This thesis deals with international portfolio choice problems, and more specifically with the equity home bias - the empirical finding that people overinvest in domestic stocks relative to the theoretically optimal investment portfolio.
The first chapter describes the recent literature on international portfolio choice. We consider various home bias measures, and we illustrate the extent and the evolution of equity home bias both with recent portfolio holdings data and longer time series. We consider and discuss the various institutional-based and behavior-based explanations for the home bias puzzle, and conclude that none of the proposed theories can explain the extent of the bias by itself. Thus, international portfolio choice problems should
probably be explained by a mixture of rational and irrational behavior.
In Chapter 2 we generalize the Cooper and Kaplanis (1994) methodology for estimating the costs that could reconcile international portfolio holdings with CAPM predictions. First, we can simultaneously estimate inward and outward investment costs and even interactions between home and host country. Second, the risk aversion parameter is estimated rather than postulated. Third, we control for exchange rate risk, inflation hedging, fixed-interest investments, round-tripping and omitted countries. Our estimates of implicit investment costs for the developed countries are much lower than those reported in prior studies. Over the period 2001-2004, our estimates of the average inward shadow costs range from 0.01 (US) to 37 (Indonesia) percent per annum. Although the estimated implicit investment costs for mature economies are low and thus plausible, the methodology of Chapter 2 leads to extremely high
deadweight cost estimates for emerging markets; a matter that we will examine in Chapter 3. We find that the equity home bias is related to a mixture of market frictions, such as information asymmetries, institutional factors and explicit costs.
In the third chapter we reconsider the estimated deadweight costs for the emerging countries implied by the mean-variance portfolio model developed in Chapter 2. We show both theoretically and empirically that estimated implicit investment costs are mostly driven by estimated risk if home bias is strong, which is particularly the case for emerging markets. For OECD countries, risk is stable and relatively easy to estimate, but for emerging markets ex post risk may be very different from