A decision-useful assessment of a company’s transition strategy often requires an asset-level and forward-looking analysis that identifies the key risks and opportunities the company faces. Many large financial institutions (FIs) already carry out transition assessments, but common practice is to focus on target ambition, climate governance, and reported transition-aligned capital expenditures in previous years, rather than planned expenditures. While this provides a useful foundation for understanding the ambition of clients, it does not necessarily surface how specific assets, technologies, or business activities are exposed to future risks and opportunities. As a result, the assessments typically do not find applications outside the sustainability or sustainable finance teams, and the impact on deploying additional transition finance may be limited.

This insight brief discusses how banks can leverage asset-level forward-looking data at scale to generate transition intelligence relevant to multiple functions — from sustainability to the front-office and credit risk teams. Although focused on banks, the analysis presented here could similarly be applied by other types of financial institutions.

We present a case study from the steel sector, analyzing the steelmaker ArcelorMittal using data from Forward Analytics’ (FA) Digital Transition Twin dataset. The intent is not to single out ArcelorMittal as they, like many steelmakers, navigate a challenging and uncertain policy and market environment. Rather, this case study demonstrates how forward-looking asset-level data from third-party providers can be applied in a systematic and scalable manner — streamlining otherwise costly assessments and mitigating the challenge of gaps in corporate transition plan disclosures.

This case study follows the structure laid out in RMI’s guide How To Assess the Alignment of Corporate Investment Pipelines and compares the investment pipeline of the ArcelorMittal to two sets of benchmarks:

• Company targets: to evaluate the feasibility of targets and internal consistency of targets with actual company strategy, highlighting potential “execution gaps” where the pipeline falls short of meeting stated ambition.
• Transition pathways: to assess alignment with different future plausible operating environments and identify the associated risks and opportunities presented by these futures. This includes comparing emissions trajectories as well as technology mixes.

Equipped with these insights, FIs can develop transition intelligence for application across several key use cases:

• Identify opportunities for engagement to close gaps between ambition and execution, and structure transition-finance products tailored to clients’ investment needs.
• Anticipate technology, policy, and market dependencies tied to investment decisions, informing risk assessments and investment opportunities.
• Evaluate the projected impact of a company on portfolio-level transition metrics and targets.

In this case study, we examine ArcelorMittal, a large-cap steel producer with a globally diversified portfolio of steel assets. ArcelorMittal plans to expand production in the coming years and has a 2030 emissions intensity target at the group level and for its European operations.

The high-level steps of this analysis are:

1. Construction of the forward-looking dataset
2. Comparison to company targets
3. Comparison to global transition pathways
4. Comparison to regional transition pathways
5. Understanding the technologies driving future emissions trajectories

1. Construction of the forward-looking data set

The FA data profile is constructed in a bottom-up way using modeled emissions intensities at the asset level, accounting for technology configuration, furnace type, and production route.[1] In cases where production is announced but technologies are unknown, an average emissions intensity is assumed. These asset-level emissions intensities have been aggregated to the parent company using production-weighted averages that account for ownership shares across all subsidiaries and intermediate entities. This ensures that each asset’s contribution to the group-level emissions intensity reflects the company’s effective ownership stake. Additionally, assets can be consolidated up ownership trees via different methodologies, which may yield different outcomes with different interpretations (see footnote for an example[2]). The map below shows the distribution of ArcelorMittal’s facilities around the world, with each facility containing one or more assets.[3]

The dataset relies on historic reported emissions up to 2023, modeled emissions based on asset-level emissions intensity and reported production in 2024, and forward-looking projections based on announced capacity additions, retirements, and asset-level emissions intensity from 2025 to 2030. From this dataset, a projected emissions intensity trajectory has been calculated for ArcelorMittal as shown below.

Given that we have geographical information, we can further break down the projected emissions intensity by geography. Of particular interest here, we can look at the company’s projected emissions intensity in Europe where it has set an emissions reduction target, and in India where the majority of its capacity expansion is targeted.[4]

It is important to note that these are estimated trajectories based on publicly available data. Similar analyses could make slightly different assumptions about the utilization rate of different assets, their respective emissions intensity over time, and potential improvements in efficiency that would yield different results. Additionally, while capital expenditure (capex) plans typically include the start year for an asset, it is often unclear when assets will retire or ramp down. In cases where that information is not available, it is assumed that the asset is operational through 2030. Non-public plans for new assets or ramp down of existing assets would change these estimated trajectories. However, given the long lead time on deploying new steelmaking assets, it is unlikely that non-public plans would make a meaningful difference to the time horizon in focus here (to 2030).

Key takeaways:

• With sufficiently granular forward-looking data, it is possible to estimate how the emissions intensity trajectory of a steel producer will evolve over time. A similar approach could be applied to other sectors where data is available.
• Bottom-up emissions trajectories built from asset-level data enable methodologically consistent, like-for-like comparisons across corporations, regardless of differences in their corporate disclosures.

2. Comparison to company targets

Company targets provide an important indication of ambition and signal strategic direction. However, it can be unclear how feasible it will be for the company to achieve them or how well the company is executing on this ambition. To answer these questions, we can compare the projected emissions trajectory based on ArcelorMittal’s current investament pipeline to their publicly stated target.

ArcelorMittal has  a group level  target to reduce Scope 1 and 2 CO₂e emissions intensity 25% by 2030 from a 2018 baseline. This target is raised to 35% for its European operations.[5] This European target is compared against ArcelorMittal’s projected emissions-intensity pathways below. Additionally, there may be slight emissions accounting and financial consolidation differences between the Forward Analytics and ArcelorMittal approaches.

With those caveats in mind, the chart above shows that ArcelorMittal is projected to fall short of target based on current public investment plans. Given the long lead-time in developing steel production assets, it seems unlikely that this shortfall would be overcome without significant changes in the company’s investment pipeline — creating an opportunity for financiers to explore options for supporting ArcelorMittal’s transition objectives. ArcelorMittal itself noted in its 2024 sustainability report that it will likely not be able to meet its 2030 emissions intensity target as the company was unable to take final investment decisions on lower carbon production routes, citing “European policy, energy and market environments had not moved in a favorable direction.”

It is important then to determine what transition investments are still possible and to avoid carbon lock-in with new carbon-intensive assets. ArcelorMittal’s XCarb innovation fund is one avenue for this, investing in breakthrough low-carbon technologies that can support decarbonization of the steel sector.

Falling short of targets can raise reputational risk for both clients and financial institutions. At the same time, it raises a clear point for engagement teams to discuss with these clients and understand what is driving this gap. We should note that the execution gap is not necessarily a sign of low ambition or greenwashing, there are often valid constraints preventing alignment with high ambition targets. Bank sustainability and front-office teams can work with relationship managers to explore questions such as: What are the market, policy, and technology conditions limiting progress? What progress is possible under current conditions and how could conditions be changed? Front-office teams could then use this as an opportunity to identify projects and corresponding financial products to support these clients to get closer to meeting their targets.

Key takeaways

• Forward-looking assessments of corporate investment pipelines allow financiers to evaluate the feasibility of meeting corporate targets.
• Execution gaps highlight potential risks, including reputational risks, and provide additional information for engagement.
• Execution gaps also raise potential financing opportunities to support clients in meeting their targets.

3. Comparison to global transition pathways

Comparing the investment pipeline to company targets highlights the gap between execution and ambition, but is not necessarily indicative of the broader transition risks and opportunities a company is facing. For this, we need to understand the broader plausible future operating environments of companies. Measuring alignment with transition pathways can offer such insights.

Transition pathways provide forward-looking benchmarks for what the future operating environment could look like under different technology, market, and policy conditions.[6] These benchmarks take various forms, with some being normative — looking at what is needed to achieve global climate goals like 1.5 °C, such as the IEA Net Zero Emissions by 2050 scenario (NZE) — and others being more exploratory — looking at more probable outcomes based on governments’ announced commitments or stated policies, such as the IEA Announced Pledges Scenario (APS) or the Stated Policy Scenario (STEPS).

To determine alignment to pathways, a convergence approach is used to establish company-specific benchmarks derived from the endpoint of a chosen pathway(s). This approach, pioneered by the Science Based Targets initiative, proportions decarbonization efforts to individual entities based on their relative starting position and requires all entities to converge on the same end point of the given transition pathway. The starting point used in this analysis is the year 2024, which has been compared to the IEA’s scenarios from the same year,[7] as seen in the results below.

We observe that ArcelorMittal’s projected emissions intensity trajectory is misaligned with the NZE global pathway, a commonly used pathway when assessing 1.5°C climate outcomes. This can be interpreted from an impact perspective as the company falling short of global climate targets. From a transition risk perspective, the misalignment can be interpreted as follows:

If global policy, technology maturity, and market dependencies were to evolve at the pace required by the IEA NZE scenario, ArcelorMittal could be exposed to transition risks.

However, the IEA-NZE may not be the most probable outcome. To assess other plausible future operating environments, we can compare the projected emissions intensity trajectory with the APS and the STEPS pathways. Misalignment with the APS can be interpreted as:

There is an inconsistency between the company’s own investment plans and the operating environment that governments plan to create to achieve their climate ambitions.

We observe that the projected intensity also falls above STEPS, which can be interpreted as:

The current investment plans are associated with a worse climate outcome than that projected under governments’ current policies.

Given the current challenges of low-carbon steel production, this is intuitive and arguably prudent from a traditional risk perspective. However, this misalignment with global climate goals implies higher exposure to future transition risks such as a tightening of policy over time as physical risks manifest further. Put differently, if governments’ current ambitions are not met and the impacts of physical risks become more apparent, a late and sudden transition may become reality. Governments may be compelled to implement measures more disruptive than the current NZE scenario implies to manage worsening physical risks.

For some technologies and sectors, this raises a difficult tension between near-term risk-return profiles, and transition risk exposure in the future. Again, this highlights the importance of determining which transition investments are still possible — such as operational and efficiency improvements — and avoiding carbon lock-in from new investments.

Key takeaways

• Transition pathways help us understand potential future operating environments.
• Evaluating alignment to these pathways builds understanding of exposure to potential transition risks in different scenarios.

4. Comparison to regional transition pathways

Above we interpreted the projected emissions intensity alignment and transition risks against a set of global scenarios. This is a helpful but limited analysis given transition risks will be regionally differentiated. To account for these geographic differences, here we compare ArcelorMittal’s regionally differentiated trajectories to a range of Europe-specific scenarios from the Mission Possible Partnership (MPP) using the same convergence approach as above.

We can see a similar pattern to the global scenarios with ArcelorMittal being misaligned with MPP’s Europe-specific “Carbon cost” and “Technology moratorium” scenarios. The carbon cost scenario assumes a carbon price that is unlikely to become a reality before 2030, but the technology moratorium scenario assumes no specific climate action in the steel sector before 2030. As such, misalignment to both pathways would imply transition risks for ArcelorMittal in the future based on MPP’s modelling assumptions.

This analysis could be repeated with other Europe-specific steel pathways as part of a scenario analysis exercise to further identify what future operating conditions ArcelorMittal is and is not aligned with, and the policy-driven transition risks this may raise. It could likewise be repeated across other regions where ArcelorMittal is active for a region-by-region assessment of alignment, and the risks and opportunities that alignment implies.

RMI’s transition pathway repository is being built to make it easier for FIs to identify and interpret a range of region- and sector-specific pathways which can serve as an input to this process. Though it currently only covers the power sector, coverage will be expanded to the steel sector this year.

Key takeaway

• Asset-level data enables mapping to region-specific transition pathways, linking assets to local dependencies and risk drivers. This increases the granularity of transition intelligence and enables regionally differentiated risk interpretation.

5. Understanding the technologies driving future emissions trajectories

To understand why ArcelorMittal’s emissions-intensity trend evolves the way it does, we next examined the underlying technology mix that shapes the forward-looking emissions trajectory. By analyzing planned capacity additions, asset retirements, and shifts between production routes, we can identify the technological drivers of decarbonization, lock-in risks, and exposure to future policy or market pressures. This same type of analysis can be applied in many sectors, and these findings then provide key decision points for FIs to engage with clients to manage their own risk exposure.

As noted in section 1, there may be non-public plans to ramp-down or retire some assets which are not captured by this dataset. Where no retirement year has been announced, it is assumed that assets will operate until 2030.

The largest share of ArcelorMittal’s projected capacity growth is in lower-carbon direct reduced iron–electric arc furnace (DRI-EAF) technology, which is positive. The company also already operates a modest share of EAF with scrap steel, around 13% of its technology mix, which remains relatively constant over time. However, ArcelorMittal also plans to expand its carbon-intensive blast furnace–basic oxygen furnace (BF-BOF) capacity, partially negating the modest emissions intensity improvements of the DRI-EAF additions. Furthermore, the long average economic life span (~20 years) of BF-BOF assets will result in carbon lock-in, affecting the company’s ability to pivot to low-carbon alternatives if the operating environment shifts toward lower-carbon technologies in the medium to long term.

There is also an “unknown” category, which includes planned capacity where the underlying technology has not been announced publicly. This unknown category presents an opportunity for front offices to engage and discuss technology build-out plans, and the enabling factors that need to be in place to deploy lower-carbon technologies. These engagements can further evolve to explore new product lines, including transition finance and/or specific project finance for low-carbon technology build-out where access to capital is a limiting factor, rather than policy, market, or technology barriers.

Key takeaways

• Asset-level data allows us to identify the technologies that are driving a company’s emissions-intensity trajectory, on what timeline, and how this interacts with their climate targets.
• This granularity supports more informed client engagement, prompting discussions on technology choices and constraints rather than just on the level of corporate ambition.

Forward-looking, asset-level data supports decision-making across multiple functions within a financial institution by providing a consistent, scalable input, creating transition intelligence for sustainability, risk, and front-office teams.

For sustainability teams

The combined use of forward-looking asset-level data and alignment assessments can inform transition planning at the financial institution level. In jurisdictions such as the EU, this approach can play a direct role in meeting supervisory expectations. Applied at the corporate level, it enables assessment of how clients’ forward-looking (mis)alignment may contribute to, or hinder, portfolio-level climate targets.

Using bottom-up asset-level third-party inputs allows sustainability teams to apply consistent and comparable benchmarks across clients, while offering a scalable approach that is not reliant on mining corporate disclosures or additional resource-intensive engagement processes.

In this case study, ArcelorMittal’s projected emissions performance would not contribute to meeting a lender’s 1.5°C portfolio targets. However, this is in the context of a steel sector where very few companies are able to meet high ambition targets. In that context, attention could turn to avoiding carbon-lock-in, which would further jeopardize future targets, and policy advocacy to support the sector’s transition in the regions of greatest exposure.

For risk teams

The use of forward-looking data is critical, enabling ex ante risk assessments that support understanding of how transition risks may materialize over time. Furthermore, asset-level data enables geographically specific assessments of risks and dependencies when paired with region-specific transition scenarios. This approach also supports the development of sectoral risk appetites and key risk indicators linked to forward (mis)alignment with transition pathways.

In this case study, risk teams could set region-specific thresholds or risk indicators on emissions or blast furnace exposure based on expected policy developments in the regions where they have exposure. These could be based on transition pathways that reflect a financial institution’s perspectives on the transition and its associated risk appetite. They could also be based on enacted or draft legislation related to national climate targets where such legislation exists.

For front-office and client-facing teams

Asset-level insights can be used in engagement conversations to discuss clients’ investment pipelines and identify potential misalignment with their own targets and/or with plausible future operating environments, as assessed through transition pathway alignment. The use of production-based asset-level data anchors these discussions in core business outputs rather than less tangible emissions metrics.

Where this analysis reveals execution gaps, front-office teams can explore potential financing solutions, including entity-level transition finance or dedicated project finance to support lower-carbon technology build-out. When combined with assumptions on capital expenditure, an asset-based approach also enables the construction of more traditional finance metrics, linking climate considerations to core business decisions.

Based on this case study, client-facing teams could engage on the barriers to deployment of low-carbon technologies in Europe and explore what deal structures could make low-carbon projects feasible where financing is a barrier. In many cases, policy engagement will also be necessary to improve bankability.

There are further opportunities to expand this analysis into financial data and metrics. This will make it easier to incorporate the results into existing FI processes and frameworks that risk and front-office teams are more familiar with. For example, the forward-looking investment pipeline and emission-factor data described above can be combined with assumptions about capex requirements for different technologies to estimate the additional capex required for these companies to meet their targets or align with the selected transition pathways. These capex figures can then be compared to historic investment volumes, market capitalization, and company balance sheets to evaluate the financial feasibility of meeting their targets or achieving alignment with different pathways. The estimated capex requirements further indicate the scale of the opportunity to provide financial products to support deployment of low-carbon assets.

The transition data landscape is evolving rapidly, with increasing granularity and transparency making it possible to analyze companies in new ways. Forward-looking, asset-level data enables financial institutions to move beyond high-level company transition assessments focused primarily on target ambition, and to better evaluate the feasibility of that ambition. It provides a scalable and decision-useful foundation for developing transition intelligence through the analysis outlined above.

This analysis gives financial institutions the information they need to understand how their clients are positioned with respect to the transition, the risks and opportunities this presents, and react accordingly. Open source and commercial data providers are building the data architecture that makes these assessments possible, and RMI is building tools and developing guidance to make use of this data and scale in-depth transition assessments. Learn more and find resources at the transition finance resource hub.

[1] Steel production assets are modeled with different emission factors drawn from Climate TRACE, which are based on their technology archetype. Scope 1 emissions are modeled based on the technology type, ranging from low- to high-carbon technologies, with adjustments made for the percentage of scrap being used in electric arc furnaces. Scope 2 emissions are modeled by taking the product of the average electricity use factor for the different technology pathways and the regional grid emissions intensity.

[2] The asset “AMNS-I” is included in the analysis despite being a joint venture that is not operationally controlled by ArcelorMittal or consolidated in ArcelorMittal’s financial statements. Its inclusion here reflects that the group remains economically exposed to this asset and any related transition risks. From a strict emissions accounting perspective, excluding the AMNS-I asset would ensure consistency with ArcelorMittal’s reported boundary and targets. However, attributing emissions on an equity-share basis (60%) provides a more complete view of economic exposure and transition risk.

[3] For ArcelorMittal, the FA data profiles capture the global operational footprint across 25 subsidiaries in 18 countries, including 73 physical assets (furnaces) across  36 steelmaking facilities as of 2024.

[4] The emissions intensity for ArcelorMittal Europe is calculated based on the asset boundary disclosed on pp. 65 and 185 of ArcelorMittal S.A.’s 2024 Annual Report. The European perimeter is defined according to the reporting entity boundary (i.e., subsidiaries classified as European within the Group), rather than the physical location of the asset. Consequently, assets owned by European entities are included even if located outside Europe (e.g., ArcelorMittal Sonasid Casablanca, Morocco), while assets classified by the Group as “Other” (e.g., Ukraine) are excluded. In addition, following the Italian Government’s February 20, 2024, decree placing Acciaierie d’Italia (ADI) into extraordinary administration and the resulting loss of control recognized by ArcelorMittal (2024 Annual Report, p. 216), the Taranto plant is excluded from the date of loss of control onward, consistent with the operational control principle applied in Scope 1 and Scope 2 greenhouse gas accounting. Scope 2 emissions for Europe are estimated by applying the 2024 disclosed Scope 2-to-Scope 1 ratio disclosed by ArcelorMittal Europe (3.3%).

[5] A 2018 baseline year Scope 1 and 2 CO₂e of 1.6 tCO₂/t steel from Forward Analytics data set for ArcelorMittal’s European operations was used to calculate this target.

[6] Transition pathway is a generic term for resources that provide a forward-looking description of how companies, sectors, and/or regions can or may evolve over time. Scenarios are a subset of transition pathways, indicating that the resulting pathway is based on a systematic modeling exercise. Scenarios offer the most systematic and detailed pathways but are not the only viable source of pathways.

[7] Scenarios are regularly updated and banks are encouraged to use the most up-to-date scenarios. In this case study, a set of three scenarios has been chosen from the IEA WEO 2024 publication, maintaining consistency with the underlying asset-level data being used. This choice is intended to be illustrative of how a bank can interpret alignment to different scenario outcomes. The IEA came out with a new set of scenarios in 2025, which would yield different interpretations.