Comparison of Optimal Portfolio Performance Based on Value at Risk and Upside Risk with Conventional Models

Abstract

The present study compares the performance of the optimal portfolio, based on Value at Risk and Upside Risk, with conventional models (the optimal portfolio based on the Markowitz model). This study aimed to select an optimal portfolio based on the proposed model in three scenarios: 1- potential and risk-averse 2- neutral potential and risk-averse 3- aversive potential and risk aversion investor, and compare them to the Markowitz model. So, we calculated the monthly returns of the 50 most active Tehran Stock Exchange companies over seven years of 91-97. These Four years are used to build the model and determine the efficient frontier. The value at risk is used as an undesirable risk measurement criterion, and covariance upper partial moment-CUPM is used as a desirable potential measurement. Also, this procedure was repeated for 36 months. To calculate the model's parameters Eviews11 software was used, and MATLAB software was used to solve the Markowitz quadratic programming. Likewise, GAMS software was used to solve the proposed nonlinear programming problem. And finally, SPSS software was used to test the hypotheses and compare the results of the two models with the longitudinal studies method with repeated measurements. The results show that updating the model, efficient frontier and applying the criteria of value at risk, and paying attention to investor orientations in terms of tendency to optimal potentials and risk aversion lead to optimized portfolio performance.

Keywords

• value at risk
• upside risk
• efficient frontier update
• portfolio performance