Least-Distance Range Adjusted Measure in DEA: Efficiency Evaluation and Benchmarking for Japanese Banks
Least-Distance Range Adjusted Measure in DEA: Efficiency Evaluation and Benchmarking for Japanese Banks
复制标题
DEA 中的最小距离范围调整措施：日本银行的效率评估和基准
DOI：
10.1142/s0217595922500063
复制
发表时间：
2022
影响因子：
1.4
通讯作者：
Hasuike Takashi
中科院分区：
文献类型：
--
作者：
Wang Xu;Hasuike Takashi
本研究旨在建立数据包络分析（DEA）中的最小距离范围调整测度（LRAM），并将其应用于评价日本银行业的相对效率，为日本银行业提供基准信息。在DEA中，传统的范围调整措施（RAM）作为一个定义良好的模型，满足一组理想的属性。然而，由于最小距离测量的实用性，我们制定的LRAM，并提出使用一个有效的混合整数规划（MIP）的方法来计算它在这项研究中。所提出的LRAM（1）满足与传统RAM相同的期望性质，（2）为低效决策单元（DMU）提供最小距离基准信息，（3）可以通过使用所提出的MIP方法容易地计算。在这里，我们将LRAM应用于2017-2019年期间的日本银行数据集。基于结果，LRAM生成更高的效率得分，并允许低效银行以比RAM所需的更小程度的输入-输出修改来改善其效率，从而表明LRAM可以为低效银行提供更容易实现的基准信息。因此，从决策单元管理者的角度，本研究为决策单元的效率评价和标杆分析提供了有价值的LRAM。
This study aims to formulate the least-distance range adjusted measure (LRAM) in data envelopment analysis (DEA) and apply it to evaluate the relative efficiency and provide the benchmarking information for Japanese banks. In DEA, the conventional range adjusted measure (RAM) acts as a well-defined model that satisfies a set of desirable properties. However, because of the practicality of the least-distance measure, we formulate the LRAM and propose the use of an effective mixed integer programming (MIP) approach to compute it in this study. The formulated LRAM (1) satisfies the same desirable properties as the conventional RAM, (2) provides the least-distance benchmarking information for inefficient decision-making units (DMUs), and (3) can be computed easily by using the proposed MIP approach. Here, we apply the LRAM to a Japanese banking data set corresponding to the period 2017–2019. Based on the results, the LRAM generates higher efficiency scores and allows inefficient banks to improve their efficiency with a smaller extent of input–output modification than that required by the RAM, thereby indicating that the LRAM can provide more easy-to-achieve benchmarking information for inefficient banks. Therefore, from the perspective of the managers of DMUs, this study provides a valuable LRAM for efficiency evaluation and benchmarking analysis.