Print
E-MailThis post comes to us from Yael V. Hochberg of the Kellogg School of Management, Northwestern University, Alexander Ljungqvist of the Stern School of Business, New York University, ECGI, and CEPR, and Yang Lu of Barclays Capital.
In our paper Networking as a Barrier to Entry and the Competitive Supply of Venture Capital, which was recently accepted for publication in the Journal of Finance, we examine whether strong networks among incumbent venture capitalists in local markets help restrict entry by outside VCs, thus improving incumbents’ bargaining power over entrepreneurs.
Our results are consistent with the hypothesis that networking among VCs reduces entry. First, we find that there is less entry in VC markets in which incumbents are more tightly networked with each other, as evidenced by their past syndication patterns. The magnitude of the effect is large: Controlling for other likely determinants of entry, a one-standard deviation increase in the extent to which incumbents are networked (using measures borrowed from economic sociology) reduces the number of entrants in the median market by around a third.
The networking patterns we observe in the data may not be exogenous; rather, they may reflect omitted variables affecting both networking and entry. For example, unobserved variation in the cost of doing business in a given industry or location could induce networking (say, to economize on information costs) and independently reduce entry. To correct for this potential endogeneity problem, we follow two approaches. First, we use instrumental variables motivated by non-strategic and mechanical determinants of syndication decisions. This strengthens our results. Second, we exploit the three-way panel structure of our data (which span time, location, and industry) to identify omitted time-varying factors that are either location-specific or industry-specific. This produces results that are very similar to the IV estimates.
Our second test focuses on the determinants of an individual VC firm’s entry decision. Strong networks among the incumbents in the target market reduce the likelihood of entry. But not every potential entrant is deterred. Controlling for industry experience and geographic proximity to the market (which each double the likelihood of entry), we find that a VC firm is significantly more likely to enter if it has previously established ties to incumbents by inviting them into syndicates in its own home market. Moreover, it is with these very same incumbents that the entrant does business in the target market. In the context of the entry deterrence game sketched out above, this suggests that incumbents deviate from the strategy of non-cooperation with entrants when the gain from deviating – reciprocal access to the entrant’s home market – is sufficiently tempting. The cost of deviation is punishment, in the form of reduced syndication opportunities with fellow incumbents. We show that after doing business with a potential entrant, an incumbent’s probability of being invited into fellow incumbents’ syndicates decreases considerably and significantly, for up to five years after the event. This effect is concentrated in markets with a small number of incumbents, consistent with the notion that a small number of players can more easily prevent free-riding when called upon to execute a punishment strategy.
Finally, we examine the price effect of reduced entry by comparing the valuations of companies receiving VC funding in relatively more protected and relatively more open markets. Controlling as best we can for other value drivers, we find significantly lower valuations in more densely networked markets: A one-standard deviation increase in our networking measures is associated with a 10% decrease in valuation, from the mean of $25.6 million. This suggests that incumbent VCs benefit from reduced entry by paying lower prices for their deals. On the other hand, the more market share entrants capture, the higher are valuations in the following year, suggesting that entry is pro-competitive and, at least in that sense, benefits entrepreneurs. An unanswered question is whether networks provide offsetting benefits to entrepreneurs. We leave an examination of the overall welfare effects of networking to future research.
The full paper is available for download here.
