GenRiskNet: A GenAI-Driven Multi-Source Heterogeneous Data Fusion Framework for Financial Risk Prediction
DOI:
https://doi.org/10.71222/3bnztc94Keywords:
GenAI, Financial Risk Prediction, Multimodal Fusion, Heterogeneous Financial Data, Large Language Models, Time-Series Modeling, Credit Risk, Market RiskAbstract
Modern financial markets are increasingly shaped by fast-evolving information flows, ranging from market micro-structure signals to corporate disclosures, macroeconomic indicators, ESG assessments, and large volumes of high-frequency textual news. Traditional risk-prediction models struggle to jointly model these heterogeneous sources, limiting their ability to capture cross-modal causal drivers and abrupt risk dynamics. To address these challenges, this paper introduces GenRiskNet, a GenAI-driven heterogeneous data fusion framework that integrates large-language-model (LLM)-based event understanding with multi-branch time-series encoding and cross-modal multi-scale attention fusion. GenRiskNet jointly leverages (1) quantitative market features, (2) LLM-extracted financial textual events, (3) macroeconomic indicators, and (4) ESG corporate profiles to improve credit-risk and market-risk forecasting. Experiments conducted on the multi-source financial dataset show that GenRiskNet consistently outperforms LSTM, Temporal Fusion Transformer, and state-of-the-art multimodal fusion baselines across all tasks. The proposed framework achieves a 15.8% improvement in AUC for credit-risk prediction, reduces VaR forecasting error by 12.6%, and delivers a 19.7% gain in F1 score for default-event detection. These results closely align with the characteristics of the heterogeneous dataset and demonstrate the effectiveness of GenAI-driven cross-modal fusion in capturing complex financial risk patterns, confirming GenRiskNet as a robust and scalable framework for next-generation risk prediction.References
1. S. Xu, L. Jiang, and B. Gu, "Design and Validation of a Smart Neuromorphic System Architecture for Algorithmic Trading," In Proceedings of the 2nd International Symposium on Integrated Circuit Design and Integrated Systems, September, 2025, pp. 127-136. doi: 10.1145/3772326.3774721
2. X. Ding, Y. Zhang, T. Liu, and J. Duan, "Deep learning for event-driven stock prediction," In Ijcai, July, 2015, pp. 2327-2333.
3. P. Chen, Z. Boukouvalas, and R. Corizzo, "A deep fusion model for stock market prediction with news headlines and time series data," Neural Computing and Applications, vol. 36, no. 34, pp. 21229-21271, 2024. doi: 10.1007/s00521-024-10303-1
4. V. D'Amato, R. D'Ecclesia, and S. Levantesi, "Firms' profitability and ESG score: A machine learning approach," Applied Stochastic Models in Business and Industry, vol. 40, no. 2, pp. 243-261, 2024. doi: 10.1002/asmb.2758
5. P. M. S. Choi, S. H. Huang, and Q. Wang, "Large language models in finance: An overview," Finance and Large Language Models, pp. 1-26, 2025. doi: 10.1007/978-981-96-5833-6_1
6. D. Araci, "Finbert: Financial sentiment analysis with pre-trained language models," arXiv preprint arXiv:1908.10063, 2019.
7. E. Siphuma, and T. van Zyl, "Enhancing credit risk assessment through transformer-based machine learning models," In Southern African Conference for Artificial Intelligence Research, November, 2024, pp. 124-143. doi: 10.1007/978-3-031-78255-8_8
8. Y. Xiao, E. Sun, T. Chen, F. Wu, D. Luo, and W. Wang, "Trading-r1: Financial trading with llm reasoning via reinforcement learning," arXiv preprint arXiv:2509.11420, 2025.
9. R. Luo, N. Wang, and X. Zhu, "Fraud detection and risk assessment of online payment transactions on e-commerce platforms based on llm and gcn frameworks," arXiv preprint arXiv:2509.09928, 2025.
10. R. F. Engle, "Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation," Econometrica: Journal of the econometric society, pp. 987-1007, 1982.
11. S. Das, X. Huang, S. Adeshina, P. Yang, and L. Bachega, "Credit risk modeling with graph machine learning," INFORMS Journal on Data Science, vol. 2, no. 2, pp. 197-217, 2023. doi: 10.1287/ijds.2022.00018
12. Z. Yang, D. Yang, C. Dyer, X. He, A. Smola, and E. Hovy, "Hierarchical attention networks for document classification," In Proceedings of the 2016 conference of the North American chapter of the association for computational linguistics: human language technologies, June, 2016, pp. 1480-1489. doi: 10.18653/v1/n16-1174
13. G. Zhang, H. Zeng, and L. Jiang, "Uni-FinLLM: A Unified Multimodal Large Language Model with Modular Task Heads for Micro-Level Stock Prediction and Macro-Level Systemic Risk Assessment," arXiv preprint arXiv:2601.02677, 2026.
14. B. Su, G. Gui, S. Xu, and S. Shen, "Study on Real Estate Investment Risk Assessment and Decision Support System Driven by Fintech," In Proceedings of the 2nd International Symposium on Integrated Circuit Design and Integrated Systems, September, 2025, pp. 168-174. doi: 10.1145/3772326.3774727
