Metode Particle SWARM Optimization untuk Optimasi Portofolio

Abstract

Stock is one of financial assets popularly invested in the present time. In spite of its unstable movement, stock price can be predicted for certain period of time to estimate the profit that will be made. The owned stocks can be systematically arranged to produce a maximum profit with minimal risks through portfolio. Portfolio can be arranged by imposing various constraints, such as sector capitalization. With this constraint, portfolio does not create a proportion for each stock, but for every sector containing various stocks. If implemented, the investor may be interested in investing their capital into the stocks that have the high value in the capital market. By investing in various stocks, investors can minimize the risks provided one of the chosen sectors suffers loss. Optimization with this constraint is a non-linear optimization problem with integer and real numbers with the help of Particle Swarm Optimization method. The optimization results in this article show the consecutive weight of five sectors of LQ45 stocks are 0,353; 0,2465; 0,222; 0,1164; 0,0955 with the risk of 0,0030.