Editor’s note: In this five-part series, Inman News examines the rising tide of mortgage fraud in the U.S., giving an overview of the problem, describing how it works, explaining obstacles to fighting it and describing technologies that combat it and legislation that aims to defeat it. To read the entire five-part report, “Inside real estate’s fraud crisis,” with two bonus stories, subscribe to Inman News.
Kathleen Wood had a funny feeling about the appraisal documentation for a loan she was reviewing in her role as operations manager and chief underwriter for The Money Warehouse in mid-2005.
Wood was nervous because the house number didn’t jive with the property address on the photo. So she ran the appraisal through fraud detection software and came up with the system’s lowest possible score. Wood drove by the house, which happened to be local, and bingo — the property she had been led to was not the home detailed in the appraisal report.
As Wood’s story demonstrates, an ounce of prevention is worth a pound of cure, and technology is one way to help prevent mortgage fraud.
Fraud prevention tools can imprint a digital watermark on an appraisal document that’s nearly impossible to duplicate; verify incomes and identities; and sniff out a plethora of fraud indicators, such as several recent sales of a house or an inflated appraisal.
Of course, “no tool is perfect,” in the words of Arthur Prieston, principal of the Prieston Group, which insures lenders against mortgage fraud. When the group asked some 300 lenders who participate in its studies how effective technology is in detecting mortgage fraud last year, 90 percent said they use one or more tools; 50 percent use them on 90 percent of their loans; and 80 percent say such tools reduce fraud less than 25 percent.
Not only that – technology is a double-edged sword, and can expedite fraud just as it does fraud-fighting. Because technology is now so sophisticated, it’s easy to use desktop publishing to duplicate almost any document, altering numbers to create, for example, an artificially inflated appraisal. Fraudsters can go to the Internet and quickly and easily obtain phony Social Security numbers. The examples are endless.
But technology can be of help in fraud detection, Prieston and other experts say, advocating a combination of employee training and technology as optimal.
Residential real estate loan fraud is a national epidemic, costing communities nationwide an estimated $1 billion in 2005, compared to $429 million in 2004, according to the Federal Bureau of Investigation. In this five-part series, Inman News examines the rising tide of mortgage fraud in the U.S. This part focuses on how technology can help prevent mortgage fraud.
One of the best-known tools in the mortgage industry is the MIDEX database of the Mortgage Asset Research Institute, or MARI, a group endorsed by the Mortgage Bankers Association that tracks mortgage fraud.
“Subscribers to MARI have access to their consortium database that reports any type of derogatory activity on the part of real estate professionals,” said Mark Kasak, assistant vice president for enterprise fraud prevention at Option One Mortgage. “When we are doing broker background investigation to approve brokers, we go to MARI and see if there are any kinds of issues reported.”
According to Prieston, automated valuation models, known as AVMs, can be useful technologies in fraud detection. Examples of fraud detection software using AVMs include Interthinx, DataVerify, Digital Risk, LoanIQ by First American Real Estate Solutions and CoreLogic.
CoreLogic was founded in 1997 and its tools are used on somewhere between 20 and 25 percent of the loan applications made in the United States, according to Cary Mark, the company’s chief information officer.
Generally, clients run a CoreLogic tool up front to score the product’s risk, and if the product looks risky, they will scrutinize it more closely using the company’s AVM, Mark said.
“Fraud for profit is what we’re focused on,” Mark said. “When we can catch that, it’s going to help the lender avoid potential losses and it also helps them pump more production through the same pipe. Now, with the efficiencies going on as the housing market slows and the yield curve and the spreads are razor thin, this is a big advantage.”
“Fraud for profit” is one of the two major categories of mortgage fraud. It refers to fraud whose goal is to make money, and usually involves industry professionals, according to the FBI. The other category, “fraud for property (or housing),” is when an individual consumer makes misrepresentations in order to buy a house to live in.
Many of the fraud detection applications using AVMs are focused on fraud for profit. Web-based LoanIQ is another example, according to Robert Walker, an executive vice president at First American Real Estate Solutions.
“The prerequisite is an artificially inflated appraisal,” Walker said. “Fraud for profit could involve collusion between a loan broker, straw borrowers, all those things. But at the heart of the matter is an artificially inflated appraisal. That’s the thing we focus on.”
Many industry professionals agree about the central role of appraisals in mortgage fraud.
“Two primary drivers behind mortgage fraud are identity theft and property valuation – typically, there’s an inflated appraisal when the appraiser is in cahoots with the broker,” said Jeff Moyer, vice president, sales, for Interthinx.
Moyer, whose company services about 1,100 lenders, including mortgage giants Fannie Mae and Freddie Mac, said Interthinx provides a number of products. These include AVMs and verification that loan applicants plan to live in the subject properties, among other things.
One of the newest fraud detection products is Fidelity Information Services’ LSI FraudDetector, slated to launch in mid-May.
The product is more oriented toward fraud for housing, at least according to an abbreviated description from Brian Hershkowitz, an executive vice president at Fidelity Information Services.
The product will use external databases to scrutinize borrowers’ backgrounds to discover whether they have been flagged as having identity theft issues, whether their Social Security numbers have been assigned to someone who is dead or too young to have that particular number, and the borrower’s inquiry history, Hershkowitz said.
Also in the “fraud for housing” arena, Rapid Reporting, a Texas firm, has two products verifying income and Social Security number verification.
As Moyer commented, one of the primary drivers behind mortgage fraud is inflated appraisals. One appraisal firm took matters into its own hands to make sure its appraisals weren’t modified to produce fraudulent appraisals.
Mitchell Maxwell & Jackson is a residential appraiser in New York and Connecticut that produces up to 30,000 appraisals a year. When a couple of reports bearing the company’s name were modified to produce falsified appraisals, the company swung into action.
“It’s possible to scan a document and convert it to a Word document you can edit,” said Michael Martin, the company’s managing director. “When we found out about the modifications, we decided to protect ourselves and the integrity of the brand.”
The company custom-made a digital watermark that now appears on the background every one of its reports – its initials, MMJ.
“We took some clues from the Treasury Department – the graphic specifications that make photocopying and scanning virtually impossible,” Martin said. These anti-counterfeiting techniques include using lines so tiny and close to each other that, when scanned or photocopied, “they don’t even show up as separate lines,” Martin said.
“It’s true that someone could still get away with counterfeiting one of our appraisals, just as they do when counterfeiting money,” Martin acknowledged. “We are doing this so we know without a doubt if anything has been done. If someone brings something to us and says, ‘This doesn’t seem right,’ we can say it has been modified and take appropriate actions from there.”
Send tips or a Letter to the Editor to firstname.lastname@example.org or call (510) 658-9252, ext. 140.