Thesis title: The price of the emissions and its anomalies: an investigation into carbon market dynamics
This dissertation addresses four key gaps in the literature on the European Emissions Trading System (EU ETS). First, while numerous studies have analyzed the price dynamics of CO2, less attention has been paid to identifying speculative behavior in the first two phases of the mechanism, i.e. 2005 – 2007 and 2008 – 2012. Moreover, only a few studies have focused on the last two phases, i.e. 2013 – 2020 and 2021 – 2030, characterized by significant regulatory adjustments, geopolitical tensions, and the consequences of the Covid – 19 pandemic. To fill this gap, the analysis extends the speculative and hedging ratio proposed by Lucia et al. (2015) to Phases III and IV, investigates cross-maturity interactions through a Vector Autoregressive Model (VAR), and assesses the impact of geopolitical shocks using Bai–Perron test, Local Projection, and GARCH model. A comparative event study of EUA futures (2022–2025) further highlights short-term market reactions to major events.
Second, the strategic behavior of market participants remains underexplored, especially in the Italian market, one of the most important in terms of emissions. By applying Gaussian Mixture Model (GMM) to ETS Registry data, this study identifies behavioral clusters unobservable with traditional techniques, highlighting trading strategies, market concentrations, and opportunistic dynamics that have been poorly documented so far.
Third, although market structure is central to the debate on integrity and transparency, empirical analyses of trading networks based on real transaction data are lacking. This work reconstructs the dynamic network of ETS trade in Italy from 2013 to 2023, offering a unique perspective through network analysis, pattern identification and anomaly detection techniques.
Finally, literature pays little attention to future developments of the ETS mechanism, especially given the introduction of the parallel ETS2 mechanism starting in 2027. The few studies available treat the current ETS1 and the new ETS2 separately. To this end, this thesis proposes an integrated model based on State-Space models and Monte Carlo simulations, which allows for a consistent comparison of alternative price scenarios and emissions trajectories in line with the 2030 targets.