Response to consultation on draft Guidelines on the overall recovery capacity in recovery planning
Go back
Also this approach could contradict stress scenarios based on events that are exceptional but plausibly possible (like e. g. according to the German MaSanV): Depending on the capitalisation of the institution at the respective reporting date, the stress scenarios must be of varying severity. For an institution with good capitalisation, this may mean that the stress effect must be disproportionately severe and thus no longer within the range of what is plausibly possible.
For this reason the approach should be based on banks specific circumstances. Recovery is a bottom-up exercise. Thus, it should be banks defining their individual conditions for “near default” given their business model, capital and liquidity situation, potential externalities, etc. Regulatory requirements should not be the lead for defining banks’ near default stage.
Furthermore, we do not consider a comparison of banks' ORC based on stress scenarios with different severity levels to be suitable for the intended benchmarking, as the probability of occurrence differs between institutions.
Requiring banks to stress to the level of the TSCR makes it also often impossible to comply with other EBA requirements for plausible scenarios (i.e. EBA Guidelines on recovery scenarios).
Finally the Macroprudential view is missing in the approach, in particular:
• We assume for recovery planning the MDA level should decrease during period of stress due to Countercyclical capital buffer (CCyb) depletion;
• The approach may lead to a vicious circle where banks trying to sell their assets would cause further deterioration in the economy;
• and make the biggest part of the buffers practically “unusable”. In the case of capital buffer releases by supervisors (as was the case at the beginning of the Covid19 pandemic), capital buffers cannot be used due to the requirement for recovery planning. The lack of flexibility in setting thresholds for recovery indicators in exceptional circumstances prevents banks from using capital buffers. This is because the use of capital buffers would trigger escalation processes.
• It creates heightened complexities (for scenario selection, timing of recovery options etc.) and extends the assessment well beyond 12 months considering also the initial phase in the run up of an indicator breach.
• The starting point will vary among banks and this could jeopardize the overall objective of comparability among institutions.
• There should be no split of timeframes of 6 months for liquidity and 12 months for capital. For both scenarios a 12m time horizon should be considered to increase comparability and provide same conditions and settings for the recovery scenario. With regard to liquidity the partially delayed effect of individual liquidity options would be taken into account in this way. Furthermore, it should not only be possible to look at the effect on capital indicators after 12 months, but also later, e. g. after 15 or 18 months. Transactions, such as the sale of subsidiaries and participations, may require a longer implementation period until they take effect in the balance sheet after the transaction has been closed.
See also Question 5.
All factors used to determine the “adjusted ORC” and the “final ORC” by the supervisor should be disclosed to the affected bank, in particular to make the use of double haircuts (by the bank as well as by the supervisor) visible. In particular, the bank concerned should be informed if haircuts have been applied due to a lack of information on the options. In addition, it should be disclosed which qualitative factors were used and included in the adjusted or final ORC. It would be helpful to disclose which comparative values (e.g. haircuts, etc.) are used by peer groups.
Question 1: do you have any comments on the general factors to be considered when assessing credibility and feasibility of the recovery options?
Not applicable.Question 2: do you have any comments on the specification of the scenario severity for the purpose of calculating the ‘scenario-specific recovery capacity’?
The Guidelines main objective is to develop a scenario in which the bank reaches a ‘near default’ stage. Furthermore, the Guidelines require that the scenarios should be severe enough to breach the Total SREP Capital Requirement (TSCR) level. However, breaching the TSCR level does not necessarily mean ‘near default’, as the regulatory requirement could be set too high, thus not at default stage.Also this approach could contradict stress scenarios based on events that are exceptional but plausibly possible (like e. g. according to the German MaSanV): Depending on the capitalisation of the institution at the respective reporting date, the stress scenarios must be of varying severity. For an institution with good capitalisation, this may mean that the stress effect must be disproportionately severe and thus no longer within the range of what is plausibly possible.
For this reason the approach should be based on banks specific circumstances. Recovery is a bottom-up exercise. Thus, it should be banks defining their individual conditions for “near default” given their business model, capital and liquidity situation, potential externalities, etc. Regulatory requirements should not be the lead for defining banks’ near default stage.
Furthermore, we do not consider a comparison of banks' ORC based on stress scenarios with different severity levels to be suitable for the intended benchmarking, as the probability of occurrence differs between institutions.
Requiring banks to stress to the level of the TSCR makes it also often impossible to comply with other EBA requirements for plausible scenarios (i.e. EBA Guidelines on recovery scenarios).
Finally the Macroprudential view is missing in the approach, in particular:
• We assume for recovery planning the MDA level should decrease during period of stress due to Countercyclical capital buffer (CCyb) depletion;
• The approach may lead to a vicious circle where banks trying to sell their assets would cause further deterioration in the economy;
• and make the biggest part of the buffers practically “unusable”. In the case of capital buffer releases by supervisors (as was the case at the beginning of the Covid19 pandemic), capital buffers cannot be used due to the requirement for recovery planning. The lack of flexibility in setting thresholds for recovery indicators in exceptional circumstances prevents banks from using capital buffers. This is because the use of capital buffers would trigger escalation processes.
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’?
The proposed starting point criteria would not be beneficial, given that:• It creates heightened complexities (for scenario selection, timing of recovery options etc.) and extends the assessment well beyond 12 months considering also the initial phase in the run up of an indicator breach.
• The starting point will vary among banks and this could jeopardize the overall objective of comparability among institutions.
• There should be no split of timeframes of 6 months for liquidity and 12 months for capital. For both scenarios a 12m time horizon should be considered to increase comparability and provide same conditions and settings for the recovery scenario. With regard to liquidity the partially delayed effect of individual liquidity options would be taken into account in this way. Furthermore, it should not only be possible to look at the effect on capital indicators after 12 months, but also later, e. g. after 15 or 18 months. Transactions, such as the sale of subsidiaries and participations, may require a longer implementation period until they take effect in the balance sheet after the transaction has been closed.
See also Question 5.
Question 4: do you have any comments on the general steps to be followed for the determination of the ORC?
The guideline seems to assume that the threshold of a recovery indicator will return to the same level after the recovery measure. This is not necessarily the case, because depending on the nature of the recovery measure, an adjustment of the threshold for the recovery indicator might be necessary.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?
A clear distinction between capital and liquidity scenarios might not always be possible due to their interlinkages, therefore it should be clarified that whereas the focus might be on capital or liquidity in some scenarios, there might still be an impact on the other (just not a material one).Question 6: do you have any comments on the scope of the assessment of the ‘scenario-specific recovery capacity’ by the competent authorities?
As suggested above, the current setting regarding prolonged time frames etc. would result in incomparability among peers.Question 7: do you have any comments on the proposed ORC score?
The ORC score assessments and benefits are not clear at this stage. In particular the “adjusted ORC” should not be included in the SREP assessment, especially not in the beginning as even with these guidelines, there will most likely still be significant variances across banks.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?
Standardization and harmonization of supervisory assessment is key. The suggested standardization is appreciated, however risks of standardizing increased complexity delivering incomparable outcomes given the individual scenarios of breached individual recovery indicators.All factors used to determine the “adjusted ORC” and the “final ORC” by the supervisor should be disclosed to the affected bank, in particular to make the use of double haircuts (by the bank as well as by the supervisor) visible. In particular, the bank concerned should be informed if haircuts have been applied due to a lack of information on the options. In addition, it should be disclosed which qualitative factors were used and included in the adjusted or final ORC. It would be helpful to disclose which comparative values (e.g. haircuts, etc.) are used by peer groups.
Upload files
DK PP EBA GL ORC 20230301 fi .pdf
(28.25 KB)