Uncovering Drivers of EU Carbon Futures with Bayesian Networks
Jan Maciejowski, Manuele Leonelli · 2025
Read Uncovering Drivers of EU Carbon Futures with Bayesian Networks by Jan Maciejowski, Manuele Leonelli (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-10384.txt.
Introduction
Maciejowski and Leonelli use in 2025 Bayesian networks to uncover what drives European carbon futures prices. The EU trading scheme issues allowances that big emitters buy and sell, a core tool for cutting greenhouse gases. Daily data span 2013 to 2025 across the scheme's third and fourth phases.
Both snapshot and dynamic networks were fit to energy, equity, currency, and bond indicators. Energy commodities, above all coal and oil futures, plus the European energy sector's performance, shaped allowance prices most. Wider market mood acted indirectly through energy demand. The dynamic model found a modest next day pull from oil, while most links were same day effects.
Catalog record
| CREATOR | Jan Maciejowski, Manuele Leonelli |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 67 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-10384
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Uncovering Drivers of EU Carbon Futures with Bayesian Networks Jan Maciejowski, Manuele Leonelli Abstract The European Union Emissions Trading System (EU ETS) is a key policy tool for reducing greenhouse gas emissions and advancing toward a net-zero economy. Under this scheme, tradeable carbon credits, European Union Allowances (EUAs), are issued to large emitters, who can buy and sell them on regulated markets. We investigate the influence of financial, economic, and energy-related factors on EUA futures prices using discrete and dynamic Bayesian networks to model both contemporaneous and time-lagged dependencies. The analysis is based on daily data spanning the third and fourth ETS trading phases (2013–2025), incorporating a wide range of indicators including energy commodities, equity indices, exchange rates, and bond markets. Results reveal that EUA pricing is most influenced by energy commodities, especially coal and oil futures, and by the performance of the European energy sector. Broader market sentiment, captured through stock indices and volatility measures, affects EUA prices indirectly via changes in energy demand. The dynamic model confirms a modest next-day predictive influence from oil markets, while most other effects remain contemporaneous. These insights offer regulators, institutional investors, and firms subject to ETS compliance a clearer understanding of the interconnected forces shaping the carbon market, supporting more effective hedging, investment strategies, and policy design. 1 Introduction The EU Emissions Trading System (ETS) is a cap-and-trade scheme designed to reduce greenhouse gas emissions across the EU, recently strengthened through the Fit for 55 package ( European Commission ) . Emissions-intensive sectors are required to hold European Union Allowances (EUAs) to offset their carbon output. One EUA grants the right to emit one tonne of CO 2 . If, by the surrender deadline, a firm does not hold enough EUAs to cover its emissions, it faces financial penalties and must make up the shortfall in the following compliance cycle. EUAs are distributed through annual auctions under a declining cap. Participants include all EU countries, Iceland, Norway, Liechtenstein, and Northern Ireland’s energy sector. Auctions occur through the European Energy Exchange in Germany, while surplus allowances may be retained for future use or traded on the secondary market, often via derivatives. Among these, futures contracts are the most liquid instruments and the focus of this study ( European Securities and Markets Authority 2024 ) . EUA futures enable participants to hedge price risk, with delivery typically set in December at the end of the compliance year. Certain carbon-intensive sectors receive free allowances to reduce the risk that firms relocate production to countries with less restrictive emissions regulations. These free allocations are gradually being phased out. As the ETS directly affects energy prices and production costs, it has been central to EU political and economic discussions. Revenues from EUA auctions are intended to fund the transition to a clean economy, especially in the energy sector. However, rising EUA prices can burden energy consumers, particularly in countries with more carbon-intensive generation. This effect is often amplified by volatility in fossil fuel markets. Understanding the drivers of EUA pricing is essential for stakeholders across policy, compliance, and finance. As an indirect form of carbon taxation, EUA revenues must be strategically allocated to support the energy transition. At the same time, compliance sectors, including energy, heat, and petroleum production, face fluctuating EUA prices and must manage risk accordingly. Financial institutions active in the market also require tools for forecasting and price discovery. A clearer understanding of EUA price drivers enables all participants to anticipate trends and volatility, improving decision-making and reducing market inefficiencies. Prior studies have examined the economic and policy factors shaping EUA prices using various statistical approaches. We extend this work by incorporating a broader set of financial and energy-related indicators and modelling their interdependencies through discrete and dynamic Bayesian networks (BNs) ( Koller and Friedman 2009 ; Murphy 2002 ; Pearl 1988 , e.g.) . These models provide a probabilistically grounded, data-driven representation of the EUA pricing system, capturing both contemporaneous and time-lagged effects. BNs are well suited to this setting: they represent multivariate relationships through conditional dependencies, offering interpretability and transparency that many black-box machine learning methods lack ( Rudin 2019 ) . A standard (static) BN captures associations among variables on the same day, while its dynamic counterpart introduces temporal structure to model next-day influences. We use both static and dynamic BNs to analyze how energy, financial, and macroeconomic signals interact to shape EUA pricing. This framework reveals the interconnected structure of the variables driving carbon prices and enables the identification of key drivers, including coal and oil futures, energy sector equity performance, and broader market sentiment, alongside their contemporaneous and time-lagged effects. The dynamic extension quantifies temporal dependencies and complements the static model, offering a richer understanding of how shocks propagate across markets. Together, the models provide actionable insights for compliance entities, financial institutions, and regulators seeking to forecast EUA prices, manage risk, or design more effective policy interventions. 2 Literature review A wide range of academic and industry studies has sought to explain the drivers of EUA prices, drawing on variables from regulatory mechanisms, commodity markets, energy production, macroeconomic conditions, and financial speculation. Most of this literature relies on standard econometric techniques such as structural vector autoregressions ( Lovcha et al. 2021 ) , quantile regressions ( Tan and Wang 2017 ) , copula-based spillover analysis ( Hanif et al. 2021 ) , and structural equation modelling ( Wang and Zhao 2020 ) . While these approaches have yielded important insights, they are often limited in capturing the joint structure of interdependencies or the time-propagation of effects. In this section, we summarize relevant findings across key domains, highlighting evidence that motivates the inclusion of specific variables in our model and clarifying where our approach extends the current state of the art. 2.1 Regulatory mechanisms The supply of EUAs is governed by a set of regulatory instruments that play a fundamental role in shaping price dynamics in the EU ETS. Among these, the Market Stability Reserve (MSR) and the Linear Reduction Factor (LRF) are the most influential mechanisms introduced to tighten the system over time and address imbalances in supply and demand. The MSR, introduced in 2019 and strengthened under the Fit for 55 package, was designed to manage the oversupply of allowances that characterized the first two trading phases. It adjusts auction volumes based on the Total Number of Allowances in Circulation (TNAC): if TNAC exceeds 833 million, 24% of the surplus is withheld and placed in reserve; if TNAC falls below 400 million, 100 million allowances are released. Under the revised rules introduced in the Fit for 55 package, any allowances held in the MSR above the 400 million threshold are invalidated, permanently tightening supply ( Borghesi et al. 2023 ; European Commission ; Salvagnin et al. 2024 ) . The LRF, which defines the annual rate at which the emissions cap declines, was increased from 2.2% to 4.3% under the Fit for 55 reform, with the goal of achieving a 55% reduction in greenhouse gas emissions by 2030 compared to 1990 levels. These supply-side constraints are critical drivers of long-term EUA price appreciation, as they limit the availability of allowances in each compliance year ( European Commission )