An Ordinal Theory of Risk and Ambiguity Aversion

Abstract

This paper shows that the representation of preferences over uncertain prospects by an ordinal utility function on multi-variate distribution functions offers an easily tractable and yet more general and descriptive way of modeling decision-making under uncertainty. This representation first makes straightforward not only the definition of ambiguity and ambiguity aversion but also a clear delineation of risk and ambiguity in a setting of decision-making under uncertainty where both risk and ambiguity are present. The representation further enables an analytical framework for decision under uncertainty that assumes no event separability of any kind but is capable of replicating and extending important results derived in the EU framework with objective risks.