The main purpose of this study is to examine whether macroeconomic variables could subsume the size and book‐to‐market (BM) anomalies for longer‐return intervals using Tokyo Stock Exchange‐listed stocks.
The Fama‐MacBeth cross‐sectional regressions of various models over time‐intervals ranging from one month to one year are performed.
The empirical results show that most macroeconomic variables explain short‐term returns within six months, with the industrial production as the only variable that persistently explains returns of all horizons ranging from one month to one year. Firm size does bear significant risk premium, but its significance diminishes for return‐intervals beyond three months when macroeconomic variables are included in the regression. BM is the only variable that significantly accounts for the cross‐section of stock returns for all horizons, regardless of the inclusion of macroeconomic variables.
These empirical findings suggest that stock returns are determined by both rational factors such as macroeconomic variables and behavioral factors such as BM.
The findings suggest that potential trading strategies indeed can be formed to exploit the persistent predictability, especially the BM regularity.
This paper is the first study that examines the competing explanatory power of various asset‐pricing models over different investment horizons.
