Extreme Risk Connectedness in China's Stock Market: Fresh Insights from Time-Varying General Dynamic Factor Models

Xiaoye Jin
DOI: https://doi.org/10.1007/s10614-024-10779-y
IF: 1.741
2024-11-23
Computational Economics
Abstract:This study develops a two-step analytical framework consisting of the one-factor GAS model and the high-dimensional connectedness method to analyze the magnitude and persistence of extreme risk connectedness in China's stock market. In summary, some interesting findings emerge from this investigation. First, we provide strong evidence for the time-varying property of extreme risk connectedness in China's stock market. Second, we find consistent evidence of asymmetric downside and upside extreme risk connectedness in China's stock market. Third, the spectral analysis shows that extreme risk connectedness in China's stock market exhibits the natures of persistence and heterogeneity. Last, extreme risk connectedness at the sector level enables us to identify the Energy, Information Technology, Financials, and Telecommunication Service as "systemically important sectors" in China's stock market. Our findings offer another layer of insightful information available to academics, practitioners, and policy makers in terms of extracting valuable information for the real economy or forecasting purposes, aligning their investment horizons with their risk attitudes, and establishing more efficient regulatory mechanism.
economics,mathematics, interdisciplinary applications,management
What problem does this paper attempt to address?