This paper presents an analysis of the European Emission Trading System as a transaction network. It is shown that, given the lack of well-identified trading institutions, industrial actors had to resort to local connections and financial intermediaries to participate in the market. This gave rise to a hierarchical structure in the transaction network. It is then shown that the asymmetries in the network induced market inefficiencies (e.g., increased bid-ask spread) and informational asymmetries, that have been exploited by central agents at the expense of less central ones. Albeit the efficiency of the market has improved from the beginning of Phase II, the asymmetry persists, imposing unnecessary additional costs on agents and reducing the effectiveness of the market as a mitigation instrument.