By Paul Kiel, ProPublica
Why has the administration’s flagship foreclosure prevention program been so ineffective in helping struggling homeowners get loan modifications and stay in their homes? One reason: The government’s supervision of the program has apparently ranged from nonexistent to weak.
Documents obtained by ProPublica — government audit reports of GMAC, the country’s fifth-largest mortgage servicer — provide the first detailed look at the program’s oversight. They show that the company operated with almost no oversight for the program’s first eight months. When auditors did finally conduct a major review more than a year into the program, they found that GMAC had seriously mishandled many loan modifications — miscalculating homeowner income in more than 80 percent of audited cases, for example. Yet, GMAC suffered no penalty. GMAC itself said it hasn’t reversed a single foreclosure as a result of a government audit.
The documents also reveal that government auditors signed off on GMAC loan-modification denials that appear to violate the program’s own rules, calling into question the rigor and competence of the reviews.
Some of the auditors’ mistakes are “appalling,” said Diane Thompson of the National Consumer Law Center, an advocacy group. “It suggests the government isn’t taking the auditing process seriously.”
In a written response to ProPublica questions , a spokeswoman for the Treasury Department, which runs the program, denied there were serious flaws in its oversight system, calling it “effective and unprecedented in many ways.”
The audits of GMAC, though revealing, give only a limited view into the program, because the Treasury has refused to release the documents for other servicers. For more than a year, through a Freedom of Information Act request, ProPublica has sought the audits of 10 of the largest program participants. The Treasury provided only GMAC’s audits, because the company consented to their release. ProPublica continues to seek all of the reports.
Abuses of the foreclosure process, in which banks and mortgage servicers cut corners or even created false documents  to move troubled borrowers out of their homes, have been extensively documented , along with failures by government  to regulate the industry. But the lapses revealed in the documents obtained by ProPublica stand out because they occurred within the government’s main effort to prevent foreclosures, the Home Affordable Modification Program.
Oversight shrouded in secrecy
For HAMP’s first two years, the government offered very little public detail about its oversight efforts. It was virtually impossible for the public — or even Congress — to know how well the banks and mortgage servicers were complying with the government’s effort to prevent struggling homeowners from losing their homes. Those years were crucial, because that’s when servicers evaluated the vast majority of homeowners eligible for a modification — about three million.
The documents obtained by ProPublica show auditors finding serious problems at a major servicer during that time. Instead of publicly revealing the findings, Treasury chose to privately request that GMAC fix the problems.
“For two years, they’ve known how abysmal servicers were performing, and decided to do nothing,” said Neil Barofsky, the former special inspector general for the Troubled Asset Relief Program, better known as TARP or the bank bailout, which provided the money for HAMP.
“It demonstrates that if you have a set of rules for which compliance is completely voluntary and no meaningful consequences for those who violate them, having all the audits and reviews in the world are not going to make a bit of difference,” he continued. “It’s why the program has been a colossal failure.”
Treasury continued to release few details about its audits until June, when it began publishing quarterly reports based on the audits’ results. The public report showed what Treasury called “substantial” problems at four of the 10 largest servicers — Bank of America, JPMorgan Chase, Wells Fargo and Ocwen — and Treasury for the first time  withheld taxpayer subsidies from three of them.
Mortgage servicers that signed up for the program agreed to follow strict guidelines on how to evaluate struggling homeowners seeking reduced mortgage payments. In exchange, the servicers would receive taxpayer subsidies. But as we’ve reported extensively, the largest servicers haven’t abided by the guidelines . Homeowners have often been foreclosed on in the midst of reviews for a modification  or denied because of the servicer’s error. For many homeowners, navigating what was supposed to have been a simple, straightforward program has proven a maddening ordeal .
HAMP has fallen dramatically short of the administration’s initial goal to help three million to four million homeowners. So far, fewer than 800,000 homeowners have received loan modifications through HAMP, or fewer than one in four of those who applied .
As part of the $700 billion bailout program, HAMP was launched in early 2009 with a $50 billion budget to encourage loan modifications by paying subsidies to servicers, investors and homeowners. But only about $1.6 billion has gone out so far .
GMAC said it agreed to release its audits under the program because the company “believes in honoring the spirit of the Freedom of Information Act process” and “elected to be transparent on our work with the [modification] program,” spokeswoman Gina Proia said.
GMAC's parent company has changed its name to Ally Financial, but its mortgage division is still called GMAC. The government owns a majority stake in Ally because it rescued the company with TARP funds, but both the company and the Treasury said that didn’t factor into the company’s decision to allow the documents to be released.
ProPublica contacted all nine servicers that objected to the reports’ release. All either declined to comment on why they wanted the audits kept secret or defended keeping them out of the public domain by saying the reports contained confidential information. Collectively, these companies have so far been paid more than $471 million — dubbed “servicer incentive payments” — through the program. They are eligible for hundreds of millions more. The country’s four largest banks — Bank of America, JPMorgan Chase, Wells Fargo and Citigroup —are also the largest servicers of mortgage loans.
In its written response, Treasury’s spokeswoman said it agreed to withhold the records in part because they could undermine “frank communications between mortgage servicers and compliance examiners” and hurt the program’s effectiveness. The department declined to provide either redacted versions or an index of the documents.
Early reviews “useless” and flawed
Since the program’s beginning, homeowner advocates have wondered where HAMP’s watchdog was  and why it was having so little effect. That watchdog is Freddie Mac, tapped by Treasury in February 2009 and working under a contract worth $116 million and rising. The Freddie Mac unit, now staffed with 121 employees and employing about 150 more through contractors, is supposed to regularly audit servicers in the program to make sure they are following the rules. Treasury is ultimately responsible for deciding whether to punish a servicer but relies on auditors’ findings to make that decision.
It took several months for the unit to even get off the ground. In August 2009, Treasury rejected Freddie Mac’s first reviews of servicers as inadequate  because they were “inconsistent and incomplete” and its staff was “unqualified,” according to a report by the TARP’s special inspector general. Freddie Mac promised to improve. That process took several more months.
As a result, for the program’s crucial first eight months, there was effectively no watchdog. Nationwide, servicers filed to pursue foreclosure on about two million loans during that time.
Treasury disputed the idea that there was no watchdog for those months, saying that auditors had performed “readiness reviews” of servicers as early as May 2009, one month after the program was begun. The documents obtained by ProPublica, however, show that Freddie Mac’s auditing unit, called Making Home Affordable – Compliance (MHA-C), didn’t issue its first report for GMAC until early December 2009 .
That audit was a modest effort that involved collecting a sample of 323 loans handled by GMAC and determining whether they’d been properly reviewed for the program. Because of the delays in starting the reviews, the report was based on a sample of loans that was five months old . Such delays continued into 2010. Another Freddie Mac review, completed at the end of March 2010, was based on GMAC loans selected in October of the previous year .
The delays make those reviews “largely useless to homeowners,” said Thompson of the National Consumer Law Center. If a homeowner lost the house to foreclosure in July, it wouldn’t help to have an auditor notice that several months later, she explained.
The December 2009 audit noted that GMAC might have already foreclosed on loans that auditors had flagged as potentially mishandled, but didn’t order remedial steps. It only requested that GMAC not take “further action.” 
GMAC said it had never reversed a foreclosure action as a result of a HAMP audit. ProPublica put the same question to the other nine servicers that objected to the audits’ release. American Home Mortgage Servicing, the only other servicer that answered the question, said it also had never reversed a foreclosure action due to a HAMP audit.
American Home handles about 384,000 loans , putting it among the 10 largest servicers in the program.
A Treasury spokeswoman said that auditors have reviewed more than 50,000 loan files, but did not directly answer whether a servicer had ever reversed a foreclosure action because of a HAMP audit. Where auditors have found problems, she wrote, the department has “required servicers to take steps to tighten controls” and “re-evaluate any borrowers who may have been potentially impacted.”
In early 2010, around the same time that the auditing unit was issuing its first reports, auditors complained that servicers’ lack of responsiveness to their requests was hampering their efforts. Getting the right documents from servicers was "a cumbersome process," the head of Freddie Mac’s audit team, Paul Heran, said in February 2010 at a mortgage industry conference. It seemed, he added, that servicers often relegated responding to the auditors to low-level staff who didn’t understand the requests. Another manager in the unit, Vic O’Laughlen, said servicers tended to respond with “at best 50 percent of what we’re expecting to see.”
However uncooperative the banks and mortgage services may have been, Freddie Mac’s auditing reports contain errors that call into question their reliability.
Every few months, the auditors examine a sample of the servicer’s loans that have been denied a HAMP modification to check whether the denials were legitimate. In each GMAC report reviewed by ProPublica, auditors found that the servicer had, with very few exceptions, given the homeowner fair and appropriate consideration. But among the justifications listed in the audits are some that violate the program’s rules or simply don’t make sense.
For instance, the December 2009 review said that 35 of the 247 loans that auditors reviewed were denied because the homeowner was “less than 60 days delinquent.”  In the report, auditors said that was the right decision in all but one case. But being less than 60 days delinquent is never on its own a legitimate reason for a servicer to deny a modification, according to the program rules. Homeowners are eligible for a modification even if they’re current on their loans, as long as they can show they’re in imminent danger of defaulting.
Another example: Auditors agreed that GMAC had correctly denied a homeowner because of a failure to sign a trial modification offer by Dec. 31, 2012, HAMP’s end date . That makes no sense, because the review took place in 2009. Treasury’s spokeswoman said this was a typo and that the homeowner was denied for a completely different reason.
There are several other examples in later reports of auditors signing off on denial reasons that have no apparent basis in the program’s rules. For instance, auditors cited “grandfathered foreclosure”  as a legitimate reason for some denials. The spokeswoman said such loans had been in the foreclosure process before GMAC signed up for the program, but the program rules explicitly stated at the time that such loans were eligible.
When ProPublica asked GMAC if it had denied homeowners loan modifications for these reasons, the company said it couldn’t comment because auditors, not GMAC, had generated those descriptions of why homeowners had been denied. In some cases, Proia said, the descriptions were simply wrong: GMAC had never denied homeowners simply because they weren’t 60 days delinquent.
But Treasury defended the questionable denials, and in so doing raised even more questions. For instance, the spokeswoman said HAMP “does not specifically require servicers to evaluate loans that are less than 60 days delinquent.” But Treasury’s official guidance to servicers said such borrowers “must be screened.”
“It makes you wonder if the Treasury even knows the rules for their own program,” said the National Consumer Law Center’s Thompson.
A congressionally appointed panel, among others, has pointed to a fundamental flaw in the way the oversight was carried out: Auditors have had no direct contact with homeowners. The program has been dogged by servicers’ inadequate document systems. Borrowers have long reported  faxing and mailing the same documents over and over, because servicers kept losing them. Servicers have denied about a quarter of all modification applications due to an alleged lack of documentation . Because HAMP’s auditors do not contact borrowers, the auditors have no way of determining whether a denial for inadequate documentation was correct.
In response to this criticism from the Congressional Oversight Panel for the TARP in December 2010 , Treasury said auditors did not contact homeowners to avoid giving them added stress. The panel rejected that reason, saying that contacting borrowers was “critical to assessing the accuracy of a servicer’s determination.”
Instead of talking with borrowers, auditors conduct on-site reviews of mortgage servicing companies, Treasury’s spokeswoman said in her written response to ProPublica. Treasury believes that focusing “on servicer processes and internal controls is the most effective deployment of our compliance efforts,” she wrote.
Detailed audit shows serious problems
It wasn’t until July 2010 — 16 months after HAMP was launched — that the unit performed its first major audit of GMAC. The review included a visit to GMAC’s offices and a detailed review of a sample of loans.
The report enumerated various rules violations, including in GMAC's evaluation of homeowners for modifications. GMAC’s practice was to begin the foreclosure process too quickly : The program required the servicer to give the homeowner 30 days to respond to a trial modification offer, but GMAC’s procedure was to wait only 20.
GMAC’s Proia said no homeowners were “negatively impacted by this issue.”
Auditors also found that GMAC was regularly miscalculating homeowners' income. In a review of 25 loan files of homeowners who had received modifications, the auditors said 21 involved a miscalculation of income . Since income is a key factor in whether a homeowner qualifies for a modification, the high error rate raises obvious questions about whether GMAC was accurately evaluating homeowners’ applications.
Asked about the frequent income miscalculations, GMAC’s Proia said the “issue was identified in the early stages of the program,” that calculating the borrower’s income is a “complicated process,” and that GMAC has improved since the mid-2010 review — an assertion backed up by recent audit results published by the Treasury.
The July 2010 review also found that GMAC had been aware of certain problems such as “incorrect income and expense calculations,”  but had not fixed them. Proia said the company does its best to fix problems when it becomes aware of them.
Penalties: late and weak
Typical of the Treasury’s oversight of the program, GMAC was never penalized for any of the rules violations. For the first two years of the program, Treasury officials publicly threatened servicers with possible penalties but instead followed a cooperative approach . When auditors found problems, servicers were asked to fix them.
The documents illustrate as much. In response to the auditors’ findings, GMAC was required to develop an “action plan.” GMAC refused to provide the action plan to ProPublica and recommended seeking it and similar documents by filing a Freedom of Information Act request with the Treasury.
Treasury has sent mixed messages about its ability to penalize banks over the course of the program , threatening “monetary penalties and sanctions” in late 2009, then saying it lacked the power to enforce such penalties. Treasury finally departed from its cooperative approach in June, when it withheld incentive payments  from three of the top 10 servicers. (GMAC was not among them.) The companies would not receive the public subsidies for completing modifications until they made certain changes. The companies were cited for some of the same problems for which auditors had criticized GMAC, such as regularly miscalculating borrowers' income. JPMorgan Chase, for instance, had erred in estimating income in about a third of the homeowner loan files reviewed.
The punishment hasn’t had much sting. Incentive payments were restored for two of the three companies when Treasury’s most recent report  declared they’d improved. Only Chase and Bank of America, the country’s largest servicer, would continue to have their incentives withheld, Treasury said.
But while those incentives have slowed, they have not stopped, according to Treasury’s monthly TARP reports . Since June, when Treasury first announced it would withhold incentives, Bank of America has received $2.5 million in taxpayer incentives. While that’s a steep reduction from the roughly $7.5 million it had been receiving monthly, the bank is supposed to get nothing. Chase received $404,000 during that same time.
Treasury responded that it has programs to encourage modifications on both first and second mortgages, and that the payments Bank of America and Chase received were related to second mortgages. “Current system limitations” meant the Treasury couldn’t withhold these payments, according to the Treasury spokeswoman. Treasury is working to fix the problem, she said.