Finance workers are 10 times as likely to share inside information and make inappropriate comments during phone calls and video chats than over email and other text-based platforms, according to new research that also highlights the persistence of bad behaviour in some corners of the industry.
Behavox, a New York-headquartered company that uses artificial intelligence and natural language processing to flag problematic conversations for clients, analysed data from 20 financial services firms through the pandemic.
One conversation flagged as problematic involved an employee boasting about how a “few engineered tweets” could send prices soaring — while the other participant praised him for being a “master manipulator” and dubbed him “the Elon Musk of food stores”.
Text communications have been a frequent source of past embarrassment in the industry as investigations into misconduct uncovered gems such as “if you ain’t cheating you ain’t trying”.
“Rogue traders are wising up and moving their conversations to video platforms and the telephone from written text communications,” said Erkin Adylov, founder and chief executive of Behavox. “Voice platforms are the new dark alley of financial fraud.”
Anonymised transcripts shared with the Financial Times also include a chat where an employee shares a tip with another market participant which he says is “something to bet on” and then asks for an order to be “put through on Mum’s account . . . I don’t want any buys linked to my account”.
In another example, one employee gave another market participant advance information about an order that he had been just asked to place, which he said would “skyrocket the price” of a company’s stock. Both participants in the chat then placed trades to buy the same stock before the client, so they could benefit from the higher price a few hours later.
The proportion of problematic messages was still tiny — 0.0024 per cent of voice-based communications were judged concerning in 2021, compared with just 0.0002 per cent of text communications — but show the persistence of worrying behaviour across the finance industry.
Transcripts from Behavox also highlighted the discrimination still rife in some parts of the sector, despite years of pledges to make workplaces more inclusive for women and minorities.
In one conversation, a worker advised a colleague “just don’t go getting pregnant or anything like that, you’ll upset him [your boss].” When the colleague asked why her boss would be upset she was told “he’s been let down by great women in his team getting pregnant and preferring that life. He didn’t like it.”
In another discussion, a worker talks about not getting a role he applied for and is told “no doubt you would’ve got it if you weren’t white”. He replies: “Ha ha, so true. It’s all box ticking isn’t it.”
Behavox said the proportion of malicious and inappropriate comments it found had stayed roughly stable over the past three years, but that the absolute volume of problematic comments in voice chats had increased dramatically because voice was used so much more as the pandemic forced people to work from home.
Industry-wide data show that voice-based communications across 100 financial services companies was 107 per higher in the first 11 months of 2021 versus all of 2019. Text based communications grew 18 per cent over the same period.
Fahreen Kurji, Behavox’s chief customer intelligence officer, said regulators were “taking their eye off the ball” on voice, pointing to guidance from US regulator Finra suggesting companies do a random sampling of just 1 per cent of their voice data “to flesh out misconduct”. European and UK rules require firms to “periodically monitor” phone conversations.
