Executive Summary
Default rates jumped in 2006 and between then and 2014 nearly 9.3 million borrowers were foreclosed on, received a deed in lieu of foreclosure, or short sold their home. To date, nearly a million of these former owners have returned to the market and many more of these “return buyers” or “boomerang buyers” are already qualified, but waiting. Overlays and credit impairment have held a significant number back and could impact thousands more potential return buyers in the coming years. Roughly a third of formerly distressed owners will ever return to the market.
NAR Research analyzed these former owners taking into account multiple factors:
- The time a buyer must wait to be re-eligible for a financing program with timing like the FHA
- The time necessary to repair the distressed seller’s credit
- Whether the distressed seller’s credit profile, at the time of purchase, was unacceptable by historic, sound underwriting standards
- Whether the return buyer would meet credit overlays in the current stringent environment
- The time needed to build down payment for a purchase
- Whether the buyer has the desire to own again
This analysis revealed that the long time to repair credit scores, time to build down payment, and overlapping post-distress factors limit a former owner’s ability to return.
- Since 2006, 950,000 of these former owners likely already purchased a home again
- However, tight conditions in financial markets limit access to 350,000 of these FHA re-purchase eligible borrowers
- An additional 1.5 million return-buyers will likely purchase over the next five years as they become eligible, but overlays will act as headwinds for 140,000.
- As many 260,000 of current and future program eligible borrowers may not return as their former ownership was facilitated by excessively loose lending in the mid-2000s
At the state level,
- California has been the largest benefactor of return buyers followed by Florida.
- Arizona, Nevada and Georgia also made the list as markets like Phoenix, Las Vegas and Atlanta experienced sharp increases in distress among homeowners during the housing downturn.
- Despite the relatively steady housing markets in Texas and solid price growth, the sheer size of the state put it in the top 10.
Over the coming nine years, the states expected to benefit from the trend will remain the same, though some will juxtapose rankings. Florida will nearly catch California, Illinois and Georgia will rise modestly, while Nevada will ease closer to the bottom of the top 10. Virginia will leave the list and be replaced by North Carolina. The shift in the future trend will also reflect a larger share of prime borrowers that were dragged into distressed events as result of price declines and weak employment, rather than risky lending.
Implications
The large number of return buyers coming to the market will continue to play an important role in the market. This demand is in addition to nascent household formation and the normal baseline demand from trade-up buyers. While overlays will hamper some borrowers, those overlays will likely normalize in the future. Mitigating some risk to Federal programs is a stronger regime of regulation on underwriting and the fact that most return buyers are of prime quality. New credit scoring models that utilize rent and utility payments can help shed light on the risk posed by these return buyers. These innovations will improve the propensity of these borrowers to return and gain access, while reducing their risk to the FHA, VA, GSEs, and private mortgage insurers.
The country and housing market are still healing from the collapse of the foreclosure and distress sale wave. As home prices rise and the economy improves, these trends will abate, but there remains a large reserve of former owners who have the desire and ability to return to the market. New credit models and financing opportunities combined with fundamental changes to the mortgage origination process will help to ensure that soundness of the market as these borrowers return.
Full Report: Return Buyers
The Distressed Wave
Foreclosures, deeds-in-lieu of foreclosure, and short sales surged from 2006 to 2014. This wave had a significant impact on the market resulting in falling home prices, transitions from ownership to rentership, loss of equity, crippled credit scores, and damaged communities.[1] However, in time these borrowers will face many of the same realities that first brought them to ownership; a desire to own, stability, better schools, and for the long-term benefit of forced savings relative to rising rents. Lax lending in the mid-2000s resulted in many purchases by buyers who would not have met a historical standard of credit worthiness, but following the decline in home prices and rise in unemployment in the late 2000s, the share of prime borrowers entering default rose dramatically. Thus, an estimate of return buyers must consider several factors:
- The time a buyer must wait to be re-eligible for a financing program
- The time necessary to repair the distressed seller’s credit
- Whether the distressed seller’s credit profile, at the time of purchase, was unacceptable by historic underwriting standards
- The time needed to build down payment for a purchase
- Whether the return buyer would meet credit overlays in the current stringent environment
- Whether the buyer has the desire to own again
Estimating Return Buyers
The incidence of foreclosures increased significantly starting in 2006. Figures for completed foreclosures, deeds-in-lieu of foreclosure, and short sales were derived from data provided by the Hope Now program.[2] This data is subdivided into prime and subprime borrowers. Where data was not available, they were imputed backward based on relationships with foreclosure starts from the MBA’s National Delinquency Survey and short sales from NAR’s RCI survey. To account for non-owner occupants, these figures were adjusted using estimates of 90+ days delinquency for investor-owned, first-lien mortgages from the work of Haughwout, Lee, Tracy, and van der Klaauw[3].
The owner occupied foreclosures, DIL and short sales figures where then separated into year-buckets based on their re-eligibility for FHA financing. As depicted below, the FHA’s waiting period is three years for both foreclosures and short sales, but can be reduced for extenuating circumstances; documented proof of hardship. Survey work by HUD of newly delinquent FHA loans provided an estimate for the share of borrowers who would qualify for extenuating circumstances.[4] This definition changed in August of 2013 to include loss of employment making a significant number of formerly distressed owners eligible for FHA financing.
The FHA programs guidelines require less wait time, lower down payments, lower FICO scores, and a shorter waiting period for potential return buyers.[5] Thus, the FHA program would provide a better avenue for re-entry to the market for these formerly distressed borrowers. However, not all return buyers will utilize the FHA. They may also utilize the VA or extenuating circumstances program at the GSEs that have similar timing. For shorthand we refer to the choice of program with this timing as FHA, though. No attempt is made to correct for those borrowers who pay cash rather than finance their second purchase, but all figures used in this sample are for borrowers who financed their original purchase suggesting a propensity to finance a re-purchase.
As depicted below, the choice of program and eligibility for extenuating circumstances shifts the distribution of re-purchasers significantly.[6] A simple estimate that assumes the use of GSE programs for eligibility and/or excludes extenuating circumstances misses many borrowers who would be eligible earlier and including investors over-estimates the number of owner occupants returning.
A distressed sale has a large impact on a household’s credit standing. The Fair Isaac Corporation reported in 2011 that a short sale could lower a borrowers’ credit score by 50 to 125 points, while a foreclosure to increase that to 85 to 160 points.[7] However, foreclosures and short sales are often a sign of a larger problem as the distressed seller may have issues with other lines of credit such as auto loans and credit cards. The cumulative impact is a larger drop of roughly 170 to 200 points[8] in credit score that takes years to recover. Furthermore, this pattern exhibits persistence and occur for years after a foreclosure. Annual estimates of the share of foreclosed borrowers that recover to pre-event credit score levels were incorporated to adjust the year-buckets by the time for credit recovery. These estimates were based on those derived by Brevoort and Cooper[9] and vary for prime and subprime borrowers. The authors also provide estimates for both a historically normal period and for the recent economic and housing downturn. The recent estimates were used. As depicted above, after 10 years the credit profile of nearly all subprime borrowers had recovered to ex-ante levels, while just roughly 70% of prime borrowers had recovered. For this analysis, a borrower is assumed ready to purchase once they achieve their pre-crisis credit score.
Based on survey work and literature, 82% of prime former owners are assumed to attempt to re-enter the market[10], though some may not. Research by Federal Reserve economist John Krainer suggests that roughly 40% of prime buyers and 10% of subprime borrowers purchase again within 10 years after foreclosure. Based on this relationship, subprime borrowers’ willingness to purchase is assumed to be 20.5%. Estimates of 77% and 21% willingness to purchase for prime and subprime return buyers, respectively, can be derived based on this model and the Krainer results. The lower willingness to buy among subprime borrowers is in line with other researchers[11] who found that the intentions to own among lower income respondents was less predictive of behavior. Note that since the method in this study includes the credit score recovery and programmatic factors incorporated above, the ability to build a down payment and reserve are still endogenous to this estimate of willingness or desire for ownership.
Tight Credit
Even following a programmatic waiting period, not all borrowers could have attained a mortgage in recent years. The average accepted credit scores for borrowers utilizing both FHA and GSE financing were 40 to 60 points above historic norms in recent years and remain elevated. Many lender and aggregators maintain overlays on loans with credit scores below 640, high debt-to-income ratios or other factors. In addition, some borrowers who received credit in the mid-2000s received credit as a result of risk layering and other practices that more recent regulatory changes like the ability to repay rule (ATR/QM) are intended to minimize. However, market dynamics made FHA financing unattractive during the boom and forced some borrowers into risky subprime products and high priced loans. Combined, these conditions create headwinds to re-entry for subprime borrowers and should be accounted for.
In their analysis of the 2012 HMDA data, Federal Reserve economists Canner and Bhhuta[12] analyzed the characteristics of borrowers that originated a purchase or refinance loan in 2006. From these figures, a profile of 2006 vintage borrowers who become 60+ days delinquent within 2 years of origination was developed with four classifications. These classifications are used as proxies for formerly distressed owners:
a) Those with high priced loans and FICO scores less than 640[13] (41%)
b) Those without high priced loans and FICO scores greater than 640 (14%)
c) Those with high priced loans and FICO scores below 640 (23%)
d) Those without high priced loans and FICO above 640 (16%)
The authors use a breakpoint of 620, but simple mathematical averaging was used to reapportion shares to 640 in order to estimate the impact of overlays. A high higher priced loan is one in which the FICO score may be below a certain level or where multiple risk factors are present. Consequently, type D loans, which are not high priced and have higher FICO scores are viewed as most likely to be originated in the recent, tight mortgage environment.
Today, FHA requires manual underwriting for applicants with credit scores below 620 and observed DTI ratios fall dramatically as credit scores move below this level. Rather than increasing pricing to absorb the losses post by risk layer, the “FHA relies on risk-based underwriting to discourage extreme risk layering for higher risk loans while still enabling the use of compensating factors, as appropriate.”[14] The new agency standards and lender behavior limit the risk overlays observed in the mid-2000s that are indicative of higher priced loans with low FICOs. Thus, a relaxation of lending standards to open access to the lower FICO spectrum that do not have layered risk would allow in type C borrowers.
A further relaxation to allow in borrowers with a higher FICO, but multiple risk factors would allow in borrowers with a type B profile. The risk layering may be due to borrower specific traits such as FICO score or DTI or it could be the product used. We assume that the reason for the higher pricing is the riskier loan products and lack of an FHA option.
Finally, type A borrowers, nearly 41% of the delinquent 2006 originations, are deemed too risky to be originated in the current or future environments. The 2006 distribution of risk profiles for sub-prime borrowers was overlain on all cohorts of subprime completed foreclosures, DILs, and short sales in this study. The bulk of delinquent borrowers in years since 2006 have been made up of borrowers from the 2005 to 2008 period[15], prior to the great tightening of credit.
Results
From 2006 to 2014, 9.3 million homeowners were foreclosed on, received a DIL or short sold. Roughly 950,000 former distressed owners have once again become eligible for the FHA or similar financing and restored their credit to pre-distress levels and thus are likely to have bought. An additional 163,000 are both program and credit qualified, but may face overlays in the current tight credit environment. However, lax lending standards during the peak book allowed in 187,000 with multiple risk factors that, though program and credit eligible, would likely not qualify in a market with normalized underwriting standards.
Between 2015 and 2023, an additional 1.63 million former distressed owners will become both program eligible and credit quality, while 140,000 will face headwinds due to their low credit quality, roughly half of whom will not qualify in a normalized underwriting environment.
As depicted below, the bulk of the subprime borrowers have already become eligible for FHA or similar financing again, but they likely face credit overlays to varying degrees based on their pre-distress credit quality. However, nearly 260,000 formerly distressed sellers, depicted in red below, will likely not qualify in the post-crisis mortgage environment following reforms like the ability to repay rule. It’s worth pointing out that the bulk of sub-prime borrowers become eligible for FHA or similar financing again in the early years and are likely already waiting to become credit eligible, while the bulk of prime borrowers will come over the next nine years. This pattern reflects the national mortgage market experience of a subprime bust driven by poor credit quality and risk overlays that drove phantom housing demand. When the subprime market bust and demand contracted, prices fell resulting in a second wave of foreclosures that dragged millions of prime borrowers into distress events.
At the state level, California has been the largest benefactor of return buyers followed by Florida. The sheer size of these markets accounts for part of their positioning, but each was a focal point of the subprime crisis and experienced sharp home price and employment declines. Arizona, Nevada and Georgia also made the list as markets like Phoenix, Las Vegas and Atlanta experienced sharp increases in distress among homeowners during the housing downturn. Michigan and Ohio also saw an uptick in distress owners that will return or have returned, but the subprime trend augmented long term secular issues of weak and transitioning regional economies in these areas. Both trends have rebounded in recent years. Despite the relatively steady housing markets in Texas and solid price growth, the sheer size of the state put it in the top 10.
Over the coming nine years, the states expected to benefit from the trend will remain the same, though some will juxtapose rankings. Florida will nearly catch California, Illinois and Georgia will rise modestly, while Nevada will ease closer to the bottom of the top 10. Virginia exits the top ten to be replaced by North Carolina. The shift in the future trend will reflect a larger share of prime borrowers dragged into distressed events as result of price declines and weak employment rather than risk lending.
Robustness of the Analysis:
According to the Survey of Home Buyers and Sellers from the National Association of Realtors®, 6% and 8% of homebuyers had previously experienced a foreclosure or short sale in 2013 and 2014, respectively. Factoring in both new and existing home markets, this suggests total return purchases of 259,000 and 352,000, respectively, compared to modeled estimates of 254,000 and 316,000 over this time period. Thus, the survey data suggests that the model is robust.
The down payment for the FHA program is just 3.5%, and the VA charges a similar upfront fee, but for some borrowers, this may take significant time to accrue. The GSEs have larger requirements. According to the 2014 Profile of Home Buyers and sellers, 71% of home buyers took 24 months or less to save for their down payment. However, this could vary significantly for distressed sellers depending on employment situation and local debt recourse laws. Consequently, this factor was not incorporated into the analysis. In addition, the high pricing at the FHA may have pushed some program eligible candidates over an affordable DTI ratio, pushing them to the VA or GSEs or limiting their ability to return. New pricing in 2015 would allow in many more borrowers. However, lenders may have overlays specific to formerly distressed sellers that were not accounted for here.
While foreclosure starts and short sale volumes have fallen dramatically in 2014. Furthermore, nearly 6 million homeowners owe more on their mortgage than their home is worth and foreclosures remain hung up in states with judicial processes, so there may be a skew in future volume towards these states. As a result, foreclosures and short sales are likely to remain historically elevated though lower than in the late 2000s. The estimates in this analysis do not attempt to forecast post-2014 foreclosures and short sales.
In this analysis, a borrower is assumed to be credit qualified to repurchase when they reach their pre-distress even credit score level unless they face overlays in the case of a subprime borrowers. However, some prime borrowers may reach an acceptable score earlier. Or a subprime borrower may continue to increase their credit score well beyond their pre-distress level and thus be eligible for re-entry. For example, while roughly 65% of subprime borrowers attain their pre-distress credit score within 3 years of the event, their score could continue to rise in the subsequent years, perhaps even to prime levels, at which point they could avoid overlays re-enter. Likewise, a borrower could reduce their multiple risk factors which led to their high-price loan status on their initial purchase. Furthermore, the FHA program was difficult to utilize in the strong sellers’ market of the mid-2000s. Many borrowers with high LTVs and low credit scores were forced into risky products and higher priced loans. In today’s more balanced market, these borrowers might qualify for non-high priced loans and not fit in the “type A” classification. Conversely, many of these borrowers may have been selected out of the foreclosure, short-sale, and DIL route and into modification programs, leaving a more concentrated group behind. Of course, the buyer could have a cash windfall and purchase without financing. Finally, the behavior of borrowers studied in the work by Brevoort and Cooper may be specific to the 1999 to 2010 time frame of their research.
Improvements in financing programs like the FHA could increase the propensity of formerly distressed sellers to return to the market. In a similar way, credit scoring models that take advantage of rental and utility payments would help to boost the standing of the buyers ahead of the time-frames used in this model and it might also reduce the pricing these borrowers face.
Finally, with the expanding economy, many distressed owners could sell and move to take advantage of better economic opportunities elsewhere. This migration could shift the local of return buying.
Implications
A large number of return buyers have already entered the market, while an even larger group will enter of the coming nine years. The bulk of the coming return buyers will re-enter in the initial five years, but overlays have and will continue to push off a significant number. These return buyers constitute demand that is in addition to nascent household formation and the normal baseline demand from trade-up buyers.
New scoring models can help shed light on the risk posed by these return buyers. Following a foreclosure many former owners rent, pay utility and telecom bills and move. New scoring models like Vantage Score 3.0 and FICO 9 along with FICO most recently proposed, but as of yet unnamed model could help to both improve the propensity of these borrowers to return, while reducing their risk to market insurers.
Since many of these buyers will seek FHA financing, the agency will need to maintain a broad and diversified pool in order to limit the impact of losses, but the bulk of return buyers will have restored their prime credit profile. These buyers have a profile similar to trade-up buyers; further along their lifecycle and likely to prefer single family housing rather than condos or coops.
The country and housing market are still healing from the collapse of the foreclosure and distress sale wave. As home prices rise and the economy improves, these trends will abate, but there remains a large reserve of former owners who have the desire and ability to return to the market. New credit models and financing opportunities combined with fundamental changes to the mortgage origination process will help to ensure that soundness of the market as these borrowers return.
[1] Academic research has also documented a link to physical and psychological impacts. See Bowdler, Quercia, and Smith (2010)
[2] http://www.hopenow.com/industry-data.php
[3] http://www.newyorkfed.org/research/staff_reports/sr514.pdf
[4] http://portal.hud.gov/hudportal/documents/huddoc?id=FHALPT_June2014.pdf
[5] Based on desire to own from Drew (2014); Drew and Herbert (2012) showed that responses for those with familiarity with distressed event were not statistically significantly different from those who did not
[6] “Crude” represents an assumption that all buyers repurchase 7 years after foreclosure completion.
[7] http://www.fico.com/en/blogs/risk-compliance/research-looks-at-how-mortgage-delinquencies-affect-scores/
[8] Brevoort and Cooper (2010) pp. 10
http://www.federalreserve.gov/pubs/feds/2010/201059/201059pap.pdf
[9] http://www.federalreserve.gov/pubs/feds/2010/201059/201059abs.html
[10] For this analysis, 85% of former owners were assumed to want to purchase again. This may be the case or not, but the timing may not match that of program eligibility. Furthermore, not all will want to purchase a home.
[11] Cohen, Lindblad, Paik, and Quercia (2009) http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2458560
[12] Table C.1.A http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2328830
[13] A simple linear estimate was used to reapportion from buckets from 620 to 640
[14] Pp. 45 http://portal.hud.gov/hudportal/documents/huddoc?id=FY2014FHAAnnRep11_17_14.pdf
[15] See Boesel http://www.corelogic.com/research/the-market-pulse/marketpulse__2014-december.pdfpdf and Goodman http://www.urban.org/publications/2000092.html