Ambiguity Matters If You Invest in Many Assets

Abstract

This study examines the practical performance of the multiple priors optimal portfolio based on the mean-variance preference. The multiple priors optimal portfolio is designed to be robust to model uncertainty, also known as ambiguity. A back test finds two properties: the multiple priors optimal portfolio tends to be efficient when the number of assets is large and it has fewer turnovers than the mean-variance-efficient portfolios. Furthermore, the presented simulation shows that the multiple priors optimal portfolio outperforms the others when the number of assets is large and number of observations to form a portfolio is small.