Editor’s note: This series shines a light on the "shadow inventory" of distressed properties that haven’t yet reached the market. This article, Part 1, explores the various definitions and projections for shadow inventory, and Part 2 ("Skeptics don’t expect REO flood") details the potential impact of this inventory on the housing market as a whole.
There are lingering fears that the fragile recovery under way in many housing markets could be derailed by a "shadow inventory" of homes that have either been repossessed by lenders, are in foreclosure, or whose owners are behind on their loans.
An estimated 7 million homeowners are currently behind on their loans or in the foreclosure process — roughly double the number two years ago.
Not all of those homes will end up on back on the market. Some homeowners will "cure" their delinquent loans through a loan modification or by refinancing.
But lenders are also managing real estate-owned (REO) inventories of between 775,000 and 1 million homes that they will need to liquidate. Only 25-30 percent of those homes are believed to be on the market.
Analysts who track the performance of home loans for investors in mortgage-backed securities are sounding alarm bells about the impact the "shadow inventory" of distressed loans and REO properties could have on markets.
But others who are keeping close tabs on the situation say fears that a sudden wave of foreclosed properties will drive down prices are overblown.
Some loans that analysts have counted as shadow inventory will actually be saved from foreclosure, they say, and those that aren’t will hit the market over an extended period of time and largely be absorbed by pent-up demand and growth in household formation.
One positive sign is that while a staggering number of borrowers are behind on their loans or in foreclosure, the growth in their ranks may finally be slowing as the number of good loans going bad begins to decline.
What alarms analysts who crunch numbers for investors in mortgage-backed securities: There is a steady growth in the number of loans that are delinquent or in the foreclosure process, suggesting that lenders have been playing a game of "extend and pretend" and putting off the inevitable day of reckoning.
The most recent data from Lender Processing Services (LPS) put the number of non-current home loans at 6.89 million at the end of February, including a record 1.79 million homes in the foreclosure process.
While 30- and 60-day delinquencies remained relatively stable throughout 2008 and 2009, LPS data showed considerable growth in foreclosures and "seriously delinquent" loans — borrowers behind on their payments by 90 days or more.
The Mortgage Bankers Association’s last National Delinquency Survey painted a similar story, with an estimated 7.33 million loans either delinquent or in foreclosure at the end of the fourth quarter.
The big question is how many of the loans that are currently delinquent will move into the foreclosure process, and how many loans that begin the foreclosure process will ultimately be liquidated.
There’s no dispute that there’s a huge number of homes in limbo, but predicting how many of them will end up on the market — and when — involves considerable guesswork.
In a paper published in September, analysts with Amherst Securities crunched numbers from the MBA’s second-quarter 2009 National Delinquency Survey to peg the shadow inventory at 6.94 million homes.
The widely cited paper, "Housing Overhang/Shadow Inventory = Big Problem," assumed very low "cure rates" among delinquent borrowers. Amherst analysts concluded that only 626,000 of the 7.57 million loans the MBA survey implied were delinquent or in foreclosure at the time would escape liquidation.
More recently, the National Association of Realtors crunched numbers from the latest MBA delinquency survey and other sources to produce a shadow inventory estimate of roughly 2.5 million homes.
Yet another recent report by analysts at Standard & Poor’s Financial Services LLC — which analyzed only "private label" mortgage-backed securities not guaranteed by Fannie Mae and Freddie Mac — put the shadow inventory at 1.75 million homes.
That report — "The Shadow Inventory of Troubled Mortgages Could Undo U.S. Housing Price Gains" — covered less than one-third of mortgages bundled into securities for sale to investors.
Defining ‘shadow inventory’
Estimates of the size of the shadow inventory "overhang" vary widely, depending both on how the term is defined and how far ahead those making the estimates are attempting to look.
When Zillow surveyed homeowners last year and found that 8 percent would be "very likely" to sell if local conditions improved, the survey results were reported by some as evidence of a "shadow inventory" of 10 million homes. …CONTINUED
Given that homes are selling at a rate of about 5 million a year, such sentiments could have an impact on housing markets, even if all of those homes did not come onto the market at the same time. Unlike lenders selling REO properties, many homeowners entering the market as sellers would also be looking to buy another home.
Shadow inventory was once widely understood to mean REO properties — homes repossessed by lenders — that had not yet been offered up for sale via a multiple listing service or other means. By that definition, the growth in shadow inventory has been significant but gradual.
Data collected by LPS shows REO inventories experiencing rapid growth in 2008, cooling in 2009, and then flattening in recent months at just over 1 million homes. That suggests lenders have been moving homes off of their property rolls at about the same pace they’ve been adding them.
NAR’s estimate of shadow inventory assumes 75 percent of the more than 1 million homes identified as REOs by LPS in January — about 770,000 homes — were not yet on the market, and therefore part of the "shadow inventory."
But it’s become common for analysts to factor in homes whose owners have fallen behind on their payments or entered the foreclosure process into their shadow inventory estimates.
NAR’s estimate of the shadow inventory, detailed in a recent paper by research economist Selma Lewis, also took a stab at predicting how many loans that were seriously delinquent or in the foreclosure process would end up as distressed sales or REOs.
Lewis took the "roll rates" at which LPS has observed loans moving through various stages of delinquency and into foreclosure, applying those rates to the numbers from the latest MBA delinquency survey.
Those calculations suggested that an additional 3.2 million homes might be considered shadow inventory candidates because they were in the foreclosure process in 2009 or could be assumed to have entered foreclosure in January and February.
But NAR estimates that distressed sales account for nearly one in four existing-home sales, so Lewis assumed that 766,708 of those 3.2 million properties were already on the market, and should therefore not be counted as "shadow inventory."
Lewis also assumed that 75 percent of the trial and permanent loan modifications under way during January under the auspices of the Home Affordable Modification Program (HAMP) will ultimately succeed — an assumption that led her to subtract another 710,051 homes from her shadow inventory estimate.
The resulting NAR estimate of a "shadow inventory" of 2.49 million homes is considerably more optimistic than Amherst’s projection of 7.57 million loans for several reasons.
The Amherst estimate doesn’t account for the fact that many distressed properties are currently listed for sale, and largely dismisses the possibility that homeowners will avoid foreclosure through loan modifications.
In a paper outlining their findings, Amherst analyst Laurie Goodman and her team noted that after a year, the redefault rate on non-HAMP loan modifications has historically been about 70 percent.
Writing at a time when the HAMP program was still picking up steam, Goodman and her team concluded that even if the success rate on HAMP modifications is much higher than 30 percent, the program couldn’t be relied on to divert many more than 1 million distressed properties from the "shadow inventory" of homes.
A spokesman for Amherst provided a copy of the paper to Inman News but Goodman and her team declined a request for comment.
In an e-mail exchange, Lewis acknowledged that her estimate of a 75 percent success rate for HAMP loan modifications was probably optimistic. Treasury officials have said they assume about 40 percent of HAMP loan mods will redefault.
But the biggest difference in the approaches taken by NAR and Amherst is how far into the future they attempt to look.
NAR’s application of LPS "roll rates" simply updates foreclosure statistics gathered by the MBA by two months, by estimating the number of loans that were delinquent by 60 days or more at the end of the year that can be assumed to have moved into foreclosure in January and February.
The Amherst analysis, on the other hand, attempts to determine the ultimate fate of all loans identified by the MBA survey as delinquent or in foreclosure, through the use of "cure rates."
Those assumptions are stark. Amherst analysts assumed that none of the 2.4 million loans believed to be in the foreclosure process at the time would avoid liquidation, and that only 1 percent of the 2.17 million borrowers behind on their payments by 90 days or more would avoid that fate.
The "cure rates" — the share of loans brought current by loan mods or other means — for loans overdue by 60 days and 30 days were also assumed to be dismal. Less than 5 percent of borrowers who’d missed two monthly payments were expected to keep their homes, and nearly three out of four homeowners who’d missed just one payment were written off as doomed.
Lewis said that when she asked Amherst analysts how they calculated their cure rates, she was told they were based on the percentage of borrowers who prepaid their loans or became more — or less — delinquent. The goal, Lewis was told, was to generate a snapshot of today and apply it through time.
Because Lewis did not know the specifics of the portfolio of mortgage-backed securities analyzed by Amherst, "their answer wasn’t sufficient for me," she said. …CONTINUED
During the boom, Amherst’s analysis showed cure rates were much higher — above 80 percent for borrowers who’d missed one payment, and nearly 70 percent for homeowners who’d missed two payments.
As late as early 2006, rising home prices helped even those who were behind on their payments by 90 days or more to "cure" at rates approaching 40 percent, Amherst’s analysis showed.
Diane Westerback, the leader of a team of analysts at Standard & Poor’s, said she agreed with Amherst’s more pessimistic outlook for cure rates going forward.
"Those earlier cure rates were in rising housing markets, where there were multiple opportunities to … get out from under" a problem loan, Westerback said. "It used to be that a large proportion of people in the 90-day bucket could cure."
It would take a significant drop in unemployment — back to around 5 percent, where it was during the boom — to see an improvement in cure rates, she said.
Westerback’s team estimated in February that a shadow inventory of 1.75 million homes is lurking in pools of "private label" mortgage-backed securities.
She said that while there’s considerable political and legal pressure on lenders to engage in workouts, so far such efforts have not been an overwhelming success. Like Amherst, Standard & Poor’s analysis assumes a 70 percent redefault rate on loan modifications.
If nothing else, what reports like those from Amherst, NAR and Standard & Poor’s illustrate is that while lenders seem to have capped growth in their REO property inventories, there’s a growing number of loans that are either in foreclosure or default.
The evidence suggests that at least part of the growth is due to the longer period of time it takes for seriously delinquent loans — those with payments three months or more in arrears — to become REO properties.
Standard & Poor’s analysts found that on average, seriously delinquent borrowers were more than 13 months behind on their payments in November, compared to seven months during 2006. By the time lenders closed out a distressed loan, payments were 14 months in arrears, on average, compared with nine months in 2006.
The latest numbers from LPS show a similar trend, with seriously delinquent borrowers typically 8.5 months behind on their payments in February, up from 6.2 months in January 2008. Homeowners in the foreclosure process were 14 months behind on their payments, on average, compared with 8.4 months at the beginning of 2008.
Although experts have been warning for some time that the floodgates holding back shadow inventory would open at any time, those predictions haven’t yet panned out.
According to data gathered by the HOPE NOW alliance of loan servicers, lenders started foreclosure proceedings on about 2.8 million homes last year, up 28 percent from 2008. But the number of actual foreclosure sales was a more modest 953,000, up only 4.2 percent from the year before.
Westerback doesn’t expect a "major flood" of shadow inventory homes into the market.
"However, the age of the loans in the shadow inventory is growing, and the older they get, the more likely they are to hit some terminal point," she said.
"It’s a good thing in terms of the housing market if (the process of releasing that inventory) is dragged out, but to the extent that they become more aged, more dilapidated, or are not being maintained, that increases losses, and impacts on local neighborhoods."
NAR spokesman Walter Molony said the group’s economists expect a similar level of foreclosures will "work their way through the pipeline" in 2010.
"If this shadow inventory continues to feed into the market in a measured way, as it has, we’ll probably be in pretty good shape," Molony said.
RealtyTrac Senior Vice President Rick Sharga, whose company tracks foreclosures and provides information on properties to subscribers, also sees a steady stream rather than a rush of foreclosures.
"You’re not likely to see the banks accelerate foreclosure proceedings and rush stuff to the market," Sharga said. "They couldn’t handle the losses."
Sharga said he expects lenders to undertake a "very, very gradual burn through their inventory over the next three to four years, so you’re probably not going to see a huge dip in home prices."
What’s your opinion? Leave your comments below or send a letter to the editor.