The European Union Emissions Trading System (EU ETS) plays a crit- ical role in Europe’s climate policy, regulating greenhouse gas emissions through a cap-and-trade mechanism. This thesis presents three analyses to deepen the understanding of the EU ETS market. First, a bibliomet- ric review of 367 research papers (2004-2024) examines citation patterns, co-authorship networks, and keyword trends to identify pivotal authors, in- stitutions, and themes, as well as uncovering research gaps. Notable find- ings include an annual growth rate of 12.99%, with research peaks corre- sponding to policy changes, and key contributions from Germany, China, and France. The review highlights emerging topics such as carbon pricing and market volatility, with a growing emphasis on financial implications. De- spite the progress, several research gaps can be identified in the development and application of non-parametric methods and the impact of macroeco- nomic variables. The second and third analyses address these gaps using innovative, non-parametric, and model- free methodologies. The Informa- tion Imbalance (II) method identifies the most predictive variables for EU ETS price forecasting, revealing that while Phase 3 was influenced by en- ergy indices and commodities, Phase 4 is dominated by financial variables and may be due to economic shocks like the COVID-19 pandemic and the energy crisis. The methodological results demonstrate that the Information Imbalance method effectively selects the most informative weekly timescale for accurate predictions and addresses the mixed-frequency problem by in- corporating macroeconomic variables into the analysis. Gaussian Processes (GP) further enhance prediction accuracy in nowcasting and forecasting sce- narios. Additionally, this thesis introduces the Differentiable Information Imbalance (DII), a non- linear, model-free approach, based on the Information Imbalance, that detects causal relationships within the EU ETS market. DII is compared with the Vector AutoRegression (VAR) model using syn- thetic data, demonstrating its ability to capture non-linear interactions. A new causal strength metric, Imbalance Gain (IG), is developed within the DII framework to measure the causal influence of predictor variables, similar to the F-statistic. The empirical analysis from 2013 to 2024 reveals signifi- cant causal effects of IBEX35 and coal futures on EUA prices, highlighting the intricate linkages between energy, financial, and carbon markets. By leveraging these advanced methodologies to fill research gaps, this work aims to enhance emissions trading systems and support robust climate policies to combat climate change.
Il Sistema di Scambio delle Emissioni dell'Unione Europea (EU ETS) svolge un ruolo cruciale nella politica climatica europea, regolando le emissioni di gas serra attraverso un meccanismo di cap-and-trade. Questa tesi presenta tre analisi per approfondire la comprensione del mercato dell'EU ETS. In primo luogo, una revisione bibliometrica di 367 articoli di ricerca (2004-2024) esamina i modelli di citazione, le reti di co-autorship e le tendenze delle parole chiave per identificare gli autori, le istituzioni e i temi più influenti, oltre a individuare lacune nella ricerca. Tra i risultati principali emergono un tasso di crescita annuo del 12,99%, con picchi di pubblicazioni in corrispondenza di cambiamenti normativi, e contributi chiave da Germania, Cina e Francia. La revisione evidenzia temi emergenti come la determinazione del prezzo del carbonio e la volatilità del mercato, con un’attenzione crescente agli impatti finanziari. Nonostante i progressi, permangono diverse lacune nella ricerca, in particolare nello sviluppo e nell'applicazione di metodi non parametrici e nell'analisi dell'impatto delle variabili macroeconomiche. La seconda e la terza analisi affrontano queste lacune utilizzando metodologie innovative, non parametriche e prive di modelli predefiniti. Il metodo Information Imbalance (II) identifica le variabili più predittive per la previsione dei prezzi dell'EU ETS, rivelando che, mentre la Fase 3 era influenzata dagli indici energetici e dalle materie prime, la Fase 4 è dominata dalle variabili finanziarie, probabilmente a causa di shock economici come la pandemia di COVID-19 e la crisi energetica. I risultati metodologici dimostrano che il metodo Information Imbalance seleziona efficacemente l’orizzonte temporale settimanale più informativo per previsioni accurate e affronta il problema delle frequenze miste incorporando variabili macroeconomiche nell’analisi. I Processi Gaussiani (GP) migliorano ulteriormente la precisione delle previsioni in scenari di nowcasting e forecasting. Inoltre, questa tesi introduce il Differentiable Information Imbalance (DII), un approccio non lineare e senza modelli basato sull’Information Imbalance, che rileva relazioni causali all'interno del mercato dell'EU ETS. Il DII viene confrontato con il modello Vector AutoRegression (VAR) utilizzando dati sintetici, dimostrando la sua capacità di catturare interazioni non lineari. Nell’ambitoel DII, viene sviluppata una nuova metrica della forza causale, Imbalance Gain (IG), che misura l’influenza causale delle variabili predittive in modo simile alla statistica F. L’analisi empirica dal 2013 al 2024 rivela effetti causali significativi dell’IBEX35 e dei future sul carbone sui prezzi delle EUA, evidenziando i complessi legami tra mercati dell’energia, mercati finanziari e mercato del carbonio. Sfruttando queste metodologie avanzate per colmare le lacune della ricerca, questo lavoro mira a migliorare i sistemi di scambio delle emissioni e a supportare politiche climatiche solide per contrastare il cambiamento climatico.
Exploring the European Union Emissions Trading System: Non-Parametric Models for Price Determinants, Forecasting, and Volatility Causality
SALVAGNIN, CRISTIANO
2025
Abstract
The European Union Emissions Trading System (EU ETS) plays a crit- ical role in Europe’s climate policy, regulating greenhouse gas emissions through a cap-and-trade mechanism. This thesis presents three analyses to deepen the understanding of the EU ETS market. First, a bibliomet- ric review of 367 research papers (2004-2024) examines citation patterns, co-authorship networks, and keyword trends to identify pivotal authors, in- stitutions, and themes, as well as uncovering research gaps. Notable find- ings include an annual growth rate of 12.99%, with research peaks corre- sponding to policy changes, and key contributions from Germany, China, and France. The review highlights emerging topics such as carbon pricing and market volatility, with a growing emphasis on financial implications. De- spite the progress, several research gaps can be identified in the development and application of non-parametric methods and the impact of macroeco- nomic variables. The second and third analyses address these gaps using innovative, non-parametric, and model- free methodologies. The Informa- tion Imbalance (II) method identifies the most predictive variables for EU ETS price forecasting, revealing that while Phase 3 was influenced by en- ergy indices and commodities, Phase 4 is dominated by financial variables and may be due to economic shocks like the COVID-19 pandemic and the energy crisis. The methodological results demonstrate that the Information Imbalance method effectively selects the most informative weekly timescale for accurate predictions and addresses the mixed-frequency problem by in- corporating macroeconomic variables into the analysis. Gaussian Processes (GP) further enhance prediction accuracy in nowcasting and forecasting sce- narios. Additionally, this thesis introduces the Differentiable Information Imbalance (DII), a non- linear, model-free approach, based on the Information Imbalance, that detects causal relationships within the EU ETS market. DII is compared with the Vector AutoRegression (VAR) model using syn- thetic data, demonstrating its ability to capture non-linear interactions. A new causal strength metric, Imbalance Gain (IG), is developed within the DII framework to measure the causal influence of predictor variables, similar to the F-statistic. The empirical analysis from 2013 to 2024 reveals signifi- cant causal effects of IBEX35 and coal futures on EUA prices, highlighting the intricate linkages between energy, financial, and carbon markets. By leveraging these advanced methodologies to fill research gaps, this work aims to enhance emissions trading systems and support robust climate policies to combat climate change.File | Dimensione | Formato | |
---|---|---|---|
Tesi Salvagnin corretto.pdf
accesso aperto
Dimensione
20.72 MB
Formato
Adobe PDF
|
20.72 MB | Adobe PDF | Visualizza/Apri |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.14242/214262
URN:NBN:IT:UNIBS-214262