Deep Learning Meets Queue-Reactive: A Framework for Realistic Limit Order Book Simulation
This work provides an incremental improvement for researchers and practitioners in finance who need accurate simulations for strategy development and backtesting.
The authors tackled the problem of simulating limit order books more realistically by extending the Queue-Reactive model to include dependencies between price levels and richer market features, resulting in a model that captures key market properties like the square-root law of market impact and realistic order size patterns while maintaining computational efficiency.
The Queue-Reactive model introduced by Huang et al. (2015) has become a standard tool for limit order book modeling, widely adopted by both researchers and practitioners for its simplicity and effectiveness. We present the Multidimensional Deep Queue-Reactive (MDQR) model, which extends this framework in three ways: it relaxes the assumption of queue independence, enriches the state space with market features, and models the distribution of order sizes. Through a neural network architecture, the model learns complex dependencies between different price levels and adapts to varying market conditions, while preserving the interpretable point-process foundation of the original framework. Using data from the Bund futures market, we show that MDQR captures key market properties including the square-root law of market impact, cross-queue correlations, and realistic order size patterns. The model demonstrates particular strength in reproducing both conditional and stationary distributions of order sizes, as well as various stylized facts of market microstructure. The model achieves this while maintaining the computational efficiency needed for practical applications such as strategy development through reinforcement learning or realistic backtesting.