Nov. 12: This post was updated to include details of documents ProPublica obtained after the story's publication.

The Obama administration launched its main program to prevent foreclosures in the spring of 2009 with $50 billion and abundant promises. What the Home Affordable Modification Program, or HAMP, lacked — and wouldn’t have for years — was effective oversight of the big banks that were crucial to the program’s success.

Documents obtained by ProPublica shed new light on this failing in 2009 and 2010, when the foreclosure crisis was at its peak and six million American homeowners were in danger of losing their homes. HAMP required mortgage servicers to offer loan modifications to eligible homeowners so that their monthly payments would be lower. The servicers — the largest of which were owned by the banks that had fueled the crisis in the first place — were in charge of reviewing homeowner applications, but the government set the rules and was supposed to supervise their work.

But the documents show that the government did not complete a major audit of the two largest banks in the program, Bank of America and Wells Fargo, until over a year after the program launched.

Such audits were rare at the other large mortgage servicers throughout 2009 and 2010, according to the documents. During these years, when the government provided little oversight and administered no sanctions, servicers reviewed 2.7 million modification applications and denied two-thirds of them. Meanwhile, homeowners regularly complained they had been mistreated by servicers in the program.

The documents also show how the Treasury Department coddled servicers that weren’t complying with the program’s rules. Once a year, servicers are required to certify that they are complying with the program’s rules. But servicers define for themselves what it means to comply. A company that admits violating the rules is allowed to merely submit a cover letter with their certification stating the exceptions and how it would fix the problems.

For instance, on September 28, 2010, PNC Mortgage submitted its certification. Along with that certification, it disclosed five “instances of noncompliance” or “gaps,” as it called them, along with its plans to address the issues (“steps”).

 

Like all of the documents ProPublica received, PNC’s letters are heavily redacted, so the nature of their “gaps” and “steps” remains secret.

The Treasury defended the oversight of HAMP. “The robust nature of our compliance program, together with the steps we have taken to strengthen protections for homeowners under the program, are critical reasons why homeowners who enter HAMP today show a strong likelihood of long-term success to avoid foreclosure,” said a Treasury spokeswoman.

Prying loose the documents

The documents were hard to obtain. They came as a result of two Freedom of Information Act requests by ProPublica. Initially, the Treasury Department, which administers HAMP, refused to release any documents at all. It was only after ProPublica appealed the denial that Treasury agreed to release some documents, although with large portions blacked out.

One of our requests has dragged on for more than two years, and even after all that time, the department continues to withhold certain documents, though it says it intends to turn over more. (See here for a full index of the documents we’ve obtained so far. If we receive more, we will add them to our collection when we receive them.)

Update: On Nov. 9, ProPublica received another batch of documents. Treasury said it was the final response to our requests. We've updated our index to reflect the new documents.

In some cases, the Treasury even withheld the documents of servicers who never objected to their release. When ProPublica informed the Treasury that certain servicers had said they had no objection to releasing the documents, the Treasury finally turned them over.

“It seems like they’re trying to prevent the information from getting out,” said Rick Blum, coordinator of the Sunshine in Government Initiative, a coalition of media groups. FOIA protects business trade secrets from being divulged, but Blum doubts whether that exception is being fairly applied in this case. The documents mainly concern how servicers were breaking HAMP’s rules, he noted, and “a screw-up is not a trade secret.”

A Treasury Department spokeswoman said HAMP has brought “an unprecedented level of transparency” to the mortgage servicing industry and servicers are “subject to an unprecedented level of compliance oversight.”

That’s damning by faint praise, say consumer advocates. “The reason that the level of transparency and accountability is ‘unprecedented’ is because no one has ever held servicers to account. Just because you have something where before there was nothing, that doesn’t mean that something works or is effective,” said Diane Thompson of the National Consumer Law Center.

Audits slow and rare

By now, HAMP’s disappointing performance is well known. The program was launched with President Obama’s promise to help three to four million homeowners avoid foreclosure. Three and a half years later, the program is only approaching 1.1 million modifications. It’s spent just $4 billion of its original $50 billion budget.

A recent study found a big reason for the program’s failure was that, despite all its rules, it didn’t change the behavior of the biggest banks. The banks did a poor job of modifying loans before HAMP was launched and weren't much better after.

To oversee the program, the Treasury awarded Freddie Mac a $209 million contract to be the program’s watchdog. Freddie Mac formed a group, called Making Home Affordable – Compliance, or MHA-C for short, but it got off to an inauspicious start. In August 2009, Treasury rejected its first reviews of servicers as inadequate because they were “inconsistent and incomplete” and its staff was “unqualified.”

The Treasury has refused to turn over MHA-C’s audit reports — with one exception. The servicer GMAC Mortgage expressly consented to the release of the documents. Those audits, we reported last year, revealed that the servicer had seriously mishandled many loan modifications. MHA-C’s audits themselves contained numerous errors, calling into question the competence of the reviews.

The Treasury didn’t dispute the fact that no major audits of the biggest banks were completed until well after HAMP’s launch. But the spokeswoman said “it is important to note that Treasury began unprecedented reviews of servicer compliance with program directives within the first months of program implementation.” Those earlier “compliance activities” included “on-site reviews” and “sampling of homeowner loan file reviews,” she said.

But the GMAC audits show how cursory those earlier reviews could be. In December 2009, MHA-C reviewed a sample of files, but when it reported its findings to GMAC, it told the servicer that the report was “being provided for informative purposes only, and no response is required from you at this time.” GMAC itself was not the subject of a major audit until July 2010. It was never penalized.

In the new batch of documents, the government has kept secret the audits themselves. But ProPublica has obtained the servicers’ written responses to the audits (see here for an index of the documents). The Treasury scrubbed or withheld almost all of the responses’ substantive content. Even so, they reveal some basic facts.

The watchdog was very slow to conduct major audits of the biggest servicers. MHA-C’s first major audit of Bank of America, the largest servicer in the program, wasn’t completed until July of 2010, more than a year after HAMP launched. The first major audit of Wells Fargo was completed in August of 2010.

By July of 2010, Wells and Bank of America had already denied about 430,000 homeowners a HAMP modification.

Even for big banks that received an audit sooner, oversight was infrequent, the documents show. JPMorgan Chase, the other mortgage servicing titan, received a major audit within HAMP’s first year, but through all of 2009 and 2010, it only responded to two major audits. CitiMortgage, the fourth largest servicer, only received three major audits in that time period.

When Treasury did conduct a major audit of one of the big banks, it often reviewed files that were many months old. A Wells Fargo audit delivered in March of 2011, for example, covered a “review period” of “May/June 2010.”

Once the audits were finished and delivered to the servicers, there was another delay: servicers had a month to respond with “action plans with implementation dates” to address the problems.

A Bank of America audit in late November of 2010 was based on information that was six months old. One month later, Bank of America replied: “In the Report, you requested that Bank of America management respond to the observations, and if we agree, provide a detailed remediation plan, and if we disagree, provide a detailed explanation and evidence to support our position." A page of redacted text follows, so the substance of the bank’s response remains secret.

Bank of America spokesman Rick Simon declined to discuss the communications: “As part of the MHA compliance reporting, servicers may provide information and statements that are of a proprietary and confidential nature, with the full expectation that the Department of Treasury and its agents will treat it appropriately.”

Banks evaluate themselves

In September 2010, the robo-signing scandal hit. A number of the nation’s biggest banks announced they were halting foreclosures to investigate whether they had submitted false filings to courts. The revelations drew immediate attention to mortgage servicers’ failings and eventually led to action by bank regulators and state and federal law enforcement.

Yet the same month that the scandal erupted, mortgage servicers submitted their first annual certification to the Treasury Department that they were complying with HAMP’s rules.

The certifications were toothless. In fact, following a fox-guarding-the-henhouse model, servicers could certify that they were complying even when they were not.

It was up to the servicer to decide whether it was in “material compliance,” according to the certification form. What rose to the level of being a material problem? A Treasury directive gave guidance that is so vague it borders on no guidance at all: “This evaluation of materiality may or may not be quantifiable in monetary terms and should include, but is not limited to, consideration of the nature and frequency of noncompliance as well as qualitative considerations, including the impact on Program goals and objectives.”

If the servicer found that it was, by its own definition, noncompliant, it was required to list the problems and its “action plan” in a separate “cover letter” to be sent with the certification filing. But that was it. There was no penalty.

The Treasury continues to withhold many of the cover letters from ProPublica. Among the documents the Treasury is still keeping secret are the letters for the four largest servicers in the program (Bank of America, Wells Fargo, JPMorgan Chase, and Citibank), although the Treasury has produced their certifications.

Update: On Nov. 9, ProPublica received the cover letters for Bank of America, Wells, Chase, and Citi, among other servicers. All four letters disclosed the banks were not in compliance with HAMP's rules, but Treasury redacted the details.

Still, the documents that ProPublica has received show that several servicers claimed they had no problems to report.

“None to the best of our knowledge,” wrote a GMAC Mortgage executive.

There are reasons to question that statement. Two months earlier, a MHA-C audit had found a number of problems at the servicer, including that it had miscalculated homeowner income in more than 80 percent of audited cases. And that same month, September 2010, GMAC had kicked off the robo-signing scandal by halting its foreclosures across the country.

GMAC spokeswoman Susan Fitzpatrick said the company has “rigorous internal and external controls in place that maintain consistent compliance with HAMP guidelines” and that the servicer was materially compliant with HAMP’s rules when it filed the certification. “The foreclosure issues identified in September 2010 are not related to servicer compliance with HAMP guidelines.”

The Treasury said the certifications were “only one part of a more comprehensive process” that included its audits. “While Treasury uses these certifications as part of the compliance process, certainly we do not rely solely on servicers to self-identify and report their weaknesses,” said the department’s spokeswoman. Servicers were allowed to define materiality for themselves in the certifications because, she said, a “‘one size fits all’ approach would not have been practical.”

“All instances of non-compliance are tracked and pursued to ensure that servicers have and are executing against remediation plans, and that any potentially-affected homeowners are identified and re-evaluated if applicable,” she said.

Penalties were late and fleeting

It was only in the wake of the robo-signing scandal that Treasury decided to take punitive action against servicers breaking HAMP’s rules. In June 2011, it withheld the program’s subsidies to the three largest servicers in the program. HAMP provides incentive payments to servicers, as well as borrowers and investors, to encourage modifications. The servicer subsidies would stop until the banks showed “substantial improvement,” Treasury said.

Wells Fargo soon received a passing grade, but the Treasury continued to withhold subsidies from Bank of America and JPMorgan Chase until the payments were restored as part of the February 2012, $25 billion national mortgage settlement between federal agencies, 49 states, and the five largest servicers. Together, the servicing subsidiaries of Chase and Bank of America have received about $509 million in subsidies through the program.

When asked for comment about HAMP’s limited oversight, the three largest servicers in the program all pointed to Treasury’s recent assessments of how servicers comply with the program’s rules. Those assessments show the banks requiring only “moderate improvement.”