistically analyzed using the SPSS statistics software package, in order to form generalizable results from a large sample. Such a decision involves the trade-off between a high degree of bias, as would be present in case of the use of expert opinions, surveys, and case studies on the one hand, and a low degree of bias, which would be present in case of (quasi-)experimental studies. The main downside of taking this approach is that qualitative details behind the reasoning of key people within firms get lost.

Regarding feasibility, this quantitative approach using databases, especially the databases that available through the UvA, is an appropriate choice, as databases such as ThomsonOne have extensive information available on specifically CVC, and for a broad

(25)

range of countries. The main aim in the evaluation is to achieve high internal and external validity, high reliability with low biases employing an objective research approach with a relatively large samples. The concepts discussed in the literature review section can be measured in different ways and therefore, a main goal is to verify the validity of these measures.

The sample that is used in the analysis covers the period of 2005 up until 2015 and is obtained via ThomsonOne, which is the standard database for research related to CVC investment (Bielesch _{et al.}

, 2012). It consists of data related to the portfolio company, CVC

and firm, used for identification, the investment date, fund stage, and other dates, used to evaluate the experience, nation, in order to match the strength of the IP regime from the data of WIPO, and information on syndication partners type to evaluate the diversity, and other data such as industry such as fund type, fund status, used as control variables.

Using this information, the composition of the syndicate investing in the portfolio company can be identified and the fund nation’s IP regime can be analyzed. Moreover, the extent to which the exits were successful is analyzed by considering exit types. The IP regime strength is evaluated using data from The International Property Rights Index 2016. This data can be analyzed by contrasting the different columns that describe the different portfolio companies, the dependent variable (CVC performance), the independent variables (syndicate partner characteristics), moderator (IP regime) and other variables.

With every restricting factor concerning data requirements, the sample gets smaller, which results in a sample that is between 7081 observations in the most demanding case, and 7900 observations in the less demanding case.

The dataset was initially cleaned by considering only portfolio companies in which more than one single party invested, as this research revolves around investment through syndication. Moreover, portfolio companies in which only non-CVC investors had invested were excluded, as the focus of this research is not so much on comparing CVC performance with the performance of other forms of venture capital, as it is about evaluating the characteristics of the investments that are successful. For the syndicates that did include a CVC party, the amount of different kinds of investors were considered as well, and by taking the total amount of different investors within the portfolio company, the diversity of the syndicate was calculated.

(26)

4.3 Empirical strategy

In order to get a clear overview of the relationships between the firm, the CVC and the portfolio company, data was gathered from Thomson One. This data was put into three separate files, containing three levels of analysis, namely the firm level, fund level, and portfolio company level.

The strength of IP regimes is examined through the database on IP by the World Intellectual Property Organization (WIPO) and the International Property Rights Index 2016. Consequently, the previously discussed relationship between syndication and investment performance can be reviewed while taking into account the strength of IP regimes.

4.4 Measures

The main focus of this research is visually depicted in the conceptual model in figure 1, describing the relationship of the syndication characteristics on success of CVC, and the influence of IP regimes. The dependent variable of success of CVC investments is measured using exit type, in which an IPO is preferred to an acquisition (Dushnitsky and Lenox, 2006). The experience aspect of syndication partners is measured by considering the amount of years the partner has been active as a (corporate) venture capital investor at the time of the investment.

The diversity and amount of partners are evaluated by considering the types of investors that the syndicates consist of and their importance. The IP regime could be measured by considering measures such as patents awarded, but as obtaining this data may not be feasible within the scope of this research, secondary classifications of countries’ IP regime evaluation are used, namely The International Property Rights Index 2016, which is added to the database using Excel’s Vlookup-function. This allows for a quantitative analysis of the investment characteristics, as the ratings of the strength of the countries’ IP regime is rated on 10-point scale. Investor partner experience is evaluated using the years of activity by other firms as indicated by ThomsonOne. Diversity and amount of syndication partners are obtained in a similar way.

The experience present within the syndicate is calculated by taking the sum of the amount of experience of each party within the syndicate up until the time of investment. The IP rating is calculated by taking the average of IP ratings of the individual source countries, as

(27)

judged by the WIPO. Moreover, the amount of countries present within the syndicate is taken from the ThomsonOne database by simply taking the sum of the unique source countries.

The dependent variable is measured using the exit of the portfolio company. To evaluate this, for each portfolio company, it is considered whether there was an IPO of this portfolio company within the time period of 2005-2015. When this is the case, this instance is coded as a 1, denoting a successful CVC investment, whereas 0 denotes the absence of such an IPO.

Regarding the industry in which the portfolio company is active, the ThomsonOne Database provides several options for classification, with some measures containing many different industry types, and some containing less. For this research, ThomsonOne’s ‘Company VE Primary Industry Class’ classification is used, as the amount of different kind of industries is manageable statistically, and the classifications are relevant for the current research. That is, portfolio companies fall within either of three categories; information technology, medical/health/life sciences, and non-high technology. For these categories different dummies are created.

The portfolio status is used as an indicator of whether the portfolio companies is currently backed by private equity or not. Moreover, several other factors also have an impact on the dependent, which are included as control variables.

Private equity investments are usually required to have a return after four to seven years (Hungarian Private Equity and Venture Capital Association, 2016) and the success of the exits of the investments is generally assessed by considering the exit type of the investment. A favorable exit type for the investments is an IPO, in which the portfolio company becomes available on the stock market. It is desirable due to the fact that it allows the investors to gain the largest possible return on their investment, even though such IPOs also come with risks. That is, when an IPO is issued, it is unlikely that all stocks are immediately sold, which means that the issuer does not necessarily exit the investment exactly at the time of IPO, subjecting the issuer to price fluctuations. Besides this risk, the process of issuing an IPO is expensive by itself.

Such high costs are not present in the case of a trade sale in which the portfolio company is sold to a third party, often within the same industry as the seller, which can be classified as merger and acquisition (M&A). It does not impose as strict regulations on the transaction as is the case with an IPO, and negotiations are often more efficient, which makes

(28)

trade sales an even more desirable exit type, although a main downside is that such a sale to an industry peer is likely to lead to competitors obtaining valuable business information.

Another possible exit type is a secondary sale in which the portfolio company is sold from one private equity fund to another, which is an instrument used mainly by CVCs that need to sell the portfolio company before it is ready for an IPO, due to for example incompatibility on a business level (CVC Capital Partners, 2016).

4.5 Dependent variable

The dependent variable is whether the investment ended in an IPO, which indicates success, or whether it did not, which indicates a lack of success. Even though in reality, success of a CVC investment can not readily be evaluated, this is a good indicator. Moreover, there are many different possible outcomes for CVC investments, but IPO is the most desirable one. This is a similar approach to that taken by other authors in the CVC literature, such as Dushnitsky and Lenox (2006). Moreover, the time period of 10 years is similar to the usual approach in academic papers on CVC.

Since the dependent variable is binary, that is, it is either IPO, or it is not, indicated by a 1 and a 0 respectively, and should therefore be estimated using binary logistic regression Binary Logistic Regression (Hidalgo _{et al.,}

2014).

4.6 Independent and moderating variables

The first independent variable is the amount of investors that invested in a given portfolio company, which indicates the size of the syndication. This variable is found by considering all investments in a particular portfolio company and then finding the amount of individual investors in this portfolio company.

The second independent variable is diversity, which is indicated by the amount different fund investor types. This is where a choice must be made between whether to look at the actual activities of the fund, or at the organizational structure of the fund. As there is more information available on the organizational structure of the funds, in terms of fund investment type, this is the measure of diversity that is used in this paper.

The third independent variable is syndicate partner experience. Initially, the amount of experience per investor is found by calculating the difference between the foundation date of

(29)

the fund, and the date of the investment. Since there are multiple investors for a given investment, the amount of years of experience of each investor is added up to arrive at a total amount of years of investment.

The moderating variable is the strength of the IP regime, which is a rating obtained via WIPO on a scale from 1 to 10.

4.7 Control variables

As discussed above there are several factors already known to affect the success of CVC performance, which are controlled for.

Firstly, it is known that, despite a shift of the center of gravity, almost all CVC investments take place in the US and therefore, fund nation is controlled for by considering the size of GDP. Secondly, related to the diversity argument brought forward before, diversity of knowledge within the syndicate allows for a more agile syndicate that is better able to respond to unexpected changes. One way in which this diversity comes about is by having a broad range of source countries within the syndicate, which allows for different insights from different cultures. Another factor that must be controlled for is whether the portfolio company is active or not, and if so, for how long, as well as whether they are currently backed by private equity. The last element of the controls concerns the industry within which the portfolio company is active, as there are large discrepancies between the amounts of CVC investment in different industries. That is, the industry in which the portfolio company is active is known to affect the outcome of the CVC investment (Dushnitsky and Shaver, 2009), and is controlled for accordingly.

Following Dushnitsky and Lenox (2005), firm size affects the approach firms take with regard to either internal or external development, for example in terms of R&D (Cohen and Levinthal, 2012), and therefore the focus of this research is on larger firms.

There are many factors that impact the success of CVC investments, such as the fit between the CVC investor and portfolio company, which are captured in the control variables and independent variables. The data in this research covers a 10-year period from 2005 up until 2015 in which many large firms tried to gain from investing in portfolio companies. The success of these investments is measured by considering the exit mode, as is done by convention (Dushnitsky and Lenox, 2006).

(30)

5. STATISTICAL MODELS

5.1 Correlation matrix

Before considering all of the variables in combination with each other in the regression analysis in the hierarchical model depicted in table 2, the individual variables are evaluated in relation to each other by considering their bivariate correlations in a correlation matrix depicted in table 1. Most of the data in this matrix looks acceptable, as there is rather no multicollinearity present.

However, it can be observed that the amount of syndicate partners and syndicate experience are significantly correlated with each other with a value of 0,796, but this is the only instance in the correlation matrix of a relatively high correlation. Moreover, the IP rating has a negative effect in the correlation matrix observed, which is not expected. However, this is not the case anymore in the hierarchical model which follows after.

5.2 Hierarchical models

In the analysis of the data, several models are applied to the data, starting with the first model consisting of the dependent variable, and the controls. This is done to evaluate the effect of these known factors on CVC performance in isolation of of any of the other possible factors. Next, the independent variables are added in model 2, which allows for the evaluation of these variables without the effect of any of the moderating effects that follow after. Model 3 adds the first moderation effect between the portfolio company rating and amount of investors, which is done by taking the product between these two variables. This is then compared with the previously added independent variables and control variables. Model 4 includes the second moderation, namely that between portfolio company rating and diversity and lastly, model 5 includes the third moderation effect between portfolio company rating and experience in the syndicate. Further evaluation of moderation between all the moderators in combination would have been possible, but such an analysis is outside of the scope of the current research, and more data points would have been required to investigate this.

Regarding the data in the hierarchical model, the number of used observations is highest in the model in which only the controls are included, with N = 7834, and in the other

(31)

4 models, in which variations of the independent variables are used, the number of used is lower, due to missing data, but the amount of data is still highly usable with N = 7081.

The amount of variance explained initially with only the controls, in terms of R-squared, is 0,398. This is already a relatively high amount of variance explained, but one should bear in mind that this model already includes several important indicators of success of IPO.

The addition of the independent variables in the second model increases R-squared to 0,438 and the addition of the moderators increases the variance explained slightly. That is, the R-squared of model 3 is 0,443; the R-squared of model 4 is 0,442; and the R-squared of model 5 is ,445. Moreover, all these models are significant with Chi-squared significance levels of below 0,000, thereby passing even more conservative significance tests of 0,01.

(32)

(33)

(34)

6. RESULTS

Since the dependent variable in this research is binary, that is, the exit mode of the investment is either an IPO or it is not, a binary logistic regression should be used (Cohen _et

al.,

1984). Moreover, whereas probit analysis is better suited for designed experiments, in

studies using observations, such as the current one, logistic regression is a better option, and combining this with the two possible outcomes of the dependent variable, a binary logistic model is used. In order to test for moderation, separate moderation variables are created by multiplying the moderating variable, IP rating, with the values of the independent variables related to syndication characteristics.

The amount of syndicate partners follows a distribution that is skewed, with syndicates consisting of mostly two partners, with every higher amount of partners becoming less present. Moreover, by observing the histograms of the data, it can be observed that the independent variables are all right-skewed, but the relevant skewness statistics are within the acceptable range of 2,0. The skewness of the IP rating, however, is exceeding the recommended margins, but the sample size is relatively large, namely higher than 7000, which has as a result that more skewness is acceptable.

As can be seen from the binary logistic regression, in the first model, the controls that are most relevant are the amount of source countries, active company status, and the portfolio company industry classification. All these control variables have significance levels of p<0,05. This is in line with theory in the sense that, for example, having more source countries within a syndicate, gives it a higher performance, due to a higher degree of agility within this syndicate.

Moving on to model 2, it can be observed that the third and fourth independent variables, related to syndicate experience and IP protection respectively, are significant (P<0,05). The first two independent variables, however, related to amount of syndicate partners and syndicate diversity, are insignificant, with significance levels of p>0,05. These first two independent variables are related to hypotheses h1a and h2a and considering their insignificance, we can conclude that there is no direct effect of either the amount of variables leading to increased CVC performance, nor is there a direct effect of diversity of the syndicate leading to increased CVC performance. This is unexpected, but as described below, adding the strength of the portfolio regime changes the role of these variables.

The influence of syndication characteristics on investment performance of corporate venture capital

Master thesis