Transaction reporting may have started off as a baby elephant but as regulators increasingly focus on the importance of accurately recorded data, and the regulatory landscape becomes ever more complex, that animal is getting bigger. MiFID II and MiFIR transformed the scope of transaction reporting for financial firms, requiring them not only to guard against market abuse but monitor the ongoing functioning of the market – and with new regulations such as SFTR and CAT on the horizon, and swingeing fines (see UBS and Goldman Sachs earlier this year) exacted for non-compliance, effective systems have become essential.
The UK’s FCA, for example, has been emphatic in its approach. ““Effective market oversight relies on the complete, accurate and timely reporting of transactions,” it asserted in response to the Goldman Sachs fine. “This information helps the FCA to effectively supervise firms and markets. In particular, transaction reports help the FCA identify potential instances of market abuse and combat financial crime.”
“There is an increased emphasis on transaction reporting these days, and especially on the timeliness of transaction reporting,” agrees James Corcoran, CTO Financial Services at data analysis software developer Kx. “It’s getting more and more difficult for financial firms, many of whom are now having to report on transactions almost as they occur, in real time. It’s a tricky technology problem to solve.”
Kx, which provides the world’s fastest time-series database, has been around for a while (since 1993 to be precise) and knows a thing or two about financial services technology – especially around the process of capturing and processes large volumes of data. From a background working in the front office on trading models and market data streams, it has been but a short step to addressing compliance functions and integrating RegTech solutions to meet their clients’ evolving demands.
“We started off with a fast, powerful, and efficient database to process market data – and over the years, we’ve expanded that use case to bring in transaction records such as emails, phone calls, and instant messaging as well. Combining multiple data streams and sources, we’ve been able to develop fairly complicated models to detect and prevent abuse by either clients or employees.”
This has translated into a consolidated surveillance product designed to assist the firm’s extensive client base to meet their compliance requirements through the capture and analysis of transaction records. Developed in partnership with clients over the past 12 months, the solution is now live but has not yet been formally marketed.
“One of the biggest challenges we see is the large quantity of alerts that are generated through most systems. Larger customers are getting hundreds or even thousands of these a day – to the point where they simply don’t have the time or the staff to investigate every single case,” says Corcoran. “We’re helping customers solve that problem by applying machine learning in order to reduce the number of false positive alerts they generate and to enable them flag the highest priority alerts for immediate investigation.”