국내 금융기관 경영실태평가모형 개선에 관한 연구: 상호금융기관을 중심으로

The failure of financial institutions not only can incur a considerable amount of social cost, but also can cause direct damages to their customer. Therefore, various forms of early warning models have been developed in the United States since late 1970s for the systematic evaluation of management status and failure prediction. Many financial institutions in Korea were also liquidated due to the strong restructuring during the financial crisis in 1997. Since then, however, the failure of non-bank financial institutions, including mutual savings banks, has significantly affected on the entire economy as well as the financial market. Accordingly, the evaluation of management status and the system improvement of the financial institutions becomes an important research subject. In this study, an empirical analysis was conducted with respect to the current management evaluation (CAEL model) of Financial Supervisory Service in Korea, which has shown a high error ratio in determining the soundness of credit union institutions, focusing on the adequacy of failure classification. To this end, based on the various management indexes and the experience of structure improvement that have been systematically managed by National Agricultural Cooperative Federation of Korea, the significance of failure classification of various indices was first reviewed, and then the CAELV model, which is better than CAEL model in its failure prediction, was presented using the method of Principal Component Analysis (PCA). In addition, the statistical power analysis using AR (AUROC) method shows that the CAELV model is excellent in its discrimination power with 0.90 (0.95) and higher in its hit ratio for failure (prediction) with 96.30%(type 1 error of 3.70%), indicating that it is useful for evaluating the management status of credit union institutions.