BMLL, the leading independent provider of harmonised, continually engineered historical Level 3, 2 and 1 data and analytics for Capital Markets, today announced a partnership with Simudyne, an advanced generative AI and agent-based simulation platform designed to model and replicate market dynamics.
The collaboration will combine BMLL's L2 and L3 harmonised historical order book datasets with Simudyne's advanced Pulse simulator, used specifically for intraday markets, to enable market participants to run highly realistic, reactive market simulations, perform market replays and test system performance.
Transitioning to AI-Driven Simulation: Training AI models on BMLL’s tick-level data
Simudyne is an established leader in the development of market simulation software, built around methods such as agent-based modelling, which relies on proprietary behavioural assumptions for market participants, and other machine learning techniques. Simudyne is now pre-training state-of-the-art generative models of intraday order flow using the same techniques as underpinning language AI models, such as ChatGPT. Because these next-generation AI models do not rely on hand-crafted expert rules, their realism is strictly determined by the quality and volume of the data they are fed.
Under this partnership, Simudyne has begun training generative AI 'Large Market Models', using granular market data, enabling synthetic tick-data generation and client-specific fine-tuning alongside its established rule-based agent-based modelling capabilities. The resulting simulator goes beyond traditional, static historical replay to offer ‘reactive’ back-testing, meaning the simulated market dynamically responds and adapts to the user's executed orders in real-time.
Empowering the Ecosystem: Pre-training Advanced AI Models with Definitive Level 2 and Level 3 Data
The integration of BMLL’s granular historical data and Simudyne’s AI-powered market simulation engine will deliver critical simulation capabilities across the entire Capital Markets ecosystem.
-
Algo traders & quants can test, optimise, and refine execution algorithms in a simulated market that dynamically responds to their trading activity (reactive back-testing) rather than static historical replay. Quants can also leverage Simudyne's calibrated models and BMLL's datasets as a benchmark to optimise and train their own machine learning execution models.
-
Execution & TCA analysts can simulate large orders prior to actual execution to estimate slippage, transaction costs, and market impact, helping to select optimal order routing strategies and study how other simulated participants react to their orders.
-
Trading technology teams can connect staging and testing systems directly to the simulated exchange (via FIX connectivity or APIs) to validate system performance, latency, and end-to-end trading software setups under highly authentic, reactive conditions.
-
Exchanges can safely test and validate modifications to matching engine logic, trading protocols, and new order types against a faithful simulator replica before deploying to production. This enables risk-free prototyping of new auction mechanics, circuit breakers, and fee structures.
-
Risk & clearing teams can run highly complex liquidity risk calculations, matching the advanced risk analysis, and stress-test margin, clearing, and default-fund models against a broad range of simulated volatility events and market regimes.
Paul Humphrey, Chief Executive Officer of BMLL, said: "We are delighted to welcome Simudyne to our BMLL Activate Data Credits Programme. High-quality training data is the single most critical ingredient for sophisticated, AI-driven financial modelling. By hosting Simudyne’s simulation models directly within the BMLL Data Lab, we will be able to give our clients a safe, highly realistic sandbox to run reactive simulations on top of our definitive historical order book. Ultimately, this collaboration serves as a powerful proof-of-concept, establishing BMLL data as the premier foundation for firms to train their own AI models, while reinforcing our ongoing commitment to driving the next wave of innovation across global Capital Markets.”
Justin Lyon, Chief Executive Officer of Simudyne, said: "We are very excited to collaborate with BMLL to push the boundaries of market simulation. As we move towards data-driven, generative AI models, simulation fidelity is only as strong as the data underpinning them. BMLL’s granular, high-resolution historical market data gives our models the depth needed to learn and reproduce real market dynamics at the order-book level. Together, we are creating the next generation of market simulation environment that bridges historical market behaviour and realistic synthetic markets.”
Partnership Anchored in Simudyne's Participation in the 'BMLL Activate' Data Credits Programme
The partnership is enabled by Simudyne’s participation in the 'BMLL Activate - Data Credits Programme', an initiative designed to help partners build, test, and launch new analytics or data-driven solutions on top of BMLL's historical Level 3, 2, and 1 market data.
The program accelerates product development and market validation by offering lower to no upfront data licence costs during the build phase. Under the program, selected partners receive a defined allowance of BMLL data credits redeemable for structured access to the BMLL Data Lab and the BMLL Data Feed.
This enables Simudyne to build, train, and host its generative foundation models and multi-agent market simulation software, creating a seamless path from research and validation to commercial adoption for joint clients.
