UCL School of Management is delighted to welcome Scott Stern, MIT/NBER, to host a research seminar discussing ‘Venture capital selection and start-up growth’.
While a long line of research highlights the central role of venture capital in facilitating the acceleration and growth of start-up firms, previous studies have not been able to systematically examine the selection of firms by venture capitalists. This paper leverages comprehensive business registration records for more than 80% of the US economy with a predictive analytics approach to the estimation of entrepreneurial quality to evaluate both the selection into and (estimated bounds on) the impact of venture capital on start-up performance. We establish several key findings. First, venture capital selection is drawn in a highly skewed way from the “right tail” of the entrepreneurial quality distribution – more than 70% of all venture capital investments are estimated to be drawn from the top 5% of the underlying observable quality distribution. By and large, the type of early-stage start-up observables (such as Delaware registration or early patents) that are associated with predicting equity firm performance (IPO or acquisition) are similar (and even have similar predictive value) to those associated with VC funding; interestingly, trademark protection is a predictor of growth but less so of venture capital selection. Observable entrepreneurial quality is more closely connected to venture capital selection outside of high-tech hubs and when capital is more scarce. These estimates allow us to compare the performance outcomes of the firms selected by VCs (versus not) for a given level of entrepreneurial quality. While venture capital backed firms are more than 180 times more likely to achieve an equity growth outcome (in the absence of our controls), matching on entrepreneurial quality reduces the performance advantage of the venture-backed group to less than 5 times, consistent with the matched sample results of Puri and Zarutskie (JF, 2012), and lower performance premiums are observed for firms with higher initial estimated levels of entrepreneurial quality. Finally, we document that, though venture capital is rare, those firms that achieve equity growth but do not receive venture capital share start-up characteristics similar to those that do receive venture capital.