| Abstract: | This paper set out to explore if the observed structure of VC syndication is conducive to pseudo-diversification in the portfolios of the VCs' sources of funds; the institutional fund managers. One hypothesis is tested that states that the observed investment network is significantly riskier than we would expect in a random allocation of risk.Data are employed on the syndication of 20 UK-based VCs, in the 11 years from 1993–2003, along with their interactions within that period with 73 different institutional funds. The hypothesis is tried by means of a z-test developed from the comparison of observed risk scores and the mean of a population of risk-proxies. This population is derived with randomised models.The findings are consistent with the observed structure of VC syndication and institutional fund-VC interactions being conducive to less diversification than expected, particularly where institutional funds invest in just two VCs. This result extends prior research. Chiefly, the study explores a gap in portfolio theory; with mutual dependency between financial agents, risk can be pseudo-diversified, affecting all parties. The data examined in this study relates VC syndication to ineffectiveness in reducing portfolio risk, in contrast with its established association to risk mitigation. Also, it is suggestive of social structure within the VC community that is replicated in interactions with both ventures and institutional fund managers. A full understanding of VC syndication and investment risk should extend beyond its prior focus on VC firms and investee ventures, and also encompass the source of funds.This paper uses case studies and a statistical model to explore the limitations of portfolio theory in the presence of investment networks with interacting agents. The research highlights that such investment networks create a kind of systemic risk that is neither particularly rare nor without disastrous precedent. |
