<p>Multi-modal financial distress prediction (FDP) has attracted widespread interest from both academic researchers and practitioners. However, existing FDP studies mostly focused on multi-modalities from the company itself, neglecting the predictive effect of external factors such as industry environment. Furthermore, the interaction between companies and industries plays a crucial role in predicting the future development trends of a company, yet it has rarely been explored. Under these circumstances, a Novel Industry-aware Hierarchical Graph Fusion model (Ind_HGF) is proposed for multi-modal FDP, which simultaneously considers the effect of industry environment, and the interaction between companies and industries. In particular, firstly, multi-modalities including financial and textual modality are extracted, with attention mechanisms employed to obtain features most relevant to the company’s financial condition. Secondly, a company-level graph is constructed to learn the complex relationships of the financially distressed companies, which is further extended to an industry-level graph. Finally, an industry-aware hierarchical graph fusion module is designed, incorporating an innovative hierarchical message passing (HMP) mechanism to handle the interaction between company-level and industry-level graphs. Extensive experiments on real-world datasets demonstrate that the proposed method achieves superior performance compared with benchmarked methods.</p>