|
Special Reports
Mortgage Fraud: Protecting
The Investment
Flexibility Viewed as Key to Detecting Fraud
By Anthony Garritano
WELDON SPRING, MO -- As the instances and kinds of fraud increase, the most effective
fraud detection system is not the one that's the most tech heavy, but the one that has the ability to change with
the birth of new types of fraud and draw from the largest number of outside data sets to check and double-check
every component of the application at the same time.
"The most common fraud is definitely white-collar fraud, which is when the
average borrower lies about something that would impact their capability to get a loan," said Stephen Gott,
chairman and chief executive officer of AppIntell Inc. here.
"In a lot of these cases the bank will eventually foreclose. Banks have huge
sums of money put aside for foreclosures, but what they don't realize is that 50% of the foreclosures could have
been prevented by doing a thorough investigation of the application before it was closed."
Appintell's automated fraud detection service looks at 350 components of the standard
application and goes to outside data sources to match the information on the application to see who has a criminal
record, who has a history of bad loans, who has lied about their income, who stated that they work 15 minutes from
their new home but actually work an hour away, etc.
Other fraud detection companies like Loveland, Colo.-based Kroll Factual Data
introduced new fraud detection products dating back two or three years ago that are just getting traction because
of the impact of Sept. 11, according to Kroll senior vice president Dave Vinson. "Fraud used to be a dirty
little secret," he said. "However, after Sept. 11, the government said that it was important to identify
people to prevent money laundering that could be funding future terrorist acts.
"As a result, we have identity products that do a social verification, a
death benefits search and a chronological search of a 1003 to ensure that the applicant did in fact live in San
Diego for the last three years for example," Mr. Vinson continued. "When a discrepancy arises and a new
name or social comes into play, we can run a search on the new social or name to decipher if an identity was stolen."
However, despite popular opinion, the most common type of identity fraud is not
identity theft. "People are concerned about 1003 fraud and people getting their identity by buying goods online,
but that's not common," said Mr. Vinson. "That type of fraud is more common at the local Denny's where
the server has direct access to your credit card and not necessarily during the mortgage process or an Internet
sale."
"Identity-theft fraud is prevalent, but it's easy to catch and it isn't the
most common type of fraud out there," Mr. Gott agreed. "The most common type of fraud is white collar."
But the face of fraud is always changing. "The most common type of new fraud
is appraiser fraud where the appraiser is either lying on the appraisal or finding comparables that will meet the
qualifications of the house because they need the property approved at a higher price," said Mr. Gott.
"This type of fraud is one reason why automated valuation models are becoming
so popular," added Mr. Vinson. "There's no subjectivity involved when using an AVM. Historic sales data
is compared to comps in the area and in 70% of cases you can come up with a legitimate value without having to
employ a physical appraiser."
New technology innovations will continue full steam ahead as the instances of
fraud are expected to increase in the coming environment as rates increase and loans are extended to less-than-stellar
applicants. The bottom line is that technology has to adapt as fraud is ever changing.
"As with any disease, fraud mutates," said Mr. Gott. "Fraud will
try to go into those areas where it can't be detected so we continue to stay up with the latest kinds of fraud
to develop hybrids to catch the new ways people are committing fraud."
|
