Graph Neural Networks for Cross-Market Contagion Forecasting with Residual-Stress Driven Risk Propagation
Keywords:
Graph neural networks, contagion forecasting, residual-stress risk, cross-market propagation, systemic risk, financial infrastructure, stress testing, machine learning governanceAbstract
The interconnected nature of modern financial markets requires forecasting frameworks that capture both structural dependencies and latent stress accumulation. This paper introduces a graph neural network architecture designed to model cross-market contagion by incorporating a residual-stress signal that quantifies unobserved drawdown risk beyond conventional volatility measures. The proposed system operates on a multilayer graph representation of financial instruments, where nodes correspond to asset classes or indices and edges encode both direct and indirect exposure channels. A residual-stress propagation mechanism is embedded within the message-passing layers to simulate how localized stress accumulates and cascades across markets under different liquidity regimes. The study emphasizes system-level considerations, including architectural trade-offs between model expressivity and computational scalability, governance implications for regulatory stress testing, and policy challenges in deploying such models within real-time surveillance infrastructures. We analyze the robustness of the framework against adversarial perturbations and data sparsity, and discuss fairness concerns arising from heterogeneous market participation. By integrating a leakage-safe residual-stress signal, the model provides a more stable and interpretable indicator of impending contagion, offering advantages over purely volatility-based forecasts. The paper concludes with a forward-looking assessment of deployment requirements, infrastructure sustainability, and the need for cross-jurisdictional coordination in systemic risk governance.
References
1. Forbes, K. J., & Rigobon, R. (2002). No contagion, only interdependence: Measuring stock market comovements. Journal of Finance, 57(5), 2223-2261.
2. Pesaran, M. H., & Pick, A. (2007). Econometric issues in the analysis of contagion. Journal of Economic Dynamics and Control, 31(4), 1245-1277.
3. Kipf, T. N., & Welling, M. (2017). Semi-supervised classification with graph convolutional networks. Proceedings of the International Conference on Learning Representations (ICLR).
4. Hamilton, W. L., Ying, Z., & Leskovec, J. (2017). Inductive representation learning on large graphs. Advances in Neural Information Processing Systems, 30, 1024-1034.
5. Cont, R. (2001). Empirical properties of asset returns: Stylized facts and statistical issues. Quantitative Finance, 1(2), 223-236.
6. Bouchaud, J.-P., & Potters, M. (2003). Theory of financial risk and derivative pricing (2nd ed.). Cambridge University Press.
7. Allen, F., & Gale, D. (2000). Financial contagion. Journal of Political Economy, 108(1), 1-33.
8. Longstaff, F. A. (2010). The subprime credit crisis and contagion in financial markets. Journal of Financial Economics, 97(3), 436-450.
9. Brunnermeier, M. K. (2009). Deciphering the liquidity and credit crunch 2007-2008. Journal of Economic Perspectives, 23(1), 77-100.
10. Acemoglu, D., Ozdaglar, A., & Tahbaz-Salehi, A. (2015). Systemic risk and stability in financial networks. American Economic Review, 105(2), 564-608.
11. Feng, G., He, J., & Polson, N. G. (2018). Deep learning for predicting asset returns. arXiv preprint arXiv:1804.03714.
12. Wang, J., Zheng, S., & Chen, J. (2019). Graph neural networks for stock market prediction: A survey. Proceedings of the AAAI Conference on Artificial Intelligence, 33, 9630-9637.
13. Xu, K., Hu, W., Leskovec, J., & Jegelka, S. (2019). How powerful are graph neural networks? Proceedings of the International Conference on Learning Representations (ICLR).
14. Withers, P. J., & Bhadeshia, H. K. D. H. (2001). Residual stress: Part 1 – Measurement techniques. Materials Science and Technology, 17(4), 355-365.
15. Sornette, D. (2003). Why stock markets crash: Critical events in complex financial systems. Princeton University Press.
16. Battaglia, P. W., Hamrick, J. B., Bapst, V., et al. (2018). Relational inductive biases, deep learning, and graph networks. arXiv preprint arXiv:1806.01261.
17. Liu, T. (2026). Beyond volatility: A leakage-safe residual-stress signal for drawdown risk monitoring. Available at SSRN 6503179.
18. Danielsson, J., & Zigrand, J.-P. (2008). Equilibrium asset pricing with systemic risk. Economic Theory, 35(2), 293-319.
19. Li, Q., Han, Z., & Wu, X. (2018). Deeper insights into graph convolutional networks for semi-supervised learning. Proceedings of the AAAI Conference on Artificial Intelligence, 32, 3538-3545.
20. Chen, M., Wei, Z., Huang, Z., Ding, B., & Li, Y. (2020). Simple and deep graph convolutional networks. Proceedings of the International Conference on Machine Learning (ICML), 119, 1725-1735.
21. Bastan, M., & O'Neil, C. (2020). The ethical algorithm: The science of socially aware algorithm design. Oxford University Press.
22. Kasy, M., & Abebe, R. (2021). Fairness, equality, and power in algorithmic decision-making. Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, 576-586.
23. Borio, C. (2003). Towards a macroprudential framework for financial supervision and regulation? BIS Working Papers, 128.
24. Goodfellow, I. J., Shlens, J., & Szegedy, C. (2015). Explaining and harnessing adversarial examples. Proceedings of the International Conference on Learning Representations (ICLR).
25. Zhang, J., & Wang, Y. (2020). Dynamic graph learning for time series forecasting. Advances in Neural Information Processing Systems, 33, 21462-21473.
26. Hinton, G., Vinyals, O., & Dean, J. (2015). Distilling the knowledge in a neural network. arXiv preprint arXiv:1503.02531.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 International Journal of Artificial Intelligence Research
This work is licensed under a Creative Commons Attribution 4.0 International License.
This article is published under the Creative Commons Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
