دانشگاه اصفهان مدیریت دارایی و تامین مالی 2383-1189 14 4 2026 12 22 Portfolio Optimization Using Quantitative Value Allocation Method in Iran's Capital Market بهینه‌سازی سبد سهام با روش تخصیص ارزش کمّی در بازار سرمایۀ ایران 47 68 30156 10.22108/amf.2025.140404.1979 FA میثم دعائی گروه مدیریت مالی، دانشگاه آزاد اسلامی واحد اسفراین، اسفراین، ایران Journal Article 2025 05 04 This study proposes and implements a two-phase portfolio optimization model in Tehran Stock Eschange (TSE) and Iran Farabourse Company, integrating a quantitative value assessment with a multi-objective optimization framework. The first phase involves a fundamental valuation of constituent companies, while the second phase constructs an optimal portfolio using a multi-objective mathematical model. The epsilon-constraint method was employed to solve the model, utilizing financial data sourced from the Codal database. Empirical results demonstrate the model's efficacy in supporting investment decision-making. Portfolios formed through this integrated approach, which systematically incorporates both fundamental and technical analysis, significantly outperformed randomly selected portfolios. Furthermore, the model successfully accommodated diverse investor preferences within the asset selection process. By developing a structured quantitative framework grounded in a defined value function and multi-objective optimization, this research provides a precise and adaptable tool for optimal asset allocation in TSE, facilitating the simultaneous application of fundamental and technical criteria in portfolio construction.<br /><strong>Keywords</strong>: Tehran Stock Exchange, Iran Farabourse Company, Portfolio Optimization, Quantitative Value Allocation.<br /><strong>JEL</strong>: G11, C61, C44, C58, G12.<br /><strong> </strong><br /><strong>Introduction</strong><br />This study examines the application of a quantitative value allocation (QVA) framework to portfolio optimization in the TSE and Iran Farabourse Company. The proposed framework integrates principles of fundamental and technical analysis into a multi-objective optimization structure, aiming to address limitations inherent in classical portfolio models. Traditional approaches, most notably the mean–variance model, often rely on simplified assumptions—such as perfect asset divisibility, negligible transaction costs, and static investor preferences—that seldom align with the realities of emerging markets. In contexts such as Iran’s capital market, challenges including varying disclosure quality, liquidity constraints, and pronounced market frictions further complicate the direct application of conventional models. To bridge this gap, the present study introduces a more realistic portfolio construction model that accounts for multi-period progression, firm-level fundamental scoring, interdependencies in risk factors, and heterogeneous investor utility functions. The study is guided by the following central research question: Can a hybrid, value-driven quantitative methodology yield portfolio performance superior to random selection, and does the integration of fundamental and technical signals contribute to a more robust and consistent investment strategy under Iran’s specific market conditions?<br /><strong> </strong><br /><strong>Methods & Materials</strong><br />This study employs a two-phase mathematical modeling framework. The first phase develops a quantitative value function to screen companies based on six fundamental dimensions: liquidity, efficiency, leverage, profitability, market valuation metrics, and systematic risk. Each indicator is converted into a unit-free score using adjustable preference thresholds, categorizing performance into acceptable, neutral, or penalty zones; only stocks with a positive aggregate score advance to the subsequent stage. The second phase formulates a three-objective optimization model designed to maximize the portfolio's temporal progression, maximize its time-weighted average value, and minimize the covariance among selected assets. The model incorporates constraints on the number of assets, minimum and maximum investment thresholds, and binary inclusion decisions. The ε-constraint method is utilized to generate the Pareto-efficient frontier, selected for its efficacy in navigating non-convex, multi-objective search spaces. The empirical analysis utilizes financial data from 420 companies listed on the Tehran Stock Exchange and Iran Fara Bourse between 2017 and 2021, sourced from Codal, Fipiran, and official market databases. For comparative performance evaluation, the proposed model is benchmarked against portfolios constructed via three alternative approaches: random selection, low-beta selection, and a strategy based on technical indicators (MFI, RSI, SMA, and MACD), with each portfolio comprising seven assets.<br /> <br /><strong>Findings</strong><br />he value-based screening process successfully identifies a subset of firms exhibiting robust fundamentals across key industries, including basic metals, pharmaceuticals, chemicals, cement, and energy. These firms, characterized by superior composite value scores, are subsequently integrated into the multi-objective optimization model. The solution generated via the ε-constraint method yields an optimized portfolio encompassing assets from over twenty-five industries, effectively balancing sectoral exposure while achieving high aggregate value and maintaining low inter-asset covariance. This allocation not only adheres to defined investor constraints but also results in stable and interpretable portfolio weights. Comparative backtesting reveals that portfolios constructed using the proposed Quantitative Value Allocation (QVA) framework consistently outperform randomly selected portfolios in terms of both absolute returns and performance consistency. While beta-based diversification strategies improve risk-adjusted returns relative to random portfolios, they generally underperform the optimized value-based model. Portfolios formed exclusively on technical signals (MFI, RSI, SMA, MACD) demonstrate periods of high returns but are accompanied by significantly elevated volatility. Overall, the integrated hybrid approach delivers the most stable and favorable performance profile, underscoring the persistent explanatory power of fundamental valuation as a core performance driver even within volatile and informationally constrained market environments.<br /> <br /><strong>Discussion and Conclusion</strong><br />This study demonstrates that a quantitative, value-centric, multi-objective portfolio optimization framework presents significant empirical and theoretical advantages over heuristic or discretionary portfolio construction methods. By integrating a fundamental valuation screen with a formal multi-objective model and augmenting this with technical signals, the proposed approach effectively synthesizes firm-level structural attributes with market-driven price dynamics. The findings substantiate contemporary theoretical arguments advocating for the fusion of economic fundamentals with quantitative techniques, aligning with the growing literature on hybrid strategies for emerging markets. From a practical standpoint, the model offers a structured, actionable tool for investors, underscoring that systematic value-based selection can enhance long-term performance and improve risk management relative to random or single-factor strategies. The study acknowledges inherent limitations, including potential data quality constraints, the static nature of the fundamental scoring, and the model's omission of exogenous macroeconomic shocks. Future research should aim to extend this framework by introducing dynamic preference mechanisms, stress-testing its performance under extreme volatility, or integrating robust or fuzzy optimization techniques to better account for parameter uncertainty. In conclusion, this research contributes a flexible and empirically validated framework that refines portfolio selection practices within the Iranian capital market while providing a foundational platform for subsequent methodological innovation. هدف اصلی پژوهش، طراحی و پیاده‌سازی یک مدل بهینه‌سازی سبد سهام با رویکرد تخصیص ارزش کمّی در بستر بازار سرمایۀ ایران است. این پژوهش در دو فاز شامل ارزیابی کمّی ارزش شرکت‌ها و بهینه‌سازی سبد سرمایه‌گذاری با اهداف چندگانه تدوین شد. داده‌ها از سایت کدال استخراج و برای حل مدل بهینه‌سازی چندهدفه از الگوریتم اپسیلون استفاده شده است. نتایج، کارایی رویکرد پیشنهادی در ارتقای تصمیم‌گیری سرمایه‌گذاران را تأیید می‌کند. نتایج نشان داد که سبدهای تشکیل‌شده براساس تحلیل بنیادی و تکنیکال در مقایسه با انتخاب‌های تصادفی بازدهی بیشتری داشتند و مدل ارائه‌شده توانست ترجیحات متنوع سرمایه‌گذاران را در انتخاب بهینۀ دارایی‌ها به‌طور مؤثری لحاظ کند. این پژوهش با توسعۀ یک چارچوب کمّی دومرحله‌ای مبتنی بر تابع ارزش و بهینه‌سازی چندهدفه، مدلی دقیق و منعطف برای تخصیص بهینۀ دارایی‌ها در بازار سرمایۀ ایران ارائه کرده است که امکان لحاظ هم‌زمان تحلیل بنیادی و تکنیکال را در فرایند انتخاب سبد سهام فراهم می‌آورد. https://amf.ui.ac.ir/article_30156_22a29f2babe7881ae12b44fe7ec59dbb.pdf