2025 - The Future of Market Abuse Surveillance - The Surveillance Control Tower: Better insight through behavioural analytics
19 January 2021
Market abuse is perpetrated by individuals, so it logically follows that to detect market abuse it’s necessary to monitor individual behaviour and the way it manifests in trading activity. Understanding behaviour means being able to identify subtle and connected events and patterns formed over time.
Trade surveillance vendor technologies are increasingly evolving and adopting machine learning alongside more traditional rules-based approaches. Our Future of Market Abuse Surveillance publication raises the question: Can any single vendor meet the aspirations for a behavioural system and ingest and combine all the requisite data sources?
For behavioural surveillance to be effective, a rather unusual approach is needed. The traditional method involves tuning surveillance models to produce only the most abusive events for review. A behavioural approach requires a broad set of input signals that can be used to build up profiles of behaviour patterns, which can be achieved through detuning surveillance models. What was once considered a false-positive alert is now a crucial data point to understand an individual’s trading patterns, and where that is unusual and worth investigating. This represents a step away from reviewing every alert generated by surveillance tools in the full knowledge that almost everything proves benign.
Our Future of Market Abuse Surveillance publication explores whether it is time to challenge the prevailing orthodoxy that it is imperative to review every alert. This thinking led to the genesis of our ‘Surveillance Control Tower’, a bespoke analytics layer that could become the nerve centre and intelligent eyes of the Surveillance function monitoring traders and trading activity, kept alive by a Conduct and Compliance centric data lake and fed by an ecosystem of vendor solutions.
The Control Tower would consolidate trader activity, surveillance alert outcomes, communications meta-data, P&L, positions data and a host of data artefacts, for example, conduct breaches, which have value in building a profile. This would allow a trader-centric view to be created covering a wider range of behaviours.
While the fictional vision of surveillance we have set out in our publication is just one possible future, technology enhancement will enable surveillance to deliver more insight through a focus on the classification of patterns of individual events, namely behaviours.
Download our ‘2025 - The Future of Market Abuse Surveillance’ publication for more in-depth analysis of this theme and others relating to the evolution of surveillance.