Mutual funds are a primary investment vehicle for households in many countries. In the U.S., the $13.0 billion mutual fund industry attracts 44.4 percent of households, among which 68 percent hold more than half of their financial assets in mutual funds. As such, mutual fund investments not only affect the wealth of a large fraction of the population, but also affect the efficiencies of financial markets. Not surprisingly, there is a strong interest in understanding the behaviors of mutual fund investors, whether they are sophisticated, and whether they invest rationally.
Previous studies in this area have produced mixed results. In a recent paper, my colleague David P. Brown and I develop a new test of investor sophistication by considering cross-fund learning within fund families. We investigate whether investors rationally use the performance of all funds in a fund family to evaluate an individual fund, instead of evaluating each fund in isolation.
Most mutual funds belong to a family, which provides rich possibilities for cross-fund learning. One source of cross-fund learning is common skill or resources shared by all funds in the same family. For example, multiple funds in a family may share a common manager or management team. As a result, a fund’s performance reflects not only its fund-specific characteristics, but also the quality of the common skill.
Another source of cross-fund learning is correlation in the unobservable noise in the returns of funds in a family. When funds rely on shared sources of information, they are likely to tilt their portfolios in similar directions relative to their benchmarks. These funds are then subject to correlated shocks to their performance.
Given the considerations above, how should a rational investor evaluate the alpha-generating skill of a mutual fund in a family? More specifically, how does an estimate of skill depend on a fund’s own performance and the performance of other funds in the family? How do the sensitivities of the estimate to fund performance and family performance change with fund and family characteristics, including the number of funds in the family? And finally, do investors respond to fund performance and family performance in a manner that is consistent with optimal learning?
To answer these questions, we developed a continuous-time model, in which a fund’s performance is driven by a combination of fund-specific skill and family-wide common skill, and is subject to correlated idiosyncratic shocks that are correlated with the family. Neither the skill nor the shocks are directly observable. Investors estimate each fund’s skill by observing the returns of all funds in the family, and allocate wealth across funds, as in Berk and Green (2004).
Our model highlights two competing effects of family performance on the estimated skill and fund flows of a member fund: a positive common-skill effect and a negative correlated-noise effect. The overall spillover effect of family performance can be either positive or negative, depending on the relative strength of these two effects. It increases with the weight of family skill in the alpha-generating process and decreases with the correlation of noise in fund returns. Its absolute value increases with the number of funds in the family and declines over time. The sensitivity to a fund’s own performance is positive and declines over time. It increases with the correlation of noise in fund returns and decreases with the weight of family skill and the number of funds in the family.
We empirically tested these model predictions about mutual fund flows, using a comprehensive sample of actively managed domestic equity funds, and found strong support. On average, good family performance has a positive effect on fund flows to a member fund, suggesting the dominance of the common-skill effect. It has a stronger impact for funds in larger families and funds with a higher manager overlap rate and a lower idiosyncratic return correlation with other funds in their families. Interestingly, for the subsample of funds with a below-median manager overlap rate, an above-median idiosyncratic return correlation, and a below-median family size, the response of fund flows to family performance is significantly negative, suggesting the dominance of the correlated-noise effect. Furthermore, the sensitivity of flows to a fund’s own performance decreases with fund age, the manager overlap rate, and family size, but increases with the correlation of idiosyncratic returns within families, as our model predicts.
We consider four alternative explanations for our findings, including effects of a star or dog fund in a family, asset allocation by affiliated funds of mutual funds, cannibalization within families, and effects of style performance, but none of them explains the rich patterns of fund flow sensitivities we document.
Our findings suggest that mutual fund investors as a whole appear to be sophisticated enough to perform rational cross-fund learning within fund families, countering common perceptions of them as being naïve.
To read the paper, which is co-authored by Youchang Wu, assistant professor of finance, and David P. Brown, professor of finance, visit the Social Science Research Network website.