DOI: 10.5176/2251-1997_AF17.51
Authors: Abdollah Pakdel
Abstract:
Uncertainty in the capital market, price volatility, and firms stock returns make investors worried about the future of their investment. Thus, of the most important solutions to reduce this concern is to select suitable securities for investment and formation of stock portfolio. Portfolio optimization problem and determining the efficient frontier of investment, when the number of investment assets and the limitations of the market are low can be solved with mathematical models. However, when the conditions and limitations of the real world are considered, since most investment opportunities are risky, portfolio optimization problem is not solved easily using mathematical methods. Thus, modern methods such as evolutionary algorithms have been used to optimize portfolios. The main objective of this study is optimizing the stock portfolio using particle-swarm optimization method and comparison with the classical algorithm (mean-variance). For this purpose, mean and variance information of 25 companies listed on the Stock Exchange for the period 2010-2016 was collected and given to software (Matlab) and the related diagram was extracted. The results show that particle-swarm algorithm method of portfolio optimization has acted much stronger in reducing the uncertainty of optimal portfolio formation.
Keywords: particle swarm algorithm, classical algorithm, optimal stock portfolio, Markowitz optimization
Updating...