This paper is drawn from the use of data envelopment analysis (DEA) in helping a Portuguese bank to manage the performance of its branches. The bank wanted to set targets for the branches on such variables as growth in number of clients, growth in funds deposited and so on. Such variables can take positive and negative values but apart from some exceptions, traditional DEA models have hitherto been restricted to non-negative data. We report on the development of a model to handle unrestricted data in a DEA framework and illustrate the use of this model on data from the bank concerned.
Bibliographical noteThis is an Accepted Manuscript of an article published by Taylor & Francis Group in Journal of the Operational Research Society in 2004, available online at: http://www.tandfonline.com/10.1057/palgrave.jors.2601768
- bank branches
- data envelopment analysis
- negative data in DEA