Loan grading assigns an institution-defined risk category to a credit exposure and updates it as repayment risk changes.
Loan grading is the process of assigning an institution-defined risk category to a loan or credit exposure and updating that category as credit quality changes. A grade summarizes repayment risk for monitoring, approval, portfolio reporting, review, and loss analysis. There is no universal 1-to-10 or A-to-C scale that applies to every lender.
A useful grade communicates both the amount and direction of credit risk. It can reflect:
The lender’s policy should define each grade and the evidence needed to assign or change it. A numeric grade of 3 at one bank can represent a different risk level from a grade of 3 at another.
Banks often use several internal pass grades to separate stronger and weaker acceptable credits, plus criticized or classified categories for heightened risk. In U.S. bank supervision, commonly referenced categories include special mention, substandard, doubtful, and loss. These terms have supervisory meanings and should not be casually mapped to consumer letter grades.
| Category type | General analytical use |
|---|---|
| Pass grades | Distinguish acceptable credits by relative risk under the institution’s system |
| Special mention | Identifies potential weaknesses deserving close attention before they become more serious |
| Substandard | Indicates well-defined weaknesses that jeopardize repayment under existing terms |
| Doubtful | Includes substandard weaknesses with collection or liquidation in full highly questionable and improbable based on current facts |
| Loss | Identifies amounts considered uncollectible or of such little value that continued recognition as a bankable asset is not warranted |
The table is a high-level educational summary. Institutions should use the controlling supervisory definitions and their own approved policy rather than this page to classify a specific credit.
At origination, underwriting may assign a proposed risk grade based on verified historical results, forecasts, structure, collateral, and stress analysis. After closing, the grade should respond to new evidence rather than remain fixed because the loan was once approved.
Possible review triggers include:
A grade change should record the effective date, evidence, rationale, approver, and any resulting monitoring or action plan.
Assume a lender approved a business loan with stable debt-service coverage, timely reporting, and moderate leverage. Six months later, the borrower’s largest customer, representing 35% of revenue, cancels its contract. Cash flow declines, the borrower misses a reporting deadline, and projected coverage falls close to 1.0x.
These facts do not dictate a universal grade. They do justify a prompt review of the repayment source, liquidity, customer replacement plan, covenant compliance, collateral, and management forecasts. The lender may move the credit to a weaker internal pass grade, place it on a watch list, assign a criticized category, or take another action under its definitions. The decision should follow evidence and policy, not a generic letter-scale formula.
If the borrower later replaces the customer and demonstrates sustained recovery, an upgrade may be appropriate. A single favorable month would usually provide less evidence than a durable improvement across several reporting periods.
| Measure | What it answers |
|---|---|
| Loan grade | How the institution currently categorizes credit risk |
| Consumer credit score | How a scoring model ranks a consumer’s predicted credit behavior |
| Probability of default | Estimated likelihood of default over a defined horizon |
| Nonaccrual status | Whether interest recognition has stopped under applicable policy or rules |
| Allowance for credit losses | Accounting estimate of expected credit losses under the applicable framework |
| Default | Whether a contractual, regulatory, or model-defined default event has occurred |
These measures can influence each other but are not interchangeable. A loan can be downgraded before payments become past due, and two loans with the same internal grade can have different expected-loss estimates because of collateral, maturity, or exposure differences.
An effective grading process commonly includes:
Technology can support consistency and monitoring, but a model output is not self-validating. Data quality, threshold design, overrides, and changes in borrower behavior can weaken results.
The OCC material applies in its stated supervisory context. Other regulators, accounting frameworks, jurisdictions, and institutions can use different terminology and rules. This article provides general financial education, not a classification of any specific loan or credit advice.