Lawmakers want to shake out TARP fraud with tech

A standardized, Web-accessible database showing how TARP money is spent could make it easier to detect fraud in the program, according to experts.

Although the Troubled Asset Relief Program (TARP) played a significant role in bringing the financial system back from the brink of collapse, it is unlikely that taxpayers will see a full return on the $700 billion in spending, according to a report by a Treasury Department independent watchdog.

Certain TARP programs, such as the mortgage modification component of the $50 billion Making Homes Affordable program, will yield no direct return, according to the report released Oct. 21 by the Office of the Special Inspector General for TARP.

For other programs, “including the extraordinary assistance to American International Group Inc. and the auto companies, full recovery is far from certain,” the report states.

As money continues to be spent under TARP, some members of the House hope to amend the program to direct Treasury to use technology to detect fraud, according to industry experts that advised Congress about the technology. Credit card companies, using such tools, can call a customer as soon as a suspicious transaction takes place because the software detects activity that doesn't match a customer’s normal behavior, said Stephen Horne, vice president of master data management and integration services at Dow Jones.

Similar technology could alert federal government officials about fraudulent spending under TARP, Horne said at a recent panel discussion.

Reps. Carolyn Maloney (D-N.Y.) and Peter King (R-N.Y.) introduced a bill, H.R. 1242, earlier this year that is now in the House Financial Services Committee. It has 40 co-sponsors from both parties.

The bill is only for TARP spending, but Maloney said it could be a model for transparency in other areas of government.

“We can only hope that normalizing our databases, as my TARP Data bill would do, will spread to other government programs,” Maloney said.

The measure would require Treasury to create a centralized, Web-accessible public database system that has a standardized format so that TARP funds are easily visible and traceable, Maloney said.

Treasury would be required to combine the reported government data with data collected by independent sources, including corporate press releases, news articles, indexes, corporate profiles and other nongovernment financial information, Maloney said. 

Technologies such as data mining are well suited to discover fraud in TARP, said Tom Kimner, risk manager the SAS Institute's Americas Risk Practice. Kimner testified before a congressional committee about the role data mining can play in TARP oversight.

“Analytic technologies can help identify patterns and anomalies, optimize where the funds are applied, and pinpoint use and abuse of funds,” Kimmer said. “Before you reach that point, though, you must have confidence in the accuracy of the data. Technologies supporting data collection, integration and quality are critical in that effort.”

Because the bill would require the database to be public, it could be a boon for private developers that come up with new ways to use the data, Maloney said.

“We all live in an information economy now,” Maloney said. “Those who build a better mousetrap to present existing information in a new way can often profit from their insight; it’s truly the American way.”