Forecasting Stock Returns During Good and Bad Times

Dashan Huang, Singapore Management University
Fuwei JIANG
Jun TU, Singapore Management University
Guofu Zhou, Washington University in St.Louis

Abstract

See https://ink.library.smu.edu.sg/lkcsb_research/5156/ for the full text. This paper proposes a two-state predictive regression model and shows that stock market 12-month return (TMR), the time-series momentum predictor of Moskowitz, Ooi, and Pedersen (2012), forecasts the aggregate stock market negatively in good times and positively in bad times. The out-of-sample R-squares are 0.96% and 1.72% in good and bad times, or 1.28% and 1.41% in NBER economic expansions and recessions, respectively. The TMR predictability pattern holds in the cross-section of U.S. stocks and the international markets. Our study shows that the absence of return predictability in good times, an important finding of recent studies, is largely driven by the use of the popular one-state predictive regression model.