Despite the well known benefits from international portfolio diversification international investors still exhibit a great deal of home bias in their equity allocations. Already in the late sixties and early seventies several authors show the large international diversification benefits (See e.g. Grubel, 1968; Solnik, 1974). Since the early seventies financial markets have liberalized, thereby reducing barriers of cross border asset holdings. However, the equity home bias anomaly remains present to date and economists have not been able to fully explain its causes. Not surprisingly, Obstfeld and Rogoff (2001) have singled out the equity home bias as one of the six major puzzles in macroeconomics.
Most research on the home bias focuses on explaining the frictions in international markets which hinder the investor to optimally diversify his portfolio. These frictions include, among others, transaction costs (Heathcote and Perri, 2004), capital controls (Chan et al., 2005), equity market development (Berkel, 2007) and informational asymmetries (Van Nieuwerburgh and Veldkamp, 2009). These papers seek to find the determinants of the lack of international equity diversification.
This paper focuses on the factors explaining cross border asset holdings, thereby investigating the properties of only the foreign invested part of resident’s equity portfolios. Early studies focus solely on the foreign equity position of the United States (e.g Ahearne et al., 2004). However, the IMF’s construction of the Coordinated Portfolio Investment Survey has enabled researchers to use detailed bilateral data on aggregate bilateral asset holdings for a large number of countries. Recent studies find that geographical proximity, trade and currency unions appear to be the main determinants (See e.g. Aviat and Coeurdacier (2007); Lane and Milesi Ferretti (2008)). Hence, many factors used to explained trade patterns are useful in describing bilateral asset holdings as well. In addition, capital controls (Bekaert and Wang, 2009) and information asymmetries (Andrade and Chhaochharia, forthcoming) appear to remain important determinants as well.
The International Capital Asset Pricing Model (ICAPM) predicts that each investor holds a share of foreign equities equal to their respective share of world market capitalization. The existence of home bias in portfolio equity holdings implies that the investor overweights domestic equities, thereby creating hedging demands for equities that comove less with the domestic equity market. Put differently, given the existence of the home bias, investors need to overweight their investment in low comoving assets compared to the ICAPM market portfolio. Hence, the investor aims to compensate for his overinvestment in domestic equities by bringing his portfolio closer to the market portfolio.
Existing studies find mixed evidence of the investor’s behavior in response to stock market comovement. Portes and Rey (2005) establish a weak negative reaction of investors to covariance, once informational frictions are controlled for. The evidence is weak and changes across specifications. A related study by Lane and Milesi-Ferretti (2008) finds no relationship between portfolio equity allocations and correlations. This result questions the rationality of investors, because investors do not reap the benefits of global diversification if they do not overinvest in low comoving equities.
However, Coeurdacier and Guibaud (forthcoming) point at an endogeneity problem when correlation is used as an explanatory variable. There is a clear simultaneity bias, because equity holdings are determined by correlations, but correlations are also dependent on investors’ equity demands. These authors find that once frictions and endogeneity concerns are controlled for, international investors behave rationally by overinvesting in low comoving assets and overinvesting in high comoving assets.
Download
Dynamic International Portfolio Adjustment: Rational Investors and the Home Bias
