Robust Portfolio Optimisation with Multiple Experts

Abstract:     
We consider mean-variance portfolio choice of a robust investor. The investor receives advice from J experts, each with a different prior for the distribution of returns. Confronted with these multiple priors the investor follows a min-max portfolio strategy. We study the structure of the robust mean-variance portfolio and empirically compare its performance with a variety of alternative portfolio strategies. The empirical tests are based on bootstrap simulations on the 25 Fama-French portfolios and on 81 European country and value portfolios. We find that the robust portfolio performs well in both settings. Robust portfolios do not exhibit the extreme weights typically observed in naive mean-variance portfolios. Robust portfolios are also better diversified than portfolios that impose short-sell constraints to suppress the symptoms of extreme weights.

Keywords: Mean-variance, model uncertainty, portfolio choice

JEL Classifications: C11, D80

Working Paper Series