In July 2007, a recently hired analyst in Standard & Poor's
structured finance group exchanged a series of emails with an
investment banking client who wanted to know how the new job was
going. Things were just great, the analyst said sardonically,
"aside from the fact that the MBS world is crashing, investors
and media hate us and we're all running around to save face ...
no complaints." Part of the problem, the analyst said in a
subsequent email, was that some people at S&P had been pushing
to downgrade structured finance deals, "but the leadership was
concerned of p*ssing off too many clients and jumping the gun
ahead of Fitch and Moody's.
I suspect that only an investment banker could find it in
his or her heart to sympathize with a credit analyst in the
summer of 2007, but this one suggested that some good might come
from S&P's internal conflict. "This might shake out a completely
different way of doing biz in the industry," the banker wrote.
"I mean come on, we pay you to rate our deals and the better the
rating the more money we make?!?! Whats up with that? How are
you possibly supposed to be impartial????"
The email exchange, recounted deep in the Justice
Department's new civil complaint against S&P and its parent, McGraw-Hill, pretty well sums up the entire theory of the
government's case against the rating agency. S&P, according to
the Justice Department, had a choice as the housing market began
to collapse and subprime mortgages began to default. Rating
agency analysts who monitored mortgage-backed securities knew
the crash was coming and warned repeatedly that previous ratings
of mortgage-backed instruments were no longer a reliable gauge.
But rather than heed those warnings and toughen standards on
mortgage-tied instruments, S&P continued to accept fees from
banks in exchange for conferring its blessing on tens of
billions of dollars of collateralized debt obligations. When
truth collided with the client relationships that generated
S&P's revenue, in other words, money won out.
That essential accusation against the credit rating agencies
is not new. Congress, the private plaintiffs' bar and some state attorneys general have been claiming for years that S&P and its
fellow rating agencies, Moody's and Fitch, were more concerned
about their share of the lucrative market for rating structured
finance products than with the reality of their ratings. The
Justice Department's complaint, filed in federal court in Los
Angeles and seeking damages under the Financial Institutions
Reform, Recovery and Enforcement Act of 1989, does add some
juicy new details to what's already been alleged about S&P. As
far back as 2004, according to the complaint, S&P was sculpting
its analytics to satisfy the banks that paid it to rate their
products. When one S&P executive, for instance, asked why a
proposed new system for evaluating mortgage-backed securities
was shelved even though it generated more accurate ratings than
the old model, he was supposedly told that if the new system
"was not going to result in S&P increasing its market share or
gaining more revenue, there was no reason to spend money putting
it in place."
On the other side of the subprime bubble, as defaults began
piling up in the spring of 2007, S&P analysts tracked the
deteriorating performance of mortgage-backed securities rated by
the agency. (One creative fellow, according to the Justice
Department complaint, even devised a dance video on the impending crash, based on "Burning Down the House" by Talking
Heads.) An internal S&P report in June 2007 disclosed that the
lower tranches of the 18,000 subprime mortgage-backed securities
rated by the agency were experiencing double the delinquencies
anticipated in the agency's precrash model. High-level S&P
executives who oversaw CDO ratings were warned that losses in
2006-vintage mortgage-backed securities could top 25 percent.
Yet officials in the CDO rating division did not pass that
information from MBS analysts along to line-level CDO analysts.
In June 2007 alone, S&P rated at least 12 CDOs priced at more
than $27 billion.
And why not? As the complaint explains with cold, hard
numbers, S&P was making good money by disregarding its analysts'
doomsday predictions. Under the issuer-pays model for credit
ratings, S&P and its fellow rating agencies earn fees when they
deliver the ratings issuers want. Issuers, in turn, willingly
chip up those fees because they need AA or AAA ratings to sell
their complex instruments to investors like state and union
pension funds and credit unions, which are frequently restricted
from buying anything but high-rated securities. That symbiosis
worked well for the credit agencies in the subprime bubble, as
long as they supplied high ratings. According to the complaint,
S&P charged issuers between $500,000 and $750,000 for each CDO
it rated, provided that the rating process was completed and the
deal went through. If the issuer withdrew its ratings request
(because, for example, the proposed rating was too low or
another rating agency offered a less stringent evaluation
model), S&P would collect "only a fraction of the rating fee it
would otherwise earn," the complaint said. In 2006 and 2007,
even as the crash loomed, S&P's Global CDO unit supposedly
generated $385 million. The Global Asset-Backed Securities Unit
- which was supposedly rating MBS at such a fast clip that S&P
rating committees spent less than 15 minutes reviewing analyst
evaluations - kicked in another $278 million in revenue in 2006
and $243 million in 2007.
Who loses in the issuer-pays model? You know the answer: all
those investors who kept buying mortgage-referenced CDOs in 2007
because they didn't see the surveillance reports S&P's analysts
prepared on delinquencies in S&P-rated mortgage-backed
securities. Sure, some CDO and MBS buyers were just as savvy as
issuers about the dreck dressed up with AAA ratings. But plenty
were not. They relied on credit rating agencies to live up to
their promise of independence and objectivity. And look where
that got them.
The Justice Department complaint against S&P dispels one of
the only justifications for the issuer-pays system of rating
securities. Supporters of the model contend that it saves
investors the cost of evaluating securities themselves. But
according to the complaint, issuers typically passed on to
investors the rating fees they paid to S&P. Such costs were
considered part of the "organizational and structuring fees and
expenses," which came out of the proceeds of CDO and MBS sales,
according to the Justice Department. So investors were doubly
deceived, not only duped by misleading ratings but also hit for
the cost of their own victimization.
We obviously don't know whether the Justice Department will
ultimately collect anything from S&P, let alone the more than $1
billion it wants, according to The New York Times. S&P has
strongly denied the government's allegations and argues that it
has already spent $400 million to "reinforce the integrity,
independence and performance" of its ratings. S&P's lead lawyer,
Floyd Abrams of Cahill Gordon & Reindel, hinted to CNBC that the
government's suit is retribution for S&P's downgrade of the U.S.
credit rating in 2011.
But based on the allegations in the complaint, we can
certainly say that the issuer-pays model doesn't work. Fallout
from the financial crisis has shown time and again that when
money is at stake, investors cannot rely on systems contingent
upon the honor of financial institutions.
Congress knows the danger of issuer-pays credit ratings. In
the Dodd-Frank Act, it instructed the Securities and Exchange
Commission, which has regulated the credit rating agencies since
2007, to conduct a study on the model and issue a report on its
findings. The SEC report came out in December but didn't get
much notice because the agency took more than 80 pages to come
to no conclusions. My friend Susan Beck at the Am Law Litigation
Daily called the SEC report a "maddeningly meek 82-page document
(that) reads like a memo written by a summer associate hopped up
on Red Bull. There's a lot of stuff in it, but not a single
clear, original thought."
The SEC, or Congress, can do better. There are alternatives
to the issuer-pays model. (The SEC report gives them a thorough
airing, even though it studiously avoids reaching any
conclusions.) Investors could directly share with issuers the
cost of rating securities, or could fund their own independent
credit rating agency. Or perhaps the government or an
independent board could administer credit ratings, either by
allocating assignments on the basis of market share or by
overseeing auctions in which rating agencies compete for
business. The point of all the alternative models is to
recalibrate the incentives of rating agencies so their revenue
doesn't depend on delivering the results issuers want.
As the investment banker told the S&P analyst back in 2007,
it's time for a completely different way of doing biz in the
industry.
Follow us on Twitter: @AlisonFrankel, @ReutersLegal | Like us on Facebook