Global Business and Economics Review, Vol. 14, Nos. 1/2, 2012 77 Copyright © 2012 Inderscience Enterprises Ltd. Bayesian portfolio selection under a multifactor asset return model with predictive model selection Tomohiro Ando Graduate School of Business Administration, Keio University, 4-1-1 Hiyoshi, Kohoku-ku, Yokohama-shi, Kanagawa, 223-8526, Japan E-mail: andoh@kbs.keio.ac.jp Abstract: This paper addresses the problem of portfolio selection under a multifactor asset return model, using Bayesian analysis to deal with uncertainties in parameter estimation and model specification. These sources of error are ignored in the classical mean-variance method. We apply two approaches: the empirical Bayes method, and Bayesian model averaging. The previous literature on Bayesian portfolio selection has paid little attention to the researcher’s choice of factors contributing to the asset return prediction. This paper uses a previously published criterion to quantify the predictive power of several candidate models and justify this choice. Using data from the US and Japanese stock markets, a comparative analysis is conducted between the two Bayesian methods and the classical mean-variance method. A major finding pertinent to investors is that the influence of each asset return factor varies with time, depending heavily on the state of the market. Both Bayesian methods perform better than the classical method, but the difference between them is not great. Keywords: Bayesian model averaging; empirical Bayes; Markov chain Monte Carlo; predictive Bayesian model selection. Reference to this paper should be made as follows: Ando, T. (2012) ‘Bayesian portfolio selection under a multifactor asset return model with predictive model selection’, Global Business and Economics Review, Vol. 14, Nos. 1/2, pp.77–101. Biographical notes: Tomohiro Ando is an Associate Professor of Graduate School of Business Administration, Keio University. In 2007–2008, he was a Visiting Scholar at Booth School of Business, the University of Chicago. As a Visiting Faculty, he has taught at University of California, Berkeley and University of California, Los Angeles. He is the author of the book, Bayesian Statistical Modelling and Model Selection, published by Chapman & Hall/CRC Press. His research papers can be found in Biometrika, Journal of Econometrics and a number of international journals. This paper is a revised and expanded version of a paper entitled ‘Bayesian portfolio selection under a multifactor asset return model with predictive Bayesian model selection’ presented at International Association for the Scientific Knowledge: Global Management 2010, Oviedo, Spain, 8–10 November 2010.