Can reference-dependent loss aversion explain choice behaviour?

Abstract

This paper examines the extent to which reference-dependent loss aversion, on its own, can explain choice behaviour under risk. To this end, a model of preferences over monetary lotteries and a rule of endogenous reference points are developed. The model is characterised by a new property of reference-dependent loss aversion. It is reduced to the expected value when the reference point is invariant across choices and satisfies first-order stochastic dominance. Our findings highlight the crucial role of loss aversion in explaining risk attitudes. In particular, the results show that the change in the reference point could be responsible for the observed well-known choice patterns, such as the common consequence effect and the common ratio effect, as well as for their reverse effects. The model predicts this behaviour by making use of only one function that weights losses with a value that increases with the reference point. Neither a probability weighting function nor a specific functional form for the utility of outcomes is required.