Abstract
The purpose of this paper is to investigate the risk-return tradeoff allowing for the presence of noise traders, i.e., a subset of investors who either base their trading strategies on sentiment or hold unjustified optimistic/pessimistic views regarding market prospects. We measure noise traders' sentiment relying on two sets of indices, namely the Baker and Wurgler sentiment index and the Michigan Consumer Confidence Index, in the US stock market. Under the assumption of the presence of noise traders' sentiment, the risk-return tradeoff is tested through two sets of models: Merton's Intertemporal CAPM and the GARCH-in-mean model. First, we find that the relationship between risk and return allowing for the presence of noise trader risk as measured by the Baker and Wurgler sentiment index is positive and statistically significant when tested through Merton's Intertemporal CAPM. Second, the risk-return tradeoff tested through GARCH-in-mean models augmented by noise traders' risk as measured through survey-based measures of sentiment establishes no clear evidence for a significant mean-variance relationship. Overall, we confirm Merton's (1973) hypothesis that the more risk an investor bears, the greater his expected returns. This paper contributes to the asset pricing literature by trying to shed some light on the risk-return tradeoff from the standpoint of behavioral finance.