The role of dynamic measurement and early warning in China’s stock-market resilience: evidence from 28 industry sectors

Tianxi Cheng, Xiaohuizi Tan, Xiangcheng Meng and Cheng Li

Abstract

This study quantifies the stock resilience of Chinese industrial sectors across two dimensions: absorption intensity and duration. Drawing on the theoretical framework of Yilin Chen et al, we estimate sectoral responses to shocks (proxied by the Chicago Board Options Exchange’s Volatility Index) using a time-varying parameter vector autoregression model. Our empirical results reveal significant time-variation in resilience, with sharp declines during the 2007–9 global financial crisis, the 2015–16 turmoil in China’s stock markets and the 2020–23 Covid-19 pandemic. We find substantial industry heterogeneity: defensive sectors such as banking are robust, while cyclical sectors such as real estate are vulnerable. Notably, the total correlation index (TCI) reaches 47.19%, indicating strong risk integration. Low-resilience sectors act as central transmitters in the Diebold–Yilmaz spillover network. Finally, a long short-term memory-based model effectively predicts low-resilience states, providing a tool for financial stability monitoring.