Academic journal article Journal of Money, Credit & Banking

Nonparametric Frontier Models for Assessing the Market and Cost Efficiency of Large-Scale Bank Branch Networks

Academic journal article Journal of Money, Credit & Banking

Nonparametric Frontier Models for Assessing the Market and Cost Efficiency of Large-Scale Bank Branch Networks

Article excerpt

In this paper we propose models for assessing the efficiency in large networks of bank branches. We distinguish bank branch efficiency into market and cost components suitably modified to capture different tiers of bank management. The paper proposes a methodology which includes the use of multivariate analysis in order to ensure the homogeneity of the branches assessed and then data envelopment analysis for assessing efficiency. The methodology is applied on a sample of 580 branches of a commercial bank in the United Kingdom. The results obtained reinforced previous claims regarding the presence of high technical inefficiencies and economies/diseconomies of scale at the branch level from a production and cost point of view. Furthermore, the decision to pre-cluster the network of branches into homogenous groups has had profound implications on the magnitude of the assessed efficiencies.

The performance of financial institutions is a topic that draws considerable attention within the business environment of many countries. This interest primarily concerns the corporate performance of individual banks (see Berger, Hunter, and Timme 1993). Less attention can be found in the literature of branch-specific efficiency studies which is the principal channel of the production work in banking. Berger, Leusner, and Mingo (1997) discuss in detail the consequences of ignoring the bank branch efficiency on issues of technical efficiency, economies of scale, and product mix.

The assessment of branch efficiency has variable significance to financial institutions that operate under different regulatory regimes. For example, in markets with no restrictions on the size of the branch network of individual banks (for example, Greece, Japan, and the United Kingdom) the corporate performance of a bank does not convey any information about the performance of its possibly three thousand branches. In the United Kingdom the focus on bank branch performance has had recent stimuli from (i) the limited ability of the main clearing banks to realize investment profits from foreign markets and (ii) the intensified competition at the national level by the entrance of the mortgage lending institutions (building societies) with expanded retail banking capabilities. This changing environment brought up issues related to the operational efficiency and the quality of services at the bank branch level. The recognition of the importance of these issues invited considerable theoretical and applied work over the last ten years (see Sherman and Ladino 1995).

Previous studies about bank branch efficiency have concentrated on their ability to control costs. In the parametric frontier literature Pavlopoulos and Kouzelis (1989) employed a translog cost function on 362 branches of the National Bank of Greece. They reported economies of scale and scope at the branch level while monopolistic phenomena were detected in the rural areas of Greece. Doukas and Switzer (1991) also used a translog cost function on 563 bank branches in Canada and reported the presence of economies of scale at the branch level. Another translog cost function was most recently employed by Zardkoohi and Kolari (1994) on 615 branches of 43 Finnish banks also reporting economies of scale at the branch level. Berger, Leusner, and Mingo (1997) analyzed 760 branches of a U.S. bank over a three-year period using a Fourrier-Flexible (instead of the translog) form. The latter study reports results regarding the presence of technical and scale inefficiencies at the branch level that explain partly the bank inefficiencies at the corporate level.

The other methodology used to assess bank branch efficiency is based on nonparametric deterministic frontier estimations whose applications have been on small sample sizes. Sherman and Gold (1985) focused on fourteen branches of a savings bank; Parkan (1987) assessed thirty-five branches of a chartered bank in Canada; Oral and Yolalan (1990) assessed twenty branches of a bank in Turkey; Vassiloglou and Giokas (1990) and Giokas (1991) studied the efficiency of twenty branches of the Commercial Bank of Greece. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.