This paper by John H. Cochrane analyzes the risk and return of venture capital (VC) investments, using a maximum likelihood estimate to correct for selection bias. The selection bias correction accounts for log returns, which are heavily skewed towards IPOs and acquisitions. Without correction, the mean log return is about 100%, with a log CAPM intercept of 90%. With the correction, the mean log return is about 7%, with a -2% intercept. Returns are highly volatile, with a standard deviation near 100%, leading to much higher arithmetic average returns and intercepts. The paper also finds that later rounds of financing are less risky, with lower volatility and arithmetic average returns. The betas of successive rounds decline dramatically from near 1 for the first round to near zero for the fourth round. The maximum likelihood estimate matches many features of the data, such as the pattern of IPO and exit as a function of project age, and the stability of return distributions across horizons. The paper discusses the implications of these findings for understanding the risk and return of VC investments.This paper by John H. Cochrane analyzes the risk and return of venture capital (VC) investments, using a maximum likelihood estimate to correct for selection bias. The selection bias correction accounts for log returns, which are heavily skewed towards IPOs and acquisitions. Without correction, the mean log return is about 100%, with a log CAPM intercept of 90%. With the correction, the mean log return is about 7%, with a -2% intercept. Returns are highly volatile, with a standard deviation near 100%, leading to much higher arithmetic average returns and intercepts. The paper also finds that later rounds of financing are less risky, with lower volatility and arithmetic average returns. The betas of successive rounds decline dramatically from near 1 for the first round to near zero for the fourth round. The maximum likelihood estimate matches many features of the data, such as the pattern of IPO and exit as a function of project age, and the stability of return distributions across horizons. The paper discusses the implications of these findings for understanding the risk and return of VC investments.