A Hybrid Model for Analyzing the Effect of the Carbon Border Adjustment Mechanism on the Historical Volatility of the EU ETS

Abstract

The 1st of October 2023 the Carbon Boarder Adjustment Mechanism (CBAM) transitional phase began. The CBAM is a tool created by the European Union (EU) to reach the target of 55% reduction in greenhouse gas emission levels in 2030 compared with 1990 levels. The CBAM imposes customs on sectors at high risk of carbon leakage. This thesis aims to analyze the historical volatility of the EU Emissions Trading System (ETS) for the end of phase 3 and for phase 4 until the 29th of February 2024. The data consists of the price of indexes related to the CBAM and the price of natural gas. Motivated by a methodology created by Amirshahi and Lahmiri (2023), Generalized Autoregressive Conditional Heteroskedasticity (GARCH), exponential GARCH, GlostenJagannathan-Runkle-GARCH models are created based on different assumptions for the distribution of the residuals. In addition, a model is created using Long Short-Term Memory (LSTM). The optimal GARCH-type model for each variable is used to create a hybrid GARCH-LSTM model. Model performance is compared based on the root mean squared error (RMSE) and the mean absolute error (MAE). The results show that the GARCH-LSTM model outperforms the alternatives in terms of RMSE and MAE. In addition, the model shows that the predictability increases in phase 4. The thesis provides EU with a tool to determine the right policies to ensure a carbon price that does not undermine the main goals of the CBAM.