THOUGHT LEADERSHIP

RBC’s Aiden VWAP: A new era of AI trading in Europe, leveraging advanced AI and deep reinforcement learning to achieve optimised execution

AI is not merely a tool, but a catalyst for change setting new standards and expectations. Amidst this transformation, the need to stay agile is paramount, writes RBC Capital Markets’ European head of multi-asset agency solutions, James Hilton.

In competitive financial markets, where every trading action can make or break a trade, advances in AI and deep real-time reinforcement learning are revolutionising algorithmic trading strategies. The accelerating impact of these technologies on low-touch trading has been profound, as market dynamics drive demand for precise execution, adaptability and efficiency.

To address trading execution issues such as slippage and alpha erosion through periods of volatility that could upend the best historical models, RBC and its Borealis AI research institute developed Aiden – an electronic trading platform that uses AI and deep reinforcement learning to adjust to market conditions in real-time. Making its debut in North America in 2020, Aiden proved its mettle by swiftly adapting to volatile market changes while preserving performance, without the need for frequent manual re-coding. Now launched in UK and European markets, the platform continues to adeptly navigate the complexities of dynamic market conditions, with the goal of delivering improved trading execution quality and insights for clients.  

Solving for VWAP with advanced AI

One key area of focus is optimising execution performance while minimising slippage against the volume-weighted average price (VWAP) benchmark, a critical reference point for traders.  Aiden VWAP is the first algorithm on the Aiden platform in Europe – a volume-weighted average price strategy that uses the power of AI and deep reinforcement learning to reward optimal outcomes, aiming to minimise the slippage against the benchmark.  Unlike traditional VWAP algorithms, Aiden VWAP considers a variety of different market data with more than 200 inputs that it processes through a deep neural network. Within guardrails, it is given discretion to optimise its behaviour, learning from its performance both during and after the trade. 

RBC has also developed Aiden Insights, an initial version of an explanation system for the Aiden algorithms. The RBC coverage team utilises Aiden Insights to provide real-time trading insights and explain to our clients how Aiden is adapting and making decisions to optimise its execution benchmark.

Beyond pre-set trading algorithms

By analysing both historical data and real-time shifts in market dynamics and liquidity patterns, Aiden learns autonomously, seeking to minimise slippage and execute orders at optimal prices. Unlike traditional pre-set algorithms, Aiden proactively explores new trading patterns within defined parameters. It has the freedom to discover and act upon new patterns that aim to deliver better execution outcomes. This adaptability ensures Aiden’s execution strategies remain effective across various scenarios, including periods of market volatility. 

When executing trades more efficiently, traders can avoid unnecessary price deviations. This efficiency allows greater focus on higher-value activities, such as customising strategies to meet specific individual needs, risk tolerance, and investment horizons.

Staying ahead of the curve

The Aiden VWAP algorithm marked the first foundational step of Aiden’s evolution. In 2022, RBC Capital Markets introduced Aiden Arrival Price in North America – the second AI-based execution algorithm that offers a comprehensive solution to the arrival price challenge. With an expanded set of trading actions, Aiden Arrival Price offers greater flexibility over the execution trajectory. It is designed to reduce slippage against the arrival price benchmark by utilising an enhanced reward system and leveraging over 300 data inputs. This allows for continuous performance optimisation.

We are actively engaged towards offering Aiden Arrival to our clients in the UK and Europe and we continue to work closely with clients to push the boundaries of trading technology. In the world of algorithmic trading, AI is not merely a tool, but a catalyst for change setting new standards and expectations. Amidst this transformation, the need to stay agile is paramount. With its adaptability and precision, the Aiden platform is a powerful choice for optimising trade execution, offering a competitive edge in the complexities of today’s fast-moving markets.

Find out more

Aiden VWAP is now live for UK and European clients via RBC Capital Markets sales and trading desk. 

To learn more about how Aiden delivers powerful trading intelligence at RBC Capital Markets, please visit rbccm.com/aiden. To discover more about the research behind Aiden from the Borealis AI team, head to www. borealisai.com or contact ElectronicSalesandTrading@rbccm.com.

Key points: 

  • Advanced AI has transformed algorithmic trading in Europe, as market dynamics drive demand for improved execution, adaptability and efficiency.
  • RBC’s Aiden trading platform uses deep reinforcement learning, a technique of machine learning, to adapt to different market or order characteristics and respond to changes in market conditions in real-time. 
  • A key area of focus is optimising execution performance while minimising slippage against the volume-weighted average price (VWAP) benchmark. 
  • Unlike traditional pre-set algorithms, Aiden proactively explores real time trading patterns seeking to reduce discrepancies between expected and executed trade prices.