Nonconcave Robust Utility Maximization under Projective Determinacy

Abstract: We study a robust utility maximization problem in a general discrete-time frictionless market. The investor is assumed to have a random, nonconcave and nondecreasing utility function, which may or may not be finite on the whole real-line. She also faces model ambiguity on her beliefs about the market, which is modeled through a set of priors. We prove, using only primal methods, the existence of an optimal investment strategy when the utility function is also upper-semicontinuous. For that, we introduce the new notion of projectively measurable functions. We show basic properties of these functions as stability under sums, differences, products, suprema, infima and compositions but also assuming the set-theoretical axiom of Projective Determinacy (PD) stability under integration and existence of $\epsilon$-optimal selectors. We consider projectively measurable random utility function and price process and assume that the graphs of the sets of local priors are projective sets. Our other assumptions are stated on a prior-by-prior basis and correspond to generally accepted assumptions in the literature on markets without ambiguity.