Maximizing the Out-of-Sample Sharpe Ratio

Analytical Approach to Sharpe Ratio Optimization Under Estimation Risk In “Maximizing the Out-of-Sample Sharpe Ratio,” the author studies how investors can optimize their out-of-sample Sharpe ratio when the mean vector and covariance matrix of asset returns are estimated from limited historical data. The author derives analytical expressions for the expectation and variance of the out-of-sample Sharpe ratio for portfolio rules located on the sample mean-variance efficient frontier. With these results at hand, the author identifies the optimal location on the efficient frontier in the presence of estimation risk. The author provides simulation and empirical evidence demonstrating the benefits of the proposed portfolio rules relative to several benchmarks. This paper extends the mathematical toolkit for portfolio selection under parameter uncertainty by addressing the challenge of optimizing the out-of-sample Sharpe ratio in a nonasymptotic setting.