Systemic Risk Modeling in Listed Chinese Financial Institutions: A Two-Layer Network Analysis with a Graph Attention Prototype

Abstract

System-wide financial risk may be obscured when return comovement and directional spillovers are examined separately. We analyzed 3,167 daily observations for 24 listed Chinese banks, securities firms, and insurers from 19 December 2011 to 31 December 2024. A two-layer multiplex network combined a thresholded Pearson returncomovement layer with a LASSO–VAR generalized forecast error variance decomposition layer, and maximum spanning arborescences represented dominant transmission backbones. Across four non-overlapping windows, verified total connectedness was 88.50%, 89.01%, 89.37%, and 86.34%, whereas return-comovement density was 0.736, 0.739, 0.833, and 0.518. Full-sample estimates identified GDB, JTB, HtS, ZXS, and ZSB as leading net transmitters and GJS and GFS as leading receivers. Correcting the archived edge direction and arborescence implementation produced period-specific roots rather than a stable single source. The accompanying GATv2 prototype merged the two relations as edge features, but only four network snapshots, circular structural targets, absent period-5 inputs, and no reproducible temporal split prevented verification of the reported clustering and classification scores. The graph-attention evidence is therefore treated as a proof-of-concept, not out-of-sample early warning tool. This audited framework separates synchronous returns from directional connectedness and may support transparent monitoring after prospective validation.

Keywords

• systemic financial risk
• return spillovers
• multiplex financial network
• LASSO–VAR
• graph attention network
• connectedness

References

• [1] Haldane AG, May RM. Systemic risk in banking ecosystems. Nature. 2011;469:351–355.
• [2] Billio M, Getmansky M, Lo AW, Pelizzon L. Econometric measures of connectedness and systemic risk in the finance and insurance sectors. Journal of Financial Economics. 2012;104:535–559.
• [3] Diebold FX, Yilmaz K. Measuring financial asset return and volatility spillovers, with application to global equity markets. Economic Journal. 2009;119:158–171.
• [4] Diebold FX, Yilmaz K. Better to give than to receive: predictive directional measurement of volatility spillovers. International Journal of Forecasting. 2012;28:57–66.
• [5] Diebold FX, Yilmaz K. On the network topology of variance decompositions: measuring the connectedness of financial firms. Journal of Econometrics. 2014;182:119–134.
• [6] Demirer M, Diebold FX, Liu L, Yilmaz K. Estimating global bank network connectedness. Journal of Applied Econometrics. 2018;33:1–15.
• [7] Koop G, Pesaran MH, Potter SM. Impulse response analysis in nonlinear multivariate models. Journal of Econometrics. 1996;74:119–147.
• [8] Pesaran HH, Shin Y. Generalized impulse response analysis in linear multivariate models. Economics Letters. 1998;58:17–29.
• [9] Tibshirani R. Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society: Series B. 1996;58:267–288.
• [10] Caccioli F, Barucca P, Kobayashi T. Network models of financial systemic risk: a review. Journal of Computational Social Science. 2018;1:81–114.
• [11] Battiston S, Puliga M, Kaushik R, Tasca P, Caldarelli G. DebtRank: too central to fail? Financial networks, the FED and systemic risk. Scientific Reports. 2012;2:541.
• [12] Poledna S, Molina-Borboa JL, Martínez-Jaramillo S, van der Leij M, Thurner S. The multi-layer network nature of systemic risk and its implications for the costs of financial crises. Journal of Financial Stability. 2015;20:70–81.
• [13] Kivelä M, Arenas A, Barthelemy M, Gleeson JP, Moreno Y, Porter MA. Multilayer networks. Journal of Complex Networks. 2014;2:203–271.
• [14] Boccaletti S, Bianconi G, Criado R, et al. The structure and dynamics of multilayer networks. Physics Reports. 2014;544:1–122.
• [15] Blondel VD, Guillaume JL, Lambiotte R, Lefebvre E. Fast unfolding of communities in large networks. Journal of Statistical Mechanics: Theory and Experiment. 2008;2008:P10008.
• [16] Edmonds J. Optimum branchings. Journal of Research of the National Bureau of Standards B. 1967;71B:233–240.
• [17] Veličković P, Cucurull G, Casanova A, Romero A, Liò P, Bengio Y. Graph attention networks. International Conference on Learning Representations. 2018.
• [18] Brody S, Alon U, Yahav E. How attentive are graph attention networks? International Conference on Learning Representations. 2022.