Relationship Benefit in Credit Ratings

President Signs S. 3850, the "Credit Rati...

President Signs the "Credit Rating Agency Reform Act of 2006". Image via Wikipedia

S&P downgraded the USA from AAA to AA+. For me, this downgrade is a no-brainer. After all, the rating is an estimation of the probability that the USA will not service its debts. Either because it is not able to or it is not willingly to follow it’s obligations. Given that all US debt is dollar denominated, it seems inconceivable that it cannot service its debt. But given the chaos in the congress its very possible that it one day may not be willingly to do so. „AAA“ indicates that this chance of this happening in the next year is 0.01% or once in 10,000 years. That seems very optimistic to me.

Therefore, from my perspective, the downgrade is no surprise. The only questionable thing is: Why not much earlier? The congress gone insane is not new. One possible answer come from Prof. Thomas Mählmann’s latest research publication „Is There a Relationship Benefit in Credit Ratings?„. He finds for the rating agency S&P that

This paper documents the existence of relationship benefits in credit ratings. In
particular, the paper shows that firms with longer rating agency relationships have
better (i.e., closer to AAA) credit ratings. […] controlling for observables, the firms with longer relationships, while having higher ratings, do not have lower default rates.

In other words, the ratings of bond issuers with a long lasting relationship with S&P, have higher default rates than equal ratings of new customers of the rating agency. In case of the USA that means: you shouldn’t have taken the AAA too serious in the first place, given the relationship length of a century or so.

2 Responses to Relationship Benefit in Credit Ratings

  1. nigecus sagt:

    interesting relationship. what is the conclusion of the paper? one explanation might be „my buddy cannot default“. another is increasing confidence regarding the credit analysis, i.e. longer data histories, less unknown problems to guess around, … the latter explanation would simply assume that people charge premiums for knightian uncertainty (e.g. if you don’t know, exspect the worst) or for ellsberg paradox like angst (i.e. you have to overcome angst to get same chances as under full information). it might be popular to focus favouritism (or „customer relationship mangement“). however, other effects might cause same/similar empirical/measurable symptoms.

    • ketzerisch sagt:

      The paper doesn’t investigate if the borrower tricks the agency, e.g. by knowing the threshold values, or if the agency hesitates to downgrade a long-standing good customer.

      I would exclude „increasing confidence“ as a reason. Confidence might be increase, but the agency would obviously be overconfident in this case as the defaults rate are too high given the rating and ex-post the confidence is thus unfounded.

Kommentar verfassen

Trage deine Daten unten ein oder klicke ein Icon um dich einzuloggen:

Du kommentierst mit Deinem Abmelden /  Ändern )

Google Foto

Du kommentierst mit Deinem Google-Konto. Abmelden /  Ändern )


Du kommentierst mit Deinem Twitter-Konto. Abmelden /  Ändern )


Du kommentierst mit Deinem Facebook-Konto. Abmelden /  Ändern )

Verbinde mit %s

%d Bloggern gefällt das: