Reliable evaluation of an obligors’ creditworthiness and the estimation of the probability of their default (PD), exposure at default (EAD) and expected recovery (RR, LGD) are crucial components of the credit risk management process. Various financial models ranging from expert assessments to advanced statistical methods are commonly used for this purpose. Considerable model building expertise and deep knowledge of the capital regulatory framework are required due to the wide-ranging impact of these models. They drive regulatory IRB capital requirements, affect the reported profit via impairment provisions and influence operational performance as primary input to the credit approval process.
Fintegral has extensive experience in the development, calibration and validation of PD, LGD and EAD models and possesses deep understanding of their regulatory context. In order to facilitate the model building process, we have designed and developed a suite of programming routines in R and SAS. Our credit risk team has expertise in the use of traditional expert and regression-based model building techniques as well as machine learning techniques. We have our own user guide to the validation of various PD, LGD and EAD models ensuring the consistency and effectiveness of the validation process.