| 467 | 0 | 518 |
| 下载次数 | 被引频次 | 阅读次数 |
标本兼治、远近结合是防范化解重大金融风险的核心手段。考虑到中美两国股市存在复杂的内外部风险关联，且受到宏观经济环境变化的影响，本文提出超高维混频DCC模型，从大维复杂经济系统视角考察中美股市的风险关联。进一步延拓传统研究范式，基于短期和长期相关系数构建复杂网络，从静态到动态、全局到局部的视角分析中美股市间的风险交互效应，进而从一级行业和个股两个层面度量风险关联强度，并捕捉重要的风险关联节点。研究结果表明，宏观经济代理变量对个股波动的平均解释程度占比为31%，是个股波动和风险关联的重要驱动因素。相较于美国股市，我国股市波动受宏观经济环境变动的影响较小。短期来看，我国铜、铝等有色金属细分行业与美国股市关联紧密，一级行业层面的风险关联渠道则位于金融、房地产和交通运输行业。长期来看，中美股市在有色金属、金融、煤炭方面存在紧密风险关联。本研究对于厘清跨国股市风险关联机制、防范金融系统性风险具有积极意义。
Abstract：Combining symptomatic and fundamental treatment and integrating short-term and long-term perspectives are essential for mitigating major financial risks. Given the complex internal and external risk linkage between the Chinese and U.S. stock markets under changing macroeconomic conditions, this paper extends an ultra-high-dimensional DCC MIDAS model to examine risk linkage from both global and individual perspectives, across short-term and long-term horizons. Using the stock market big data, we incorporate U.S.-China economic policy uncertainty and global geopolitical risk as proxies for macroeconomic fluctuations to quantify bilateral risk linkages. Going beyond conventional approaches, we construct complex networks based on short-term and long-term correlations and apply the TMFG filtering algorithm to map them onto a low-dimensional manifold. This enables a dynamic, multi-level analysis of risk interactions, identifying key nodes and measuring risk linkage at both the sectoral and stock levels. Results show that macroeconomic proxies explain on average 31% of individual stock volatility,highlighting their importance in driving volatility and risk linkage. Compared to the U.S. market, Chinese stock volatility is less sensitive to macroeconomic changes. In the short term, China's non-ferrous metal sectors(e.g. copper, aluminum) show strong connections with U.S. stocks, while key sectoral channels include finance, real estate, and transportation. In the long term, significant risk linkages exist in non-ferrous metals, finance, and coal. These findings shed light on cross-border risk transmission mechanisms and provide valuable insights for preventing systemic financial risks.
[1]李政,李丽雯,刘淇.中国行业间尾部风险溢出的测度及时空驱动因素研究[J].统计研究, 2024, 41(2):64–76.
[2]杨科,郭亚飞,田凤平.经济政策不确定性冲击下全球系统性金融风险的跨市场传染——基于TVP-FAVAR和TVP-VAR模型的研究[J].统计研究, 2023, 40(7):70–84.
[3]郑挺国,巩璐,叶仕奇.时变视阈下在险通货膨胀的跨国溢出研究[J].世界经济, 2024a, 47(12):35–70.
[4]郑挺国,张宏音,叶仕奇.低碳转型风险的“涟漪效应”[J].中国工业经济, 2024b(4):37–56.
[5]周开国,邢子煜,杨海生.宏观经济信息与金融市场关联性——来自混频动态条件相关系数模型的证据[J].金融研究, 2021, 497(11):41–59.
[6]Allard A F, Iania L, Smedts K. Stock-Bond Return Correlations:Moving away from“One-Frequency-Fits-All” by Extending the DCC-MIDAS Approach[J]. International Review of Financial Analysis, 2020, 71:101557.
[7]Asgharian H, Christiansen C, Hou A J. Macro-Finance Determinants of the Long-Run Stock-Bond Correlation:The DCC-MIDAS Specification[J]. Journal of Financial Econometrics, 2016, 14(3):617–642.
[8]Bardoscia M, Barucca P, Battiston S. The Physics of Financial Networks[J]. Nature Reviews Physics, 2021:1–18.
[9]Barigozzi M, Hallin M. A Network Analysis of the Volatility of High Dimensional Financial Series[J]. Journal of the Royal Statistical Society Series C:Applied Statistics, 2017, 66(3):581–605.
[10]Brownlees C, Guemundsson G S, Lugosi G. Community Detection in Partial Correlation Network Models[J]. Journal of Business&Economic Statistics, 2022, 40(1):216–226.
[11]Colacito R, Engle R F, Ghysels E. A Component Model for Dynamic Correlations[J]. Journal of Econometrics, 2011, 164(1):45–59.
[12]Conrad C, Loch K. Anticipating Long-Term Stock Market Volatility[J]. Journal of Applied Econometrics, 2015, 30(7):1090–1114.
[13]Diebold F X, Y?lmaz K. On the Network Topology of Variance Decompositions:Measuring the Connectedness of Financial Firms[J]. Journal of Econometrics, 2014, 182(1):119–134.
[14]Dietrich A M, Kuester K, Müller G J. News and Uncertainty about COVID-19:Survey Evidence and Short-Run Economic Impact[J]. Journal of Monetary Economics, 2022, 129:S35–S51.
[15]Engle R. Dynamic Conditional Correlation:A Simple Class of Multivariate Generalized Autoregressive Conditional Heteroskedasticity Models[J]. Journal of Business&Economic Statistics, 2002, 20(3):339–350.
[16]Engle R F, Ghysels E, Sohn B. Stock Market Volatility and Macroeconomic Fundamentals[J]. Review of Economics and Statistics, 2013, 95(3):776–797.
[17]Engle R F, Ledoit O, Wolf M. Large Dynamic Covariance Matrices[J]. Journal of Business&Economic Statistics, 2019, 37(2):363–375.
[18]Fang L, Yu H, Li L. The Effect of Economic Policy Uncertainty on the Long-Term Correlation Between US Stock and Bond Markets[J].Economic Modelling, 2017, 66:139–145.
[19]Gabaix X, Gopikrishnan P, Plerou V, et al. A Theory of Power-Law Distributions in Financial Market Fluctuations[J]. Nature, 2003, 423(6937):267–270.
[20]Hu Z, Kutan A M, Sun P W. Is US Economic Policy Uncertainty Priced in China’s A-Shares Market? Evidence from Market, Industry, and Individual Stocks[J]. International Review of Financial Analysis, 2018, 57:207–220.
[21]Ledoit O, Wolf M. Nonlinear Shrinkage Estimation of Large-Dimensional Covariance Matrices[J]. Annals of Statistics, 2012, 40(2):1024–1060.
[22]Ledoit O, Wolf M. Analytical Nonlinear Shrinkage of Large-Dimensional Covariance Matrices[J]. Annals of Statistics, 2020, 48(5):3043–3065.
[23]Li X, Ye C, Bhuiyan M A, et al. Volatility Forecasting with an Extended GARCH-MIDAS Approach[J]. Journal of Forecasting, 2024, 43(1):24–39.
[24]Massara G P, Di Matteo T, Aste T. Network Filtering for Big Data:Triangulated Maximally Filtered Graph[J]. Journal of Complex Networks,2017, 5(2):161–178.
[25]Pakel C, Shephard N, Sheppard K. Fitting Vast Dimensional Time-Varying Covariance Models[J]. Journal of Business&Economic Statistics,2021, 39(3):652–668.
[26]Perego E R, Vermeulen W N. Macro-Economic Determinants of European Stock and Government Bond Correlations:A Tale of Two Regions[J].Journal of Empirical Finance, 2016, 37:214–232.
[27]Schwert G W. Why Does Stock Market Volatility Change over Time?[J]. Journal of Finance, 1989, 44(5):1115–1153.
[28]Song W M, Di Matteo T, Aste T. Hierarchical Information Clustering by Means of Topologically Embedded Graphs[J]. PLoS ONE, 2012, 7(3):e31929.
[29]Yu X, Huang Y. The Impact of Economic Policy Uncertainty on Stock Volatility:Evidence from GARCH-MIDAS Approach[J]. Physica A:Statistical Mechanics and Its Applications, 2021, 570:125794.
[30]Zaremba A, Cakici N, Demir E, et al. When Bad News Is Good News:Geopolitical Risk and the Cross-Section of Emerging Market Stock Returns[J]. Journal of Financial Stability, 2022, 58:100964.
(1)相关网址为https://www.gov.cn/yaowen/liebiao/202512/content_7050963.htm。
(2)港股通是沪港股票市场交易互联互通机制的组成部分，全称为港股直通车。
(3)相关网址为https://www.gov.cn/xinwen/2023-03/06/content_5745092.htm。
(1)本文亦考虑了滚动窗口的长期成分估计法，结果稳健。
(2)转化函数将月度的低频长期成分通过填补法填补为日度的高频时间序列，得到的高频时间序列每个月的值在当月内保持不变，与低频长期成分在该月的值相对应。
(3)去均值化算子对于矩阵中的每一个元素，减去其在时间维度上的均值。去均值后的相关系数矩阵对角元素为0。
(1)因篇幅所限，收缩估计量以附录A（一）展示，见《统计研究》网站所列附件。下同。
(1)因篇幅所限，似然函数具体表示以附录A（二）展示。
(1)因篇幅所限，复杂网络统计特征介绍以附录A（三）展示。
(2)因篇幅所限，数据选取与处理方法以附录B（一）展示。
(3)英为财经相关网址为https://cn.investing.com/。
(4)因篇幅所限，数据描述性统计与波动性分析以附录B（二）展示。低频变量的时间窗口设定较高频变量提前三年，用于在MIDAS框架下提取长期波动成分。
(1)因篇幅所限，部分模型参数估计结果和相关分析以附录C（一）展示。
(2)随着滞后阶数K的增加，长期成分所包含的信息也将随之增加，参数估计结果趋于稳定。因篇幅所限，滞后阶数K的选择及相关稳健性分析以附录A（四）展示。
(1)因篇幅所限，相关讨论以附录A（五）展示。
(1)本文采用中美股市一级行业分类标准为申万一级行业2021版（SWI2021）和全球行业分类标准（Global Industry Classification Standard,GICS）。因篇幅所限，中美股市一级行业分类详细信息以附表B.4展示，高维网络聚合方法以附录A（六）展示。
(2)因篇幅所限，复杂网络分析与可视化结果以附录C（二）展示。
(1)我国缺乏铁矿产资源，铜精矿的外贸依存度高达80%，有色金属是国民经济发展的重要基础材料。2020年9月我国提出“碳达峰碳中和”目标后，新能源汽车行业的蓬勃发展扩大了我国对于铜矿资源的需求。
基本信息:
DOI：10.19343/j.cnki.11-1302/c.2026.01.003
中图分类号:F832.51;F837.12;F125;F171.2
引用信息:
[1]郑挺国,叶仕奇,范小龙,等.大维视角下的中美股市动态关联与宏观经济环境变动[J].统计研究,2026,43(01):41-55.DOI:10.19343/j.cnki.11-1302/c.2026.01.003.
基金信息:
国家社会科学基金重大项目“大数据方法在宏观经济预测中的应用研究”（23&ZD074）
2026-01-25
2026-01-25