When mortgage fraud schemes unravel, their lack of sophistication is often surprising. Borrowers’ incomes may have been falsified and property values grossly inflated, but nobody seems to notice anything’s amiss until the loan payments stop coming.

Banks routinely open their vaults and make bad loans based on false information provided by real estate professionals, including mortgage brokers, appraisers and closing agents.

When mortgage fraud schemes unravel, their lack of sophistication is often surprising. Borrowers’ incomes may have been falsified and property values grossly inflated, but nobody seems to notice anything’s amiss until the loan payments stop coming.

Banks routinely open their vaults and make bad loans based on false information provided by real estate professionals, including mortgage brokers, appraisers and closing agents. Because they handle a tremendous volume of applications, it’s difficult for human loan officers to spot trends that could indicate a loan application or mortgage broker deserves closer scrutiny.

Using pattern-recognition techniques employed by the credit card and insurance industries, Carlsbad, Calif.-based BasePoint Analytics claims to have developed software that can help lenders and mortgage firms identify problem brokers and problem loans. 

Only a small percentage of brokers intentionally submit bad information or engage in fraudulent activity, it seems. BasePoint’s own study of 3 million loans originated by 50,000 brokers found that 12 percent of brokers accounted for all bad loans.

While some lenders rely on “watch lists” that attempt to identify problem brokers, honest brokers may end up on those lists by mistake. Lenders who rely solely on watch lists may lose out on business from highly productive brokers, said Tim Grace, BasePoint’s president and chief executive officer.

BasePoint’s fraud detection products include FraudMark, which is used to score individual loan applications for fraud risk. In the first six months of this year, BasePoint says its clients used FraudMark to put the brakes on $300 million in suspicious loans.

The newly unveiled BrokerWatch is designed to complement FraudMark, tracking historical broker behavior and looking for patterns of fraud on a loan-by-loan basis.

FraudMark and BrokerWatch, for example, look at each borrower’s occupation — whether it be landscaper or white-collar professional — and check to see if their stated income falls within the norms for their profession.

BrokerWatch looks for deviations from a broker’s past practices, such as the number of loans handled in a month or the rate of early pay defaults. Instead of arbitrary benchmarks, the program uses each broker’s historical record to establish baseline norms.

“We don’t do hard and fast rules — it’s always according to norms,” Grace said.

In one instance, BasePoint software helped derail approval of $3 million in suspect loans submitted by one broker in 34 loan separate applications. All the applicants worked for just four firms, and their average income was 20 percent higher than the mean, Grace said. After the loans were flagged, a risk manager determined the broker had used payroll software to produce fake paychecks for the applicants in order to document their ability to repay the loans. The checks were even in sequence, Grace said.

In another instance, 24 loans totaling $8 million were flagged when pattern recognition software determined that an unusually high percentage of the borrowers — 60 percent — were self-employed, Grace said.

For now, clients can only use BrokerWatch to monitor a broker’s history within their own firm. That’s a potential weakness, since perpetrators of mortgage fraud often use multiple lenders to avoid detection. But a future phase of the product’s development will allow clients to become members of a consortium that shares data on broker behavior, Grace said. 

“Fraud does travel. It will take the path of least resistance,” Grace said. “As a lender, you just have to make sure your defenses are the strongest in the industry so that fraud does not affect you.”

Many mortgage fraud schemes rely on inflated appraisals to flip property. Grace said. BasePoint’s software compares property values provided in loan applications with historical data that’s collected from lenders. Beginning in October, he said, the company plans to offer an enhanced service with access to the vast databases of First American Real Estate Solutions. First American, which owns a minority stake in BasePoint, claims to be the nation’s largest collector of real estate public records, tracking 2 million property and mortgage transactions a month.

So how do mortgage brokers view BrokerWatch and other methods of tracking their behavior?

Jack Williams, president of the California Association of Mortgage Brokers, said watch lists and software programs shouldn’t be used to penalize brokers for honest mistakes.

“The bad thing (about a watch list) is there may be something in the file that you cleaned up, or maybe you cancelled the loan out. If the lender wants to, they can put you down as being suspected of fraud, when in reality, it was done unintentionally or corrected,” Williams said. “I have a problem with that type of reporting.”

Williams said he wasn’t familiar with BrokerWatch. But in general, he said, if a program is sophisticated enough to catch fraudulent mortgage brokers without penalizing the innocent, “We’ll fully support that. We’re consumer oriented — we’re trying to protect the consumers to make sure they can get the best deal they can. We sure don’t want people out there being predatory or committing fraud. It goes against the grain of our code of ethics.”

Grace said BasePoint’s software won’t flag brokers just because they do business with high-risk borrowers, such as those with low credit scores or incomes. He noted that BasePoint’s clients include subprime lender New Century Financial Corp. BasePoint also counts brokerages, including Aegis Mortgage Corp., as clients.

Show Comments Hide Comments

Comments

Sign up for Inman’s Morning Headlines
What you need to know to start your day with all the latest industry developments
Success!
Thank you for subscribing to Morning Headlines.
Back to top