ETSs are gaining momentum, and design choices vary regarding the nature of the cap, the method of free allowance allocation (section 4.1) and the compliance possibilities (including the use of carbon credits) (section 4.2). Some design choices provide more flexibility to firms and can help ease competitiveness and affordability issues. For instance, output-based free allowance allocation methods provide flexibility on production levels; the use of carbon credits for compliance provides sectoral and geographical flexibility, while the possibility to borrow or bank permits provides temporal flexibility. Different designs could reflect national circumstances and priorities but may require coordination to ensure interoperability for linking (Verde et al., 2022[1]; Galdi et al., 2022[2]) or for recognition of carbon pricing policies or prices.1
Effective Carbon Rates 2025
ETSs may be distinguished according to whether they set a pre-determined cap on covered emissions (as in cap-and-trade systems) or not (as in intensity-based systems). When the cap is pre-determined2 the total quantity of allowable emissions for each compliance period is fixed.3 This is the case of cap-and-trade systems, in which case the cap is set as an overall emission limit at the system level – e.g. the EU ETS. It can also be the case of baseline-and-credit systems with pre-determined baselines, in which case the cap may be calculated as the sum of installation-level emissions limits across covered facilities – e.g. the Tokyo Cap-and-Trade system. Intensity-based systems4 do not set a limit on emissions; rather, covered emissions are allowed to vary with production and the limit implicitly applies to emission intensities. Intensity-based systems are baseline-and-credit systems where the main allocation of allowances5 method depends on output-based benchmarking. This implies the reliance on emission intensity factors (a benchmark which may be country-, sector- or emitter-specific) and entities’ current year’s production (Fischer, Qu and Goulder, 2024[3]): hence, to reduce their average carbon costs, covered entities need not adjust their production as long as the emission intensity of their production is below that set by the benchmark. By, in effect, easing constraints on production (Fischer, 2001[4]), this design can help support the competitiveness of industry. However, intensity-based systems generally do not provide certainty on the total level of emissions covered by the system.
The emissions trading systems introduced in recent years have shifted away from having a pre-determined cap and are now increasingly intensity-based (Figure 4.1). Starting in 2005 with the introduction of the EU ETS, there has been a steady increase in the number of ETSs. Up until 2018, the majority of new ETSs have a pre-determined cap. The year 2019 marks the introduction of several Canadian province or territory-level ETSs, all intensity-based. The majority of ETSs introduced since then have been intensity-based – these include the Chinese national ETS, the Australian Safeguard Mechanism and the Indonesian Economic Value of Carbon Trading Scheme (Figure 4.1). In 2023, the majority (70%) of GHG emissions covered by an ETS are covered by intensity-based ETSs (Figure 4.1) – an effect mainly driven by the Chinese national ETS being intensity-based.
Number of ETSs with pre-determined caps and intensity-based ETSs over the years
Note: In 2023, systems with pre-determined caps include: most Chinese Pilot ETSs (Fujian, Guangdong, Hubei, Shanghai, Shenzhen, Tianjin), the EU ETS, the German national ETS, the Japanese subnational ETSs (Saitama Target Setting ETS and Tokyo Cap-and-Trade System), the Kazakhstan ETS, the Korean Emissions Trading System, the Mexico national ETS, the New Zealand ETS, the Québec Cap-and-Trade system, the Swiss ETS, the UK ETS, all US subnational ETSs (California Cap-and-Trade, RGGI, Massachusetts Limits on Emissions from Electricity Generators, Washington Cap-and-Invest). The Austrian national ETS is also classified in this category: while it currently operates as a hybrid system, it is in the process of transitioning to the EU ETS2, at which point it would be classified as a cap-and-trade system (i.e. an ETS with a pre-determined cap).
Intensity-based systems include: the Australia Safeguard Mechanism, the Canada Alberta Technology Innovation and Emissions Reduction (TIER) Regulation, the Canada Federal Output-Based Pricing System (FOBPS), the Canada New Brunswick Output-Based Pricing System, the Canada Newfoundland and Labrador Performance Standards System (PSS), the Canada Nova Scotia Output-Based Pricing System for Industry, the Canada Ontario Emissions Performance Standards (EPS), the Canada Saskatchewan Output-Based Performance Standards, the China National Emissions Trading System, the Beijing and Chongqing Pilot ETSs and Indonesia’s Economic Value of Carbon (Nilai Ekonomi Karbon) Trading Scheme.
Source: International Carbon Action Partnership Status Reports (ICAP, n.d.[5]), complemented by authors’ desk research.
Systems may transition from one regime to another. For example, the Chinese national ETS is expected to include non-binding control targets on total covered emissions and eventually transition to a cap-and-trade system by 2030 (Carbon Pulse, 2024[6]; General Office of the State Council, 2024[7]; Carbon Pulse, 2025[8]). The Chongqing Pilot ETS, on the other hand, transitioned to an intensity-based system in 2021 (from a cap-and-trade system from 2014 to 2020).
Emission allowances may be freely allocated following different methods, with grandparenting or benchmarking being the most common. In many cases, ETSs may use a mix of methods. The first approach relies on historical emission levels, while the second depends on production levels and emission intensity factors. With cap-and-trade systems, grandparenting tends to be found more frequently in earlier phases of ETSs, with a move to benchmarking as the system evolves (Kuneman et al., 2022[9]). Moreover, some ETSs use both – for instance, the California Cap-and-Trade Program uses output-based benchmarking for industrial facilities, while it uses grandparenting for natural gas suppliers. Annex B provides more details on free allowance allocation methods across ETSs.
Emissions allowances may be freely allocated using grandparenting or benchmarking – with benchmarking being based on either past production or production in the current year (output-based benchmarking). There are a few cases where free allowances are allocated ex-post and are set equal to verified emissions (see examples in Annex B – this can be interpreted as these emissions facing reporting obligations but no carbon price.
Grandparenting and benchmarking approaches
The formulas for free allowance allocation under the grandparenting and benchmarking methods are of the following form:
Grandfathering: past emissions x adjustment factors
Benchmarking: production x emissions intensity [benchmark] x adjustment factors
Historically, free allowances based on benchmarking have been calculated using production in previous years, but output-based benchmarking now uses current year production data.
Adjustment factors
Adjustment factors may be included in the above formulas for various purposes:
Factors related to the risk of carbon leakage: this risk is generally assessed using the emission-intensity and trade-exposure (EITE) of sectors. The factor may be based on a binary assessment with all activities above a threshold of leakage risk having a factor of a 100% (e.g. the EU ETS in Phase 3) or based on a tiered assessment, with the application of what is commonly referred to as an assistance factor for different levels of emissions intensity and trade exposure (e.g. New Zealand, Québec).
Factors related to the decrease in the cap: such factors are used to bring allowance allocation in line with the general cap decline trajectory (e.g. in California).
Factors related to the decrease in the benchmark: such factors are used to bring allowance allocation in line with the general emissions intensity decline trajectory (e.g. in many Canadian systems).
Adjustment factors may also be used to stick with a pre-determined cap when multiple allocation methods are used. This is the case in California for instance, where allocations for the industry are output-based and are adjusted if the total sum of freely allocated allowances exceeds the pre-determined cap.
Incentives provided by different approaches
Free allowances can affect economic rents and thus influence investment decisions, but the different methods also have different impacts on the channels through which marginal abatement incentives operate. Grandparenting and benchmarking based on past production levels provide marginal abatement incentives both to reduce the emission intensity of production as well as production (the product of both being equal to covered entities’ emissions). Output-based benchmarking, however, only provides an incentive to reduce the emission intensity of production.
In terms of investment, regarding grandparenting, setting the base year not too far back in time enables emissions estimations used to calculate allocation amounts to be more in line with current technologies and abatement opportunities. At the same time, setting the base year or period sufficiently back in time can avoid providing incentives to firms to increase emissions before the implementation of the ETS so as to increase the allocation they receive (OECD, 2023[10]).
The benchmarks used in benchmarking methods can impact investment incentives (Flues and van Dender, 2017[11]; Kuneman et al., 2022[9]). In particular, the more granular benchmarks are, the narrower the range of abatement options they promote: tying them to specific fuels, processes, or technologies can distort incentives to adopt the most cost-efficient means of achieving emissions reductions.
Finally, there are also cases where free allowances may not be traded. If they can then be used for compliance, this can be interpreted as the marginal price signal not being maintained, since there is no incentive to reduce emissions below the allowance allocation (since they cannot be sold). If they should be used for other purposes than compliance (e.g. redistribution) then this can be interpretated as both the marginal and average price signals being equal (the free allowance allocation does not directly reduce the ETS-related costs for firms).
Source: Authors based on ICAP (2025[12]), ETS legislations and announcements, and Kuneman et al. (2022[9]).
Intensity-based systems may also require total covered emissions to decline and may not entirely rely on output-based benchmarking. This is the case for instance of the Australia Safeguard Mechanism6 (Australian Government DCCEEW, n.d.[13]). Intensity-based systems also do not necessarily entirely rely on output-based benchmarking: for instance, the Beijing and Chongqing Pilot ETSs, which are intensity-based, partly distribute allowances according to grandparenting and benchmarking based on past years’ production values.
There is a rising practice of accounting for current production levels in free allowance allocation methods, even in cap-and-trade systems. Output-based benchmarking is not exclusively used in intensity-based systems as it also appears in some cap-and-trade systems. This was historically the case in the California cap-and-trade system for covered entities in the industry sector (see example in Box 4.1). More recently the Kazakhstan ETS has started relying on this method as well (operating a shift from free allowance allocation based on grandparenting). Free allowance allocation in the EU ETS now also accounts to some extent for current production levels: revised rules applying from Phase 4 onwards include adjustments to free allocation when an installation makes a significant change to its production (at least a 15% increase or decrease in production) (European Commission, n.d.[14]). Such a provision has also been included in the UK ETS.
Auctioning or fixed price selling of allowances may complement the allocation of free permits, both in systems with pre-determined caps and intensity-based systems. Cap-and-trade systems generally sell allowances at auctions, while intensity-based systems, when they sell allowances, do so at a fixed price in many cases.7 As the adjustment factors used in the allocation of free allowances across systems decline (Box 4.1), entities covered by ETSs may need to increasingly rely on the purchase of permits for compliance. Section 4.2 delves deeper into the use of auctioning or fixed price funds as one of the compliance options across systems.
4.2.1. Different compliance options
Entities covered by an ETS have a variety of compliance options to cover their verified emissions, which can help provide flexibility.8 These include the use of permits received for free, of permits purchased from other entities (trading) and of permits purchased from the government, the use of banking and borrowing and the use of carbon credits (Figure 4.2).
Note: Some systems may allow for auctioning to take place in the legislation, but in effect no auction may take place in a given year. This is the case, for instance, of some Chinese Pilot ETSs, where auctioning takes place on an ad-hoc basis.
Source: ICAP (2025[12]) complemented by authors’ own desk research.
The possibility of trading is what defines an ETS. For instance, the Australian Safeguard Mechanism, which had been in place since 2016, introduced tradeable permits in its system in July 2023, which has thus classified it as an ETS. Allowing trade, however, does not necessarily entail that it takes place on a large scale. In many systems, trade is limited.
Free permits are available in almost all ETSs. Only four ETSs do not allocate permits for free (see Annex B for more detail) – these concern ETSs covering the power sector or ETSs applying exclusively upstream (mainly to the road and heating sectors).9 Permits may be distributed according to different rules (section 4.1), and while they may not impact marginal abatement incentives, they can impact rents and investment incentives (OECD, 2023[10]; Flues and van Dender, 2017[11]).
Permits may be purchased from government auctions or from fixed price funds in a majority of ETSs. While there are provisions for auctions in 28 systems, these are not systematically offered – in 2023, auctions were offered in only 22 systems. Alternatively, the government can also sell permits at a fixed price (instead of prices being determined by the outcome of auctions). For instance, the purchase of permits at a fixed price from government funds is an option for all Canadian intensity-based ETSs.
Banking and borrowing enable temporal flexibility within ETSs – and while banking is typically allowed in ETSs, borrowing is seldom an option (ICAP, 2023[15]). Banking refers to an entity using permits from previous compliance periods (i.e. banked permits), while borrowing refers to borrowing permits that the entity expects to receive for free in future periods. Banking unused permits from one compliance period for use in future periods can be used to meet own compliance obligations or to sell to other market participants. By allowing entities to carry forward unused allowances, banking can incentivise early emissions reductions. Banking is allowed in most ETSs, albeit with limits in time and quantity. Borrowing could help facilitate investment choices by providing flexibility in timing, but could also be seen as delaying the emission reductions needed to achieve the ETS’s objectives. Most ETSs do not allow borrowing, and when they do, they only allow it to a limited extent.
Offsetting through the use of carbon credits10 offers entities covered by ETSs the possibility to cover their compliance obligations by purchasing credits generated by emission reduction and GHG removal projects undertaken outside the scope of the ETS – and present a widespread compliance option among ETSs (La Hoz Theuer et al., 2023[16]). Carbon credit use is allowed in more than 60% of ETSs, generally with restrictions on quantity (what share of compliance obligations can be met through carbon credits) as well as on quality (which criteria these credits should fulfil). This is further discussed in section 4.2.2. Reduction and removal credits generated within an ETS’s scope are not considered carbon credits in this report (in line with the definition provided in e.g. La Hoz Theuer et al. (2023[16])). For instance, the New Zealand ETS includes forestry and some other removal activities11 within the scope of the ETS, so that entities generating emissions can trade with entities removing emissions within the scope of the ETS. As discussed in section 3.1.2, the interaction of GHG removal credits with ETSs is increasingly being considered by jurisdictions – and different options are available for that (see extensive discussion in La Hoz Theuer et al. (2021[17])).
4.2.2. Carbon credit use in ETSs12
Carbon credits can be generated from emissions reduction and GHG removal projects (Figure 4.3, Allen et al. (2024[18]), La Hoz Theuer et al. (2023[16])). GHG emission reduction credits are generated by activities that reduce the amount of GHG emissions that enter the atmosphere, compared to a baseline scenario of how large emissions would have been in the absence of the credit-generating activity. This includes the deployment of renewable energy (e.g. solar, wind), programmes to deploy energy-efficient cookstoves, the capture and utilisation of methane from landfills and the destruction of ozone depleting substances with very high global warming potential.
GHG removal relates to taking GHGs from the atmosphere. This includes nature-based solutions (e.g. sequestering carbon through afforestation or reforestation) and technology-based solutions (e.g. bioenergy with carbon capture and storage (BECCS), direct air capture with geological storage (DACCS), converting atmospheric carbon back into rock through remineralisation). These latter negative emissions technologies (NET) have an important role to play in reaching net zero emissions, both by removing residual emissions (e.g. emissions that may be too difficult, too costly, or impossible to abate in the time required) and in scenarios with an overshoot in emissions – i.e. scenarios where the GHG emissions consistent with the 1.5°C or 2°C goals of the Paris Agreement are exceeded (La Hoz Theuer et al., 2021[17]; Intergovernmental Panel on Climate Change, 2021[19]). This may help explain their increasing consideration for inclusion in carbon pricing schemes, either through carbon credit use or through the inclusion of removal activities within the scope of the schemes.
The use of carbon credits for compliance diversifies the sources of ETS compliance. Diversifying the sources of compliance can be especially important in jurisdictions or sectors where abatement opportunities are narrowing, with implications for compliance costs and ETS functioning, due to declining liquidity. Moreover, allowing for the purchase of credits from emission reduction or GHG removal projects can provide incentives for their development. This can help also stimulate mitigation in sectors that may be harder to price – e.g. Agriculture, forestry and other land use, AFOLU, especially in countries where a large share of domestic emissions come from those sectors.
Note: CCS: Carbon Capture and Storage
DACCS: Direct Air Capture with Geological Storage
Source: The Oxford Principles for Net Zero Aligned Carbon Offsetting (revised 2024) (Allen et al., 2024[18]).
To help ensure effective emission abatement incentives and environmental integrity, carbon credit use for compliance in ETSs is typically subject to both quantitative and qualitative limits. Since the use of carbon credits reduces the requirement on regulated entities to reduce their own emissions, careful attention should be paid by regulators to the (i) quantity of credits that can be used and (ii) qualitative criteria for eligible carbon credits. In particular, quantitative limits can be placed to ensure regulated entities’ incentives to reduce on-site emissions are maintained.
Qualitative criteria may be introduced to ensure the environmental integrity of carbon credits.13 Environmental integrity encompasses several elements, such as additionality, permanence and quantification of impacts (La Hoz Theuer et al., 2023[16]; Wetterberg, Ellis and Schneider, 2024[20]). Additionality refers to the requirement that mitigation activities should only generate credits if they would not have occurred in the absence of the added incentive created by such credits.14 Permanence refers to ensuring that emission reductions or GHG removals of the project will not be reversed (e.g. in the case of forest-related projects) or that the project is accompanied with a way to mitigate and compensate for potential reversals. Finally, emission reductions should be conservatively quantified – i.e. more likely to be underestimated than overestimated.
In most ETSs, carbon credits used for compliance should fulfil qualitative criteria, related to the projects’ location. Most ETSs allow for “domestic carbon credits” – referring to credits generated from projects within the geographical boundaries of the country in which the ETS operates (or that of a linked ETS) – while the use of “international carbon credits” originating from projects outside of the ETS is currently only allowed in the Korean ETS15 (ICAP, 2025[12]). In the case of Korea, a restricted set of international carbon credits may be used for compliance (with criteria related inter alia to the type of projects and their ownership). Criteria related to the location of projects may go beyond the domestic versus international dichotomy: for instance, in the Chongqing Pilot ETS, at least 80% of the credits used must be generated by projects within Chongqing city. The Alberta TIER only allows Alberta-based emissions carbon credits.
Many ETSs place other qualitative restrictions, including on the nature of the projects generating credits or the types of credits allowed for. For example, for the Québec Cap-and-Trade System, a new regulatory framework allows four carbon credit-generating activities: reclamation and destruction of methane from landfill sites, destruction of certain halocarbons contained in insulating foam from refrigeration, freezer or air-conditioning equipment,16 carbon sequestration through afforestation or reforestation on private lands and anaerobic digestion of manure. Finally, many ETSs only allow the use of credits from specific crediting mechanisms – e.g. Australian Carbon Credit Units (ACCUs) for the Australian Safeguard Mechanism, credits from the Chinese Certified Emissions Reduction scheme (CCER) for the Chinese national ETS – these are generally governmental crediting mechanisms (Wetterberg, Lanzi and Gómez, 2025[21]).17
Almost all ETSs that allow for the use of carbon credits for compliance place a limit on the quantity which can be used. Across ETSs, when allowed, carbon credits can be used in a range between 3.3% and 100% of compliance obligations18 (Figure 4.4), though compliance obligations may be defined differently for baseline-and-credit systems and for cap-and-trade systems. Even in systems where no limit is placed on the quantity of carbon credits that may be used for compliance, their use beyond a limit may need to be justified. For instance, in the Australian Safeguard Mechanism, entities surrendering ACCUs equivalent to 30% or more of their baselines are required to provide a statement explaining why they have not undertaken more on-site abatement activities (Australian Government Clean Energy Regulator, 2025[22]). In some ETSs, a limit on the share of carbon credits used for compliance by entities is complemented by a cap on the total quantity of carbon credits which can be used at the system-level. This is for instance the case of the Guangdong Pilot ETS, where 10% of covered entities’ annual emissions can be covered by carbon credits and where a limit is also set on the total amount of carbon credits which can be used for compliance in a year: in 2021 and 2022, this amounted to one million carbon credits (ICAP, 2025[12]).
In some ETSs, the quantitative limit is linked with the qualitative criteria carbon credits should fulfil. This may relate to the type of credits used. For instance, in the Fujian Pilot ETS, the use of both domestic (Chinese) project-based carbon credits (CCERs) and Fujian Forestry Certified Emission Reduction credits (FFCERs) is allowed. 5% of the annual compliance obligation may be met through CCERs, while this limit is increased to 10% for entities that use both CCERs and FFCER. This may also relate to the type of project which generated the credit or the location of the project. Both the Saitama Prefecture Target Setting ETS and the Tokyo Cap-and-Trade Program place limits on carbon credits not generated within the respective prefectures: for instance, for Saitama, outside Saitama credits can be used for compliance for up to one-third of offices’ reduction obligations and to 50% for factories. For instance, the Washington Cap-and-Invest Program places a limit of 5% of an entity’s compliance obligations for projects not located on federally recognised tribal land and an additional 3% from projects located on federally recognised tribal land.
2023
Note: This figure aims at highlighting that most systems place quantitative limits on the use of carbon credits, but the shares displayed may not be comparable since the limits relate to “compliance obligations” defined in different ways. This may either relate to an entity’s total verified emissions (e.g. the Québec Cap-and-Trade system) or to an entity’s difference between verified emissions and free allowance allocation (e.g. the Canada FOBPS). In some restricted cases, these are shares of an entity’s free allowance volume (e.g. the Hubei Pilot ETS).
(1) In Nova Scotia, regulations contain provisions for the potential use of carbon credits, but carbon credits are currently not enabled in the system.
(2) For the Saitama Prefecture Target Setting ETS (resp. Tokyo Cap-and-Trade Program), quantitative limits apply only for outside Saitama (resp. Tokyo) credits. These credits can be used for compliance for up to one-third of offices’ compliance obligations (and in Saitama factories can use up to 50%).
(3) Note that while carbon credits generated from activities undertaken outside the scope of the ETS are not allowed in the New Zealand ETS (NZ ETS), New Zealand Units (NZUs) generated from NZ ETS removal activities can be used for compliance. No limit applies on their use and they can be banked indefinitely.
(4) For the Fujian Pilot ETS, the 5% limit applies if using only CCERs; the limit is increased to 10% for companies that use both FFCER and CCER carbon credits.
(5) For the Washington Cap-and-Invest Program, the 5% limit applies to projects not located on federally recognised tribal land and an additional 3% can be used for projects located on federally recognised tribal land.
* Over the course of 2023, the federal OBPS was operational in Manitoba, Nunavut, Prince Edward Island and Yukon.
** RGGI (Regional Greenhouse Gas Initiative): in 2023, it operates with ten the states of Connecticut, Delaware, Maine, Maryland, Massachusetts, New Hampshire, New Jersey, New York, Rhode Island, Vermont and Virginia.
Source: ICAP (2025[12]) complemented by authors’ own desk research.
In many ETSs, the quantitative limit on the use of carbon credits is, in effect, not reached. This can relate to many factors, including free allowance overallocation (Dechezleprêtre, Nachtigall and Venmans, 2018[23]) implying a reduced need for further compliance options, low supply of eligible credits or carbon credit prices being higher than primary and secondary ETS market prices (e.g. in Kazakhstan where secondary market prices are currently lower than EUR 1/tCO2). While data is not always available, where it is, it displays a lower use of carbon credits than what is allowed for (Table 4.1).
Note: RGGI: Regional Greenhouse Gas Initiative; KCU: Korean Credit Unit.
Source: Canada FOBPS: Table 10 in Government of Canada (2022[24]); Alberta TIER: Alberta Government (2024[25]); Québec Cap-and-Trade: Québec Ministère de l’Environnement, de la Lutte contre les changements climatiques, de la Faune et des Parcs (2025[26]); Tokyo Cap-and-Trade: Mizuho Research & Technologies, Ltd. (2024[27]); Kazakhstan ETS: ICAP (2025[12]); Korea ETS: ICAP (2024[28]; 2025[12]), KRX (2025[29]); California Cap-and-Trade: California Air Resources Board (California Air Resources Board, 2025[30]); RGGI: ICAP (2025[12]); Washington Cap-and-Invest: Department of Ecology - State of Washington (2024[31]); Australia Safeguard Mechanism: Australian Government Clean Energy Regulator (2025[32]).
Carbon credit prices typically depend on the type of projects they come from. While price data on ETS-eligible carbon credits is limited, prices in other market segments suggest that carbon credit prices differ by project type. Exchange-traded prices differ by project type; they are higher for removal projects as compared to reduction projects – and within reduction projects they are highest for nature-based projects. In April 2025, reduction projects traded between USD 1/tCO2e for renewable energy projects and USD 5.3/tCO2e for nature-based projects (World Bank, 2025[33]). Estimated over-the-counter prices of carbon credits related to carbon dioxide removal (CDR) projects are higher than those of reduction projects: for technology-based removals they have averaged at about USD 180/tCO2 over the 2022-2025 period and for nature-based removals, prices are on the rise, with an increase from USD 17 to USD 35/tCO2 from the end of 2024 to mid-2025 (AlliedOffsets, 2025[34]).19 Moreover, the World Bank (2025[33]) finds a price premium for credits eligible to be used for NDC achievement (Article 6.2) and international compliance markets (e.g. CORSIA) relative to voluntary markets.
The price of carbon credits used for compliance in ETSs is not readily observable. Since credits may be bought and sold directly between entities, in many cases price estimates are not available. Survey evidence can provide such data, but due to confidentiality constraints, detailed information by project and mechanism is lacking. Hence, the average price of credits relating to GHG removal and reduction projects, as outlined in the previous paragraph, are not necessarily representative of the prices of credits allowed for compliance in ETSs, since as seen above carbon credits generally come with qualitative restrictions. Some initiatives seek to provide project-specific estimates for these prices through modelling, though these price models are uncertain and have limited coverage of many compliance-eligible mechanisms, and may be less reliable than survey evidence (Wetterberg, Lanzi and Gómez, 2025[21]).
When ETS compliance related carbon credit price data is available, it does not necessarily show lower prices than primary and secondary market ETS permit prices (Table 4.1). For Tokyo, the trading prices of renewable energy credits averaged at JPY 5 625/tCO2 in 2023 – much higher than excess emission reductions credits trading between JPY 650 and 700/tCO2 (secondary market prices) in the same year. In Korea, in July 2025, Korean Credit Units (KCU) prices for 2024 and 2025 are of KRW 9000/tCO2, similar to recent auction clearing prices (e.g. KRW 9070/tCO2 in July 2025). In Québec, the weighted average price of carbon credit transactions in 2023 was of CAD 28.19/tCO2, 37% lower than the primary market price in 44.65. In 2025, similar differences are found in California, between the prices of California Carbon Allowances and California Carbon Offsets (Carbon Pulse, 2025[35]).
4.2.3. Evolutions and implications of the variety compliance options on emissions covered by ETSs and covered entities’ compliance costs
Free allowance shares are decreasing in many systems. Relatedly, auctioning is taking on a greater role. For instance, for the Chinese national ETS, Interim Regulations state that auctioning is to be introduced and gradually expanded. With the introduction of the many intensity-based Canadian ETSs, the sale of allowances at a fixed price as opposed as through auctions has also taken on a greater role in ETSs. The German and Austrian ETSs have also been selling allowances at a fixed price during their initial transition period. Auctioning should be introduced in the German ETS from 2026, with a price corridor. Indonesia is considering the introduction of a carbon tax through which entities may fulfil part of their compliance obligations (through a hybrid “cap-tax-and-trade” system) (ICAP, 2025[12]).
Carbon credit use for compliance has evolved and is set to continue evolving. The use of carbon credits for compliance has been an option since the onset of ETSs, with domestic carbon credits taking on a greater role than international carbon credits. While the use of international carbon credits used to be a compliance option in the EU ETS and the New Zealand ETS, these became ineligible in, respectively, 2021 and 2015. The Korean ETS, on the other hand, initially only allowed for domestic credits, and introduced the possibility to use international credits three years after its introduction (2018). The qualitative criteria and quantitative limits for carbon credit use are subject to evolutions as well. For instance, qualitative criteria have recently been updated in Québec, California and China. The share of compliance obligations that can be met with carbon credits is set to increase in some systems – e.g. from 4% per year for 2021 to 2025 to 6% for 2026 to 2030 in the California Cap-and-Trade Program; from 60% in 2023 to 90% in 2026 in the Alberta TIER20 Regulation.
By increasing temporal, spatial and sectoral flexibility, banking, borrowing and carbon credit use directly affect the total amount of emissions covered by ETSs. Borrowing and banking allow for the use of permits from a different period. Hence, these options along with the use of carbon credits affect the total amount of emissions which can occur within the ETS in a given year, potentially making them higher than the cap, even in cap-and-trade systems. Moreover, the surrender of carbon credits for compliance leads to a potential increase of emissions of covered entities (La Hoz Theuer et al., 2023[16]), since they can compensate by abatement taking place in other sectors, other locations or at any rate, outside of the ETS scope. This provides more flexibility to covered entities but also reduces the incentive to reduce emissions on-site.
The different compliance options affect compliance costs for covered entities (in other words, the average price paid per tonne of CO2e). The prices corresponding to permits borrowed or banked and to carbon credits can substantially differ from primary and secondary market ETS prices, adding to the already existing heterogeneity in prices within systems and across jurisdictions. Box 4.2 presents a conceptual example of an EACR profile that would account for all compliance options – since these options ultimately affect the average price faced by entities.
Borrowing, banking and carbon credit use can also impact primary and secondary market prices in an ETS by (i) increasing the supply of emissions permits within an ETS and by (ii) introducing potentially cheaper compliance options. The first channel refers to the increase in the supply of permits through a variety of compliance options beyond using free permits, trading permits and buying permits from government, which can then drive prices down. The second channel occurs when compliance options with different (and potentially lower) prices directly interact with primary and secondary market prices. This could especially be the case with carbon credits, particularly when their use has no or loose quantitative limits. For instance, evidence was found for the New Zealand ETS (NZ ETS), that when it was operating with an unconstrained international linkage, the decline of international carbon credit prices from 2011 was accompanied by a decline permit prices in the NZ ETS (Leining, 2022[36]).
The Effective Average Carbon Rate (EACR) accounts for how free allowances affect the costs faced by entities covered by an ETS. This indicator could be augmented by integrating the other compliance options offered to covered entities, which also affect the average carbon price:
Prices corresponding to banked permits will either be equal to 0 if these correspond to permits that had been received for free or will be equal to the price on the secondary or primary market in an earlier year (if they were purchased from the government or another trading entity). Over the long term, permit prices have tended to increase (ICAP, n.d.[37]), but this is not always the case (e.g. the Korean ETS) nor is it necessarily the case in the shorter term (e.g. average permit prices have declined in the California Cap-and-Trade Program between 2024 and 2025, and in the EU ETS between 2023 and 2024 – Figure 3.2).
Borrowing can raise the cost since it generally comes with an interest (e.g. 2 to 10% in the Australian Safeguard Mechanism (Australian Government CER, 2025[38])). Moreover, since it implicitly applies the forthcoming prices in the ETS, it can tend to increase prices through this way also, though not necessarily (e.g. in the past years prices have decreased in certain systems – see Figure 3.2).
Where carbon credits are used for compliance, the prices at which they are purchased tend to be lower than secondary and primary market ETS prices, though going forward this could also depend on the type of credits available in the market and allowed for compliance (section 4.2.24.2.2).
Integrating these compliance options into an augmented EACR indicator is not straightforward. First, information on the types of permits used for compliance is generally not available. Second, even when they are available, tying them to a price may not be straightforward, as this would require information on the year from which they date for banked permits or on the price at which carbon credits were purchased. Regarding this latter point, the price of credits used for compliance with government-mandated policies is not readily available (Wetterberg, Lanzi and Gómez, 2025[21]).
Source: Authors’ own elaboration.
References
[25] Alberta Government (2024), Alberta industrial greenhouse gas compliance, https://open.alberta.ca/publications/alberta-industrial-greenhouse-gas-compliance (accessed on 27 March 2025).
[18] Allen, M. et al. (2024), “The Oxford Principles for Net Zero Aligned Carbon Offsetting (revised 2024)”, University of Oxford, https://www.smithschool.ox.ac.uk/sites/default/files/2024-02/Oxford-Principles-for-Net-Zero-Aligned-Carbon-Offsetting-revised-2024.pdf.
[34] AlliedOffsets (2025), “CDR: State of the Market (July 2025)”, https://alliedoffsets.com/reports/.
[38] Australian Government CER (2025), Managing excess emissions (15/04/2025), https://cer.gov.au/schemes/safeguard-mechanism/managing-excess-emissions#borrow-from-a-future-baseline (accessed on 22 May 2025).
[22] Australian Government Clean Energy Regulator (2025), 2023–24 baselines and emissions data (15 April 2025), https://cer.gov.au/markets/reports-and-data/safeguard-data/2023-24-baselines-and-emissions-data (accessed on 22 May 2025).
[32] Australian Government Clean Energy Regulator (2025), Schemes, https://cer.gov.au/schemes (accessed on 30 April 2025).
[13] Australian Government DCCEEW (n.d.), Safeguard Mechanism overview;, https://www.dcceew.gov.au/climate-change/emissions-reporting/national-greenhouse-energy-reporting-scheme/safeguard-mechanism/overview#the-emissions-reduction-task (accessed on 22 May 2025).
[30] California Air Resources Board (2025), “Cap-and-Trade Program Data”, Fourth Compliance Period (CP4) Compliance Report, https://ww2.arb.ca.gov/our-work/programs/cap-and-trade-program/cap-and-trade-program-data (accessed on 30 April 2025).
[35] Carbon Pulse (2025), California’s latest offset issuance thins to 112k, no new DEBs, https://carbon-pulse.com/402425/ (accessed on 22 June 2025).
[8] Carbon Pulse (2025), China outlines timeline for futher ETS extension, shift to absolute cap (25/08/2025), https://carbon-pulse.com/428616/ (accessed on 25 August 2025).
[6] Carbon Pulse (2024), China ETS seen likely to introduce absolute emissions cap around 2030 (05/08/2024), https://carbon-pulse.com/310200/ (accessed on 27 May 2025).
[41] Cárdenas Monar, D. (2024), “Maximising benefits of carbon pricing through carbon revenue use: A review of international experiences”, I4CE Institute for Climate Economics, https://www.i4ce.org/en/publication/maximising-benefits-carbon-pricing-through-carbon-revenue-use-review-international-experiences-climate/.
[23] Dechezleprêtre, A., D. Nachtigall and F. Venmans (2018), “The joint impact of the European Union emissions trading system on carbon emissions and economic performance”, OECD Economics Department Working Papers, No. 1515, OECD Publishing, Paris, https://doi.org/10.1787/4819b016-en.
[31] Department of Ecology - State of Washington (2024), “Ecology Publications & Forms”, 2023 Cap-and-Invest Compliance Summary Report, https://apps.ecology.wa.gov/publications/summarypages/2414086.html (accessed on 27 March 2025).
[14] European Commission (n.d.), Allocation to industrial installations, https://climate.ec.europa.eu/eu-action/eu-emissions-trading-system-eu-ets/free-allocation/allocation-industrial-installations_en (accessed on 22 May 2025).
[4] Fischer, C. (2001), “Rebating Environmental Policy Revenues: Output-Based Allocations and Tradable Performance Standards”, Resources for the Future Working Paper, Vol. Discussion Paper 01–22, https://www.rff.org/publications/working-papers/rebating-environmental-policy-revenues-output-based-allocations-and-tradable-performance-standards/.
[3] Fischer, C., C. Qu and L. Goulder (2024), Rate-Based Emissions Trading with Overlapping Policies: Insights from Theory and an Application to China, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w33197.
[40] Flues, F. and K. van Dender (2020), “Carbon pricing design: Effectiveness, efficiency and feasibility: An investment perspective”, OECD Taxation Working Papers, No. 48, OECD Publishing, Paris, https://doi.org/10.1787/91ad6a1e-en.
[11] Flues, F. and K. van Dender (2017), “Permit allocation rules and investment incentives in emissions trading systems”, OECD Taxation Working Papers, No. 33, OECD Publishing, Paris, https://doi.org/10.1787/c3acf05e-en.
[2] Galdi, G. et al. (2022), “Linking emissions trading systems with different offset provisions: implications for linking”, Florence School of Regulation, Robert Schuman Centre for Advanced Studies, European University Institute, https://doi.org/10.2870/795960.
[7] General Office of the State Council (2024), “Work Plan for Accelerating the Construction of a Dual Control System for Carbon Emissions (08/02/2024)”, Reference number: Guo Ban Fa [2024] No. 39, https://www.gov.cn/zhengce/content/202408/content_6966079.htm (accessed on 27 May 2025).
[24] Government of Canada (2022), Greenhouse Gas Pollution Pricing Act: Annual report for 2022, https://www.canada.ca/en/environment-climate-change/services/climate-change/pricing-pollution-how-it-will-work/greenhouse-gas-annual-report-2022.html.
[12] ICAP (2025), Emissions Trading Worldwide: Status Report 2025., Berlin: International Carbon Action Partnership., https://icapcarbonaction.com/system/files/document/250409_icap_sr25_final.pdf.
[28] ICAP (2024), “Emissions Trading Worldwide - Status Report 2024”, https://icapcarbonaction.com/en/publications/emissions-trading-worldwide-2024-icap-status-report.
[15] ICAP (2023), Banking and Borrowing (Brief 8) (14/12/2023), https://icapcarbonaction.com/system/files/document/icap_briefs-en-brief-8.pdf (accessed on 22 May 2025).
[37] ICAP (n.d.), ICAP Allowance Price Explorer, https://icapcarbonaction.com/en/ets-prices (accessed on 22 May 2025).
[5] ICAP (n.d.), Status Report, https://icapcarbonaction.com/en/status-report (accessed on 21 May 2025).
[19] Intergovernmental Panel on Climate Change (2021), “Sixth Assessment Report, Chapter 3, Frequently Asked Questions”, https://www.ipcc.ch/assessment-report/ar6/, https://www.ipcc.ch/report/ar6/wg3/downloads/faqs/IPCC_AR6_WGIII_FAQ_Chapter_03.pdf.
[29] KRX (2025), Emissions Market Information Platform, https://ets.krx.co.kr/contents/ETS/03/03010000/ETS03010000.jsp (accessed on 31 July 2025).
[9] Kuneman, E. et al. (2022), “Benchmark-based allocation in emissions trading systems: experiences to date and insights on design”, Berlin: ICAP, https://icapcarbonaction.com/en/publications/benchmark-based-allocation-emissions-trading-systems-experiences-date-and-insights.
[17] La Hoz Theuer, S. et al. (2021), “Emission Trading Systems and Net Zero: Trading Removals”, Berlin: ICAP., https://icapcarbonaction.com/en/publications/emissions-trading-systems-and-net-zero-trading-removals.
[16] La Hoz Theuer, S. et al. (2023), “Offset Use Across Emissions Trading Systems”, Berlin: ICAP, https://icapcarbonaction.com/en/publications/offset-use-across-emissions-trading-systems.
[36] Leining, C. (2022), “A Guide to the New Zealand Emissions Trading Scheme: 2022 Update”, Motu Economic and Public Policy Research, https://www.motu.nz/assets/Documents/our-research/environment/climate-change-mitigation/emissions-trading/A-Guide-to-the-New-Zealand-Emissions-Trading-System-2022-Update-Motu-Research.pdf.
[42] Leining, C. and S. Kerr (2018), “A Guide to the New Zealand Emissions Trading Scheme. Report prepared for the Ministry for the Environment”, Wellington: Motu Economic and Public Policy Research, https://www.motu.nz/assets/Documents/our-research/environment/climate-change-mitigation/emissions-trading/A-Guide-to-the-New-Zealand-Emissions-Trading-System-2018-Motu-Research.pdf.
[27] Mizuho Research & Technologies, Ltd. (2024), “Tokyo Emissions Trading Seminar”, Reference Quotations for Transaction Prices in the Tokyo Cap-and-Trade Program, https://www.kankyo.metro.tokyo.lg.jp/documents/d/kankyo/2024_05 (accessed on 10 April 2025).
[43] OECD (2025), “Towards interoperable carbon intensity metrics: Assessing monitoring, reporting and verification systems”, Inclusive Forum on Carbon Mitigation Approaches Papers, No. 9, OECD Publishing, Paris, https://doi.org/10.1787/b185bcfa-en.
[39] OECD (2024), Pricing Greenhouse Gas Emissions 2024: Gearing Up to Bring Emissions Down, OECD Series on Carbon Pricing and Energy Taxation, OECD Publishing, Paris, https://doi.org/10.1787/b44c74e6-en.
[10] OECD (2023), Effective Carbon Rates 2023: Pricing Greenhouse Gas Emissions through Taxes and Emissions Trading, OECD Series on Carbon Pricing and Energy Taxation, OECD Publishing, Paris, https://doi.org/10.1787/b84d5b36-en.
[26] Québec Ministère de l’Environement (2025), The Carbon Market - Documentation, https://www.environnement.gouv.qc.ca/changements/carbone/documentation-en.htm (accessed on 27 March 2025).
[1] Verde, S. et al. (2022), “Linking emissions trading systems with different measures for carbon leakage prevention”, Florence School of Regulation, Robert Schuman Centre for Advanced Studies, European University Institute, https://doi.org/10.2870/651097.
[20] Wetterberg, K., J. Ellis and L. Schneider (2024), “The interplay between voluntary and compliance carbon markets: Implications for environmental integrity”, OECD Environment Working Papers, No. 244, OECD Publishing, Paris, https://doi.org/10.1787/500198e1-en.
[21] Wetterberg, K., E. Lanzi and N. Gómez (2025), “Exploring governments’ efforts to shape carbon credit markets: Possible actions to enhance integrity”, OECD Environment Working Papers, No. 263, OECD Publishing, Paris, https://doi.org/10.1787/0bafe9af-en.
[33] World Bank (2025), “State and Trends of Carbon Pricing 2025”, https://www.worldbank.org/en/publication/state-and-trends-of-carbon-pricing (accessed on 22 May 2025).
← 1. A number of other aspects contributing to increased flexibility and also impacting interoperability of systems are not discussed here. Many of them are documented in ICAP (2025[12]), e.g. price stability mechanisms (see also OECD (2023[10]), Flues and Van Dender (2020[40])), the use of revenues (see also OECD (2024[39]), Cárdenas Monar (2024[41])), auction design, underlying Monitoring, Reporting and Verification systems (OECD, 2025[43]) or fines for non-compliance.
← 2. Also referred to as an “absolute cap”, e.g. ICAP (2024[28]).
← 3. With some flexibility for adjustment mechanisms at the margin in many cases.
← 4. Intensity-based systems (e.g. ICAP (2025[12])) are also referred to as “rate-based systems” (Fischer, Qu and Goulder, 2024[3]) or “output-based systems” (e.g. the Canadian Federal OBPS).
← 5. While these are referred to as “baselines” and not free allowances, they are in practice the same as free allowances.
← 6. “Total emissions from all Safeguard facilities are also required to reduce over time, measured on a 5-year rolling average. From the financial year commencing on 1 July 2024, the rolling average of Safeguard covered emissions over the previous 5 years is required to be lower than the 5-year rolling average from three years earlier, and from 1 July 2027, the 5-year rolling average of Safeguard covered emissions is required to be lower than the 5-year rolling average from two years earlier.”
← 7. This may not always be the case. For instance, both the Chinese national ETS and the Beijing Pilot ETS are output-based systems. The introduction of auctioning is being considered for the Chinese national ETS, and up to 5% of allowances for irregular auctions may be set aside in the Beijing Pilot ETS.
← 8. This section discusses options to meet compliance obligations on verified emissions, i.e. once reductions in emissions have already been achieved.
← 9. These are sectors where international competition and carbon leakage risks may be limited and where free allowances could result in economic rents (see discussion in section 3.4 of OECD (2023[10])).
← 10. Also referred to as offsets in certain jurisdictions or publications (e.g. ICAP (2025[12])).
← 11. Entities can opt to receive units for embedding emissions in products or for destroying or exporting synthetic GHGs (Leining and Kerr, 2018[42])
← 12. This section discusses carbon credits used for compliance in ETSs, but it should be noted that some carbon taxes allow for this compliance option as well – e.g. the Colombia, Singapore and South Africa carbon taxes.
← 13. Environmental integrity requires several conditions to be in place for the generation of carbon credits, their trading environment, and their use (Wetterberg, Ellis and Schneider, 2024[20]). The discussion here focuses on integrity elements related to the generation of carbon credits (‘supply-side integrity’).
← 14. Further supply-side integrity considerations are described in (Wetterberg, Ellis and Schneider, 2024[20]).
← 15. This may evolve with the rules for Article 6 finalised in 2024, providing regulatory certainty for other countries to consider the potential inclusion of international carbon credits in their ETSs (see section 3.1).
← 16. https://www.environnement.gouv.qc.ca/changements/carbone/credits-compensatoires/destruction-halocarbures-en.htm#1, as accessed on 29 August 2025.
← 17. While most ETSs recognise that environmental integrity is essential for the inclusion of carbon credits, their environmental integrity guardrails differ considerably (Wetterberg, Lanzi and Gómez, 2025[21]), and the difference in standards followed can affect interoperability across systems but also with other policies (e.g. cross-border policies). Many ETSs only recognise carbon credits from specific crediting mechanisms. These crediting mechanisms, in turn, have different standards, methodologies and tools that guide the development of carbon credit supply. Most crediting mechanisms include provisions to assess additionality, ensure conservative quantification and permanence, but the standards differ: the stringency of such provisions vary greatly. In response to these integrity issues, several initiatives have sought to create international benchmarks for carbon credit quality. These include the UNFCCC-supervised Paris Agreement Crediting Mechanism, ICAO’s Eligible Emission Units, and the ICVCM’s Core Carbon Principles. The alignment of ETS-eligible carbon credit supply with these international quality benchmarks could help foster interoperability.
← 18. Compliance obligations generally refer to verified emissions in cap-and-trade systems and the difference between verified emissions and free allowances in baseline-and-credit systems.
← 19. These carbon credit prices are largely driven by the demand from voluntary corporate buyers, who have diverse preferences for carbon credits, compared to ETS-regulated buyers, whose primary motivation is to use credits for compliance.
← 20. Technology Innovation and Emissions Reduction.