Paper: arXiv 2406.05094

Authors: Cristiano Salvagnin, Aldo Glielmo, Maria Elena De Giuli, Antonietta Mira

Abstract

The European carbon market plays a pivotal role in the European Union’s ambitious target of achieving carbon neutrality by 2050. Understanding the intricacies of factors influencing European Union Emission Trading System (EU ETS) market prices is paramount for effective policy making and strategy implementation. We propose the use of the Information Imbalance, a recently introduced non-parametric measure quantifying the degree to which a set of variables is informative with respect to another one, to study the relationships among macroeconomic, economic, uncertainty, and energy variables concerning EU ETS prices. Our analysis shows that in Phase 3 commodity related variables such as the ERIX index are the most informative to explain the behaviour of the EU ETS market price. Transitioning to Phase 4, financial fluctuations take centre stage, with the uncertainty in the EUR/CHF exchange rate emerging as a crucial determinant. These results reflect the disruptive impacts of the COVID-19 pandemic and the energy crisis in reshaping the importance of the different variables. Beyond variable analysis, we also propose to leverage the Information Imbalance to address the problem of mixed-frequency forecasting, and we identify the weekly time scale as the most informative for predicting the EU ETS price. Finally, we show how the Information Imbalance can be effectively combined with Gaussian Process regression for efficient nowcasting and forecasting using very small sets of highly informative predictors.

Complexity vs Empirical Score

  • Math Complexity: 6.5/10
  • Empirical Rigor: 7.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs advanced non-parametric measures (Information Imbalance) and Gaussian Processes, indicating moderate-to-high mathematical density. The analysis involves real-world EU ETS price data across two phases with specific variable selection, forecasting, and nowcasting results, demonstrating strong empirical backing and implementation relevance.

Research Flowchart

  flowchart TD
  A["Research Goal:<br>Determine EU ETS price<br>determinants"] --> B{"Key Methodology:<br>Information Imbalance &<br>Gaussian Process Regression"}
  
  C["Data Inputs:<br>Macro, Economic,<br>Uncertainty, Energy Vars"] --> B
  
  B --> D["Computational Process:<br>Non-parametric Analysis<br>of Variable Relationships"]
  
  D --> E["Phase 3 Outcome:<br>Commodity/ERIX Index<br>most informative"]
  
  D --> F["Phase 4 Outcome:<br>Financial Fluctuations<br>& EUR/CHF Uncertainty dominant"]
  
  B --> G["Forecasting Outcome:<br>Weekly Scale optimal<br>for prediction"]

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper proposes a non‑parametric approach using the Information Imbalance measure, combined with Gaussian Process regression, for mixed‑frequency nowcasting and forecasting of EU ETS prices.

Source sentence
We propose the use of the Information Imbalance, a recently introduced non‑parametric measure quantifying the degree to which a set of variables is informative with respect to another one, to study the relationships among macroeconomic, economic, uncertainty, and energy variables concerning EU ETS prices.

Assumes 4

  • The Gaussian Process mean function is assumed to be zero. model
    Source sentence
    Without loss of generality, the GP mean function is typically assumed to be zero as the GP is typically applied to standardised data.
  • A smoothness parameter ν is set to 1.5, the largest degree of smoothness compatible with all the time series available. model
    Source sentence
    In all our experiments, we set the smoothness parameter, ν, to 1.5, the largest degree of smoothness compatible with all the time series available
  • Mixed‑frequency data can be aggregated using Gaussian Processes with a low noise level σ²_n = 10⁻³ for regularisation. data
    Source sentence
    a GP is fit to the data using a low noise level of σ2 n = 10−3 just needed to regularise the inversion in Eq. 4, and the posterior GP mean is then used to compute the value of the time series at the needed
  • Information Imbalance does not depend on any model assumptions about the underlying time series relationships. statistical
    Source sentence
    Unlike conventional parametric econometric models, the Information Imbalance does not depend on any assumption on the model behind the time series and their relationships.

Data

  • EUA (EUA) price series · Bloomberg® market-prices commercial
    January 2014 - April 2023
    Source sentence
    EUA (EUA) January 2014 - April 2023 Bloomberg®
  • ERIX index · Bloomberg® other commercial
    January 2014 - April 2023
    Source sentence
    ERIX index January 2014 - April 2023 Bloomberg®
  • Euro-area GDP (current value) · Eurostat macro public
    January 2014 - April 2023
    Source sentence
    Euro-area GDP (current value) January 2014 - April 2023 Eurostat
  • Euro-area inflation (HICP) · Eurostat macro public
    January 2014 - April 2023
    Source sentence
    Euro-area inflation (HICP) January 2014 - April 2023 Eurostat
  • Uncertainty EUR/CHF (CAFZUECH) · Bloomberg® other commercial
    January 2014 - April 2023
    Source sentence
    Uncertainty EUR/CHF (CAFZUECH) January 2014 - April 2023 Bloomberg®

Recognised sources: Bloomberg

Costs

Not applicable The paper reports no trading or execution result.

Finds 5

  • In Phase 3 the most informative variables are the ERIX index, EUROSTOXX Electricity index and the 3‑month bond yield. statistical
    Source sentence
    ERIX index, EUROSTOXX Electricity index, 3-month Bond yield were the top 3 most informative variables for Phase 3.
  • In Phase 4 the most informative variables are uncertainty on the EUR/CHF exchange rate, coal prices and the EUR/USD spot rate. statistical
    Source sentence
    For Phase 4, the most informative variables include Uncertainty on EUR/CHF exchange rate, Coal prices, and EUR/USD Spot rate.
  • Commodity‑related variables are highly correlated with EUA price in Phase 3, while uncertainty‑related variables are poorly correlated in Phase 3 but become more positively correlated in Phase 4. statistical
    Source sentence
    during Phase 3, the uncertainty-related variables are poorly correlated with the target variable. Especially during Phase 4 most of the considered variables have a positive correlation
  • The Information Imbalance approach identifies weekly frequency as the most informative time scale for predicting EU ETS prices. statistical
    Source sentence
    we identify the weekly time scale as the most informative for predicting the EU ETS price.
  • Combining Information Imbalance with Gaussian Process regression enables efficient nowcasting and forecasting using very small sets of highly informative predictors. performance
    Source sentence
    Finally, we show how the Information Imbalance can be effectively combined with Gaussian Process regression for efficient nowcasting and forecasting using very small sets of highly informative predictors.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 19 related papers). Coverage is partial, so this is neither replication nor refutation.

Cited by 3 papers in the archive (1 by independent authors)

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.

Cited References (60)

  1. Price determination in the EU ETS market: Theory and econometric analysis with market fundamentals (2013) — link
  2. Predicting carbon allowance prices with energy prices: A new approach (2021) — link
  3. Price drivers and structural breaks in European carbon prices 2005–2007 (2008) — link
  4. Robust determinants of CO2 emissions (2021) — link
  5. Bond returns and market expectations (2014) — link
  6. Coal phase-outs and carbon prices: Interactions between EU emission trading and national carbon mitigation policies (2020) — link
  7. Exchange rate expectations and economic policy uncertainty (2017) — link
  8. Modeling the price dynamics of CO2 emission allowances (2009) — link
and 52 more
  1. Is the ETS an effective environmental policy? Undesired interaction between energy-mix, fuel-switch and electricity prices (2022) — link
  2. Europe beyond coal–An economic and climate impact assessment (2022) — link
  3. Forecasting carbon futures volatility using GARCH models with energy volatilities (2013) — link
  4. Carbon futures and macroeconomic risk factors: A view from the EU ETS (2009) — link
  5. A model of carbon price interactions with macroeconomic and energy dynamics (2011) — link
  6. Macroeconomics, finance, commodities: Interactions with carbon markets in a data-rich model (2011) — link
  7. Options introduction and volatility in the EU ETS (2011) — link
  8. A conditional dependence approach to CO2-energy price relationships (2019) — link
  9. The relationship between carbon-intensive fuel and renewable energy stock prices under the emissions trading system (2022) — link
  10. European Commission - EU ETS (2023) — link
  11. Renewable Energy Directive (2018) — link
  12. Carbon price drivers: Phase I versus Phase II equilibrium? (2012) — link
  13. Geopolitical Risk Index (2022) — link
  14. Measuring exchange rate risks during periods of uncertainty (2022) — link
  15. Macroeconomics and the reality of mixed frequency data (2016) — link
  16. Mixed data sampling (MIDAS) regression models (2020) — link
  17. DADApy: Distance-based analysis of data-manifolds in Python (2022) — link
  18. Ranking the information content of distance measures (2022) — link
  19. Analyzing spillover effects between carbon and fossil energy markets from a time-varying perspective (2021) — link
  20. Does R&D investment in renewable energy technologies reduce greenhouse gas emissions? (2022) — link
  21. Energy prices and CO2 emission allowance prices: A quantile regression approach (2014) — link
  22. What explain the short-term dynamics of the prices of CO2 emissions? (2014) — link
  23. Nonlinear dependence and connectedness between clean/renewable energy sector equity and European emission allowance prices (2021) — link
  24. Allowance price drivers in the first phase of the EU ETS (2010) — link
  25. Information linkage, dynamic spillovers in prices and volatility between the carbon and energy markets (2018) — link
  26. What happens to the relationship between EU allowances prices and stock market indices in Europe? (2019) — link
  27. Does carbon asset add value to clean energy market? Evidence from EU (2021) — link
  28. Causalities between CO2, electricity, and other energy variables during phase I and phase II of the EU ETS (2010) — link
  29. Carbon Prices and Inflation in the Euro Area (2024) — link
  30. Wavelet feature extraction for high-dimensional microarray data (2009) — link
  31. The determinants of CO2 prices in the EU emission trading system (2022) — link
  32. CO2 prices, energy and weather (2007) — link
  33. A deep learning-based method for forecasting gold price with respect to pandemics (2021) — link
  34. Orthogonal matching pursuit: Recursive function approximation with applications to wavelet decomposition (1993) — link
  35. Carbon prices forecasting in quantiles (2022) — link
  36. The interrelationship between the carbon market and the green bonds market: Evidence from wavelet quantile-on-quantile method (2022) — link
  37. Gaussian processes for time-series modelling (2013) — link
  38. Sparse orthogonal variational inference for Gaussian processes (2020) — link
  39. On the use of the wavelet decomposition for time series prediction (2002) — link
  40. Does renewable energy affect fossil fuel price? A time–frequency analysis for the Europe (2023) — link
  41. Forecasting European carbon returns using dimension reduction techniques: Commodity versus financial fundamentals (2022) — link
  42. How connected is the carbon market to energy and financial markets? A systematic analysis of spillovers and dynamics (2020) — link
  43. Extreme linkages of carbon futures, energy markets, and economic indicators: A copula approach (2023) — link
  44. Multilevel wavelet decomposition network for interpretable time series analysis (2018) — link
  45. Which exogenous driver is informative in forecasting European carbon volatility: Bond, commodity, stock or uncertainty? (2023) — link
  46. The dynamic spillover between carbon and energy markets: new evidence (2018) — link
  47. The impact of the global stock and energy market on EU ETS: A structural equation modelling approach (2021) — link
  48. Gaussian processes for machine learning (2006) — link
  49. Deep kernel learning (2016) — link
  50. On the causality between export growth and GDP growth: an empirical reinvestigation (1996) — link
  51. Asymmetric risk spillover between financial market uncertainty and the carbon market: A GAS–DCS–copula approach (2020) — link
  52. Usefulness of economic and energy data at different frequencies for carbon price forecasting in the EU ETS (2018) — link

0 of 60 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.