Abstract: We develop a methodology that characterizes the set of CARA utility functions that may rationalize a decision over discrete and objective uncertainty. We show that if one assumes that a revealed preference (decision) remains invariant over wealth, then there cannot exist any EU representation of a set of decisions unless it permits a CARA representation. We discuss how this methodology can be used to test the existence of probability-weighting risk preferences such as Cumulative Prospect Theory. We also empirically analyse how the implied curvature of choices between lotteries drastically changes between multiple choice types and what implications this has for representing preferences over uncertainty through utility curvature.
Sir Clive Granger BuildingUniversity of NottinghamUniversity Park Nottingham, NG7 2RD
Contact us