A Risk-Aware Rule-Based Portfolio Management Framework for the Tehran Stock Exchange Using Genetic Algorithm Optimization

Algorithmic portfolio management has attracted significant attention due to its potential to reduce emotional decision-making and improve risk-adjusted investment performance. However, many conventional approaches based on deep neural networks and reinforcement learning are not well-suited for the Tehran Stock Exchange because of structural limitations such as low liquidity, price fluctuation limits, buy and sell queues, and the one-sided nature of the market. In addition, the relatively small and noisy dataset of the Iranian market increases the risk of overfitting in complex machine learning models. To address these challenges, this study proposes a rule-based portfolio management framework optimized using a Genetic Algorithm (GA). The model utilizes technical indicators, smart money activity, cash flow metrics, and buyer-seller power to generate trading signals. Initially, twelve trading rules were designed, and the GA selected the most effective rules based on Sharpe ratio optimization. The optimization process was performed on approximately ۲۰۰ randomly selected stocks with more than ۱,۰۰۰ daily observations covering the ۲۰۲۲-۲۰۲۶ period. Furthermore, the framework incorporates a dynamic risk allocation mechanism that adjusts the portfolio composition among equities, gold exchange-traded funds, and fixed-income assets according to investor risk tolerance. Back testing results indicate that the proposed framework outperformed the Tehran Stock Exchange index in most scenarios while achieving more stable risk-adjusted returns and lower drawdowns. The findings suggest that interpretable rule-based systems combined with evolutionary optimization and adaptive risk management can provide a practical and robust solution for portfolio management in structurally inefficient emerging markets such as the Tehran Stock Exchange.

کلیدواژه ها:

Portfolio Management ، Genetic Algorithm ، Tehran Stock Exchange ، Risk Management ، Rule-Based Trading System