JAX-LOB: A GPU-Accelerated limit order book simulator to unlock large scale reinforcement learning for trading

TL;DR

JAX-LOB is a GPU-accelerated limit order book simulator that enables large-scale reinforcement learning research in trading by processing thousands of books simultaneously with high efficiency.

Contribution

It introduces the first GPU-enabled LOB simulator, JAX-LOB, optimized for parallel processing of multiple books without sacrificing realism.

Findings

Supports processing thousands of books in parallel

Reduces per-message processing time significantly

Demonstrates integration with RL training workflows

Abstract

Financial exchanges across the world use limit order books (LOBs) to process orders and match trades. For research purposes it is important to have large scale efficient simulators of LOB dynamics. LOB simulators have previously been implemented in the context of agent-based models (ABMs), reinforcement learning (RL) environments, and generative models, processing order flows from historical data sets and hand-crafted agents alike. For many applications, there is a requirement for processing multiple books, either for the calibration of ABMs or for the training of RL agents. We showcase the first GPU-enabled LOB simulator designed to process thousands of books in parallel, with a notably reduced per-message processing time. The implementation of our simulator - JAX-LOB - is based on design choices that aim to best exploit the powers of JAX without compromising on the realism of LOB-related mechanisms. We integrate JAX-LOB with other JAX packages, to provide an example of how one may address an optimal execution problem with reinforcement learning, and to share some preliminary results from end-to-end RL training on GPUs.