Regularized Portfolio Optimization Using Constrained Hierarchical Bayes Models

• Activity Number: 132
• Type: Contributed
• Date/Time: Monday, August 5, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Business and Economic Statistics Section
• Author(s): Jiangyong Yin*+ and Xinyi Xu
• Companies: Ohio State University and Ohio State University
• Keywords: Bayesian portfolio optimization ; Hierarchical Bayes model ; mean-variance framework ; shrinkage effect ; regularization

Abstract

Traditional plug-in estimator of mean-variance optimal portfolio weights delivers rather erratic and unsatisfactory out-of-sample performance due to estimation errors associated with the sample mean and sample covariance of excess asset returns. Regularized solutions, such as the minimal variance framework, "no-short-sale" constrained or norm-constrained optimization, and informative Hierarchical Bayes priors, can achieve superior out-of-sample performance due to shrinkage effect. By inducing priors directly on optimal portfolio weights and implementing no-short-sale constraints a priori in Hierarchical Bayes models, we can construct portfolios that are better diversified with significantly better out-of-sample performance than their frequentist counterparts. Our method is tested on a number of Fama-French industry portfolios and it consistently beats the Hierarchical Bayes model, no-short-sale constrained optimization, the naive diversification strategy as well as several existing Bayesian models that are motivated by economic theory.