Controllable Financial Market Generation with Diffusion Guided Meta Agent

TL;DR

This work tackles controllable financial market generation by introducing DigMA, a two-stage framework that unifies a diffusion-based meta-controller with a market-aware order generator. The meta-controller learns intraday market-state dynamics under scenario targets using a conditional diffusion model, while the order generator employs a meta agent with financial priors to produce realistic orders within a simulated double-auction exchange. Across two real tick datasets, DigMA demonstrates strong controllability and fidelity to stylized facts, and it provides an effective, efficient generative environment for downstream high-frequency trading tasks. The approach enables scenario-based testing and counterfactual analyses with real-time generation speed, offering practical value for research and development in finance.

Abstract

Generative modeling has transformed many fields, such as language and visual modeling, while its application in financial markets remains under-explored. As the minimal unit within a financial market is an order, order-flow modeling represents a fundamental generative financial task. However, current approaches often yield unsatisfactory fidelity in generating order flow, and their generation lacks controllability, thereby limiting their practical applications. In this paper, we formulate the challenge of controllable financial market generation, and propose a Diffusion Guided Meta Agent (DigMA) model to address it. Specifically, we employ a conditional diffusion model to capture the dynamics of the market state represented by time-evolving distribution parameters of the mid-price return rate and the order arrival rate, and we define a meta agent with financial economic priors to generate orders from the corresponding distributions. Extensive experimental results show that DigMA achieves superior controllability and generation fidelity. Moreover, we validate its effectiveness as a generative environment for downstream high-frequency trading tasks and its computational efficiency.

Figures (4)

• Figure 1: Limit order book and order flow
• Figure 2: Overview of DigMA model. Raw order flow is proccessed into market states for the meta controller to learn. The meta agent is guided by the meta controller, and generates simulated order flow.
• Figure 3: Aggregated price curves (first row) and distributions of targeted indicators (second row). In the first row, each curve represents the price trajectory derived from one day of generated order flow. In the second row, each colored density corresponds to the distribution of a targeted indicator computed from the generation results.
• Figure 4: Comparison of stylized facts distribution across baselines. The x-axis is the stylized facts and the y-axis is the density.