Are investor-paid ratings really better?

Let me take you all back to prehistoric times. A time so long ago, most readers will not have first-hand memories of it. Let me take you back all the way to 2009. Back then we had a lively discussion about credit ratings. Because rating agencies like S&P or Fitch are paid by the issuer of a bond, the incentive for them is to give the bonds a better rating than they deserve. Back then, one of the solutions proposed was to have ratings paid for by investors whose incentive is to get the best indication if a bond is about to default.

Since then, investor-sponsored credit ratings agencies have mushroomed, and I certainly can’t speak for every one of them and the quality of their ratings. Besides, these companies have expensive lawyers and I want to make sure they don’t send me nasty letters.

So, let me simply quote the results of a study Nan Qin and Lei Zhou and ask any lawyer not to sue me. I am just the messenger.

Qin and Zhou compared the ratings of S&P and Fitch (issuer-sponsored ratings) and investor-paid Egan-Jones Ratings (EJR). They found that S&P ratings on average were more stringent than the EJR ratings and the Fitch ratings. And while the EJR ratings were sometimes better able to warn of impending defaults than Fitch, the comparison of the ratings accuracy between S&P and EJR is very lopsided. The chart below shows the hit rate and the false alarm rate for S&P and EJR over one-year time horizons (left) and three-year horizons (right).

Receiver operating curves for S&P and EJR

Source: Qin and Zhou (2023)

The further up a point is the higher the share of bonds in default the ratings agency predicted in advance. The further to the right a point is the more false alarms were issued by the ratings agencies, i.e. they claimed a bond was at risk of defaulting when it didn’t. The fact that the curve for S&P ratings is above and to the left of the curve for EJR ratings shows that S&P ratings are more reliable and have fewer false alarms than EJR ratings.

As I said, I can’t speak for all investor-sponsored ratings agencies, but at least in this case, it looks like issuer-sponsored ratings like those of S&P are at least as good if not better. Which doesn’t make me hopeful that in the next credit crisis we are going to be able to avoid similar losses as during the financial crisis.