Risk Aversion in Cumulative Prospect Theory

Abstract

This paper characterizes the conditions for strong risk aversion and second-order stochastic dominance for cumulative prospect theory. Strong risk aversion implies a convex weighting function for gains and a concave one for losses. It does not necessarily imply a concave utility function. The latter does follow if the weighting functions are continuous. By investigating the exact relationship between loss aversion and strong risk aversion, a natural index for the degree of loss aversion is derived.