发表机构
Simudyne; King’s College London（思慕迪恩; 伦敦国王学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文提出FlowLOB，一种基于流匹配的LOB轨迹生成模型，经HKEX数据训练可泛化至未见过的交易品种，采样效率与可控性优于基准模型，且能零样本泛化。
AI 中文摘要
限价订单簿（LOB）模拟器对从业者的实用价值，在于其能兼具真实市场动态、计算高效采样、可控场景生成，以及泛化至训练时未见过的金融工具的能力——这些特性是现有基于智能体的模拟器和深度生成式模拟器仅部分具备的。本文提出FlowLOB，这是一种条件流匹配LOB轨迹生成模型，在香港交易所（HKEX）多个交易品种的三种采样频率（0.1秒、1秒、10秒）下，以tick相对表示形式训练，可泛化至未见过的交易品种。由于流模型与扩散模型具有共同的公式，我们使用相同的数据、架构和计算预算训练两类模型，并通过相同的固定步长常微分方程（ODE）求解器采样，从而对采样效率和保真度进行可控比较。流匹配仅用10个ODE求解器步长即可达到最佳质量，而扩散模型需要更多函数评估才能接近相同保真度。在这一高效运行点，FlowLOB在两种更高采样频率下的多数分布指标上，相较于两个学习模型和两个基于智能体的基准模型，提升了真实性。我们通过一项分布测试评估反事实可控性，该测试要求改变场景条件时，生成的统计量是否向对应真实尾部制度移动；FlowLOB在多数测试设置中满足该标准。真实性和可控性效果均能零样本泛化至保留的交易品种。我们还对网络架构和学习率进行了消融研究。
英文摘要
Limit order book (LOB) simulators are most useful to practitioners when they combine realistic market dynamics, computationally efficient sampling, controllable scenario generation, and the ability to generalize beyond the instruments seen during training---properties that existing agent-based and deep generative simulators provide only partially. We present \textbf{FlowLOB}, a conditional \textbf{flow}-matching generator of \textbf{LOB} trajectories, trained on multiple Hong Kong Exchange (HKEX) symbols at three sampling frequencies ($0.1$s, $1$s, $10$s) in tick-relative representation that transfers to unseen instruments. Because flow and diffusion models admit a common formulation, we train both with identical data, architecture, and budget, and sample both through the same fixed-step ODE solvers, yielding a controlled comparison of sampling efficiency and fidelity. Flow matching attains its best quality with only $10$ ODE-solver steps, whereas diffusion needs many more function evaluations to approach the same fidelity. At this efficient operating point, FlowLOB improves realism over baselines, two learned and two agent-based models, in most distributional metrics at the two finer sampling frequencies. We evaluate counterfactual controllability with a distributional test that asks whether changing a scenario condition moves the generated statistic toward the corresponding real tail regime; FlowLOB satisfies this criterion in most tested settings. Both realism and control effects transfer zero-shot on a held-out symbol. We additionally conduct ablation studies on the network architecture and the learning rate.
Comments8 pages, 3 figures, 2 tables
