A massive piece of legislation — the Housing and Economic Recovery Act of 2008 — was signed into law by President Bush on July 30. Some economists, such as Paul Krugman, argue the legislation doesn’t go far enough in addressing the problems caused by the subprime mortgage crisis, especially those caused by imprudent and loose lending standards. Others argue it goes too far and provides inappropriate aid to borrowers who should have known better than to assume so much credit.
One of the most contentious provisions funds the reduction of mortgage debt owed by some borrowers facing foreclosure. The plan offers borrowers the chance to obtain a new loan with the Federal Housing Administration equal to 90 percent of the current market value of the house. Lenders who participate agree to accept the proceeds of this loan as complete payment for the original mortgages owed by these borrowers and accept the differences as losses. FHA will receive $300 billion in additional lending authority to make these new loans. What the legislation does not do is offer relief to the many other homeowners who have suffered substantial declines in the values of their properties but continue to pay their mortgages.
Although the full details of the implementation are not yet available, we used the FHA Fact Sheet and some educated guesses about how the program might work and how the aid might be distributed. With this information and some data analysis, we obtained a sense of how the declines in house values and the new relief will be distributed among a group of homeowners. As we like to do, we focused on a particular market — a neighborhood in the Bronx — and made use of the extraordinary data underlying the Cyberhomes website and other public record data.
Case study of ZIP code 10461 in the Bronx
Barbara’s side of the family hails from a neighborhood in the Bronx in ZIP code 10461. Her grandparents moved to this area, near Morris Park and Pelham Parkway, in the mid-1920s and raised 10 children in a strong community of Italian immigrants. We visited the home on several occasions before it was sold a few years ago.
Bronx 10461 has a median home value of just under $500,000, which is slightly above the median for the rest of the Bronx. Like many parts of the country, this market has had substantially slower housing sales, and lower prices, in the past couple of years. Sales are running well below 25 per month, which was the typical number of sales per month in 2005 through early 2007. Median annual household income among the 20,000 or so households in this area is $64,000. Most of the houses, including the home of Barbara’s grandparents, were built in the early part of the 20th century, prior to the Great Depression.
We analyzed a subset of single-family residential properties purchased in 2005 through 2007 that have not sold again. The average purchase price for these 215 sales was about $450,000, which is close to the average of all sales in this ZIP code. The average loan-to-value ratio at the time of purchase was around 80 percent in each year; however, 20 percent of the purchases had 100 percent loan-to-value ratios at the time of purchase.
We sought to measure whether the homebuyers increased their equity or wealth positions — assessing the “pain” of the recent fall in house prices. We examined both the absolute dollar amount of the change in equity and the change in the equity relative to the value of their homes. The latter is captured by changes in the loan-to-value ratio and incorporates all mortgages obtained by the property owners. Current property values are as of June 2008 and equal the valuation estimates used by Cyberhomes.
A unique and tedious part of this study involved tracking the additional mortgages associated with these properties since the purchase date. A particularly tricky part was deciding whether a mortgage was a second lien or replaced the original lien, since we did not have complete information. We made some conservative judgments and have a pretty good handle on this issue. We also ignored amortization, which we cannot observe, but suspect this is minor since the loans are less than four years old and amortization would be relatively small for a typical 30-year mortgage. We were unable to measure whether the additional debt was used to make improvements to the property and increase its value.
Our efforts produced what we consider to be rather stunning results. The value of homeowner equity in these properties has declined by nearly half from $18 million to $9 million. Current loan-to-value ratios now average about 90 percent, versus 80 percent at the time of purchase. Eighty-six of these owners, about one third, now have negative equity. We label these results as stunning because home ownership is typically considered to be a wealth-building activity, whereas for many within this group of owners it has been just the opposite. Additionally, negative equity is one of the strongest predictors of foreclosure.
Searching for the root causes
Our next step was an attempt to identify the main causes of the changes in equity, because we think one’s assessment of the fairness of the relief plan may depend upon the specific causes of the equity declines.
- Lower house prices. One obvious potential source of reduced equity is the decline in house prices during this period. In fact, house prices did decline for most of the homebuyers in the past 18 months, and the decline has left many owners with a less valuable asset than when it was purchased. The decline in house prices explains about $4 billion of the $9 billion reduction in equity. There is, however, much variation in this effect. On average, house prices declined by about $6,000, but more than a third experienced declines of more than $25,000 since the time of purchase.
- Additional debt. Another potential contributor to the decline in equity is the additional debt taken on by some homeowners through second mortgages and cash-out refinances. We estimated this additional mortgage debt to be the primary cause of the decline in equity. In total, 311 loans are secured by these properties; 56 of these loans originated after the purchase of the home. The additional loans total more than $5 billion in additional mortgage debt.
- Timing of the purchase. Those who bought relatively recently were more likely to have negative equity because house prices generally began their plunge here in the spring of 2007. This is a little like the old rule of hiring: last hired, first fired. The analogy here is that those who came to the market late were the ones most adversely affected by the declines in house prices since early 2007. Earlier buyers had the chance to build up some equity during the house price rises of 2005 through 2006.
- Price range of the house. The bulk of the lost equity is disproportionately concentrated in the middle price range of $400,000 to $550,000 — the lost equity in this range was almost 2.5 times the share of houses in this category. The two extreme house price categories incurred a disproportionately low portion of the losses. In other words, the damage of lost equity was focused on those in the heart of the distribution of house prices for this ZIP code.
Potential impact of the relief
Our goal was to approximate the total amount of the relief, the beneficiaries of the relief and some of their characteristics. This exercise sheds light on the magnitude of the relief effort embedded in the legislation and offers some insights about its perceived fairness.
We assume that relief is limited to those with negative equity who obtain a new loan equal to 90 percent of the current value of the property but no more than $550,000.
These assumptions generate a group of 95 homebuyers who qualify for relief. The aggregate amount of relief will be about $9 million, or about $93,000 per owner. This relief amounts to just over 10 percent of the total mortgage debt outstanding for these owners that banks would have to forgive or write down. The FHA program would at the same time guarantee 95 new loans totaling about $37 million.
What are the characteristics of those receiving the bulk of this relief? First, those who took out high loan-to-value loans are both common in this ZIP code and prominent among those being provided relief. Second, those who took out additional debt are also disproportionately represented. Third, those who bought relatively recently also receive a disproportionate share of the relief, about 37 percent.
Some of the implementation plans are still being developed, and some criteria are not in our database. For example, only owner-occupants will be eligible; we were unable to incorporate this factor. Neither could we tell whether borrowers intentionally missed making their scheduled mortgage payments nor whether the borrowers will be able to afford the 10 percent down payment for the new FHA loan, both of which may be criteria for obtaining relief. Another key issue we cannot determine is whether lenders will go along with the relief plans for each borrower. We suspect that lender participation may be particularly difficult for those borrowers who have both a first and a second lien since second lien holders will be less likely to accept a plan that probably involves a complete write-off of the second lien. For all of these reasons, our estimates likely overstate the amount of the relief, but we suspect our estimates do capture key elements of those who will receive relief.
Is the relief plan fair?
In our view, there is plenty of blame to go around in the subprime mortgage crisis, and the “mix” of blame probably varies widely among markets and market participants. We also believe that the relief program just passed will help stem the negative influence of the foreclosure problem on nearby property owners and provide relief to people most deeply affected by the subprime crisis.
Our analysis does raise a question about whether the relief plan is as fair as it can be. On the one hand, relief is targeted to those who experienced the largest house price declines and who happened to get on the house price boom train a little late, which seems fair. On the other hand, the plan also benefits those who purchased houses with little or no down payment and who chose to withdraw substantial amounts of equity in the form of new debt. Perhaps this is fair if these folks were inappropriately coaxed into these new mortgage arrangements by aggressive lenders. We don’t have enough information to make this call.
We do offer a suggestion to those developing the final implementation process that can be summed up in two words: foreclosure forensics. Yes, we are fans of CSI and Bones, but the kinds of forensics we have in mind are a bit different.
We encourage the review process to include two components. First, examine the original loan documents and process to see if the loans being forgiven violated predatory and fair lending laws. If they did, consider ways to hold the lender responsible for its mistakes. Second, examine the borrowers’ non-housing assets. We suspect these other assets are small and illiquid for most of the borrowers. For borrowers with substantial assets who extracted equity to purchase them, however, the other assets should be at least partly accessible to the lender or the government. If these kinds of steps are taken, the final distribution of the relief aid will be closer to what most folks would describe as fair.