The value of the right distribution in stochastic programming with application to a Newsvendor problem

In this paper we introduce the concepts of the Value of the Right Distribution (VRD), the Performance Bound (PB) and the Worst-Case Performance Bound (WPB), which allow us to quantify how much we lose if we guess the wrong distribution of the uncertain parameters affecting a stochastic optimization problem. In order to show how they apply, we introduce a cost-based variant of the classical Newsvendor problem and model it as a two-stage stochastic programming model. For this problem, we first provide optimal solutions in closed form for different probability distributions and then compute, both analytically and computationally, the VRD measure and the corresponding performance bounds PB and WPB. Finally, systematic numerical results are provided.