With regard to the above-mentioned problems, rating migrations seem to be a more reliable indicator of changes in credit quality than default rates. Given that the risks of downgrade as well as default vary over time, the question is whether credit spreads compensate investors adequately.
Since the sample for the calculation of rating transition matrices is much broader than for default rates, they are less likely to be biased by changes of the rating agencies’ universe. To measure changes of credit quality over time, the ratings drift, that is the number of upgrades minus the number of downgrades, as a proportion of the total number of entities rated, can be a valuable indicator. A sample of high-quality issuers, however, will tend to have more downgrades than upgrades, and vice versa. Hence, variations of the ratings drift partly reflect changes in average credit quality over time.
As one would expect, credit spreads tend to rise when the ratio of upgrades to downgrades becomes worse.
The question, however, is, whether the credit spreads widen enough to compensate investors sufficiently for the deterioration of average credit quality that is reflected by a falling ratings drift. While predicting the direction of spread changes may help to make money on a mark-to-market basis, it is not adequate for buy-and-hold investors. They have to estimate the magnitude of the spread widening that corresponds to an observed deterioration of credit quality. Hence, the focus is purely on credit risk, while credit spreads also incorporate liquidity premia, and are influenced by technical factors and market sentiment.