Stopping rogue traders in their tracks
UBS trader Kweku Adoboli cost his employer $2.3 billion in losses and $47.5 million in fines with his fraudulent trading. He will pay for it with seven years in prison according to the sentence he received last week. But firms are hoping big data can stop such fraud before it does damage, according to the Securities Technology Monitor.
Companies like Cataphora, which models individual and organizational behavior, thinks they can stop it. Cataphora is using big data to model employee behavior. It can show a contextual relationship between data--email, spreadsheets, instant messages, phone calls, voice mail, tweets, Facebook (NASDAQ: FB) status updates, expense reports, etc.--and build a digital character for each employee that is mapped against a model of the organization's normal behavior. And then, detect deviations.
The company looks at things such as the consistency of routine, which can indicate some sort of stress, distraction or disgruntlement, or the consistency of channel, which indicates an employee's desire to avoid leaving a written record through emails or instant messages.
It also relies on something called centrality, which measures an employee's overall sense of engagement, or lack of it. Other patterns of behavior include consistency of hierarchy, which identifies employees who have control over roles usually reserved for two or more people, as a means of checks and balances. And the way an employee spends his or her time can also raise a red flag, if it deviates from established patterns virtually or in the real world.
Earlier this year, Threadneedle, an asset management company based in London, said its IT systems successfully caught a junior employee who attempted to make a suspected $150 million rogue trade.
- see Securities Technology Monitor article