ABSTRACT
The new energy industry is crucial for solving the problem of pollution, and its development requires support from the stock market. This paper proposes a Chinese investor sentiment index based on the Long Short-Term Memory (LSTM) deep learning method and investigates the effect of investor sentiment on new energy stock returns as well as value at risks (VaR) behavior before and during COVID-19. It also compares these effects on traditional energy companies to identify differences between the new energy and traditional companies. The empirical results show that investor sentiment has significant effects on stock returns and VaR of both new and traditional energy companies but the effects are stronger in the new energy industry. The effects of investor sentiment have increased during COVID-19, and investors pay more attention on risks than returns during COVID-19. These results provide guidance for small and medium-sized investors in China to optimize their investment strategies and avoid heavy losses associated with extreme risks.