John Smith's four-bedroom house stands tall on Cleveland's East Side, its tidy cream siding and green lawn oblivious to the devastation that has scarred the surrounding neighborhood. It is everything thousands of foreclosed homes in the area are not: occupied, intact, and still an asset to the family that lives in it. Smith purchased the home in 2005 through a nonprofit dedicated to repopulating the city with working- and middle-class homeowners. But Smith's investment was one of a few drops in a bucket with no bottom: The census tract he lives in, with about 600 homes, has seen more than 200 foreclosure sales in the past 15 years. "We know the inner city is the inner city, and it's no big surprise to us," Smith says. But even he is stunned at the level of devastation around him. "Whole neighborhoods are almost totally down."
The Smith family has come close to the precipice. In 2008, Smith, then 38, lost his job as a financial analyst for a bank and had trouble paying his mortgage. His credit-card debt shot into six figures. When his lender initiated foreclosure proceedings that December, the notice was a month late because the bank did not have the correct address on file for its own borrower.
Smith could have ended up like another Cleveland homeowner, truck driver Larry Saulsberry. In 2004, Saulsberry got a letter from a mortgage broker, offering to help him extract cash from the home he had owned since 1972. Saulsberry refinanced his old mortgage, on which he owed just $16,000, for a new one at $88,000 and higher interest rates, intending to use the proceeds to buy a new truck and bankroll badly needed repairs on his 90-year-old home. "Desperate people," Saulsberry explains, "do desperate things." Only after he had signed the papers did Saulsberry discover that he had paid his broker nearly $10,000 in fees that had not been previously disclosed and that his payments to Argent Mortgage were $1,100 a month, not the promised $800 -- and what's more, the interest rate would adjust in the future. In 2008, the county sheriff sold his beloved home.
A twist of fate saved Smith. After receiving the foreclosure notice, he called his lender, which sent him to Fannie Mae, which would only say that the mortgage was held by a "private investor." A month later, Fannie finally gave him a number in Durham, North Carolina. On the other end of the phone was a counselor from Self-Help, the nonprofit organization that unbeknownst to Smith -- and nearly 52,000 borrowers like him -- had adopted his mortgage under a program that bought affordable home loans from banks under strict guidelines Self-Help had created. Self-Help shepherded Smith through a loan modification.
Smith is employed again, but far from liberated. The house that he bought in 2005 for $218,900 is now worth half that, putting Smith among the nearly one in four of all mortgage borrowers who owe more than their home is worth. It's no accident that Smith has been able to weather the storm of the foreclosure crisis while similarly indebted neighbors succumbed. Thanks to Self-Help, his mortgage had a fixed interest rate, where as Saulsberry's was adjustable. Modest closing costs were clearly disclosed. And Smith obtained his loan from a bank, not a broker. The features of Smith's loan were standard on mortgages sponsored by Self-Help.
Self-Help was founded in 1984. In 1998, it teamed up with Fannie Mae to make custom-built safe loans available to high-risk borrowers from coast to coast. The Ford Foundation made the project possible through a $50 million grant that Self-Help used to guarantee the mortgages. The partnership sought not just to help the relative handful of borrowers the organization could aid through its purchases of affordable mortgages from its bank partners but to transform the lending business just when subprime lending was surging as a malign alternative. Self-Help demonstrated that it was a sound business proposition to design mortgages that vulnerable borrowers -- many already struggling with high debts and low incomes -- could actually pay.
Since then, with additional funding from Ford, a team of researchers at the University of North Carolina has tracked a sample of borrowers whose loans were purchased from their original lenders by Self-Help. The 3,700 Self-Help borrowers under the microscope look very much like subprime America. Four in 10 are black, Latino, or Asian, and most have annual incomes between $20,000 and $50,000; the typical borrower earns just 62 percent of his area's median income.
At first glance, the Self-Help loans look rather like the toxic mortgages that devoured John Smith's neighborhood. The median borrower took out a loan for 97 percent of the purchase price, leaving home buyers heavily leveraged. A near majority had to pay more than 38 percent of their income toward debts. Many had questionable credit ratings.
But if Self-Help mortgages appeared similar to subprimes in their loose qualifying standards for borrowers, the mortgages themselves couldn't have been more different. Self-Help mortgages carry fixed interest rates, and they do not include the toxic features that sabotaged many subprime borrowers, such as penalties for refinancing. What's more, borrowers' income and assets were fully documented. Risks were based on fact, not fantasy.
Thirteen years later, in the rubble of the foreclosure crisis, the Self-Help experiment makes it possible to answer the great "what if": What if the borrowers who succumbed to subprime mortgages had received soundly structured loans instead? Tracking borrowers in real time, first during the housing bubble and then through the bust, the research has given the world unique insight into why some homeowners were equipped to weather the credit bubble while others quickly went bust.
The answer, quite simply, is that borrowers who had fixed-rate loans without hidden costs were far likelier to hold on to their homes. Subprime borrowers with adjustable-rate mortgages have cumulatively faced a "serious delinquency" rate approaching 40 percent, meaning that four out of 10 borrowers ended up at least three months late on their mortgages. From there, it's usually very hard for them to avoid foreclosure. By contrast, only 8.5 percent of Self-Help borrowers -- in the same cities and with the same financial profiles -- have fallen into such deep trouble, and fewer than 5 percent have ended up losing their homes to foreclosure.
"In the statistical analysis, we were able to isolate the risks of the mortgages from the risks of the borrowers. What we learned is that mortgage products can amplify the risks or minimize the risks," says Roberto Quercia, principal investigator and director of the Center for Community Capital at the University of North Carolina. "The right product is crucial."
The experiment has much to offer the Obama administration as it rebuilds a broken system of housing finance, in the midst of ongoing partisan warfare over what caused the mortgage market to collapse in the first place. Blaming the Community Reinvestment Act -- the 1977 law that calls on federally regulated lenders to meet the credit needs of entire communities -- remains popular blood sport among Republicans, pandering to racist stereotypes about minority homebuyers. The facts are poised to speak more loudly, if anyone chooses to listen.
The researchers and their sponsors hope Congress and the administration will pay attention to their findings: Without a forceful commitment from those rebuilding the credit system, borrowers like John Smith -- who have everything it takes to pay a mortgage except a pot of gold in the bank -- may be shut out of owning property in the future.
The idea that homeownership could be a force to better low-income people's lives drove the Ford Foundation to call the University of North Carolina's Michael Stegman in 1998, asking him to lead the research on the results of Ford's investment in affordable homeownership. The solution, it seemed to all involved, would have to include a big role for Fannie Mae and Freddie Mac, which, along with government-run Ginnie Mae, bankrolled more than half of the entire mortgage market. "The idea was to just take the existing machinery, making it work better and close the gap," says Janneke Ratcliffe, executive director of the Center for Community Capital.
In Durham, Self-Help Community Credit Union had already been financing loan products for high-risk borrowers, both through its own lending and through purchases of batches of other banks' loans. But it was impossible for Self-Help to sell these mortgages to Fannie or Freddie. That obstacle kept this market at a tiny scale, while millions of eligible borrowers remained unserved. To Self-Help CEO Martin Eakes, it seemed absurd. His borrowers were making their payments on time, and his credit union was making money. Why wouldn't Fannie and Freddie want a piece of the action? With his early results and Ford's support, Eakes convinced Fannie Mae to commit to a national-scale loan-financing program, in which it would sell $2 billion in Self-Help-backed mortgages to investors, and Ford's money would absorb any losses.
The project sought to outdo the subprime market -- engineering and disseminating a better product that would eventually force its way mainstream. It was reasonable to expect that Fannie and Freddie -- operating under a mandate from Congress to finance loans for low- and moderate-income borrowers -- would embrace the kind of innovative lending Self-Help was pioneering. "We hoped that the rigorous analysis of the performance of the mortgages over time would reveal that the actual credit risk associated with these mortgages was significantly less than the risks perceived by the lending industry," says Frank DeGiovanni, director of financial assets at the Ford Foundation. "The ultimate goal of the demonstration, beyond assisting 40,000 low-income households to become homeowners, was to change the way that the financial industry approached the provision of mortgage financing for low-income and minority households." At the time the project started, just four in 10 African American and Latino households owned their homes, in contrast to seven in 10 white households.
But this promising experiment, though immensely helpful to homeowners like John Smith, has been swamped by the wider damage of subprime. Fannie and Freddie are now wards of the federal government -- largely because they took the low road to reaching underserved borrowers, following Wall Street's subprime-lending stampede. (Fannie stopped buying Self-Help mortgages in 2009.) While the crisis thwarted the Self-Help program's original goal, however, the research on its prime loans for subprime borrowers continues to pour out -- and may be more important than ever. The scholars possess data that shows that the popular wisdom on subprime lending was wrong. The problem wasn't that the borrowers couldn't afford to be homebuyers; it was that they had been set up to fail.
What's more, low-income homebuyers realized meaningful benefits. As real-estate prices soared, owners who bought their homes before 2005 -- before the entire market veered into unreality -- saw sharp rises in their equity that have not been entirely wiped out. Low-income buyers were much less burdened by the expenses of home maintenance than earlier assumed. The research has also contradicted the standard ways that lenders label low-income borrowers "high risk": It turns out that credit ratings are a poor predictor of which low-income homeowners will catch up on their mortgage debt after late payments and which are doomed to lose their homes.
We now know that it's possible to lend to high-risk, low-income borrowers with minimal chance of default. Yet the kinds of loans that help those borrowers succeed are precisely those that are now most endangered. The North Carolina team found that just three perverse loan features overwhelmingly determined whether a borrower would default: a loan is sold by a mortgage broker; has adjustable interest rates; or imposes prepayment penalties on borrowers who seek to sell or refinance.
The 2010 Dodd-Frank Act has drained some of the poison: Mortgage brokers now can't be incentivized to sell harmful loans, and adjustable-rate mortgages now must be based on borrowers' ability to pay the real interest rate, not a teaser. But adjustable rates and prepayment penalties are not going anywhere; more likely, they will become more prevalent than ever.
"It may be more difficult for many Americans to afford the traditional pre-payable, 30-year fixed-rate mortgage," warns the Obama administration's blueprint for housing finance reform. Historically, U.S. borrowers could get into long-term, fixed-rate mortgages cheaply and get out of them easily, because an implied government guarantee effectively absorbed the resulting risk -- that fluctuating interest rates over time will reduce the value of those mortgages to investors as they're packaged into securities. For mortgage products lacking a government guarantee, Wall Street came to rely on prepayment penalties and adjustable rates to deflect the risk back onto borrowers -- and as we now know from the North Carolina research, these features were the most harmful to consumers.
In the current spasms of reform, Fannie Mae and Freddie Mac could well disappear, shuttering the institutions that for more than seven decades sponsored the widespread availability of fixed-rate, long-term mortgages. Even if some form of government guarantee remains, home mortgages are likely to become more expensive, and beyond the reach of low-wealth renters. Proposed guidelines under Dodd-Frank are likely to establish a 20 percent down payment for most loans financed through mortgage-backed securities -- excluding millions of would-be homebuyers who can't scrounge together tens of thousands in cash up front. Even the Federal Housing Administration insurance program, historically a low down-payment option for upwardly mobile working-class buyers, is now requiring down payments of at least 10 percent for those with less than sterling credit.
Such rules are built not on statistical evidence but on political jockeying and on broad-stroke formulas to protect banks and investors. Quite likely, they overreach. And no one has yet figured out what it would take to ensure loans remain available to the millions of households whose credit ratings decayed in the recession. "There's no information out there, and that's startling," says Ratcliffe of the Center for Community Capital. "The fact that people are making these big policy decisions without research is really problematic."
The U.S. may soon lose a generation of homeowners, as households who would have once qualified to buy a home get locked out. Stegman, who now heads policy for the MacArthur Foundation, is comforted that the research he instigated isn't going away. "What's really exciting about it is that the longer it goes, the more it's going to be able to inform policy," he reflects. "Ultimately, when the political environment becomes less poisonous and evidence begins to matter again, then this becomes a very powerful voice."