Why Does Home Bias Exist: An Experimental Analysis of the Role of Familiarity, Information and Social Interaction Effects

Standard international financial portfolio theory predicts that investors should hold a sizeable portion of their portfolio in international assets. Actual portfolios of individual and institutional investors exhibit strong home bias i.e. heavy overweighing of local stocks relative to foreign stocks.Theoretical reasons for home bias include the potential for more efficient hedging against home-country specific risks, formal barriers to international diversification and informal barriers relating to informational asymmetries and behavioral biases. While arguments on hedging behavior and formal barriers have been empirically revealed to be unlikely candidates in explaining the full extent of the home bias phenomenon, empirical research has demonstrated indirect support for information costs and asymmetries or behavioral biases as plausible causes. However, standard empirical tests have an inherent problem that the measurable variables (geographical proximity, correlated buying, etc.) used as a proxy for the above concepts have a confounding effect on the results. The researchers therefore conduct tests in a controlled laboratory setting in order to investigate the underlying causes of home bias and also to disentangle the effects of familiarity, information and social interaction.In the controlled experiment, the subjects are assigned the role of asset market traders. The trading set-up is a standard limit-order process. By introducing various treatments that reflect the characteristics of the competing theories, the researchers can infer how each of these concepts affect the holding and trading patterns of their subjects. The researchers use two control groups: one where traders are initially assigned holdings of randomly selected shares, and a second where sub-groups of traders are assigned specific initial portfolios of stocks. The effect of three specific treatments are then measured against these controls. In the first treatment, the researchers create a degree offamiliarityto the traders by restricting trades during the initial stages of trading to specific stocks per sub-group. The second treatment allows the subgroups to collect refinedinformationabout the stock’s return distribution. In the final treatment, the researchers create aneighborhoodor “local social interaction” effect by allowing the participants within each sub-group to access each others holdings and trading patterns.