Performance of Portfolios Optimized with Estimation Error

Andrew F. Siegel
University of Washington Business School

Abstract: We use stochastic Taylor series matrix methods to explain the poor out-of-sample performance of mean-variance optimized investment portfolios, developing theoretical bias adjustments for the effects of estimation risk by asymptotically expanding future returns of portfolios formed with estimated weights. We provide closed-form adjustments of classical estimates of portfolio mean and standard deviation. The adjustments significantly reduce bias in international equity portfolios, increase economic gains, and are robust to sample size and to non-normality. Dominant terms grow linearly with the number of assets and decline inversely with the number of past time periods in the data set. Under suitable conditions, optimized portfolios become more diversified. Using these approximation methods it may be possible to assess, before investing, the effect of statistical estimation error on portfolio performance.