Data Envelopment Analysis And Multivariate Analysis Set Decision-making In Applied Research, Network Optimization In The Commercial Banking Institutions

In the resources allocation optimization progress conducted by Chinesecommercial banks aimed at improving their core competences, there is an importantissue about the optimization of bank branches' setting and adjustment. As far as thisissue is concerned, the related problem in the optimization setting of existing bankbranches can be categorized as status quo evaluation, diagnosis and adjustmentdecision-making. It's of prominent practical significance to study on the decisionsupport methods for such problems.Based on the operational characteristics and operating status quo of Chinesecommercial banks, and focus on decision support for existing branches' settingoptimization from the viewpoint of profitability evaluation and diagnosis and theestimation of potential profitability, a systematic decision support approach based onDEA and multivariate analysis is developed in this study. This approach containsseveral modules including profitability evaluation and diagnosis, return-to-scaleanalyzing and adjustment decision-making. Profitability diagnosis based on thepresent scales of branches and scale efficiency analysis is combined to make decisionon commercial bank branches optimization setting. The holistic analyzing idea,analyzing methods in different modules, the systematic decision-making approach andapplication in a case study are presented gradually in this dissertation.Firstly the holistic analyzing idea is introduced. The idea is to diagnosis, viz. tofind the main sources of a bank branch's poor profitability by analyzing the maindriving factors of it. Thus the potential profitability of it may be estimated, andadvices on adjustment may be made. The main driving factors of bank branches'profitability can be classified into several kinds, viz. factors that can be control bybranches themselves such as efficiency and operation technology, scale factor whichare controlled by their supervising bank, and the external environment factors. Thusthe diagnosing of profitability and potential profitability estimating are based on theanalyzing of the influences of these factors on bank branches profitabilityrespectively.And then, to clear each factor's effect on profitability, a framework ofcross-sectional analyzing based on DEA and Tobit regression is built. After that, amethod also based on DEA and Tobit regression is presented for branch profitability evaluation and diagnosis based on panel data analyzing. And an approach based onDEA is presented to evaluate the scale efficiency of bank branch. Subsequently, thetactics correspond to different combinations of branch profitability expectations andscale efficiencies are presented respectively. Besides, an idea of using the DEA modelbased on interval input-output data to estimate the possible effects of differentadjusting solutions is presented.Finally, a comprehensive empirical study including aforementioned parts ofanalyzing is conducted based on 3 years' data from 36 branches of some Chinesecommercial bank to show the application process of the holistic approach. And theresults show the decision support significance of the approach built in this study.