School for (Avoiding) Scandal10.06.2014
Five steps Compliance Officers should consider when choosing a market surveillance system.
By Theo Hildyard
A recent article in the FinOps Report – Chief Compliance Officers: Five Steps to Hiring the Right Team – started me thinking. It said that compliance officers: “may as well act as though you’re responsible for everything, because you’re going to have to live with the results.” This has never been truer than today.
With law enforcement agencies such as the US’ FBI and the UK’s Serious Fraud Office watching financial markets, the game just got more dangerous. Compliance officers have the responsibility to ensure their firms do not cross the line into fraudulent or manipulative behavior. And they are now more accountable than ever if it does happen.
Changes to corporate culture, practices and technology are needed or more people will end up in prison. Detecting “unwanted behavior” such as rogue trading, system errors, insider trading, front-running, wash trading or quote stuffing will help them prevent a raft of trade-related meltdowns from UBS to Knight Capital. Monitoring trader behavior can help avoid another LIBOR or FX fixing scandal.
So, I propose five steps that compliance officers can take to make choosing their surveillance software easier.
Step 1 – Buy vs Build
Consider the extent to which you want to buy versus build. You need to decide this first. Do you have the team to build something 100% bespoke? Many firms do not have either the experience or the budget. But equally, a commercial off-the-shelf system can too restrictive, especially for global multi-jurisdiction operations. Maybe you are looking for a middle-path – a hybrid of the two.
Step 2 – Pace of Evolution
Consider how easily you can evolve your technology in face of the next big scandal. Look at FX – out of nowhere it became the hot topic when regulators began to investigate traders with around 10 banks suspected of manipulating the 4pm FX benchmark assessment.
FX joins LIBOR, the gold fixing, short selling, front running, rogue algos and dark pools in the School for Scandal. There is always another transgression just around the corner.
Step 3 – Profiling Behavior
Consider how to best use big data analytics on historical data to find patterns and trends that can enhance on-the-fly detection of anomalies. Profiling of ‘normal’ behavior from historical data can give highly granular insights into activity by trader/instrument/client/time of day that can materially improve real-time alerting.
These insights can elevate alerting from one-size-fits-all parameters, with high volumes of noise, to granular alerting using logic and calibration tailored to specific trader/instrument/client combinations. Better results, less false positives, more productive teams – a win-win.
Step 4 – Interface to GRC
Consider how comprehensive a case management platform you need. Do you want to wire your surveillance activities into the wider governance, risk management and compliance structure or have it stand alone? To what extent is the wider GRC structure capable of handling on-the-fly output from surveillance?
Step 5 – To Host or not to Host
Consider a hosted versus on-premise system. A hosted system is necessarily cookie-cutter in order for vendors to achieve economies that can be passed on. An on-premise system doesn’t have to be expensive and would be better suited to firms who want ownership of the future direction of their surveillance.
There is obviously far more that could be said here and, in the current regulatory climate, “Five Steps” could easily have been “50 Steps”.
But if you start with this road, you might just avoid scandals and keep your company and your boss out of the headlines; not to mention the attention of people in dark suits who are usually found dealing with serious crimes and organized syndicates.
For more information on Software AG’s Market Surveillance Solutions click here