Estimating and analyzing the efficiency and productivity of bank branches

Abstract

Purpose The purpose of this paper is to examine the efficiency and productivity of a Greek bank's branches. Design/methodology/approach The sample consists of 458 branches of a Greek commercial bank, operating in 13 regions of Greece over the period 2002‐2005, a total of 1,795 observations. Data envelopment analysis was used to explore the efficiency and productivity of the branches. Then, fixed and random effects models were used to determine the impact of internal and external factors on the efficiency and productivity scores. Findings The results indicate that the branches in the sample could have achieved improved overall performance during 2002‐2005. Also, that the inclusion of loan loss provisions as an input variable increases the efficiency score, but for the total factor productivity (TFP) change, the results are mixed. The second stage regressions indicate that both the logarithm of personnel and the logarithm of income per capita in the local market have a significant impact on efficiency, while the loans to total assets ratio has a significant impact on pure technical efficiency only. When the various productivity change measures were regressed over the explanatory variables, it was found that the logarithm of per capita gross fixed capital formation has a positive and statistically significant impact on all measures. Also, that the return on assets, the loans to deposit ratio, the logarithm of personnel, and the logarithm of income of per capita, all have a positive and statistically significant impact on overall efficiency change. Originality/value This paper is the first study on Greek branches which examines the impact of market conditions. It examines the impact of risk‐taking on the efficiency of the branches and examines the productivity growth of the branch network using the Malmquist TFP index.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 49 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.