This paper measures the mean, standard deviation, alpha, and beta of venture capital investments using a maximum likelihood estimate that corrects for selection bias. Firms go public when they have achieved a good return, leading to optimistic estimates without correction. The selection bias correction accounts for log returns, resulting in a mean log return of about 7% and a -2% intercept, with high volatility (standard deviation near 100%). Arithmetic average returns and intercepts are much higher than geometric averages. Without correction, arithmetic returns are around 700% and CAPM alpha is nearly 500%. With correction, arithmetic returns are about 53% and CAPM alpha is about 45%. Later rounds are less risky, with lower volatility and arithmetic returns. The maximum likelihood estimate matches data features, including IPO and exit patterns and stable return distributions across horizons. The paper analyzes the risk and return of venture capital investments, using the VentureOne database. It addresses selection bias by modeling the probability of going public or being acquired as value increases. The model captures return distribution characteristics, including the impact of time to IPO. Returns are highly skewed, with a few exceptional high returns and many modest ones. Lognormal distributions describe returns well, with a mean log return of 108% and standard deviation of 135%. Annualized returns are highly skewed, with a mean of 4×10⁹% and standard deviation of 2×10¹¹%, though the median is 62%. The paper also estimates betas, finding high values without correction but lower with log returns. Maximum likelihood estimates show a mean log return of 5.2% and standard deviation of 98%, with a leverage parameter of 5.4%. These results highlight the high volatility and risk of venture capital investments.This paper measures the mean, standard deviation, alpha, and beta of venture capital investments using a maximum likelihood estimate that corrects for selection bias. Firms go public when they have achieved a good return, leading to optimistic estimates without correction. The selection bias correction accounts for log returns, resulting in a mean log return of about 7% and a -2% intercept, with high volatility (standard deviation near 100%). Arithmetic average returns and intercepts are much higher than geometric averages. Without correction, arithmetic returns are around 700% and CAPM alpha is nearly 500%. With correction, arithmetic returns are about 53% and CAPM alpha is about 45%. Later rounds are less risky, with lower volatility and arithmetic returns. The maximum likelihood estimate matches data features, including IPO and exit patterns and stable return distributions across horizons. The paper analyzes the risk and return of venture capital investments, using the VentureOne database. It addresses selection bias by modeling the probability of going public or being acquired as value increases. The model captures return distribution characteristics, including the impact of time to IPO. Returns are highly skewed, with a few exceptional high returns and many modest ones. Lognormal distributions describe returns well, with a mean log return of 108% and standard deviation of 135%. Annualized returns are highly skewed, with a mean of 4×10⁹% and standard deviation of 2×10¹¹%, though the median is 62%. The paper also estimates betas, finding high values without correction but lower with log returns. Maximum likelihood estimates show a mean log return of 5.2% and standard deviation of 98%, with a leverage parameter of 5.4%. These results highlight the high volatility and risk of venture capital investments.