| Abstract: |
| We analyze the short-term spot prices of carbon dioxide (CO2) emission allowances of China Hubei Emission Exchange (CHEE, 2015-2019) and EU Emissions Trading System (EU ETS, 2008-2019). We feed predictive variables, monetary policy proxy variables including interest rates and statutory reserve ratio, energy consumption indicators, macroeconomic indicators including GDP, economic recession indices, and prices related features including lagged prices and realized volatility to machine learning models including linear regression, xgboost, and LSTM. We examine the approaches by conducting a one-step-ahead forecasting analysis. By comparing the results from these two markets, we analyze how different features effect these markets and for EU ETS, we see how the economic recession effects the trading. |
|