Collective risk resonance in Chinese stock sectors uncovered through higher-order network analysis
Shanghai Jiao Tong University Journal Center
image:
(a) Volatility of sector x_i at time t; (b) Synchronized co-movements across sectors of varying orders; (c) Pre-filtered higher-order network representation, where isolated nodes represent 0-order co-movement, gray edges indicate 1-order co-movement, yellow triangles denote 2-order co-movement, blue quadrilaterals signify 3-order co-movement, and green pentagons represent 4-order co-movement. (d) higher-order co-movement relationships retained after threshold filtering.
view moreCredit: Zisheng Ouyang and Yaoxun Deng (Hunan Normal University, China) Tianlei Zhu (Beijing Jiaotong University, China)
Background and Motivation
Systemic financial risk remains a critical challenge for modern economies, underscored by recurring crises such as the 2008 global financial meltdown, the 2015 Chinese stock market crash, and the COVID-19 pandemic. Traditional research has often examined sectors in isolation or focused on pairwise risk spillovers, overlooking the complex, multi-sector dependencies that can amplify systemic threats. This study addresses that gap by exploring higher-order interactions—where risks resonate simultaneously across multiple sectors—within China’s stock market. By moving beyond conventional dyadic models, the research provides a more nuanced understanding of how collective risk behaviour shapes financial stability.
Methodology and Scope
Using the Reconstructing the Higher Order Structure of Time Series (RHOSTS) method, the authors construct dynamic higher-order networks to capture risk co-movement among 24 Chinese stock sectors from 2007 to 2024. Sectoral volatility is estimated via GJR-GARCH models, and hyperedges represent synchronised risk resonance across multiple sectors. Network topology metrics—such as higher-order degree, systemic importance, and clustering coefficient—are analysed at both sector and system levels. The study further integrates these metrics into a coupled-map-lattice model to quantify time-varying resilience during major crises, including the 2008 financial crisis, the 2015 market crash, and the COVID-19 pandemic.
Key Findings and Contributions
- Dominant Third-Order Resonance: The most prevalent risk pattern involves synchronous resonance among four sectors (third-order hyperedges), highlighting limitations of traditional pairwise models.
- Sectoral Heterogeneity: The insurance (INS) sector consistently shows high systemic importance, while energy (ENE) becomes central during geopolitical crises like the Russia-Ukraine conflict.
- Crisis-Specific Clusters: Core resonance groups shift with each crisis—e.g., {ENE, INS, DFI, TSE} post-2008, {TSE, RES, ACO, INS} post-2015, and {DFI, TSE, THA, SSE} post-COVID-19.
- Network Resilience: System-wide resilience exhibits an upward long-term trend, though it fluctuates significantly during stress periods. Financial sectors generally demonstrate higher shock-absorption capacity, while retailing (RET) and capital goods (CGO) are among the most vulnerable.
- Structural Shifts: Major events drastically alter network density, connectivity, and cluster formation, confirming that external shocks reconfigure risk transmission pathways.
Why It Matters
The study offers a paradigm shift in systemic risk analysis by capturing group-level risk synchronisation that traditional models miss. This approach reveals how multi-sector co-movements can accelerate contagion and create hidden vulnerabilities. By identifying crisis-specific resonance clusters and tracking resilience in real time, the research provides a more precise tool for monitoring and mitigating systemic threats in increasingly interconnected financial systems.
Practical Applications
- For Regulators: Enables dynamic monitoring of higher-order risk clusters and informs targeted policies, such as cross-sector exposure limits or circuit-breaker mechanisms for highly synchronised sectors.
- For Investors: Highlights the danger of over-concentrating portfolios in sectors prone to collective resonance—e.g., avoiding simultaneous heavy exposure to TSE, RES, ACO, and INS during turbulent periods.
- For Risk Management: Provides a framework to design hedging strategies that account for multi-sector dependencies, particularly for energy and climate-related financial risks.
- For Global Financial Stability: Demonstrates a scalable methodology for building real-time risk resonance surveillance systems in other markets.
Discover high-quality academic insights in finance from this article published in China Finance Review International. Click the DOI below to read the full-text! Open access for a limited time!
Journal
China Finance Review International
Method of Research
News article
Article Title
Collective risk resonance behavior and network resilience in Chinese stock sectors: evidence from higher-order financial network
High-frequency investor sentiment from online forums enhances stock return predictions
Shanghai Jiao Tong University Journal Center
image:
It represents mean posts per half-hour during a day, showing the intraday seasonality of message posted on stock forum. Intraday posting activities usually increase rapidly after the opening, decrease at lunch time and increase again before market close. This result is very similar to Renault (2017), which shows that the average number of messages during non-trading hours is lower than during trading hours.
view moreCredit: Xiaojun Chu and Yating Gu (Nanjing University of Information Science & Technology, China).
Background and Motivation
In behavioural finance, the relationship between investor sentiment and stock returns has long been recognised. However, most studies rely on sentiment data at the same frequency as the returns being forecast—such as daily, monthly, or quarterly. With the rise of digital platforms, high-frequency intraday sentiment data has become accessible, yet its potential to improve low-frequency return forecasts remains underexplored. Against this backdrop, China Finance Review International (CFRI) brings you a study titled “Does intraday high-frequency investor sentiment help forecast stock returns? Evidence from the MIDAS models”, which investigates whether high-frequency sentiment extracted from Chinese online stock forums can enhance the predictability of daily stock returns.
Methodology and Scope
The authors employ Mixed Data Sampling (MIDAS) models to integrate intraday high-frequency investor sentiment with daily stock returns of Chinese A-shares. Sentiment is constructed from over 6.7 million posts on the Eastmoney stock forum between 2014 and 2022, using a tailored Chinese financial sentiment dictionary. The study distinguishes between sentiment during trading hours (TS) and non-trading hours (PS, LS, AS), and compares the performance of various MIDAS specifications—including U-MIDAS, Beta, and Almon lag models—against a daily sentiment (DS) baseline.
Key Findings and Contributions
- High-frequency intraday sentiment significantly outperforms daily aggregated sentiment in forecasting daily stock returns.
- Sentiment during non-trading hours has stronger predictive power than sentiment during trading hours.
- Among MIDAS-class models, the U-MIDAS model delivers the best forecasting accuracy, both in-sample and out-of-sample.
- The study also introduces a novel, high-frequency sentiment proxy tailored to the Chinese market, filling a gap in the existing literature.
Why It Matters
China’s A-shares market is the world’s second-largest by capitalisation and is dominated by retail investors, who are more prone to sentiment-driven trading. Improving return predictability in such a market has significant practical implications for both domestic and international investors. This research demonstrates that intraday sentiment—especially from non-trading periods—can capture nuanced market dynamics that daily measures miss, offering a more timely and granular tool for forecasting.
Practical Applications
- Investors and Fund Managers: Can use intraday sentiment signals, particularly from non-trading hours, to refine trading strategies and asset allocation.
- Financial Analysts: May incorporate U-MIDAS models with high-frequency sentiment data for more accurate equity return forecasts.
- FinTech and Data Providers: Can develop sentiment tracking tools that segment data by trading vs. non-trading periods to enhance predictive analytics.
- Academic Researchers: The methodology and findings offer a framework for applying high-frequency sentiment analysis in other emerging markets.
Discover high-quality academic insights in finance from this article published in China Finance Review International. Click the DOI below to read the full-text original! Open access for a limited time!
Journal
China Finance Review International
Method of Research
News article
Article Title
Does intraday high-frequency investor sentiment help forecast stock returns Evidence from the MIDAS models
No comments:
Post a Comment