Deutsche Bank

The draft Guidelines are generally appropriate, especially where they build on the SREP. However, we would flag - as we have before - the need for greater transparency on communicating SREP scores to banks whilst safeguarding the confidentiality of such scores. This will help banks to avoid requiring early intervention and, ultimately, resolution. We therefore also underline the importance of maintaining a continuous open dialogue with individual banks throughout the SREP review process, as well as in relation to the early intervention process.

We strongly welcome the statement in the guidelines that a determination that an institution is failing or likely to fail does not automatically lead to resolution and that no single element should automatically lead to the determination that a bank is failing or likely to fail. We also appreciate the recognition that macro-economic developments and market indicators should always be used in conjunction with other indicators, but suggest the elements should also consider the institution’s prospects for recovery and whether the failing or likely to fail determination is therefore temporary in nature. Requiring authorities to conduct (and therefore also document) a more “forward-looking” analysis will be important to avoid potential legal challenges, given the determination involves a high degree of discretion for the relevant authority.

This discretionary nature also means coordination between authorities is key, especially given the complexity of potentially having two authorities involved in each country involved.. As such, we also strongly welcome Title III on consultation and information exchange between competent authority and resolution authority. However, even in a simple cross-border resolution, with one host jurisdiction, there could be four authorities each making an independent determination. This significantly increases complexity with regards to decision-making at the point of resolution, and could even make agreement on the group resolution plan more challenging, if there is a lack of transparency on approach to resolution triggers between home and host authorities. Therefore, we suggest that the EBA goes further and also addresses coordination between home and host authorities. This also helps to avoid the risk that different triggers in different jurisdictions could result in disorderly resolution due to time inconsistency. This coordination will be as important in the case of a single point of entry as a multiple point of entry resolution, as the home authority will be reliant on the host authority to support their resolution action, even within the EU.
Yes, but as described above, we believe they should also address coordination between home and host authorities. Furthermore, we believe that the EBA Resolution Committee could promote such transparency and coordination. For example, it could encourage authorities to publish high level statements with more detail about their approach to such determinations, to help investors and firms better understand when resolution may occur. For example, the German BRRD implementing law empowers the Ministry of Finance to issue more detailed conditions for resolution, while the Bank of England recently published a document on its approach to resolution; these are very helpful developments. In addition, in the context of resolution planning, the EBA should also encourage authorities to discuss ex-ante as part of the group resolution planning process the circumstances under which they would consider a specific bank to be failing or likely to fail, and the point at which resolution action is likely or possible.
Yes, we agree that hypothetical examples are useful. However, we suggest that, rather than including them in the final Guidelines, alongside or shortly after their publication the EBA could consider publishing a complementary document which includes a high level description of different Member States’ approaches to deciding whether to place banks into resolution, including any variation in “failing or likely to fail” determinations and examples of when resolution may not apply. This would help promote transparency and support coordination between authorities.
We agree these are the right elements to look at but suggest they should be rephrased to be more forward-looking, based around the banks’ ability to reverse the situation in a sufficiently timely way. For example, the SREP score should not be sufficient on its own to invoke resolution, nor should the valuation. The element based on whether the recovery plan has failed should be further specified to avoid subjectivity. For example, rephrasing to read “the institution has activated its recovery plan and has exhausted all recovery measures that can feasibly be deployed in the current market circumstances in the relevant timeframe, in particular where the activation...” etc. Without the authority demonstrating robustly that it has considered whether there are no possible private sector alternatives, it potentially opens up the decision to legal challenge.

We are also concerned that while the resolution and supervisory authorities are generally required to coordinate, there is a lack of continuity between the two determinations. In addition, as not all Member States will require an additional determination by the resolution authority, we suggest combining sections 2. & 3. and framing the objective elements that the resolution authority should take into account as: i) its own analysis of those considered by the competent authority and ii) the additional elements on capital position, liquidity position and other requirements for continuing authorisation. This limits repetition, promotes continuity in the process and supports the requirement for coordination between the two authorities when determining if a bank is failing or likely to fail.
No. Apart from anything else, we disagree that point (e) relating to results of asset quality reviews is an appropriate element to include in these Guidelines. Any negative results will have to be incorporated into the capital position of the bank and should only result in a bank being considered failing or likely to fail when this would result in the outcomes described under 23. Point (e) is therefore redundant. It also overrides supervisory processes, which are the appropriate approach to reflecting outcomes of asset quality reviews in capital position. We also believe that no quantitative threshold for a significant decrease in asset value should be set, as it may not reflect that a bank is failing (e.g. if sufficient capital to cover the loss is available). In any case, what is a significant decrease will depend on the individual bank and authority’s approach to risk appetite and management.
As described above, we believe 24. points (a)-(d) can be covered by combining section 3 with section 2, and replacing them with “Where relevant under national law, the resolution authority should base its determination on the same elements as the competent authority, as far as these are known to the resolution authority”. Then, elements (f)-(j) are an additional set of elements relating to capital position that either the competent or resolution authority should consider, depending on responsibilities in national law. As described above, we believe element (e) should be deleted as it is redundant and overrides supervisory processes.

We agree that points (f)-(j) cover the right elements, but to make them more forward-looking, we suggest that “Additional elements ... when carrying out this determination include” should be rephrased to read “Additional elements should be considered by the relevant authority when carrying out this determination, insofar as they would result in depletion of the capital position to the point that it would infringe the institution’s own funds requirements and the institution is unable to address this in the relevant timeframe. Where relevant to the characteristics of the institution, these additional elements should include...” etc. Point (h) should align with point (j) and specify “significant non-temporary adverse developments”.
We believe the language in paragraph 26 needs to be strengthened to be more forward-looking about the prospects for recovery. This would also better reflect how the regulatory liquidity buffers operate under the CRR, which sets out the requirements for firms to have a plan for the timely restoration of their liquidity position where they breach either the LCR or the NSFR. As such, we suggest rephrasing the two bullet points to “Indefinitely incapable of meeting regulatory liquidity requirements, including requirements imposed according to Article 105...” and “indefinitely unable to pay debts and liabilities as they fall due”, This reflects that a breach of liquidity requirements may lead to a firm being unable to pay debts and liabilities, but this does not mean that they are failing or likely to fail, as management action or use of liquidity reserves could restore them.

As under capital position, sections 2 & 3 should be merged to reduce repetition and promote coordination and continuity. If so, points (a)-(c) would be deleted as redundant. Paragraph 27 should also be amended to read “Additional objective elements that should be considered by the relevant authority for the purposes of the determination that the institution is indefinitely incapable of remediating a breach of its regulatory liquidity requirements or paying its debts and liabilities as they fall due include...” etc. Point (d) is also largely redundant given it largely replicates paragraph 26. If (a) is retained it should be clarified to say “critical risks and indicators within the liquidity assessment of the SREP process”.

Points (e)-(k) would then be the additional elements the relevant authority should consider. In order to make these more forward-looking, some changes are necessary. Point (e) should amend “significant non-temporary adverse evolution of the institution’s liquidity buffer” to add “which would render the institution indefinitely incapable of paying its debts and liabilities as they fall due”. Assessments should be obliged to consider the bullet points, so “shall consider” should be used.

Point (g) should also refer to “a significant non-temporary adverse evolution” and also require the assessment to consider the bullet points. The final bullet point should specify “any irrevocable contingent obligation”. Likewise, point (h) should be “experiencing any non-temporary difficulties”.

Finally, point (i) should specify significant rating downgrades “of more than three notches”, as grounds for a determination of failing or likely to fail should exceed the LCR downgrade scenario.
We believe that governance arrangements are appropriate to include, as serious weaknesses in this area may justify withdrawal of authorisation, but we consider that in all cases, not only “most”, this should be in conjunction with other objective elements related to capital and liquidity. For governance weaknesses to justify withdrawal of authorisation, the Guidelines should specify that this should always be in conjunction with material deteriorations in the capital and liquidity position, although it may be the case that own funds and liquidity requirements are not yet breached.

We agree with the first two objective elements under Governance arrangements in paragraph 30, but the third element is currently very subjective. We appreciate that the EBA has included Box 2 to try and specify key “material deficiencies” in governance further, but these are very general and not necessarily linked to risk of failure. More objective and more consistent with the overall approach of the draft Guidelines would be to link this third element to the SREP score for internal governance and institution-wide controls. We therefore suggest rephrasing as “an accumulation of material deficiencies in key areas of governance arrangements, resulting in a SREP score of 4 for internal governance and institution-wide internal controls, where this would have serious prudential impact on the institution”. Box 2 should not then be necessary.
We agree that lack of operational capacity to provide regulated activities can lead to bank failure, as the resulting loss of trust can trigger “runs”. However, in order to justify withdrawal of authorisation, as required in the level 1 text, as with governance arrangements these elements should usually have to be linked to weaknesses and material deterioration in the capital and liquidity position.

We strongly welcome the recognition that these circumstances and events should only be considered where they are not contingent and cannot be addressed in a timely manner. However, they should only be considered to justify withdrawal of authorisation where they would also result in a loss of customer confidence and “runs”. As such, we suggest adding to “the institution becomes unable to make or receive payments” that this should be “on a non-temporary basis, and thereby unable to conduct its banking activities in the foreseeable future, leading to a loss of market and depositor confidence”.
Katie Melville