Abstract

The alpha of an investment reflects its ability to increase the Sharpe ratio of a benchmark portfolio allocation based on tradable factors. We argue that, in the context of private capital, the usual approach to estimate alpha is misleading because it ignores the economic realities of investing in private markets. We then combine a large sample of 5,028 U.S. buyout, venture capital, and real estate funds from 1987 to 2022 to estimate the alphas of private capital asset classes under realistic simulations that account for the illiquidity and underdiversification in private markets as well as the portfolio allocation of typical limited partners. We find that buyout as an asset class provided a positive and statistically significant alpha during our sample period. In contrast, over our sample period, the venture capital alpha was large and positive but statistically unreliable whereas the real estate alpha was very close to zero.

Authors

Gregory Brown, UNC Kenan-Flagler Business School
Andrei S. Goncalves, The Ohio State University
Wendy Hu, MSCI