A dynamic analysis of price forecast models of carbon credits

Abstract

Climate change is a global issue concerning scientists and governments alike. Dealing with this issue is at the top of the international agenda. The primary cause of the human-induced greenhouse effect is found to be the carbon dioxide (CO2), deriving mainly from burning fossil fuels, industrial activity and deforestation. Under the Kyoto Protocol, the EU has committed to reduce its total greenhouse gas emissions through a flexible mechanism, the EU Emissions Trading System (EU ETS). Here, emission allowances in the form of a new tradable asset, the European Union Allowance (EUA), can be traded in organised financial markets. Reducing effectively carbon emissions depends upon the success of such a carbon market, which requires sufficient prediction of the price behaviour. The tradable carbon credits appear to be influenced by fuel prices in the energy sector and by certain economic indicators. This thesis presents statistical models for efficient price forecasting based on the relationships between the emission spot and futures prices with energy and several industrial and economic indicators to empirically study the emission price behaviour. Due to the different Kyoto-established trading commitment phases of emission spot prices and their volatility behaviour, we propose appropriate AR–GARCH models for stochastic futures price modelling. We find that various energy and industrial variables affect the formation of the emission futures prices and should be incorporated in price forecasting models.