Empirical Research on the Prediction Effect of Volatility Model Based on the Perspective of Investor Sentiment Health and Market Liquidity

Healthy investor sentiment has an important impact on stock market volatility. To avoid extreme volatility in the stock market, it becomes crucial to focus on the sentiment health of investors. Therefore, we establish the realized volatility model, the implied volatility model, and the historical volatility model, respectively, introduce investor sentiment and market liquidity into these models, and empirically study the forecast effects of future daily and weekly frequency volatility to play a role in the development of the physical industry and the improvement of organizational performance. The study found that the HAR-RV model has the strongest predictive ability, while the GJR-GARCH model's predictive effect is not ideal; market liquidity and high investor sentiment can improve the predictive effect of most models; compared with long-term (one-week) volatility, the effect of predicting short-term (one-day) volatility is better.