Institutional portfolios exhibit a home bias that causes them to overweight stocks from the investor’s home market and underweight foreign stocks. What is the cause of this home bias? Is it a result of a behavioral bias, or is it an optimal response to possessing superior information about the home market? The authors offer evidence in support of the latter explanation.

The authors find that portfolio performance is higher when there is a greater degree of home bias, foreign concentration, and industry concentration. In addition, these returns are amplified when the investor’s capacity to learn—which the authors measure as higher “skill” at achieving excess or abnormal returns—is higher, as in the case of hedge funds. Specifically, institutions with higher skill are found to construct more concentrated portfolios at both the industry and the country levels, but they show more foreign bias than home bias. The authors also find that institutional home bias is more prevalent in the case of higher home market uncertainty, which they measure in terms of variance and a few other proxies.

The authors build on the information advantage theory proposed by S. Van Nieuwerburgh and L. Veldkamp (Journal of Finance 2009), which argues that it is optimal for rational investors to choose to concentrate their portfolios in countries or in industries within a country where they have an initial information advantage relative to the average investor. This lack of diversification allows these investors to achieve higher risk-adjusted returns.

The information advantage theory also suggests that the greater the investor’s capacity to learn, the higher the portfolio’s concentration and expected performance. In addition, the information advantage theory argues that a higher level of uncertainty in the investor’s home market tends to cause investors to exhibit more home bias because the benefits from an information advantage are greater.

Portfolio concentration is measured in terms of three factors: (1) home bias, which measures whether the institutional investor’s portfolio is overweight or underweight with respect to the investor’s home country compared with the world market portfolio; (2) foreign concentration, which indicates whether the foreign share of the investor’s portfolio is well diversified among foreign countries or is concentrated in one or a few countries; and (3) industry concentration, which indicates whether the assets are concentrated in a few industries or are well diversified among all industries, as in the hypothetical world market portfolio.

The authors use quarterly filings of more than 10,000 institutions representing 72 countries from the FactSet database covering the last quarter of 1999 to the first quarter of 2010. They use widely accepted measures of portfolio return and risk variables to calculate institutions’ risk-adjusted performance. The authors then regress institutional performance on the various proxies for portfolio concentration. The regression results statistically and economically support the hypothesis that (1) portfolio performance is positively correlated with portfolio concentration and (2) this positive correlation is not specific to the United States and holds for institutional investors worldwide. They find a positive correlation between “home market uncertainty” and home bias and a strong positive correlation between “skill” and foreign bias. Where both skill and home market uncertainty are present, there is a significant degree of home bias, lending support to the authors’ suggestion that uncertain markets allow skilled investors to make more profitable use of their information advantage.

The authors’ results support a theory that is more reflective of institutional investors’ investment behavior than the traditional asset pricing theory, which focuses mainly on diversifying systemic risks and assumes that all investors have access to the same information needed to construct an optimum portfolio. The real world has information asymmetry, with larger institutions in a superior position to obtain better information, which they use to earn higher risk-adjusted returns. The only limitation in this research is the use of “skill”—measured as the decile of excess or abnormal returns—as a proxy for the capacity to learn. The authors regress the skill variable against portfolio concentration variables to arrive at conclusions regarding how the capacity to learn might affect portfolio holdings. But because skill itself is measured as portfolio outperformance, the results are preemptive and redundant. This research will foster more research, and it will be interesting to see how asset class concentration affects portfolio performance.