بهره وری فنی بانک های ژاپنی : اندازه گیری عملکرد جهت دار غیر رادیال با خروجی نامطلوب

The paper analyses technical efficiency of the Japanese banks from 2000 to 2007. The estimation technique is based on the Russell directional distance function that takes into consideration not only desirable outputs but also an undesirable output that is represented by non-performing loans (NPLs). The results indicate that NPLs remain a significant burden as for banks' performance. We show that banks' inputs have to be utilised more efficiently, particularly labour and premises. We also argue that a further restructuring process is needed in the segment of Regional Banks. We conclude that the Japanese banking system is still far away from being fully consolidated and restructured.

The aim of this study is to examine technical efficiency of the Japanese banking sector from 2000 to 2007. We apply a non-radial directional performance measurement that quantifies the impact of non-performing loans (NPLs) on bank efficiency. The inclusion of NPLs in the model is underpinned by recent empirical studies on the Japanese economy and its banking system. Fukao [1], Hoshi and Kashyap [2] and [3] among others unambiguously confirmed that the extent of NPLs in Japan have been at the centre of its lengthy banking crisis. The Japanese banking system was brought almost at a standstill mainly due to credit misallocation in the late 1980s. A large volume of NPLs that had mounted on bank balance sheets were hidden by the Government in the early 1990s. Further, banks also concealed the true level of NPLs by accounting wizardry. The unsustainable volume of NPLs remained a dominant factor that has negatively influenced bank performance. Fukao [1] and [4] showed that the banking sector reported a negative operating profit from 1993 to 2002. The contribution of this study to current empirical research on bank efficiency and the Japanese banking system are as follows: first, the applied methodological concept provides a comprehensive analysis of efficiency levels of the individual bank inputs/outputs and compare them with the estimated optimal levels. The decomposition of total technical efficiency into inputs and outputs is an important addition to the current research on bank efficiency. Second, we examine nexus between NPLs and bank efficiency. We apply the innovative methodological approach that makes possible to quantify the negative impact of NPLs on bank efficiency. Third, we analyse bank efficiency during the extensive bank consolidation process across different types of banks over the period 2000–2007. We then examine the impact of the restructuring programme on bank efficiency. The methodology is based on an innovative non-parametric model introduced by Chen et al. [5], who showed that the non-radial model can compute efficiency inputs and outputs individually. That is not possible with the radial models since they are based on the assumption of proportional changes in inputs and outputs. The model is discussed in more details in Section 4. The paper is organised as follows: Section 2 provides an overview of the Japanese banking system in the last two decades with the emphasis on the recapitalisation and consolidation process. Section 3 then summarises studies on efficiency in the Japanese banking sector and the methodological papers on efficiency with bad outputs. Section 4 presents the Data Envelopment Analysis (DEA) that is applied in this study. Section 5 discusses the data and empirical results. Section 6 provides summary of the results.

This paper presents the innovative methodological model to estimate bank efficiency that is based on the Russell directional distance function that takes into consideration desirable and undesirable outputs. We disentangle efficiency by inputs, outputs and bad outputs and identify the contribution of each input for bank efficiency. The model is applied on the Japanese banks and analysed the changes in bank technical efficiency over the period 2000–2007. Firstly, we find that CRS and VRS display similar results. This means that the variable return to scale is a multiple of the constant return to scale and VRS are monotonic multiple of CRS. This is consistent with traditional results found in DEA models. Secondly, the decomposition of efficiency levels into individual input/output components provides a comprehensive picture on bank efficiency in Japan. Thirdly, the average inefficiency scores indicate a further deterioration in total technical inefficiency that increased from 0.28 to 0.36 over the analysed period. The inefficiency levels remain particularly high as for the number of employees and the use of banks' premises. As for banks deposits, the changes in inefficiency scores are rather marginal. We have identified an overall increase in inefficiency levels due to NPLs. Fourthly; we show that there exist large differences among all three types of banks, i.e., City Banks, Regional Banks I and Regional Banks II. City Banks present a marginal increase in their technical efficiency levels for employment and premises. Regional Banks I display a larger deterioration in employment efficiency scores, while premises and banks deposits maintain almost constant values. The aggregate measure of inefficiency increases along the observed time period. Further, we show that inefficiency levels for NPLs remain high. This clearly indicates that the problem with NPLs continues to be an issue and the consolidation and restructuring process has not yet been fully materialised. We report the similar results for Regional Banks II. The study confirms that the Japanese banking system is still far from being fully consolidated and restructured. Concerns remain about the non-restructured banks balance sheets with a large volume of NPLs, particularly in the segment of Regional Banks. These issues need to be addressed by banks and the regulatory authorities. The presented study confirms that the implementation of non-performing loans (NPLs) into the applied model may provide bank mangers and policy makers (regulators) an additional dimension in their decision process of how to implement optimal strategies and measures in order to improve efficiency of DMUs and the banking sector as a whole.