Published: December 2, 2014
As part of his punishment for what authorities call the most profitable insider trading scheme in history, Mathew Martoma consumed his Thanksgiving meal at FCI Miami—the prison that he checked into on November 20th. The former hedge fund manager for S.A.C. Capital Advisors had generated a $275 million profit in 2008, which New York federal judge Paul G. Gardephe called “hundreds of millions of dollars more than ever seen in an insider trading prosecution.” He’ll be serving nine years—that’s about an hour for every $3,500 he stole—unless released sooner for wholesome behavior.
In sentencing Martoma in September, the judge called his illegal trading edge “deeply corrosive to our markets,” and noted that “the conduct was well-planned and Mr. Martoma knew the amount of avoided losses or profits were likely to be staggering.”
But what if the S.E.C. and FBI had been able to utilize cutting-edge software that could have detected illegal trading at the company as it was starting to materialize in 2008? For one thing, Martoma wasn’t alone. Former S.A.C. trader Michael Steinberg was also convicted for inside trading, and the firm itself pled guilty and paid a $1.8 billion penalty for failing to prevent its employees from engaging in the illegal activity. Could all of this skulduggery have been spotted sooner?
In recent years, there’ve been gigabytes of chatter inside Wall Street’s compliance departments about the need for better surveillance technology (employed by the firms, or regulators—or both) to keep a closer tab on traders. FINRA, the country’s largest independent securities regulator, recently proposed a rule that would require the production of trading information on a granular level at financial service firms, including from customer accounts. There’s no question that the firms and regulators have access to ever-expanding mountains of data they didn’t have in the past, but creating software to sift through it all (especially on a real-time basis) and presenting the findings in a visually understandable format is the ultimate challenge.
Now a year-old startup called AIMPaaS, which develops trading execution platforms, has come up with what may be a viable solution: A comprehensive software system that appears to be unique in the financial services industry. It not only executes the stock orders, but it also employs sophisticated behavorial modeling of traders and portfolio managers that can detect insider trading and other nefarious or impermissable activities. Moreover, it ranks and rates analysts in a simulated environment that can also guard against insider trading. This three-pronged approach, combined with deep expertise on Wall Street, is the key thing that differentiates AIMPaaS from competitors.
AIMPaaS’s tech triggers alerts if patterns are altered that could indicate insider trading—for example, larger dollar trades; greater quantities of stock; buying more aggressively almost regardless of price; or perhaps suddenly trading in nontransparent venues such as dark pools.
The concept of behavorial variance tracking in and of itself is nothing new. Such modeling started nearly a decade ago in the fraud and financial crimes arena, with companies such as NICE Actimize, Oracle Mantas, Detica (BAE Systems), and IBM. All of them have some form of profile-driven monitoring of patterns in order to find abnormal activities. In geek-speak, such analytics are designed to detect unknown behaviors that are anomalous and deviate statistically from historical (or expected) activity—as opposed to algorithm-driven monitoring which is designed to detect a known behavior. “I call that a rifle shot, because they are targetted on specific, defined behaviors,” says James Heinzman, one of the industry’s top experts on regulatory compliance technology. (Heinzman developed compliance and money-laundering solutions for NICE Actimize, a leading player.)
In addition, firms such as Thompson Reuters have created a data feed that identifies the impact of news events on the price of publicly-traded securities that’s known as “directional news sentiment.” Put simply, for news stories that cover corporate announcements of, say, takeovers or clinical trials for drugs, they can determine the (positive or negative) impact of the stories on the stock price. They do this with algorithms designed to detect trading activity prior to that announcement that may have benefitted from the impact of the news article—a form of insider trading.
Read the full article here.
Originally Posted on Forbes.com on December 2, 2014
Richard Behar is the Contributing Editor, Investigations, for Forbes magazine. He can be reached at firstname.lastname@example.org