Abstract
Examining and analyzing the role of breakpoints in carbon price prediction can help people better understand the carbon market's structural changes to carry out technological predictions. However, the current carbon price prediction models lack comprehensive utilization of structural breakpoints. Then, a hybrid forecasting model is proposed by combining the Bai&Perron test, Iterated Cumulative Sums of Squares algorithm (ICSS), wavelet transform, and long short-term memory neural network (LSTM). According to the results, there are fifteen breakpoints in the EU allowance (EUA) carbon price. And the breakpoints can improve the prediction accuracy of the proposed model by 10–20 %. The ICSS algorithm is superior to the Bai&Perron test in detecting the breakpoints and improving the prediction accuracy. Compared with other benchmark models, the proposed hybrid model has the best prediction accuracy. We suggest that two types of breakpoints information could be accounted for in future prediction models.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have