摘要：This research uses macro factors to explain four standard U.S. stock market risk premia, i.e. the market excess return (RM-RF), size (SMB), value (HML), and momentum (WML). We find in-sample predictive power of macro factors, in particular at a one-year horizon. Differentiating between bull and bear market states roughly doubles forecast performance compared to neglecting market states. All four stock market risk premia can be explained with R-squares of 10% to 25%. However, macro factors have limited predictive power in a true out-of-sample setting
关键词：stock market; risk premia; factor analysis; market states
