This paper offers an innovative approach to capture the trend of oil futures prices based on the text‐based news. By adopting natural language processing techniques, the text features obtained from online oil news catch more hidden information, improving the forecasting accuracy of oil futures prices. We find that the textual features are complementary in improving forecasting performance, both for LightGBM and benchmark models. Besides, event studies verify the asymmetric impact of positive and negative emotional shocks on oil futures prices. The generated text‐based news features robustly reduce forecasting errors, and the reduction can be maximized by incorporating all features.

No abstract

This study explores the predictive effect of climate change attention on carbon futures returns. Using climate‐related Google Trends and news, we construct five dimensions of the public climate attention index and media climate attention index. After feature selection, we incorporate the optimized combination with lagged order into the machine learning model to predict EU Emission Allowance futures returns. Our empirical results show that the forecasting models with climate attention outperform the corresponding benchmark models, indicating that climate attention does provide predictive information for carbon futures returns. In addition, we carry out trading simulations to investigate the economic performance of the forecast results. It turns out that the market strategies based on the prediction models with climate attention can deliver more benefits than the counterpart market strategies. More specifically, the cumulative returns reach 140% during the out‐of‐sample period, much higher than 79% of the cumulative returns of the buy‐and‐hold strategy.

No abstract

No abstract