This paper applies the dynamic mixture copula model method and proposes a mobility measure of the marginal expected shortfall (MES) to depict the changing systemic risk in China’s mainland stock market and Hong Kong’s stock market. The empirical analysis focuses on the rank of individual stocks’ MES and the changes in the rank. The empirical results suggest that the mobility of MES is strong in the stable stage of the market, while the MES ranking of stocks tends to be disordered. When risk events occur, the MES ranking is relatively stable and the market risk is always affected by the same specific stocks. This property gives us a new perspective on risk management and deepens our understanding of the characteristics of systemic risk. In addition, most of the MES time series have strong long-term memory, which indicates that the risks of individual stocks have persistent impacts on the market’s systemic risk.