Novel Monitoring Techniques Could ID the Next SAC Capital

SAC Capital Advisors will have to pony up $1.2 billion—$1.8 billion if you include a previous fine of $616 million—to settle the insider trading case levied against it by the US Justice Department. The hedge fund pled guilty to five counts of fraud, and it will no longer manage money for outside investors.

SAC founder Steve Cohen was not included in the settlement and will still be allowed to trade on his personal capital. But there is still a civil suit filed by the US Securities and Exchange Commission (SEC) against Cohen, which could result in more fines and the loss of his trading license.

The questions will continue to linger about whether Cohen knew what was happening inside his house, whether he was intentionally blind toward the illegal activity, and whether there is a witch-hunt against SAC and Cohen, himself.

From a technology perspective, combating insider trading involves trade monitoring, surveillance, and the ability to capture and analyze voice and electronic communications.

"Almost everyone keeps records of their voice and electronic communications—some are mandated to archive it, some are not—but obviously, just keeping records and not deleting them for a long time is not enough. The next step is what do you do with them? How do you parse them?" says Alexei Miller, executive vice president at DataArt, who has been examining communication monitoring technologies.

Miller says that inspecting streams of text—whether for instant messages, text messages or email—is relatively easy and that the cost to do this has diminished significantly thanks to open-source solutions. Voice, though, is much more challenging.

"Voice is an interesting one because there are a few approaches to making sense out of it," he says. "You have to take voice and put it into text and then try to analyze the text. It's here where technology is, unfortunately, not quite there yet."

Miller says one of the more interesting solutions that he's heard about are firms looking at sound, rather than merely transcribing and analyzing text. The solution "listens" for unnatural pauses or whispers, with the thinking that the natural flow of conversation would be broken if something illicit was being talked about.

Further Examination
Trade book analysis could also help to prevent insider trading. SAC was known as a powerhouse when it came to event-driven trading for earnings releases or new product releases. As it pertains to the case against SAC, this is where some inside information could prove advantageous.

"One field that we're watching is trying to apply designs of behavioral psychology to trading log analysis," Miller says. "The penetration of these technologies is very, very low on Wall Street, but we've seen a few firms that are trying to make moves in this space."

Miller, as an example, points to Daniel Kahneman's book "Thinking, Fast and Slow," which says that often, people do not act rationally because we all have cognitive biases. Reverse engineering certain behaviors can often identify certain biases.

"You can use that approach to trade analysis," Miller says. "If you have 100 traders on the floor and you analyze the data that's available rationally and legally to every trader, by applying some of those methods you can figure out who is behaving irrationally and why."

While the behavior is not always sinister, you can get an idea of the traders that are reacting nervously or poorly, and potentially see who is acting illegally.

"This is where psychology and technology may actually be able to offer some insight to compliance officers," he says.

This is a topic that will continue to evolve, in part because of SAC findings and in part because of new regulatory mandates. My colleague Jake Thomases is working on a story looking at the evolution of voice and text capture. If you have some added insight, feel free to reach out to him at jacob.thomases@incisivemedia.com.