One of the most popular tools to monitor credit risk is a standardized risk rating system. A credit risk rating system provides banks and credit unions the opportunity to grade transactions in their commercial loan portfolio by level of risk. CEIS Review, a New York-based bank consulting firm, recently published an article on these systems in their newsletter, The CEIS Quarterly. The article cites a lack of specific requirements and standard models for credit risk rating systems. The only requirement from regulators is assigning transactions, when appropriate, to four criticized categories: special mention, substandard, doubtful and loss.
Because there is no standard model, each institution is allowed to customize a risk rating system to fit its own needs. However, each institution should pay special attention to the level of detail included in its model. According to CEIS, the “number of grade levels will primarily depend on the breadth of the spectrum of risk embedded in the portfolio, and where within that range it lies both in dollar terms and in terms of number of transactions and/or clients.” Finding the right balance between having insightful information and avoiding unnecessary complexities is key.
In 2001, the OCC published the Comptroller’s Handbook on Rating Credit Risk, which highlighted the expectations of credit risk rating systems:
1. The system should be integrated into the bank’s overall portfolio risk management.
2. The board of directors should approve the credit risk rating system.
3. All credit exposures should be rated.
4. The risk rating system should assign an adequate number of ratings.
5. Risk ratings must be accurate and timely.
6. The criteria for assigning each rating should be clear and precisely defined using objective and subjective factors.
7. Ratings should reflect the risks posed by both the borrower’s expected performance and the transaction’s structure.
8. The risk rating system should be dynamic.
9. The risk rating process should be independently validated.
10. Banks should determine through backtesting whether the assumptions are valid.
11. The rating assigned to a credit should be well supported and documented in the credit file.