MarketGPT: Developing a Pre-trained transformer (GPT) for Modeling Financial Time Series

TL;DR

MarketGPT introduces a pre-trained transformer model that accurately simulates financial market order flows, capturing key statistical features and enabling realistic, high-fidelity market simulations for research and analysis.

Contribution

This paper presents the first GPT-based model tailored for financial time series, capable of generating realistic order book dynamics and reproducing market stylized facts.

Findings

Successfully reproduces key features of order flow data

Captures stylized facts of real financial markets

Produces long sequences of order messages in streaming mode

Abstract

This work presents a generative pre-trained transformer (GPT) designed for modeling financial time series. The GPT functions as an order generation engine within a discrete event simulator, enabling realistic replication of limit order book dynamics. Our model leverages recent advancements in large language models to produce long sequences of order messages in a steaming manner. Our results demonstrate that the model successfully reproduces key features of order flow data, even when the initial order flow prompt is no longer present within the model's context window. Moreover, evaluations reveal that the model captures several statistical properties, or 'stylized facts', characteristic of real financial markets and broader macro-scale data distributions. Collectively, this work marks a significant step toward creating high-fidelity, interactive market simulations.