Using Deep Learning to Predict Energy Stock Risk Spillover Based on Co-Investor Attention

The risk spillover of the energy stock market has become a research hot spot in the field of energy finance. Investors can avoid risks through the risk spillover relationship between listed companies, and regulators can warn of systemic risks. Therefore, it is necessary to predict the risk spillover of the energy stock market. This paper proposes a new forecasting framework to predict the risk spillover between different listed energy companies using the common concerns of investors for such companies. The results show that the deep neural network with the targeted loss function constructed in this paper has higher prediction accuracy