Response to consultation on Guidelines on ESG scenario analysis
Question 1: Do you have any comments on the interplay between these Guidelines and the Guidelines on the management of ESG risks?
We are in favour of harmonising and integrating these Guidelines with the Guidelines on the management of ESG risks. The terminology should be consistent.
In principle, there are close links between the relevant regulatory documents. As a result, attention should be paid to the consistency of the statements, for example on access to proportionality-related simplifications for institutions of a certain size, and the terminology used, and redundancies in implementation should be avoided. In particular, there is a close link between the transmission channels mentioned in the draft and the materiality analysis in accordance with the EBA Guideline on the management of ESG Risks (EBA/GL/2025/01). In order to consider the required transmission channels as a whole, they should be explicitly taken into account in the upstream materiality analysis, as this forms the basis for scenario analyses. A clearer link between the directives would be helpful here.
In addition, it remains unclear to what extent there should be harmonisation with other European rulebooks, such as the Corporate Sustainability Reporting Directive and the topics listed there, particularly on transition planning. When harmonising the requirements, which is desirable in order to achieve synergies between different EU frameworks, it must also be ensured that, against the background of the Omnibus initiative and the resulting diminishing influence of CSRD, CSDDD and the taxonomy for certain sizes of institution, smaller institutions in particular do not have to fulfil requirements laid out in these EBA Guidelines, which they will probably be exempted from, for good reason, as part of the bureaucratic relief provided by the Omnibus. A clear definition of ESG risks with a distinct hierarchy and subcategories by the regulator in the glossary would be helpful to avoid any ambiguity.
Since this guideline is intended to support the building of capacities and knowledge, the question arises as to when they should be introduced. They could either be introduced too early – due to the ongoing implementation of the general guideline on the Management of ESG Risks (EBA/GL/2025/01) and the as yet open outcome of the Omnibus initiative – or too late, as full implementation of the requirements by the beginning of 2026 seems unrealistic due to EBA and NCAs processes, but also for the institutions.
As announced in the virtual public hearing on 17 March 2025, an appropriate implementation period is required. Since the consultation paper was only published on 16 January 2025, the consultation phase will run until 16 April 2025 and the final paper is not expected to be published before the third quarter of 2025, we believe that requiring the institutions to implement these guidelines by 11 January 2026 is too ambitious. The implementation deadline should be at least by the end of 2026.
Question 2: Do you have comments on the proposed definition of scenario analysis and its various uses as presented in Figure 1?
The draft guidelines offer methodical guidance on developing and applying ESG scenario analyses in risk management, with a focus on short to medium-term climate stress tests in line with ICAAP requirements. However, it should be noted that there is a strong focus here on specific use cases, e.g. the analysis of climate-related risks. At the same time, the draft guidance stipulates not only taking into account material risks in relation to climate but also including other ESG factors in scenario analyses, which would also mean environmental risks, social risks and governance risks in some cases. However, this is not yet sufficiently quantifiable. It is positive to note that the supervisory authority itself does not require detailed quantification and that potential influences are sufficient. This should be retained. SCNIs can initially also do this qualitatively.
The draft guidelines largely omit specific methodological guidance on other ESG factors. There is also no comprehensive representation of a risk driver analysis or materiality analysis of ESG risk drivers, although these results from the materiality analysis are generally the starting point for the scenario analysis.
Question 3: Do you have comments on the proposed distinction made between short-term scenario analysis (CST) and longer-term scenario analysis (CRA) as illustrated in Figure 3?
Differentiating between Climate Stress Test (CST) and Climate Resilience Analysis (CRA) makes sense. However, the required granularity and complexity of the analyses (CRA) must be reduced. Furthermore, the scenario outcomes are subject to a high degree of uncertainty. They should, therefore, serve merely as an indicator for possible long-term strategies and for the robustness of the business models.
The question also arises as to whether the same scenario may be used for CST and CRA – with different characteristics in terms of short and long-term effects. While qualitative analyses should suffice in principle for the CRA, this does not appear to be consistent with the description in Article 74 where quantitative analyses over a period of five years are also mentioned. In our view, the comments in paragraph 25 (“The general approach consists of a detailed analysis of their business environment …, on which risk-adjusted profitability analyses are drawn up for each area and/or activity…”) also indicate that qualitative analyses may not be sufficient.
Furthermore, it should be clarified what is to be understood by ‘alternative scenarios’. This might be referring to the usual NGFS scenarios, but possibly also to other scenarios.
Question 4: Do you have any comments on the interplay between these Guidelines and the Guidelines on institution’s stress testing?
We do not see sufficient need for a separate guideline on ESG scenario analyses. Our understanding is that this topic has already been covered by the existing EBA Guidelines.
It remains unclear whether ESG scenarios need to be fully integrated into the institution-wide stress test programme or whether they could also be taken into account additionally or in parallel. It is equally unclear whether a stress test that covers a number of different risk types is sufficient or whether the integration of specific types of risk is necessary.
Embedding in the ICAAP framework is only dealt with briefly in the draft. A stronger link to ESG guidelines with existing stress test and ICAAP processes would be desirable. An explicit representation of the interplay between ESG stress tests and macroeconomic stress tests could help identify and exploit existing synergies.
Question 5: Do you have comments on the Climate Scenario Analysis framework as illustrated in Figure 4?
The framework outlined in Figure 4 relating to the CST is based on known methods and processes, the fundamentals of which are clear. With regard to the CRA as a new instrument, however, a number of questions remain unanswered. The use of a central scenario is specified, but the implementation of the requirements in paragraph 69 et seq. for the formulation of such a scenario represents a considerable challenge due to its high level of complexity. In particular, the selection of parameters and taking uncertainties into account require further methodological considerations. Additional specific examples for defining key scenarios, including sector-specific challenges and macroeconomic assumptions, taking into account proportionality, could improve the practical applicability of the CRA.
The objective of the CST – the evaluation of the effects on total capital and liquidity – does not appear fully in line with envisaged management measures (e.g. encouraging counterparties to reduce risk), since different levels of granularity are used.
Furthermore, individual requirements for the definition of transmission channels and data collection are already comprehensively regulated through existing EBA Guidelines and ECB guides. Additional requirements do not, therefore, appear to be necessary.
Question 6: While respecting the definitions provided in other parts of the regulation, is there any concept/s used in these guidelines that it would be useful to include in an annexed glossary?
- Yes, a glossary would be very helpful.
Question 7: Do you have comments on section 4.1 Purpose and governance?
The draft guidelines describe governance on scenario analyses in rather general terms, make references to other EBA frameworks and generally recommend a cross-functional approach. However, paragraph 16, for example, remains vague with regard to the practical integration into existing management systems. There is a lack of concrete information about coordinating between strategies and operative departments, in particular between risk management and sustainability management.
There is also a lack of recommendations on governance structures, including responsibilities, escalation channels and processes for regularly reviewing the scenarios. A more precise definition of these aspects would ease implementation considerably and strengthen the effectiveness of the scenario analysis.
In addition, ‘forward-looking approaches’ should be specified more clearly with regard to the use of models, especially if they are based on historical data to determine future developments. There is also no clear link to other relevant EBA Guidelines in this area, which makes consistency and practical applicability more difficult.
It should be made clear that the results of the CST are not to be seen as an add-on to the normal RTF/ICAAP calculation, but that the results are to be considered in isolation from each other.
Reference to paragraph 41: “not to overinterpret”. As there are no direct capital requirements, we request clarification.
Question 8: Do you agree that the proposed proportionality approach is commensurate with both the maturity of the topic and the size, nature and complexity of the institution’ s activities?
The proposed proportionality approach is largely appropriate and takes into account the different sizes and complexities of the institutions. However, e.g. accessibility to size-dependent simplifications for the initial phase (paragraph 26) should also be made consistent with the EBA guideline on the management of ESG risks (i.e. implementation simplifications not only for SNCIs, but also for other non-complex institutions).
However, after taking into account the size, type and complexity of the institution’s business in the scenario analysis, a distinction should also be made between different scenarios and the frequency with which they are updated. While the EBA stress test guidelines (EBA GL/2018/04) and the ECB guidelines on the ICAAP and ILAAP (2018, Principle 7) provide for regular plausibility checks and, if necessary, adjustments to the scenarios (usually quarterly), climate scenarios often only contain annual data points (e.g. on natural risks and emissions). An annual review and any resulting necessary updating would therefore be more appropriate. The guidelines should clearly take these differences into account in the requirements for the frequency of updates (see paragraph 25 in Chapter 4). In addition, we do not believe the reverse stress tests should be in the scope of the analysis.
Question 9: Do you agree with the proposed references to organisations in paragraph 28? Would you suggest alternative or complementary references?
The references listed in paragraph 28 are helpful, however, they could be complemented by additional and in particular sector-specific sources in order to further increase the relevance and practicability of the CSA. For example, the comprehensive climate scenarios of the IPCC (Intergovernmental Panel on Climate Change), in particular the Representative Concentration Pathways (RCPs) and Shared Socioeconomic Pathways (SSPs), could be used to strengthen the scientific basis of the scenarios.
In addition, references could be complemented by other relevant pathways, such as the Net-Zero Banking Alliance (NZBA), the EU Green Finance System (EU GFS) and scenarios of the US Federal Reserve.
Universities and colleges could also be included as scientific sources, especially those that have published their own scenarios based on the scientific consensus on global climate change. One example of this might be a local university that has modelled specific flood scenarios for a certain region. Integrating such scientifically based and practicable scenarios could further increase the meaningfulness and adaptability of the CSA.
In general, we believe the references listed in paragraph 28 are helpful, however, it should be taken into account here that paragraph 27 sets out very comprehensive requirements. These requirements could, in our opinion, be understood as simply being examples because they allow for a great deal of room for interpretation. In some cases, there is, at least at the moment, a lack of sufficient scientific evidence for the design of these scenarios to make reliable assumptions (e.g. development of consumer preferences over a longer period of time, development of technological change, etc.). The text in paragraph 27 should therefore include the phrase, “where appropriate and possible”.
Question 10: Do you have additional comments on section 5.1 Setting climate scenarios?
The need to take into account the combined effects of macroeconomic shocks and climate shocks is questionable, particularly against the background of an explicit, isolated measuring of ESG risks. It would be helpful to more precisely concretise the combination of climate scenarios with other macroeconomic shocks – taking particular account of the proportionality approach. Taking into account the various transition pathways, as described in paragraph 35, also raises the question as to whether all these pathways need to be taken into account or whether it would suffice to use a selection, taking into account the proportionality approach and the relevance to the individual institution. A more detailed explanation of the consistency of scenarios, particularly with regard to the interplay between climate risks and macroeconomic factors and their long-term effects (paragraph 41), would also be desirable.
Paragraph 27 sets out the minimum requirements for scenarios, while paragraph 29 allows scenarios to be individually designed and adapted to the respective requirements. It remains unclear how these requirements can be harmonised. In addition, paragraph 31 contains a reference whose meaning and context should be further clarified.
Paragraphs 38, 39 (Chapter 5). In banking practice, the baseline scenario is used in particular in connection with the institution’s capital planning (see e.g. ECB Guideline ICAAP, 2018). In our view, this differs from the baseline scenario of a possible independent climate stress test mentioned here and should therefore be outlined more clearly in the consulted EBA guidelines. We recommend changing “with and without” to “with or without”, otherwise this would mean an additional scenario for institutions.
Paragraph 40 (Chapter 5) in conjunction with paragraph 15 (Chapter 4): In our opinion, the meaning of “central scenario” is not clearly defined. The institutions usually forecast up to five years in line with capital planning. We assume that external scenarios from the NGFS, for example, will be used for the CRA, which can be expanded to include individual components if necessary. In our view, independent development of long-term scenarios for the CRA by the institutes would be very costly.
It should also be noted that a number of requirements relating to factors to be taken into account (such as in paragraph 27) would already have to be considered in the scientific scenarios. Corresponding scientific scenarios (e.g. with geodata, modelling of second-round effects etc. with the required granularity) are often not yet available.
Question 11: Do you have comments on the description of the climate transmission channels?
The list of climate transmission channels in the draft guidelines is extremely detailed and very complex, and should not be understood as a kind of checklist. The multitude of micro and macroeconomic transmission channels pertaining to transition and physical risks will represent enormous amounts of work for institutions.
Here, it would be helpful if the final guidelines contained more specific notes on practical implementation and prioritisation of relevant transmission channels. This would help the institutions to concentrate on significant risks while preventing them from becoming overwhelmed. Here, too, it will be important to take proportionality into account; there should be no mandatory modelling of second-round effects for institutions, and further reductions in complexity for smaller institutions.
In order to take the various transmission channels into account, the financial institutions require more guidelines from organisations that create reference scenarios (such as the NGFS). There must be more granularity in these scenarios in regard to sectors and regions.
Paragraphs 46-48: The requirements appear to be much too detailed and specific, and there is therefore need for the specification “where possible and applicable”.
Question 12: Do you have comments on climate stress test (CST) tool and its use to test an institution’s financial resilience?
The requirements for climate stress tests and their integration into existing stress test programmes contain ambiguities at multiple points. In paragraph 56 (Chapter 6) it is noted that recession stress tests in accordance with adverse scenarios for capital planning for the ICAAP/ILAAP are usually carried out quarterly by the institutions. At the same time, however, paragraph 55 references independent climate stress tests specified by a yearly stress test programme. There is a need to clarify what is meant by “regularly” in this context, in particular for IRBA institutions. In addition, there is a need to specify whether the requirements for ESG integration in (regular) adverse ICAAP scenarios pertain to other stress tests or to stand-alone climate stress tests. In general, there is a need to review whether or not integration is strictly necessary (paragraph 57), or whether, for reasons of proportionality and the fact that not all data is currently available, these factors can be dealt with separately. As already explained, the integration of different time horizons and risk drivers into existing processes does not currently seem possible; even a step-by-step integration based on expert judgement is considered extremely challenging.
The gap analysis of existing models described in paragraph 57 represents a challenge for the institutions. The exact scope and practical implementation of these requirements are currently not adequately specified. This will require a change in methodological reasoning, as climate risks are very difficult to integrate into existing models without historical data and with the requirement to use a “forward-looking” approach. In addition, there is no specific information on practical adjustment of the models, which further complicates execution and implementation.
In paragraph 58 it is unclear why the guidelines include recommendations for IT infrastructure (in this case as a sort of sandbox for testing new developments). Decisions regarding IT architecture should be left up to the banks. It should go without saying that they have a vested interest in ensuring that untested software components are not simply integrated into their technical infrastructure, and will follow existing regulations pertaining to software.
Question 13: Do you have comments on the Climate Resilience Analysis (CRA) tool and its use to challenge an institution’s business model resilience?
The Climate Resilience Analysis (CRA) requirements are very granular and present many institutions with significant conceptual and technical challenges. In the case of evaluation of a CRA result based on a time horizon of over 10 years, there is a need to take into account the fact that the long-term nature of the analysis will result in significant uncertainties in the result.
It is unclear to what extent the CRA is in agreement with other European regulatory frameworks such as the CSRD. Parallel implementation of these analyses could reduce the amount of effort required and help avoid redundancies. If these processes were more closely interconnected, institutions would be able to perform a more consistent analysis of their climate resilience and develop targeted adaptation strategies. Weakening CSRD requirements could mean that less data is available for use in CRA implementation, leading to increased use of proxy data.
In addition, paragraph 59 contains wording in need of clarification: the first sentence references model segmentation into sectors and countries, while the second sentence references developing new models or extending the granularity of existing models. These two issues should be clearly separated from one another. In paragraph 81, there is still a need for clarity as to what, exactly, is meant by “strategy”.
The need for banks to implement an additional CRA with a dynamic balance sheet over a time horizon of at least ten years is questionable. It remains unclear how any results arising from this uncertainty could serve as solid proof of long-term strategies and resilience in business models.
There is also a need to clarify that qualitative analyses for periods of more than five years are adequate. Modelling macroeconomic dependencies and second-round effects should be based on scientifically sound source scenarios.
Question 14: Do you have any additional comments on the draft Guidelines on ESG Scenario Analysis?
The draft is an important starting point, but unfortunately includes significant challenges for the institutions. On the one hand, the requirements are very detailed, in particular as pertaining to the transmission channels. On the other hand, there is an astounding degree of freedom, requiring institutions to create comprehensive plans for developing, implementing and embedding the CRA within their processes. The combination of these two factors means that the effort required to adapt to these guidelines will be significant. A clear delineation between practical implementation approaches and synergies with other regulatory frameworks could simplify the process.
In addition, the question as to the application date should be clarified further: we do not believe application is realistic before the end of 2026 at the earliest. Here, supervisory bodies should take into account that the institutions will have even less time than it seems, considering that the consultation process is ongoing and, after it is complete, there will be subsequent processes for national implementation (similar to those for the EBA guidelines on managing ESG risks, EAB/GL/2025/01). This makes it unlikely for them to be able to achieve implementation by the deadline at all, as institutions will have their own development and implementation work to complete to meet the requirements.
Full compliance with all requirements by January 2026 is practically impossible. Implementation should not take place before the end of 2026.
The principles of proportionality and consistency must be obeyed more strongly within the guidelines. It is important that there is recognition that there are high degrees of uncertainty in CSTs performed without adequate data history. ESG risks are not a new type of risk, and as such, additional stress tests for specific risk drivers should serve rather as sensitivity analyses as a means of supporting the validity of existing stress tests. The focus of the CST approach should be on this application.
In addition, it is important to clarify that – until further notice and in particular until sound scientific evidence is available and has been integrated into the scenarios of scenario providers – no other environmental factors apart from those related to climate are to be subject to detailed analysis.
Supervisory bodies should also review the guidelines specifically to examine their compatibility with the EU omnibus initiative. Some data sources may fall away (e.g. due to expected reduction of CSRD, taxonomy etc. application), meaning that the ability of the institutions to gather relevant information from their clients (e.g. transition plans) and use it for their own risk management will also be reduced. In light of future regulatory changes, therefore, it is important to guarantee that the requirements (even after the Omnibus initiative is implemented) can still be implemented in a meaningful manner. They must also, given the effort required, remain adequately informative and not negate the fact that bureaucracy has been reduced by increasing the bureaucratic workload in another area.