Sell-Side Technology Awards 2020: Best Sell-Side Market Surveillance Product—NICE Actimize

By winning the best sell-side market surveillance category in this year’s SST Awards, NICE Actimize created a bit of history: It became only the second company to have won the category since Nasdaq started its domination back in 2014. The Hoboken, NJ-based firm’s win is thanks to its Surveil-X platform, an artificial intelligence-powered, cloud-native offering designed to allow firms to monitor and detect anomalous trading behavior in compliance with global regulations.  

Trading and communication surveillance has always been a challenge for capital markets firms, given the volumes of data they need to capture and analyze in ever-decreasing timeframes. That undertaking became significantly more complex in the early days of the Covid-19 pandemic as employees hunkered down across a variety of locations, including existing offices, traditional disaster recovery sites and their homes. Such moves posed operational challenges to capital markets firms and the tech providers supporting them, although, according to Lee Garf, general manager of NICE’s compliance line of business, the firm adjusted well to the new dispensation. “In terms of scale and performance, we had no problem at all,” Garf explains. “We did see significant spikes in trading volumes—we witnessed multiples of traditional trading volumes—but the platform being cloud-native scaled without a problem. We did hear that some providers had scalability issues, and the problem is that if it takes more than 24 hours to process and analyze communication and trades, you quickly get behind and you can’t catch up.”

Another challenge facing firms when monitoring trades and communication is their surveillance platforms producing false positives, all of which need to be investigated. NICE Actimize’s solution to that problem entailed embedding machine learning technology within Surveil-X, which, Garf says, resulted in two use-cases among its clients. “The first is model tuning,” he says. “One of the unique capabilities of Surveil-X is the ability to tune [surveillance] models without the need for extensive services. Parameters can be adjusted to clients’ specific workflows within the platform.” 

Different capital markets firms trade different volumes across a range of asset types, which means each has slightly different surveillance requirements. Surveil-X’s machine learning functionality supports that customizability by evaluating large numbers of parameter permutations in order to establish the right balance between true and false positives. “The second use-case is around anomaly detection,” Garf continues. “It’s really important to use both traditional expert rules and machine learning anomaly detection to identify potential market abuse. We need to look for known behaviors, but where machine learning helps is when it comes to unknown behavior.”