MaGNet: A Mamba Dual-Hypergraph Network for Stock Prediction via Temporal-Causal and Global Relational Learning

TL;DR

MaGNet tackles stock movement forecasting under volatility by unifying advanced temporal modeling with dynamic higher-order relations. It introduces the MAGE block for bidirectional temporal context, enhanced by sparse Mixture-of-Experts and multi-head attention, plus feature-wise 2D spatiotemporal fusion. The dual-hypergraph learning framework (Temporal-Causal Hypergraph and Global Probabilistic Hypergraph) disentangles local temporal-causal signals from global market structures, enabling multi-scale relational learning. Across six major indices, MaGNet delivers superior predictive accuracy and profitability with robust risk management, validating its practical impact for trading and portfolio optimization.

Abstract

Stock trend prediction is crucial for profitable trading strategies and portfolio management yet remains challenging due to market volatility, complex temporal dynamics and multifaceted inter-stock relationships. Existing methods struggle to effectively capture temporal dependencies and dynamic inter-stock interactions, often neglecting cross-sectional market influences, relying on static correlations, employing uniform treatments of nodes and edges, and conflating diverse relationships. This work introduces MaGNet, a novel Mamba dual-hyperGraph Network for stock prediction, integrating three key innovations: (1) a MAGE block, which leverages bidirectional Mamba with adaptive gating mechanisms for contextual temporal modeling and integrates a sparse Mixture-of-Experts layer to enable dynamic adaptation to diverse market conditions, alongside multi-head attention for capturing global dependencies; (2) Feature-wise and Stock-wise 2D Spatiotemporal Attention modules enable precise fusion of multivariate features and cross-stock dependencies, effectively enhancing informativeness while preserving intrinsic data structures, bridging temporal modeling with relational reasoning; and (3) a dual hypergraph framework consisting of the Temporal-Causal Hypergraph (TCH) that captures fine-grained causal dependencies with temporal constraints, and Global Probabilistic Hypergraph (GPH) that models market-wide patterns through soft hyperedge assignments and Jensen-Shannon Divergence weighting mechanism, jointly disentangling localized temporal influences from instantaneous global structures for multi-scale relational learning. Extensive experiments on six major stock indices demonstrate MaGNet outperforms state-of-the-art methods in both superior predictive performance and exceptional investment returns with robust risk management capabilities. Codes available at: https://github.com/PeilinTime/MaGNet.

Figures (3)

• Figure 1: Overview of MaGNet architecture. It consists of: (1) MAGE Block, combining bidirectional Mamba, adaptive gating, MoE, and multi-head attention for temporal modeling; (2) Feature-wise 2D Spatiotemporal Attention to capture cross-feature dependencies while preserving spatiotemporal structure; (3) Dual Hypergraph Module, with TCH modeling fine-grained temporal-causal relations and GPH capturing global market patterns via soft assignments and JSD-based weighting.
• Figure 2: Backtesting performance of the MaGNet on DJIA, HSI and NASDAQ 100 indices
• Figure 3: Backtesting performance of the MaGNet on S&P 100, CSI300 and Nikkei225 indices