Spotting a Rogue Trader

When the news broke that UBS equity trader Kweku Adoboli had racked up $2.3 billion in unauthorized trades, everyone had questions. Why did UBS' risk systems fail to find this rogue? How could a single young trader go rogue and the bank not know? Again? Advanced Trading spoke with James Heinzman, managing director of Global Compliance Solutions, for some answers.

Advanced Trading: Kweku Adoboli was an equities derivatives trader who went rogue for months before he turned himself in. UBS had no idea until he turned himself in lst week. How is this possible? Was this an error of his human managers or the failing of UBS' risk systems?

James Heinzman, Global Compliance Solutions: Trading losses such as these don't just pop up overnight. They are usually accumulated over an extended period of time. Traders or employees with knowledge of how back office and internal risk controls work are able to circumvent those controls over a period of time to hide unauthorized trading activity. Often this activity involves the creation of fake orders that never settle.

One example of this kind of rogue trading involved a trader who would accumulate a large position in a security and at the end of the day would write a series of sell tickets equal to 95 percent of the position. When the firm's reconciliation process ran overnight, it would deduct the sell transactions from the real position and only reflect 5 percent of the real position on the books. The trader repeated this day after day, cancelling the fake sell transactions each day and rebooking them with updated quantities to mask the real underlying position. This kind of activity is difficult to detect without the correct monitoring system in place. Managers may have been able to see certain red flags such as higher than normal profits for a trader with a relatively conservative investment profile, or a high number of cancelled and corrected trades.

There may also be some control processes that could have deterred (are at least made it harder) to perpetrate this behavior such as a separation or order entry and order correction functions, enforced two-week vacation policy, or settlement date position reconciliations, etc. Systems such as ours are designed with specific analytics to monitor trading activity, as well as sophisticated profiling techniques to identify unusual or aberrant activity combined with the ability to connect data from different silos and systems to ensure you are able to connect all the right dots.

Advanced Trading: Traders are more tech savvy than ever before. How can IT safeguard their systems from traders who once worked in the back office? What can they do to better shore up their defenses?

Heinzman: The first step, which I am certain all the banks are in the process of doing now, is to review current policies, procedures, and processes to identify any potential gaps or weaknesses. The next step would be to remediate any of those gaps and start to formulate a plan of how to monitor and identify unauthorized trading activity.

There are two specific monitoring tools that should be used: Specific monitoring algorithms to identify known behaviors that may be suspicious or inappropriate such as excessive trade corrects, extended settlements, and high cost to carry ratio, and behavior profiling to identify unknown behaviors that are statistically different than historical or expected behavior or those of peers. This combination of surveillance and profiling techniques is foundation that our compliance and fraud monitoring systems are built upon. Banks will also need to identify and collect all the data necessary to enable these monitoring capabilities. They will need to consider not just the standard data sets such as orders, executions, and positions, but also sources such as system access, HR records, and clearance and settlement data.

Putting together a comprehensive plan to analyze, scope, and deploy a system to adequately monitor for rogue traders is no small undertaking, but it will be well worth the investment. It is difficult to quantify a bad thing that doesn't happen because you have good controls in place ... that is, until it happens to someone else. Additionally, training of back office personal, clear escalation procedures, and training around red flags and how to respond will enable firms to better identify and manage emerging issues

Advanced Trading: What are the long-term effects of this rogue trading incident? Will banks pay lip service to this problem or will they make real changes? Does every bank have a rogue trader out there? Or are they whistling past the graveyard?

Heinzman: The long term effects of this event will be more profound than the last one. I think that regulators will now put pressure on firms to make real and lasting changes to how they approach this issue. The answer is yes ... and no. After Societe Generale happened a few years ago, all the firms were on "high alert" and had lots of meetings to discuss how this happened and what could be done to make sure it didn't happen to them. I was in many such meetings during that time, and I think there were some really positive ideas raised, processes proposed, and systems reviewed -- all well intentioned, but then the financial crises hit in full force in 2008. The industry was caught in a tremendous amount of turmoil and the focus on this particular topic may have been somewhat lost.

This problem is completely solvable, but it will require an investment of time and resources to deploy a targeted monitoring and control system. While it is possible the industry may have lost some focus on this issue during the financial crises, I am confident that an event like this will bring renewed focus and attention.

ABOUT THE AUTHOR

Phil Albinus is the former editor-in-chief of Advanced Trading. He has nearly two decades of journalism experience and has been covering financial technology and regulation for nine years. Before joining ...

See more from Phil