The rollout of the EU’s CBAM has recently driven a great deal of criticism, as the European Commission has been slow in delivering key implementation details. CBAM benchmarks are one of the missing pieces that is driving a great deal of consternation for importers and third country producers. Without these values, industrials within the scope of the CBAM Regulation have little visibility as to what costs they will face as the border levy enters its ‘definitive phase’ in 2026.
The plan, as announced by the European Commission, is to release ‘provisional CBAM benchmarks’ in Q4 2025 and the final values in Q1 2026. The delay is due to the ongoing, parallel process by which the Commission is determining the new EU ETS benchmark values for 2026-2030, also scheduled for release in Q1 2026.
CBAM benchmarks will determine the portion of emissions exempt from carbon costs. These values are intended to mirror the product-level benchmarks that set the coverage of free allocation under the EU ETS.
Benchmarks are one of the key elements of the formula that determines how much free allocation an industrial installation receives. The ETS benchmark values are determined based on the emissions intensity of the 10% most efficient installations. Initial benchmark values were established for the 2013-2020 allocation period, with subsequent benchmark updates determined based on industrial progress towards improved emissions intensity and policy-determined minimum and maximum reduction rates. These minimum and maximum rates were set at 0.3% and 2.5% respectively during the last EU ETS reform process.
For the current update process, the actual emissions intensity for best performers in 2023 has been determined through a data gathering process. A line is drawn between the initial benchmark and the 2023 actual values, the slope of which is used to extrapolate out to 2028. If this extrapolated value falls between the minimum and maximum values described above, it becomes the new benchmark (see Figure 1). If it falls outside that range, then either the minimum or maximum threshold becomes the benchmark.
For the EU ETS benchmarks, we have used a multi-tiered approach. First, if the actual values for the years 2016 and 2017 (from the 2021 update) were outside of the new min/max range, the minimum or maximum value was used as appropriate.
Second, in order to provide a conservative estimate, we assume that the efficiency rate observed in 2016-2017 remains the same until 2023. Under this premise, if the extrapolated value in 2028 falls within the min/max range, this value is used.
A few benchmarks have their own rules. For example, the hot metal benchmark uses the minimum reduction rate from the 2021 update (0.2% per year rather than 0.3%). Similarly, the synthesis gas and aromatics benchmarks are reduced at the same rate as the refinery benchmark rather than following the standard methodology.
Our estimated EU ETS benchmark values can be found here, under the methodology section.
For some sectors, the change to benchmarks will have little impact. For example, in the cement/lime sector only the lime and dolime benchmarks are set to the maximum reduction level, whereas clinker (white and grey) benchmarks, representing the bulk of emissions and allocation in the sector, are adjusted downwards at or around the minimum reduction rate.
Conversely, pulp and paper as a sector will see all benchmarks cut to the maximum 50% of the initial values. Similarly, heat and fuel benchmarks will be reduced to 50% of those used during the first allocation period (2012-2020).
See Figure 3 for an overview of changes to benchmark levels by sector.
The impact of updated benchmarks on free allocation distribution as a whole will be dependent upon the degree to which industry has improved emissions efficiency during the last allocation period. Higher investments in efficiency improvements across the sector will result in higher free allocation and vice versa.
Assuming a flat improvement rate of 10% between the data collection periods, our modelling finds that the EU ETS benchmark updates will result in a 6% reduction in free allocation compared to a scenario in which benchmarks were not updated.
Simultaneously during this period, the impact of the CBAM phase-out will also serve to reduce the amount of free allocation afforded to EU industry. During the first years of the coming period, the impact of benchmarks outweighs that of CBAM, however in 2029 and 2030, CBAM will reduce allocation by 10.3% and 22.3% compared to a no-CBAM scenario, outweighing the impact of benchmark reductions.
Figure 4 below shows the relative impact of the benchmark update and the CBAM phase-out on free allocation volumes relative to their respective non-application.
The CBAM benchmarks are intended to provide the same relative protection from a carbon price on imports as free allocation does for industrial production within the EU. As such, they must be as close to the EU ETS benchmarks as possible. This presents a difficult conundrum for policymakers, as the EU CBAM products are determined at an HS code level (Harmonised System, used for classification of trade products), whereas the ETS benchmarks are much broader and not strictly tied to product codes.
The mismatch between ETS benchmark values and the classification of CBAM products requires some broad estimates to be made.
In order to translate the EU ETS benchmarks to CBAM benchmarks, we first established the percent difference between the estimated ETS benchmark values and the average EU emissions intensity reported in the 2021 ETS benchmark update. This difference is then applied to the Joint Research Council’s estimates for the EU’s per-product average emissions intensity to arrive at a per-product CBAM benchmark estimate.
Our estimated CBAM benchmark values can be found here, under the methodology section.
A weekly cap of what moved EUA prices and a clear view of the week ahead. We set out the drivers, their directional impact, and what matters next.
Specialising in data, analysis, and insights for all significant low-carbon markets and renewable energy.
