Inverse data envelopment analysis for parallel production systems: an analysis in four types of Chinese banks

The conventional inverse DEA model is usually applied to all component units of a system, but there exist many problems in the real world that include some internal divisions operating as a parallel system, so this paper proposes a novel parallel inverse DEA model that can open the black box with individual component unit. Methodologically, two cases without and with price information are considered respectively. The proposed model is used to evaluate performance through a real-world application in the banking sector. We find that, in the real application, more important information is provided from the internal structure of different types of banks, and the interpretation of empirical results is reasonable. Furthermore, our model can be used in other contexts as well.