Extrapolation and Risk-Return Trade-offs

Abstract

This study investigates the impact of return extrapolation on risk-return trade-offs in both the aggregate time series and the cross section of stocks. We find that the relation between the market's expected return and expected variance is positive during periods with a low degree of extrapolative weighting (DOX), as proposed by Cassella and Gulen (2018), but this relation is negative during high-DOX periods. In the cross section, we find that low-risk anomalies are significantly more pronounced following high-DOX periods and among firms with high firm-level DOX. These findings are consistent with the interpretation that, relative to a fully rational framework, the existence of extrapolators tends to raise equilibrium stock prices in response to a positive risk shock. The relative increase in stock prices is meant to sustain high perceived future returns to compensate for this enhanced risk for extrapolators and thus leads to overpricing and a weakening of the traditional risk-return trade-off.

Keywords: Behavioral finance, Extrapolation, Risk-return trade-off, Sentiment

JEL Classification: G11, G12, G40