Portfolio Optimization Using Data Envelopment Analysis-Based Asset Preselection and Segmented DEA Evaluation in the Iranian Capital Market

Document Type : Original Article

Author

department of finance, Islamic Azad univeristy

Abstract

This study introduces a two-stage non-parametric framework for portfolio optimization in the Iranian capital market. In the first stage, efficient assets are identified using the Slacks-Based Measure Data Envelopment Analysis (SBM-DEA) model. In the second stage, the constructed portfolios are evaluated through the Segmented DEA model, which assesses their efficiency relative to a piecewise efficient frontier under cardinality constraints. Based on data from 126 publicly traded companies between 2019 and 2023, the proposed hybrid model demonstrates superior performance in terms of Sharpe ratio, standard deviation, and proximity to the efficient frontier compared to benchmark approaches such as the Markowitz model, equal weighting, and PCA-based methods. Sensitivity analysis across different weighting schemes, asset counts, and market periods further confirms the robustness of the model. The findings support the practical applicability of the DEA–Segmented DEA framework in emerging markets and underscore its potential utility for portfolio managers and financial institutions.The findings support the practical applicability of the DEA–Segmented DEA framework in emerging markets and underscore its potential utility for portfolio managers and financial institutions.

Keywords

Main Subjects



Articles in Press, Accepted Manuscript
Available Online from 09 December 2025
  • Receive Date: 28 June 2025
  • Revise Date: 28 August 2025
  • Accept Date: 09 December 2025