A few months ago, we highlighted research showing how much chasing returns in equity mutual fund investing can reduce performance, noting that this behavior had cost investors around 2 percent per year over the period 2000-2012. New research over a longer period of time shows similar results.
In a recent Economic Synopses essay, YiLi Chien, a senior economist with the Federal Reserve Bank of St. Louis, examined investment strategies and the relationship between returns and equity mutual fund flows for the period 1984 through 2012. Chien found that current equity flows were positively correlated with stock returns from one and two quarters ago, with correlation coefficients of 0.38 and 0.31, respectively.1
In contrast, the correlation of current equity flows with future stock returns was negative for the next three quarters, though not sizably negative. The correlation of current mutual fund flows with stock returns of the following quarter was -0.06, of two quarters later was -0.07 and of three quarters later was -0.13. Chien wrote, “The positive correlation between current flows and past returns suggests equity mutual fund investors tend to buy when past returns are high and sell otherwise.”
Chien then compared this returns-chasing behavior to a buy-and-hold strategy, which assumes investors simply buy equity and hold it for an extended period of time. In this case, Chien looked at rolling seven-year windows. Over the full period studied, the return-chasing strategy resulted in relative losses each year except for a few quarters in the 1990s, and the differences in returns could be substantial. The buy-and-hold strategy outperformed the return-chasing strategy by up to 5 percent, meaning that the cumulative difference in returns over seven years could be as high as 40 percent.
Chien concluded, “Ultimately, this analysis shows that poor investment timing caused by return-chasing behavior has a significant impact on portfolio performance.”
1 Coefficients indicate a stronger positive correlation the closer they are to 1.0 and a stronger negative correlation the closer they are to -1.0.