Source: Eventus Systems
Eventus Systems, Inc., a provider of innovative regtech software solutions for the capital markets, today announced that it has integrated machine learning (ML) into its Validus surveillance and risk software platform, augmenting capabilities for identifying potential disruptive trading practices such as spoofing and layering.
Eventus CEO Travis Schwab said: “We’ve taken a unique approach, adding a machine learning module to our already strong procedural market manipulation detection tool. This integration creates smarter, more actionable alerts that enable clients to dig deeper and take resolution action faster. We are the first in this space to offer clients such an intuitive, user-friendly choice, providing this new machine learning capability as a no-cost, optional overlay to our existing product. In Validus, we apply ML as an enhanced layer to existing alerting, thereby highlighting the most pressing issues that need to be addressed.”
Validus helps clients improve their compliance capabilities, identifying and monitoring any potentially suspicious or risky trading behavior, as well as creating alerts and intelligent resolution based on parameters identified by compliance, risk management, trading desks and operations personnel. The machine learning technology can further hone the parameters automatically without requiring user input, enhancing the quality of alerts and reducing already stretched resources.
Schwab said: “Overall, it’s about applying the right technology for the surveillance job at hand. Some procedures, such as Wash Trades or Order Marking, would not be productive by employing ML technologies. This is why we make sure and apply these technologies to problems that can benefit from their unique advantages.”
The firm kicked off this project several months ago by first conducting extensive supervised learning on existing scored data sets. As more and more customers upgrade to the new software, they can score their own results, which creates continuous improvement in the learning process. The firm expects to quickly expand on the use cases for applying machine learning approaches now that the data and infrastructure is in place.
Validus includes features such as sophisticated alerting and visualization; a suite of tools for advanced reporting; delivery of reconciled, multi-party counterparty data; monitoring of key trading functions; and the ability to scale the platform across asset classes and connect into existing systems to meet client needs.