How the SEC is using automation and targeting for exams
SEC examiners have new tricks up their sleeves and others under development to help make inspections quicker and more effective, according to the regulator.
“Markets are supposed to make sense. And when you see anomalies in the market, it’s probably a place where we should look further — whether it’s the way a firm operates or a particular type of trade,” Chairman Jay Clayton said Tuesday night during a panel at New York University. “The use of big data to identify those anomalies has really helped us.”
Clayton and top officials from the SEC’s Enforcement Division and Office of Compliance Inspections and Examinations hinted at the agency’s methods during the hour-long discussion. Tools like automated trade and text analysis, along with a more targeted approach, have changed SEC oversight.
MORE WITH LESS
The SEC projects OCIE exams of RIAs, broker-dealers and others to grow by 1% this year to 2,449 and another 7% to 2,609 in 2018 despite its trimmed budget request for the unit. Compliance experts have predicted fewer but more focused SEC and FINRA exams, noting records set for enforcement fines last year.
“We’re really trying to leverage our staff using technology to get the biggest bang for our buck and get the best coverage, all the while helping investors,” Peter Driscoll, OCIE’s acting director, told an audience that included former SEC Chairwoman Mary Jo White, law students and securities lawyers.
OCIE tech advancements in the past five years have resulted in a “revolution” over its capabilities compared to when Driscoll started there in 2004, he added. Driscoll described the regulator’s national exam analytics tool — known as ‘NEAT’ — as its most significant tech accomplishment to date.
Rather than trying to analyze trade blotters using spreadsheets, the tool ingests the data from one or more firms to run over 50 reports at once. NEAT also uses a so-called "news feed" to probe whether firms carried out big transactions in the days leading up to important announcements and other news.
Driscoll also revealed text analytics projects the regulator plans to use for looking at firms’ compliance manuals, websites and Form ADV Part 2 brochures. The tools would allow OCIE to compare one firm's filings to those of others, along with searching for problematic words like “guarantee” or “risk-free.”
Lastly, OCIE’s risk analysis examination team has launched tools allowing it to cull through trading data from clearing firms more easily and quickly. The small team can analyze trades by 300 to 400 firms at a time through the capability, Driscoll said.
“That leads to exams. It leads to potential enforcement referrals,” he said. “It’s a tremendous tool and work that they do on a large-scale basis, and it’s seven people.”
The Enforcement Division has embraced similar tools in its efforts to bring actions against rogue brokers and firms, according to Stephanie Avakian, the division’s co-director. Automated analysis has figured into cases involving insider trading, churning and cherry picking, among other types, she said.
“It does the work of many people that would normally take eons to do in a very short, compressed time frame,” Avakian said. “And the cases are very compelling on the data.”
The shifts at the SEC discussed on the panel extended beyond tech tools. OCIE’s risk-based approach has led to some “fundamental changes on how we conduct our exams,” Driscoll said. Examiners now look closely at one or two issues rather than many of them like they did 10 years ago, he said.
The method has helped make exams easier for all involved, according to Driscoll.
“You’re seeing us ask for a lot more documents ahead of time. So, when we come in, assume that we’ve read the documents, assume that we know the documents and assume that we have a list of questions to ask you,” he said.
“That makes for a much more efficient fieldwork process for us and much less of a disruption for you all when we show up at the door.”