This course specifically deals with developing PD (Probability to Default) Models-Credit Risk Scorecard Models under BASEL II. We will briefly discuss the basic fundamentals and related concepts in the Credit Risk, but our main focus is going to be on developing Credit Risk Scorecards Models-from HOW TO DEVELOP CREDIT RISK MODEL perspective, using popular techniques like Logistic Regression, and not from WHAT IS CREDIT RISK perspective.
We will deal with data preparation, understand the typical variables in Credit Risk Models (like delinquency in 60/90/120 day period, tradelines, total defaults, total assets, balance to debt ratio, demographic variables, trend variables on account/tradeline activity etc and others), binning/coarse classification of the variables, Information value and Weight of Evidence, imputation and missing value treatment, variable transformations, using Data Mining techniques like Logistic Regression and others to build a PD Model, Model Evaluation, Validation and Monitoring, understanding PSI (population stability index) and Scaling/Calibration in Credit Scoring.
Online Availability for overseas candidates-Yes
Go back to Business Analytics Training