|
Special Reports
Fraud Detection
AppIntelligence Automates IRS Verification Process
By Jennifer Harmon
ST. LOUIS -- In response to the growing need for fraud detection tools,
AppIntelligence Inc. developed and implemented an electronic solution for processing 4506-T requests. The 4506-T
is a form sent to the Internal Revenue Service to request a transcript of a tax return and is used to verify income
for an individual.
According to the company, demand for the service has increased by over 300% over
the past four months.
Mortgage professionals simply submit their request to AppIntelligence, who handle
all of the processing and perform a quality assurance check before electronically submitting the request to the
IRS. The transcript is obtained within days vs. the six to 10 weeks it takes to obtain a copy of a complete tax
return.
When the service was initially launched, AppIntelligence received between 30-40
requests per day. Now just four months later these requests have increased to over 100 requests per day and the
numbers keep growing, the company said.
The ease of this processing combined with the benefit of the transcript as a fraud-detection
tool has demand for the service skyrocketing, according to Don Effertz, vice president of operations.
"Traditional methods were somewhat faster, but were rife with fraud. Today
verification of employment letters, check stubs, W2s and even tax returns can easily be falsified. An actual copy
of an individual's tax return can be obtained from the IRS. However, each request costs $39 and takes about six
weeks to process. The transcript contains all the necessary information for income verification and is delivered
in approximately 72 hours."
Stephen Gott, chief executive officer of AppIntelligence, gives a preview of what's
next for the 4506-T requests. "Our teams worked directly with the IRS to establish this process and set up
quality controls. Now that we've made the process easy and economical, we're developing a scoring system to apply
to the transcripts, removing even more time from the process and empowering lenders with a quantitative tool for
fraud detection."
Income verification often indicates a large number of fraudulent loans, said Mr.
Gott, with someone else such as a secretary filling out tax returns for a boss. The forms never make it to the
personnel office.
"We see a lot of cases of fraud, especially self-employed people. It tells
us how much you submitted to the IRS vs. what you said on the W2."
The company scans billions of dollars of loans every day, searching people's histories
such as all of the addresses for homes that a person has ever lived in. Four hundred different checks are completed
in the span of 15 seconds.
AppIntelligence, which provides data integrity, fraud prevention, risk assessment
and compliance tools for the residential lending industry, has set out to evangelize the market on the growing
problem of fraud. From small companies to larger institutions, Mr. Gott said write-off percentages are significant
in every single firm he deals with across the country.
In retrospect, he explains, defaults could have been prevented if only the bank
or mortgage company had done their homework. "The industry has a long way to go. That's why we run loans free
to show them 50% are stoppable," he said.
"The problem could be containable if they had looked at these loans before
they were originated. We believe any early payment default in excess of 90 days...50% of those are fraudulent.
Through our automated systems, we scan these loans prior to the closing and eliminate these write-offs."
Right now, the mortgage market is seeing different shades of gray when it comes
to fraud. "Some view the problem as simply villains who falsify documents, they steal money and specialize
in organized crime," said Mr. Gott.
"But in many cases, white-collar fraud is the guy who promises his wife a
bigger house and 'by hook or by crook' stretches the documents for a fund he can't afford. As rates increase, he
can't make payments on the house and the bank loses $100,000."
Part of the secret to detecting fraud is to get as much data about the consumer
and property as possible. "We know if your cell phone is linked to the address of the house you claim to live
in. Many think with a cell phone, no one knows who you are," he said.
When AppIntelligence scans data, if it sees a discrepancy between the city and
address, the loan is questioned. There may be an explanation in certain situations, according to Mr. Gott. "If
not, we will catch the problem. That's part of our service. We're going to get the history of these people to show
if an individual has problems in their past. For example, we can see how many times their houses have burned down.
Or if they live in one city and commute to work in another. If the distance is unrealistic, the probability is
slim to none the loan is fraudulent."
Using fraud technology tools, if a borrower's Social Security number turns out
to be wrong, that doesn't necessarily mean it is identity theft. The number might simply have been keyed in wrong.
But if a combination of red flags is raised, fraud is likely to have occurred, and the bank or mortgage company
is called to say there might be a problem with the loan. If on a scale of 0-1,000 the borrower receives a score
of 200, that means the probability of fraud is almost 100%, said Mr. Gott. If the score is in the 400-500 range,
some questions have been raised and researched. And when the score is over 600, you know it's a good loan, he said.
|
