A novel ranking method in data envelopment analysis: a real case on Chinese banking industry

Purpose: This study aims to address the lack of discrimination between fully efficient decision-making units in nonparametric efficiency analysis models by introducing a new ranking technique that incorporates contextual variables. Design/methodology/approach: The proposed method combines Data Envelopment Analysis (DEA) and Ordinary Least Squares (OLS). First, DEA evaluates the partial efficiency of each unit, considering all inputs and only one output. Next, OLS removes the influence of contextual variables on the partial efficiencies. Finally, a ranking criterion based on modified partial efficiencies is formulated. The method is applied to data from 100 Chinese banks, including state-owned, commercial and industrial institutions, for the year 2020. Findings: The ranking results show that the top six positions are assigned to highly esteemed banks in China, demonstrating strong alignment with real-world performance. The method provides a comprehensive ranking of all units, including nonextreme efficient ones, without excluding any. It resolves infeasibility issues that arise during the ranking of efficient units and ensures uniqueness in efficiency scores, leading to a more reliable and robust ranking process. Contextual variables exerted a greater influence on the first partial efficiency compared to the second. Notably, Total Capital Adequacy (TCA) significantly impact bank efficiency. Originality/value: This study introduces a novel ranking method that effectively integrates contextual variables into DEA-based efficiency analysis, addressing limitations of existing methods. The practical application to Chinese banks demonstrates its utility and relevance.