Time-Varying Risk Attitude and Behavioral Asset Pricing

Abstract

We propose a new behavioral asset pricing model that can flexibly express the time-variability or state-dependence of aggregate risk attitude. We suppose that individual preferences are heterogeneous, and each preference changes depending on the state of the economy. While most previous studies, including Chan and Kogan (2002) which is the basis of our model, imply that aggregate risk aversion is counter-cyclical by assuming either heterogeneity or changeability of preferences, considering both of these can result in aggregate risk aversion being procyclical in a particular economic situation. The status quo hypothesis we propose suggests that the aggregate risk attitude is procyclical during recessions and counter-cyclical during booms and depressions. This stems from the fact that in our setting only risk-tolerant agents face losses large enough to be loss-averse in order to maintain the status quo. We analyze the weekly returns of 10 major stock markets (Western markets including Canada, France, Eurozone, Germany, the UK, and the US, and Asian markets including China, Hong Kong, India, and Japan) and provide some evidences to support our status quo hypothesis in Western stock markets, where stock markets are relatively mature. We conclude that it is essential at least in such matured stock markets to allow both heterogeneity and changeability of preferences in order to obtain an accurate aggregate risk attitude.