Random Sampling does not work to efficiently detect Market Abuse in 2021.

6th February 2021 by Brielle Hewitt

Back to

The basis for the random sampling of Electronic Communications to discover bad actors has always been sketchy at best, evidenced by its dismal hit rate in leading to convictions and fines. Now, the shift to remote work and proliferation of communications channels has made it even more unreliable. 

There’s an old joke about a scientist walking to his car after a late night in the lab. Halfway across the car park, he drops his keys, so he walks back to the door of the lab and begins frantically looking for the keys. A colleague saw him return and asks what he’s doing: 

“I dropped my keys in the car park.” 

“Then why are you searching by the door?” his colleague asks, perplexed. 

“Because this is where the light is.” 

It’s hardly Seinfeld, but it illustrates a useful principle. When looking for things, humans have a tendency to search where it is easiest to look, not where they are most likely to find what they are actually looking forThis feels increasingly true in the financial services sector.   

Especially when it comes to how firms meet the recording obligations as set out in the FCA’s Senior Management Systems and Control Handbook a.k.a SYSC Rule 10A 

With the proliferation of electronic communications channels and increasingly geographically dispersed teams, random sampling feels like an approach chosen for the single reason that compliance teams can implement it, rather than being an approach that will genuinely discover wrongdoing. 

Proactive monitoring via random sampling is a long-standing approach to compliance, but the growth in electronic messaging has created profound challenges. The volume of communication has become a haystack in which the needle of wrongdoing is highly unlikely to be discovered by chance  

As long ago as 2008, lawyers were raising concerns about the costs call recording and sampling was placing on small firms, while questioning its efficacy for securing convictions.  

Today there are many more communications channels available than in 2008 – WhatsApp wasn’t launched until 2009 – and when compliance departments seek to monitor by random sample unstructured data from the newer communications channels, they face significant challenges in stamping out suspected bad behavior, as we saw late last year. 

Covid driven business change has only increased the complexity of capturing and sampling communications, as firms have been forced to shift to remote working for months at a time. As the FCA noted in Market Watch 66, homeworking significantly increases the threats firms face. 

“Risks from misconduct may be heightened or increased by homeworking. This includes increased use of un-monitored and/or encrypted communication applications such as WhatsApp for sharing potentially sensitive information connected with work. Use of such apps can present challenges and significant compliance risks, since firms will be less able to effectively monitor communications”. FCA Market Watch 66 – read our blog on the FCA’s expectations here 

Before the pandemic, industry experts were arguing that manual surveillance was inadequate to meet both regulatory and evidential requirements Of course, proactive monitoring can be extremely useful in constructing a case once wrongdoing has been discovered, but typically the initial investigation is spurred by whistle-blowers or trade surveillance rather than happened across by random sampling. 

Random Sampling is expensive and time consuming for compliance officers, played against the fact that they typically fail to discover wrongdoing. So, what can compliance officers do? 

With the multiplicity of communication channels now in general use, and the fact that all data needs to be supervised the task is far too big for an individual or team handle, but new technology offers a way out.  

Voice-To-Text, Natural Language Processing and Machine Learning mean proactive monitoring can become truly proactive. The Fingerprint Platform analyses and risk ranks each and every item of communications.  Our technology can filter, analyse and escalate suspect behavior, allowing firms to focus their efforts only where they are most likely to make a difference. 

We can take you through how Fingerprint can help your firm move away from manual, ineffective process, and empower your compliance team to find more effective ways of running day to day. Drop us an email: hello@fingerprint-supervision.com

Related posts