MEAN-ABSOLUTE DEVIATION PORTFOLIO SELECTION MODEL WITH FUZZY RETURNS

Abstract

In this paper, we consider portfolio selection problem in which security returns are regarded as fuzzy variables rather than random variables. We first introduce a concept of absolute deviation for fuzzy variables and prove some useful properties, which imply that absolute deviation may be used to measure risk well. Then we propose two mean-absolute deviation models by defining risk as absolute deviation to search for optimal portfolios. Furthermore, we design a hybrid intelligent algorithm by integrating genetic algorithm and fuzzy simulation to solve the proposed models. Finally, we illustrate this approach with two numerical examples.