Summary: | 碩士 === 國立高雄應用科技大學 === 金融資訊研究所 === 95 === ABSTRACT
The corporate default risk is a major composition of credit risk. Bank management have more concerned with corporate default risk management to abate business risk and enhance stockholders’ equity. Taiwan belongs to a high credit risk financial market, the great part of bank official according to the at least requirement of regulation to make loan provision. It may cause too low coverage rate than other country, hence it is most important for bank management to reinforce forecasting’s accuracy of credit risk. In this paper, we propose a model to achieve efficient and feasible to predict probability of default of corporate of credit risk based on Binary Logistic Regression Model. This model will produce a risk signal to advise bank management to make a right decision. This paper data from Taiwan Economic Journal Database, from December of 1996 to June of 2006 in season’s the financial variables, the external rating variables, company information variables, the accountant's variables, the macroeconomic variables and the corporate governance variables on the general industry of TSE and OTC of Taiwan, the way of mating by1:1 of the normal company and the default company in the same industry, the same data period and similar asset size, forecasting correct rates use Logistic Regression Model. The major research findings include:forecasting accuracy are high for the financial variables model and synthesize variables model, the effect valuation-Receiver Operating Characteristic and Kolmogorov-Smirnov Test those confirmed are suitable of correct rates of model. This paper make a contribution to choose significant variables and build a predict corporate default risk model accurately , but it is to be short of Override variables, also it has not used the merge financial reports, and segment credit rating grades of corporate loan by probability of default yet, It is satisfy risk management requirement for bank to suggest the future researcher improve the model refer to the banker opinion, outside credit rating and BaselⅡ.
Key word:Probability of Default of Corporate、Credit Risk、Logistic Regression
|