A credit rating is formally an opinion about the likelihood of repayment. Its practical power comes from the rules that reference it rather than from the analysis it contains.
What a rating is measuring
Ratings assess the probability of default and, in some scales, the expected recovery afterwards. They are not judgements about price, value or the risk of price movement.
The scale is ordinal rather than numerical. A given grade indicates relative standing rather than a specific probability, and the meaning is calibrated over long historical periods.
Sovereign ratings usually act as a ceiling for companies in that country, on the reasoning that a government under stress can restrict access to foreign currency. Exceptions exist but are limited.
Why the investment grade boundary matters so much
Many institutional mandates permit holdings only above a defined threshold. Crossing below it forces sales regardless of what the manager thinks.
Insurance and banking capital rules also reference ratings, requiring more capital against lower-rated holdings. The cost of holding the asset rises for regulated buyers.
The result is a step change in demand at a boundary that the underlying credit crosses gradually. Yields move more than the change in fundamentals would justify.
How the cost reaches the borrower
A wider yield on existing bonds sets the reference for new issuance. The borrower pays the higher rate on everything it refinances thereafter.
Some loan agreements contain pricing grids linking the interest margin directly to the rating. A downgrade raises the cost automatically without renegotiation.
Collateral requirements in derivative contracts are frequently rating-linked as well. A downgrade can therefore trigger immediate cash demands alongside higher interest.
Why ratings lag the market
Agencies aim for stability, avoiding changes that would soon be reversed. Bond prices react continuously while ratings move in discrete steps after review.
Market-implied measures derived from prices therefore anticipate rating changes. By the time a downgrade is announced, much of it is already in the price.
The announcement still matters because of the mandate and capital effects. The forced selling happens on the announcement, not on the anticipation.
The structural criticisms that persist
Issuers pay for their own ratings, which creates an obvious conflict that agencies manage through separation of commercial and analytical functions. The arrangement remains contested.
Embedding ratings in regulation was identified after past crises as amplifying procyclicality, and several regimes have sought to reduce mechanical reliance on them. Progress differs by jurisdiction.
Registration and oversight of rating agencies vary between countries and have been revised repeatedly. The underlying commercial model has largely survived those reforms.