Stock Prediction via a Dual Relation Fusion Network incorporating Static and Dynamic Relations
Long Chen, Huixin Bai, Mingxin Wang, Xiaohua Huang, Ying Liu, Jie Zhao, Ziyu Guan
TL;DR
The paper tackles stock price forecasting by addressing the evolving inter-stock relationships that drive price movements. It introduces the Dual Relation Fusion Network (DRFN), which jointly models long-term relatively static relations and short-term dynamic relations, aided by News-Indicator contrastive alignment and a three-stage relation fusion mechanism (static, dynamic, relative-static) followed by adaptive residual output for next-day prediction. The approach delivers three core contributions: a distance-modulated dynamic relation learning framework, an adaptive relative-static relation construction to capture overnight information, and a bidirectional fusion strategy that leverages static and dynamic complementarities. Across US and China markets, DRFN achieves superior predictive accuracy and demonstrates strong sensitivity to relational changes co-moving with price, indicating robust cross-market applicability and practical potential in multimodal, relational stock forecasting.
Abstract
Accurate modeling of inter-stock relationships is critical for stock price forecasting. However, existing methods predominantly focus on single-state relationships, neglecting the essential complementarity between dynamic and static inter-stock relations. To solve this problem, we propose a Dual Relation Fusion Network (DRFN) to capture the long-term relative stability of stock relation structures while retaining the flexibility to respond to sudden market shifts. Our approach features a novel relative static relation component that models time-varying long-term patterns and incorporates overnight informational influences. We capture dynamic inter-stock relationships through distance-aware mechanisms, while evolving long-term structures via recurrent fusion of dynamic relations from the prior day with the pre-defined static relations. Experiments demonstrate that our method significantly outperforms the baselines across different markets, with high sensitivity to the co-movement of relational strength and stock price.