Analyzing efficiencies of city commercial banks in China: An application of the bootstrapped DEA approach

Abstract

City commercial banks (CCBs) have continued to grow in support of small- and medium-size enterprises (SMEs), thus improving the misallocation of capital and helping to accelerate China's tremendous economic growth. This study uses bootstrapped data envelopment analysis (DEA) to statistically test the property of returns to scale by taking into account undesirable outputs in order to employ the appropriate model for statistically inferencing the characteristics of these CCBs. Employing a dataset from Bankfocus that covers 101 CCBs for the period 2015–2017, the returns to scale test supports technology as exhibiting variable returns to scale. The partitioning around medoids (PAM) algorithm, based on bootstrapped DEA scores, clusters China's CCBs into three groups: Cluster 1 contains half of the superior foreign-owned banks and half of those competitive local CCBs being located in the coast region, while Cluster 2 and Cluster 3 consist primarily of domestic CCBs associated with similar characteristics in locations and owner-types, where the former outperforms the latter. Other findings show that (1) biased upward efficiencies may provide incorrect information and misguide managerial and/or policy implications, and that (2) ignoring the effort by CCBs at absorbing deposits leads to significantly different results.