RBC’s activity in AI began in 2016 with the establishment of Borealis AI, an AI research centre. In partnership with Borealis AI, the bank integrated various AI-powered features into its consumer banking products, including bill scanning and predictive analytics for spending. These innovations earned RBC the award for Best Use of AI for Customer Experience at the Digital Banker Digital CX Awards. However, RBC’s commitment to AI innovation extends beyond consumer banking.
Introducing Aiden: AI-powered trading
Borealis AI has since merged with RBC’s data platform and architecture teams to form RBC Borealis. This team leads AI research and development across the bank, with the Aiden trading platform standing as a flagship product in intelligent trading solutions.
Initially launched in 2020, Aiden is RBC’s AI-powered electronic trading platform. The platform uses deep reinforcement learning to learn about stock trading and automatically execute buy and sell transactions in the client’s best interest. Aiden can adapt to changes in the stock market and adjust trade executions accordingly.
VWAP: Aiden’s original execution algorithm
A key component of Aiden, or any other automated trading system, is the execution algorithm. This is the set of rules Aiden follows when executing trades. Aiden’s original execution algorithm was based on the Volume-Weighted Average Price (VWAP) strategy.
Through deep reinforcement learning, Aiden VWAP is optimised to reduce slippage against the VWAP benchmark. This means Aiden will execute trades at prices close to the average price of a security, weighted by volume, over a given period. The VWAP approach helps minimise slippage—the difference between the intended execution price of a trade and the actual execution price—by aligning trades with the market’s natural liquidity patterns.
Reinforcement learning in Aiden’s trading strategy
Reinforcement learning is a machine learning technique where an agent learns to make decisions by interacting within its environment. It employs a trial-and-error approach, where the agent receives feedback in the form of rewards or punishments based on its actions.
How well do you really know your competitors?
Access the most comprehensive Company Profiles on the market, powered by GlobalData. Save hours of research. Gain competitive edge.
Thank you!
Your download email will arrive shortly
Not ready to buy yet? Download a free sample
We are confident about the unique quality of our Company Profiles. However, we want you to make the most beneficial decision for your business, so we offer a free sample that you can download by submitting the below form
By GlobalDataUnlike traditional VWAP algorithms that follow static rules, Aiden VWAP can adapt to real-time market conditions by leveraging deep reinforcement learning. It processes over 200 inputs through a deep neural network, allowing it to learn from each action and continuously optimise execution quality. This adaptability enables Aiden to navigate complex and volatile market environments effectively, delivering improved trading outcomes for clients.
Arrival: The next step in intelligent trading
Building on the success of its initial VWAP algorithm, RBC introduced a second execution algorithm in 2022 called Aiden Arrival. Aiden Arrival functions similarly to its predecessor; it increases the number of inputs from 200 to 300 and swaps the VWAP benchmark for the arrival price benchmark. The arrival price represents the market price of a security the moment an order is initiated and serves as a critical metric for evaluating execution performance. Reducing slippage against the arrival price is particularly difficult when executing large trades that can impact the price.
Related Company Profiles
Royal Bank of Canada