Abstract
The objective of this article is 2-fold. First, it provides an empirical assessment of the cost and profit stochastic frontiers based on a panel dataset of Greek commercial banks over the period 1993 to 2005. Second, on the basis of the same sample, it also compares the most widely used parametric and nonparametric techniques to cost efficiency measurement, namely, the Stochastic Frontier Approach and Data Envelopment Analysis. The results suggest greater similarities between the predictions of cost and profit efficiency methods than between parametric and nonparametric techniques. Such evidence is new in the literature and calls for a more technically level playing field for estimating bank efficiency.