RBC rolls out AI-based trading platform

Global clients can deal in U.S., Canadian equities with help of algorithm

Royal Bank of Canada is launching an artificial intelligence-based electronic-trading platform to clients around the world, starting with an algorithm designed to help reduce trading slippage.

The Aiden trading platform, the product of five years of research and development, will now be available to global clients trading U.S. and Canadian equities, the Toronto-based bank said Wednesday. Some Canadian clients have already been using the system and some U.S. clients had been beta testing it.

The platform’s first product is an algorithm to help clients execute trades as close as possible to the day’s volume-weighted average price. To hit that target, which moves as the day progresses, Aiden employs “deep reinforcement learning” that doesn’t require recoding like traditional machine-learning techniques, said Shary Mudassir, co-head of global equities execution for RBC Capital Markets.

“That form of AI allows us to very intelligently navigate the challenges of a market that is very fluid and dynamic,” Mudassir said in an interview.

“You can’t just rely on traditional approaches of machine learning that only rely on historical data but don’t learn from the dynamic nature of the world with which they’re engaged.”

Aiden was jointly developed by RBC Capital Markets and the Borealis AI research centre, which is backed by the bank and led by inventor Foteini Agrafioti. RBC and Borealis have previously introduced AI-based solutions for the bank’s retail operations.

RBC’s Aiden trading platform, based on artificial inteliigence, is the product of five years of research and development.