Response to consultation on draft Guidelines on the overall recovery capacity in recovery planning
Go back
In this context, the assessment of the credibility and feasibility should remain exclusively qualitative and institution-specific, as it befits this kind of assessment that can only be performed while duly considering the specific capabilities and organization of every institution. Accordingly, it should be at least openly discussed with the institution, and, in case of disagreement by the authorities on the credibility or feasibility, ideally agreed upon.
The point laid down in item 19.a, page 24 of the EBA Draft Guidelines about ORC needs further clarification. The assessment of the impact of the recovery option on the financial system shall be made through the respective analysis of the continuity of critical functions. The financial system can be at risk if a business carrying out a critical function was being sold to an acquirer, who would be unable to ensure its continuity, would go through a disorderly wind down or be abruptly discontinued. However, such risk appears extremely remote as options concerning such businesses would be closely monitored and, in the case of sale in particular:
- the acquirer’s intent is normally to pursue the critical functions, since the latter contribute to the value of the acquired business;
- the seller would most likely assess the capacity of the acquirer to continue the critical functions;
- the acquisition of regulated businesses is subject to supervisory approval. The capacity of the acquirer to ensure such continuity and comply over time with regulatory requirements is thoroughly assessed by the supervisor approving the transaction.
With regard to item 19.b, page 24 of the EBA Draft Guidelines about ORC, EBF Members understand the conceptual point, but they believe it is more of theoretical value. Reference to previous experience would certainly be useful to assess the credibility and feasibility of the recovery options. However, it is likely that only few institutions have gone through recovery situations. In addition, when institutions experience recovery situations, this is generally not public information. Nevertheless, in case of a recovery option consisting of a sale of business or entity, reference to any similar, recent Mergers and Acquisitions (M&A) transactions performed by the institution itself or its peers would make sense.
Even though final assessment of credibility and feasibility of recovery options should remain qualitative and institution-specific, an indicative list of predefined criteria relevant for the feasibility and credibility analysis could be useful.
In regard to item 19.c, page 25 of the EBA Draft Guidelines about ORC, EBF Members believe that more details should be provided about the factor ‘level of preparedness for implementing the recovery option’.
The requirements on capital appear to be a tightening of existing regulation and would imply that well-capitalised or well-diversified banks would need even harsher and faster moving scenarios than less well-capitalised or well-diversified banks. Such a request would generate an unlevel playing field among European banks, in which well-capitalised banks or well-diversified banks would have to design very extreme and unrealistic scenarios and identify more and more potent recovery options than less well-capitalised banks or less well-diversified banks, which also seems counterintuitive. Furthermore, the conditions in which they would be supposed to activate those options would be very extreme too, potentially with detrimental effects on the value of such options.
In paragraph 21, in addition to the fact that solvency thresholds expected to be breached (TSCR or TSLRR) are considered as unrealistically severe and well beyond recovery situations (see below), “or” should be redrafted as “either … or” to avoid any possible misunderstanding. Indeed, as evidenced during the previous financial crisis, crises scenarios are rarely purely ‘capital scenarios’ or ‘liquidity scenarios’, but generally a mix of the two. For large well-diversified institutions, with recourse to wholesale funding, the probability to breach TSCR or TSLRR within 12 or 18 months is highly unlikely. Even in a slow-moving scenario beginning with capital losses, at some point of time, credit ratings would deteriorate below investment-grade, market confidence would disappear, and liquidity outflows would accelerate, causing severe liquidity problems well before the TSCR or TSLRR is breached. Moreover, for this type of banks, designing scenarios that would lead to a breach of TSCR or TSLRR would mean assuming macro-economic conditions never seen in Europe since last war. In such conditions, the estimation of the impacts of the options would be highly hypothetical, their credibility and feasibility would be questionable, and the concept of testing the options against a specific scenario would become meaningless.
Most importantly, the breach of TSCR or TSLRR would correspond to a ‘Failing or Likely to Fail’, i.e., a resolution and not a recovery situation whereas EBA/GL/2014/06 specify that “scenarios should be designed as ‘Near Default Point’ (NDP) situations; i.e., they should bring an institution close to failure but no further”.
It is also worth noting that scenarios in which institutions would let their capital ratios deteriorate to TSCR or TSLRR levels without taking any action or exercising any recovery option are extremely unrealistic; beyond the severity of the scenario that would lead to such breach, the absence of corrective actions is no realistic. Furthermore, it is inconsistent with other EBA guidelines such as the guidelines on recovery plan indicators which state that “capital indicators should be calibrated above the combined capital buffer requirement.” There is a contradiction between the two guidelines as, if banks follow the EBA guidelines on recovery indicators, the scenarios that are required to be performed mechanically cannot happen. The EBA cannot impose on banks guidelines that contradict each other.
If the need for scenarios specifically breaching regulatory capital requirements is nevertheless maintained, EBA should specify the NDP for all relevant indicators, and, in particular, it should consider including buffers for the solvency and leverage indicators (rather than using the TSCR, and by the same token using another name than NDP). In addition, EBA should consider the starting position and business model of the institution to determine whether this requirement to reach NDP applies or not. It would also be particularly important that the EBA clarify and objectivise the notion of ‘exceptional circumstances’ allowing an institution to be exempted from reaching the NDP. An alternative approach could be to define an acceptable NDP range (e.g., lower than Overall Capital Requirement (OCR) and higher that TSCR) which guarantees (1) very severe scenario, but not totally unrealistic when considering the relative importance of capital buffers, (2) continuum between Internal Capital Adequacy Assessment Process (ICAAP) and recovery in terms of actions and processes, (3) internal usability of the framework and (4) consistency with existing guidelines.
Moreover, EBF Members also question the relevance of comparing ORCs based on scenarios whose severity varies according to the initial positions and business models of the banks. European banks point out that, even with the same basis of calculation, the differences in business models, variety of options and starting points would lead to different resistance in the event of a crisis. The better the starting position, the more severe the scenarios should be and the more options as well as the ability of institution to recover could be invalidated, which is inconsistent.
Finally, EBF Members emphasize again that the ORC, in general or under a given scenario, is by nature institution-specific, as is the recovery plan and as are many recovery options.
In view of the multiple drawbacks and ineffectiveness of using the same scenarios to all banks, EBF Members believe that if the EBA or the European Central Bank (ECB) persist to compare ORCs, which should be done independently from a scenario. This could be done using the so-called (‘Business as Usual-Recovery Capacity’) mentioned in footnote 9 page 16 of the EBA Draft Guidelines about ORC, which is, in the banks’ view, the only reasonably relevant comparison basis that could possibly be used.
However, should the EBA persist and set predefined measurement timeframes, EBF Members underline that by limiting them to 6 months for liquidity and 12 months for capital, the number of options would be significantly reduced, and subsequently more focus could be placed by institutions on short-term solutions, instead of on long-term viability. Options enhancing the longer-term viability often require more drastic measures and tend to require a longer throughput time, especially in distressed markets (e.g., reorganisations like divestments). This concern is more prominent in fast moving, system-wide scenarios than in slow moving scenarios, as in the latter actions will already be taken anticipatively upon breaches of early warnings and limits and thus, before entering in recovery. Anyway, in case the EBA persists, EBF Members would recommend measuring the recovery capacity at different points in time and capturing thereby the longer-term resistance and viability of institutions.
‘Starting Point’
With regard to the proposed ‘Starting Point’ criteria, EBF Members believe that they would not be beneficial, given that heightened complexities are being created, while the assessment is being extended well beyond 6 or 12 months, respectively. The EBA Draft Guidelines about ORC should provide clear guidance as to whether the recovery analysis should extend after the 12-month window.
Certainly, the ‘Starting Point’ will vary among banks, thus this adds to the general incomparability of recovery planning among institutions, as it would potentially jeopardise the level playing field between institutions.
We would welcome clarifications on the concept of ‘Starting Point’. In particular, whether it refers to the calibrated recovery threshold of an indicator, or to the value reached by that indicator when the recovery threshold is breached. For example, if the CET1 recovery threshold is 9% but in a specific scenario the CET1 reaches an 8%, what should be the ‘Starting Point’? 9% or 8%? Secondly, if a bank estimates that it is going to breach the recovery threshold in the near future, we believe it is not realistic to wait until that happens in order to activate the recovery plan and start implementing recovery measures.
Timeframe
A 6-month horizon and a 12-month horizon for ORC calculation as regards liquidity indicators and solvency indicators respectively appear to be too restrictive. Flexibility – as mentioned during the EBA public hearing of 14 February 2023 - and longer time horizons should be taken into consideration since numerous effective measures may take much longer than 12 months to be implemented or to produce their full effect. Measuring recovery capacity over the medium to long-term is often sensible in terms of viability.
With the focus on 6 months and 12 months for liquidity and capital recovery options respectively, institutions will be driven to prioritize options that have a short-time impact only but may be sub-optimal in the longer term.
For instance, one of the major types of recovery options, notably, subsidiary disposal, might be largely excluded over a time horizon limited to 12 months. Even when the sale is completed within that timeframe, the full effect may not be reached yet in particular if the selling entity continues to provide funding to the sold one. A longer timeframe would be the best option. In any case, the ECB, as well as national competent authorities (NCAs), should take specific measures to speed up the regulatory approval processes for recovery (as for resolution) situations. Currently, this often extends a sale process by 4 months or more, which is not compatible with a crisis context.
Banks would strongly recommend avoiding measuring recovery capacity over a fixed time horizon per type of metric and adopting a more flexible and comprehensive approach (and the Standardised Reporting Template used by the ECB should be adapted accordingly).
Indeed, the ability to recover beyond 6 or 12 months is critical for assessing the longer term viability of an institution and it should not be neglected (e.g., an institution that would recover 50% of its LCR ratio in 6 months and 50% over the following 12 months should not be considered less solid than one that would recover 60% over the first 6 months and 10% over the following 12 months, rather the opposite).
Representative Indicators
EBF Members believe that too many indicators have to be followed and that it would be more relevant for the impacts of the scenarios to be quantified on a limited number of indicators and not on an as large number as proposed. In particular, we do not support the inclusion of the Net Stable Funding Ratio (NSFR) in the list of ‘relevant RP indicators’ because this is not a short-term liquidity risk indicator. It is actually a structural issue that shall be handled at a post-recovery stage.
Also, EBF Members do not support to add Minimum Requirement for Own Funds and Eligible Liabilities (MREL) to the list of ‘relevant RP indicators’ in view of the following:
- In general, MREL is intended as a buffer. A temporary breach of the MREL requirement alone is in itself not necessarily a sign of increased risk or financial weakness that should trigger the activation of the recovery plan.
- The capital elements of MREL are already captured by existing RP indicators on capital ratios like Total Capital.
- An MREL breach in relation to eligible liabilities only is most likely caused by the inaccessibility of capital markets. Activation of the recovery plan will not cure this.
Authorities already have specific powers that can be applied in case of a MREL breach by way of the Maximum Distributable Amount in relation to MREL (M-MDA), which is noteworthy not automatic. Upon notification of a breach, the relevant bank is already required to inform the authorities on the actions and timeline to restore the level of MREL eligible resources, including, where relevant, a funding plan.
We would also kindly require the EBA to clarify its envisaged approach in case of combined scenarios (capital and liquidity) with regard to the breach of indicators and the related ORC computation. In such scenarios, would a breach of capital indicator also start the Liquidity ORC computation? If this were not the case, banks should assume two different starting points for capital and liquidity in the same scenario, which would be contrary to the effective crisis management. We suggest EBA to better clarify that in combined scenarios the ‘Starting Point’ for capital and liquidity ORC is the first breach of a relevant indicator (whatever capital or liquidity indicator).
EBF Members do not agree with this step that departs from the concept of ORC, as it would not even measure a ‘capacity’ but just the ‘impact’ of a limited number of selected options. Thus far, the ECB has defined ORC as the measure of a bank's maximum recovery capacity at a given time. This involves a priori all the options available in the scenario considered, and not only the few options that would be selected as the most suitable to remediate this scenario. Therefore, all credible and feasible recovery measures should be included in the ORC (whether or not applied in a scenario).
In addition, the notion of appropriateness (“…options that would be available and appropriate…”) needs to be clarified. Our understanding is that ‘appropriate’ means that the selection of options should consider provisions of paragraph 30. and aims to reach the provisions of paragraph 42.a. However, there are several combinations of options resulting in different magnitudes to measure the scenario-specific recovery capacity, hence the ORC.
Step 2
Please find below some comments on items d. and e. regarding increased reputational effects and consequences to business models and profitability.
Item d: “Increased reputational risks”.
In the view of EBF Members, the potential constraining factor related to reputational effects is highly questionable: in a recovery situation, any effective and credible recovery action will be considered positively. Selling several subsidiaries is part of the constraints and the strategy adopted, yet the likelihood of numerous and massive disposals happening simultaneously is quite remote as, very often, just a few decisive actions are sufficient to cure even quite severe crises. As such, the implementation of a combination of several recovery options would generate reputational consequences and would modify the existing (pre-recovery) bank’s business model.
It is difficult to anticipate the effect of massive disposals, but it is important to highlight that ‘cause’ and ‘consequence’ should not be confused. Reputational risk would result essentially from the crisis situation triggering entry in recovery itself and the implementation of recovery actions may well be the visible sign of a crisis situation to the public, but it cannot be considered as a ‘cause’ of reputational risk.
Moreover, it should be noted that the Market Abuse Regulation requires a listed bank to disclose without delay any material element that could impact its share price, meaning that in the case of a listed bank, the visible sign would be the announcement of significant losses or problems, not the activation of the recovery plan. Furthermore, adequate communication should help making the reputational effect of a clear and credible recovery plan positive. In any case, the disposal of non-core entities should not pose any major reputational problems. In such context, decisions aiming at restoring the bank’s financial situation would be considered favourably by the bank’s stakeholders. If a bank approaches recovery territory, its reputation will automatically be affected by the visible deterioration of its financial positions, noticeable in the market and by its counterparties; the more serious the crisis, the more serious the reputational issue. If so, a clear and credible action plan commensurate to the crisis and executed consistently can only improve the reputation. The absence of such clear, visible, and credible reaction would, by contrast, be extremely detrimental to the reputation of the bank.
To sum up, introducing reputational effects within the constraining factors should be avoided and moreover, in our view, it is fundamentally inconsistent with the concept and purpose of ORC.
Item e: “Consequences to the business model or profitability”
In our view, in an ORC calculation, the consequences on the bank’s business model should not be considered. As specified by the ECB, the ORC a priori involves all the options available in the scenario considered and not only the options that would be the most suitable for this scenario (in particular by taking into account the consequences of the exercise of all options on the group’s business model). The implementation of an ORC-type (i.e., maximalist, and often widely over-reacting) strategy would definitely have a significant impact on the bank’s business model. This analysis is purely conventional for the purposes of the calculation; therefore, this element of business model impact should not, in our opinion, be considered in the ORC methodology. In a nutshell, it is a different approach that should be considered on its own in due time, when relevant, but that should not interfere with the ORC.
Step 3
EBF Members believe that step 3, as presented in the example (page 13 and page 17), should be clarified, since the manner the recovery capacity shall be computed requires further explanation and the reasons provided in paragraphs 25 and 23 are not sufficiently clear. Indeed, referring to the figure 12 included in page 17, it seems that the recovery capacity in terms of CET1 ratio is computed as the difference at a 12-month horizon (where the ‘Starting Point’ is the breach of indicator as defined in paragraph 23) between:
- the CET1 ratio as observed on the CET1 ratio trajectory with implementation of recovery options (green line);
- the CET1 ratio as observed on the CET1 ratio trajectory when no recovery options have been implemented (blue line); This is clearly what the vertical blue arrow in figure 12 suggests but this is not explicitly confirmed elsewhere in the EBA Draft Guidelines about ORC. Indeed, recovery capacity could also be calculated as the difference between:
- the CET1 ratio as observed at the ‘Starting Point’ (value of the ratio at the time of the breach of the indicator) and
- the CET1 ratio as observed at a x-month horizon (after the ‘Starting Point’) on the CET1 ratio trajectory with implementation of recovery options (green line).
The results of these two approaches differ significantly whereas both are dynamic as required in paragraph 31 (meaning that the level of the ORC at a given time horizon will be dependent on the strategy of recovery options implementation).
Additionally, the calculation of the ORC on a ‘dynamic balance sheet approach’ on the capital and liquidity ratios implies that significant differences in results could arise from different prioritization of options over each other. The impact on the ratios of each implementable action in a specific scenario depends on the dynamic positioning in the relation between numerator and denominator. The order of implementation of the actions reported in the recovery plan for maximizing the ORC calculated in such a way should definitely not be retained as a constraint for the management if ever having to implement the recovery plan.
EBF Members would rather refer to the footnote 9 on page 16 of the EBA Draft Guidelines about ORC. It allows the institutions “to use a reference value for their recovery capacity under the application of no scenario (Business as Usual – Recovery capacity): i.e., the sum of the impacts of the list of credible and feasible recovery options under no scenario while also adjusted for mutual exclusivity between certain options and any other constraining factors”. Such an ORC computation would be independent from scenarios, which are very hypothetical and ad-hoc, rather artificial constructions. In our view, it does represent a realistic measure of ORC in the actual meaning of the term.
Secondly, we do not support the application of haircuts, which can be quite subjective, while the reference to a peer group analysis to calibrate haircuts does not seem suitable either, as suggested by paragraph 38 of the EBA Draft Guidelines about ORC. In our opinion, the ORC should be analysed in relation to the business model and to the diversity of options available, and not (only) in relation to other institutions. Additionally, “cross-institutional comparison of the expected time required to implement a recovery option” (paragraph 38, point c of the EBA Draft Guidelines about ORC) does not appear appropriate, because some banks might have already developed in-house M&A expertise, as well as proven experience in processes such as acquiring, selling, disentangling parts of their businesses, or executing significant capital market transactions, whereas others may not. As a result, the latter would need long-learning time and significant external support towards this purpose.
Moreover, banking groups have different business models and different organizational structures, some of which facilitate the implementation of options, while some others render the same type of options less easy to implement. Thus, these characteristics can hardly be taken into account in the calibration of the haircuts, if it is made by reference to a peer group. Especially for relatively large banking groups, a case-by-case analysis – with an institution-specific ORC - appears to be the only one that would make sense.
All in all, we believe that no one is better placed than institutions themselves to determine the ‘feasibility’ of recovery options, as we also doubt how a peer group/cross-institutional analysis could tell if the implementation of recovery options is feasible. The authorities’ assessment should essentially be a consistency and plausibility check.
We also note that, paragraph 41 of the EBA Draft Guidelines about ORC reads as follows: “[…] that competent authorities’ ‘adjusted ORC’ should be either lower or equal to the ORC determined by institutions.” In this regard, EBF would suggest granting the power to competent authorities to potentially increase the ORC of the institutions deemed too conservative in their approach.
In case the proposed ORC score would eventually be implemented, the calculation and haircuts underlying the ORC adjustment should be adequately justified on a case-by-case basis, allowing banks to properly understand the haircuts applied and to replicate the calculations. Furthermore, the ECB should disclose its internal calculations (preferably, detailed by recovery option) to allow banks to figure out the criteria and subsequently improve the gaps.
In addition, we argue that the ‘adjusted ORC’ should not be included in the Supervisory Review and Evaluation Process (SREP) assessment. If, however, ORC was included in the SREP assessment, after a transition period, the way it would be integrated in the SREP score should be clearly specified well in advance. In the EBA public hearing of 14 February 2023, the EBA clarified that the judgement of the competent authority will be of qualitative nature and no quantitative haircuts will be applied. We strongly suggest the disclosure of the criteria used for reaching such qualitative judgement.
This calibration should not rely on an ORC comparison between banking groups because there will most likely always be significant yet justified variations across banks. The variety, substitutability, and extent of the recovery options by category, type of business, geographic dispersion are key elements in the appreciation of the capacity of a given banking group to recover from a deep crisis. As such, it should be also taken into account by the ORC score.
Also, we would like to reiterate that, from a financial stability perspective, the ability of banks to build up excess capital shall not be construed by the competent authorities as a detrimental factor.
Last but not least, we would kindly ask for a clearer and more precise communication from the competent authorities to institutions. The communication provided is often very vague or not straightforward and may lead to interpretation discrepancies as to what the competent authorities expect from institutions.
Other Comments:
Regarding page 22, Article 8
“For institutions that are part of a group subject to consolidated supervision pursuant to Articles 111 and 112 of Directive 2013/36/EU, these guidelines apply at the level of the Union parent undertaking and at the level of subsidiaries.”
Institutions that are part of a group subject to consolidated supervision typically prepare one recovery plan for the whole group, including its subsidiaries. Accordingly, the assessment and calculation of the ORC should be at group level. To avoid any confusion, an explicit clarification that the scope of application does not result in a requirement for subsidiaries to develop their own recovery plan or ORC would be highly appreciated.
Large and diversified banking groups have the ability to overcome severe problems, in particular geographies or business lines, thanks to their geographical spread and diversity of businesses. It is by essence the role of the parent undertaking to allocate support and resources where and when necessary, amongst its subsidiaries. As a consequence, it is best placed both to document and maintain the Recovery Plan and to implement it where this is required. In such group, local subsidiaries are not expected to develop their own Recovery Plans with specific recovery options and specific ORC calculation.
As a final remark, EBF Members believe that the macroprudential view could be more explicitly considered, particularly in systemic crises. Indeed, in a crisis, the Countercyclical Capital Buffer (CCyB) and/or Combined Buffer Requirement (CBR) would logically decrease (e.g., reference to the Covid crisis) and the Overall Capital Requirement (OCR) level to be reached after recovery should decrease accordingly. Hence, the Overall Recovery Capacity (ORC) required to cure a crisis should be measured against a reduced “post-crisis” Overall Capital Requirement (OCR) and not against the current one.
Question 1: do you have any comments on the general factors to be considered when assessing credibility and feasibility of the recovery options?
The assessment of recovery options’ credibility and feasibility is highly dependent on the discretionary interpretation of the person that is assessing the information. This assessment may lead to diverse interpretations and viewpoints; it is subjective by nature and, therefore, potentially conductive of an unlevel playing field.In this context, the assessment of the credibility and feasibility should remain exclusively qualitative and institution-specific, as it befits this kind of assessment that can only be performed while duly considering the specific capabilities and organization of every institution. Accordingly, it should be at least openly discussed with the institution, and, in case of disagreement by the authorities on the credibility or feasibility, ideally agreed upon.
The point laid down in item 19.a, page 24 of the EBA Draft Guidelines about ORC needs further clarification. The assessment of the impact of the recovery option on the financial system shall be made through the respective analysis of the continuity of critical functions. The financial system can be at risk if a business carrying out a critical function was being sold to an acquirer, who would be unable to ensure its continuity, would go through a disorderly wind down or be abruptly discontinued. However, such risk appears extremely remote as options concerning such businesses would be closely monitored and, in the case of sale in particular:
- the acquirer’s intent is normally to pursue the critical functions, since the latter contribute to the value of the acquired business;
- the seller would most likely assess the capacity of the acquirer to continue the critical functions;
- the acquisition of regulated businesses is subject to supervisory approval. The capacity of the acquirer to ensure such continuity and comply over time with regulatory requirements is thoroughly assessed by the supervisor approving the transaction.
With regard to item 19.b, page 24 of the EBA Draft Guidelines about ORC, EBF Members understand the conceptual point, but they believe it is more of theoretical value. Reference to previous experience would certainly be useful to assess the credibility and feasibility of the recovery options. However, it is likely that only few institutions have gone through recovery situations. In addition, when institutions experience recovery situations, this is generally not public information. Nevertheless, in case of a recovery option consisting of a sale of business or entity, reference to any similar, recent Mergers and Acquisitions (M&A) transactions performed by the institution itself or its peers would make sense.
Even though final assessment of credibility and feasibility of recovery options should remain qualitative and institution-specific, an indicative list of predefined criteria relevant for the feasibility and credibility analysis could be useful.
In regard to item 19.c, page 25 of the EBA Draft Guidelines about ORC, EBF Members believe that more details should be provided about the factor ‘level of preparedness for implementing the recovery option’.
Question 2: do you have any comments on the specification of the scenario severity for the purpose of calculating the ‘scenario-specific recovery capacity’?
Paragraph 21 specifies that the scenario should result in the institutions breaching their Total SREP Capital Requirement (TSCR) or Total SREP Leverage Ratio Requirement (TSLRR) or their minimum regulatory liquidity requirements.The requirements on capital appear to be a tightening of existing regulation and would imply that well-capitalised or well-diversified banks would need even harsher and faster moving scenarios than less well-capitalised or well-diversified banks. Such a request would generate an unlevel playing field among European banks, in which well-capitalised banks or well-diversified banks would have to design very extreme and unrealistic scenarios and identify more and more potent recovery options than less well-capitalised banks or less well-diversified banks, which also seems counterintuitive. Furthermore, the conditions in which they would be supposed to activate those options would be very extreme too, potentially with detrimental effects on the value of such options.
In paragraph 21, in addition to the fact that solvency thresholds expected to be breached (TSCR or TSLRR) are considered as unrealistically severe and well beyond recovery situations (see below), “or” should be redrafted as “either … or” to avoid any possible misunderstanding. Indeed, as evidenced during the previous financial crisis, crises scenarios are rarely purely ‘capital scenarios’ or ‘liquidity scenarios’, but generally a mix of the two. For large well-diversified institutions, with recourse to wholesale funding, the probability to breach TSCR or TSLRR within 12 or 18 months is highly unlikely. Even in a slow-moving scenario beginning with capital losses, at some point of time, credit ratings would deteriorate below investment-grade, market confidence would disappear, and liquidity outflows would accelerate, causing severe liquidity problems well before the TSCR or TSLRR is breached. Moreover, for this type of banks, designing scenarios that would lead to a breach of TSCR or TSLRR would mean assuming macro-economic conditions never seen in Europe since last war. In such conditions, the estimation of the impacts of the options would be highly hypothetical, their credibility and feasibility would be questionable, and the concept of testing the options against a specific scenario would become meaningless.
Most importantly, the breach of TSCR or TSLRR would correspond to a ‘Failing or Likely to Fail’, i.e., a resolution and not a recovery situation whereas EBA/GL/2014/06 specify that “scenarios should be designed as ‘Near Default Point’ (NDP) situations; i.e., they should bring an institution close to failure but no further”.
It is also worth noting that scenarios in which institutions would let their capital ratios deteriorate to TSCR or TSLRR levels without taking any action or exercising any recovery option are extremely unrealistic; beyond the severity of the scenario that would lead to such breach, the absence of corrective actions is no realistic. Furthermore, it is inconsistent with other EBA guidelines such as the guidelines on recovery plan indicators which state that “capital indicators should be calibrated above the combined capital buffer requirement.” There is a contradiction between the two guidelines as, if banks follow the EBA guidelines on recovery indicators, the scenarios that are required to be performed mechanically cannot happen. The EBA cannot impose on banks guidelines that contradict each other.
If the need for scenarios specifically breaching regulatory capital requirements is nevertheless maintained, EBA should specify the NDP for all relevant indicators, and, in particular, it should consider including buffers for the solvency and leverage indicators (rather than using the TSCR, and by the same token using another name than NDP). In addition, EBA should consider the starting position and business model of the institution to determine whether this requirement to reach NDP applies or not. It would also be particularly important that the EBA clarify and objectivise the notion of ‘exceptional circumstances’ allowing an institution to be exempted from reaching the NDP. An alternative approach could be to define an acceptable NDP range (e.g., lower than Overall Capital Requirement (OCR) and higher that TSCR) which guarantees (1) very severe scenario, but not totally unrealistic when considering the relative importance of capital buffers, (2) continuum between Internal Capital Adequacy Assessment Process (ICAAP) and recovery in terms of actions and processes, (3) internal usability of the framework and (4) consistency with existing guidelines.
Moreover, EBF Members also question the relevance of comparing ORCs based on scenarios whose severity varies according to the initial positions and business models of the banks. European banks point out that, even with the same basis of calculation, the differences in business models, variety of options and starting points would lead to different resistance in the event of a crisis. The better the starting position, the more severe the scenarios should be and the more options as well as the ability of institution to recover could be invalidated, which is inconsistent.
Finally, EBF Members emphasize again that the ORC, in general or under a given scenario, is by nature institution-specific, as is the recovery plan and as are many recovery options.
In view of the multiple drawbacks and ineffectiveness of using the same scenarios to all banks, EBF Members believe that if the EBA or the European Central Bank (ECB) persist to compare ORCs, which should be done independently from a scenario. This could be done using the so-called (‘Business as Usual-Recovery Capacity’) mentioned in footnote 9 page 16 of the EBA Draft Guidelines about ORC, which is, in the banks’ view, the only reasonably relevant comparison basis that could possibly be used.
Question 3: do you agree with the proposed criteria for the relevant starting point, timeframe (in particular with regard to the 6-month period for the LCR and NSFR) and representative indicators (in particular with regard to the explicit consideration of potential other/substitute indicators – e.g. MREL) for the ‘scenario-specific recovery capacity’?
Setting a-priori a maximum timeframe for measuring the recovery capacity in terms of capital and liquidity positions is not perceived as a suitable solution. Indeed, setting ex ante such predefined timeframes would hardly ever be compatible with the timeframes over which the scenarios and the options develop, and it would add further artificial constraints removing relevance to the exercise. For these reasons, the timeframe for the assessment of the ORC should be more coherently evaluated in accordance with the deployment of the underlying scenario and could vary from different scenarios and different institutions.However, should the EBA persist and set predefined measurement timeframes, EBF Members underline that by limiting them to 6 months for liquidity and 12 months for capital, the number of options would be significantly reduced, and subsequently more focus could be placed by institutions on short-term solutions, instead of on long-term viability. Options enhancing the longer-term viability often require more drastic measures and tend to require a longer throughput time, especially in distressed markets (e.g., reorganisations like divestments). This concern is more prominent in fast moving, system-wide scenarios than in slow moving scenarios, as in the latter actions will already be taken anticipatively upon breaches of early warnings and limits and thus, before entering in recovery. Anyway, in case the EBA persists, EBF Members would recommend measuring the recovery capacity at different points in time and capturing thereby the longer-term resistance and viability of institutions.
‘Starting Point’
With regard to the proposed ‘Starting Point’ criteria, EBF Members believe that they would not be beneficial, given that heightened complexities are being created, while the assessment is being extended well beyond 6 or 12 months, respectively. The EBA Draft Guidelines about ORC should provide clear guidance as to whether the recovery analysis should extend after the 12-month window.
Certainly, the ‘Starting Point’ will vary among banks, thus this adds to the general incomparability of recovery planning among institutions, as it would potentially jeopardise the level playing field between institutions.
We would welcome clarifications on the concept of ‘Starting Point’. In particular, whether it refers to the calibrated recovery threshold of an indicator, or to the value reached by that indicator when the recovery threshold is breached. For example, if the CET1 recovery threshold is 9% but in a specific scenario the CET1 reaches an 8%, what should be the ‘Starting Point’? 9% or 8%? Secondly, if a bank estimates that it is going to breach the recovery threshold in the near future, we believe it is not realistic to wait until that happens in order to activate the recovery plan and start implementing recovery measures.
Timeframe
A 6-month horizon and a 12-month horizon for ORC calculation as regards liquidity indicators and solvency indicators respectively appear to be too restrictive. Flexibility – as mentioned during the EBA public hearing of 14 February 2023 - and longer time horizons should be taken into consideration since numerous effective measures may take much longer than 12 months to be implemented or to produce their full effect. Measuring recovery capacity over the medium to long-term is often sensible in terms of viability.
With the focus on 6 months and 12 months for liquidity and capital recovery options respectively, institutions will be driven to prioritize options that have a short-time impact only but may be sub-optimal in the longer term.
For instance, one of the major types of recovery options, notably, subsidiary disposal, might be largely excluded over a time horizon limited to 12 months. Even when the sale is completed within that timeframe, the full effect may not be reached yet in particular if the selling entity continues to provide funding to the sold one. A longer timeframe would be the best option. In any case, the ECB, as well as national competent authorities (NCAs), should take specific measures to speed up the regulatory approval processes for recovery (as for resolution) situations. Currently, this often extends a sale process by 4 months or more, which is not compatible with a crisis context.
Banks would strongly recommend avoiding measuring recovery capacity over a fixed time horizon per type of metric and adopting a more flexible and comprehensive approach (and the Standardised Reporting Template used by the ECB should be adapted accordingly).
Indeed, the ability to recover beyond 6 or 12 months is critical for assessing the longer term viability of an institution and it should not be neglected (e.g., an institution that would recover 50% of its LCR ratio in 6 months and 50% over the following 12 months should not be considered less solid than one that would recover 60% over the first 6 months and 10% over the following 12 months, rather the opposite).
Representative Indicators
EBF Members believe that too many indicators have to be followed and that it would be more relevant for the impacts of the scenarios to be quantified on a limited number of indicators and not on an as large number as proposed. In particular, we do not support the inclusion of the Net Stable Funding Ratio (NSFR) in the list of ‘relevant RP indicators’ because this is not a short-term liquidity risk indicator. It is actually a structural issue that shall be handled at a post-recovery stage.
Also, EBF Members do not support to add Minimum Requirement for Own Funds and Eligible Liabilities (MREL) to the list of ‘relevant RP indicators’ in view of the following:
- In general, MREL is intended as a buffer. A temporary breach of the MREL requirement alone is in itself not necessarily a sign of increased risk or financial weakness that should trigger the activation of the recovery plan.
- The capital elements of MREL are already captured by existing RP indicators on capital ratios like Total Capital.
- An MREL breach in relation to eligible liabilities only is most likely caused by the inaccessibility of capital markets. Activation of the recovery plan will not cure this.
Authorities already have specific powers that can be applied in case of a MREL breach by way of the Maximum Distributable Amount in relation to MREL (M-MDA), which is noteworthy not automatic. Upon notification of a breach, the relevant bank is already required to inform the authorities on the actions and timeline to restore the level of MREL eligible resources, including, where relevant, a funding plan.
We would also kindly require the EBA to clarify its envisaged approach in case of combined scenarios (capital and liquidity) with regard to the breach of indicators and the related ORC computation. In such scenarios, would a breach of capital indicator also start the Liquidity ORC computation? If this were not the case, banks should assume two different starting points for capital and liquidity in the same scenario, which would be contrary to the effective crisis management. We suggest EBA to better clarify that in combined scenarios the ‘Starting Point’ for capital and liquidity ORC is the first breach of a relevant indicator (whatever capital or liquidity indicator).
Question 4: do you have any comments on the general steps to be followed for the determination of the ORC?
Step 1EBF Members do not agree with this step that departs from the concept of ORC, as it would not even measure a ‘capacity’ but just the ‘impact’ of a limited number of selected options. Thus far, the ECB has defined ORC as the measure of a bank's maximum recovery capacity at a given time. This involves a priori all the options available in the scenario considered, and not only the few options that would be selected as the most suitable to remediate this scenario. Therefore, all credible and feasible recovery measures should be included in the ORC (whether or not applied in a scenario).
In addition, the notion of appropriateness (“…options that would be available and appropriate…”) needs to be clarified. Our understanding is that ‘appropriate’ means that the selection of options should consider provisions of paragraph 30. and aims to reach the provisions of paragraph 42.a. However, there are several combinations of options resulting in different magnitudes to measure the scenario-specific recovery capacity, hence the ORC.
Step 2
Please find below some comments on items d. and e. regarding increased reputational effects and consequences to business models and profitability.
Item d: “Increased reputational risks”.
In the view of EBF Members, the potential constraining factor related to reputational effects is highly questionable: in a recovery situation, any effective and credible recovery action will be considered positively. Selling several subsidiaries is part of the constraints and the strategy adopted, yet the likelihood of numerous and massive disposals happening simultaneously is quite remote as, very often, just a few decisive actions are sufficient to cure even quite severe crises. As such, the implementation of a combination of several recovery options would generate reputational consequences and would modify the existing (pre-recovery) bank’s business model.
It is difficult to anticipate the effect of massive disposals, but it is important to highlight that ‘cause’ and ‘consequence’ should not be confused. Reputational risk would result essentially from the crisis situation triggering entry in recovery itself and the implementation of recovery actions may well be the visible sign of a crisis situation to the public, but it cannot be considered as a ‘cause’ of reputational risk.
Moreover, it should be noted that the Market Abuse Regulation requires a listed bank to disclose without delay any material element that could impact its share price, meaning that in the case of a listed bank, the visible sign would be the announcement of significant losses or problems, not the activation of the recovery plan. Furthermore, adequate communication should help making the reputational effect of a clear and credible recovery plan positive. In any case, the disposal of non-core entities should not pose any major reputational problems. In such context, decisions aiming at restoring the bank’s financial situation would be considered favourably by the bank’s stakeholders. If a bank approaches recovery territory, its reputation will automatically be affected by the visible deterioration of its financial positions, noticeable in the market and by its counterparties; the more serious the crisis, the more serious the reputational issue. If so, a clear and credible action plan commensurate to the crisis and executed consistently can only improve the reputation. The absence of such clear, visible, and credible reaction would, by contrast, be extremely detrimental to the reputation of the bank.
To sum up, introducing reputational effects within the constraining factors should be avoided and moreover, in our view, it is fundamentally inconsistent with the concept and purpose of ORC.
Item e: “Consequences to the business model or profitability”
In our view, in an ORC calculation, the consequences on the bank’s business model should not be considered. As specified by the ECB, the ORC a priori involves all the options available in the scenario considered and not only the options that would be the most suitable for this scenario (in particular by taking into account the consequences of the exercise of all options on the group’s business model). The implementation of an ORC-type (i.e., maximalist, and often widely over-reacting) strategy would definitely have a significant impact on the bank’s business model. This analysis is purely conventional for the purposes of the calculation; therefore, this element of business model impact should not, in our opinion, be considered in the ORC methodology. In a nutshell, it is a different approach that should be considered on its own in due time, when relevant, but that should not interfere with the ORC.
Step 3
EBF Members believe that step 3, as presented in the example (page 13 and page 17), should be clarified, since the manner the recovery capacity shall be computed requires further explanation and the reasons provided in paragraphs 25 and 23 are not sufficiently clear. Indeed, referring to the figure 12 included in page 17, it seems that the recovery capacity in terms of CET1 ratio is computed as the difference at a 12-month horizon (where the ‘Starting Point’ is the breach of indicator as defined in paragraph 23) between:
- the CET1 ratio as observed on the CET1 ratio trajectory with implementation of recovery options (green line);
- the CET1 ratio as observed on the CET1 ratio trajectory when no recovery options have been implemented (blue line); This is clearly what the vertical blue arrow in figure 12 suggests but this is not explicitly confirmed elsewhere in the EBA Draft Guidelines about ORC. Indeed, recovery capacity could also be calculated as the difference between:
- the CET1 ratio as observed at the ‘Starting Point’ (value of the ratio at the time of the breach of the indicator) and
- the CET1 ratio as observed at a x-month horizon (after the ‘Starting Point’) on the CET1 ratio trajectory with implementation of recovery options (green line).
The results of these two approaches differ significantly whereas both are dynamic as required in paragraph 31 (meaning that the level of the ORC at a given time horizon will be dependent on the strategy of recovery options implementation).
Additionally, the calculation of the ORC on a ‘dynamic balance sheet approach’ on the capital and liquidity ratios implies that significant differences in results could arise from different prioritization of options over each other. The impact on the ratios of each implementable action in a specific scenario depends on the dynamic positioning in the relation between numerator and denominator. The order of implementation of the actions reported in the recovery plan for maximizing the ORC calculated in such a way should definitely not be retained as a constraint for the management if ever having to implement the recovery plan.
Question 5: do you have any comments on the definition of the ORC as a range between the lowest and the highest ‘scenario-specific recovery capacity’ both in terms of capital and liquidity?
In short, EBF Members do not reject the idea of determining the ORC as a range. However, the proposed range is not fully clear because it is based on scenarios of different nature (systemic, idiosyncratic, among others).EBF Members would rather refer to the footnote 9 on page 16 of the EBA Draft Guidelines about ORC. It allows the institutions “to use a reference value for their recovery capacity under the application of no scenario (Business as Usual – Recovery capacity): i.e., the sum of the impacts of the list of credible and feasible recovery options under no scenario while also adjusted for mutual exclusivity between certain options and any other constraining factors”. Such an ORC computation would be independent from scenarios, which are very hypothetical and ad-hoc, rather artificial constructions. In our view, it does represent a realistic measure of ORC in the actual meaning of the term.
Question 6: do you have any comments on the scope of the assessment of the ‘scenario-specific recovery capacity’ by the competent authorities?
First of all, EBF Members consider that, subject to realistic values and implementation timelines, all the options presented in a well-prepared recovery plan are feasible a priori; ‘feasibility’ is an intrinsic quality of the option that is independent of any scenario. In particular, for the purposes of the plan, a certain number of options assessed as relatively easy to implement, is selected (and not a maximum number of options with uneven degrees of feasibility). Normally, valuations and implementation timelines (in benign and stressed contexts) do take into account all the identified potential issues that could impact their implementation. Based on these valuations and within these implementation timelines, all the recovery options presented can be considered ’highly feasible’. Therefore, in the plan, the ‘feasibility’ of an option cannot be considered a distinguishing characteristic of one option over another.Secondly, we do not support the application of haircuts, which can be quite subjective, while the reference to a peer group analysis to calibrate haircuts does not seem suitable either, as suggested by paragraph 38 of the EBA Draft Guidelines about ORC. In our opinion, the ORC should be analysed in relation to the business model and to the diversity of options available, and not (only) in relation to other institutions. Additionally, “cross-institutional comparison of the expected time required to implement a recovery option” (paragraph 38, point c of the EBA Draft Guidelines about ORC) does not appear appropriate, because some banks might have already developed in-house M&A expertise, as well as proven experience in processes such as acquiring, selling, disentangling parts of their businesses, or executing significant capital market transactions, whereas others may not. As a result, the latter would need long-learning time and significant external support towards this purpose.
Moreover, banking groups have different business models and different organizational structures, some of which facilitate the implementation of options, while some others render the same type of options less easy to implement. Thus, these characteristics can hardly be taken into account in the calibration of the haircuts, if it is made by reference to a peer group. Especially for relatively large banking groups, a case-by-case analysis – with an institution-specific ORC - appears to be the only one that would make sense.
All in all, we believe that no one is better placed than institutions themselves to determine the ‘feasibility’ of recovery options, as we also doubt how a peer group/cross-institutional analysis could tell if the implementation of recovery options is feasible. The authorities’ assessment should essentially be a consistency and plausibility check.
Question 7: do you have any comments on the proposed ORC score?
As a general principle, EBF Members would recommend competent authorities to be transparent vis-à-vis all quantitative and qualitative adjustments they make in assessing the ORC. In particular, such adjustments should be explicitly disclosed to the institutions with sufficient level of detail in order to be concrete and allow them to understand the rationale, while avoiding any misinterpretations.We also note that, paragraph 41 of the EBA Draft Guidelines about ORC reads as follows: “[…] that competent authorities’ ‘adjusted ORC’ should be either lower or equal to the ORC determined by institutions.” In this regard, EBF would suggest granting the power to competent authorities to potentially increase the ORC of the institutions deemed too conservative in their approach.
In case the proposed ORC score would eventually be implemented, the calculation and haircuts underlying the ORC adjustment should be adequately justified on a case-by-case basis, allowing banks to properly understand the haircuts applied and to replicate the calculations. Furthermore, the ECB should disclose its internal calculations (preferably, detailed by recovery option) to allow banks to figure out the criteria and subsequently improve the gaps.
In addition, we argue that the ‘adjusted ORC’ should not be included in the Supervisory Review and Evaluation Process (SREP) assessment. If, however, ORC was included in the SREP assessment, after a transition period, the way it would be integrated in the SREP score should be clearly specified well in advance. In the EBA public hearing of 14 February 2023, the EBA clarified that the judgement of the competent authority will be of qualitative nature and no quantitative haircuts will be applied. We strongly suggest the disclosure of the criteria used for reaching such qualitative judgement.
This calibration should not rely on an ORC comparison between banking groups because there will most likely always be significant yet justified variations across banks. The variety, substitutability, and extent of the recovery options by category, type of business, geographic dispersion are key elements in the appreciation of the capacity of a given banking group to recover from a deep crisis. As such, it should be also taken into account by the ORC score.
Question 8: do you have any comments on the possibility to identify areas of improvement or material deficiencies related to the competent authorities’ assessment of the ORC?
The suggested standardisation is appreciated by EBF Members, however it might lead to increased complexity, delivering incomparable outcomes derived from individual scenarios of breaching individual recovery indicators different from bank to bank. This being said, we believe that standardisation of the supervisory assessment approach is welcome, though such assessment should remain a case-by-case exercise particularly for large banking groups. In particular, the initial situation of the institution, the diversity of its business model, and the severity of scenarios that potentially affect the ability to recover, given the extremely deteriorated environment in which the institution would theoretically operate, should be taken into account, as well as the number, variety, and availability of recovery options.Also, we would like to reiterate that, from a financial stability perspective, the ability of banks to build up excess capital shall not be construed by the competent authorities as a detrimental factor.
Last but not least, we would kindly ask for a clearer and more precise communication from the competent authorities to institutions. The communication provided is often very vague or not straightforward and may lead to interpretation discrepancies as to what the competent authorities expect from institutions.
Other Comments:
Regarding page 22, Article 8
“For institutions that are part of a group subject to consolidated supervision pursuant to Articles 111 and 112 of Directive 2013/36/EU, these guidelines apply at the level of the Union parent undertaking and at the level of subsidiaries.”
Institutions that are part of a group subject to consolidated supervision typically prepare one recovery plan for the whole group, including its subsidiaries. Accordingly, the assessment and calculation of the ORC should be at group level. To avoid any confusion, an explicit clarification that the scope of application does not result in a requirement for subsidiaries to develop their own recovery plan or ORC would be highly appreciated.
Large and diversified banking groups have the ability to overcome severe problems, in particular geographies or business lines, thanks to their geographical spread and diversity of businesses. It is by essence the role of the parent undertaking to allocate support and resources where and when necessary, amongst its subsidiaries. As a consequence, it is best placed both to document and maintain the Recovery Plan and to implement it where this is required. In such group, local subsidiaries are not expected to develop their own Recovery Plans with specific recovery options and specific ORC calculation.
As a final remark, EBF Members believe that the macroprudential view could be more explicitly considered, particularly in systemic crises. Indeed, in a crisis, the Countercyclical Capital Buffer (CCyB) and/or Combined Buffer Requirement (CBR) would logically decrease (e.g., reference to the Covid crisis) and the Overall Capital Requirement (OCR) level to be reached after recovery should decrease accordingly. Hence, the Overall Recovery Capacity (ORC) required to cure a crisis should be measured against a reduced “post-crisis” Overall Capital Requirement (OCR) and not against the current one.