Time-Variation in Expected Returns

Abstract

This article characterizes the stochastic behavior of expected retu rns on common stocks. The authors assume market efficiency and postulate an autoregressive process for conditional expected returns. They use weekly returns of ten size-based portfolios over the 1962-8 5 period and find that (1) the variation through time in expected returns is well characterized by a stationary first-order autoregression process; (2) the extracted expected returns explain a substantial proportion (up to 26 percent) of the variance in realized returns and the magnitude of this proportion has a monotonic (inverse) relation with size; (3) the degree of variation in expected returns also changes systematically over time; and (4) the forecasts subsume the information in other potential predictor variables. Copyright 1988 by the University of Chicago.