This repository contains code and resources for analyzing a portfolio of consumer mortgages, building predictive models for probability of default (PD), constructing scorecards, and performing risk analytics aligned with regulatory requirements (such as Basel regulations).
The workflow in this project covers:
- Data Analysis: Exploration and cleaning of consumer mortgage portfolio data.
- Probability of Default Modeling: Building statistical and/or machine learning models to predict the likelihood of default for each loan.
- Scorecard Development: Translating model outputs into a practical credit scorecard for risk segmentation.
- Cutoff Optimization: Determining the optimal score threshold for lending decisions.
- Loan Binning & Monitoring: Grouping loans by score, and tracking default rates across bins over time.
- Economic Factor Analysis: Studying how observed default rates relate to macroeconomic variables.
- Provisioning & Regulatory Compliance: Calculating provisions needed by the bank to meet Basel regulatory standards.
- End-to-end workflow for credit risk modeling using consumer mortgage data.
- Tools for scorecard creation and cutoff optimization.
- Analysis modules for linking credit performance to economic factors.
- Guidance for regulatory provisioning (Basel compliance).