Abstract： This paper measures the stock index return comovements between China’s mainland and 7 critical countries or regions from 2006 to 2021 under a unified framework by using the Asymmetric Dynamic Conditional Correlation Model. Then we compare the comovements changes during the periods of Global Financial Crisis in 2008， European Debt Crisis in 2010， the stock market crash of China’s mainland stock market in 2015， and after the COVID-19 outbreak. Furthermore， we test the nonlinear threshold effect of comovements， and explore the equilibrium convergence or random walk characteristics of stock index return comovements in different thresholds. The main conclusions of this paper are as follows： The comovement between China’s mainland stock market and 7 major global stock markets conforms to the“geographical hypothesis”， and presents the characteristics of heterogeneity， complexity and variability； Comparing the characteristics of the comovements during market crisis period， the previous market crises did not strengthen the linkage between China’s mainland stock market and the global stock markets， and the comovements of the stock market even decreased significantly during the global financial crisis in 2008 and China’s stock market crash in 2015. After the CO VID-19 outbreak， China’s mainland stock market was gradually affected by the spillover effect of global market fluctuations， the correlation with the US stock market even increased by 66.84%on average； In addition， the stock index return comovement shows an unbalanced convergence or random walk state in different regions， especially the random walk state of the correlation between China’s mainland stock market and the stock markets of the United States and European countries in the high regional range， indicating that if there is a strong external shock， China’s mainland stock market may suffer a violent shock.

Keywords： Stock Index Return Comovements; ADCC-GARCH Model; SETAR Model; Time-varying Characteristics