RETRACTED ARTICLE: Forecasting carbon emissions future prices using the machine learning methods.

Area of Science:

• Environmental Economics
• Energy Markets
• Climate Policy

Background:

• Uncertainty exists in the coupling/decoupling of natural gas, oil, and energy commodity futures.
• Forecasting energy and carbon futures is crucial for environmental sustainability.
• Understanding price interactions informs climate and energy policy.

Purpose of the Study:

• To investigate the interactions between energy commodity futures, oil price futures, and carbon emission futures.
• To analyze these interactions from a forecasting perspective.
• To provide insights for environmental sustainability and policy-making.

Main Methods:

• Utilized daily data from January 2018 to October 2021 for natural gas, crude oil, carbon, and energy commodity futures prices.
• Applied machine learning techniques: Multiple Linear Regression (MLR), Artificial Neural Network (ANN), Support Vector Regression (SVR), and Long Short-Term Memory (LSTM).
• Compared the performance of linear and nonlinear modeling frameworks.

Main Results:

• Nonlinear machine learning frameworks (ANN, SVR, LSTM) significantly outperform linear models (MLR) in capturing relationships between oil prices and carbon emission futures.
• Extreme price movements in oil and natural gas futures exhibit nonlinear effects on carbon emission futures prices.
• Machine learning models effectively identified complex, nonlinear dynamics in energy and carbon markets.

Conclusions:

• Nonlinear dynamics are essential for accurately forecasting energy and carbon futures, especially during periods of extreme price volatility.
• Findings support policymakers in designing effective climate and environmental strategies and managing energy market fluctuations.
• The study highlights implications for achieving Sustainable Development Goals (SDGs) 7 (Affordable and Clean Energy) and 12 (Responsible Consumption and Production).