CORPORATE VENTURE CAPITAL: DOES EXPERIENCE MATTER?
James
Henderson, Babson College, Wellesley
Benoit
Leleux, IMD, Lausanne, Switzerland
ABSTRACT
INTRODUCTION
LEARNING, CORPORATE
VENTURE CAPITAL, AND HYPOTHESIS GENERATION
RESEARCH DESIGN
EMPIRICAL RESULTS AND
IMPLICATIONS
CONTACT
REFERENCES
FIGURE 1
TABLE 1
TABLE 2
This paper argues that given the substantial evidence of organizational learning, firms may be expected to improve their skills in resource transfer through previous experience in venture investing. Hypotheses are developed both for proprietary learning, learning at the industry level and forgetting on a database consisting of 34 companies operating corporate venture capital programs in the worldwide networking related industries in the US, Europe and Japan. The findings suggest that while resource transfers are associated with a greater likelihood of liquidation events, proprietary learning does not seem to be the main driver. Rather, there is substantial evidence of vicarious learning or copying what others do.
This paper examines how firms may learn to improve their corporate venture capital investing through their own and their industry cohorts’ past experiences. Corporate venture capital (CVC) programs, through which large, established companies make minority equity investments in promising start-up enterprises, have long been recognized as important and strategic activities in either sustaining or renewing profitable growth in large corporations. These programs typically involve two value adding roles in addition to acting as independent venture capitalists: leveraging the resources of the corporation; and accessing the resources of the venture. These roles would thus suggest that the returns for CVC programs be at least as high as independent venture capital programs. While recent studies by Gompers and Lerner (1998) and Maula and Murray (2000a; 2000b) demonstrate how these roles add value, many corporations have still been very frustrated by their corporate venture capital programs. Indeed, a recent Bain study showed corporate venturing as one of the least applied and least satisfying strategic programs used (Bain, 2001).
Researchers have provided numerous reasons why CVC programs might not have been fully effective. First, a well-defined mission for the corporate venture capital activity may not have been provided (Fast, 1978; Siegel, Siegel and MacMillan, 1988). Top management often seeks to accomplish multiple potentially incompatible objectives, such as gaining access to emerging technologies and, at the same time, generating attractive financial returns. Second, the commitment to corporate venturing has often been limited, disappearing when the executive champion was reassigned (Hardymon, DiNino and Salter, 1983; Rind, 1982; Sykes, 1990). Third, business unit managers may resist supporting venture capital efforts as they either prefer funds to be allocated to their internal programs, don’t recognize the potential resource transfers, lack the incentives to work with the start-ups, or view the start ups with suspicion (Henderson and Leleux, 2002). Fourth, corporate venture capitalists may hold onto their losing investments longer than independent VCs as there is a greater “strategic” interest in it. Finally, corporations have frequently been reluctant to compensate their venture managers through “carried interest” provisions, i.e. direct equity stakes in the ventures, fearing (1) that they might need to make huge payments if their investments were successful, (2) that it might create a double culture in the company and a lot of disruptive envy between those working hard with and without equity stakes, and (3) that it may elevate revenue expectations for all in the company (Block and Ornati, 1987; Barr, 2000). As a result, corporations were often unable to attract top people to their venture funds, leading to even less commitment to the activity (Hardymon, DiNino and Salter, 1983; Rind, 1982; Sykes, 1990). As a result of these reasons, CVC programs have often been short-lived. These past two years are no exception given the recent rapid decline in corporate venture capital investing and the shuttering of many programs. Indeed, it seems that CVC programs have not learnt much from their past mistakes.
Yet, this conclusion is somewhat surprising given the evidence that managers and their organizations try to learn. For example, the literature on the learning or experience curve shows the importance of learning for improving cost positions (see e.g., Yelle, 1979 for a review). The literature on the returns to R&D (see e.g. Mansfield, 1991) has consistently found that the social return to R&D is substantially higher than the private return to R&D, indicating that firms learn from each other. More recent research on R&D spillovers has documented how firms learn and that some firms have a higher “absorptive capacity” or learning rate than others (Cohen and Levinthal, 1990). This type of learning is further supported by evidence collected by Keil (2000) on Scandinavian external corporate venturing programs. Extending this view, the “resource based” view would argue that firms’ skills in corporate venture capital investing should improve with time as they accumulate a capability or stock of experience in making those decisions (see e.g., Dierickx and Cool, 1989).
The objective of this paper is to examine the incidence of learning in corporate venture capital programs. We focus on one aspect of learning: whether firms learn to transfer resources in corporate venture capital programs. The setting for the empirical analysis is the network-related industries including telecommunications, cable, wireless and satellite network operations. These industries have experienced extraordinary rates of change over two decades including, deregulation, privatisation, new technologies and the introduction of the Internet. To cope with this environmental change, companies in these industries have resorted to a number of strategic programs including internal research and development, joint ventures, vertical partnerships, technology licenses, product market licenses, acquisitions, internal corporate venture and corporate venture capital activities. Some rivals, such as ATT, MCI and the Regional Bell Operating Companies have invested large amounts of capital in major acquisition programs either to consolidate their markets or to commit to particular access technologies. At the same time, corporate venture capital investments were initiated as a way to gain access to innovations in the marketplace for items such as security, e-commerce engines to improve their services, or optical components and other networking hardware to improve their network operations. Out of the 300 firms we identified in these industries, approximately 10% had started a venture capital activity, some as early as 1974. Those 33 firms identified have made approximately 500 venture investments some in the same companies, ranging from a low of 1 investment to a high of 91 investments for one company.
LEARNING, CORPORATE VENTURE CAPITAL, AND HYPOTHESIS GENERATION
The paper first briefly reviews the major findings in the corporate venture capital literature. Then, we derive three main hypotheses drawing on organizational learning research, which is described below:
Firms are said to learn from several sources, the main one being from the experience of doing (Yelle, 1979). However, this proprietary learning by doing may be subject to “leaks” due to spillovers i.e. less experienced firms may learn from observing more experienced firms (Huber, 1991, Miner and Haunschild, 1994) or due to “forgetting” or to the depreciation of the stock of knowledge (Levinthal and March, 1993). Hypotheses about these three areas of learning are developed next.
Learning By Doing (Knowledge Creation)
Organizations are said to ‘learn by doing’ when the ‘efficiency’ of a routine, such as planning or manufacturing, improves (see e.g., Yelle, 1979 for a review). In general, to become more efficient, a routine requires repetition. However, as a routine is repeated, firms may also develop capabilities for ‘effectiveness’ (Levitt and March, 1988). For example, Cohen and Levinthal (1990) found that firms with larger stocks of research and development experience developed a greater “absorptive capacity” or a capability to see the value of external research and development information, to understand it and deploy it commercially.
Learning by doing can be applied to corporate venture capital investments. For example, after an investment has been made, many venture capital managers discuss how the investment process can be improved as well as how additional value can be created (Henderson and Leleux, 2002). As a result, they may develop better ways to ensure some sort of resource transfer actually occurs between the business units and venture investment. For example, they may gain up front commitment by including business unit managers in the due diligence process; or they may suggest that the business unit manager sit on the board of the start up. As a result of this experience, they may be more likely to encourage a relationship and resource transfer between the venture company and the business unit. Furthermore, as more investments are made, the process to ensure a resource transfer process may become more effective. If this were the case, long standing CVC programs would have a significant advantage over short-term programs. This leads to the following hypotheses:
Hypothesis 1: Ceteris paribus, resource transfers will more likely occur the larger the stock of previous proprietary CVC investments.
Learning Through Observation (Knowledge Externalities)
However, we have assumed there is no “leak” or decay in the stock of experience. In other words, it simply accumulates over time through successive corporate venture capital investments. If this were the case, firms with a larger stock would always be at an advantage over firms with fewer investments when studying the next investment opportunity. However, this stock of “proprietary” experience may leak due to spillovers, i.e. less experienced firms may learn from more experienced firms (Levitt and March, 1988). It is indeed quite plausible that learning may come from observing the outcomes of other companies (Miner and Haunschild, 1995) or being involved in knowledge-sharing networks (Lee, Lee and Pennings, 2001). In the R&D literature, for example, Mansfield (1991) noted that the private returns from innovation were much lower than the social returns and argued that the difference was due to intended or unintended spillover effects. Furthermore, the link between actions and outcomes may not be entirely clear. Since a result may be several years out, feedback is delayed. Sterman (1989) argues misperceptions of feedback or the failure to understand delays between action and outcomes can result in sub-optimal behaviour in complex settings. For this reason, players may look to each other for answers.
Applied to corporate venture capital investing, firms may be able to collectively learn from each other through benchmarking exercises and seminars. As a result of these spillovers, late entering firms in corporate venture capital may learn to better leverage their venture investments more quickly than their more experienced counterparts leading to the following hypotheses:
Hypothesis 2: Ceteris paribus, resource transfers will more likely occur the larger the stock of all industry participants’ CVC investments.
Forgetting (Knowledge Depreciation)
Yet, learning tends to be myopic (Levinthal and March, 1993). Based on the availability heuristic, actors are apt to weigh recent events more heavily than events more distant in a long history of experiences (Tversky and Kahneman, 1986). Thus, if routines are not repeated, memory or the stock of experience may depreciate.
Corporate venture capital investing can be an infrequent event. Any lessons from them may be forgotten by the next investment exercise especially if the investments are spaced apart over long periods of time. For example, the venture managers responsible for previous investments may have left the firm, or moved to another position. Given that few corporate venture capital programs conduct formal post mortem analyses, most of the learning could be lost (Henderson and Leleux, 2002). As a result, corporate venture capital programs may continue to suffer from the same problems that have plagued them in the past. The following hypotheses reflect this limit to learning:
Hypothesis 3: The longer the time from their last investment decision, the less likely resource transfer will occur in the subsequent decision.
In sum, two hypotheses have been proposed that suggest that firms with more proprietary experience will learn to encourage resource transfer. However, we suggest that this learning is subject to two forms of knowledge leakage. First, firms may collectively learn from each other’s investment experiences. Second, since corporate venture capital decisions can be infrequent, the knowledge gained from any previous investment experience may be subject to significant depreciation. From one decision to another, firms may simply forget the lessons of the past.
Sample
To empirically example the possible effects of learning in corporate venture capital programs, we examined the venture investments made in the fast paced worldwide network related industries between 1974 and 2000. We can cite several reasons for choosing this industry, which includes telecommunications, cable, wireless and satellite network operations. First, these industries have experienced extraordinary rates of change over the last several years, making them particularly attractive for this study. Advances in broadband technology have initiated a convergence of several industries: media and broadband, data communications and mobile, and information technology and telecom, to name a few. The emergence of the Internet combined with the growing numbers of telecommuters has meant increasing demand for access technologies with greater bandwidth. Yet, there has been no clear winner as to who will provide the best, most cost effective access technology, so that competition remains intense between cable, wireless, DSL, and satellite networks. In parallel, the established order has also incurred substantial changes in regulations, such as the 1984 break up of AT&T, the 1996 Deregulation Act passed in the US and widespread privatisation witnessed in Europe, including British Telecom, Teledanmark, France Telecom, and Deutsche Telekom. As a result, many of these companies have had to become far more customer oriented than during the period of regulation. As a result, these companies have resorted to many different business development initiatives to cope in this new environment. Corporate venture capital investments have been one of these strategic initiatives.
Secondly, the data is available on corporate venture capital programs in this industry. With the help of Venture Expert, and public information sources (Lexis-Nexis, Dow Jones etc.) we constructed a database tracking each of the venture investments made by these large corporations during the life of the corporate venture capital program. The database is therefore at the venture level of analysis containing data on founding date, country of origin, number of rounds, investment per round, the number of investors, the number of corporate investors, the number of telecom investors, the date of the relationship announcement (if any), the type of relationship made, patents before and after the corporate investment and the situation of the company at the end of 2002. We also have collected financial data on the each of the corporations. Finally, we added information on the venture capital market in general, when the investments were being made. In total, a dataset of 499 venture investments was developed spanning 34 network related companies operating in North America, Europe and Asia. Of those 499 investments, 350 were used in the estimations (due to missing data.)
Dependent Variable
To study the incidence of learning in corporate venture capital programs, we need to define the variable “resource transfer.” One could argue that in any investment there are resource transfers based on the various meetings taking place between the corporate venture capitalists, business units and the start up company. However, since there is no “event” surrounding them, they may not warrant much attention after the fact. Rather, we assume that a “resource transfer” must be an “event,” or recognizable after the fact. We, thus, defined a “resource transfer” as an agreement made public that the firms were going to work together. These public announcements were found in Lexis-Nexis and Dow Jones business wire and newswire information sources by searching on the name of the start-up at the time and the names of the various business units that were linked with the corporation. For example, when we searched for potential relationships between WebEx (a net meeting company previously known as Active Touch) and Deutsche Telekom, we keyed in the names of the business units (e.g. T-Online, T-Mart, T-Mobile etc.) to ensure that we covered all potential relationships. The articles were then read and highlighted where a specific relationship had been developed between the corporate investor and start-up. A binary variable was created which was coded as one when these start-ups had a relationship with their corporate investors. Further information was collected concerning the type, and the date of the relationship. The relationships tended to be clustered around three areas: the corporation as a customer of the product/service; the corporation as a distribution channel for the product/service or the business units and ventures working together to develop new products/services based on the technology from the venture. Two coders independently rated the first 10 start-ups. The coders then discussed the reasons for their ratings to ensure consistency. They then rated the remainder of the observations. For 20 start-ups there was disagreement (90% agreement on the other 479 firms). More extensive text search was done to resolve these disagreements. Where there were further disagreements, the coders called the start-ups to resolve the issue.
To start with the simplest model we set the dependent variable as a binary choice measure equalling 1 for all observations in which there was a relationship and 0 otherwise.
Explanatory Variables
The explanatory variables concern the measures of direct interest, i.e. learning and forgetting in corporate venture capital programs and suggest three effects. The first hypothesis linked prior experience in venture investing to higher likelihood of resource transfer in subsequent investments. To analyze this relationship, a stock variable was constructed. Unfortunately, there is no clear method for determining the content of the stock: the accumulation of time, or the accumulation of investments? Both have been used in previous operationalizations in the learning curve literature (see e.g. Levy, 1965; Lieberman, 1984, Henderson and Cool, 2003). As a result, the first, the accumulation of time was measured in months from the date that the corporation started its corporate venture capital program. The second, the accumulation of investments, simply counted up over time the number of investments that it had made prior to the one that it was investing in.
The second hypothesis was related to the effect of the venture investment experience at the industry rather than firm level. In this case, prior industry experience in venture investing could spill over to a higher likelihood of resource transfer for the firm. Once again, there is no clear method for determining the content of the stock variable. We resorted to the accumulation of previous investments in the industry (less the investments made by the focal corporation).
The third hypothesis concerns the possible declining effect of experience over time. In this case, a measure of forgetting was developed. Some studies on learning have estimated the decay of a stock of experience either by applying a depreciation rate (Henderson and Cockburn, 1984) or by imposing a geometric lag structure (Griliches, 1984). However, the annual flows in these studies are smooth. In this study, however, corporate venture capital decisions can be rapid or infrequent. As a result, we resorted to a simple measure that counts the number of months between investments.
Control Variables
Control variables were clustered around four areas: characteristics of the corporate venture capitalist: the organization structure of the corporate venture capital program, and the level of commitment to the venture; characteristics of the venture: the age of the venture, the round of the investment, the number of patents before and after the investment, and finally characteristics of the ventures’ environment: the number of other corporate venture capital investors, the number of other investors in network operator industries, and finally the amount of venture capital funding at the time of the investment. Each one is briefly discussed.
Characteristics of the Corporate Venture Capitalist: It could be argued that the organizational structure of the corporate venture capital program rather than learning or the accumulation of experience would bias them toward investing for return on investment (pick good investments) rather than investing for the development potential (seek future partners.) As a result, we placed the corporate venture capitalists in three categories (as coded in Venture Expert): Wholly Owned Subsidiary, Independent Affiliate and Direct Investor. Wholly Owned Subsidiaries referred to those programs (such as AT&T Ventures) who were still 100% owned by their mother corporations and belonged in the corporation’s organizational structure. Independent Affiliates were those programs (such as Venture Management Services) which had a relationship but arm’s length with the corporation (in this case, AT&T). Finally, direct investors were those who invested in the companies from headquarters where no separate subsidiary had been developed (such as Williams Communications.) It was expected that those programs that were Wholly Owned Subsidiaries were more “strategic” in nature and would, thus, search for investments in which they could engage in resource transfer.
We argue that the level of commitment to the venture would more likely result in a relationship being developed. Unfortunately, since we did not have the contributions of each investor by round, we had to resort to overall investments made. Commitment was therefore constructed as the percentage of corporation’s investment to the total investment in the company. While this measure is crude, it likely approximates the corporation’s commitment to the start-up, relative to the other investors.
Characteristics of the Venture: Many of the variables that have been included in previous studies on the venture were added in this one as well. These included the age of the venture (prior to the corporation’s first investment), and the round of the investment (see Gompers and Lerner, 1998). Consistent with interviews held with these corporate venture capitalists in this industry, we expected that the later the round and the older, more established the start up the more likely the corporation would initiate a relationship with it. There has to be sufficient incentives for the business units of the corporations to engage in any form of relationship. In our interviews, we found anecdotally more transfers occurred with investments beyond the seed and start-up phases, i.e. second and later rounds of financing. As some of the respondents mentioned, these ventures already had a prototype or a product that was ready for market introduction, reducing significantly the required time commitments and incubation services the business unit manager were willing to provide.
Furthermore, we expected that those companies that had already developed a large stable of patents were better candidates for a strategic relationship that those that did not. It has been shown in previous literature, that technology based knowledge (in the form of patents) are positively associated with the formation of strategic relationships (Hagedoorn and Schakenraad 1994; Grandstrand, Patel and Pavitt, 1997 and Kelley and Spinelli, 2001). While we would have preferred to use patent citations as our measure of patent quality, due to data limitations (unavailability of patent citations for European patents), we resorted to a simple patent count prior to and after the investment.
Characteristics of the Environment: We also included a measure of endorsement, competition and slack in the environment. Endorsement was calculated by counting the number of other corporate investors in the start-up prior to the focal corporation’s investment. Competition for the start up’s resources was measured by counting the number of other investors in the same industry who funded the start up prior to the focal corporation’s investments. Finally, slack in the environment was determined by using the dollar flows in the venture capital at the time of investment. We expected endorsement and competition from other investors from the same industry to be positively related to the development of a strategic relationship. Conversely, more slack in the environment was expected to be negatively associated with the development of a relationship.
Estimation Method
The relationship between the likelihood of establishing a strategic relationship and its explanatory variables was analyzed through probit regression analysis. The general form of the estimated equation is:
where EXPi stands for the three learning variables, connected to the three hypotheses: proprietary learning, learning from others and forgetting.
EMPIRICAL RESULTS AND IMPLICATIONS
Descriptive Statistics
Does resource transfer in corporate venture capital programs matter? We decided to answer that question before proceeding to the testing of the hypotheses. Two arguments can be made. First, the resource transfer helps out the business unit, either in terms of research and development (e.g. a window on new technology), or in lowering costs (e.g. a software engine for e-commerce transactions) or in terms of increasing direct sales (e.g. the increased use of web based meetings over higher broadband networks). Secondly, the resource transfer may aid the start-up by increasing its added value, through endorsement and credibility (see e.g. Maula, 2001), potentially increasing its liquidation potential. While we have no data to answer the first question, we do have some to address the second. Figure 1 provides some clues on whether forming relationships in corporate venture capital programs matter for subsequent liquidation of the companies. While we have not yet calculated the size of the liquidation distributions (see e.g. Gompers and Lerner, 1998), we do have some interesting observations using the change in status of the start up as an indicator. As one can see from Figure 1, for those start-ups that had received corporate venture capital funding prior to 2000 (3 years lag), they were first more likely to be liquidated either through acquisition or through an IPO. These relationships, in general, hold for a 3 – 8 lag in a probit analysis. (Note that they do not hold for more recent investments (2000, 2001, 2002.) This finding, while not fully explored, certainly has implications for the type of corporate venture capital program that corporations may want to exploit. It also suggests that past experience in developing these alliances may matter even more, giving more experienced corporate venture capital firms a leg up on those less experienced.
Table 1 shows summary statistics for the variables that are used in the estimations. All of the key variables show substantial variation. On average, approximately 56% of start-ups had engaged in a strategic relationship with their corporate investor. The majority of the investments came from corporate venture capital programs which were wholly owned subsidiaries (57%) rather than independent affiliates (27%) or direct investors (16%). The commitment to the venture also ranged from a negligible amount to 100% with the average at 18%. Typically the ventures were about 3 years old before receiving an investment from the corporate investor, which on average was slightly less than Gompers and Lerner’s (1998) sample of 3.9 years. However, the round average was approximately the same at 2.10. On average few of the start-ups had patents granted prior to the investment; however, a mean of 2 patents were granted after the investment. The number of other corporate and telecom investors ranged dramatically from 0 to as many as 16 prior to the focal firm making its investment in the start up. The stock of investment experience variable shows the number of investments made. While the average shows approximately 20 investments, the maximum illustrates the size of the largest corporate venture capital program at 90 venture investments. Interestingly, the time between investments on average was approximately 5 months; however the range was quite surprising from 0 (the same month) to 204 months later (i.e. 17 years.) Finally on average the time in a corporate venture capital program averaged around 68 months (i.e. 5 years) with the maximum being 300 months (i.e. 25 years).
Probit Analysis
Regarding the explanatory and control variables, previous research and hypotheses led to the following predictions for the estimated effect on the probability of a strategic relationship being developed with a corporation’s venture investment:
Positive Sign Negative Sign
Time
Forgetting
Previous corporate experience
Slack
Previous industry experience
Organization Structure: Wholly Owned Subsidiary
Commitment
Age
Round
Patents
Endorsement
Competition
Table 2 shows the results of the probit analysis. We show two models corresponding to the types of proprietary experience: time, and the stock of previous investments. The first column of Table 2 lists the independent variables and their values at the sample mean. The following columns report the parameter estimates and their t-statistics. The parameter estimates are the partial derivatives of the probability of developing a relationship with respect to the independent variable, calculated at the sample mean.
First, it is important to take note of the various control variables. We find, as expected, that corporate venture capital programs, which are wholly owned subsidiaries, committed to the venture and invest in later rather than earlier rounds more likely establish a strategic relationship with the start up organization. Curiously patents and the age of the start up did not turn out to be significant. While we can only speculate, age may be increasingly less important a criterion as entrepreneurs over time may be better able to assemble a team of highly skilled individuals to develop a commercializable products. Patents may not be significant because their patent quality is not reflected in the variable that we constructed. Unfortunately, European data on patent citations is not available.
Secondly, when we examine the impact of the environment some very interesting results appear. First, endorsement from other corporate investors seems to create a bandwagon effect. This finding is consistent with research already performed in the field (see e.g. Maula, 2001). However, this endorsement is even stronger when we look at the number of other network related companies that also invested in the same company. Clearly this competitive endorsement creates even more incentive for these corporate venture capital programs to develop a relationship with their venture investment. Furthermore, it seems that the amount of venture capital funding may also play a role in whether corporations seek a relationship with their venture investments. While only a weak negative relationship, the results do suggest that more funding may result in more opportunistic, non-relationship building investments. Indeed as “good investment” prospects were plentiful, as they presumably were in 1998, and 1999, the height of the Internet boom, then the discipline for choosing the right ventures for developing relationships with the business units may have waned.
Finally, the variables of interest also show curious results. The learning and forgetting hypotheses (1 and 3) were outright rejected, whereas learning from the industry could not be. In other words, corporate venture capitalists seem to look to each other for answers rather than rely on their own proprietary knowledge built through experience. While we can only speculate, this finding may be due to the fact that venture investments tend to be longer term in nature (i.e. 4 to 5 years); hence, the feedback regarding their ultimate success is delayed. In many cases, due to uncertain and long term payoffs, corporate investment managers may not have much incentive (until the venture has been successfully liquidated) to truly examine the benefits of effecting the establishment of strategic relationships. As a result, they may look to others for their cues as the empirical results illustrate.
Conclusions
This paper examined whether experience from previous venture investment decisions was a good teacher using a sample on the worldwide network related industries during the years 1974–2002. It was assumed that firms, which accumulate stocks of experience related to venture investments, would form more relationships in their subsequent investments. Thus, given some evidence that those investments with a strategic relationship indeed are more likely to be liquidated, experienced firms could take advantage of their less experienced rivals by initiating strategic relationships with their investments more often. However, following the R&D literature, we suggested that learning may not be fully proprietary and that firms may learn through the observation of each other’s outcomes.
The results were surprising. Rather than rely on proprietary experience, it seems due to the highly uncertain nature of venture investing, corporate venture capital managers turn to others outside for answers: other corporate investors, investors from their own industry, or simply the experience of the industry as a whole. These results also suggest that proprietary learning in corporate venture capital programs is either non-existent or subject to significant spillovers.
Clearly, there are limitations to this research. First, we have not yet determined the return on investment for those ventures that had a relationship and for those that did not. This weakness is certainly the next avenue of research. Secondly, while proprietary experience may not aid in increasing the likelihood of establishing subsequent strategic relationships, it may aid in the speed to which these relationships are established. Furthermore, it may also lead to more quality relationships and higher likelihood of liquidating the ventures. No doubt, this is another avenue of research that we must pursue before making any definitive conclusions. Finally, this research is only focusing on one rather broad industry: network operations. In order to generalize any of our findings, more research would have to be conducted on other industries. In spite of these limitations, we hope that we have shed further light on the topic of learning in corporate venture capital programs.
CONTACT: James Henderson, Babson College, Babson Park, MA, 02457; (T) 781-239-4290; (F) 781-239-4556; henderson@babson.edu
Bain and Co., (2001) Annual Survey on Management Tools, Bain Report.
Block, Z. and I. MacMillan, (1995). Corporate Venturing, Creating New Businesses Within the Firm, Boston, MA, Harvard Business School Press.
Block, Z. and O. Ornati, (1987). “Compensating Corporate Venture Managers,” Journal of Business Venturing, 2: 41–52.
Cohen, W. M. and D. A. Levinthal (1990). “Absorptive Capacity: A New Perspective on Learning and Innovation.” Administrative Science Quarterly 35: 128–152.
Dierickx, I. and K. Cool (1989). “Asset Stock Accumulation and the Sustainability of Competitive Advantage.” Management Science 35(12):
Fast, N. (1978). The Rise and Fall of Corporate New Venture Divisions, Ann Arbor, Michigan, UMI Research Press.
Gompers, P. and J. Lerner (1998). “The Determinants of Corporate Venture Capital Success: Organizational Structure, Incentives and Complementarities,” NBER Working Paper Series.
Granstrand, O., P. Patel, & K. Pavitt. (1997) “Multi-Technology Corporations: Why They Have ‘Distributed’ Rather Than ‘Distinctive Core’ Competencies.” California Management Review. 39(4): 8–25.
Griliches, Z. 1984. Patents and Productivity. Chicago University Press: Chicago, IL.
Hagedoorn, J., & J. Schakenraad. (1994) “The Effect of Strategic Technology Alliances on Company Performance.” Strategic Management Journal. 15(4): 291–309.
Hardymon, G., DeNino, M. and Salter, M. (May–June 1983), “When Corporate Venture Capital Doesn’t Work,” Harvard Business Review, 61: 114–120.
Hellmann, T. (2001). “A Theory of Strategic Venture Investing.” Journal of Financial Economics (forthcoming).
Henderson, J. and K. Cool (2003). “Learning to Time Capacity Expansions: An Empirical Analysis of the Worldwide Petrochemical Industry, 1975–1995, Strategic Management Journal, 25 (5): 393–415.
Henderson, J. and B. Leleux (2002). “Corporate Venture Capital: Effecting Resource Combinations and Transfers” Babson Entrepreneurial Review.
Henderson, R., Cockburn, I. 1996. “Scale, scope and spillovers: the determinants of research productivity in drug discovery.” RAND Journal of Economics 27: 32–59.
Huber, G. (1991). “Organizational Learning: The Contributing Processes and the Literatures.” Organizational Science 2: 88–115.
Keil, T. (2000). “External Corporate Venturing: Cognition, Speed, and Capability Dvelopment,” Doctoral Dissertation, Institute of Strategy and International Business, Helsinki University of Technology.
Kelley, D. and S. Spinelli, (2001). The Role of Corporate Investor Relationships in the Formation of Alliances in Corporate Venture Capital Funded Startups. Paper Presented at the Babson College Kauffman Foundation Entrepreneurship Research Conference.
Lee, C., Lee, K. and J.M. Pennings (2001). “Internal Capabilities, External Networks, and Performance: A Study on Technology-Based Ventures.” Strategic Management Journal 22: 615–640.
Levinthal, D. and J. March (1993). “The Myopia of Learning.” Strategic Management Journal 14: 94–112.
Levitt, B. and J. March (1988). “Organizational Learning.” Annual Review of Sociology 14: 319–340.
Maula, M. and G. Murray (2000a). “Corporate Venture Capital and the Creation of US Public Companies: The Impact of Sources of Venture Capital on the Performance of Portfolio Companies”, forthcoming in Creating Value: Winners in the New Business Environment. Hitt, M.A., Amit, R., Lucier, C. and R.D. Nixon (Editors). Blackwell Publishers: Oxford, UK.
Maula, M. and G. Murray (2000b). “Corporate Venture Capital and the Exercise of the Options to Acquire.” R&D Management (Forthcoming).
Maula, M. (2001). Corporate Venture Capital and the Value Added for Technology Based Firms. Unpublished Ph.D. Dissertation, Helsinki University of Technology.
Miner, A. S. and P. R. Haunschild (1995). Population level learning. Research in Organizational Behaviour. B. M. Staw and L. L. Cummings. Greenwich, CT., JAI Press: 117–166.
Mansfield, E. (1991). “Social Returns from R&D: Findings, Methods and Limitations” Research Technology Management, 34: 24–28.
Rind, K. (1981). “The Role of Venture Capital in Corporate Development,” Strategic Management Journal, 2: 169–180.
Siegel, R., Siegel, E. and I. MacMillan (1988). “Corporate Venture Capitalists: Autonomy, Obstacles and Performance,” Journal of Business Venturing, 3: 233–247.
Sykes, H. (1990). “Corporate Venture Capital: Strategies for Success,” Journal of Business Venturing, 5: 37–47.
Tversky, A. and D. Kahneman (1986). “Rational Choice and Framing of Decisions.” Journal of Business 59: 251–284.
Yelle, L. (1979). “The Learning Curve: Historical Review and Comprehensive Survey” Decision Sciences, 10: 302.