Mean-Variance Optimization of Factors and the Cross-Section of Stock Returns

Abstract

Currently, popular asset pricing factors utilize ad-hoc weighting procedures. By synthesizing mean-variance theory with multi-factor pricing, we generate an asset pricing factor that uses optimal portfolio weights that maximize the portfolio Sharpe ratio. Our factor works out-of-sample and can be utilized by a real-time investor. We show that the optimal factor is priced in the cross-section of stock returns even after controlling for nine other popular factors in the literature. The optimal factor is priced in international stock markets and can also explain the cross-section of bond returns.

Keywords: Mean-Variance Optimization, out-of-sample, optimal factor, asset-pricing

JEL Classification: G11,G12