Artificial Bee Colony Algorithm for Portfolio Optimization: Comparative Performance Analysis on Two different Indices
COMPUTATIONAL ECONOMICS, 2026 (SCI-Expanded, SSCI, Scopus)
- Yayın Türü: Makale / Tam Makale
- Basım Tarihi: 2026
- Doi Numarası: 10.1007/s10614-026-11427-3
- Dergi Adı: COMPUTATIONAL ECONOMICS
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Social Sciences Citation Index (SSCI), Scopus, IBZ Online, ABI/INFORM, Aerospace Database, EconLit, INSPEC, zbMATH, Pharma Collection (ProQuest), Technology Collection (ProQuest)
- Ondokuz Mayıs Üniversitesi Adresli: Evet
Özet
Portfolio optimization has become increasingly important in today's volatile markets, as traditional methods often face challenges when dealing with complex financial datasets. This has led to growing interest in flexible metaheuristic algorithms that offer global search capabilities. In this study, the performance of traditional and artificial intelligence-based optimization algorithms was comparatively evaluated for the portfolio optimization problem. The Sharpe ratio was adopted as the optimization criterion to balance portfolio return and risk. The analysis compared the Artificial Bee Colony (ABC), Genetic Algorithm (GA), and Particle Swarm Optimization (PSO) algorithms with the classical Quadratic Programming (QP) approach using historical data obtained from the Borsa Istanbul and Dow Jones Industrial Average indices. The results indicate that the metaheuristic algorithms achieved higher Sharpe ratios than the Quadratic Programming approach for the analyzed datasets. Furthermore, independent executions conducted under different random initializations showed that the ABC algorithm produced more stable and reproducible results with lower variability than the other metaheuristic methods. In addition, parameter sensitivity analyses demonstrated that the ABC algorithm consistently reproduced the optimal portfolio with the maximum Sharpe ratio across different parameter configurations, indicating superior robustness with respect to parameter selection. Overall, the findings suggest that the ABC algorithm represents a competitive and robust alternative for Sharpe ratio-based portfolio optimization under different market conditions.