Sentiment and the CAPM

We consider whether sentiment affects the validity of CAPM. We hypothesize that pessimistic periods have low levels of noise trading because costly short-selling mutes trading on negative sentiment. On the other hand optimistic sentiment stimulates long positions that are easier to establish, and optimistic periods may also attract more naïve investors, stimulating greater noise trading. Thus, optimistic periods may suppress rational pricing while pessimistic ones may accentuate the phenomenon. Empirically, using a standard Fama and French (1992) approach, and controlling for other well-known cross-sectional return determinants, we find that beta is strongly and positively significant in pessimistic periods (but not in optimistic ones), with a tstatistic exceeding three, and a reasonable estimate of the market risk premium. The evidence survives several robustness checks and supports the notion that rational asset pricing holds when the populace of agents trading in the market is more likely to be predominantly rational. * Antoniou, [email protected], Xfi Centre for Finance and Investment, University of Exeter Business School, Streatham Court, Rennes Drive, Exeter, EX4 4ST, U.K., Tel: 44-(1392) 726256. Doukas, [email protected], Old Dominion University, Graduate School of Business, Constant Hall, Suite 2080, Norfolk, VA 23529-0222, USA, Tel: 1-(757) 683-5521. Subrahmanyam (corresponding author), [email protected], The Anderson School, University of California, Los Angeles, CA 90095-1481, Tel: 1-(310) 825-5355. Sentiment and the CAPM