Carbon price prediction model based on EMD-ARMA-GRACH

The carbon market serves as a potent tool for attaining the objectives of Carbon Peaking and Carbon Neutrality, and factors influencing the carbon price and their forecasts have a direct impact on the functioning of this market. In this paper, the Empirical Modal Decomposition (EMD) technique is employed to categorize carbon prices into high-frequency, low-frequency, and trend-based carbon prices. Subsequently, Autoregressive Moving Average (ARMA) and Generalized Autoregressive Conditional Heteroskedasticity (GARCH) models are applied to fit these categories, respectively, to establish a carbon price forecasting model; finally, the empirical research and comparative analysis are carried out on the daily transaction data of the Hubei Carbon Trading Market. The results show that the combination of EMD-ARMA-GARCH can be effectively used for carbon price prediction. The results of this research are conducive to improving the level of carbon price prediction and perfecting the development of carbon market in China.