Investors face an increasingly complex decision-making environment due to the growing number of alternatives in financial markets. Portfolio optimization has emerged as a hot research area over the past few decades, aiming to determine how much to invest in various assets. However, introducing real-world conditions renders the optimization problem NP-hard, leading researchers to adopt evolutionary algorithms for approximate solutions.
This paper presents strengthening strategies for multi-objective evolutionary algorithms to achieve faster convergence rates and extensive search capabilities in portfolio optimization under cardinality constraints. A unique solution representation, a novel operator, and new repair mechanisms are introduced to address the problem of setting lower and upper limits on the number of assets in the portfolio.
New mating strategies, along with the aforementioned enhancements, are implemented in well-known multi-objective evolutionary algorithms for testing. The customized algorithms demonstrate better approximations and faster convergence compared to traditional methods, even as the number of market assets increases, without compromising performance.
Blogger's Review: This paper significantly boosts the efficiency of multi-objective evolutionary algorithms in portfolio optimization through innovative solution representation and mating strategies, showcasing the powerful potential of evolutionary computation in financial applications.