This study proposes to research how venture capital firms (VCF's) "learn" from their portfolio companies (PFC's) using basic learning and behavioral theories. The authors establish four primary hypotheses:

1. The prior investment experience of the VCF is positively related to the amount of learning by the VCF.
2. The relationship between the level of knowledge overlap between the VCF and the PFC and the amount of learning by the VCF is curvilinear, i.e. up to a certain point increases in the knowledge overlap increases the amount of learning. Beyond that point, however, further increases in knowledge overlap leads to a decrease in learning.
3. The VCF's trust in the PFC is positively related to the amount of learning by the VCF.
4. The PCF's current performance is positively related to the amount of learning by the VCF.

Only one of the hypotheses (Hypothesis 4) was supported by the study. The authors offer the following explanations for lack of correlation with their predicted learning theory model.

Hypothesis 1. The results may indicate that the author's measure for organizational learning captured the extent to which the knowledge base of the VCF could be updated and extended rather than how this knowledge base could be used to acquire new knowledge. Therefore, the perception of additional learning may be conditioned to only certain types of changes in the knowledge base or experienced venture capitalists are subject to the believe that, due to their experience, they have little to learn.

Hypothesis 2. As with the first hypothesis, the authors believe that VCF's may perceive that less learning is required the more they share industry and stage experience with the PFC. This may be an artifact of "new" learning occurring only in new environments.

Hypothesis 3. The authors suggest that the negative correlation may be the result of extremely high levels of trust resulting in lower need for penetrating discussions and information exchange. When the investor has full confidence in the PFC's honesty and truthfulness, the investor is less likely to challenge or scrutinize the PFC's decisions and, therefore, less learning takes place.

Hypothesis 4. While behavioral theory suggests that poor performance may lead to more organizational learning; in the venture capital context, however, it was expected that performance would have a positive relationship with learning because of the chance to score a "homerun" would promote more intensive VCF effort. The authors also postulate that VCF's learn more from such ventures because the human capital in such ventures both drives their success and provides more valuable knowledge to be assimilated.

There were a number of aspects of this study that this reviewer found confusing and misdirected. First is the definition of "learning." It is unclear what the authors were using in terms of the "learning" focus. Were they attempting to measure "learning" in terms of the technology, industry, business model, business and/or marketing strategies, investment strategies, etc? The concept of learning could be significantly different for each of these different elements.

Second, the results and value of the study may have been substantially altered had a multi-variant analysis been performed on any number of potentially related variables, however, there was no apparent effort made to evaluate these interrelated factors.

Third, there are a number of factors related to the venture capital industry that could provide addition explanations for the results obtained in the study.

1. An understanding of the evolution of the venture capital industry over the past 25-30 years may have provided better direction in the study's hypotheses, format and objectives. Thirty years ago, the venture capital industry was composed, almost entirely, of individuals that had evolved from the security analyses and/or investment banking communities due to their "fund management" and industry analytical capabilities. Few, if any, of these individuals had a background in building and/or operating startup or early stage companies let alone evaluating new and untested technology platforms. As a result, their learning curve was very steep and, therefore, critical to their success. As a result, they were highly dependent on outside resources to provide sufficient knowledge to make investment decisions.
   
   Today's venture capital community, however, is largely composed of former successful venture capitalists, successful entrepreneurs, industry and technology sector experts, and retired industry executives that bring a significantly different skill and knowledge base to the position. The highly specialized focus of individuals and, in some cases firms, results in few investments being made outside the individual's/firm's field or area of expertise. The venture capitalists tend stay within their technical fields and industry areas of expertise and knowledge, thus minimizing the need for learning new knowledge. As a result, the learning curve has been significantly flattened or, possibly, even ignored due to a certain level of hubris within some sectors of the venture capital community. Consequently, the learning curve may be different depending on the variable it is tested against. It would be valuable to know the profile of the respondents and analyze variations based on background, experience, expertise etc.
   
   
2. While the venture capital industry is highly respected for its risk tolerance, industry data actually suggests the opposite. For example, as indicated by data published by the National Venture Capital Association, the investments by venture capitalists in seed stage companies (those requiring the greatest knowledge base and support as well as the highest risk factor) has declined from over 17 percent of total venture capital investments in 1995 to under 1.5 percent in 2003. During the same period, total investments in early stage companies (those companies one-step removed from seed stage companies) declined from 23.4 percent in 1995 to just over 17 percent in 2003. During the same period, however, investments in expansion stage companies went from 42.1 percent to nearly 52 percent of total venture capital investment. These trends would suggest that the preferred sector for venture capital investment would be in expansion stage companies that have effectively reduced (or eliminated) significant risks associated with technology, market, management and financing. While learning may be involved in expansion and/or later stage companies, it is far less critical than in seed and early stage company development. As a result, the primary investment focus of the industry is in the stage of development where "learning" requirements may be at a minimum. Consequently, knowing the stage of development of the PFC's would be helpful in understanding the variation in responses based on potential investment risk and knowledge requirements.
   
   
3. Another factor affecting the venture capitalists learning capacity may be the firm's investment cycle. Venture capital firms are, most commonly formed with a life cycle of 10 years. During that 10 year period, the firm must (1) review 100's if not 1,000's of business plans, (2) identify the top investment candidates, (3) perform due diligence on a select few of those opportunities identified, (4) prepare a term sheet, (5) negotiate the investment, (6) serve on the board of directors of possibly 12-15 companies, (7) repeat steps 1-5 until the firm's fund is invested over the first 4-5 years, (8) identify and initiate exit plans during years 6-10 of the life cycle, (9) assist in the negotiation of the exit strategy for IPO, merger or acquisition, (10) begin raising the next fund. As a result, there is little time and limited opportunity to expand one's learning capacity or knowledge base. The learning process may be more related to interactions, which occur during the normal course of their daily functions and include other venture capitalists, attorneys, accountants, technical experts and other external advisors.
   
   The relatively short life cycle of the venture capital firm may also help explain the positive correlation between VCF focus and PFC performance. By the end of year 5 or 6, portfolio companies that have failed to establish technical proof of concept, strategic relationships with potential business partners, stabilized their management teams, initiated market entry, or secured additional funding have, most likely, been identified and eliminated from consideration as potential portfolio successes. This elimination of slow-developing companies leaves more time to concentrate on high performing companies that offer greater potential for high returns. Further knowledge of the VCF's investment cycle as well as portfolio structure may have provided insight into the function of performance and VCF interaction.
   
   
4. There is also what this reviewer has heard members of the venture capital community refer to as the "lemming" factor. During the early stages of development, the venture capital industry was heavily dependent on "syndication" for completion of venture deals. Several VCF's would combine their resources to fund an investment opportunity in order to reduce their potential risk and expand their investment resources. The "lead" investor was expected to provide "due diligence" to other syndicate members. As a result, lead investors had the primary responsibility to investigate the investment potential and establish the terms of the investment. As the industry became more engorged with investment capital in the 90's, syndication appeared to disappear in favor of an "everybody for himself" approach. Too much capital chasing too few deals was the word on the street. The result seemed to be that many venture capital firms followed other firms into what appeared to be "hot" industry or technology sectors. The investment frenzy surrounding the early successes of the dot.com industry provides a good example of the industry following each other off the "cliff." As a result, the industry has long had a tendency to "trust" other venture capitalists more than any other sources available to them, sometimes to their detriment. Knowing if the responding VCF's participated in the PFC's investment as a lead or follow-on investor would provide insight into the VCF's need for knowledge and the relative importance of learning.
   
   
5. Thirty years ago it was feasible for a relatively "unknown" entrepreneur to gain an audience with a venture capitalist to make a presentation and raise capital for a startup. In today's venture capital industry, it is increasingly more difficult for an unknown entrepreneur to raise venture capital. Unless an entrepreneur knows a venture capitalist or someone else (e.g. attorney, accountant, entrepreneur) that knows a venture capitalist, it is extremely difficult to gain access into their sanctum. As a result, "trust" between the VCF and PFC is established very early in the process of raising capital through other network contacts. If at any time during the due diligence effort a venture capitalist finds reason to distrust someone seeking capital, the deal will be terminated. As a result, one might speculate that trust is probably a more important factor in making an investment decision than it is related to the learning process. Knowing the previous relationship of the PFC and the VCF would provide insight into the "trust" factor that could, feasibly, help differentiate whether the factor was important in the investment decision, learning or both.

In summary, there appears to be a number of explanations for the results that the authors have not considered. It is the view of this reviewer that a better understanding of the venture capital industry, its structure, life cycle, history, and interrelationships within the industry might be used to better define and understand the processes which impact and change the venture capital community and their investment decisions. While the learning process may be of interest at the academic level, it might be more beneficial to evaluate the venture capital investment decision process and determine how different variables affect investment choices.

Link to full study