Technology

Currency Traders Are Becoming ‘Algo DJs’ as AI Alters Roles

Employees of BharatPe work at the company headquarters in New Delhi, India on Tuesday 05 October 2021. Photographer - Anindito Mukherjee/Bloomberg (Anindito Mukherjee/Bloomberg)

(Bloomberg) -- Algorithmic trading is radically changing the role of currency dealers, leading them to question who will be responsible for the growing reliance on computer models.

Buy-side currency traders say the advancements in technology have already altered their day-to-day jobs, but they are skeptical that artificial intelligence will fully eradicate the need for human oversight, market participants said at the TradeTech FX conference in Amsterdam this week.

“Let’s be honest about it, at least for buy-side asset managers, we’ve become merely algo DJs at this point,” said Alan Martin Lucero, currency trader at Norges Bank Investment Management, Norway’s sovereign wealth fund. “About 85% of trading processes could be successfully automated using technology we currently have at our firm.”

Trading in the $7.5 trillion-a-day global currency business is increasingly being handled by machines, with hedge funds, asset managers and central banks lured by cheaper transactions and greater anonymity. This is driving a technological arms race between banks such as BNP Paribas SA and Citigroup Inc.

Algorithms are already handling more than 75% of the trading in spot markets for currencies. That means less and less people will be required to run a trading desk, and those remaining will need to have other skills, Lucero said.

“These developments will give rise to a new kind of role, some kind of jack of all trades,” he said. “We’ll be expected to be experts in many areas of the business from legal to settlement, transaction cost analysis to execution.”

Others at the event, Europe’s biggest foreign-exchange market conference, cautioned that artificial intelligence will not make the human trader completely redundant, pointing to questions of who will be responsible if things go wrong. Algorithmic programs were blamed for exacerbating a “flash crash” in the yen in 2019, and for a plunge in the pound in 2016.

“If I haven’t built the tool, is it my fault if it goes wrong? Is it compliance? Is it the fund manager? So figuring all of that out is a huge part for us, before we really go down the AI route,” said Thomas Roberts, foreign-exchange dealer at abrdn plc.

©2024 Bloomberg L.P.

Top Videos