Optimal saving rules for loss-averse agents under uncertainty

Most empirical studies assume only monotonic preferences for households. Behavioral research, however, provides substantial evidence that preferences for wealth are measured relative to a reference point. In this paper we introduce and solve a two-period consumption and savings model for a loss-averse agent who measures utility from consumption relative to a benchmark level. The solution is given as a parametric decision rule with one unknown parameter that depends on the distribution of the return on saving. We find non-linearity in the fraction of wealth saved, where the specific saving pattern depends on the sign of the real return on savings. The amount of saving is nondecreasing in initial wealth and the riskiness of the return distribution. © 2002 Elsevier Science B.V. All rights reserved.