Skip to main content

S&P: Obviously No One Thought Our Ratings Were Objective Or Unbiased Or, For That Matter, True

  • Author:
  • Updated:

The Journalchecked in on the government's lawsuit against S&P today, since there's a hearing coming up next week where S&P will ask a court to dismiss the case. You can read the arguments here, here, here, and here. My sympathies are mostly with S&P, though I'm also not holding my breath that a judge will bounce this case entirely without giving the government a chance to take some depositions and bluster a bit.

What's fun here though is to consider S&P's two main arguments which are:

  • When we say we have a policy of objectivity and independence and avoiding conflicts of interest, we don't want anyone to take us seriously about that, and
  • What do you mean our ratings of RMBS CDOs in 2007 were wrong? They were fine, stop whining.

They really say these things; e.g. (page 8 of the motion to dismiss):

Third, the Complaint alleges that S&P misled the public through pronouncements—made in a November 2005 policy statement and in codes of conduct released in September 2004, October 2005 and June 2007—concerning how it should rate securities. The Complaint recites S&P statements that “No employee of Standard & Poor’s/McGraw-Hill shall attempt to exert improper influence on the opinions of an Equity Analyst or a Rating Analyst” and “In all analytic processes, Ratings Services must preserve the objectivity, integrity and independence of its ratings.” These statements, however, are prescriptive statements S&P made to its employees concerning how they should carry out their business activities. As with the aspirational statements above [e.g. that S&P's “mission” is to “provide high-quality, objective, rigorous analytical information to the marketplace”], they are commands for how business should be conducted. They are not actionable representations of objective facts.

And (pages 10 & 13 of S&P's reply):

B. The Complaint Fails to Plead S&P’s Ratings Were False. ...

The Government argues that because the Complaint identifies conflicting, internal opinions within S&P regarding how ratings should be made, the Complaint has sufficiently pled that S&P’s ratings on all CDOs rated between March and October 2007 were false and actionable under the fraud statutes. Not so. Those allegations still fail to plead with the required particularity that S&P knew how future RMBS performance would impact RMBS ratings used in the CDO ratings process, and knew that the CDO ratings were false when made. The Government has not properly pled fraud; it has merely highlighted differing opinions that arose within S&P during the analytic process.1

(It's helpful to read that last paragraph with the context of "pretty much all of those investment-grade-rated CDOs defaulted with huge losses soon after they were rated.")

Both of these arguments are ... pretty good arguments?2 Like, in a court case and stuff? I mean, I wouldn't want to use them in S&P's advertising materials or whatever, but they're good for a court case. They cite a bunch of cases and stuff. And they come from a pretty sensible place, which is a place of:

  • everyone more or less thought that S&P's job, in rating structured credit, was to provide some mathematical rating in line with the ratings provided by Moody's so that the structured credit machine could continue churning and S&P and Moody's could continue getting fees, and
  • it was, and
  • no one particularly thought the other thing: that S&P's job was to avoid conflicts of interest, say, or to live up to its advertised code of conduct, or to provide objective ratings of credit quality, or to provide true ratings of credit quality. Or, one is tempted to add, to provide any ratings of credit quality at all. S&P's job was to produce certain sequences of letters that structurers and underwriters and buyers of CDOs were happy with at the time they were produced. Going back and saying "oh well that sequence of letters turned out to be wrong" is a category mistake.

I LIKE IT A LOT, of course. Its cynicism is bold and pleasing and twofold. There's the immediate cynicism: you can't hold us to our promises of objectivity and independence, and you can't prove that our ratings were wrong just because they were, y'know, horribly wrong.

But there's a subtler cynicism, which situates S&P in its exact place on the CDO construction chain and disclaims any responsibility for anything else in the chain. "We just gave you a sequence of letters, with lots of warnings about how you shouldn't use it for any purposes," says S&P. "We didn't do any credit work because S&P rated it," say investors. "We just put together what investors wanted and what S&P was willing to rate," say the banks. Whatever went wrong is The System's fault: you can't blame it on S&P. Or anyone.

S&P Says Judge Should Dismiss Government’s Case Against It
U.S. v. McGraw-Hill Companies et al. - Complaint [C.D. Cal. via Reuters]
S&P Motion to Dismiss [C.D. Cal.]
Government Opposition [C.D. Cal.]
S&P Reply [C.D. Cal.]

1.More in the weeds but sillier is page 16 of the reply:

For example, the Government alleges that 52% of the collateral backing the Plettenberg Bay CDO was non-prime RMBS rated BBB or below. The CDO was rated in March 2007, and it was affirmed two months before S&P announced large-scale downgrades to non-prime RMBS ratings in July 2007. The Government does not allege, and it cannot, that those large-scale downgrades—the very actions that it alleges S&P “knew” it was going to take back in March 2007—had any impact on the rating of the CDO. To the contrary, the Government claims that those downgrades impacted a mere 3% of RMBS collateral and does not allege any impact on the rating of the CDO. If downgrades to the underlying RMBSs did not have any impact on the CDO when they were taken in July, there is no reason to believe that S&P’s supposed “knowledge” that they were coming back in March would have had any impact then either.

The point here is that (1) the government claims S&P knew large-scale RMBS downgrades were coming in March, but didn't flow those downgrades through to its CDO ratings until July, but (2) when they did flow them through there was no impact, so (3) the government's allegation that the delay from March to July constituted fraud is bullshit. A good argument! But consider (page 60 here) that Plettenberg Bay defaulted in March 2008, and that its AAA-rated tranches - rated AAA in March and July 2007, that is - apparently had a zero recovery, and you might conclude that something was wrong with those AAA ratings anyway.

2.Ooh there's a third argument (pages 18-21 of the reply) that goes like this:

The Government’s Complaint, however, fails to allege that S&P had the specific intent to deprive the CDO investors—the supposed victims—of any money or property. Instead, it focuses on the alleged evils of the “issuer-pays model,” in which the banks that issued the CDOs actually paid S&P’s fees. But because those issuing banks are not the victims of the alleged scheme to defraud, each of the Government’s fraud claims lacks an essential element and must therefore be dismissed.

In other words: "We can't have defrauded investors! We only got paid by the banks! Why would we have cared about investors?" OKAY.