跳转至

价格与交易动态的可解释深度学习：从黑箱预测到高效参数化模型

文章背景与核心概要

本文探讨了市场微观结构中价格与交易的联合动态，采用深度神经网络进行结构性发现，而不仅仅局限于黑箱预测。作者分析了大标价（large-tick）和小标价（small-tick）股票的高频收益率与符号交易量，并应用基于Shapley值的可解释性方法，揭示了其中的非线性依赖关系和状态依赖机制。

研究发现，滞后的符号交易量会产生符号保持且趋于饱和的效应（这与非线性价格冲击一致），而滞后的收益率则充当决定持续还是反转的状态变量。基于这些深刻见解，该研究引入了一种精简的、受SHAP启发的非线性参数化模型，该模型在保持经济可解释性的同时，匹配了深度学习的预测性能。这项工作为从黑箱预测走向具有经济意义的价格与交易动态参数化模型提供了一条有效途径。

Metadata

  • Authors: Manuel Naviglio, Fabrizio Lillo
  • Subjects: Trading and Market Microstructure (q-fin.TR); Artificial Intelligence (cs.AI); Data Analysis, Statistics and Probability (physics.data-an)
  • Submitted on: September 5, 2026
  • arXiv Identifier: arXiv:2609.06085
  • DOI: 10.48550/arXiv.2609.06085

Abstract

理解价格与交易的联合动态是市场微观结构的核心，其中收益率和订单流通过非线性及状态依赖的机制相互作用。线性模型具有可解释性，但可能会漏掉这些效应，而深度神经网络在提高预测能力的同时却牺牲了透明度。

Understanding the joint dynamics of prices and trades is central to market microstructure, where returns and order flow interact through nonlinear and state-dependent mechanisms. Linear models are interpretable but may miss these effects, while deep neural networks improve forecasting at the cost of transparency.

我们使用神经网络作为结构发现的工具，而不仅仅用于预测。我们在针对大标价和小标价股票的高频收益率与符号交易量上训练了一个深度前馈网络，并将其与线性VAR基准模型进行了比较。神经网络提高了预测性能（尤其是对收益率的预测），揭示了超出线性规范的非线性依赖关系。

We use neural networks as tools for structural discovery rather than only for prediction. A deep feed-forward network is trained on high-frequency returns and signed volumes for large- and small-tick stocks and compared with a linear VAR benchmark. The neural network improves predictive performance, especially for returns, revealing nonlinear dependencies beyond the linear specification.

通过基于Shapley值的可解释性方法，我们发现主要贡献集中在最近的滞后期上。模型隐含的响应与从数据中重建的条件平均值相一致。然而，与经验平均值不同的是，神经网络分解孤立出了各个回归变量对总体依赖关系的具体贡献。

Using Shapley-based explainability, we show that the dominant contributions are concentrated at the most recent lags. Model-implied responses are consistent with conditional averages reconstructed from the data. Unlike empirical averages, however, the neural-network decomposition isolates individual regressor contributions to the aggregate dependence.

  • 滞后符号交易量（Lagged signed volume）会产生符号保持且饱和的效应，这与非线性价格冲击和订单流持续性相一致。
  • 滞后收益率（Lagged returns）充当状态变量：当先前的交易没有引发价格变动时，模型预测会沿着过去订单流的方向持续；而当收益率非零时，则会产生衰减或反转。
  • Lagged signed volume generates sign-preserving and saturating effects, consistent with nonlinear price impact and order-flow persistence.
  • Lagged returns act as state variables: when the previous trade does not move the price, the model predicts continuation in the direction of past order flow, whereas non-zero returns generate attenuation or reversal.

基于这些发现，我们引入了一种精简的、受SHAP启发的非线性参数化模型。它重现了主要的收益率-交易量依赖关系，优于线性VAR基准模型，并达到了与深度神经网络相当的性能。多滞后扩展版本在保持可解释性的同时，捕获了剩余的长期记忆效应。总体而言，可解释性为从黑箱预测转向具有经济意义的价格与交易动态参数化模型提供了一条途径。

Building on these findings, we introduce a parsimonious SHAP-inspired nonlinear parametric model. It reproduces the main return-volume dependencies, outperforms the linear VAR benchmark, and achieves performance comparable to the neural network. A multi-lag extension captures residual longer-memory effects while preserving interpretability. Overall, explainability offers a route from black-box prediction to economically meaningful parametric models of price and trade dynamics.

Full-Text & References