Mean–variance models for portfolio selection subject to experts’ estimations

Abstract

Since the security market is complex, sometimes the future security returns are available mainly based on experts judgements. This paper discusses a portfolio selection problem in which security returns are given subject to experts’ estimations. The use of uncertain measure is justified, and two new mean–variance and mean–semivariance models are proposed. In addition, a hybrid intelligent algorithm for solving the optimization models is given. To illustrate the application of the new models, the method to obtain the uncertainty distributions of the security returns based on experts’ evaluations is given, and two selection examples are provided.