APPLICATION OF STATISTICAL PHYSICS TO OPTIMISE LQ45 PORTFOLIO IN INDONESIA STOCK MARKET USING RANDOM MATRIX THEORY METHOD

• Main Author: MAHARDHIKA MAULANA (NIM : 10212009), ALIFIAN
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/20950
• Institution: Institut Teknologi Bandung

Random Matrix Theory (RMT), introduced by Dyson and Mehta, is a method that can explain the energy levels of complex nuclei. It It has recently been applied to the noise filtering in financial time series, especially on systems with a large dimension, such as the stock market, by several authors including Plerou et al. and Laloux et al. Both authors had conducted an analysis on the US stock markets and found that the correlation matrix eigenvalues of return were consistent with calculations using random return, with the exception of few large eigenvalues. Modern portfolio theory is a theory used to maximize profits from a portfolio. However, the problems rising in the current economic circumstances have led to the modern portfolio theory not being quite able to explain how to maximize the portfolio. In this paper, author would apply the RMT method to eliminate the effect of noise from the historical data, and then the filetered results would be applied in the calculation of efficient frontier portfolios and distribution analysis, as stated in modern portfolio theory. Application of the RMT has been successfully carried out on stocks in the list of Indonesias LQ45 portfolio, resulting in maximum return at 1.9 percent with the probability of the risk at maximum of 1, as well as on the chart of efficient frontier which can be seen that the optimum return is at 1.4 percent with a probability of maximum risk at 0.5.