Validation of Credit Risk Rating and Scoring Models
Developed credit scoring models, being the core of risk estimation process, need to be timely and properly validated at least annually, but usually even more often. Credit scoring models are one of the most important aspects for business in financial industry for both people inside and outside of any organization. As such, their quality, precision and compliance need to be and stay at the highest level. Validating process for credit scoring models can be very time demanding and complex task, but we can help you achieve this task in best practice possible.
How can Deloitte help?
We are here to provide for you detail and thorough validation for every credit scoring model or similar tasks. We have specialized team of experts with banking experience knowledgeable in validation subjects that can deliver validation tasks in the best way for you and your business.
We are here to be confident support in:
- Validation of Rating and Scoring Models for IFRS9, Basel and managerial purpose. Special focus on clients’ segments (Corporate, SME, Private Individuals, Project Finance, Financial Institutions) and product’s types (Working Capital, Investment Loans, Multi-Purpose Lines, Credit Cards, Overdrafts, Mortgages, Consumer Loans, Agriculture loans, Contingent Products) including techniques Back-testing, Discriminatory power, Analyses of representativeness, Analyses of overrides, Stability analyses of the internal ratings and risk parameters over time, Analyses of model design stability, Evaluation of input data, Benchmarking analyses, Data cleansing analyses, Review of the model specification, Quality assurance of the computer codes used and Additional qualitative analyses.
- Rating Policy and Model Validation Methodology - Application models, Behavior models and Expert models (Risk Differentiation, Risk Quantification, Calibration/Recalibration, MoC Estimation, Good and Bad Years and Downturn analysis, Model Theory, Target Segmentation, Variables Estimation, Model Estimation).
- Credit Risks Models Documentation Validation.
- Validation of development data quality.
- Validation of credit model maintenance framework.