碩士 / 國立高雄第一科技大學 / 風險管理與保險所 / 93 / Alone in terms of the measurement of the credit risk in BASEL Ⅱ, Taiwan is going to face an impact since BASEL Ⅱ will be implemented at the end of year 2006. As to the measurement process for the risk factors in the credit rating, this study is in an attempt to apply the ways specified under BASEL Ⅱ to conduct the ordinal logit model as the empirical method. For this research, it is anticipated to evaluate the credit risk more accurately and find out various possible groups of high risk, and thus, to classify risks according to the risk level.
Samples in the current study were collected from a branch of some bank in southern Taiwan. Total data achieve to the number of 2332. The dependent variables are the 3-dimentional result of the non-performing loans situation, including normal, overdue below 90 days, and overdue beyond 90 days, respectively. The independent variables are age, gender, seniority, the number of latest banks inquired, and the income grade difference. The empirical results had showed that all 5 independent variables could affect the dependent variables significantly. Then the test samples were introduced and the predicting ability was examined in the evaluated model. It was found that the overall predicting ability was about 82-85%, but the ability to predict the non-performing loan accounts was low. Next, the independent variables were used to characterize in order to find out groups with high credit risk, so that groups with high credit risk could be highlighted and could be a reference for the banks during risk control and management.
Though banks are allowed to set up their own internal rating model under BASEL Ⅱ, the predicting ability when new data are introduced should be attended in using the model. The complicated model is not really more flexible than the simple model. Besides, severe bias may be caused if items on the credit rating list are simply used to predict as variables when you predict the violation cases by the logit model. In addition to the statistic model, the combination of different independent variables can help mark the high-risk groups by classification, which is another way for the banks to measure the credit.
| Identifer | oai:union.ndltd.org:TW/093NKIT5218024 |
| Date | January 2005 |
| Creators | Chia-fu Chang, 張家福 |
| Contributors | Hsien-Chueh Peter Yang, 楊顯爵 |
| Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
| Language | zh-TW |
| Detected Language | English |
| Type | 學位論文 ; thesis |
| Format | 60 |
Page generated in 0.0082 seconds