A convicted felon once told MARI’s Bill Matthews that he never would go into a bank’s branch office because he feared employees might recognize him. Instead, he preferred the less personal telephone to commit fraud.
Now, the Internet takes that depersonalization a step further. That shift has been both a blessing and a curse to the mortgage industry, said Matthews, who works as VP and general manager for Mortgage Asset Research Inc. The Web and other automated systems have added efficiency to the industry, but they have also made it easier for perpetrators to take advantage of the faceless relationships the technology creates.
The challenge lies in developing new tools to detect and prevent mortgage fraud, which experts say is rampant and on the upswing. Mortgage fraud can cost lenders and consumers millions of dollars. Perpetrators often include sophisticated rings of professionals, and the crimes can take years to investigate, making prevention even more important.
The technology used in fraud detection and prevention is ever changing. Some new tools focus on confirming borrowers’ identities, while others focus on a property’s background.
“Where there’s money, people get creative and people have gotten much more creative in the mortgage industry,” Matthews said. “A lot of people did fraud before to get commissions. Now they’re doing fraud to rip people off.”
Real estate fraud usually requires collusion among different professionals, from loan originators to appraisers, Matthews said. But it also requires a buyer. Fraudsters may find buyers by duping them into going along, stealing their identities or by using straw buyers, which are sham buyers who are part of the fraud ring.
MARI recently developed an identification verification tool to confirm borrower’s identities before loan approval. Subscribers to the company’s MIDEX database can enter certain key data, including date of birth, Social Security number and address, and the database searches its records for anything that might trigger a red flag, such as an applicant’s SSN showing up on the master death index or an applicant’s supposed home address showing up as an office address.
The database also can check whether that address has been used in other suspected fraud cases, Matthews said.
“If you look at one piece of data, it could look fine, but then you look at 10 pieces of data and start comparing them, you start realizing you have a catastrophic problem, not just a minor one,” Matthews said.
The MIDEX database contains public information, along with the non-public suspected fraud data that subscribing companies report. Subscribers send MARI descriptions, which include the company that originated the loan, the loan officer and any other information they want to reveal. Subscribers can only check the database if they agree to contribute when encountering something suspicious.
MARI also is working on a feature that will allow database subscribers to filter out certain types of records that don’t interest them. For example, some lenders may not care about loan originators who have been fined small amounts or a property that has been flipped only once. That service, available next year, will allow lenders to focus their anti-fraud efforts where they believe they will have the most impact, Matthews said.
AppIntelligence’s latest innovation addresses another major area of concern for lenders – making sure applicants earn what they say they earn. The product enables lenders to verify an applicant’s employment for about $5 per loan. Such reports usually fetch about $20 per loan, according to Stephen Gott, AppIntelligence’s chairman and CEO.
High-speed optical scanners and artificial intelligence that can read documents have helped knock down the price, Gott said. The company spent a year and “substantial capital investment” on developing the technology, which digitally scans a document from a fax machine, automatically “reads” the document and sends it to the appropriate office at the Internal Revenue Service for verification. Results generally take up to 24 hours, Gott said.
In addition to technology, AppIntelligence also works on developing algorithms that help detect potential fraud on applications. One algorithm compares the location of an applicant’s employer with the house being bought to determine the commuting distance and whether it’s possible. Certain occupations, such as sales, are exceptions.
Inman Blog: Industry pros give their take on everything from home ownership, foreclosures, mortgages and real estate life to the weather, our economy, and politics. Check in daily to get the scoop.
Other advances in fraud prevention look for clues in the property for sale.
Hansen Quality, a division of Fidelity National Financial, recently launched Insight, a Web-based screening tool to identify hidden collateral risks at the loan origination level and in secondary market transactions.
The tool combines Fidelity’s data and analytics to assess risk based on REO (real estate owned) activity, recent number of flips and high property turnover volumes in specific areas. The electronic reports include a summary and a range of raw data for further analysis. Details include a 10-year transaction history of the property in question, 15 comparable neighborhood sales, along with a cluster of properties located within close proximity, regardless of sale price.
The information is used to determine risk based on property type, ownership and values of surrounding property. The risk assessment translates into a rating to determine whether the property is viable collateral for a loan. Each rating includes a recommendation for further due diligence if necessary.
Mortgage fraud technology is helpful in detection and prevention, but lenders shouldn’t rely upon it alone, said Arthur Prieston, founder of The Prieston Group. Prieston encourages lenders to implement standards and best practices through programs such as TPG’s LoanCert.
“It still boils down to who you do business with,” Prieston said.
TPG considers a lender’s overall business and lending practices before approving them. The company conducts on-site visits and examines the lender’s history and types of employee training it offers, among other things.
Lenders that purchase TPG’s insurance as a safety net agree to adopt the company’s standards for loan reviews, quality control and closing processes, said Carleton Prince, the company’s EVP and chief marketing officer. TPG provides ongoing support, which includes training quality control personnel on how to detect fraud.
TPG recently partnered with AppIntelligence’s DISSCO, its front-end fraud detection and prevention product. Loans that meet certain DISSCO scores and are insured by TPG receive preferred pricing on insurance premiums.
Technology has significantly advanced in mortgage fraud detection and will continue to change along with the methods of crime. But convincing lenders that fraud happens to them will be the biggest advance to be made, according to Gott. Fraud is not the Sopranos; it does not involve organized crime, he said.
“A substantial problem in the industry today is that most lenders do not believe that their write-offs are due to fraud. They just think it’s the cost of doing business,” Gott said.
Send tips or a Letter to the Editor to firstname.lastname@example.org or call (510) 658-9252, ext. 140.