Do dynamic linkages exist among European carbon markets?

In this paper we examine statistical relationships among European carbon markets from 2005 to 2010. We use a time-series approach using 1,220 daily (spot and forward) price data observations from Phase I and Phase II of the European Union Emissions Trading Scheme (EU ETS). Procedures such as unit root, cointegration, vector error correction models (VECMs), Granger causality, and generalised impulse response functions are employed in the analysis. The results reveal dynamic linkages among spot and forward carbon prices in Phases I and II, indicating that joint price discovery is taking place in carbon markets (at least in the short-run). However, evidence of constrained long-run information flows in Phase II, as indicated by the joint short- and long-run Granger causality testing, may be problematic for policy makers. This finding suggests that the coordinated policies designed to improve the operation and transparency of the EU ETS in Phase II may have actually been counter-effective. If carbon pricing mechanisms are dysfunctional, this has implications for the informational efficiency of carbon markets in Phase II and beyond, thus signalling the possibility of arbitrage and other profitable trading opportunities for market participants. Further research could attempt to address these issues.