Financial Early Warning System Models of Taiwanese Banks

• Main Authors: Kang-Rung Chen, 陳康容
• Other Authors: Chi-Hsiang Huang
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/4qk365

碩士 === 國立虎尾科技大學 === 經營管理研究所 === 98 === Since 1980, Taiwan’s government relaxes restrictions of banks gradually, new banks establish induce the competition intense in bank-industry. And government strengthened banking supervision and practiced financial reform bills actively. Therefore, banks to face the sudden changes in the financial environment that management of risk is an important issue. This article using financial information among 33 domestic banks from first season in 2004 to the fourth season in 2008, using CAMELS ratings to assess the bank operating performance. The early warning model using Logit regression analysis to set up the non-risk model and risk model. That can be reference for financial inspection units and related financial institutions. The Empirical research results indicate as follows： (1) Select by factor analysis variables that consistent with CAMELS (Capital Adequacy、Asset Quality、Management、Earning、Liquidity、Sensitivity to market risk) that means this article can objectively evaluate the bank performance. Hence, that can reference for the relevant institutions and investors. (2) Use Logit regression analysis set up original sample models that prediction ability are：non-risk model A is 72.73% and risk model B is 78.79%. Predictive sample models of prediction ability：non-risk model A is 90.91% and risk model B is 96.97%. (3) The CAMELS of non-risk models has good predictive ability, the model add risk indicators that predictive ability and type I error work better than non-risk models.