A More Human-Like Portfolio Optimization Approach: Using Utility Function to Find an Individualized Portfolio

Abstract

In this paper, a multi-objective model based on the decision maker’s (DM) utility function is proposed to find an optimized portfolio that fits with the desires of the DM. The proposed algorithm is developed in three stages. First, a fundamental analysis on accounting criteria is done using TOPSIS-DEA method. By this method, companies’ efficiency ranks according their fundamental reporting sheets are achieved. Second, the specific utility function of DM is found using the UTASTAR method. Third, a two-objective model is solved to find the stocks’ proportion for the individual investor. In this study, different criteria and decision making tools are used to make human-like decisions that meet investor’s expectations as well as possible. This approach is illustrated in this paper by a real-world case study concerning the evaluation of stocks in the Iran stock exchange. The suggested portfolio not only made a higher level of utility with the minimum level of risk but also is consistent with investor’s interests.