ACT: Anti-Crosstalk Learning for Cross-Sectional Stock Ranking via Temporal Disentanglement and Structural Purification

2026-04-22

AI Summary

This study addresses factor crosstalk in cross-sectional stock ranking caused by entangled temporal representations and improper integration of graph structures. To mitigate this, the authors propose the Anti-Crosstalk (ACT) framework, which first decomposes individual stock time series into trend, volatility, and shock components and models them separately. The trend component then undergoes progressive structural purification to eliminate interference from heterogeneous relationships. Finally, multi-branch representations are adaptively fused for ranking. ACT is the first to systematically distinguish crosstalk arising from temporal scales versus structural layers, enabling transferable trend modeling and relation-specific signal extraction. Experiments on CSI300 and CSI500 demonstrate significant improvements in ranking accuracy and portfolio performance, with annualized returns on CSI300 increasing by up to 74.25%.

Technology Category

• Machine Learning: Learning Preferences or Rankings
• Cognitive Modeling & Cognitive Systems: Symbolic Representations
• Reasoning under Uncertainty: Stochastic Optimization

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: Personalized content ranking and presentation

Abstract

Cross-sectional stock ranking is a fundamental task in quantitative investment, relying on both temporal modeling of individual stocks and the capture of inter-stock dependencies. While existing deep learning models leverage graph-based approaches to enhance ranking accuracy by propagating information over relational graphs, they suffer from a key challenge: crosstalk, namely unintended information interference across predictive factors. We identify two forms of crosstalk: temporal-scale crosstalk, where trends, fluctuations, and shocks are entangled in a shared representation and non-transferable local patterns contaminate cross-stock learning; and structural crosstalk, where heterogeneous relations are indiscriminately fused and relation-specific predictive signals are obscured. To address both issues, we propose the Anti-CrossTalk (ACT) framework for cross-sectional stock ranking via temporal disentanglement and structural purification. Specifically, ACT first decomposes each stock sequence into trend, fluctuation, and shock components, then extracts component-specific information through dedicated branches, which effectively decouples non-transferable local patterns. ACT further introduces a Progressive Structural Purification Encoder to sequentially purify structural crosstalk on the trend component after mitigating temporal-scale crosstalk. An adaptive fusion module finally integrates all branch representations for ranking. Experiments on CSI300 and CSI500 demonstrate that ACT achieves state-of-the-art ranking accuracy and superior portfolio performance, with improvements of up to 74.25% on the CSI300 dataset.

Problem

• crosstalk
• cross-sectional stock ranking
• temporal disentanglement
• structural purification
• quantitative investment

Innovation

• temporal disentanglement
• structural purification
• crosstalk mitigation
• cross-sectional stock ranking
• graph-based learning