Is Fully Holistic Surveillance the Holy Grail?
Under MAR and MiFID II, firms are required to detect and report unlawful behaviour in a timely manner by putting preventative measures and solutions in place. To achieve compliance, first and second line surveillance functions need to be created that provide more holistic, forward-facing surveillance solutions.
Regulators now expect market participants to be able to identify exceptions, trends and patterns and to capture related data such as trades, holiday schedules, emails, phone calls, instant messages and market prices. For many firms, there is still a large gap between their current combination of surveillance activities and a fully automated and holistic surveillance solution that can cover all asset classes, products and functions.
A holistic surveillance solution sources both structured and unstructured data to form a broad view of relevant trading behaviours and decisions and the patterns that link relatively innocuous phone calls or IM chats with transaction decisions.
Many firms will initially see use of such a solution to be expensive, difficult to implement, and perhaps, even operationally impossible. However, much can be achieved using any or all of the advancing technologies such as Big Data Analytics, Artificial Intelligence (AI) & Machine Learning (ML) and voice recognition & transcription in combination with firms’ legacy surveillance tools and vendor solutions available in the market that offer trading, trader, and eCommunication surveillance. Integrating these technologies creates the opportunity for firms to develop and deploy powerful holistic surveillance solutions, gain additional focus and target attention on abnormal activities.
Technology deployment is not a panacea that will resolve all surveillance needs – human intervention remains a necessary component. As per the European Securities and Markets Authority in the MAR Technical Standards, “The arrangements, systems and procedures to be put in place and maintained under this Regulation shall allow for human analysis to play an important role in the monitoring, detection and identification of suspicious trading activity, including transactions and orders that could constitute market abuse or attempts of market abuse.”
AI and ML have developed significantly in recent years and there is certainly a growing desire within the Capital Markets to find realistic opportunities for the mainstream implementation of such technologies. While they won’t replace human surveillance activities, they will certainly supplement them by, for example, adjusting the alert parameters to reduce the number of false positives that are being alerted. ML techniques can be leveraged to develop profiles of trader activities by analysing data to determine a trader’s normal profile and by greatly improving the probability of identifying abnormal activities.
The key to holistic surveillance is not the systems – it is sourcing and contextualising the data to understand the story it is telling. The combination of the structured data (transactions, market data) and unstructured data (messages, emails, voice messages) delivers the true holistic surveillance outcome. Individually, these datasets may not appear to provide anything of significance but together they tell a whole new story. However, blending structured and unstructured data from multiple sources presents a whole new challenge.
GreySpark believes that surveillance is the next big topic. Regulation has certainly increased visibility on surveillance activities within firms, and many software vendors are developing and expanding their surveillance tools to help them address the shortfalls they may have. Holistic surveillance may be a pipedream for many firms, but for some institutions, the goal may be much simpler and more achievable. Surveillance does not have to be all-or-nothing; a step-by-step approach towards holistic surveillance integration can deliver significant improvement to legacy surveillance activities. Financial institutions, the buyside and the sellside have been making significant investments to ensure their regulatory compliance – GreySpark expects the attention will now pivot towards refinement, automation and improvement.
Rogue conduct is unavoidable. Surveillance will continue to be a combination of prevention and detection, evolving to adapt to the complex and changing market structures and the technologies that the markets now demand.