Evidence and explanations of home bias in portfolio choice

Abstract

Most share portfolios are substantially overweight their owner's home sharemarket. Many papers in the literature present evidence of this phenomenon, known as home bias. Home bias is observed even in multi-billion dollar portfolios that have their asset allocation selected by investment professionals. This challenges simple behavioural explanations of home bias, such as that investors feel uncomfortable about foreign markets, put patriotism ahead of wealth maximisation, don't know how to access foreign markets, or don't understand diversification. This thesis argues that home bias occurs due to a rational response by investors to advantages they have investing at home. This thesis presents evidence of large and statistically significant increases in most pairwise correlations between the returns of the world's six largest stockmarkets, during the last 15 years. Increased correlations greatly reduce the risk-reduction benefit to investors of diversifying their investments internationally. Next, this thesis develops an equilibrium model of both portfolio choice and asset prices. It uses this model to generate investors' portfolios, as an optimised trade-off between each investor's advantage investing at home and the risk-reduction benefit of international diversification. The model of portfolio choice is calibrated using up-to-date estimates of international market correlations and home market advantage. Predicted home bias is found to be comparable in magnitude to home bias observed in practice.