Response to consultation on Implementing Technical Standards on amended disclosure requirements for ESG risks, equity exposures and aggregate exposure to shadow banking entities

Go back

1. Do you have any comments on the proposed set of information for Large institutions?

While we support the EBA’s initiative to shift certain templates from a semi-annual to an annual reporting frequency, we strongly advocate for an annual frequency to be applied universally to all ESG Pillar 3 disclosures, including those for large, listed institutions (where applicable), based on the principle of materiality. Adopting an annual disclosure approach would harmonize with institutions’ reporting cycles for other ESG-related disclosures, such as those required under the CSRD and the Taxonomy Regulation. Since these disclosures are utilized in Pillar 3 reporting, aligning them on an annual basis would promote consistency and minimize duplication. Furthermore, it is worth noting that the BCBS voluntary framework for disclosing climate-related financial risks also adopts an annual reporting approach. Semi-annual disclosure of EUT templates requires a significant effort from disclosing companies while providing limited insights among others also because EUT KPIs of counterparties are only available on an annual basis (emissions, EPC...); targets are set on an annual basis.

Furthermore, we have noted that under article 433a of CRR 3, large non-listed institutions are expected to publish ESG disclosures annually. Yet, in the EBA consultation, a semi-annual frequency is proposed for these institutions, which represents a tightening of the framework.

If the EBA does not adopt this general approach, we believe that, at the very least, the frequency for large non-listed institutions should be revised back to annually to ensure consistency with CRR 3. It seems essential to remain aligned with CRR 3 to avoid unnecessary complexity and reporting burden for these institutions.

In case annual frequency be granted to large institutions, it should be clarified when publishing the final draft, ITS that June 2026 disclosure is not required.

The Pillar 3 ITS, should be carefully reviewed to also avoid duplication with existing reporting and disclosures. It should be carefully considered whether information in the P3 ITS would provide market participants with useful information.

2. Do you have any comments on the simplified set of information for Other listed institutions and Large subsidiaries?

As a general comment, we do not see where the proposed revised Pillar 3 templates, which add granularity and complexity, bring simplifications for large institutions.

A simplified set of disclosure for large subsidiaries and Other listed institutions is welcomed and in line with the aim to reduce overall disclosure burden for entities in scope. However, consistency with reporting requirements of large institutions is essential to ensure comparability between the disclosure of the subsidiary and its parent. In particular, we would appreciate the following clarifications:

that the “proportionate approach” disclosed on annual basis also applies to large, listed subsidiaries;
on whether a Bank that is both a large non-listed entity and a large subsidiary—being the largest financial institution in its Group, controlled by an EU parent financial holding company that prepares CRR Part 8 disclosures at the highest level, and also producing sub-consolidated disclosures—must apply Article 449a disclosure obligations at the sub-consolidated level, and whether it should make Pillar 3 ESG risk disclosures as a large non-listed entity, as a large subsidiary, or under both categories.

3. Do you have any comments on the simplified set of information proposed for SNCI and other non-listed institutions?

The scope of application for Pillar 3 ESG disclosure should be harmonized with that of the CSRD.

Exclusion of SNCIs and institutions that are not subject to the CSRD reporting also from the scope of ESG reporting under Art 449a CRR should be considered. Should they be kept in the scope, their reporting obligation should indeed be simplified, and more proportionate approach should be adopted as proposed by the EBA.

Removal of subsidiaries from Pilar 3 ESG disclosure requirements of banking groups with large subsidiaries should also be considered as it will substantial ease the reporting burden (while banking groups will continue reporting on their ESG risks at the top-consolidated level.) In our view, subsidiary exemption should be applicable to all subsidiaries regardless of their size under both CSRD and CRR.

In addition, the consultation refers to the “omnibus” (which will enter into force for SNCI banks on 31/12/2028), but does not mention the ESG Risk Management Guidelines, which will only become legally binding for SNCIs from 11/01/2027. However, the disclosure proposed in this consultation (which also includes requirements regarding the scenarios used, as well as social and governance risks) will become binding with reference date 31/12/2026. We ask to align the dates with those of the ESG Risk guidelines (ITS entry into force from January 2027), in order to ensure consistency in the treatment of social and governance topics).

4. Do you have any comments on the proposed approach based on materiality principle to reduce the frequency (from semi-annual to annual) of specific templates (qualitative, template 3, and templates 6-10) for large listed institutions?

We recommend that the EBA apply a consistent materiality approach, set in the Guidelines on the management of environmental, social and governance (ESG) risks, to these disclosures in order to reduce the reporting burden on institutions and ensure that only material information is disclosed.

Based on the content of the templates and the reporting experience of the past years, a semi-annual reporting of these templates has limited to no additional value: Template 3 alignment metrics are usually updated on an annual basis, disclosure on semi-annual basis does not provide additional or new information (as also mentioned in article 29 of the consultation paper).

Also, should the link to the Regulation 2021/2178 (EUT Art 8) remain, it requires yearly reporting. As a result, semi-annual disclosures in Pillar 3 ESG would not be aligned with EUT article 8 requirements, nor the aim to reduce reporting burden for companies in scope.

Semi-annual reporting of EUT templates requires a significant effort from reporting companies while providing limited insights due among others to the fact that EUT KPIs of counterparties are only available on an annual basis. We therefore suggest adjusting the disclosure requirement of all templates (including Templates 1, 2, 4 and 5) to annual disclosure frequency without any conditions. Based on the nature of these templates, such frequency would be sufficient for the users of the information, similarly to the frequency suggested in the voluntary ESG Pillar 3 disclosure framework recently published by the Basel Committee.

5. Do you have any comments on the transitional provisions and on the overall content of section 3.5 of the consultation paper?

We appreciate the inclusion of the transitional provisions to adapt to the revised ESG disclosure requirements, particularly the suspension of Green Asset Ratio (GAR) and EU Taxonomy-related disclosures in templates 6-10 until the end of 2026. However, the final decision for banks to make use of this transitional provision is dependent upon the competent authorities agreeing to provide the flexibility as proposed in the consultation and until then the banks will have to continue reporting the GAR Templates (6-10).

The transitional provisions postpone the disclosure obligation under templates 6–10 until 31 December 2026 for large listed and large non-listed institutions, but only for those required to report under the Delegated Disclosure Act (Article 8 of the Taxonomy Regulation), which in turn applies only to CSRD-reporting entities. Due to the “Stop the Clock” Directive delaying CSRD obligations for some banks until 2028, many may have to disclose under templates 6–10 in 2026 despite not yet reporting under CSRD. We therefore propose to align the deferral period for these templates with the CSRD timeline.

Further, the European Commission allows financial companies to no longer report detailed taxonomy information and KPIs until 31 December 2027, until the Commission reviews in detail the Taxonomy disclosure rules and technical screening criteria. The EBA should thus align the suspension deadline for taxonomy templates with the Taxonomy DA, until 31 December 2027.

We also welcome the proposal allowing bank’s large subsidiaries to refrain from standalone disclosures for GAR and alignment metrics. This is particularly helpful for large banks, managing sustainable finance and net zero targets at a group level

We also would like to make the following more specific comments:

Given that EU Taxonomy alignment inputs are still required in Templates 1 and 4 (column ‘of which environmentally sustainable CCM’), we kindly request the EBA to confirm that these columns are also covered by the transitional provisions. In this context, we would like to request the full deletion of Taxonomy-related data requirements from Pillar 3 disclosures—specifically the CCM columns in Templates 1 and 4—in order to avoid unintended consequences, such as misalignment, confusion, or unnecessary reporting burdens during the transition period.
Furthermore, we note potential overlaps between Template 10’s "Mitigating actions outside the EU Taxonomy" and various sustainability disclosures under CSRD. We suggest that the EBA recognise these overlaps and refrain from requiring duplicative data under Pillar 3.
Lastly, we would like to highlight that Templates 1 and 4 still rely on data from templates proposed for suspension. This continued dependency would still necessitate significant effort in calculation and disclosure, which ultimately undermines the objective of suspending the original templates.

6. Do you have any comments on the proposed amendments to Table 1 and Table 3?

The qualitative disclosure items are duplicative of CSRD corporate disclosures, which is a more appropriate location for such disclosures than the Pillar 3 ESG risk disclosures.

The information required in tables 1, 2 and 3 is mostly available in the annual report (CSRD). For the sake of simplification, this information should not be required twice. Therefore, we would suggest reducing the information available in the annual report at the same time reviewing the usefulness of the remaining information.

In the event that the deletion of the duplicate information already contained in the annual report is not considered, we would suggest cross-referencing [1] the annual report in order to reduce the burden for the large entities.

Furthermore, we would like to make the following, more specific comments:

In the instructions to the tables, paragraphs 4, 6, and 8 state that “Institutions shall disclose this table on an annual basis, based on materiality assessment”. It is unclear whether this should be interpreted as allowing an institution, which as part of its materiality assessment has concluded that it is not exposed to material environmental, social, or governance risks, to omit disclosure in Tables 1, 2, or 3 respectively. To eliminate this uncertainty, we suggest that EBA aligns the materiality assessment with the guidelines on ESG risk management.
Currently institutions are required to respond to all potential governance topics listed in Table 3, row (c)(i-vi) and row (d)(i-vi). We propose increasing the flexibility of institutions' responses by rephrasing the questions as:
- “(c) Institution’s integration in governance arrangements of the governance performance of their counterparties, considering e.g.:” and
- “(d) Risk management arrangements of the institution to assess and consider the governance performance of their counterparties, considering e.g.:”.
The related instructions should be updated accordingly.
Regarding the EBA GL on Risk Management we would like to make the following comments:
Table 1 (environmental risk):
a) the resilience of strategy and business model to environmental changes should be able to be implemented by the end of 2026. To define the horizon terms used by the institution should only concern short and medium terms, since long term is already defined as being at least 10 years in the EBA GL on ESG RM
k) tools by which the institution identifies and monitors its activities and exposures which are sensitive and vulnerable to environmental risks, or financially material should be able to be implemented by the end of 2026 for EBA GL ESG RM 28 and 29 partly because the list of data points required needs efforts for further development of AI-based tools, and proxy-tools, and by end of 2027 for GL 36 on nature need to develop tools to measure financial impacts
n) data and information available to carry out the risk management of environmental risks should be able to be implemented by the end of 2027 for EBA GL ESG
Table 3 (governance risk) as well as for Table 2 (social risks):
assess information related to their counterparties’ (i) alignment with international standards and (ii) negative material impacts on own workers, workers in the value chain... should be able to be implemented by the end of 2027

[1] When cross referencing is granted, it should be ensured that absolutely present and future alignment with Taxonomy Regulation is provided. In terms of content but also in terms of frequency.

7. Do you have any further suggestions on Table 1A?

We recommend that for consistency, the changes will be implemented for other reporting requirements which share the similar sectors’ breakdown.

8. Do you have any comments on the proposed additions and deletions to the sector breakdown?

ETH Zurich classification is not so easy to be implemented. This methodology increases the difficulty of classifying entities as there are entities with different economic activities. We seek clarification on the rationale behind the inclusion of ETH Zurich classification, as it appears to duplicate the existing framework concerning PAB excluded exposures.

We suggest not to include:

G 46.81 and G 47.3: If so, we would be linking the sale to the contributing sector. It is not a production economic activity as such. This would be inconsistent and would entail a risk of double counting. Same litre of oil is extracted, refined, transported and sold at the petrol station.
D 35.4: Similar reasoning would apply for brokers and agents’ activities

Additionally, we would find it helpful to have a list of NACE codes that could be used to consistently report exposures to fossil fuel sector entities in templates 1 and 3 of Pillar 3 disclosures and template 2 of EU taxonomy regulation. On comparison, there are gaps and inconsistencies across these templates as designed to report exposures to fossil fuel sector entities. Also. we would welcome non-duplicative reporting across Pillar 3 and EU taxonomy regulation.

The final amendments should be implemented in the other regulatory reporting exercises which share a similar sector breakdown.

Given the effort required to fully implement the proposed amendment (including at the IT level) it would be reasonable to provide a grace period within which institutions can continue to disclose the template using NACE 2.0

9. Do you have any views with regards to the update of the templates to NACE 2.1?

We support the proposed updates to the sector breakdown in Template 1 Credit quality of exposures by sector, emissions and residual maturity. However, we find the agriculture breakdown complex without added value.

We recommend adopting NACE Revision 2.1 while supervisory reporting frameworks like FINREP incorporate it. A synchronized rollout is necessary to ensure consistency and make cross-referencing easier.

As acknowledged by the EBA in its press release, the harmonisation of NACE codes is essential to reduce costs for banks and to enhance the analytical quality of reported data.

Moreover, we would like to highlight that it is important for banks to implement the revised NACE codes at the same time across all the reporting framework and to avoid sequential changes that would require banks to maintain several versions of NACE codes in parallel and consequently, that would imply additional undue operational costs.

Following the JBRC- Advice on the implementation of the NACE Rev2.1, it appears that the revision of the NACE codes will start from 1 January 2026. The ECB and EBA confirm that they will follow this advice and that banks subject to reporting under ECB regulations and/or the EBA supervisory reporting framework will apply the new NACE Rev. 2.1 classification as of January 1, 2026. Therefore, what nomenclature should be applied for the P3 reports concerning the financial years closed on 30/06/2026? NACE 2.0 following the current templates 1 & 5, or NACE 2.1?

10. Do you have any views with regards to NACE code K – Telecommunication, computer programming, consulting, computing infrastructure and other information service activities, and in particular K 63 - Computing infrastructure, data processing, hosting and other information service activities, whether these sectors should be rather allocated in the template under section Exposures towards sectors that highly contribute to climate change?

We understand the reallocation of NACE code K. Given the growing reliance on artificial intelligence, cloud computing, and data infrastructure services, this sector has increasing potential to contribute significantly to carbon emissions, primarily due to the intensive computational requirements and the continuous operation of large-scale data centres. Since the global ICT sector, including NACE K-63 activities, is estimated to contribute heavily to global C02 emissions, it already significantly contributes to climate change, especially as digital demand grows (cloud, AI, streaming, etc.). However, we would also like to point out that at this stage there would be difficulties to identify the activity and its energy source.

We believe majority of the GHG emissions in this sector derive from used energy mix. The energy producers already report their GHG emissions and the sustainability transition should be directed towards energy production from renewable sources. Allocation of K-sector under Exposures towards sectors that highly contribute to climate change would be inconsistent and would entail a risk of double counting. Thus, we suggest NOT to allocate the K-sector under Exposures towards sectors that highly contribute to climate change.

We also recommend “greying out” GHG financed emissions disclosures (i.e. columns i to k) for the Total line (Row 56), since the requirement for reporting of GHG financed emissions are only relevant for those sectors that highly contribute to climate change (i.e. rows 52 to 55 in the columns i to k are also greyed out).

We also believe D35.12 production of electricity from renewable sources should be removed from 35.1 considering this is not a sector that a contributes to the climate change.

11. Do you have any comments on the inclusion of row “Coverage of portfolio with use of proxies (according to PCAF)”?

Column K and the new row (row 57) capture overlapping information regarding the share of financed emissions by data type, rendering them duplicative. Row 57 better accommodates mixed data scenarios—such as company-specific data for Scopes 1 and 2 combined with proxy data for Scope 3—and thus provides a more flexible and precise reporting framework.

Accordingly, we propose retaining row 57 and deleting column K. Row 57 facilitates clear, scope-specific reporting and is simpler to calculate, while column K is more complex and offers limited added value. Removing column K will streamline the template without sacrificing essential information, aligning with the rationale detailed under question 12 (“Financed emissions”).

There is also a new sentence added in the narrative: "every non-reported emission that is not based on a report from a counterparty, is by definition a proxy". Does it mean that CRE/RRE with valid EPCs will no longer be treated as "customer reported data"?

Finally, it would be helpful if the EBA could provide further clarification of the definition and scope of the term “proxies”, given the PCAF methodology uses the term “estimates” instead. In particular, it would be helpful if the EBA could clarify whether “proxies” here is intended to align with the PCAF term “estimates” or a broader concept of “estimates”.

12. Do you have any further comments on Template 1?

Regarding PAB exclusion

There are issues in reference to PAB exclusions under benchmark regulation. One of the issues is that the exposure to undertaking not complying with the DNSH is not an available data (even by the data providers). Article 12.2 from Delegated Regulation (UE) 2020-1818 indicates that Administrators of EU Paris-aligned Benchmarks shall exclude from those benchmarks any companies that are found or estimated by them or by external data providers to significantly harm one or more of the environmental objectives. The information that an undertaking would be detrimental to an environmental objective is not available in the URD or in the data made available by the data providers.
While we welcome the willingness of EBA to simplify as much as possible the reading of the column, we recommend adding precision regarding the limited scope of the definition of PAB exclusion for Pillar 3 ESG purpose in the column title.
We seek clarification on the rationale behind the inclusion of ETH Zurich classification, as it appears to duplicate the existing framework concerning PAB excluded exposures.
As a general comment, we welcome the introduction of materiality approach on scope 3.
We suggest the instructions to be updated with a non-exhaustive list of NACE codes (rev 2.1) of sectors that fall into this category. Considering the EBF response “Simplification of the EU Sustainable Finance Framework – Omnibus Legislative Proposal”, dated 21.1.2025, page 21, we suggest the following 4-digit NACE codes:

(d) companies that derive 1 % or more of their revenues from exploration, mining, extraction, distribution or refining of hard coal and lignite:

05.10

Mining of hard coal

05.20

Mining of lignite

19.10

Manufacture of coke oven products

(e) companies that derive 10 % or more of their revenues from the exploration, extraction, distribution or refining of oil fuels:

06.10

Extraction of crude petroleum

09.10

Support activities for petroleum and natural gas extraction

19.20

Manufacture of refined petroleum products and fossil fuel products

46.81

Wholesale of solid, liquid and gaseous fuels and related products

47.30

Retail sale of automotive fuel

49.50

Transport via pipeline

(f) companies that derive 50 % or more of their revenues from the exploration, extraction, manufacturing or distribution of gaseous fuels:

06.20

Extraction of natural gas

09.10

Support activities for petroleum and natural gas extraction

20.11

Manufacture of industrial gases

35.21

Manufacture of gas

35.22

Distribution of gaseous fuels through mains

35.23

Trade of gas through mains

35.24

Storage of gas as part of network supply services

46.81

Wholesale of solid, liquid and gaseous fuels and related products

49.50

Transport via pipeline

(g) companies that derive 50 % or more of their revenues from electricity generation with a GHG intensity of more than 100 g CO2 e/kWh:

35.11

Production of electricity from non-renewable sources

Financed emissions

As a general remark, we would like to note that GHG emissions data are duplicative of CSRD. Portfolio-level financed emissions do not equate to credit risk exposure so could be misleading in the P3 context. These data also face many availability/quality issues as very small share of companies are regulatory obliged to report GHG emissions.
For a counterparty, if in the calculation of funded GHG emissions we use both actual data (on scope 1 and 2 for example) and estimated data (PCAF for scope 3 for example), then how to calculate column K "percentage of the gross book value of the portfolio according to company-specific statements"?
- => CASE 1: are the data considered estimated once we use PCAF data for 1 of the 3 scopes?
- => CASE 2: are data considered real when we use data directly from our counterparts for 1 of the 3 scopes?
- => CASE 3: should the gross book value of the portfolio be prorated according to the company’s own statements?
As outlined in our answer to Q11, column K is a duplicative of line 57 and should be deleted.
It seems that requirements rely only partially on the PCAF Standard (unlike CSRD that refers to PCAF explicitly as the norm to be followed). We have noticed the following discrepancies:
- If the EBA intends to diverge from the PCAF Standard, we suggest that it explicitly outlines and justifies these deviations, in order to ensure clarity, comparability, and alignment with existing reporting frameworks such as CSRD.
- "Exposures towards the counterparty compared to the total liabilities (accounting liabilities and shareholders’ equity) of the counterparty." PCAF requires differentiated approaches.
- If Scope 2 emissions are to be reported, we suggest an update to the instructions so it is clear which Scope 2 emissions are to be considered when reported by clients: Market or Location based?

Taxonomy

In Column C, if not deleted, institutions are required to state the share of their exposures that can be classified as environmentally sustainable according to the Taxonomy Regulation. A significant portion of institutions required to report under Template 1 is not expected to be subject to the Taxonomy Regulation any longer (post-Omnibus), we suggest that the template and accompanying guidance clarify that this column is only to be filled in if the institution is subject to the Taxonomy Regulation. This would align with EBA’s approach in Templates 6-10.
In addition Column C in Template 1 and 4 should not be required during the GAR disclosure requirements suspension until end-2027 for large entities.

Narrative information

“2. Institutions shall include in the narrative accompanying the template, explanations on the information disclosed and the changes compared to previous disclosure periods, as well as any implications that those exposures may have in terms of credit, market, operational, reputational and liquidity risks for the institutions.”

Since this is qualitative information, we believe that such descriptions should already be included in Table 1A and should not be requested in this section as well.

13. Do you have any comments or alternative suggestions for SNCIs and other institutions that are not listed, regarding the sector breakdown?

There appears to be inconsistency between Templates 1, 1A and 3 regarding which sectors are considered “fossil fuel sectors”. For example, sectors B09.1, G46.81, and G47.3 are specified in Template 1, but not in Template 1A. We recommend that EBA ensures a consistent definition across all templates.

14. Do you have any additional suggestions how to adjust Template 1A for SNCIs and other institutions that are not listed?

The main purpose of specifying certain sectors at NACE levels 2 and 3 is to gather information on institutions’ exposure to the fossil fuel sector. However, template 1A in its current form also requires reporting of physical risks at this granular level. To align requirements for large institutions, we suggest removing or greying out the physical risk cells for NACE level 2 and 3 sectors — specifically the area between rows 7-11 and 20 and columns b-g.

Since Template 1A targets smaller institutions, which typically operate in limited geographical areas, a granular geographical breakdown combined with sectoral breakdown may result in data that can be traced back to individual companies. Therefore, we propose that Template 1A should not require a breakdown into five geographical areas (columns c-g), and that only the total exposure to physical risk in column b be required. Columns c-g should thus be deleted.

The lighter templates for these entities deviate from the templates planned for large institutions. Given that small and other institutions, as well as large subsidiaries, are included in the scope of consolidation of the bank, the information is already available at the terminals of these institutions. It therefore appears more complex, in terms of configuration and operational management, to produce a light report than a complete report, particularly in terms of configuration or access to information. It should be therefore specified that the light publication of the Pillar 3 ESG templates is a minimum obligation and allow institutions to publish a full Pillar 3, regardless of their size.

15. Do you have any further comments on Template 1A?

Following a materiality-based approach, the line “of which: part of a cover pool of covered bonds” in Template 2 could instead be reflected as part of the qualitative disclosures accompanying the templates

16. Should Template 2 in addition include separate information on EPC labels estimated and about the share of EPC labels that can be estimated?

While the share of estimated EPC labels is already requested in other exercises such as in the ECB’s Short-Term Exercise on Climate Risk (Template 2), we do not recommend introducing a requirement to estimate EPC labels in the context of Pillar 3 disclosures. This would place a disproportionate burden on institutions that are not already subject to such a requirement and would be inconsistent with the SREP, which is based on actual EPC data.

The lack of available data, along with the lack of standardisation criteria for these disclosures, would make producing reliable, comparable or meaningful disclosures near-impossible. We also note that the BCBS does not mandate the use of EPC labels. Instead, it emphasizes the importance of providing information related to energy efficiency.

We note that in the final EBA GL on ESG RM the reference on EPC label has been removed (Art 81 (f) for energy efficiency (A breakdown of portfolios secured by real estate according to the level of energy efficiency of the collateral).

In the same vein, interoperability should be applied more systematically across all templates for large entities, including within the qualitative templates.

17. Should rows 2, 3 and 4 and 7, 8 and 9 for the EP score continue to include estimates or should it only include actual information on energy consumption, akin to the same rows for EPC labels?

See Q16.

18. Do you have any comments on the inclusion of information on covered bonds?

We do not agree with the inclusion of information on covered bonds and recommend removing it, i.e. removing rows 1.1 and 6.1 from this Template 2.

Covered bond pools are subject to change and complex to track. Including information on covered bonds would simplify reporting; rather, it would imply additional burden for large institutions without clear added value. Moreover, the information provided by the covered bond issuer and the covered bond purchaser may be duplicated and therefore appear in the Pillar 3 disclosure of two different entities.

19. Do you have any comments on the breakdown included in columns b to g on the levels of energy performance?

We would appreciate confirmation that when al collaterals have measured or estimated EP Score, 0 should be disclosed in column G2.

20. Do you have any further comments on Template 2?

There are differences between EU countries in EPC label calculation methodologies, especially regarding the label thresholds. In some EU countries a building with Heating energy demand of 74,55 kWh/m2a fits into class D whereas in others it fits into class C. Thus, we suggest deleting columns h) to o) or at least postpone the reporting of EPC label until all countries adopt the renewed EPBD directive and the calculation of EPC label is based on primary energy demand and the class thresholds are unified within EU.

21. Do you have any comments on Template 3?

This information is already published in the annual report which we believe is better placed. Portfolio decarbonization targets set for alignment purposes (e.g., with NZE 2050) may be misleading for assessing banks’ exposure to climate risk, as impact is not equivalent to risk. While we understand the relevance of this information under CSRD, it should not fall within the Pillar 3/prudential perimeter.
Moreover, for large institutions and their subsidiaries, transition targets are typically set at Group level. Translating these Group-level targets into meaningful, operationalized targets at the local level is complex—particularly when trying to align them with the structure and expectations of Template 3. This creates implementation challenges and risks inconsistent or unclear disclosures. We therefore suggest removal of Template 3 for all institutions.
If not removed:
- the scope of the template should be adjusted to only include exposures to companies required to disclose under the CSRD – with a voluntary option to make use of estimates or real data for exposures towards non-CSRD companies;
- the frequency of template 3 should be adjusted on a yearly basis for large institutions;
- we would appreciate clarification on the term "targets" — whether it refers to firm commitments or to indicative metrics;
- we would propose removing columns h and i, as the 2030 targets will soon be outdated, and maintaining only the columns for additional targets (j–l), with the target year specified in the accompanying narrative - this would future-proof the ITS and reduce the need for frequent revisions;
- we suggest that a consistent application of the materiality approach — as defined in the EBA ESG risk management guidelines — be applied to this template.
We also have comments in relation to the methodology:
- The requirements to use the latest available IEA NZE scenario i.e. an annual update after the relevant WEO is out and is not in line with banks NZBA targets that are fixed over the observation period.
- For information related to exposure, we suggest giving the possibility to describe what has been considered for target setting (e.g. on balance sheet lending, investments, ...) instead of requiring the mandatory use of GCA as metric and the mandatory inclusion of debt and equity instruments.
- As per CRR3 banks have to test their alignment towards EU Objectives for EU exposures but also to third countries objectives for non-EU exposures. However IEA NZE scenario does not provide third countries pathways.
- We would suggest deleting column Distance to IEA NZE scenario. Additionally, the IEA NZE scenario isn’t suitable for every sector. For instance, in the shipping industry, the Poseidon Principles provide more appropriate benchmarks, offering decarbonisation trajectories tailored to specific ship types and sizes. If retained, we recommend EBA to adopt a flexible approach by allowing institutions to select the most suitable sector-appropriate decarbonisation scenarios and selection of targets in absolute (CO2e/y) or relative values (CO2e/physical unit) in column d.
- Where template 3 is only populated where there are group level targets, extending this approach to look at all 18 sub-industries will require significant uplift and may not be consistent with the approach at the group level. Additionally, the materiality outcome may differ at the group level. We recommend providing flexibility in reporting to consider these challenges and adopting a comply or explain approach.
- Banks are requested to report sectoral "targets" in the 3 years after the year of reference. This could be inconsistent with the actual time horizons retained by Banks when setting their alignment targets.
- The use of proceeds approach should be eliminated, as it introduces significant operational complexity and is nearly unfeasible to implement effectively in practice.

22. Do you have any comments with the proposals on Template 4 and the instructions?

We recommend this template be removed. Basel Committee Pillar III ESG is based on EBA’s but has specifically removed top 20 most pollutants from their framework. In fact, this template requires disclosure of the top 20 carbon-emitting companies in the world. However, as explained in our feedback on template 1, exposures to financed emissions (which would here include exposures to specific “top” carbon-emitting companies financed by the institution) do not directly correlate to an institution’s credit risk and therefore are not relevant for Pillar 3 purposes. Including a breakdown of exposures to the top 20 polluters may lead to redundancy without significantly improving the overall insight provided.

Shall it be impossible to remove the template altogether, we would suggest:

the ITS to specify a series of sources (e.g., Carbon Majors) to ensure comparability between banks, a yearly frequency for the reporting of the template and especially the column on alignment (cfr. answer to Q4).

23. Do you have any views on whether this template could be improved with some more granular information in the rows, by requesting e.g. split by sector of counterparty or other?

No, we do not consider additional information is needed.

The cost-benefit for any additional breakdown is not justified. We suggest removing this template or at least reduce its frequency to annual.

24. Do you have any further comments on Template 4?

There is still prevailing comparability issue when institutions use different source for TOP20 carbon-intensive firms. We would suggest the ITS specify a series of sources – such as the Carbon Majors database – for identifying the top 20 carbon-emitting companies, to ensure comparability between institutions’ disclosures.

25. Do you have any comments on the proposal using NUTS level 3 breakdown for Large institutions and NUTS level 2 for Other listed institutions and Large subsidiaries? Would NUTS level 2 breakdown be sufficient for Large institutions as well?

Regarding the breakdown for the top 10 NUTS level 3 regions, we consider that the proposed change contradicts the main objective of the consultation, which is to simplify the reporting process for institutions. The suggested granularity at the NUTS level 3 is, in our view, excessively detailed (could lead to disclose 12 templates only for physical risks) and may even compromise the privacy of clients in smaller countries.

For extended portfolios the requirement to disclose the first 10 NUTS-3 might cover a percentage of the bank portfolio that is not material. In light of this, the benefits of implementing the proposed amendment would not justify the effort required by the bank.

For a banking group, it is complex and impractical to obtain consistent information from each individual legal entity, particularly those located outside national borders. Furthermore, the requirement to map each production facility of every company to which the bank is exposed—down to the regional level—would entail significantly higher operational costs. Constructing and maintaining such a detailed database would be disproportionately burdensome.

Further, as location-specific disclosures are encouraged but not mandated at NUTS3 level under CSRD and ESRS, we recommend the same approach for Pillar 3. A voluntary approach provides institutions the flexibility to report at the level of detail that reflects their risk exposure meaningfully and ultimately be more useful to those relying on it.

26. Do you have any comments on the instructions for the accompanying narrative and on whether they are comprehensive and clear?

No comments.

27. Do you have any further comments on Template 5 and on its simplified version Template 5A?

For a banking group, it is complex and often impractical to obtain consistent physical risk data from each individual legal entity, especially those operating outside national borders. In particular, mapping each production facility to the regional (NUTS) level would entail significant operational burden and require databases that are not readily available or reliable.

For smaller or listed non-SIFI institutions with relatively limited geographical footprints, requiring a granular geographic and sectoral breakdown could raise confidentiality risks, potentially leading to the indirect identification of individual clients. We therefore propose clarifying that such institutions may omit the breakdown into five geographical areas (z-axis) and instead report only total exposures in Templates 5 and 5A.

Additional simplification options include:

Limiting NUTS templates to Top 3 regions, especially for institutions operating in a few regions;
Allowing summation of sectoral exposures to preserve confidentiality; or
Reducing the number of the tables for Template 5 from twelve to three: total exposures, EU exposures, and an aggregated Top 10 NUTS 3 table, in line with the objective of simplification and the need for comparability across institutions.
Removing references to “acute” and “chronic” events in the Template 5 table heading, unless the EBA intends to align with the Taxonomy and ESG Guidelines where such terminology is used. If convergence is not envisaged, the terminology should be excluded to avoid confusion.
Removing the requirement to disclose impairments related to physical risk, as this could mislead stakeholders by implying a direct link between impairments and climate hazards that may not exist.

It should be clarified that the total of exposures subject to physical risk in column G is not expected to equal the sum of columns H to K, as the same asset can be exposed to multiple climate hazards.

Finally, we propose aligning sector classifications across templates. Based on EBA expert and industry feedback, the sector “I – Accommodation and Food Service Activities” should be included under “other sector” in Template 1, as it is no longer considered highly climate-contributing. The same logic applies for sector “K – Telecommunication, computer programming, consulting, computing infrastructure, and other information service activities” where the categorization (classification under “other sectors” or “sectors that highly contribute to climate change”), as we understand, will be based on the EBA’s assessment of answers under Question 10 above.

28. Do you have any comments on the proposal to fully align templates on the GAR, that is, templates 7 and 8, with those under the Taxonomy delegated act by replacing the templates with a direct cross reference to the delegated act?

We request the EBA to reconsider the inclusion of the GAR and Banking Book Taxonomy Alignment Ratio (BTAR) within the Pillar 3 disclosures and permanently remove Taxonomy-related data requirements from Pillar 3 disclosures in order to avoid unintended consequences.

Being developed as a sustainability alignment metric rather than a measure of financial or prudential risk, GAR is not a risk-based KPI and the templates 6, 7 & 8 are a duplication of CSRD disclosures. In accordance with article 24 of the consultation paper, disclosure requirements already established in other applicable union law is to be avoided, however disclosing templates 6, 7 & 8 with P3 would duplicate information that is already required to be disclosed under CSRD and not aligned with the aim to reduce reporting burden.

P3 ESG disclosure is focused on providing risk information on ESG exposures however the templates 6, 7 & 8 disclose on financial information and do not include any risk information.

The purpose of the templates 6, 7 & 8 is therefore not aligned with P3 ESG disclosure. In line with Omnibus simplification efforts and the principle to disclose the information only once, the GAR should be only reported in the CSRD sustainability disclosures under Article 8 DDA.

We request further clarification on our interpretation in respect of template 8 that only refers to replicating the Template 4 of EU Taxonomy to reflect “GAR KPI flow”, on that basis we would not be required to replicate template 3 of EU taxonomy to reflect “GAR KPI stock”.

Should the EBA decide to retain the GAR templates within the Pillar 3 framework, it should limit it to Templates 6-8 only and clarify that the scope of alignment with CSRD sustainability disclosures is limited to on-balance sheet exposures as per the current ITS (the consultation wording suggests full alignment with the EU Taxonomy templates which might be interpreted as including off-balance sheet items, such as financial guarantees and AuM). A more explicit wording that a cross reference to what is already published for taxonomy regulation purpose might be done as it could be misleading if we have to “copy/paste” for P3 uses the template GAR of taxonomy regulation or if a simple narrative explaining where to find the template GAR is sufficient.

29. Do you have any comments on the proposal related the BTAR and to keep it voluntary?

With respect to the voluntary BTAR disclosure in Template 9, we remain of the view that this ratio will offer little decision making and market relevance.

While article 82 of the consultation paper states that BTAR information continues to be disclosed on a voluntary basis, the ITS was adjusted and now use the term “shall” instead of “may choose”. This should be revised back to “may choose” to preserve the voluntary nature as originally intended.

Consistency of the BTAR with the GAR templates is essential. While a first wave of amendments to EUT reporting expected to be finalized within Q2/Q3 2025 are already reflected in the draft ITS, a second wave of amendments is expected as a systematic and thorough review of the EUT article 8 disclosures by the EC is planned (timeline unknown). At this stage, it is unclear how the second wave of amendments will be reflected in the ITS to ensure consistence between the updated GAR and the BTAR. If such consistency cannot be achieved, it should be envisaged to delete the BTAR template.

In line with the Omnibus (draft delegated regulation amending Commission Delegated Regulation (EU) 2021/2178) proposal for EUT art 8 reporting, BTAR should explore the option of reporting on partial alignment.

We therefore propose to:

Amend draft ITS to ensure that BTAR disclosures remain voluntary.
Ensure that the latest upcoming GAR adjustments (timing unknown for the 2nd wave of adjustments) are correctly reflected in the BTAR template to ensure consistency. If such If such consistency cannot be achieved, it should be envisaged to delete the BTAR template.
Explore the option to report partial alignment under BTAR.

In alignment with the amendments on Article 7 of EU Delegated Regulation 2021/2178; in Template 9.1 derivatives, on-demand interbank loans, cash and cash-related assets, and other assets that are not referred to in Article 7 (6), including goodwill and commodities, shall be excluded from the denominator of BTAR. Those adjustment allow to have BTAR template fully consistent with the reviewed GAR templates in EU Taxonomy Delegated Acts (2021/2178). We would like to receive confirmation that this understanding is correct.

30. Do you have any comments regarding the adjustments to template 10?

We appreciate that starting from June 2025 and for subsequent publications up to the reference date of 31/12/2027, the publication of template T10 is suspended.

However, we note that Template 10 conflates exposures to counterparties whose activities are EU Taxonomy-aligned with assets contributing to sustainability and transition finance, resulting in an inaccurate reflection of exposures in these areas. For example, a sustainable loan to a counterparty with minimal EU Taxonomy alignment will have limited impact on an institution’s GAR while being fully excluded from Template 10. Given these fundamental issues and the lack of clarity on environmental objectives—where a sustainable loan may fall outside or span multiple objectives—we propose that Template 10 be removed altogether, as it does not provide meaningful or accurate information.

Should the decision be made to retain Template 10, we recommend T10 to be disclosed on a voluntary basis – the same as BTAR –, and implementing the following enhancements to improve clarity, consistency, and usability:

For “renovation loans,” we recommend aligning the definition with that of “building renovations” as outlined in the Social Climate Fund (Regulation EU).
It is also unclear what type of instruments should be included in Template 10 under the rows referring to equity exposures? Next to equity exposures, clarification is also needed for the treatment of transition finance. Any definition should be aligned with existing EU regulations and provide clear guidance on reporting expectations.
Furthermore, how can an equity exposure be classified as a 'green exposure'?
The revised ITS introduced ‘Of which: small and medium sized enterprises’ breakdown under loans and advances to non-financial corporations within Template 10. Such a breakdown is not part of the GAR and BTAR templates. To ensure simplification, as well as alignment between different templates, we recommend removing this breakdown
According to EBA Q&A 2023_6878, general purpose financing to pure-play companies can be reported under Template 10. We suggest that this clarification to be explicitly included in the revised ITS. In addition, to ensure consistent interpretation and application, a definition of pure player would be welcome.
As Template 10 focuses on assets contributing to sustainability and transition finance, it would be helpful to clarify whether and how financing provided to companies with credible transition plans that are on track to meet their targets should be reported within this template. This would contribute to the consistency of treatment and comparability on this matter. Please refer to ‘transition finance’ definition which touches upon this matter, on Commission Recommendation (EU) 2023/1425 of 27 June 2023 on facilitating finance for the transition to a sustainable economy.
In this Template, when covering the assets classified as green beyond the Taxonomy, might we also include lending/financing with Known Use-of-proceeds, including Specialized Lending, which have been classified just partially-Taxonomy-Aligned?
In the Instructions for disclosure of ESG risks, when defining the narrative and qualitative column h “Qualitative information on the nature of the mitigating action”, it is not quite clear what do you expect for “duration of the actions”. It would be useful to receive further information about this.

31. Do you have any further comments on the Consultation Paper Pillar 3 disclosures requirements on ESG risk?

We support EBA’s proposed proportionality approach, which distinguishes not only by size and complexity but also by whether the institution is listed. We encourage EBA to apply similar proportionality considerations in its other ESG areas, including risk management and scenario analysis.
The ESG-related tables and templates use generic names such as “Table 1” or “Template 1”, when all other Pillar 3 tables and templates use a different naming convention. Other qualitative tables under Pillar 3 are typically named “Table EU [Topic Abbreviation][Sequential Letter]”, and quantitative templates are named “Template EU [Topic Abbreviation][Sequential Number]”. We suggest applying the same naming approach to the ESG-related templates, so that the first qualitative table is named “Table EU ESGA” and the first quantitative template “Template EU ESG1”, and so on.
In paragraph 31 of the consultation paper, Table 1 implies that Large Institutions – regardless of Listed or Non-Listed classification - will disclose Article 449a templates 1, 2, 4, and 5 semi-annually. This would increase operational burden and is contra to the proportionality concept in Pillar 3. Clarity is requested on scope and frequency of disclosure for Large Non-Listed institutions.
We note that the numbering of consultation questions in section 5.3 does not match the numbering in the consultation paper itself.
We would appreciate a clarification as to what is expected to be delivered to FIN FSA as P3 ESG xbrl package as several templates are now discontinued.
We welcome the occasion of this consultation to ask for the review of the inclusion of investments in subsidiaries, joint ventures and associates within equity instruments in counterparty breakdown requested in Pillar III – ESG templates.

In fact, while we acknowledge that investments in subsidiaries, joint ventures and associates, as defined by Regulation (EU) 2021/637, exist within the scope of reporting, it is nevertheless misleading to include the mentioned positions within the counterparty breakdown as presented in the Pillar III ESG tables, as this approach is not aligned with other reporting requirements:

FinRep: these exposures are treated separately in an ad hoc portfolio (row 260 of FinRep template 1.1). The breakdown by counterparty category of investments in subsidiaries, joint ventures and associates are not available in FinRep and the breakdown requested by Pillar III would make impossible to reconcile data among the two reporting’s (e.g. NFC in Pillar template 1 would be by definition greater than NFC in FinRep breakdowns);
- final Delegated Act amending the EU Taxonomy disclosure regulation 2025/4568 adopted by the Commission on July 4 clearly includes in the counterpart breakdown of GAR templates only the IFRS 9 accounting portfolios (Financial assets at amortized cost, financial assets at fair value through other comprehensive income, investments in subsidiaries, joint ventures and associates, financial assets designated at fair value through profit or loss and non-trading financial assets mandatorily at fair value through profit or loss, and real estate collaterals obtained by credit institutions by taking possession in exchange in of cancellation of debts). In case this is not aligned, NFC in Pillar template 1 would differ from NFC in GAR template.

Can you please confirm the exclusion of those investments in subsidiaries, joint ventures and associates from Pillar Transitional/Physical risk templates thus discarding also Q&A 2023_6954 in order to guarantee full alignment between the perimeter in GAR template and Transitional/Physical risk templates?

34. Are the amended template EU CR 10.5 and the related instructions clear to the respondents? If no, please motivate your response.

After reading the instructions and guidelines, it is unclear whether non-listed SNCIs are also required to complete this template and from which reference date.

If the template includes only the “total” row, what kind of breakdown should banks provide? It would be helpful to better specify the “categories” expected in the template.

35. Do the respondents agree that the amended template EU CR 10.5 and the related instructions fit the purpose and meet the requirements set out in the underlying regulation?

For the sake of simplification, we would suggest amendments to the equity exposures template (Template EU CR10.5 Section 12 Annex I) applying as soon as possible. The starting reference date should be brought forward from 31/12/2026 to 31/12/2025.

36. Do the respondents consider that the “mapping tool” appropriately reflects the mapping of the quantitative disclosure templates with supervisory reporting templates?

EBA should establish a fast channel of communication for identified errors and inconsistencies and IT tools and rules for disclosure and reporting.
In several parts of the paragraph on SNCIs, there is exclusive reference to “Listed SNCIs” and to “other institutions” regarding templates CQ1, CQ3, CR1, and CQ7. We would welcome confirmation that non-listed SNCIs are exempted from the disclosure requirement, as Article 433b of the CRR does not appear to suggest this, indicating a potential discrepancy between the ITS and the regulation.

Credit quality of loans and advances to non-financial corporations by industry (template EU CQ5):

The same starting reference date as for Pillar 3 should be envisaged to ensure entities are not required to maintain two different NACE classifications depending on the purpose of the data
While the reporting consultation paper has not yet been published, it is not fully clear (considering the publication of the EBA Reporting framework 4.3) whether the starting date could be later than 31/12/2026, therefore misaligned with the P3 disclosure.

Attachment

EBF response to the EBA CP on ESG Pillar 3 Disclosure under CRR.docx (967.06 KB)

Name of the organization

European Banking Federation