HAR model first stage estimation.

Climate risk has become a hot topic of global concern. This paper aims to explore the impact of investor climate sentiment (ICS) on China’ s crude oil futures market, covering the period from March 27, 2018, to December 30, 2022. Firstly, this paper employs the Thermal Optimal Path (TOP) method and discovers that the guiding effect of ICS on the volatility of crude oil futures (RVoil) intensifies over time, progressively becoming a pivotal factor in determining volatility. Secondly, based on the lead-lag relationship between ICS and RVoil, this study divides the sample period into five stages and confirms through the HAR model that ICS has a significant inhibitory effect on crude oil volatility during the guiding phase. In addition, incorporating ICS into the HAR model not only improves the model’ s goodness of fit but also significantly reduces the prediction error in out-of-sample forecasts. Finally, by comparing with the full-sample analysis, the volatility prediction results of the segmented samples show that during the guiding phase, the predictive power of ICS for crude oil market volatility is significantly improved. Even in the non-guiding phase, ICS can reduce the prediction error to a certain extent. This result further highlights the advantages of the TOP method in revealing the impact of ICS on the prediction of crude oil volatility.