Stochastic p-robust DEA efficiency scores approach to banking sector

This paper presents a stochastic p-robust data envelopment analysis (DEA) model for decision-making units (DMUs) efficiency estimation under the input uncertainty. This allows a deleterious effect to the objective function to better hedge against the uncertain cases that are commonly ignored in classical DEA models. This paper has been arranged to launch along with the following steps: we briefly reviewed the classical CCR DEA model. After that, the p-robust DEA model is introduced and then we calculated the priority weights of each scenario for CCR-DEA output-oriented model using analytical hierarchy analysis process (AHP). To tackle the uncertainty of experts’ opinion, we applied a synthetic technique based on both robust and stochastic optimizations. In the sequel, we proposed stochastic p-robust models for the estimation of efficiency, with particular attention being paid to DEA models. The proposed model provides a more encompassing measure of efficiency in the presence of synthetic uncertainty. According to the results, we obtained the expected score, relative regret score, and stochastic p-robust score for DMUs. Also, we showed that the stochastic p-robust DEA model is a proper generalization of traditional DEA models and gained the desired robustness level. The applicability of the proposed model is illustrated in the context of the analysis of Iranian commercial bank performance. Finally, by an example, we showed that the objective values of the input and output models are not inverses of each other as in classical DEA models.