Performance measurement in multifunctional bank branches: A nonparametric approach

Due to the combined effects of desintermediation, deregulation, and technology, competition in the financial services industry has intensified worldwide. Hence, banks can no longer limit themselves to their traditional reactive approach to retail banking and the sales activity has become increasingly important as non-interest incomes generated from fees and the sales of financial products of the bank capture a major share of banks' total income. In this thesis, we propose various models to evaluate the performance of a bank branch. We adjust the models to account for the fact that bank branches engage in sales activities that are distinct from the traditional service function. We also acknowledge that inputs, in particular resources; are often shared among these functions. To this end, Chapter 3 modifies the usual Data Envelopment Analysis (DEA) model in order to derive the multiple optimal efficiency scores (sales, service, and aggregate) contrary to traditional DEA applications that provides single measures of efficiency for functions that are clearly distinct and non overlapping. The proposed model recommends an optimal split of the shared resources that maximises the aggregate efficiency. Chapter 4 extends this analysis to non-volume related activities. Chapter 5 presents a methodology to choose best candidates for audits from the set of efficient branches derived in Chapter 3 and 4.