| It is often asserted that venture capital (VC) funds specialize by industry, stage, or geographic region. Little research in finance, however, has empirically examined how specialized VC funds really are, and how they make their specialization choices. Using a principal-agent model, I analyze why VC funds display various degrees of specialization from a theoretical perspective. In addition, I test the predictions of my model using a sample of 1586 funds with 64168 venture investments. My study shows that there exists great heterogeneity in fund specialization. Fund size, proxies for VCs' risk aversion, and proxies for the risk associated with the excess return of the fund all have negative effects on specialization. I construct two measures distinguishing VCs' specialized talent from their general talent. I find that it is VCs' specialized talent that really matters in determining specialization. There is also evidence indicating a positive relationship between specialization and fund performance. |
