Hao Bai

University of Manchester · Economics · 2026–27 job market

Fields

Theory, Behavioral & Experimental, Finance

Job market paper

A Separation of Uncertainty Attitudes and Belief Updating

This paper proposes a natural and tractable updating rule for Choquet expected utility with probabilistic sophistication, grounded in Bayesian updating. A probabilistically sophisticated decision maker expresses beliefs by a subjective probability measure, which serves as a Bayesian prior and is updated to a posterior according to Bayes' rule when new information becomes available. Under certain conditions, a non-additive capacity can be represented as a transformation of an additive probability measure. Leveraging this representation, beliefs are updated by applying Bayes' rule to the underlying probability measure, with the updated capacity derived by applying the same transformation to posterior probabilities. We provide an axiomatisation of this updating procedure and develop a simple and intuitive updating rule that neatly separates beliefs from attitudes towards uncertainty, with beliefs evolving in a Bayesian manner and uncertainty attitudes remaining unchanged.