Document Type
Article
Abstract
Financial statement analysis is widely used for credit risk analysis. This method was developed at the end of 19’s for the purpose of surveying credit reliability of credit customers. However, the result of analysis with financial statements is likely to be controlled by the work experience of analysts. As the result, it is difficult to maintain consistency of risk measurement, and moreover, it is expensive and time consuming to perform the vast amount of precise evaluations required in a short time. Many researchers have studied financial statement analysis scientifically applying mathematical and statistical methods, especially with use of multivariate statistical analysis.
This report presents the history of preceding risk control studies and considers issues of the preceding default forecasting model based mainly on the binominal logit model on compensatory rule. Moreover we propose the improving scoring model on the non-compensatory rule and verify that our model is superior to the binominal logit model, and is useful in business using based on actual financial data.
Recommended Citation
Sakamaki, Yoshikazu, "Proposal of Improving Model for Default Probability Prediction with Logit Model on Non-Compensatory Rule" (2004). ICEB 2004 Proceedings (Beijing, China). 196.
https://aisel.aisnet.org/iceb2004/196