Multi-Objective Portfolio Optimization Strategy Using the SPEA-II Algorithm

Abstract

In the finance world, the precise selection and optimization of stock portfolios are of paramount importance. This study explores the application of intelligent algorithms, particularly the multi-objective of Strength Pareto Evolutionary Algorithm II (SPEA-II), alongside traditional methods to determine optimal portfolios. Using the monthly stock prices, the Markowitz model is developed, focusing on the return and semi-variance criteria. Realistic constraints are applied to formulate a multi-objective optimization problem. SPEA-II and traditional multi-objective optimization methods are used to solve this problem, resulting in a set of optimal portfolios. The results show that the SPEA-II algorithm can generate portfolios with higher returns and lower risks compared to the Markowitz model and traditional methods, taking into account the complex and nonlinear conditions of the capital market. In addition, the SPEA- II algorithm showed significant efficiency and stability across different frequencies and time periods. The study highlights that the SPEA-II algorithm can serve as an effective and efficient method for stock portfolio selection and optimization, helping investors to identify portfolios with lower risk and higher return.

Keywords

• Multi-objective strategy
• Pareto front
• SPEA-II optimization
• Markowitz model