STOCK PORTOFOLIO VALUE AT RISK MINIMIZATION USING COMBINED METHOD OF GENETIC ALGORITHM AND SIMULATED ANNEALING

• Main Author: Unsulangi, Kevin
• Format: Final Project
• Language: Indonesia
• Subjects: Fisika
• Online Access: https://digilib.itb.ac.id/gdl/view/38104
• Institution: Institut Teknologi Bandung

In a stock market, an investor must have an investment strategy that has many benefits with little risk. It becomes difficult because of the many variables that are taken into account. One solution is to calculate Value at Risk (VaR), or the calculation of the maximum loss probability of an investment based on a level of confidence. The optimal stock portfolio is a stock portfolio that has a great opportunity to generate profits with low risk. A good stock portfolio is also a diversified portfolio and has a share of investment burdens that tend to be more in good companies / assets. The purpose of this study is to create an optimal stock portfolio, by dividing the percentage of the investment burden of each asset based on the optimization method, so that a low risk stock portfolio is obtained (low VaR). The formed stock portfolio comes from companies listed in the LQ45 index for 10 periods (August 2013-July 2018), because they’re believed to have high valuations and capitalization values. The optimization method used in this study is Simulated Annealing (SA) (optimization method with the Annealing principle in Metallurgy), and Genetic Algorithm (GA) (optimization method with the principle of evolution and natural selection). The two optimization methods have their own advantages and disadvantages, so in this study we will try to combine the two methods in the hope of producing a more optimal solution. The results of this study indicate that the combined GA-SA method can provide an far optimal solution, with an average VaR value of almost 1.5% less than the portfolio VaR with the division of investment expenses equally for each asset. This shows that the combined GA-SA method can be used to form a stock portfolio by dividing the investment burden in such a way that it can be said to be safe for investment. The hope is that the results of this study can be useful for investors or hedge fund managers in terms of investment risk management.