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?
According to the Guidelines on the management of ESG risks, materiality assessment of ESG risks should be performed every year (every 2 years for small and non-complex credit institutions (SNCI). Therefore, the European Banking Authority (EBA) should specify in these Guidelines that every ESG risk driver identified as material should be assessed in the Climate Scenario Analysis (CSA).
Secondly, the EBA mentions reverse stress testing in its Guidelines on the management of ESG risks and not in these Draft Guidelines. We suggest adding a reverse stress test in the CSA framework, as it can be a powerful tool to identify less evident risk vulnerabilities in banks’ portfolios.
Question 2: Do you have comments on the proposed definition of scenario analysis and its various uses as presented in Figure 1?
Given the limitations of the current scenario analysis, scenario analysis will be of limited use to achieve certain objectives outlined in Figure 1 (in particular, checking capital and liquidity adequacy), adaptations to the current approaches will be needed to reach meaningful conclusions. See our responses to subsequent questions for further details and proposals.
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?
Finance Watch welcomes the forward-looking and longer-term approach adopted by the EBA. We understand the distinction between a quantitative short-term approach focusing on financial impact with fewer modelling uncertainties with a qualitative longer-term approach focusing on robustness and compatibility with transition policies. However, it is essential to stress that both approaches should be complementary. Therefore, scenario narratives used in CRA should be “an extension” of/consistent with those used in CST. This will allow financial institutions to capture long-term risk resulting from the chosen short-term scenarios and vice versa.
Moreover, tipping points and chronic physical risks, which are irreversible and long-term effects of global warming, depend on current decisions (a late transition with a higher increase of temperature will increase the probability of reaching a tipping point for example). Hence, the EBA should extend the time horizon for CST to better capture climate change impacts. Otherwise, CRA time horizon should be aligned with the institution’s transition plan and be extended to 2050 (in line with the EU climate objectives).
Moreover, we warn against the exclusive use of the dynamic balance sheet assumption. The major issues with this assumption include the risk of downplaying future vulnerabilities, the speculative nature of judgements about future balance sheet developments and constraints in implementing the intended adjustments at the macroeconomic level (risk of herding behaviours and abrupt market corrections if many institutions assume similar derisking behaviours in the future). Furthermore, the dynamic balance sheet approach neglects by construction the second round and the disruptive effects of balance sheet adjustments on the economy, whereas those effects are the ones which will have the higher impact on the economy and the financial system.
Question 4: Do you have any comments on the interplay between these Guidelines and the Guidelines on institution’s stress testing?
The EBA should clarify the interaction between climate scenarios and stress tests. In the Guidelines on institution’s stress testing, scenario analysis is considered as a type of stress test and scenarios are built always assuming a certain degree of severity/stress. This is inconsistent with the CSA approach of these Draft Guidelines, where climate scenarios are thought outside the stress testing framework and stress tests are viewed as sub-type of climate scenario analysis. In order to ensure sound and consistent practices of financial institutions, the EBA should clarify the definition of CSA by aligning it to stress test Guidelines scenario characteristics which are :
- aligned in an internally consistent way;
- the risk factors forming the relevant set presuppose the simultaneous occurrence of forward-looking events covering a range of risks and business areas;
- the set of risk factors also aim to reveal, to the maximum extent possible, the nature of linked risks across portfolios and across time, system-wide interactions and feedback effects.
Referring to the above definition, adverse climate scenarios should embed stress events on all material risks simultaneously.
Moreover, the EBA should clarify if the CRA exercises should fulfil the stress testing Guidelines too, in particular given the upcoming work of the ESAs on the mandate to deliver Joint guidelines on methodologies for the stress testing of environmental, social and governance risks as per Art.100(4) CRD.
Question 5: Do you have comments on the Climate Scenario Analysis framework as illustrated in Figure 4?
See Q3 regarding comments on time horizon and balance sheet assumptions.
See Q4 regarding comments on the need to clarify scenario definition.
See Q6 regarding comments on the need to clarify other definitions.
Finance Watch advises adding a step in the process about “impacts adjustments”. As drafted in the Guidelines and recommended by the NGFS, institutions need to adjust their adverse scenarios by adding “impact adjustments” to their computation of climate impacts in scenarios. When adverse scenarios are incomplete, adjustments will be made to address limitations such as time horizons, gaps/flaws in economic modelling, lack of data, missing tipping points, among others. As sizing adjustments will face similar issues when applying a quantitative approach, these adjustments should be determined using a qualitative approach and expert judgment of bank activities and actual scenario shortcomings. The approach and assumptions should be made transparent to facilitate regulators’ understanding of the adjustment.
To emphasize the transparency point, Finance Watch urges the EBA to insist on the importance of transparency and disclosure of the CSA exercise (transmission channels, use of data sources, proxies, limitations, adjustments).
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?
These Guidelines should clearly define the expected levels of granularity of scenarios (low vs high). The EBA should specify whether it pertains to the counterparty, sector, geographical area, or other criteria.
Moreover, Finance Watch warns the EBA against the use of different denominations between CST and CRA scenarios (ie baseline/adverse vs central/alternative). CSA exercise needs continuity between short-term and long-term scenarios. Therefore we suggest using the terminology of baseline and adverse scenarios in CRA too.
Question 7: Do you have comments on section 4.1 Purpose and governance?
Finance Watch welcomes the transparency obligation mentioned by the EBA on expectations and limitations of scenario analysis conducted by the institutions. Complexity and nascent stage of CSA exercise lead to a need for close cooperation among all stakeholders. Therefore, the EBA has to ensure that scenarios used by financial institutions are consistent. We suggest that the EBA proposes several common trajectories based on IPCC, IEA or NGFS work which will drive institutions' work on scenarios. In that context, institutions could still add further elements to build more suitable scenarios to embed idiosyncratic risks and their expertise while ensuring comparability of the results across institutions. Based on their qualitative and expert approach, institutions will need to report all assumptions and shortcomings of their scenarios. This idea is discussed in paragraph 35, but it needs to be mandatory and pathways need to be more precise to avoid inconsistencies in institutions' modelling, which would be warranted. Regarding the proposed proportionate approach for small and less complex institutions, SNCI could use simplified scenarios (sector-reduced or region-reduced) but consistent with common pathways as mentioned above.
Finance Watch welcomes the EBA decision to enhance consistency, robustness and include requirements on the review of scenarios across the business functions of an institution.
Finally, concerning paragraph 16, Finance Watch emphasizes that there cannot be a resilient business model without financial resilience as well. Therefore, we reiterate the importance of consistency between CRA and CST.
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?
Proportionality approach shouldn’t be done at the expense of risk mitigation. Although SNCI absolute exposure may be smaller and less complex, their sensitivity to climate risk could be material. For example, their exposure can be concentrated on one sector or one geographical area, which could increase their sensitivity to a specific climate risk. As SNCI will need smaller and less complex scenarios to cover their exposure, proportionality is already inherent to the CSA exercise. A financial institution with fewer and less diversified counterparties across sectors and regions will only need to build climate scenarios covering these areas.
Question 9: Do you agree with the proposed references to organisations in paragraph 28? Would you suggest alternative or complementary references?
As discussed in Question 7, reliable, realistic common scenarios are a precondition for the success of CSA. Besides NGFS and EU JRC, institutions should refer to the work of the IEA and the IPCC.
Question 10: Do you have additional comments on section 5.1 Setting climate scenarios?
Finance Watch agrees with the elements for climate scenario construction listed by the EBA. In addition, nature loss events such as deforestation and biodiversity losses should be taken into account when assessing intertwined factors as they may influence future policies or consumer preferences, for example. Then, institutions should use science-based, internationally defined climate scenarios such as the one listed in Q9 as a starting point.
As pointed out by the EBA (Paragraph 9), the time horizons of the existing risk management framework (1 to 3 years) and ESG-related risks (going over decades) do not match. This is a critical point as it can lead to huge risk underestimation in short-term scenarios with unrealistic conclusions, which will prevent timely risk mitigation. Therefore, for short-term adverse scenarios, the EBA and institutions should make sure to simulate sufficiently severe short-term shocks, such as asset stranding (fossil fuel assets, real estate price falls in flooding zones). In the case of CRA, the time horizon should be aligned with the worldwide Paris Agreement objectives of 2050.
Regarding the baseline scenario, as mentioned in our response to Q7, Finance Watch advises the use of realistic and common scenarios. There is now a strong consensus among scientists that society will face severe impacts from climate change provided current policies are adopted (and until there are signs of credible additional actions being implemented). Therefore, the baseline scenario should reflect this reality. We propose, as an example, to use NGFS “Current Policies” scenario as a baseline scenario. We reiterate also the importance of baseline scenario consistency between all forward-looking approaches within an institution.
Concerning adverse scenarios, integrating other possible stress events/shocks (geopolitical tensions, pandemics, technological developments, economic downturns) should be mandatory in any case (paragraph 39 “institutions should consider climate scenarios with and without additional adverse macroeconomic shocks.”). In fact, scientists have reached consensus on a strong correlation between climate change and all other disruptive events that could occur in the future. Indeed, climate change will exacerbate other economic shocks (compound risk). Therefore, there is no adverse climate scenario conceivable without macroeconomic shocks.
Finally, Finance Watch welcomes the EBA requirement for institutions to make adjustments to overcome lack of climate risk factors in adverse scenarios.
Question 11: Do you have comments on the description of the climate transmission channels?
Finance Watch welcomes the work of the EBA on identifying climate risk transmission channels. Examples of micro and macroeconomic channels are clear and relevant. This guidance offers a good starting point for institutions to think about transmission channels in their portfolio. Nevertheless, modelling of transmission channels and translating them into financial impacts will be a big challenge for banks due to non-linearity of impacts and compounding impacts of multiple risk drivers. Therefore, the EBA must ensure that transmission channels are accurately reflected in the institution's scenarios.
In addition, financial institutions should also consider nature loss threats when identifying transmission channels, such as stranded assets from deforestation activities or nature loss impact on health and productivity.
Question 12: Do you have comments on climate stress test (CST) tool and its use to test an institution’s financial resilience?
Within the existing risk management framework, stress tests are a good tool to assess institutions’ resilience to shocks, spot shortcomings in current risk management and potential capital shortfalls of an institution. With respect to climate-related risks, the stress testing approach needs to evolve to remain a useful and effective risk management tool. In particular, this requires:
- A clear framework with risk management direct effect: rely on ICAAP/ILAAP framework.
- Precise and granular information about their exposure: Institutions need precise climate-related information about their counterparties' business. The current EU discussions on the Omnibus package (CSRD, EU Taxonomy) will likely result in less data available for financial institutions. The extensive use of proxies will persist if not increase.
- Reliable time horizon: the current stress test time horizon (< 5 years) will not reflect the entire reality of climate change (physical risk amplification may occur in several decades but needs risk mitigation now).
- Plausible but adverse scenario: climate scenarios currently underestimate the severity of climate change (tipping points, irreversibility, magnitude) due to the use of unadapted economic models and lack of historical data.
- Modelled impact of scenarios on risk parameters: as climate change is a forward-looking event, it is complex to model its effects on classical risk metrics (PD, LGD, EAD). The assumption that the current relationship between risk metrics and economic downturns remains in a warmer world may be unrealistic. The calibration and validation of models (backtesting) may be biased as they suffer from a lack of historical data.
- Transparency and reporting: we welcome the requirements for full transparency of CST and disclosure by institutions.
- Clear actions: capital adequacy review and conclusions as part of ICAAP/ILAAP process.
In a nutshell, climate stress tests are, in their current form and definition, not sufficient for institutions to rely on their conclusion.
Moreover, traditional backtesting approaches (based on the use of historical data) will not be able to assess the quality of a CST. Therefore, Finance Watch supports the sensitivity approach and adjustment calibration requested by the EBA to overcome internal models shortcomings. Further, the EBA should also request adjustments for extreme climate events that are missing in current models, such as tipping points and chronic physical risks. All of these adjustments should be disclosed and well-explained to allow a review at a regulatory level.
Given the interconnectivity between climate risk and the other risks, effects from compound and second-order risks, and consistency with the definition of stress test characteristics, adverse scenarios in the CSA framework should embed macroeconomics downturns.
Then, a static balance sheet should be used in order to identify potential vulnerabilities based on their actual financial positions and current business strategies. A dynamic balance sheet could lead to an underestimation of losses in adverse scenarios, notwithstanding the fact that it ignores macroeconomic effects and disruption risk.
In addition to the CST, the EBA should require the use of Climate Reverse Stress Tests. Given the difficulties of short-term climate scenario modelling, this approach would improve financial institutions' understanding of potential vulnerabilities in their portfolios. By identifying potential combinations of factors that could trigger regulatory failures, it would make designing adverse scenarios easier. Sharing the magnitude of these shocks would offer crucial insights, helping to better anticipate future breaking points.
Question 13: Do you have comments on the Climate Resilience Analysis (CRA) tool and its use to challenge an institution’s business model resilience?
Finance Watch welcomes the forward-looking and longer-term approach adopted by the EBA. To keep consistency among the financial institutions, the EBA should at least align the CRA with financial institutions’ transition plan time horizon (i.e. 2050). Moreover, as for CST, the EBA should ensure that conclusions can be understood and compared among the financial institutions by requesting transparency on the assumptions and shortcomings of every scenario used. Finance Watch reiterates its concern about the consistency between all scenarios used by an institution.
Furthermore, the EBA should clarify the concrete actions that would result from the conclusion of CRA exercise. The EBA and financial industries should be aware that mitigating the long-term climate risks requires short-term actions. Specifically, such actions should focus in the first instance on real-world risk mitigation aligned with (and reflected in) the institution’s transition plan, including engagement with counterparties to review their resilience towards ESG risks, and portfolio reallocation in case engagement does not yield the target results.
Even if a dynamic balance sheet approach has merits at a micro-level, it leads to underestimating and misunderstanding the future risks due to climate change and, most importantly, misses the macro-dimension. In a context where most banking institutions behave similarly, a dynamic balance sheet approach will miss the fact that when all credit institutions adjust their exposures in a similar manner, new macro risks arise. This, by definition, cannot be spotted by an entity-level analysis, and it makes dynamic balance sheet approaches contribute to an unfounded and dangerous sense of climate risk-related resilience of the financial sector. Moreover a static balance sheet will highlight the potential vulnerabilities of current business strategies. Finance Watch therefore recommends that, at the very least, static and dynamic balance sheet analysis be conducted in parallel and their results compared in order to derive meaningful conclusions.
Moreover, both CRA and CST need to be consistent. Therefore, scenarios used in CRA should be in the continuity of the CST scenarios. In addition, CRA scenarios (as CST ones) should be continuously monitored and institutions should be expected to review them annually or in case of material change of their risk profile. Institutions should be able to identify new vulnerabilities or changes in the environment quickly and then update the impact of these exercises.
Question 14: Do you have any additional comments on the draft Guidelines on ESG Scenario Analysis?
Finance Watch supports the work of the EBA on the climate risk assessment. By aiming to extend the time horizon, to consider transmission channels, intertwined risk factors and possible adjustments, the EBA requirements in these Guidelines are an attempt to mitigate current framework limitations in embedding ESG risks. Although Finance Watch welcomes the EBA proposals and appreciates the opportunity to contribute to enhancing them, we emphasize the following essential aspects:
- Data availability and quality, complex modelling and scenario shortcomings are a real challenge for financial institutions and might not be achievable in the near future. It is essential that regulators establish a framework and allocate resources to monitor and validate the quality of institutions’ CSA exercises.
- Major limitations and approximations of the CSA framework remain: irreversibility of severe ESG impacts such as tipping points, rising sea level, and chronic physical risk are still not embedded although scientists are warning about their occurrence and their unprecedented impact; compound risk and non-linear effects are also underestimated. The time horizon of the exercises still does not match climate change reality and therefore leads to huge underestimations of the future cost of climate change for financial institutions.
The EBA should set up a review framework and process for CSA exercise by monitoring models, scenarios, assumptions, data quality, information availability and adjustment factors. It could be inspired by key indicators and reviews made in the Targeted Review of Internal Models (TRIM). It is crucial for banks that the CSA exercise is validated by the supervisors, so they can rely on its conclusions. Moreover, this type of review will identify the improvements banks need to better assess their climate risk.
Nevertheless, as long as the limitations of climate scenario analysis and economic modelling of climate change are not overcome, Finance Watch urges supervisors and regulators to implement a precautionary approach. This could be done at the microprudential or macroprudential level. These are propositions to mitigate transition risk from fossil fuel exposure. It could be implemented quickly and easily to ensure financial stability, which is already threatened.