Freddie Mac relied on a botched loan review process for key data used to justify its landmark settlement with Bank of America Corp., the Federal Housing Finance Agency's inspector general has concluded.
The alleged problems stem from how Freddie Mac officials evaluated the underwriting of interest only and option adjustable-rate mortgages underwritten by Countrywide, now a part of B of A. Despite warnings that Freddie was failing to account for the wave of defaults that would likely follow the expiry of the mortgages' teaser rates, Freddie Mac went ahead with a settlement and the FHFA approved a $1.35 billion deal.
In response to the inspector general's inquiry, the report says, Freddie Mac and the FHFA have agreed to hold off on such settlements until they have reconsidered how it selects mortgages for review.
In its written response to the report, FHFA defended the settlement agreement and disputed any "inferences" that the deal might have been a raw one for the Treasury Department.
"After months of review regarding this particular transaction, FHFA has not changed its view that the settlement reached in late December was appropriate and reasonable," the agency wrote.
Beyond the size of the settlement, however, the report raises questions about oversight of Freddie and the robustness of its efforts to pursue compensation for bad loans. Freddie's internal auditors considered the government-sponsored enterprise's internal controls to be unsatisfactory as recently as June, the inspector general revealed, and the mortgage review process described in the report struck some industry observers as obviously flawed. Even when Freddie was apprised of the problems, the report finds, the GSE resisted correcting them out of concern for alienating Bank of America.
The basis of the inspector general's concern over Freddie Mac's practices is the GSE's tendency to subject only those loans which defaulted within the first two years to a high level of scrutiny. This was a longstanding practice in the mortgage industry, based on the assumption that, if a loan had been underwritten shoddily, it would default within a fairly short period of time. If the loan made it past a few years, however, it was a good bet that the borrower's fortunes – and not underwriting defects -- were to blame.
"That used to be truly the industry standard," says Jeff Naimon, an attorney for BuckleySandler LLP. "In the private market years ago, the reps and warrants generally terminated in two years."
But around 2005, changes in the mortgage market rendered suspect the assumption that bad loans would surface within 24 months. The reason was that many mortgages sold during the later years of the housing boom were option ARMs or interest only loans which permitted borrowers to pay only a fraction of what an equivalent fixed-rate mortgage payment would require for several years or more. By staving off the day when borrowers would have to make their full mortgage payment, these loan features allowed struggling borrowers to hold on for much longer.
Early last year, an FHFA examiner began looking at how this might affect the GSE's review process for defaulted loans. As one might expect, he found that delaying the date at which borrowers had to make a full mortgage payment also tended to delay the date at which they defaulted – most doomed option ARMs took at least three years to default, according to Freddie Mac's own data.
Given that Freddie tended to pass on inspecting defaulted loans that aged more than two years, this was a problem.
The examiner warned higher-ups that the oversight could potentially cost Freddie billions of dollars. As many as 100,000 loans were never given proper review, the inspector general's report states.
The examiner warned his colleagues and superiors of the oversight, the IG alleged.
"By not taking a good look at these defaulted [interest only and alternative-A] loans over the next 2-3 years …. Freddie Mac could be passively absorbing billions of dollars of losses," the examiner wrote in a Sept. 15, 2010 memo. "Since the savings in credit losses would dwarf the incremental expenses incurred in reviewing additional loan files, the fundamental question that Freddie Mac and FHFA should be addressing is this: How may of the ineligible loans sold to Freddie Mac in the 2005-2007 origination years should Freddie Mac accept the loss on?"
Multiple levels of both Freddie's leadership and the FHFA as a whole failed to address the examiner's concern, the inspector general concluded.
"One senior manager told FHFA-OIG that he never read the memorandum because he had never opened the email attachment containing it," the report states. "Two managers … acknowledged that they had reviewed the memorandum, but they did not remember that the issue could potentially involve substantial losses to Freddie Mac."
Other FHFA officials had similar concerns, the inspector general's report states, but they were ignored as well. So were internal auditors who raised concerns that the sampling of loans in Freddie's portfolio was insufficient.
Despite the internal concerns, in December Freddie Mac agreed to a tentative settlement of repurchase claims against Bank of America that failed the address the full scope of the examiner's concern, the inspector general's report says.
In part, the IG alleged, Freddie was eager to approve the deal so as to maintain good relations with Bank of America.
"Management made a deliberate decision not to consider changes to our sampling procedures," a Dec. 14 internal memorandum stated, citing the need to "maintain relationships."
"In other words, Freddie Mac management asserted that the need to maintain relationships with loan sellers such as Bank of America was a factor weighing against implementing more expansive loan review and repurchase policies," the inspector general's report states.
Many of the concerns in the IG report appear to stem from a single vocal examiner. But a case can be made, says Buckley Sandler's Naimon, that by the time of the Bank of America settlement, most bad loans should have defaulted. "The loans which were really poorly underwritten, many of them probably didn't make it through the first six months."
Likewise, in its response to the inspector general's report, the FHFA denied the premise that option ARM loans and mortgages with teaser rates should be subjected to a lengthier period of scrutiny than traditional mortgages.
"Mortgage defaults do not equate to a basis for repurchase requests, but they may be a reason to examine a loan for possible contractual violations," the agency specified. "This is not about the riskiness of loans but about contractual violations at the time of the loan origination."
FHFA acknowledged that there were areas where it could make improvements in its examination process, however. According to the Inspector General's report, it and Freddie Mac have agreed to refrain from entering into or approving similar settlements until Freddie has thoroughly reconsidered its sampling methodology for putbacks.
