Scenario generation for multi-objective stochastic portfolio selection

Abstract

Scenario generation allows Stochastic programming the expression of uncertainty. In fact scenarios allow the representation of stochastic elements by discrete distributions with many outcomes. In [15] authors proposed to use a nonlinear program to satisfy specified statistical properties. In this paper, we formulate a goal programming model for scenario generation for several random objectives. Scenarios generation help decision maker in solving stochastic goal programming model for portfolio selection problem. We applied our models on some Tunisian stock market security's data where we consider two random objectives that are the rate of return and liquidity.