A Standard & Poor’s analyst in 2004 sent an e-mail to executives at the rating company’s structured- finance group. It had lost a job to Moody’s rating a mortgage- backed security because S&P criteria were more demanding, and something had to be done, the analyst allegedly wrote.
“This is so significant that it could have an impact on future deals,” the analyst said, according to this week’s U.S. Justice Department’s Feb. 4 complaint. “There’s no way we can get back on this one, but we need to address this now in preparation for future deals.”
The May 2004 message is one of the earliest cited in the government lawsuit in which McGraw-Hill Cos.’s S&P is accused of letting competition for business take over its credit ratings calculations for securities backed by residential mortgages.
In this week’s complaint, the government sets out meetings, messages and memos that it says show S&P analysts assigning investment-grade ratings to securities based more on a desire to win business than to be accurate.
The Justice Department is seeking as much as $5 billion in damages from New York-based S&P. Its ratings allegedly helped create the U.S. housing bubble, whose bursting led to the worst financial crisis since the Great Depression.
“Put simply, this alleged conduct is egregious, and it goes to the very heart of the recent financial crisis,” U.S. Attorney General Eric Holder said at a Feb. 5 press conference.
The company from September 2004 to October 2007 issued ratings on more than $2.8 trillion worth of residential mortgage-backed securities and about $1.2 trillion worth of collateralized-debt obligations, according to the complaint.
S&P is paid by issuers that pick the rating company to rate their securities. Fees reached $150,000 for rating a residential mortgage-backed security and $750,000 for a so-called synthetic collateralized-debt obligation, in which the underlying collateral is credit derivatives, including credit-default swaps, according to the complaint.
S&P’s understating of risks allowed the banks to issue the securities with less credit protection, making them more profitable, according to the Justice Department.
S&P said in a Feb. 5 statement that the Justice Department’s lawsuit is without merit.
“There was robust internal debate within S&P about how a rapidly deteriorating housing market might affect the CDOs, and we applied the collective judgment of our committee-based system in good faith,” S&P said in the statement, referring to collateralized debt obligations. “The e-mail excerpts cherry- picked by DOJ have been taken out of context, are contradicted by other evidence, and do not reflect our culture, integrity or how we do business.”
Every CDO cited by the Justice Department in the complaint received the same rating from another company, S&P said.
The Justice Department’s allegations, which include both identified and unidentified executives, point to an April 2004 meeting as the earliest evidence that business considerations were influencing S&P’s decisions on rating criteria.
A draft document was circulated that described a process to change rating criteria. Included would be “market insight,” “rating implications,” and the polling of investors and investment bankers for a “full 360-market perspective,” according to the complaint.
An unidentified executive objected, asking in an e-mail to colleagues and analysts, “Are you implying that we might actually reject or stifle ‘superior analytics’ for market considerations?” the government said. The objection was ignored, and he never got a response to his e-mail, it said.
In July 2004, the government said, Joanne Rose, S&P’s executive managing director of structured finance, and Thomas Gillis, the head of structured finance research and criteria, circulated a memo saying managers who work with clients should be consulted to get “market perspective” on rating criteria.
S&P that year delayed updating a software program, called the Loan Evaluation & Estimate of Loss System, used to analyze the risks of proposed mortgage-backed securities, by not including data of 642,000 loans that included riskier mortgages, according to the complaint.
The more accurate updated program would have required issuers of the securities with relatively risky Alt-A or subprime loans to provide more loss-coverage protection to win investment-grade ratings, the U.S. said.
That would have made the securities less profitable for the issuers, according to the complaint.
Instead of the proposed update, S&P “slightly” modified the existing version in a way that didn’t significantly increase loss coverage, the U.S. said.
The same executive who objected to using “market considerations” in ratings was told by colleagues in structured finance that the proposed update wouldn’t increase market share or revenue and there was “no reason to spend money” putting it in place, according to the complaint.
At the same time, S&P continued to use an “inaccurate default matrix” in evaluating the credit risks of collateralized debt obligations, because the changes needed for accuracy didn’t meet market-share goals, the U.S. said.
In 2005, when S&P released an updated CDO evaluation program called E3, it also created a less-stringent version called E3 Low to rate certain synthetic CDOs because of the potential negative business impact, the government said. Analysts were told to use E3 Low if E3 didn’t produce the desired rating, it said.
In 2006, when delinquencies started to rise in nonprime mortgages, S&P analysts decided to develop new criteria for watching mortgage-backed securities S&P had already rated, the Justice Department said. At a meeting in late summer or early fall that year, the analyst group decided to select criteria that gave them the desired result: fewer and less-severe downgrades, it said.
An analyst objected to the idea that the “ends justify the means,” the government said. The analyst was no longer invited to meetings of the group, the U.S. said.
By the fall of 2006, mortgages underlying securities issued that year were preforming so poorly that analysts thought the data contained typographical errors, according to the complaint. To see realized losses in the first six to 10 months of a 30- year loan was unprecedented, according to the complaint.
One unidentified executive regularly complained to colleagues from the fall of 2006 to the spring 2007 that she was prevented by other executives from cutting S&P ratings “because of concern that S&P’s ratings business would be affected if there were severe downgrades,” according to the complaint.
“This market is a wildly spinning top which is going to end badly,” David Tesher, managing director of S&P’s cash CDO group, is quoted as saying in a confidential working note on Dec. 11, 2006.
S&P analysts recommended in February 2007 that 50 so-called tranches of mortgage-based securities be placed on negative credit watch, or internal watch, the government said. The CDOs that were exposed to those tranches would generally receive a lower credit rating as well, according to the complaint.
S&P personnel agreed to place only 18 tranches on negative credit watch, according to the complaint.
The company continued to issue credit ratings for CDOs that it knew didn’t accurately reflect the risk because they failed to account for the risk of the underlying nonprime-mortgage- backed securities tranches, the U.S. said.
On June 28, 2007, S&P gave a AAA rating to $2.8 billion of Ridgeway Court Funding II Ltd., a CDO linked to subprime mortgages.
The CDO defaulted in January the following year, resulting in almost total losses for the investors, according to the complaint.
“There was a lot of internal pressure in S&P to downgrade lots of deals earlier on before this thing started blowing up,” an S&P analyst said in a July 5, 2007, e-mail to an investment banker client. “But the leadership was concerned of p*ssing off too many clients and jumping the gun ahead of Fitch and Moody’s.”
The case is U.S. v. McGraw-Hill, 13-00779, U.S. District Court, Central District of California (Los Angeles).
To contact the editor responsible for this story: Michael Hytha at firstname.lastname@example.org.