🤖 AI Summary
Existing stock market simulation models suffer from overfitting to historical data, poor reproduction of stylized facts, and limited capacity for counterfactual analysis. Method: This paper proposes a behavioral-economics–inspired agent-based modeling (ABM) framework. We introduce the novel “search-free calibration” paradigm, addressing ABM’s non-differentiability via a differentiable agent loss function, and design a conditional-aware variable estimator that dynamically adapts to five market state indices—PPI, PMI, CPI, trend, and noise. Leveraging deep neural networks and a surrogate-trading loss, the method enables order-flow–level fidelity. Results: Evaluated on full-year real limit-order-book data, the model achieves a Kolmogorov–Smirnov statistic < 0.36, accelerates calibration by several orders of magnitude, and robustly uncovers intrinsic links between agent behavior and macro-market conditions.
📝 Abstract
Multi-agent market model is a stock trading simulation system, which generates order flow given the agent variable of the model. We study calibrating the agent variable to simulate the order flow of any given historical trading day. In contrast to the traditional calibration that relies on the inefficient iterative search, we propose DeepCal, the first search-free approach that uses deep learning to calibrate multi-agent market model. DeepCal learns from a novel surrogate-trading loss function to address the non-differentiable issue induced by the multi-agent model and introduces a condition-aware variable estimator, adapting the trading simulation to different market conditions to enhance explainability. Through extensive experiments on real order-book data over a whole year, DeepCal has demonstrated comparable simulation accuracy (<0.36 in Kolmogorov-Smirnov statistic) to traditional search-based approaches without the need for variable search, and can effectively capture the correlation between agent variable and multiple market-condition indexes~(PPI, PMI, CPI, market trend and market noise).
