Paper: arXiv 2503.15965
Authors: Hang Kin Poon
Abstract
Portfolio optimization is a critical area in finance, aiming to maximize returns while minimizing risk. Metaheuristic algorithms were shown to solve complex optimization problems efficiently, with Genetic Algorithms and Particle Swarm Optimization being among the most popular methods. This paper introduces an innovative approach to portfolio optimization that incorporates pre-assignment to limit the search space for investor preferences and better results. Additionally, taking margin trading strategies in account and using a rare performance ratio to evaluate portfolio efficiency. Through an illustrative example, this paper demonstrates that the metaheuristic-based methodology yields superior risk-adjusted returns compared to traditional benchmarks. The results highlight the potential of metaheuristics with help of assets filtering in enhancing portfolio performance in terms of risk adjusted return.
Complexity vs Empirical Score
- Math Complexity: 7.0/10
- Empirical Rigor: 3.0/10
- Quadrant: Lab Rats — high math complexity, limited empirical testing
Why this score: The paper uses mathematical formulations like the multi-objective optimization structure and Pareto optimality, but the empirical demonstration is an illustrative example without code, backtests, or heavy statistical metrics.
Research Flowchart
flowchart TD
A["Research Goal: Optimize Portfolio with Constraints"] --> B["Data & Parameters<br>Asset Returns, Risk, Preferences"]
B --> C["Pre-assignment Filtering<br>Limit Search Space"]
C --> D["Metaheuristic Optimization<br>GA & PSO Algorithms"]
D --> E["Compute Risk-Adjusted Returns<br>& Rare Performance Ratio"]
E --> F{"Compare against<br>Traditional Benchmarks"}
F --> G["Key Findings<br>Superior Risk-Adjusted Returns<br>& Enhanced Portfolio Performance"]
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper proposes an innovative metaheuristic portfolio optimization method that uses pre‑assignment constraints to limit the search space and incorporates margin‑trading strategies.
Source sentence
This paper introduces an innovative approach to portfolio optimization that incorporates pre-assignment to limit the search space for investor preferences and better results.
Assumes 4
- Pre‑assignment constraints reflect real‑world investor limitations and preferences. model
Source sentence
By incorporating pre-assignment constraints, investors can narrow down the search space to reflect real-world limitations and preferences.
- Margin trading can be modeled with initial, maintenance margins and the risk of margin calls. execution
Source sentence
Margin trading offers the potential for amplified returns by allowing investors to borrow funds to purchase assets. However, it also introduces significant risks, particularly the risk of a margin call.
- Single‑objective optimization (e.g., maximizing expected return) is computationally efficient compared to multi‑objective approaches. model
Source sentence
This approach simplifies the optimization process by concentrating on a single criterion, making it computationally efficient and easier to implement compared to multi-objective optimization.
- The benchmark portfolio is the SPY index held 100% with annual rebalancing. data
Source sentence
Benchmark Period 01/01/2015 to 01/01/2025 Rebalancing frequency Annually Weights
Data
2015-01 – 2025-01 · US equity market · equities
- Optimized Portfolio vs Benchmark market-prices unknown
Source sentence
Table 1. Optimization Results vs Benchmark Optimized Portfolio Benchmark Period 01/01/2015 to 01/01/2025 Rebalancing frequency Annually Weights LLY 68.2%, NVDA 31.8% SPY 100% CAGR 50.64% 13.01% Maximum Drawdown
Costs
Not modeled
Finds
Sharpe 1.18 (benchmark 0.38) · annual return 50.6%
Checked by others
No other paper in the archive has been found to support or contradict these findings. Coverage is partial, so this is neither replication nor refutation.
Code
No repository by the authors found in the paper.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.