The Effect of DEA-Based Stock Preselection on the Performance of Equally Weighted and Markowitz Portfolios: Evidence from the Tehran Stock Exchange

Articles in Press

Document Type : Original Article

Authors
Amir Salehi 1
Hashem Nikoomaram 2
Hamidreza Kordlouie 3
Mohsen Rostamy Mal Khalifeh 4
1 PhD Candidate in Financial Engineering, Science and Research Branch, Islamic Azad University, Iran
2 Professor, Department of Financial Management, Accounting, and Financial Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran.
3 Associate Professor, Department of Financial Management, Islamshahr Branch, Islamic Azad University, Tehran, Iran
4 Associate Professor, Department of Applied Mathematics, Science and Research Branch, Islamic Azad University, Tehran, Iran
10.22034/jik.2026.79288.4960
Abstract
The purpose of this study is to investigate whether asset preselection using Data Envelopment Analysis (DEA) can improve the performance of conventional portfolios, namely the equally weighted portfolio and the Markowitz mean–variance optimized portfolio. In this framework, an initial set of stocks listed on the Tehran Stock Exchange (the base sample) is first constructed. Then, using three DEA approaches—including DEA based on historical return/risk data, DEA based on MACD technical indicators, and DEA based on RSI technical indicators—15 “efficient” stocks are selected under each approach.
In the next stage, both equally weighted and Markowitz portfolios are constructed for the base sample as well as for each subset of the 15 efficient stocks. Annual return, annual risk, and comparative performance indicators are then calculated. The empirical results indicate that DEA-based preselection—particularly the historical-data-based DEA approach—can simultaneously increase expected returns and, in some scenarios, reduce portfolio risk. In other words, DEA acts as an “efficiency filter” that enhances the quality of the investment universe for portfolio allocation models. The findings have important practical implications for professional investors and financial policymakers, especially in emerging markets with trading constraints.
Keywords
Data Envelopment Analysis (DEA)
Stock Selection
Markowitz Portfolio
Equally Weighted Portfolio
MACD
RSI
Tehran Stock Exchange

  1. 1.        پی‌کنی، پ.، و همکاران. (2021). تحلیل پوششی داده‌های فازی خوش‌بینانه و بدبینانه: شواهد تجربی از بورس تهران

    2.        خسروی،و همکاران. (2024 . [مقاله درباره ترکیب ماتریس بردار پشتیبان خوشه‌بندی وتحلیل پوششی داده ها  و مدل مارکویتز در انتخاب سبد بهینه. نشریه مهندسی کیفیت و مدیریت، 14(4)، 358378.

    3.        رسول‌زاده، م.، و همکاران. (2024). مدل ترکیبی برای انتخاب سبد سهام بهینه با ترکیب اصول مارکویتز و تحلیل پوششی داده‌ها. فصلنامه/نشریه مالی

    1. ولی‌زاده، ا.، و همکاران. (2024). چارچوب دو مرحله‌ای تحلیل پوششی داده ها و یادگیری عمیق برای کارایی شرکت‌های بورسی ایران. 1(4), 226–235.
    2. Additional recent technical-indicator evidence (2023). Optimal and non-optimal MACD parameter values… Journal of Risk and Financial Management, 16(12), 508. (MDPI)
    3. Hosseinzadeh, A., Ortobelli, S., Lozza, A., & others. (2023). Portfolio optimization with asset preselection using data envelopment analysis. Central European Journal of Operations Research. (Springer Link)
    4. Peykani, P., et al. (2024). A novel robust network data envelopment analysis approach for mutual fund performance evaluation under uncertainty. Annals of Operations Research. (Springer Link)
    5. Santos, A. A. P. (2022). Markowitz meets technical analysis: Building optimal portfolios using RSI and MACD. Journal of Banking & Finance / Finance Research Letters (ScienceDirect). (ScienceDirect)
    6. Wei, C., & Zhang, W. (2020). Performance evaluation and portfolio selection using data envelopment analysis and Markowitz model: Evidence from the Chinese stock market. Journal of Computational and Applied Mathematics, 380, 112967. (DergiPark)
    7. Wu, X., & Fang, W. (2023). An improved تحلیل پوششی داده ها cross-efficiency method and application in fund performance analysis. Mathematics, 11(3), 585. (MDPI)
    8. Zatwarnicki, M., et al. (2023). Effectiveness of the Relative Strength Index signals in algorithmic trading/backtesting. [Open-access article in PMC]. (PMC)
    9. Zhou, W., et al. (2021). Big data and portfolio optimization: A novel approach integrating DEA with multiple data sources. Omega. (ScienceDirect)