The Financial Industry Regulatory Authority's process to identify high-risk brokers is evolving with market activities, according to officials with FINRA's High Risk Representative Program.
What makes a rep high risk? What is FINRA looking at when it's making this determination?
Brooks Brown, senior director, and Eric Hebert and John Salerno, investigative directors in the high-risk program, set out the parameters in a recent FINRA Unscripted podcast episode, "Keeping an Eye on Individuals Posing a Heightened Risk of Misconduct."
A FINRA spokesperson told ThinkAdvisor in separate comments that there's an "overlap between the types of registered representatives that the HRR team focuses on and the registered representatives who have the types of U4 disclosures that trigger Restricted Firm designation under Rule 4111."
FINRA has "a number of different ways" to spot such individuals, Brown said, including "a machine learning model" to gauge reps' "predicted likelihood to engage in future misconduct."
The model also "ranks those individuals using a number of different factors and attributes that are part of each representative's risk profile, such as complaint disclosures, termination disclosures, financial disclosures, could be associations with risky firms, and it could be their associations with other risky representatives," Brown said.
The machine learning aspect "views more recent events and associations, and weighs those more heavily than events that occurred many years ago," Brown explained.
FINRA has made improvements to the model "over the years to improve its accuracy, whether by adding additional data elements to it or tweaking the machine learning approach."
However, FINRA also relies on "internal intelligence and interactions with various parts of FINRA to identify other individuals who may be high risk and may warrant a detailed review based on what FINRA is observing," Brown continued.
FINRA is "very mindful of the notion that birds of a feather flock together, and association risk is certainly a significant factor that we look for to identify problematic individuals," Brown said.
