有序多分类logistic模型在违约概率测算中的应用

The Application of Ordered Logistic Regression Model in the Default Probability Measure

初始违约概率的测算是商业银行实施经济资本管理的必要环节。针对我国商业银行的现状,结合贷款五级分类,通过对银行的公司类客户的财务指标作时间加权化处理、因子分析、ROC检验以及使用有序多分类logistic模型对初始违约概率的测算作了有价值的探索,并通过算例分析论证了其可行性。

The initial default probability measure is an important aspect in the economic capital management of commercial bank. In the light of the present status of China＇s commercial banks, combined with loan＇s five - level classification, the paper makes valuable research to initial default probalility measure through time -weighted treatment on the financial indicators of the bank＇s corporate customers, the factor analysis, the ROC examination and the use of ordered logistic regression model, and proves its feasibility through the example analysis.