TradeFM: A Generative Foundation Model for Trade-flow and Market Microstructure

TL;DR

TradeFM is a large generative model trained on extensive trade data that captures market microstructure features and generalizes across different markets, enabling applications like synthetic data generation and trading strategies.

Contribution

We introduce TradeFM, a novel scale-invariant, multi-modal generative Transformer for market microstructure modeling, capable of zero-shot cross-market generalization.

Findings

TradeFM achieves 2-3x lower distributional error than baselines.

It reproduces key stylized facts of financial returns.

It generalizes well to out-of-distribution markets.

Abstract

Foundation models have transformed domains from language to genomics by learning general-purpose representations from large-scale, heterogeneous data. We introduce TradeFM, a 524M-parameter generative Transformer that brings this paradigm to market microstructure, learning directly from billions of trade events across >9K equities. To enable cross-asset generalization, we develop scale-invariant features and a universal tokenization scheme that map the heterogeneous, multi-modal event stream of order flow into a unified discrete sequence -- eliminating asset-specific calibration. Integrated with a deterministic market simulator, TradeFM-generated rollouts reproduce key stylized facts of financial returns, including heavy tails, volatility clustering, and absence of return autocorrelation. Quantitatively, TradeFM achieves 2-3x lower distributional error than Compound Hawkes baselines and generalizes zero-shot to geographically out-of-distribution APAC markets with moderate perplexity degradation. Together, these results suggest that scale-invariant trade representations capture transferable structure in market microstructure, opening a path toward synthetic data generation, stress testing, and learning-based trading agents.