Responding to critics who say its reports overstate the magnitude of the housing downturn, RealtyTrac next week plans to begin providing quarterly, state-by-state breakouts of the number of unique households that are in some stage of the foreclosure process, while continuing to provide its monthly reports on total foreclosure filings.
RealtyTrac’s reports have been criticized by groups such as the Mortgage Bankers Association and the Colorado Division of Housing as misleading, because some properties may be counted more than once as they move through the foreclosure process.
In many states where the foreclosure process occurs in stages, legal notices are filed when a lender issues a notice of default, and again before a property is put up for sale. If the property does not sell at auction, another foreclosure filing is also made when it is repossessed by the lender.
Although RealtyTrac says it’s unlikely for a property moving through the foreclosure process to generate two filings in a month, the company will begin providing counts of unique households in its quarterly reports to provide a clearer picture of the actual number of homes in the foreclosure process.
“For anybody who understands how the foreclosure process works — it’s a process, not an event,” and RealtyTrac’s reports of total foreclosure filings were not misleading, said Rick Sharga, the company’s vice president of marketing. “But for people who would like more pure assessments of how many households affected, this should give them the number they are looking for.”
Sharga said RealtyTrac had planned to release its first quarterly report with unique household counts today, but that the release has been pushed back for a week to double-check data.
The ability to provide unique household counts is the outgrowth of collaboration between RealtyTrac and researchers from the Federal Reserve Bank of Kansas City, Sharga said. The researchers are conducting a risk assessment for member banks by mapping the individual addresses of properties in foreclosure using data collected by RealtyTrac.
The foreclosure filings RealtyTrac collects from public sources includes addresses, and now that the company has figured out how to automate the process of isolating that information, it’s able to count the number of unique households in the foreclosure filings it collects nationwide, Sharga said.
When RealtyTrac begins releasing those numbers on a quarterly, biannual and annual basis, “You’ll have two ways of looking at the report — the way it was done in the past, with the total number of foreclosure filings, and the number of actual households,” Sharga said. “I think this new measurement will reduce the amount of misinterpretation of the numbers.”
The way it was done in the past overstated the number of foreclosures by about 30 percent, according to the Mortgage Bankers Association. When the Congressional Joint Economic Committee issued a report on subprime lending this spring, the MBA complained that the “report relies heavily on foreclosure estimates from RealtyTrac, an Internet-based firm specializing in marketing foreclosed properties.”
The MBA produces its own quarterly report on delinquencies and foreclosures that’s based on a survey of its members. The group says the survey covers 80 percent of first-lien mortgage loans, but it does not include nonbank, “private-label” lenders who financed many subprime loans during the housing boom.
The Colorado Division of Housing has also taken issue with RealtyTrac’s numbers, and has begun providing its own quarterly statistics to counter them.
While RealtyTrac reported 54,747 foreclosure filings in Colorado — or one for every 33 homes — the Division of Housing counted 28,453 foreclosure filings, or one for every 58 homes.
“This disparity is likely a function of different methods used in counting foreclosures, and while we assume good faith on the part of RealtyTrac, we have concluded that their method overcounts foreclosures in Colorado,” state officials said in releasing their first-quarter 2007 report.
Sharga said that even when the new unique household count is applied to 2006 data, Colorado had the highest foreclosure rate of any state. Although he did not provide numbers for 2007, Sharga said, “Our preliminary (unique household) numbers show theirs are very close to ours.”
The sensitivity over RealtyTrac’s numbers is understandable, Sharga said, because “you don’t want that kind of negative publicity if you’re a state looking to grow economically and expand,” he said.
RealtyTrac’s numbers are widely cited in the media, and in reports and testimony before Congress — which is considering new restrictions on lenders because of the rise in delinquencies and foreclosures.
So RealtyTrac strives to be “as open and transparent as possible about what we count and how we count,” Sharga said.
Tracking the documents filed at each stage of the foreclosure process can be useful because it can provide an indication of whether borrowers and lenders are able to work out loan modifications, short sales, or take other steps to prevent the foreclosure process from being completed.
The percentage of lenders who are forced to take possession of a foreclosed property compared to initial notices of default and notices of sale, for example, can be an indication of a market’s strength or weakness.
The percentage of real estate-owned (REO) filings RealtyTrac reported for California in June was only about 11 percent — suggesting that lenders only end up taking possession of homes in one in 10 foreclosure filings. About 71 percent of foreclosure filings in California were notices of default, and only 18 percent were notice of trustee sales — an indication that the market is strong enough for borrowers to refinance their debt or engage in workouts or short sales with lenders.
In Ohio, however, REO filings made up about 34 percent of total foreclosure filings, suggesting that lenders were having less success in conducting workouts with borrowers or auctioning properties off.
Sharga said the ability to look at unique households will also allow RealtyTrac to report the average number of foreclosure filings per property. That will provide another useful clue as to what’s happening to homes that enter the foreclosure process.
“If you are in a state like California, where (most) of the filings are initial notices of default, it would be unusual to see the average number of filings per property get above two,” Sharga said. “But it’s going to be another one of those statistics where you can argue whether a state with a high number of filings, with an average of 2.1 filings per property, is worse off than a state where there are a lower number of filings, but the average is 2.7 filings per property. That’s where the economists come in.”
That’s still a statistic, rather than a historical record of what happens to properties as they move through the foreclosure process. Getting that level of information will require more work on the back end of the database, Sharga said, and resolving the disparities in the ways records are kept at the state and county level.
“You’d think it would be straightforward to track a property all the way through the process, but with more than 3,000 counties, it’s a lot more difficult than it sounds,” Sharga said.