Estimating Probability of Default for Individual Obligors Based On Basel II

Probability of client's default is the most focal parameter in the credit risk Evaluation based on Basel II Accord. In this research, in order to estimate the probability of default, first, a scoring model applying the Logit Regression Technique using the 1343 client's personal data, credit history, employment information in one of the country's private banks within 1391-1392 was estimated. Through this study, the client's credit history was identified as the most important factor influencing the probability of default. The statistic ROC (Receiver Operating Characteristic) of the Model reflected 71.1% implying its high discriminant power. Cluster analysis, using K-means Algorithm, of the scoring estimates of the model lead to 7 rating classes. Next, applying historical default frequency approach, the probability of the default associated with each rating class was estimated. Finally, through the calibration test, the accuracy of the probability of default was confirmed for a sample of 327 credit profiles in 1393. Undoubtedly, compliance with the international agreements on the banking area is inevitable.