Investor sentiment and predictability for volatility on energy futures Markets: Evidence from China

Abstract

This research specifically reveals the predictability for the volatility on energy futures markets when involving investor sentiment, using the newly launched China’s INE crude oil futures as an incremental evidence. First, we propose a novel investor sentiment index captures the feature of instant sentiment conversion and the internet attention for energy futures markets. Second, we present the nexus between the proposed novel investor sentiment and conditional volatility, which shows that the proposed investor sentiment significantly impacts on return and volatility. It indicates that the emerging markets filled with noise traders like China are more deeply influenced by investor sentiment and bring about more volatility. Third, the results of dual leverage effects in modeling energy futures volatility shows that the optimistic (pessimistic) sentiment shift has negative(positive) effects on volatility, and downward (upward) shocks from bad news (good news) are followed by greater volatility. Fourth, incorporating the novel investor sentiment into models can significantly improve volatility forecasting accuracy because of its smaller loss function results compared with the benchmark models without sentiment. Moreover, high frequency realized volatility and range-based volatility are regarded as real volatility to examine the predictability, and Diebold Mariano test is used to present the difference of forecasting ability. Our results are robust by using the data from China’s Thermal Coal Futures. Overall, we present that the novel investor sentiment can be utilized for monitoring and predicting volatility shocks for emerging markets with lots of noise traders.