Anticipating correlations between EUAs and CERs: a Dynamic Conditional Correlation GARCH model

Previous literature has studied the empirical characteristics of European Union Allowances (EUAs) and Certified Emissions Reductions (CERs) time series by using vector autoregression, impulse response function, and cointegration analysis (Chevallier (2010)). This paper extends the analysis by modelling the inter-relationships between EUAs and CERs in a multivariate GARCH econometric framework, so as to reflect the dynamics of the correlations between the variables overtime. Using the DCC MGARCH model by Engle and Sheppard (2001) and Engle (2002) on daily data from March 09, 2007 to January 26, 2010, we confirm the presence of strong ARCH and GARCH effects. Besides, we provide strong empirical evidence of time-varying correlations in the range of [0.01;0.90] between EUAs and CERs that have not been considered by previous studies. Thus, our study shows that the correlations between EUAs and CERs extracted from the DCC MGARCH model appear as a useful tool to comprehend the nature of the inter-relationships between these two markets, and to reach optimal risk management, portfolio selection, and hedging as called by Engle (2009).