Real Estate Research provides analysis of topical research and current issues in the fields of housing and real estate economics. Authors for the blog include the Atlanta Fed's Jessica Dill, Kristopher Gerardi, Carl Hudson, and analysts, as well as the Boston Fed's Christopher Foote and Paul Willen.
Comments are moderated and will not appear until the moderator has approved them.
Please submit appropriate comments. Inappropriate comments include content that is abusive, harassing, or threatening; obscene, vulgar, or profane; an attack of a personal nature; or overtly political.
In addition, no off-topic remarks or spam is permitted.
July 2, 2020
An Update on Forbearance Trends
So far, the principal policy response to the COVID-19 pandemic in the U.S. mortgage market has been forbearance. On March 18, 2020, the Federal Housing Finance Agency (FHFA) directed the government-sponsored enterprises (GSE) Fannie Mae and Freddie Mac to suspend foreclosures on single-family mortgages and start offering forbearance and modification plans to distressed borrowers.1
The Coronavirus Aid, Relief, and Economic Security (CARES) Act, signed into law on March 27, codified the Fannie Mae and Freddie Mac forbearance programs, and included the government housing agency Ginnie Mae. The act directed the agencies to grant borrower requests for forbearance "with no additional documentation required other than the borrower's attestation to a financial hardship caused by the COVID-19 emergency..." In general, servicers of loans not covered by the law have set up similar forbearance programs.
What is mortgage forbearance?
Mortgage forbearance means a mortgage lender (that is, the holder of credit risk) allows a borrower to stop making mortgage payments for a fixed period of time. During that time, the lender does not charge late fees nor initiate foreclosure proceedings but does consider the borrower delinquent on the loan. Under normal circumstances, the lender would report the delinquency and the forbearance plan to credit bureaus. Under not-so-normal circumstances, including natural disasters, lenders often do not report loan status to credit bureaus. The CARES Act explicitly stipulates that forbearance resulting from the COVID-19 pandemic cannot negatively affect a borrower's credit score, so lenders cannot report borrowers in forbearance as being delinquent.
Once the forbearance period expires, the borrower has to make up the missed payments. For example, if the lender implements a six-month forbearance policy, then borrowers would be allowed to defer all payments for up to six months. After the forbearance period ends, the borrower and lender would typically negotiate a repayment plan that makes the lender whole in the long run but does not send the borrower back into distress. During this COVID-19 forbearance episode, Fannie, Freddie, and Ginnie Mae must provide borrowers affordable repayment options, such as converting the arrears into a partial claim, which is a non-interest-bearing second lien due at termination of the loan.
Forbearance is meant to provide short-term relief to financially distressed borrowers without inducing moral hazard by borrowers who do not need assistance. Since no debt is forgiven, the policy, in theory, should not appeal to borrowers who are not liquidity-constrained. Nevertheless, some empirical evidence indicates that some borrowers may be willing to engage in this "strategic forbearance."
While an acceptable exit strategy exists for borrowers, it is contingent on employment and income being restored before the forbearance period ends. For mortgage servicers, the magnitude of forbearance take-up is crucial to their liquidity—especially for nonbank servicers.
A liquidity squeeze arises because mortgage servicers must advance scheduled principal and interest payments to investors regardless of whether the borrower is paying. In addition, servicers must also pay tax and insurance payments.2 Either the borrower or guarantor will eventually reimburse the servicer, but in the short run, servicers must have sufficient liquidity to be able to bridge this gap. Such a squeeze is especially acute for nonbank servicers who not only are relatively thinly capitalized but also do not qualify for liquidity support programs that have been set up for banks. While this was a concern early in the pandemic, the GSEs and Ginne Mae have implemented polices that have helped to alleviate liquidity concerns of nonbank servicers.3
Forbearance take-up rate
In March, estimates of the fraction of households that would use forbearance in the coming months varied widely. On the one hand, optimists believed forbearance take-up would be fairly low because the CARES Act called for a generous increase in unemployment insurance benefits, which should provide many unemployed households with enough income to continue making payments. For example, FHFA director Mark Calabria estimated that only about 2 million GSE borrowers—a take-up rate of less than 5 percent—would seek forbearance.4
On the other hand, pessimists believed the numbers would be significantly higher. Laurie Goodman of the Urban Institute predicted that close to 12 percent, or 5.75 million borrowers, would request forbearance, and Mark Zandi of Moody's Analytics predicted that around 15 million households, or roughly 30 percent of mortgage borrowers, would miss payments.
To gauge the extent of forbearance, the Mortgage Bankers Association (MBA) recently initiated a weekly Forbearance and Call Volume Survey of its members. The survey provides a lagged picture of forbearance rates. The latest survey covers more than three quarters of first-lien mortgages, so we believe it is representative of the overall market.
The earliest data from the survey indicate that as of March 8, 2020, the forbearance rate on all loans was below 1 percent, with both Ginnie Mae and Fannie/Freddie loans having forbearance rates of less than one quarter of 1 percent. Forbearance rates rose in the month of April, according to the survey (see the chart). On April 26, overall forbearance rates were at 7.55 percent, up by more than 480 basis points since the beginning of April. The rate of increase slowed in May, and the most recent survey (June 21) shows that the overall take-up percentage fell from the previous two weeks, going from 8.55 percent to 8.47 percent and marking the first decrease in the survey.
The data also indicate that there is significant heterogeneity across market segments. The rate for Ginnie Mae loans, 11.83 percent, has been flat for four weeks, while the rate for Fannie and Freddie loans, 6.26 percent, has fallen over the past three weeks. The rate for other loans—including those in private label securitizations (not government guaranteed) and held on portfolio—is 10.07 percent, which is essentially flat for the month of June. For Ginnie loans, this means an increase of more than 1,000 basis point since March 8.
Overall, the actual take-up rates are squarely between the optimistic and pessimistic forecasts.
The forbearance take-up rates remain elevated, but recently, they have been flat or have fallen. This pattern is consistent with what we've seen in the labor markets—the rate of new unemployment claims has been fallen while the number of unemployed continues to be elevated. The generous increase in unemployment insurance benefits in the CARES Act certainly has helped take-up rates to be lower than the pessimistic forecasts. But the threat that forbearance will transition to foreclosure has regained power because the number of COVID-19 infections is increasing and the CARES Act unemployment insurance benefits will expire at the end of July.
1 [go back] You can find the official press release on the FHFA website. On the same day, the Department of Housing and Urban Development (HUD), in consultation with the federal government, implemented a 60-day moratorium on foreclosures and evictions for single-family homeowners with FHA-insured mortgages. In addition, HUD encouraged FHA mortgage servicers to offer various loss-mitigation options to distressed borrowers. The options include short- and long-term forbearance options and mortgage modifications.
2 [go back] Servicers of agency loans are required to maintain first-lien status for the GSEs. This implies ensuring that the borrower has paid all property taxes, whether an escrow account exists or not, and that, in certain states, the borrower has met homeowner association commitments.
3 [go back] On April 10, Ginnie Mae set up a lending facility that allows servicers experiencing liquidity issues to borrow funds to make forbearance-related principal-and interest advances. On April 21, the FHFA announced that it would limit the requirements for servicers of Fannie and Freddie loans to forward principal-and-interest payments to investors to four months of forbearance.
March 31, 2010
Anti-foreclosure policy and aggregate house price indexes
A new paper by researchers at the New York Fed and New York University argues that the Federal Housing Authority (FHA), the government's insurer of relatively high-risk loans, is seriously understating the amount of risk in its portfolio. The paper makes a number of different points, but we want to comment on one claim in particular that has policy relevance beyond the issue of FHA risk. In fact, if this claim is correct, then any policy designed to reduce foreclosures by eliminating negative home equity could face significant problems when put into effect.
Repeat sales indexes a poor predictor of individual home price
The specific issue we want to address is how well an aggregate house price index can predict the price of an individual home. A number of aggregate indexes measure average house prices for a particular area, from the national level to the ZIP-code level. Often, these indexes are based on repeat sales, meaning that they combine the price changes of individual homes over time. If a house sold for $200,000 in 1997 and $220,000 in 2001, this repeat sale provides a data point indicating that house prices rose by 10 percent from 1997 to 2001. It is true that the 2001 buyer might have gotten a great deal in that the house really should have sold for more than $220,000 at the second sale. However, the assumption is that the influence of good and bad deals washes out when data from many repeat sales are aggregated together. If they do, then researchers can infer the average, overall path of house prices.
The problem occurs, the authors of the paper say, when one uses the resulting aggregate index to predict the price of an individual home. Consider someone who purchased a home for $200,000 in 2007. Now assume that over the next two years the aggregate house price index for that particular area declined by 10 percent. The authors point out that the decline in the index does not necessarily mean that this particular homeowner would have sold the house for $180,000 in 2009. The owner may have taken extremely poor care of the house, or a beautiful park that was across the street from the house at the time of purchase may have become a strip mall. In either case, the homeowner was likely to have sold for less than $180,000. On the other hand, the homeowner may have made some improvements to the home that would have resulted in a sale of more than $180,000.
Research paper provides careful analysis of valuation errors in aggregate indexes
Potential problems with repeat-sales indexes were well known before the FHA paper was written. What the new paper contributes is a careful analysis of how large these so-called valuation errors can be and how they might relate to the probability of having negative equity. Using residential sales from Los Angeles County, the authors compare the actual sales prices of houses with predictions generated by different aggregate price indexes. The authors make two important findings.
First, the repeat-sales indexes are often biased, in the sense that the mean of the predictions does not match the mean of the recorded prices. For 2008 and 2009, repeat-sales indexes tended to overpredict house prices by 7 to 18 percent. In 2007, the indexes underpredicted house prices by about 4 percent. Second, dispersion in individual valuation errors is large—the standard deviation of valuation errors is about 20 to 25 percent, depending on the aggregate index used. Putting these two facts together gives a clear message: Using standard methods, it is difficult to predict what any individual house will sell for at any particular time.
Valuation errors undermine mortgage balance reduction policies
On a general level, this observation is not an indictment of the FHA, since a lot of other people also use aggregate indexes to infer prices of individual homes—including us. Moreover, without knowing the ins and outs of the FHA's default-prediction model, it is hard to know the quantitative importance of valuation errors in the calculation of FHA risk. But moving beyond this issue, it is not hard to see how large valuation errors could undermine the effectiveness of a policy that attempted to ease foreclosures by reducing mortgage balances for individual negative-equity homeowners. As we have blogged recently, some observers have claimed that many, if not most, foreclosures occur because owners with large amounts of negative equity simply walk away from their homes. The ostensible policy implication is to reduce these homeowners' mortgage balances to give them more of an incentive to stay.
If valuation errors are large, however, it is very difficult to know who has severe negative equity and who doesn't. This problem undermines the effectiveness of balance-reduction policies. Effective policymakers must know how to price individual homes to assess the depth of negative equity for those homes. Consider two homes that, according to an aggregate price index, have 30 percent negative equity. That amount may or may not be severe enough to get an owner thinking about walking away. If it is, then an appropriate policy might reduce both homes' mortgage balances by 20 percent, thereby reducing the negative equity to about 10 percent. (Leaving a little bit of negative equity is probably a good idea in practice because it may prevent homeowners from selling the moment that a balance reduction is made.)
Effective foreclosure prevention would consider both job loss and negative equity
If in reality one of these homes actually has 50 percent negative equity and the other has 10 percent negative equity, then the balance-reduction policy is likely to prevent no foreclosures. The owner with 50 percent negative equity remains underwater, to the tune of 30 percent, so is probably still thinking about walking away—according to the theory of default that motivated the balance-reduction policy in the first place. On the other hand, the owner with 10 percent negative equity was not going to walk, unless perhaps a job loss went along with the negative equity. But if the combination of job loss and negative equity is the real problem in the housing market rather than severe amounts of negative equity alone then we can devise much more cost-effective policies to reduce foreclosures than large-scale balance reductions.
The authors of the paper do not discuss balance reductions. However, in other papers, they argue that anti-foreclosure policy should consider balance reductions. We believe that the valuation-error results uncovered in the FHA paper indicate that balance-reduction policies face substantial hurdles in actual practice.
Real Estate Research Search
- Affordable housing goals
- Credit conditions
- Expansion of mortgage credit
- Federal Housing Authority
- Financial crisis
- Foreclosure contagion
- Foreclosure laws
- Governmentsponsored enterprises
- Homebuyer tax credit
- House price indexes
- Household formations
- Housing boom
- Housing crisis
- Housing demand
- Housing prices
- Income segregation
- Individual Development Account
- Loan modifications
- Monetary policy
- Mortgage crisis
- Mortgage default
- Mortgage interest tax deduction
- Mortgage supply
- Multifamily housing
- Negative equity
- Positive demand shock
- Positive externalities
- Rental homes
- Subprime MBS
- Subprime mortgages
- Supply elasticity
- Upward mobility
- Urban growth