The Impact of Size Heterogeneity on Performance Analysis in Data Envelopment Analysis Approach: An Application in Banking Sector

Data envelopment analysis is a powerful benchmarking tool in evaluating the relative efficiency of homogeneous firms with multiple inputs and outputs. The homogeneity assumption in DEA means that the firms consume the same inputs to produce the same outputs. In large samples and when there is heterogeneity in size, the benchmark points obtained from DEA models are infeasible and unattainable in practice. This paper focuses on the importance of firm size in performance evaluation in the framework of DEA. We define three size scores and by using these size scores, the set of all firms is partitioned into different groups. Corresponding to each firm, we define two efficiency scores, an Intra-group efficiency score and a Meta-efficiency score. Then the theoretical contribution is then applied to evaluate the performance and productivity changes in Asian banks from 2014 to 2020 using the DEA-based Malmquist productivity index. The study covered the 321 banks that were active during the study period. Our findings revealed that all six countries in our analysis have progressed from 2014 to 2020.