The Spanish banking system has gone through a period of profound change over the last fifteen years, mainly as a result of deregulation which, along with other important features such as technological advances or an increasing culture of finance are contributing to significantly reshape the industry. These puzzling trends, together with the traditional importance of the banking system for the economy as a whole, lead us to consider that change to be an outstanding topic of policy relevance and research interest.
One of the most intensively studied topics has been the efficiency of both commercial and savings banks. The aforementioned reasons have led many researchers to analyze how banks’ efficiency is affected by this new competitive environment. More precisely, it is widely accepted that the removal of entry barriers forces banking firms to become more efficient, in order to raise industry competitiveness.
However, bank efficiency research studies vary widely in their aims and results, as there are differences regarding the techniques used, the definitions of outputs and inputs, the firms being analyzed (savings banks and/or commercial banks), the sample period uncovered, etc. With regard to the technique, some of them consider a nonparametric approach, while others choose econometric methods. In some cases only savings banks are focussed on, essentially because of their greater homogeneity and reliability of data, while others consider the whole industry, with commercial banks also analyzed. In addition, the period uncovered by research studies varies widely, which constitutes a further source of variation.
But even if a long period is considered, the conclusions drawn on efficiency dynamics are usually based on only two moments of the distribution of the efficiency scores, namely, the mean and the standard deviation. These statistics help a great deal to acquire a general picture of the overall efficiency trends in the industry, but they do not fully inform about the differences at the firm level. In particular, a time-invariant dispersion index may hide very different data structures, which may have important economic implications. Features such as multi-modality, non-normality or asymmetry of the data are impossible to uncover by any dispersion indicator.
In this study, a model of explicit distribution dynamics will enable us to fully characterize the dynamics of cost efficiency scores. And, in particular, due to the tighter competitive conditions currently facing firms, an upward trend should emerge. Yet if, for instance, a pattern such as bi-modality persisted (i.e., some firms being more efficient or inefficient than the average) and relatively cost inefficient firms did not abandon the industry, appropriate explanations should be sought out.
Among these, we will consider one which has not been sufficiently stressed in the literature: the role of each firm’s output mix. Concerning the Spanish case, very few studies have explicitly considered that different specializations may entail different cost efficiency levels, as some activities are more costly than others (Prior and Salas, 1994). In particular, if different product mixes are not controlled for, an upward bias towards inefficiency could exist. Another way to consider the relevance of the output mix consists of defining different banking output specifications, as it involves assigning greater importance to different specializations. However, despite it being generally accepted that different output measures bias efficiency scores, this has been done only in Grifell-Tatjé et al. (1992). Our approach to control for specialization when analyzing efficiency and efficiency dynamics will not be the same, as the model considered to assess the dynamics of efficiency scores requires a different way of controlling for some variables.
The study is structured as follows. Section 2 briefly reviews the sources of debate and controversy when measuring efficiency issues and estimates the cost efficiency for banking firms through a DEA approach. Section 3 introduces the methodology to analyze efficiency dynamics, while Section 4 assesses how banks’ output mixes may bias efficiency results and efficiency dynamics. Finally Section 5 concludes.
Download
PDF Ebook Bank Cost Efficiency As Distribution Dynamics: Controlling For Specialization Is Important
