Return Horizon and Mutual Fund Performance

Abstract

Measures of investment performance, including alpha, the canonical measure of mean return after allowing for systematic (beta) risk, depend on return measurement horizon, because beta depends on horizon. We introduce a procedure that relies on a combination of theory and simulations to estimate long-return-horizon alphas and betas from short-horizon estimates. Compared to a benchmark of 41% estimated in monthly returns, the percentage of mutual funds with positive alpha estimates against the SPY decreases to 21% (increases to 46%) at the decade return horizon for funds with high (low) estimated monthly market betas. Alphas estimated from short-horizon (e.g. monthly) returns may be uninformative or even misleading regarding fund performance for investors with longer horizons.

JEL Classification: G10; G23