Integrated Prediction of Carbon Price in China Based on Heterogeneous Structural Information and Wall-Value Constraints
Qingqing Sun,Hong Chen,Ruyin Long,Jiawei Chen
DOI: https://doi.org/10.1016/j.energy.2024.132483
IF: 9
2024-01-01
Energy
Abstract:Carbon price is not a "free-flowing" product of market transactions, but an economic symbol with political attributes that has a bearing on the endogenous security of China's national economic development. This study takes into account the key factors affecting carbon price, such as international carbon quota trading price, changes in macroeconomic environment, basic energy price and public concern about climate change. Based on the high-frequency, low-frequency, trend term decomposition and integrated noise reduction processing of carbon price data by DWT, CEEMDAN and SSA, the Transformers prediction model is constructed. Further, based on the combination of scenarios and the current status of power enterprises' revenues, a real-world carbon market "tolerance wall value" and a cut-off point are determined. The results show that (1) the predictions of the models using combined modal data are all better than those using unimodal and raw data. (2) The "wall value" of carbon price is 117.6 yuan/ton, 142.3 yuan/ton, and 182.7 yuan/ton under the low carbon follow (LCF) basic as usual (BAU), and economic follow (EF) scenarios, respectively, and all of them are reached by 2029. (3) Measured with realistic data, China's electric power enterprises have paid more than the value of profits (4 % of profits) with a low tolerance level to take part in carbon trading in the market. And under the EF and BAU scenarios, with high tolerance as a parameter, the trading price in China's carbon market will reach the wall value level (119.5yuan/ton) in June 2025 and October 2026, respectively. This research contributes to a more thorough understanding and forecasting of the function and future trajectory of China's carbon market and carbon pricing system.