Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor | Department of Industrial and Systems Engineering | en_US |
| dc.contributor.advisor | Tang, Chak Yin (ISE) | en_US |
| dc.creator | Wang, Xianhe | - |
| dc.identifier.uri | https://theses.lib.polyu.edu.hk/handle/200/14624 | - |
| dc.language | English | en_US |
| dc.publisher | Hong Kong Polytechnic University | en_US |
| dc.rights | All rights reserved | en_US |
| dc.title | Enhancing retail investor portfolios with behavior-based models and particle swarm optimization | en_US |
| dcterms.abstract | Portfolio selection is a critical decision-making problem in financial engineering, where investors strive to optimize the balance between risk and return. While classical models, such as Markowitz's mean-variance theory, have laid the foundation for portfolio optimization, they often fail to account for the psychological biases and complex behaviors exhibited by real-world investors, particularly retail investors. These models also struggle under extreme market conditions and high-dimensional, non-convex optimization spaces, where computational challenges such as premature convergence and local optima arise. This research addresses these significant limitations by developing innovative portfolio selection models and enhanced optimization algorithms. | en_US |
| dcterms.abstract | The novelty of this work lies in its integration of behavioral finance with advanced optimization techniques, specifically tailored to the needs of retail investors. First, a multi-objective portfolio selection model which simultaneously incorporates prospect theory, disappointment theory, and expected utility theory is proposed. This novel combination allows the model to better reflect investor behavior under uncertainty by balancing the pursuit of returns with the avoidance of extreme losses. Additionally, a second portfolio model incorporates multiple fuzzy reference points to capture the dynamic behavior of return chasing, a common phenomenon in investor decision-making. This model offers a more adaptive approach, avoiding overly conservative or aggressive investment strategies and improving risk control. | en_US |
| dcterms.abstract | To address the computational challenges of local optimal and premature convergence, two advanced variants of the particle swarm optimization algorithm will be introduced. These algorithms overcome the common problems of premature convergence and local optima by decomposing the swarm into sub-swarms with distinct iteration strategies and introducing adaptive parameter optimization via deep deterministic policy gradient techniques. These enhancements significantly improve the efficiency and accuracy of particle swarm optimization in solving complex optimization problems in behavioral portfolio selection. | en_US |
| dcterms.abstract | The results of this study show that the proposed portfolio models outperform traditional approaches in terms of return-risk balancing, particularly under extreme market conditions. The enhanced particle swarm optimization algorithms demonstrate superior performance in high-dimensional, non-convex problems, consistently finding optimal solutions more efficiently than standard particle swarm optimization methods. The effectiveness of the models and algorithms is verified through extensive comparisons with state-of-the-art techniques, proving their practical applicability and advancing the state of knowledge in portfolio selection and optimization. | en_US |
| dcterms.abstract | The contributions of this study lie in how a behaviorally ground, multi-objective portfolio selection model, a multi-reference point extension of prospect theory, and two collaborative-adaptive particle swarm algorithms collectively address the empirical gaps of investor irrationality and algorithmic premature convergence. Case studies on real-market data validate that the proposed portfolio selection models consistently deliver higher risk-adjusted returns than classical and contemporary benchmarks, while computational studies demonstrate the superiority and generality of the proposed particle swarm algorithm variants. This study charts future research on real-time reference point calibration, hybrid heuristic-convex optimization, and resource-aware optimization algorithms, laying a roadmap for more adaptive and resilient portfolio selection framework under uncertainty. | en_US |
| dcterms.extent | xii, 179 pages : color illustrations | en_US |
| dcterms.isPartOf | PolyU Electronic Theses | en_US |
| dcterms.issued | 2026 | en_US |
| dcterms.educationalLevel | Ph.D. | en_US |
| dcterms.educationalLevel | All Doctorate | en_US |
| dcterms.accessRights | open access | en_US |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/14624