Primary tabs

Deutsche Bank

Dear Mr. Farkas,

Deutsche Bank welcomes the opportunity to respond to the EBA’s draft RTS setting out requirements for classes of instruments that may be used in variable remuneration in addition to equity instruments. To ensure that material risk takers’ variable pay is aligned with long-term risks to the financial institution, we believe that equity and similar instruments are the best way to ensure remuneration plays a role in incentivising good risk management. Not only does equity absorb losses first, but equity incentives play a long-recognised role in encouraging greater stewardship by employees, as they benefit from ownership rights.

Non-equity instruments can also play a useful role in remuneration where institutions or regulators wish to use them for operational or cost reasons - e.g. where the size or corporate structure of the bank makes equity-linked schemes difficult. In addition, using instruments able to absorb losses on a going-concern basis in remuneration will help to build up banks’ pools of additional capital and liabilities that are eligible and available for bail-in. However, we do not believe that they can or should replace the important role of equity instruments in aligning employees’ interests with those of shareholder owners of the firm.

The production of RTS by the EBA is therefore welcome as this developing market requires a harmonised approach across the EU. However, there is a risk that some features of current proposals will undermine the intention of Basel III to create consistent definitions that are transparent to the market. In particular, the final RTS should be harmonised with CRD IV / CRR on triggers for conversion, maturity requirements and write-up mechanisms. In addition, bearing in mind the desire to promote an active secondary market in banks’ loss-absorbing instruments, some adjustments are required to maintain investor appetite even where they are primarily issued for variable remuneration purposes. As the draft RTS currently stands, institutions will be better off using synthetic instruments. Our suggestions to address these issues are outlined in more detail in our response to the questions in the draft RTS below.

Yours sincerely,
Andrew Procter
Global Head of Compliance, Government and Regulatory Affairs
We do not believe that a trigger level of 7% is appropriate. The intention of Basel III and its EU implementation in the Capital Requirements Regulation (CRR) is to introduce a harmonised definition of capital which is well understood by investors in the market. Article 54 of the CRR has already introduced a “going concern” trigger of 5.125% CET1 for AT1 instruments to be automatically written down. Introducing a different going-concern trigger level for the purpose of variable remuneration will effectively create another tier of AT1 instruments, adding complexity for external investors as well as the issuer. In the case of externally issued instruments, a higher trigger could actually act to destabilise the bank. The act of writing down these higher trigger AT1 instruments would have a signalling effect for investors in AT1 instruments with the lower trigger, who would start to sell off - or short the bank’s equity - in expectation that they are next in line.

We also do not believe that 7% would reflect credit quality. Where we see these trigger levels in the market already, it reflects the caution of individual regulators about a new and evolving type of instrument. This is primarily the case in Swiss and other domestic regimes which have a very high loss-absorbing capacity requirement for major banks on top of the Basel III framework. The introduction of the CRR next year will create a consistent and well-understood definition and minimum requirement for AT1 and we expect the EU market to converge towards that level. The EBA should not deviate from the CRD framework.
Yes, as this would reduce complexity. By aligning with CRR and reducing complexity, losses can be absorbed more effectively and more easily, thereby facilitating appropriate incentives for variable remuneration. By harmonising the trigger levels - and setting it at 5.125% CET1 - this would simplify the use of instruments eligible for variable remuneration and allow institutions to choose the appropriate instrument for their issuance and remuneration strategy.

As it currently stands, the complexity and other restrictions in the draft RTS would have a significant negative effect on secondary market pricing and investor appetite for these instruments. It is questionable what investor appetite there would be for higher trigger instruments if the market norm is nearer to 5.125% (which is going concern) without banks paying a significant premium to issue these instruments. As such, institutions may only use synthetic instruments and while firms should have this possibility, the legal framework should not actively disincentivise issuance of variable remuneration instruments. These comments also apply to Articles 3 and 4.
See our response to Q2. We support one harmonised trigger level for all classes of instruments. An appropriate level would be 5.125% CET1 trigger set out in the CRR, to harmonise with the capital framework. Conversion / write-downs would occur in line with existing capital “waterfalls” and, once the EU Recovery and Resolution Directive is implemented in line with the hierarchy of claims which will be harmonised across the EU. This will ensure going-concern conversion and write-downs are appropriate to the seniority of the claim, which was a key investor concern around the introduction of the statutory bail-in regime. These comments also apply to Articles 3 and 4.
No - see detailed response provided under the next question
We do not believe that the cap on the distribution rate for instruments issued for the sole purpose of being awarded as variable remuneration is appropriate. The proposed definition, of 6 percentage points above inflation, is not appropriate to reflect arms-length pricing of bank instruments. Instead, it should be based on a market-based rate reflecting bank funding costs, plus an appropriate spread. The EBA has already considered such a measure in its consultation on draft RTS on own funds (Part III) under CRR, which proposed criteria for “broad market indices” to avoid the interest rate / dividend paid by institutions on floating rate capital instruments increasing when the credit standing of the institution decreases. These comments apply to similar provisions in Articles 3 and 4.
In general, we are concerned that there will be little appetite among external investors for these instruments if they are not harmonised with the CRR framework. The lack of a wide investor base for instruments with such a high conversion trigger will restrict the secondary market, affecting pricing compared with normal AT1 instruments. In addition, the knowledge that employee investors may exit en masse once deferral and vesting periods end may result in a permanent price pressure on the instrument from participating external investors.

If the trigger was aligned with the CRR (5.125%) then the capital instrument would be less expensive and more likely to be aligned with market pricing for AT1 instruments. If, however, the aim of setting the external investor requirement is to ensure arms-length pricing, then the relevance of maintaining the 60% threshold throughout the lifetime of the instrument is not clear. This would make monitoring and administration throughout the lifetime of the instrument very burdensome and negatively impact market liquidity as e.g. a buyback request from an external investor may have to be turned down to maintain the share of externally issued instruments. These comments apply to similar provisions in Articles 3 and 4.
No - see detailed response provided under the next question
No, as set out in our answers to Q2 and Q3, we believe the current proposed trigger events are not appropriate as they introduce complexity and cost to capital issuance and should be harmonised to a uniform going concern trigger of 5.125% CET1 in line with the CRR.
Yes, as the write down and write-up mechanism is fully aligned with that for AT1 in CRR. However, as outlined in Q7 and Q8, the trigger event should be adjusted to refer to CET1 rather than Tier 1 capital.
No - see detailed response provided under the previous question
Yes, as it creates further flexibility for institutions to choose appropriate instruments - including synthetic ones - for use in variable remuneration. In particular, such an instrument can be used on a subsidiary level where instruments are directly issued by the parent institution but which may not be recognised for capital purposes due to local regulation.
Yes, as these instruments are to be used in variable remuneration we agree they should reflect the institution’s credit quality as a going concern.
Yes, we agree that other instruments should be subject to the same write-down and write-up features as for AT1 / T2 instruments, including conversion to CET1.
No, temporary write-down should also be allowed in line with AT1 and T2 requirements and to ensure that other instruments are also “going concern” instruments. The write-up mechanism should, as with AT1 and T2, only happen when capitalisation is sufficient and subject to the same restrictions as AT1 and T2, including distributions counting as reductions in CET1.
No - see detailed response provided under the next question
No, as set out in our answers to Q2 and Q3, we believe the current proposed trigger events are not appropriate as they introduce complexity and cost to capital issuance and should be harmonised to a uniform going concern trigger of 5.125% CET1 in line with the CRR. It is especially important to align the trigger for conversion of other instruments with AT1 and T2 triggers, rather than introduce yet further complexity around AT1 / T2 linked instruments.
As in our response to Q5 and Q6, the conditions in Article 4 paragraph 1 point (f) around caps on distributions and minimum external investor requirements need adjusting to reflect arms-length pricing and to allow for changes in investor composition over time.

Other than adjusting the trigger event to refer to CET1 instead of total capital, as outlined in Q15 and Q16, we agree that the other conditions will align instruments with AT1 and T2.
Yes, as it aligns with the write-down mechanism for AT1 / T2 instruments. However, it is also important to include a temporary write-down mechanism to align with AT1 and T2 further and to ensure that other instruments are “going concern”. The write-up mechanism should, as with AT1 and T2, only happen when capitalisation is sufficient and subject to the same restrictions as AT1 and T2, including distributions should count as reductions in CET1.
With respect to Article 6, requirement are clear, but as outlined above write-downs should be able to be temporary, not “irrevocable”, and that falling below a CET rather than total capital ratio level should be the trigger event. This is important to align with the CRR, with treatment of AT1 and T2 and to ensure that remuneration instruments are “going concern” instruments.

With respect to Article 7, although the EBA only refers to the matter in the explanatory box, we believe that all instruments eligible for variable remuneration should be able to be called after the applicable deferral or retention period has lapsed (subject of course to CRR requirements for AT1 / T2 instruments). As the EBA states, it should also be possible to call the instrument and replace it with one of similar quality. This should meet all the relevant requirements for use as part of variable compensation, including that it does not increase the value initially granted to the respective employee as set out in paragraph 3. This will help ensure that employees can liquidate their deferred compensation after the vesting period.
To a certain extent - see detailed response provided under the next question.
We agree with the EBA’s assessment of benefits from establishing harmonised classes of AT1, T2 and other instruments for the purposes of variable remuneration. The EBA is also right to point to increase issuance and ongoing administration costs, but fails to take into account several other aspects which could contribute to costs:

- Issuance costs in terms of reduced investor demand or investor base for such instruments if trigger levels are higher than those used in the CRR;
- Ongoing costs from permanent price pressure on variable remuneration instruments if their characteristics differ substantially from the existing capital framework;
- Potential damage to the institution’s capital base from accelerated investor flight should AT1 investors sell following write-down of variable instruments set at a higher trigger point;
- Increased administrative cost from having to maintain the proportion of external investors over the lifetime of the instrument; and
- Increased costs from the lack of flexibility to call and replace instruments with similar ones or to write up investor claims should the institution return to financial health.

These costs will vary according to where the trigger point is set, whether the write-down mechanism is temporary and the conditions attached the instruments. Many of these can be eliminated if the final RTS align the trigger to write-down AT1 instruments with CRR at 5.125% CET and extend this trigger and a write-up mechanism to all instruments used for variable remuneration. A standard “going-concern” trigger and mechanism will ensure it is clear to investors at what point and how they are to be impacted, maintaining investor appetite in the institution and allowing it to recover effectively. In addition, better market-based pricing of internal instruments will ensure remuneration rules cannot be circumvented. Lack of coherence or complexity in the market will undermine these goals, which were central to the development of Basel III and its subsequent implementation in the EU through CRR.
Yes - see detailed response provided under the previous question.
Deutsche Bank