Optimization of Investment Portfolio Mean-Variance Model Using Genetic Algorithm

Abstract

The optimization of investment portfolio is aimed at finding the optimal combination of each stock with the goal of maximizing returns while minimizing risk through diversification. However, the question is how much funds should be invested to achieve the minimum risk. One of the approaches that has proven effective in building an optimal investment portfolio is the Mean-Variance model. The aim of this research is to determine the weights of the optimal portfolio components with the minimum risk. The data used consists of stocks included in the LQ45 index for the period from February 2020 to July 2021. Based on the research results, there are five stocks that form the optimal portfolio, namely ADRO, AKRA, BBCA, CPIN, and EXCL. The allocated weights for each stock are ADRO 9.896%, AKRA 32.049%, BBCA 30.749%, CPIN 13.949%, and EXCL 13.357%. The optimal portfolio generated by the Genetic Algorithm method has a risk of 0.000472 and an expected return of 0.000492.

Keywords

• mean-variance, lagrange multiplier, genetic algorithm.