An empirical study of stochastic DEA and financial performance: the case of the Turkish commercial banking industry

This study breaks important new ground in the analysis of financial institutions. It is one of the first empirical uses of Stochastic Data Envelopment Analysis (SDEA) in the efficiency literature. The pattern of efficiency is examined for the year 1999. The purpose of stochastic setting of DEA is two-fold: to accommodate both the inefficiency and the presence of measurement errors; and to convert the resulting stochastic linear programmes for DEA into deterministic non-linear DEA programmes. The results show that there are wide variations in the DEA efficiency scores and SDEA results suggest that these are due to measurement errors or other stochastic factors in the raw data, probably attributable to macroeconomic shocks and issues of changes in banking regulations.