An economic and econometric analysis of market sorting with an application to venture capital.

摘要

Venture capitalists (VCs) invest in entrepreneurial companies, develop them into mature businesses, and sell them, typically through an initial public offering (IPO). Companies funded by experienced VCs are found to have higher IPO rates than companies funded by inexperienced in vestors. There are two principal reasons for this: More experienced investors match with better companies (sorting) and bring these companies public at higher rates (influence), and the objective here is to quantify the relative importance of these two effects. The fundamental empirical problem is a selection problem. The investment outcomes are only observed for the matched companies and investors. This matching is endogenous, and the observed outcomes are a systematically selected sample.; The matching of investors and companies is captured by a two-sided matching model. This model is a special case of the College-Admissions Model with a new restriction that ensures a unique equilibrium. It forms the basis of an econometric discrete choice model with two-sided matching. A class of selection models that allows for more general specifications of the selection process is introduced, and sufficient conditions for their identification are derived. The class contains the selection model above, in which the selection is based on the matching model. Matching introduces computational complications, but Bayesian estimation, using Markov Chain Monte Carlo, is a numerically feasible estimation method. Estimates of the model show that both sorting and influence are significant and economically important, with sorting being somewhat more pronounced than influence in this market.