Fair Isaac, developer of the ubiquitous FICO score, has a new warning for homeowners plotting a strategic default or walkaway: We can now spot you in advance. We've developed a black-box risk-identification tool that enables lenders and mortgage servicers to tag you months in advance -- and then pursue their own strategic measures to intervene.
The tool is so effective, according to FICO, that it can "capture nearly 67 percent of strategic defaulters" who are otherwise unremarkable and undetectable, paying their mortgages on time.
Sound a little spooky? Not for the major lenders who are working with FICO to install the new statistical risk-scoring model, aimed at some of the costliest and most perplexing defaulters in the marketplace: people who just stop paying on their loan abruptly, without ever previously being late, even though they have the income to pay.
Strategic walkaways are a multibillion-dollar headache to banks and investors. A study by researchers at the University of Chicago's Booth School of Business found that during last September alone, 35 percent of mortgage defaults in the U.S. were strategic -- up sharply from 26 percent in March 2009.
With an estimated 23 percent of all residential mortgages underwater as of March of this year, according to data from consulting firm CoreLogic, spotting -- and dealing with -- walkaways has become a high priority for the biggest banks.
Walkaways are also more than a slight concern to default risk-scoring giants like Fair Isaac and Vantage Score LLC, the joint venture created by the three national credit bureaus: Equifax, Experian and Trans Union.
Both companies have been stunned to find that the very consumers they deemed the least likely to go into default -- people with 800-plus FICOs and 900-plus Vantage scores -- are statistically more likely to default strategically, with no outward signs of impending payment stoppages, than the lower-scoring masses.
People with low FICO scores still default more often than high scorers, but when high scorers do default, they are far more likely to do so out of the blue. In the lowest score category (300 to 499) more than twice as many people default nonstrategically -- they begin missing payments over time, typically because of income declines -- than strategically.
These walkaways are especially vexing to score-modeling experts like Andrew Jennings, Fair Isaac's chief analytic officer and head of FICO Labs. "They open up new credit accounts" before stopping their mortgage payments, he told me in an interview last week. "They prepare."
They intentionally default on their mortgages in part "because they believe it is in their best financial interest, and because they believe the consequences will be minimal," Jennings said.
Jennings supervised Fair Isaac's work in developing a special tool that pinpoints likely strategic defaulters while they're still cocooning and haven't yet revealed their intentions to lenders.
Some of the research involved examining massive samples of credit bureau data -- 5 percent of all U.S. mortgage accounts -- during a recent one-year period, looking for telltale clues, month by month, that would separate out strategic defaulters from ordinary defaulters.
What the project turned up, said Jennings, helped formulate the model that FICO has now created for lenders and servicers.
So what's in the black box? Obviously the complex statistical model and exactly how it works is proprietary. But Jennings said it looks at a composite of separate risk factors from credit and real estate databases, and enables servicers to identify borrowers whose profiles match those of strategic defaulters most closely.
Some of the key characteristics include:
--How long have the borrowers owned the house? The shorter the time span, the higher the risk.
--Are they good to excellent managers of their household finances and credit relationships? Do they make modest and responsible use of credit cards and other revolving debt? Do they pay their accounts on time as a rule? Do they rarely, if ever, go over the limits on their cards -- or even come close?
--Have they departed from their past credit usage patterns in recent months by opening up multiple new accounts?
--Based on local property-value indexes, is it likely that they have slipped into negative equity territory? Remember: How deeply underwater is only a moderately predictive factor. Lots of owners whose properties are worth far less than their mortgage balances do not strategically default, but keep plugging away paying every month, while borrowers who fit the FICO strategic defaulter profile may be only slightly underwater but still walk away abruptly.
By the way, location is not a key factor in the equation. FICO found that 40 percent of all strategic defaulters live in "recourse" states where lenders can -- and do -- pursue defaulters for any un-recovered debts following a foreclosure.
Of course, the model cannot peer into would-be walkaways' minds and motivations. "We're not trying to explain their psyches," Jennings said, "but you see the patterns" and certain borrowers' profiles light up like flashing neon signs.
The top bracket of high-risk homeowners identified by FICO's new model are 110 times more likely to strategically default than other borrowers -- even though they otherwise appear to be solid customers, according to Fair Isaac.
Armed with these risk profiles, what are banks and servicers likely to do as they scan their portfolios? Fair Isaac recommends that they intervene early with what it calls "pre-delinquent treatments."
These include contacting high-risk borrowers to warn them about the consequences of strategic defaults: Their credit scores will tank by 150 points or more, they'll be hampered or penalized in applications for rentals, employment, car loans or leases, and they can forget about buying another home for at least several years, possibly as long as seven.
If they live in a state that allows deficiency recoveries, servicers will probably emphasize their determination to do so in the event of any default.
Will all this work? Major banks and FICO think it should help. The jury is out at the moment, but if the early detection concept is valid, who knows?
Maybe it will cause some homeowners to think twice and discourage them from taking that first, crucial step: Secretly plotting their walkaway, months in advance.