Abstract
We examine a large number of potential home bias determinants, including some novel ones, using extensive panel data. We distinguish between the actual home bias (over investment in domestic securities) and foreign investment bias, for which we propose a new measure. For foreign investment bias, we also demonstrate how "size biases" significantly affect the results. We find that the old empirical results based on the U.S. data alone do not generalize to the panel data set; information and familiarity variables and proxies for the degree of capital market openness play an important role in explaining both home and foreign investment biases.