Academic Journal of Business & Management, 2023, 5(16); doi: 10.25236/AJBM.2023.051613.
Erya Zheng1, Yan Tu2, Zhiyue Huang
1School of Information, Central University of Finance and Economics, Beijing, China
2China Academy of Internet Economics, Central University of Finance and Economics, Beijing, China
In recent years, the continuous outbreak of public health events has intensified the risk linkage effect between China's real economy and financial markets. In order to describe the risk correlation mechanism among various entities, this paper uses the time-varying Copula - △ CoVaR model to quantify the risk spillover intensity between the real economy and the financial market in the context of the COVID-19, identifies the main risk contagion sources and risk contagion paths, and constructs a comprehensive early warning indicator system and a graph attention model to optimize the early warning work. The empirical research results indicate that: firstly, the risk spillover level generated by the impact of the epidemic has characteristics of time delay, repetition, and wave like, and the initial intensity is the highest; Secondly, the impact of the epidemic will exacerbate the risk of cross market transmission and upstream and downstream transmission along the industrial chain; Thirdly, compared to traditional machine learning models, the accuracy of graph attention early warning models has been improved by 10-20%, and their predictive ability for risk events is stronger. The research conclusion of this article will provide reference for regulatory authorities to identify major risk industries and markets, monitor multi-stage contagion of risks in sudden public health events, and establish a sound risk warning mechanism. It has important practical significance for promoting coordinated development between China's real economy and financial markets.
real economy; financial system; graph attention network; emergency events
Erya Zheng, Yan Tu, Zhiyue Huang. Research on Risk Spillover Effects and Early Warning Optimization of China's Real Economy and Financial System. Academic Journal of Business & Management (2023) Vol. 5, Issue 16: 90-99. https://doi.org/10.25236/AJBM.2023.051613.
[1] Adrian T, Brunnermeier M K. Conditional Value at Risk[R]. Federal Reserve Bank of New York, 2008.
[2] Adrian T, Brunnermeier M K. Conditional Value at Risk[R]. National Bureau of Economic Research, 2011.
[3] Acharya V, Engle R, Richardson M. Capital Shortfall. A New Approach to Ranking and Regulating Systemic Risks [J]. American Economic Review,2012, 3: 59-64.
[4] Acharya V, L H Pedersen, T Philippon, and M Richardson. “Measuring Systemic Risk” [J]. The Review of Financial Studies, 2017, 30(1):2-47.
[5] Brownless C, Engle R. A Conditional Capital Shortfall Measure of Systemic Risk [J]. The Review of Financial Studies, 2017, 1:48-79.
[6] Fang Yi, Shao Zhiquan. Research on the Measurement of Input Financial Risks in China under Major Shocks [J]. Economic Science, 2022(02):13-30.
[7] Xie Chi, Mo Tingcheng, Li Kelong. Research on Extreme risk Dependence and Risk Spillover among Financial Industries in the Context of Major Emergencies [J]. The Theory and Practice of Finance and Economics, 2021, 42(03):2-10.
[8] Ren Yinghua, Jiang Jingfeng, Ni Qingshan. Research on Financial Crisis Warning Based on Graph Neural Network—From the perspective of information spillover across industries [J]. Annals of Statistics Research, 2022, 39(08):141-160.
[9] Li Xiaohan, Wang Jun, Jia Huading, Xiao Liu. A Graph Neural Network Method for Predicting Stock Market Volatility Based on Multiple Attention Mechanisms [J]. Journal of Computer Applications, 2022, 42(07):2265-2273.
[10] Fang Yi, He Wenjia, Jing Zhongbo. Research on Risk Spillovers in China's Real Economy and Financial Market [J]. The Journal of World Economy, 2021, 44(08):3-27.