RICHLAND -- Software developed at Pacific Northwest National Laboratory to make sense of national security data has been adapted for a new use -- searching for trends or signs of fraud in financial data.
In a demonstration, it picked out anomalies in the returns reported by disgraced financier Bernard Madoff, according to the national lab in Richland.
If the financial software, called Anomalator, had been available, "I think it would have blown the whistle on Madoff much earlier," said Burton "Bud" Sheppard, president of V-Indicator Analytics, a startup company in Spokane.
Sheppard approached Battelle, which operates the national lab for DOE, after learning of PNNL and Battelle's long history of analyzing large amounts of disparate data for homeland security applications and then presenting complex results in visual ways that made them easier to grasp.
Researchers and commercialization managers agreed to help him develop visual analytics software that would allow large amounts of financial data to be analyzed and explained.
Battelle used its private technology maturation funds to finance the development and now has granted a license to V-Indicator to market Anomalator for use in the financial services industry.
"The Great Recession of the late 2000s has shown how questionable financial practices can place America's economy at serious risk," said John McEntire, a Battelle commercialization manager, in a statement. "The Anomalator provides the unbiased, fact-based analysis needed to identify those problematic practices and help protect the nation's economy."
Traditional financial analysis reports provide either a list of numbers or a simple line graph to represent value of just one investment over time.
But the Anomalator uses mathematical algorithms to identify atypical data in databases that record the movement of funds or their management over time. It then creates line graphs representing the progress of anomalous funds or management compared with others of interest.
In the example of Madoff, V-Indicator and Battelle used Anomalator to compile the financier's stated returns from one of his leading feeder funds. While other funds spiked up or down, Madoff's funds showed a steady upward slope on the graph, never losing money.
In another look at his returns from 2001 through 2004, an analysis showed they remained far steadier than other funds that spiked up and down.
"This software shows over time very graphically what happened along the way," Sheppard said.
Madoff, who is serving a 150-year prison sentence, had told his investors that he was performing a very complicated auctions program with a "split strike conversion strategy." Instead, he was running a Ponzi scheme that scammed investors out of as much as $65 billion in a fraud that was concealed for more than 15 years.
Now V-Indicator is working with financial industry leaders to apply Anomalator to problems addressed by the Dodd-Frank legislation, the 2010 U.S. financial regulation reform law, and its regulations.
The software has applications for financial regulatory and forensic uses, but it's most valuable use may be to help make investment decisions, Sheppard said.
Fund managers may use it to create better funds and financial advisers may use it to improve advice to clients on fund selection and management, he said.
He is also interested in developing the software for other industries with large amounts of data, such as health care.
"When you cut to the chase, it's anomaly detection," he said.
In the health care industry, it could see unenvisioned anomalies in data for hospital error, patient safety, genomics, clinical and pharmaceutical trials and Medicare and Medicaid fraud, he said.
-- Annette Cary: 582-1533; firstname.lastname@example.org