A multi-model approach for carbon price forecasting: methods and applications

Mitigating the climate change is important for sustainable development, and carbon emission trading market plays an important role in optimizing the allocation of different emission reduction resources. To provide decision support for the low-carbon technology investment, it is necessary to obtain an accurate prediction of the future carbon emission prices. However, this task is challenging because various factors can affect the carbon emission prices, such as the new climate policies, emergence of advanced emission reduction technologies, updated emission market rules, etc. In contrast to the previous studies using a single prediction model, this study constructs a multi-model framework to predict the future prices of China's national carbon emission trading market, which consists of high-order polynomials model, grey prediction model and support vector machine (SVM) model. The results show that the carbon emission price will exhibit an increasing trend in the future and reach 130 yuan/ton in 2030. Moreover, we have also compared the prediction performance of different methods, and found that both grey prediction model and high-order polynomial method perform well in the long-term prediction of carbon prices.