Carbon price prediction for China's ETS pilots using variational mode decomposition and optimized extreme learning...

Area of Science:

• Environmental Economics
• Climate Policy Analysis
• Financial Forecasting

Background:

• China's national goals for carbon peak (2030) and carbon neutrality (2060) drive interest in Emissions Trading System (ETS) pilot carbon prices.
• Understanding carbon price dynamics is essential for effective carbon resource management and emissions allocation.
• Carbon price fluctuations indicate resource scarcity, necessitating accurate prediction models.

Purpose of the Study:

• To propose a hybrid feature-driven forecasting model for China's ETS pilot carbon prices.
• To address the non-stationary, nonlinear, and chaotic characteristics of carbon price data.
• To enhance carbon asset management and inform emissions allocation strategies.

Main Methods:

• Decomposition of carbon price series using Variational Mode Decomposition (VMD).
• Reconstruction of decomposed series based on Sample Entropy (SE).
• Prediction of subsequences using Extreme Learning Machine (ELM) optimized by Particle Swarm Optimization (PSO), followed by ensemble summation.

Main Results:

• Empirical validation on China's ETS pilot data demonstrates the proposed model's superior efficiency compared to benchmark models.
• The model effectively captures the complex characteristics of carbon price series.
• Forecasted carbon prices are expected to rise post-COVID-19, impacting policy and corporate commitments.

Conclusions:

• The hybrid decomposition-reconstruction-prediction-ensemble model offers an efficient approach to forecasting carbon prices in China's ETS pilots.
• The findings support improved government policies, including potential Border Carbon Adjustments (BCA).
• The model aids global companies in meeting "net zero" commitments and navigating evolving carbon markets.