Behavioral portfolio selection: asymptotics and stability along a sequence of models (Q2788690)

The paper studies behavioral portfolio choice and its stability relatively to the market model. The behavioral nature of choice is captured by the objective functional which consists of the integral of a non-concave utility function when probability is distorted by a weighting function \(w\). The idea of a weighting scheme is inspired by prospect theory of \textit{D. Kahneman} and \textit{A. Tversky} [Econometrica 47, 263--291 (1979; Zbl 0411.90012)]. Maximization is achieved in the respect of an expected value constraint, as in a model with complete financial markets. The model uncertainty is modeled as a sequence of probability spaces and of risk-neutral measures which are assumed to converge to a limit probability space and risk-neutral probability, respectively. The author then proves that the value function along such sequence converges to the limit value function and that the sequence of optimal controls admits a subsequence which converges weakly to some control which is optimal for the limit problem.