A credit rating is a forward-looking opinion of the relative creditworthiness of an issuer or a particular financial obligation. It summarises credit risk in an ordered category. A portfolio manager can use it to group exposures or apply a mandate, and a quant can use it as a state in a credit model. The category itself is neither a payment guarantee nor a numerical default probability.
The word relative matters. Ratings rank credit quality within a defined rating system; they do not measure equal distances between categories. Coding categories as successive integers can be convenient for software, but it does not make the economic gap between neighbouring grades constant. A higher-rated bond can still lose market value, and a lower-rated borrower may continue paying in full.
Read the label with its scope
A letter without its accompanying definition is incomplete data. Before comparing observations, keep these fields together:
| Field | What to establish |
|---|---|
| Agency and scale | Who assigned the opinion? Is it a global or national scale, and long-term or short-term? |
| Rated object | Is the rating attached to the legal issuer or to the particular obligation being held? |
| Currency and terms | Which payment commitments, currency and contractual features does the opinion cover? |
| Observation date | When was the rating effective, and was it subsequently changed or withdrawn? |
Issuer and issue ratings answer different questions. An issuer assessment concerns the borrower; an issue assessment concerns a specified claim. Collateral, guarantees and subordination can matter to the latter. Fitch’s definitions explicitly allow issue ratings to differ from the issuer rating and distinguish scales that incorporate recovery from those addressing default probability only. Moody’s global long-term ratings reflect both default or impairment likelihood and expected financial loss. Do not assume every agency or rating product measures exactly the same thing. See the checked Fitch definitions and Moody’s explanation.
For a familiar scale example, S&P’s long-term categories place A above BBB and BB. Its description treats BBB as investment grade and BB as speculative. That is a statement about that scale, not an instruction that every investor may buy BBB or must sell BB. Eligibility depends on the actual mandate, index methodology or applicable rule, including how it handles multiple agency ratings. S&P describes ratings as one input to investment decisions, rather than a judgment about investment merits. S&P’s scale and explanation.
A downgrade is an observation, not a default event
Consider a fictional agency that ranks A above BB. It assigns A to the same issuer at date 0 and BB at date 1, using the same scale and rating scope. This is a downgrade: the agency’s opinion of relative credit quality has deteriorated. It is an instance of credit migration.
The observation alone does not establish that the issuer missed a payment or otherwise experienced default. It also does not determine the size of a price change, an eventual recovery or a precise probability of default. A portfolio team would review exposure and mandate implications, then inspect the underlying credit information and contract.
Original synthetic illustration. A and BB are ordered categories in a fictional agency’s stipulated scale; the dates identify observations, not a specified elapsed horizon. No default probability is implied.
An outlook or watch designation also needs its own definition: it communicates possible direction or review status and is not itself a completed rating change. Fitch’s definitions distinguish these designations from rating actions.
The extra step from category to probability
To turn a category into a default probability, specify a default definition, horizon, population, observation period and estimation method. Historical agency default studies describe particular cohorts and periods. Applying them to a current borrower adds a modelling assumption; using a transition matrix adds assumptions about how ratings evolve. The letter does not supply those assumptions.
McNeil, Frey and Embrechts make that step explicit in their credit-migration approach: a chosen transition matrix supplies the probability of reaching default from the current rating. The default probability belongs to that specified model. It is not encoded as an immutable percentage in the rating symbol.
Nor does a rating specify the return for bearing the risk. A credit spread is a market-price comparison and can move while a rating remains unchanged. Fitch explicitly excludes general market-price and liquidity assessments from its credit ratings. Credit quality, purchase price and the investor’s realised return therefore require separate analysis.
For the surrounding bond mechanics, the SelfMadeQuant lesson Corporate debt, ratings and credit spreads connects rating categories and migration with bond terms and spreads. A public preview is available; the full lesson requires membership.
Sources
- S&P Global Ratings, Understanding Credit Ratings: relative creditworthiness, rating changes, long-term category descriptions and use in decisions; checked 2026-10-27.
- Moody’s Ratings, Understanding Credit Ratings: issuer/instrument scope and global long-term scale; checked 2026-10-27.
- Fitch Ratings, Rating Definitions, 19 September 2025: usage and limitations, outlooks/watches, issuer and obligation ratings; checked 2026-10-27.
- Alexander J. McNeil, Rüdiger Frey and Paul Embrechts, Quantitative Risk Management: Concepts, Techniques and Tools, revised edition, Princeton University Press, 2015, Sections 10.2–10.2.1, pp. 374–375.