Abstract: A scientific prediction of stock volatility is of significant theoretical and practical importance for risk measurement and management by investors and regulatory authorities. Based on investor sentiment and the characteristics of time-varying volatility, asymmetry, fat-tailed distribution, and long memory, an SMA-Realized AHAR GARCH CICSI model under the skewed-t distribution is constructed for volatility forecasting and risk measurement. Using high-frequency data of the CSI 300 Index, we estimate the model parameters through the maximum likelihood method. VaR posterior test, MCS test, out-ofsample R² test, and the economic value evaluation are employed to comprehensively assess the volatility forecasting performance of the model. The empirical results indicate that the proposed SMA-Realized AHAR GARCH CICSI model achieves superior volatility fitting performance and higher predictive accuracy.