Essays on Uncertainty: An Axiomatization and Economic Applications

Abstract

This thesis deals with individual decision making under uncertainty and extends standard economic and financial models with risk in order to model attitudes towards ambiguity. Empirical violations of the leading theories of choice have pointed out the relevance of ambiguity on individual choices. Much empirical evidence, inspired by Ellsberg's experiment, shows that people prefer bets whose odds of winning are known, thus suggesting aversion to ambiguity. The relevance of uncertainty is pervasive in economic literature as well as in the real world. Investment decisions and asset pricing, insurance contracts, voting in elections, gambling, buying a car or planning a trip can all be thought as choices under both risk and ambiguity. The aim of this work is normative and descriptive. In the first part of the work, I concentrate on a theoretical model of focused regret as an extension of the classical paradigm of choice theory. From the alternatives in this literature, I focus on the multiple prior model, originally axiomatized by Gilboa and Schmeidler, where ambiguity is formalized as a set of plausible probability distributions to represent agents' beliefs. Then I turn to economic and financial applications and challenge the descriptive power of classical economic models when explaining, for example, asset pricing or insurance contracting. Lastly, the most innovative part of this work is an experimental investigation of the multiple priors model in a financial market, through which I find evidence for both risk and ambiguity to affect individual decision making and asset pricing.