Saturday, August 16, 2008

Investigating NJ Numbers Part 1

This is a follow-up on our previous post regarding RealtyTrac's numbers.

First we like the services that RealtyTrac offers. We like their output and their statistics are very interesting. But upon reading the differing numbers between their second quarter 2008 numbers and the July numbers we felt something was askew. From a little research it appears we are just another in a long line of questioners.

Unfortunately the media treats their press releases as industry experts, however the main goal of RealtyTrac is to sell properties. From their website -

Welcome to RealtyTrac, the nation’s leading online foreclosure marketplace and the most trusted source of foreclosure information.

Over the last year there have been several articles regarding the reliability of their statistics. Here are two excellent articles -

Here is an excerpt from Business and Media -

“RealtyTrac's monthly foreclosure data is still inflated, because they haven't changed their methodology on a month-by-month basis,” Sam Ali, a reporter for the Star-Ledger, told BMI. “So a single property may still get counted several times based on the number of filings it has accrued over time. They did release a six-month (January-June) roundup where they did include what they described as ‘Unique Households.’ They claim this figure eliminates the double and triple counting of the same properties.”


But even with the adjustments in the RealtyTrac data, Ali determined it would not be possible to compare their “tweaked” data of this year to last year’s and come up with an accurate number of increased foreclosures.


“In New Jersey that reduced the number of foreclosure filings from January through June from something like 27,000 to 13,000 – so it cut the figure in half, but for some reason they do not break down ‘Unique Households’ on a month-by-month basis,” Ali said. “Nor did they include a ‘Unique Household’ number for January through June 2006, so it’s impossible to compare year-over-year figures.”


Yet the media continue to use RealtyTrac’s questionable numbers.

“Sadly, I think all of us in the media are to blame for that,” Ali said. “Call it the lemming factor. RealtyTrac has been pretty aggressive about sandblasting their press releases to every media outlet on the planet with catchy, eye-popping headlines. When one person reports the figures, pretty soon, everyone starts reporting the figures. RealtyTrac first and foremost is not a research firm – it’s a company that sells foreclosure lists to real estate investors.”


Here is an exerpt from the Atlanta Journal Constitution (unfortunately one needs a subscription form original so here is a copy)-

RealtyTrac, one of the nation's leading sources of foreclosure statistics, reported 12,602 July foreclosure actions for Georgia. But that total counted more than 2,000 properties twice, and sometimes more, The Atlanta Journal-Constitution found in a review of the data.

There is little dispute that Georgia faces a foreclosure crisis, but the company's July report overstated the magnitude of the problem. After a preliminary investigation, the company said Friday that its data show foreclosure filings in July actually rose by 14 percent — not by 75 percent.

...
RealtyTrac said it also provides data to the Federal Reserve, the Federal Deposit Insurance Corp. and the FBI.

At a time when mortgage failures top the public agenda, no one knows the precise scope of the problem.

No government agency collects nationwide foreclosure statistics. The Mortgage Bankers Association reports quarterly on delinquent mortgages and foreclosures, but its survey covers only 80 percent of mortgages and does not include a breakdown below the state level. Other organizations use credit files and lender surveys to produce estimates.


The numbers are touted as gospel from the major networks and main national papers, however when people do some digging in the numbers there are some questions. Various states have various timelines so being in the foreclosure process with states that have timelines that are over a year will, by default, have higher numbers just since people stay in the foreclosure process longer. There is also the issue involving actual counting and double counting. We will look at NJ actual numbers in an upcoming post.

No comments: