FAQ on climate change-related indicators

Q1 Why are some of the indicators published as analytical indicators and some as official statistics?

The sustainable finance indicators have been published since September 2024 as official statistics, the highest level of quality assurance for ECB statistics. The transition and physical risk indicators remain labelled as “analytical”, meaning that they are still under development and have not yet reached the quality of experimental or official statistics. This is because there are gaps in firms’ sustainability-related data and financial information, and there are also uncertainties in climate modelling. Analytical indicators can still be relevant and helpful if they are used with care and taking into account the characteristics of the underlying data and methodologies. Further details on data caveats and methodology are provided in the Statistics Paper and the Technical Annex.

Q2 How can the climate change-related statistical indicators support the ECB policy functions?

The indicators are designed to support the ECB’s policy functions – monetary policy, financial stability and banking supervision. They are based on harmonised data sources and a transparent methodology, enabling consistent cross‑country monitoring of climate‑related vulnerabilities and their development over time.

The indicators feature in policy‑relevant publications such as the ECB/ESCB report on macroprudential frameworks for managing climate risk.

Q3 The ECB and euro area national central banks hold a large number of green bonds. Will these holdings be reported?

The Eurosystem’s holdings of green bonds or green debt securities are published at an aggregate level (see the ECB Data Portal). They account for roughly 16% of holdings of green debt securities within the euro area and around 4% of the Eurosystem’s total securities portfolio.

Q4 Green and other sustainable debt securities also include instruments where the sustainability status is self-declared by the issuer. Would it make a difference if we only considered sustainable debt securities which were “certified” or which had a “second-party opinion”?

Since November 2023 data on sustainable debt securities have been published for two levels of assurance: i) instruments with a second-party opinion validating the sustainability claims of the issuer, and ii) all sustainable instruments, i.e. with all degrees of assurance, including self-labelled instruments. Most of the sustainable debt securities issued and held by euro area residents have obtained a second-party opinion, with virtually all green debt securities having been externally reviewed. The number of “certified” green debt securities worldwide is still low (both for issuances and holdings). This is expected to change gradually in the euro area following the application of the European green bond standard in December 2024.

Q5 Why do calculations of financed emissions not include the ECB/national central banks?

The climate impact of both the corporate sector assets held by the Eurosystem for monetary policy purposes and the non-monetary policy portfolios of the ECB and national central banks has been disclosed in dedicated publications.

Q6 In June 2025 the ECB released its third climate-related financial disclosures, detailing the carbon footprint and other key carbon metrics on the climate risk exposures of the Eurosystem’s monetary policy portfolios, as well as the ECB’s foreign reserves and the ECB’s non-monetary policy portfolios. How do these differ from the statistical climate indicators on transition risk?

The ECB indicators on the Eurosystem’s monetary policy portfolios, as well as the ECB’s foreign reserves and the ECB’s non-monetary policy portfolios have a different scope and methodology from the climate indicators. They report the climate-related metrics of the assets held in these portfolios based on issuer-specific carbon emissions data and targets.

By contrast, the statistical climate indicators cover a larger sample of debtors and issuers – missing data have to be imputed, given the limited reporting of emissions at the firm level. Those indicators are only published at the aggregate country level and are used in macroeconomic assessments of transition risks.

Q7 Not all direct or indirect emissions are included in the ECB’s financed emissions indicators. Why not?

As data availability is limited, we are currently focusing on scope 1 emissions for loan-based emissions indicators and scope 1 and 2 emissions for securities-based emissions indicators. Scope 2 and 3 emissions cover a significant proportion of firms’ total emissions. However, information on scope 2 emissions is usually only available for large firms and at the group level. Scope 3 emissions are value chain emissions and are not even widely reported by large firms. To maintain comparability when reporting loan-based indicators, group-level consolidation is limited to scope 1, even if scope 2 emissions data are available for a subset of firms. Disclosure is expected to be enhanced in the coming years (e.g. in the context of the Corporate Sustainability Reporting Directive).

Q8 Is it possible to compare the different hazard data that the ECB publishes as part of the physical risk indicator?

It is possible to compare hazard data for windstorms, coastal flooding and river flooding, as the data for these hazards are all calculated using the same unit of measurement – expected monetary damage. Risk scores for other hazards are obtained from primary data sources based on different units and cannot be compared directly.

Q9 Why are the expected-loss indicators presented for only three types of hazard?

The normalised exposure at risk (NEAR) indicator and the collateral-adjusted exposure at risk (CEAR) indicator take the intensity of a hazard into account and would be a realistic measure of how a disaster translates into losses in the portfolios of the financial institutions. However, they require additional data on the value of a company’s fixed assets as well as an estimation of the potential loss caused by a hazard (i.e. how vulnerable each asset is to damage at the projected intensity of a disaster). This is currently only conducted for coastal flooding, river flooding and windstorms. Additional research is needed to expand the scope of the expected-loss indicators to other hazards.

Q10 Some physical risk metrics appear to be overly contained and seem to pose only a limited risk. Could there be measurement issues?

The ESCB physical risk indicators are compiled according to state-of-the-art methodologies and have been further improved with the latest release. Nonetheless, risks may be underestimated due to a range of data and methodological limitations, across both climate and financial dimensions.

Current climate models can understate potential impact because they often do not fully capture compound risks or climatic tipping points. First, hazards and the damage they cause are typically assessed in isolation. In reality, events can co-occur – such as a drought coinciding with heat‑driven wildfires – amplifying the impact beyond the sum of individual hazards. These compound events are difficult to model, even for climate scientists, because they are rare. Historical records are therefore limited, making robust statistical inference challenging. Second, tipping points are thresholds in the climate system beyond which small additional changes can trigger large, abrupt and potentially irreversible shifts (for example, rapid ice-sheet loss or changes in ocean circulation). Because these non-linear transitions are hard to predict and are often not fully represented in models, risk assessments that omit them can significantly understate potential impact. This is a broader challenge across climate science and is not unique to the banking sector’s modelling of physical risks.

From a financial perspective, expected-loss indicators capture only direct damage to assets and omit secondary impact such as business interruption, higher operating costs and supply-chain disruptions. Underestimation may also arise from factors like heat stress reducing labour productivity, as well as broader macroeconomic risks in the economies where firms generate revenue, with spillovers into the financial sector. In addition, our statistical climate indicators for euro area financial institutions are published at the country level. Banks are reminded of the importance of conducting detailed, institution-specific climate risk assessments and consulting relevant supervisory guidance.

Q11 Why does the potential exposure at risk indicator show such high values for water stress and windstorms but much lower values for flood risk?

The potential exposure at risk (PEAR) indicator indicates how widespread a hazard is. If an indebted company is located in a risk area, the full amount of its loans, issued debt and equity is incorporated into the PEAR indicator. Given that large areas of Europe are expected to be under water stress in the future, the respective PEAR indicator suggests that a relatively high number of financial portfolios will be affected. On the other hand, there are only a few companies in areas exposed to flooding, given the restricted geographical scope of the phenomenon (coastal and river areas). This translates into a relatively low PEAR indicator value. However, the indicators do not account for flash floods, which are sudden, high-intensity flooding events triggered by intense rainfall. These events can occur far from rivers or traditional flood zones and often impact urban areas, where water cannot infiltrate the ground and instead flows rapidly over the surface.

Windstorms are common in Europe and are characterised by a relatively high PEAR indicator value. However, the NEAR indicator has lower values, reflecting the robustness of building stock and limited potential damage. The NEAR indicator accounts for the potential losses caused by a natural hazard and is better at capturing potential financial risk than the PEAR indicator.

Q12 Over recent years, there were several strong windstorms that caused significant damage (e.g. storm Éowyn or storm Eunice). How does this square with windstorm exposures being mostly classified in the low risk category?

While it is true that some recent windstorms have caused significant damage, windstorm exposures are still classified as low risk. This can be explained by how risk scores are calculated and the relative resilience of infrastructure across regions.

For most hazards, risk scores are based on metrics specific to each hazard, e.g. for water stress this would be the ratio of water demand to water supply. As such, these scores are not comparable across hazards. However, for hazards where damage can be quantified – specifically flooding and windstorms – the risk scores are determined by the share of expected annual loss relative to a company’s assets. This approach enables risk categories to be compared across these hazards, which is desirable but may result in low variation in scores based on relative damage levels. This is particularly evident for windstorms, where most exposures are classified as low risk – partly due to relatively robust building design in Europe – whereas flooding tends to result in more damage.