Since 2013, EU Member States have generated approximately €176 billion in ETS auction revenues, rising from €5 billion in 2017 to nearly €30 billion at the 2022 peak. Over the same period, the explicit allocation to industry decarbonisation from those revenues amounted to around €1.7 billion, less than 1% of the total collected. The contrast with what the power sector, buildings, and transport received is stark, and it sits at the centre of a detailed policy brief published by Agora Energiewende in July 2026 that frames the upcoming EU ETS review as the most consequential carbon market reform since the system’s introduction in 2005.
The inversion is not accidental. ETS revenues have historically flowed toward sectors where mature, low-cost abatement options exist: energy supply, grids and storage received €18.7 billion; public transport and mobility €16.9 billion; energy efficiency in buildings €13.9 billion. Industry, where the decarbonisation options require first-of-a-kind capital investment, fundamental reconfiguration of production processes, and technologies that carry both technical and commercial risk, has been systematically underserved by the revenue allocation. The consequence is visible in the investment data: almost half of all announced industrial decarbonisation projects remain uncompleted, and fewer than one in three have entered construction or started operating.
The Cap Recalibration Problem
The ETS cap is currently calibrated to the 2030 target. It declines annually by 4.3% of base year emissions, increasing to 4.4% from 2028, or just below 90 million tonnes per year. If left unchanged, allowance issuance would approach zero after 2039 with only small aviation volumes remaining. This trajectory was derived from the EU’s 2030 impact assessment, which predated the revised EU climate law that now enshrines a 2040 target of 90% emissions reduction below 1990 levels, of which at least 85% must be achieved domestically.
The 2026 review must therefore recalibrate the cap to a target that was not yet defined when the current linear reduction factor was set. Agora Energiewende’s modelling identifies the LRF transition point as the central design question. Lowering the LRF before 2030 during the current trading period would substantially interfere with an established forward market where EU allowances trade several years ahead, damaging policy credibility at a moment when many companies have started committing capital to decarbonisation. Lowering it after 2036 risks letting the cap approach zero before the reduction effect takes hold. The technically and politically coherent window is 2031 or 2036, corresponding to the start of new allocation periods.
The preferred pathway in the Agora analysis is to maintain the current LRF of 4.4% until 2035, then lower it to 2.0% from 2036 onwards, resulting in ETS emissions 90% below 2005 base year levels in 2040 and net zero in 2045. This would place the nominal ETS cap at 200 million tonnes in 2040. However, the effective supply of allowances available to the market is substantially higher than the cap alone implies: the Total Number of Allowances in Circulation currently stands at just above 1 Gt; the Market Stability Reserve holds significant additional volumes; the New Entrants Reserve carries a substantial buffer; and carbon dioxide removals are projected to contribute between 40 and 121 million tonnes by 2050 in the modelled scenarios. Factoring in these parallel sources, the effective LRF applying to actual market supply would average 3.6% in 2031 to 2035, 1.3% in 2036 to 2040, and 2.4% in 2041 to 2045, substantially below the nominal rate in each period.
The arithmetic creates a hidden loosening risk. If the LRF is reduced early, in combination with MSR threshold changes, NER releases, and additional allowance sources entering the market, the cumulative effect could produce a structural surplus in the 2030s. A lower LRF from 2031, combined with NER releases, could see the TNAC increase in absolute terms throughout much of the decade, applying 2025-level surplus volumes to a carbon market at most a third of its current size. That scenario would either push emissions above the projected trajectory or depress the carbon price for years, undermining the investment cases of exactly the industrial frontrunners who have already committed billions.
The MSR’s Expanding Role and Its Limits
The Market Stability Reserve, introduced in 2018 to address the large oversupply that had accumulated during the 2010s and depressed carbon prices for nearly a decade, currently withholds 24% of amounts foreseen for auction when the TNAC exceeds the upper threshold. Its invalidation mechanism permanently cancels allowances when the reserve exceeds 400 million units, a rule that has been essential in removing structural oversupply from the market. In 2026, the Commission proposed suspending this invalidation mechanism. If implemented, the volume of allowances in the reserve could grow above 900 million, more than double its current size, with the effects becoming visible during the late 2030s when those allowances would re-enter the market.
The MSR faces three technical updates that are individually reasonable but collectively risky. Linking its thresholds to the declining cap trajectory would prevent the reserve from releasing allowances when the market is not experiencing genuine scarcity. Replacing the fixed annual release of 100 million allowances with a dynamic function would spread releases over a longer period and better match actual scarcity conditions. Redefining the TNAC to incorporate historical aviation net demand before 2024, and the 100 million allowances Member States can use under the Effort Sharing Regulation, would reduce the measured surplus and better reflect real market availability. Each adjustment has merit on its own terms. The risk is that implementing all three simultaneously, alongside a relaxed LRF and NER releases, could overload the system and produce interactions that no single adjustment was designed to cause.
The Agora brief argues that the MSR’s rule-based, quantity-driven design is precisely what shields it from political interference and ensures predictability for market participants. Converting it to a price-based trigger system, as some stakeholders have proposed, would risk reclassifying it as a fiscal measure under EU law, which would require unanimity rather than qualified majority voting for future changes, significantly raising the political barrier to necessary adjustments. Price volatility is better addressed through the dedicated Article 29(a) price containment mechanism in the ETS Directive, which draws on MSR allowances but operates as a distinct instrument activated by different parameters.
The Free Allocation Failure and the Conditionality Proposal
The numbers behind free allocation tell a story the ETS was not designed to tell. Between 2013 and 2025, industrial installations received almost 7 billion allowances free of charge, a commercial value exceeding €200 billion. In 2025 alone, they received 416 million allowances worth more than €30 billion at market prices. The majority of this allocation compensated industry for competitive disadvantage against non-ETS competitors abroad. Evidence suggests that some sectors received more than they needed and generated windfall profits from the surplus. And while some companies invested in climate-neutral technologies using the economic headroom the allocation provided, most continued operating existing production assets without transformative capital commitment.
Until 2018, industrial emissions in ETS-covered sectors were slightly higher than in 2005. Since 2019, they have fallen by more than a quarter, but the Agora analysis is explicit that much of this reduction reflects lower output and capacity closures rather than investments in low-emission production processes. The ETS was collecting revenues and distributing free allocation for two decades while industrial emissions remained essentially flat or fell only because demand contracted. The reform now under discussion must address why €200 billion in free allowances did not produce the investment it was theoretically intended to support.
The proposed solution is conditionality: making remaining free allocation available only to installations that commit to and verifiably undertake qualifying decarbonisation investments. From 2026, the ETS already requires the worst-performing 20% of installations within each benchmark category to submit a Climate Neutrality Plan or lose 20% of their free allocation, and imposes similar obligations on installations subject to energy audits. This is characterised in the policy brief as a step in the right direction that nevertheless falls short: the current conditionalities encourage incremental efficiency improvements rather than the transformative process changes required for climate neutrality.
What the Agora proposal adds is a threshold calibrated to net-zero trajectories by sector. For iron and steel, the qualifying threshold would start at natural gas-based direct reduced iron, which reduces emissions by approximately 47%, increasing as the technology rolls out with fossil-free hydrogen. For cement, CCS applied to clinker kiln process emissions achieves a 90% reduction and should constitute the qualification threshold. For process heat that can be electrified, full electrification should be the standard. Installations qualifying under these thresholds would receive free allowances at benchmark levels in excess of their residual emissions exposure, creating a deliberate overallocation that generates surplus allowances they can sell, strengthening the investment incentive through an additional revenue stream. Installations that do not qualify would absorb a larger share of the cross-sectoral correction factor, creating a deduction mechanism that directly links the level of support for decarbonisation frontrunners to the investment pace across the sector.
The Revenue Mismatch and the Industrial Decarbonisation Bank
Conditional free allocation addresses the incentive problem but not the financing problem. The Commission’s own impact assessment for the 2040 target projects that EU industry will require €327 billion over 2031 to 2040, with chemicals alone accounting for €150 billion, iron and steel €80 billion, and non-metallic minerals including cement and glass €48 billion. Against these requirements, the current deployment of ETS revenues toward industry decarbonisation at €1.7 billion total since 2013 is inadequate by any measure.
The proposed Industrial Decarbonisation Bank and Investment Booster are intended to address this gap by channelling ETS revenues toward industrial capital expenditure at EU level rather than dispersing them through Member State spending that has historically favoured energy supply and transport. Between 2021 and 2024 alone, Member States spent more than €14 billion on Indirect Cost Compensation to energy-intensive industries to offset higher electricity prices attributable to ETS carbon costs. ICC serves a legitimate function in reducing electricity price disadvantages for industry, but the current methodology uses standardised emission factors that frequently exceed the actual carbon intensity of consumed electricity, creating overcompensation in sectors already receiving free allocation.
A more efficient architecture would frontload ETS revenues rather than relying on annual auction proceeds as they accrue. The model already being developed for ETS2, covering buildings and transport, envisages the European Investment Bank leveraging future ETS revenues to make upfront funding available. Applying the same principle to industrial decarbonisation through the Industrial Decarbonisation Bank would allow capital to be deployed at the scale and speed that €327 billion in required investment over a decade demands, without the year-by-year revenue constraint that annual auction proceeds impose. The risk the Agora brief identifies in this approach is that bringing large volumes of allowances to the carbon market to capitalise such a bank would depress the carbon price, undermining the investment cases the bank is meant to support. Frontloading revenues from future regular auctions rather than releasing accumulated reserves is the design principle that threads between these competing requirements, and it is the tension that the Commission’s forthcoming reform proposal will need to resolve.
