Using DEA and Worst Practice DEA in Credit Risk Evaluation

The purpose of this paper is to introduce the concept of worst practice DEA, which aims at identifying worst performers by placing them on the frontier. This is particularly relevant for our application to credit risk evaluation, but this also has general relevance since the worst performers are where the largest improvement potential can be found. The paper also proposes to use a layering technique instead of the traditional cut-off point approach, since this enables incorporation of risk attitudes and risk-based pricing. Finally, it is shown how the use of a combination of normal and worst practice DEA models enable detection of self-identifiers. The results of the empirical application on credit risk evaluation validate the method. The best combination of layered normal and worst practice DEA models yields an impressive 100% bankruptcy and 78% non-bankruptcy prediction accuracy in the calibration data set, and equally convincing 100% and 67% out-of-sample classification accuracies.     

Intra- and inter-country bank branch assessment using DEA

Increasingly globalized financial markets with considerable activity in the multinational sector have created the need to understand inter-country bank branch performance. This topic is relatively unstudied, primarily due to the immense difficulty encountered in gathering reliable data. Fortunately, we have been able to obtain data on a group of banks operating in one geographical market area, but in different countries. In this paper we critically assess bank branch profitability and productivity in seven national branch networks owned and operated by a multi-national financial services corporation. The corporate head office (owner) imposes its management philosophy equally on all of its subsidiaries, thus removing executive managerial and corporate disparity. Results suggest that countries in which branch performance is quite consistent amongst domestic branches are less productive and less profitable when compared to other countries that have more disparity in their efficiency scores. In addition, we discovered that, surprisingly, branches do not have to be productive in order to be profitable and this led us to somewhat of a major breakthrough in inter-country branch analysis. Significant managerial advice may be derived from these results vis-à-vis trans-national benchmarking and opportunity for performance improvements both at the branch level and nationally as well.

The examination of pseudo-allocative and pseudo-overall efficiencies in DEA using shadow prices

The original Data Envelopment Analysis (DEA) models developed by Charnes et al. (Eur J Oper Res 2:429–444, 1978), Banker et al. (Manag Sci 30:1078–1092, 1984) were both radial models. These models and their varied extensions have remained the most popular DEA models in terms of utilization. The benchmark targets they determined for inefficient units are primarily based on the notion of maintaining the same input and output mixes originally employed by the evaluated unit (i.e. disregarding allocative considerations). This paper presents a methodology to investigate allocative and overall efficiency in the absence of defined input and output prices. The benchmarks determined from models based on this methodology will consider all possible input and/or output mixes. Application of this methodology is illustrated on a model of the financial intermediary function of a bank branch network.     

Marginal Rates and Two-dimensional Level Curves in DEA

Of great importance to management, the computation of trade-offs presents particular difficulties within DEA since the piecewise linear nature of the envelopment surfaces does not allow for unique derivatives at every point. We present a comprehensive framework for analyzing marginal rates, and directional derivatives in general, on DEA frontiers. A useful characterization of these derivatives at given points can be provided in terms of the ranges they can take; equivalently, the bounds of these ranges correspond to derivatives to the rightand to the left at these points. We present two approaches for their computation: first, the dual equivalents calculation of minimum and maximum multiplier ratios / finite differences, and then a modified simplex tableau method. The simplex tableau method provides a more general application of the method introduced by Hackman et al. (1994) to generate any two-dimensional section of the isoquant and is a practical tool to generate level plots of the frontier. By giving a complete picture of trade-offs and allowing a better visualization of high dimensional production possibility sets, these tools can be very useful for managerial applications.     

Adjusting for cultural differences,a new DEA model applied to a merged bank

The usefulness and application of Data Envelopment Analysis (DEA) efficiency measurements is usually limited by the requirement of consistent operating circumstances. However, in many real world situations this is not the case, so to overcome this problem, this paper reports on a new strategy by inventing a Culturally Adjusted DEA model to benchmark business units that operate under different cultural (business) environments. This is especially useful when these environmental factors are partial causes of inefficiency and can not be simply incorporated into a DEA model as inputs or outputs. A simulation analysis is conducted to examine the effectiveness of the CA-DEA model for controlling these environmental effects. This model is applied to a real life efficiency study of two major financial firms in Canada in 2000, when the two entities started to consolidate and merge their branch networks. Two cultural indices are identified to represent a firm’s unique operating environment, one to capture the nature of a firm’s corporate strategies (Corporate Index), and the other to estimate the effectiveness of a firm’s operational systems (Service Index). The results show that a firm’s corporate culture has a significant influence on its branches’ efficiency and this, we found, is often neglected in such studies. This paper also makes a contribution to the bank merger literature by providing an internal view of the potential benefits that may result from sharing cultural advantages while identifying the true managerial inefficiencies.     

Combining DEA Window Analysis with the Malmquist Index Approach in a Study of the Canadian Banking Industry

The banking industry in Canada is essentially an oligopoly with five large participants controlling about 90% of the market. To evaluate the industry's performance over time, we need to deal with the problem of a small number of DMU's compared to the number of relevant inputs and outputs. To overcome this problem we use data envelopment analysis (DEA) window analysis, whereby efficiency scores for the 20 year period 1981–2000 are obtained. To measure productivity changes over time, Malmquist indices can be calculated from DEA scores. Using DEA window analysis scores, however, raise the question of how to define the same period frontier in a DEA window analysis. We show that for both the adjacent and the base period Malmquist index and for all suggested definitions of same period frontier, the standard decomposition into frontier shift and catching up effects gives inappropriate results when Malmquist indices are based on DEA window analysis scores.     

DEA-R: ratio-based comparative efficiency model, its mathematical relation to DEA and its use in applications

We propose a new mathematical model for efficiency analysis, which combines DEA methodology with an old idea—Ratio Analysis. Our model, called DEA-R, treats all possible ratios “output/input” as outputs within the standard DEA model. Although DEA and DEA-R generate different summary measures for efficiency, the two measures are comparable. Our mathematical and empirical comparisons establish the validity of DEA-R model in its own right. The key advantage of DEA-R over DEA is that it allows effective integration of the model with experts’ opinions via flexible restrictive conditions on individual “output/input” pairs.