Response to eBA consults on technical standards on impracticability of contractual recognition of bail-in
Go back
An example in the context of trade finance may be a governmental body located in the third country issuing a tender and requiring that the “national” standard guarantee text (e.g. bid bond) be used.
We actually believe that it is generally not very likely that authorities will specifically/expressly prohibit that counterparties accept bail-in contractual recognition clauses other than in very exceptional cases or for very specific capital instruments, or that any third country jurisdiction laws, ordonnances or regulatory rules specifically/expressly declare such clauses to be illegal or ineffective.
However, some specific examples were signalled by our Members: India, China, probably Singapore and South Africa. In the case of India, the Reserve Bank of India (RBI) has stated in 2017 that foreign banks incorporated in EU jurisdictions (i.e. obliged to follow the BRRD) and operating in India, shall ensure that their liabilities in India do not include any contractual term that has the effect of bailing-in the liabilities of the bank in the event of resolution.
It is also worth considering that the RBI is not a resolution authority, since there is not yet a resolution scheme implemented in India. This case also demonstrates that the situation where an authority imposes instructions not to include contractual bail-in provisions should not be restricted to resolution authorities. In many third countries, such resolution authorities do not yet even exist.
The more likely and practically relevant scenarios are the following:
• The authorities (resolution or other) impose restrictions on the acceptance of contractual recognition (both in respect of Art. 55 and Art. 71a BRRD2) implicitly or informally.
• The clauses are deemed to conflict with or to outrightly contravene local mandatory laws (such as investor or consumer protection laws) or general legal principles (e.g. governing unfair term), especially in view of their one-sided nature and the fact that the content of the clauses can effectively not be negotiated by the counterparty. For instance, in China, adding the bail-in clause may be interpreted to imparity clause by the court and cause law conflicts between China Bankruptcy Law and BRRD.
Local authorities and/or local laws restrict in some cases or circumstances the extent to which counterparties, especially public entities, are permitted to subject themselves to the powers of a foreign public authority or are allowed to give effect to extraterritorial powers of a foreign authority. Local authorities or governments have requested/encouraged the industry to develop standards and templates, for example templates of local guarantees/or counter-guarantees, in order to ensure a better protection of their national interests. This is the case in Algeria, where, by order of the government in the 1990’s, the banking industry has implemented templates of local guarantees and counter guarantees and these cannot be amended in practice. This specific case shows that the wording of condition b) could be extended to include the cases where the institution is unable to amend a contract/template because of a third country practice related initially to an authority decision.
• These potential limitations and restrictions will in most cases not be definitive/clear-cut (sometimes simply because the issue has never been tested in courts) and some are likely to be invoked only in court proceedings against the effects of the eventual resolution actions.
For these reasons, we believe that the condition addressed in Art. 1 (1) (a) and (b) of the draft RTS could turn out to be of little practical relevance.
In any event, it should not matter which authority prohibits the acceptance of contractual recognition clauses, since any prohibition would always be an obvious and insurmountable impediment.
We therefore believe that Article 1 should be amended to cover any authority which is competent to make such a declaration as well as other cases of impracticability resulting from legal or practical impediments, without being limitative.
As to the broader need to provide for a more open/flexible approach in order to address the very real situation that institutions will effectively not be able to impose such contractual clauses on their counterparties and that it will be impossible to clearly identify all relevant types of conditions constituting impracticability, see our comments on Question 4 below.
This would include (non-exhaustive list):
• Any contract/agreement which does not cover an outright payment obligation and where the “liability” which could be subject to any bail-in would only be a secondary liability (damage claims or equivalent monetary secondary claims under the laws of the relevant jurisdiction) of an undetermined value/amount (which would be decided by a court for example). As to the need to clearly exclude such secondary claims from the scope of Art. 55 (limiting the scope to outright/direct/primary payment claims) - see also our comments below on Question 4.
• Deposits/certificates of deposits (concluded/confirmed via telephone or SWIFT-messages on the basis of established market practices and standards and therefore without a contractual documentation).
• Interbank guarantees loans/lines of credit (concluded/confirmed via telephone or SWIFT-messages on the basis of established market practices and standards and therefore without a contractual documentation).
• Spot transactions in securities or FX (concluded/confirmed via telephone or SWIFT-messages on the basis of established market practices and standards and therefore without a contractual documentation).
• Similar types of transactions where the terms are based on established market practices and are concluded/confirmed via SWIFT-messages or similar services.
• Guarantees or counter-guarantees issued for non-trade finance transactions (raising similar issues as for trade finance transactions) and which are widely often used in domestic commercial transactions/projects between clients and contractors in third-countries (also confirmed via SWIFT messages) such as tender bonds, advance payment bonds, retention bonds, performance bonds or warranty bonds.
• An assignment agreement by means of which an existing lender transfers part of or all its position in a syndicated facility agreement to a new lender (the transfer agreement will be executed by the assignee, the assignor and the facility agent but not by the borrower and the other parties under the facility agreement) and so therefore the borrower (and the other parties to the syndicated facility agreement) cannot give recognition that they may be bailed-in, or
• An adherence letter to a non-disclosure agreement (NDA) will not be signed by the beneficiary of the underlying NDA (but by the original recipient under said document and the new recipient) and so therefore the counterparty cannot give recognition that they may be bailed-in.
- At page 9, in the box number 6, the RTS indicates that the liabilities that are contingent to a breach of contract are not in the scope of the conditions of impracticability.
However, according to our Members, liabilities arising out of a breach of contract are not to be excluded from the impracticability criteria for various reasons:
• This approach seems to be in contradiction with the intention of the EU legislator formulated in recital 26, par. 2, of BRRD2: “For example, under certain circumstances, it could be considered impracticable to include contractual recognition clauses in liability contracts in cases where it is illegal under the law of the third country for an institution or entity to include such clauses in agreements or instruments creating liabilities that are governed by the laws of that third country, where an institution or entity has no power at the individual level to amend the contractual terms as they are imposed by international protocols or are based on internationally agreed standard terms, or where the liability which would be subject to the contractual recognition requirement is contingent on a breach of contract or arises from guarantees, counter-guarantees or other instruments used in the context of trade finance operations.”
Indeed, we understand that contingent liabilities such as NDAs, Reps and Warranties etc. which cause a financial compensation to be required only in the case of a contract’ breach, are hereabove included. As consequence, for such liabilities (i.e. liabilities contingent to a breach of contract) no inclusion of a bail-in clause should be necessary.
• The potential liabilities are not quantifiable as such and usually are being settled by courts and tribunals without any control on the outcome.
• The nature of these liabilities is purely contingent, they usually do not even come to existence.
• Most contingent liabilities will be completely irrelevant for a bail-in and it would be unreasonable to burden the institution with the obligation to impose a contractual recognition clause if such clause will not be of any practical relevance. This holds particularly true in case of contingent liabilities with mirroring counterclaims: in many cases, the contingent liability vis-à-vis the counterparty is counterbalanced by a mirroring counterclaim against a client; any bail-in into the contingent liability will therefore directly reduce the corresponding counterclaim, resulting in a zero-sum game.
The EBF therefore would suggest a general waiver, given without any detailed notification to the resolution authorities.
- At page 8, in box number 1, the consultation paper indicates that “the request of the counterparty to renegotiate the contract and/or an increase in pricing or a refusal by the counterparty to agree to be bound by a contractual bail-in recognition clause” is not a reason for impracticability. Our Members do not agree with this statement. In the context of the service contracts that a bank negotiates, mutual agreement plays a role and banks will always try to negotiate in order to include the bail-in clause. If the impracticability is defined as not ‘legally or otherwise’ possible to include the conditions in the agreement, it is a fair observation that a final refusal by a contractor effectively means that the 'otherwise' condition cannot be fulfilled. After all, both parties must agree. Same reasoning if the contracting party only wants to include the bail-in clause if important clauses/guarantees are weakened to the disadvantage of the bank or if this is used to contract at higher prices. If there are no alternatives on the market for the bank, it is not possible to include the bail-in clause without substantially weakening its position.
- In terms of low value contract/liability under box 4 (materiality threshold), we signal (SEE ALSO RESPONSE TO Q10 and 11) that a materiality/value threshold would be an extremely useful instrument to reduce unnecessary burdens and allow institutions, as well as resolution authorities, to concentrate their efforts on the practically relevant liabilities/instruments. In the event of a bail-in, the resolution authorities will – in the interest of efficiency - necessarily focus their efforts on higher value liabilities without individually pursue liabilities of low value based on contractual recognition clauses which would have to be enforced against counterparties in third countries, simply because the cost will significantly outweigh the benefits. This, of course, would not preclude the authority from simply applying the bail-in to such low-value/smaller liabilities based on their resolution powers and without resorting to contractual recognition.
First and foremost, and as already indicated above, it is important to underline that it will not be possible to identify all potential conditions of impracticability. It needs to be recognised that the impracticability can result from a combination of circumstances which cannot be easily categorised or reduced to specify predetermined types of conditions or cases.
In this context, it is necessary to take into account that the experience over the past years has clearly demonstrated the difficulties EU institutions have had with such clauses and the very considerable competitive disadvantages they face in the international markets because of this requirement (particularly in view of the still extremely broad scope which is much broader than that of any other jurisdiction). As already mentioned in our general remarks, a too restrictive application of the newly established possibility to waive the need to introduce contractual recognition clauses may force EU institutions to withdraw from certain markets or transactions with third country counterparties. We do not see how such an effect is supposed to be justified in the name of a formal improvement of the resolvability, especially since the absence of a contractual recognition clause does not preclude the authorities from applying the bail-in tool.
Against this background, we would suggest a much more flexible and open approach regarding the specification of conditions which indicate a legal or other impracticability.
This could be best achieved by setting out a list of non-exhaustive examples of such conditions. A similar solution would also, in our opinion, enable the resolution authorities - together with institutions - to better take into account developments and practical experiences made.
At the very least, we believe it is necessary to set out more clearly that the listed conditions do not preclude an institution to refer to other conditions, indicating such (de facto or de jure) other impracticability cases.
As regards the specific conditions suggested in the draft RTS, and as already indicated above, we believe that the following should also be considered as a condition indicating impracticability:
• Contingent liability.
• Low-value contract/liability. Examples are the so-called "click-wrap" agreements, (i.e. agreements that a bank enters into online and where a set of conditions must be accepted; for instance, the purchase of licenses of very limited value).
As already also indicated (Question 2), we strongly believe that it is essential to expressly clarify that liabilities/contracts which are not direct/primary contractual payment obligations and where the value/amount will regularly be undetermined, are – as such - not captured by the Art. 55 requirements. Anything else would require institutions to notify such liabilities/contracts clearly producing unnecessary burdens for both institutions and authorities, when receiving these notifications.
Depending on the specific circumstances and the composition of the portfolio of liabilities, three months is likely to be very challenging. With the current draft, our Members, especially those with trade-finance activities in third countries, would have to notify individually tens of thousands of contracts over the course of a year. In this context, it has to be taken into consideration that institutions are currently facing numerous challenges requiring far-reaching operative changes and adjustments to existing procedures and contractual arrangements, including the ongoing benchmark-replacement projects, the impending coming into force of Brexit and the parallel implementation projects in respect of Art. 71a BRRD. We therefore see the need for a longer implementation period, ideally combined with a more flexible/staggered approach allowing for a risk-based implementation by institutions. Of course, clarifications clearly limiting the scope of potentially affected contracts/liabilities (such as the suggested clarification concerning the exclusion of indirect /secondary payment obligations) would greatly help the implementation process.
It would also be much more practicable if the resolution authority would determine upfront exempt categories of liabilities – i.e. types of liabilities for which the inclusion of the contractual bail-in clause is deemed impracticable – which would eliminate the need to notify all contracts separately.
As already indicated in our response to Question 1, we believe that the two situations addressed under lit. (a) and (b) will not be of much practical relevance: the more probable and practically relevant situation will be that the imposition of contractual recognition clauses on counterparties is likely to conflict with local rules or practices such as investor (or consumer) protection rules, or mandatory legal principles concerning clauses which are deemed unfair/unilaterally imposed (without allowing for any meaningful negotiations) and attempt to extend extraterritorial effect to measures by foreign public authorities.
In many cases, it will therefore not be possible to determine with certainty whether the clauses can effectively be implemented or not. At best, it will be possible to conclude that there is a considerable likelihood of challenges or, conversely, that it is more likely than not that the clauses will be effective. However, there will never be any legal certainty in this regard.
One way to address this issue would be to take this situation into account, if not as a condition which directly implies unconditional impracticability, at least as a relevant criterion for the assessment that the non-inclusion of the clauses does not adversely affect the resolvability.
In lit. (a) the reference to “law” should also include any rule or regulation (whether or not having the force of law) issued by an authority or court of the third country and applicable to the liability.
Art. 1 (1) (c) - Liability arising out of instruments or agreements concluded in accordance with and governed by internationally standardised terms or protocols which the institution or entity is unable to amend
The condition addresses the highly pertinent situation that existing and established standard terms and practices in the relevant market effectively preclude institutions from including contractual recognition clauses into the applicable terms of the agreement. However, and in order to ensure that this condition does not become practically redundant, the additional condition that the institution was unable to amend these terms cannot be interpreted too rigidly and should not cover a purely theoretical possibility to amend terms: for example, in general transactions on the basis of established terms and practices, ICC rules for guarantees and letters of credit (but not limited to these!) allow for some level of individual amendments. However, the market participants are - in practical terms - not able to unilaterally impose the clauses. The inability to amend in practice such kind of instruments or agreements (whether or not governed by ICC rules) should be expressly recognised in this §, in compliance with Recital 26 of BRRD2.
Art. 1 (1) (d) - Liability governed by contractual terms to which the institution is bound pursuant to its membership of, or participation in, a non-Union body, including financial market infrastructures, and which the institution or entity is in practice unable to amend
The comments to lit. (c) apply correspondingly.
Art. 1 (1) (e) - Liability is owed either to a commercial or trade creditor and relates to goods or services that, while not critical, are used for daily operational functioning and where the institution or entity is in practice unable to amend the terms of the agreement concluded on standard terms
The condition is too rigid and restrictive: first, as already mentioned, it should be clarified that contracts/liabilities which are not direct/primary payment obligations are excluded as such: the contractual recognition requirement can only apply to agreements which provide for payment obligations and cannot apply to any secondary /indirect monetary obligations which may or may not arise out of, or in connection with, the relevant agreement and where the value/amount is not even known. Anything else would effectively mean that any legal or even non-contractual agreement, yet contract-like relations (depending on the applicable law), would be captured by Art. 55 BRRD. This would be a clear and unnecessary overreach and also result in considerable uncertainty.
In addition, and as in the above cases, the additional condition of an inability of the institution to amend the terms is particularly unnecessary in the context of these agreements/contracts: it simply cannot be expected that institutions need to demonstrate that they tried to include contractual recognition clauses in types of agreements where it is already unclear whether a liability will ever exist.
In accordance with Recital 26, we would suggest adding another condition: “f) the liability which would be subject to the contractual recognition requirement is contingent on a breach of contract” (SEE RESPONSE TO Q3).
Art.1 (2)
The requirements established under par. 2, determining when an institution can deemed to be unable to amend an agreement, are far too restrictive. As already pointed out in our comments on lit. (c) to (e) above, this additional condition is unhelpful and should be deleted: experience has shown that institutions face great difficulty in unilaterally imposing contractual recognition clauses on counterparties where this is not customary in the relevant market and for the relevant products. These difficulties are not limited to cases where the counterparties set the terms. Rather, it also applies to all cases where the terms are grounded on customs and market practices.
Besides, our Members propose to add to the RTS a further generic criteria (for instance, in the third paragraph in article 2) where the bank could demonstrate that the bail-in of an instrument has no impact/no benefit on its resolvability. In this case, the resolution authorities shall not require the inclusion of the bail-in clause. This condition would not put at risk the process of notification since the threshold provides itself, for liabilities below the threshold, a line for a resolvability assessment. Furthermore, the BRRD itself provides specific criteria to assess the impact on resolvability based on a 10% threshold within a liability class.
Also, the draft RTS does not provide indications for situations where the RA would require the inclusion of such clause whereas it is operationally impossible (e.g. does the bank then need to terminate the contract or accept a compliance breach?).
We understand the proposal regarding Art. 2 of the RTS on the cases where the resolution authorities have to require or may require institutions to include contractual recognition clauses – together with Art. 1 of the draft RTS ¬– to effectively provide for the following avenues for institutions to address the issue that they will not be able to impose these contractual recognition clauses on their counterparties:
• The draft RTS set out two different avenues regarding the permissibility of the non-inclusion of contractual recognition clauses:
o The first one is based on Art. 1 of the draft RTS, in the form of a positive determination of impracticability which automatically and irrefutably makes it permissible for institutions to omit the inclusion of contractual recognition clauses.
o The second one is based on Art. 2 of the draft RTS, setting out criteria under which the authorities either have to or can require an institution to include the contractual recognition clauses. Under this second approach, institutions are effectively allowed to omit the inclusion of the contractual recognition clauses until/unless the authorities require them to do so on the basis of the criteria set out in Art. 2 of the draft RTS.
This Art. 2 approach distinguishes between two sets of circumstances which effectively mean the following for institutions:
• Art. 2 (1), as currently drafted, would effectively preclude institutions from omitting contractual recognition clauses where the amount of the liabilities in question is equal or higher than EUR 20 million, or where the maturity is equal or higher than six months. This would effectively mean that a liability with a longer maturity than 6 months and regardless of the amount, or in an amount of EUR 20 million and regardless of the maturity, would always be subject to the contractual recognition requirement.
• Art. 2 (2) would effectively allow an institution to omit the contractual recognition clauses as regards liabilities below the thresholds of Art. 2 (1) of the draft RTS, except where the resolution authorities determine that the omission to include the clauses significantly affects the resolvability of the institution (regarding the thresholds approach in Art 2 (2) see also answer to Q11).
While we understand the approach, we believe that the resulting framework for institutions is far too restrictive and formalistic and will not resolve the very real and considerable problems institutions are facing.
By narrowing down the scope of liabilities which may effectively benefit from Art. 2 of the draft RTS (i.e. the possibility to omit contractual clauses to liabilities below EUR 20 million with a maturity of less than six months), a significant portion of other types of liabilities with longer or unclear maturities, which are irrelevant for a bail-in and/or where the failure to include contractual recognition clauses would not adversely affect resolvability in any meaningful way, will continue to be subject to the contractual recognition requirements.
A clearer focus on the impact on the resolvability is essential to grant the authorities and the institutions the necessary flexibility to ensure that the contractual recognition requirements under Art. 55 BRRD do not result in unreasonable burdens for institutions, authorities and counterparties, without actually improving the resolvability.
As already indicated above, this would have unreasonable consequences not only because of the resulting considerable burdens and competitive disadvantages for EU institutions in international markets, but also considering that impeding the access of EU institutions to essential international markets, or even requiring a withdrawal from such markets, will not improve the resolvability of institutions in any way.
We therefore believe that a more flexible and open approach which focusses on the actual impact on resolvability would be the only reasonable and appropriate approach.
One element could be – as already suggested in responses to earlier questions – a non-exhaustive list of conditions/examples of impracticability under Art. 1 of the draft RTS.
Specifically referring to the thresholds (in terms of liability/contract amount or maturity) introduced under Art. 2 of the draft RTS, we would like to also signal the following:
• These thresholds create an unlevel-playing field among EU banks: indeed, large banks will be penalised as they have more high-value contracts than smaller banks.
• These thresholds are not part of the level 1 text and reduce unnecessarily the flexibility of judgement of the RA in their assessment of resolvability.
Therefore, they should be removed from the RTS in their current form, i.e. as a way to automatically trigger the inclusion of the bail-in term. Materiality thresholds (in terms of maturity and amounts) should be used as a way to waive the obligation to notify the RA for liabilities below them. Moreover, considering the various sizes and capital situations of the banks, the amount-based threshold should be increased for larger banks and maturity-based threshold should also be raised:
• In terms of amount, a more proportionate approach should be used with the use of higher thresholds (or a % of the balance-sheet for instance) for the largest banks.
• In terms of maturity, it seems quite unreasonable not to consider the very short duration contracts (non-renewable) as automatically outside the scope of the art.55 requirement, taking into account the time for the RA to assess whether the inclusion of the bail-in clause is deemed necessary or not (3 to 9 months).
The assessment of the maturity of renewable contracts should also be clarified. For instance, a threshold of 3 years would make much sense given the nature of the liabilities in the scope of the impracticability conditions currently considered.
In any case, these threshold conditions should be independent, and not cumulative conditions to trigger the flexibility of the RA described in the second paragraph of Article 2. The latter explains that the resolution authority may require the inclusion of the bail-in clause when none of the conditions of impracticability described in Article 1 are fulfilled and when “the agreement’s or instrument’s value creating the liability is less than EUR 20 million and the remaining maturity of that agreement or instrument is shorter than 6 months.
We would strongly recommend the EBA that in this paragraph (p.17 of the CP) both conditions are made non-cumulative, which means that the flexibility of the resolution authority would be possible either when the amount is less than EUR 20 million or the residual maturity is shorter than 6 months. This would also alleviate the burden of the very constraining notification system proposed in this consultation as already mentioned above.
We consequently think the reference to the fact that “the institution cannot notify contracts for impracticability based on their value or maturity” (page 10 of the draft RTS), should for operational reasons be omitted.
Finally, we consider unclear whether one document or every principal document in a financing transaction should include a bail-in clause. Some industry guidance suggests all security documentation should include it in addition to the principal facility agreement. This potentially means more cost, more legal time and more negotiation for no additional benefit.
Experience has shown that the imposition of the contractual recognition clauses is very burdensome and time-consuming (for both parties). Rigid timetables will not be helpful in this context. We therefore strongly urge EBA to consider a significantly more flexible and risk-based approach.
In terms of implementation, we would welcome further clarifications on the timeframe for the resolution authorities to require the inclusion of the clause, specifically referring to the time given to the Authorities to take their decisions and the time given to the banks to comply with their request.
Besides, further information on grandfathering provisions for existing transactions would be appreciated. In this respect, we would like to be sure that new rules will only apply to the new transactions flow, and that for the stock of transactions banks will have at least one year for implementation after the entry into force of the RTS.
Finally, the answer of the IRT on a waiver request may take place after the sign-off of the contract. Should it be the case, the bank would be put in very awkward positions in front of its clients. A fast track process would be very much welcome to avoid such situations and ensure that contracts that need to be signed swiftly can also be processed by the resolution authorities at an equivalent speed. This would be especially relevant for banks’ trade finance activities which consist in a high volume of transactions in any given year, and which compete against US and Asian players which are not subject to the same rules.
The time necessary for the institutions to comply with these requirements depends also on the way in which the RAs will ask for the template submissions (if the submission is required through XBRL, it will probably take even more time to be implemented).
The tables imply that institutions may only notify cases of impracticability which can be connected to any of the listed conditions. As already indicated, the conditions suggested in the draft are too narrow and/or of little practical relevance. It is therefore necessary to allow for a notification of other conditions and this needs to be reflected in the tables.
Individual ID for liabilities:
The proposal assumes that at least some liabilities will and can be notified on an individual basis. Considering the very broad scope of Art. 55, which captures a vast array of liabilities (even if one - as strongly urged – excludes agreements which do not provide for direct/primary payment obligations), many of them will need to be notified as it will be impossible for institutions to include contractual recognition clauses. It would be extremely difficult to identify all these liabilities individually. It is thus unlikely that institutions will make regular use of an individual notification but would prefer to rely on categories defined by the resolution authority as per Article 55(7) BRRD.
Material amendment:
The purpose of a notification of a material amendment is not clear: as long as the amendments do not change the liability in such a way that the original assessment no longer applies (primarily if the amount of the liability is changed and now breaches the thresholds because of the change), amendments will be irrelevant and a requirement to notify them would be clearly unreasonably burdensome. This should be clarified, i.e. by confirming that material amendments for the purposes of the notification requirements are only amendments which substantially increase the amount of the liability.
Legal opinions:
The proposal assumes that the institutions will, at least in some cases, obtain legal opinions confirming the impracticability. Such legal opinions are not a requirement set forth by BRRD2.
Legal opinions are, of course, an important instrument to analyse the effectiveness and enforceability of standard agreements and clauses used by institutions. However, in this context, the opinions would not be relevant. First, most of the conditions of impracticability are not of a legal nature and can thus not be addressed by a legal opinion. To the extent that the impediment is indeed a legal one (Art. 1 (1) (a) and (b)), this would, however, have to be a negative opinion (confirming the illegality/ineffectiveness of a clause).
Many of the reasons will be practical and matters of fact which arise either in the industry sector, or in the countries in which the borrower and financial institutions are operating. They are not legal matters which a law firm could opine on. If an opinion were needed, a law firm would be likely to specifically exclude these matters from the scope of its opinion and liability.
As already indicated above, we can expect that only in exceptional cases will a clause be inadmissible/illegal/prohibited as such. The more relevant case will probably be that the legal analysis results in an inconclusive assessment identifying existing legal risks and uncertainties.
Insolvency ranking:
This item makes sense for liabilities governed by third-country laws and issued by third-country subsidiaries. For those issued by branches, the ranking in the European parent’s creditor hierarchy should apply. We presume that a specific chart of insolvency ranking (0080) in the third countries will be communicated by the RA.
This sentence, read alongside the Recital 26 of BRRD2, which reads “it could be considered impracticable to include contractual recognition clauses in liability contracts […] where the liability which would be subject to the contractual recognition requirement is contingent on a breach of contract”, suggests only financial liabilities are considered by resolution authorities, which is absolutely logical regarding the objectives of article 55 BRRD, i.e. resolvability of the entity. Nevertheless, and despite the efforts made from the banking sector on that subject since 2016, and to the best of our knowledge, the term “liabilities” is not defined. It is therefore not possible for entities to decide to introduce the bail-in clause only in agreements containing financing liabilities (as opposed to any other kind of liabilities such as respect of representations and warranties for example). As communicated above, it is therefore necessary to exclude with certainty all non-financial liabilities, which would for instance enable entities to avoid painful and useless negotiations to include a bail-in clause in non-disclosure agreements (NDAs).
• Does the estimated amount column imply that the volume of liabilities needs to be monitored to remain within the indicated amount? And how long the estimation is supposed to be valid?
• We presume that ‘Contracts/instruments creating new liabilities’ means: the contracts under which new liabilities may arise, and not the individual transactions or drawings in execution of those contracts.
• What about new contracts/guarantees? Should they first be reported before the bank can deliver the guarantee/conclude the contract? In practice, this will not work.
• Re. the final maturity date (0030): what in case of contracts/instruments valid until further notice (initially agreed or issued for an indefinite period of time)?
• We presume that a specific chart of insolvency ranking (0080) across the third countries will be communicated by the RA (see also Q15), to cover the case of subsidiary liabilities.
• Re. governing law (0120): for certain trade finance instruments (such as letters of credit) it is not market practice to insert a clause dealing with governing law.
• More information about the “legal opinion” (0160) that seems to be required would be welcome. It apparently overrides the level 1 text.
• Re. counterparty – name (0170): this is problematic, because this may lead to problems re. banking secrecy and the duty of confidentiality of a bank.
• Governing law: in the case of trade-finance liabilities (i.e. guarantees and standby letter of credit), many third-country contracts are based on ICC rules (e.g. URDG458, URDG758 or ISP98). Counterparties for such contracts have access to the ICC dispute resolution mechanism which rely on the ICC International Court of Arbitration. Therefore, banks’ IT systems record neither the underlying governing law nor the competent court.
• There is general lack of clarity on the time necessary for the institutions to comply with these requirements (see Q14), which depends also on the way the RA will ask for the templates’ submission (e.g. if the submission is required through XBRL, it would probably take even more time to be implemented).
Regarding Template N 01.02, it is highly preferable and much more practicable that the resolution authority will determine “exempt categories”.
Regarding Template N 02.00 (Liability Insolvency Classes), it is very burdensome to fill in this template (cf. “all outstanding amounts of liabilities governed by the law of a third-country”).
In addition, no comparison to the cost of a proper equivalent system (between EU and non-EU resolution regimes) has been made. The balance between the gain in terms of legal security and operational burden/costs is obvious, for instance, on the securities and debts, where we welcome the article 55’s effect. On the contrary, for some liabilities, this is very cumbersome and its benefits deeply limited (contingent liabilities when depending on the breach of a contract, non-monetary liabilities, etc.).
Finally, no assessment regarding the absence of a level-playing field between EU entities, subject to this requirement, and non-EU entities has been made.
In our opinion, the proposed draft of IA is therefore missing some crucial points, which may have led to a more comprehensive assessment of impracticability.
Q1. Are there any third country authorities, other than resolution authorities, that might impose instructions not to include the contractual bail-in recognition term?
FOR OUR GENERAL REMARKS, PLEASE SEE THE FILE ATTACHED.An example in the context of trade finance may be a governmental body located in the third country issuing a tender and requiring that the “national” standard guarantee text (e.g. bid bond) be used.
We actually believe that it is generally not very likely that authorities will specifically/expressly prohibit that counterparties accept bail-in contractual recognition clauses other than in very exceptional cases or for very specific capital instruments, or that any third country jurisdiction laws, ordonnances or regulatory rules specifically/expressly declare such clauses to be illegal or ineffective.
However, some specific examples were signalled by our Members: India, China, probably Singapore and South Africa. In the case of India, the Reserve Bank of India (RBI) has stated in 2017 that foreign banks incorporated in EU jurisdictions (i.e. obliged to follow the BRRD) and operating in India, shall ensure that their liabilities in India do not include any contractual term that has the effect of bailing-in the liabilities of the bank in the event of resolution.
It is also worth considering that the RBI is not a resolution authority, since there is not yet a resolution scheme implemented in India. This case also demonstrates that the situation where an authority imposes instructions not to include contractual bail-in provisions should not be restricted to resolution authorities. In many third countries, such resolution authorities do not yet even exist.
The more likely and practically relevant scenarios are the following:
• The authorities (resolution or other) impose restrictions on the acceptance of contractual recognition (both in respect of Art. 55 and Art. 71a BRRD2) implicitly or informally.
• The clauses are deemed to conflict with or to outrightly contravene local mandatory laws (such as investor or consumer protection laws) or general legal principles (e.g. governing unfair term), especially in view of their one-sided nature and the fact that the content of the clauses can effectively not be negotiated by the counterparty. For instance, in China, adding the bail-in clause may be interpreted to imparity clause by the court and cause law conflicts between China Bankruptcy Law and BRRD.
Local authorities and/or local laws restrict in some cases or circumstances the extent to which counterparties, especially public entities, are permitted to subject themselves to the powers of a foreign public authority or are allowed to give effect to extraterritorial powers of a foreign authority. Local authorities or governments have requested/encouraged the industry to develop standards and templates, for example templates of local guarantees/or counter-guarantees, in order to ensure a better protection of their national interests. This is the case in Algeria, where, by order of the government in the 1990’s, the banking industry has implemented templates of local guarantees and counter guarantees and these cannot be amended in practice. This specific case shows that the wording of condition b) could be extended to include the cases where the institution is unable to amend a contract/template because of a third country practice related initially to an authority decision.
• These potential limitations and restrictions will in most cases not be definitive/clear-cut (sometimes simply because the issue has never been tested in courts) and some are likely to be invoked only in court proceedings against the effects of the eventual resolution actions.
For these reasons, we believe that the condition addressed in Art. 1 (1) (a) and (b) of the draft RTS could turn out to be of little practical relevance.
In any event, it should not matter which authority prohibits the acceptance of contractual recognition clauses, since any prohibition would always be an obvious and insurmountable impediment.
We therefore believe that Article 1 should be amended to cover any authority which is competent to make such a declaration as well as other cases of impracticability resulting from legal or practical impediments, without being limitative.
As to the broader need to provide for a more open/flexible approach in order to address the very real situation that institutions will effectively not be able to impose such contractual clauses on their counterparties and that it will be impossible to clearly identify all relevant types of conditions constituting impracticability, see our comments on Question 4 below.
Q2. Can you provide concrete examples of instruments, such as letters of guarantee, governed by the law of a third country which are not used in the context of trade finance and which would be subject to conditions of impracticability?
There are many instruments which are not necessarily connected to trade finance transactions but which, nevertheless, are subject to comparable internationally established market practices and customs and/or are concluded in a manner (i.e. by telephone) which makes it practically impossible to include the required contractual recognitions clauses.This would include (non-exhaustive list):
• Any contract/agreement which does not cover an outright payment obligation and where the “liability” which could be subject to any bail-in would only be a secondary liability (damage claims or equivalent monetary secondary claims under the laws of the relevant jurisdiction) of an undetermined value/amount (which would be decided by a court for example). As to the need to clearly exclude such secondary claims from the scope of Art. 55 (limiting the scope to outright/direct/primary payment claims) - see also our comments below on Question 4.
• Deposits/certificates of deposits (concluded/confirmed via telephone or SWIFT-messages on the basis of established market practices and standards and therefore without a contractual documentation).
• Interbank guarantees loans/lines of credit (concluded/confirmed via telephone or SWIFT-messages on the basis of established market practices and standards and therefore without a contractual documentation).
• Spot transactions in securities or FX (concluded/confirmed via telephone or SWIFT-messages on the basis of established market practices and standards and therefore without a contractual documentation).
• Similar types of transactions where the terms are based on established market practices and are concluded/confirmed via SWIFT-messages or similar services.
• Guarantees or counter-guarantees issued for non-trade finance transactions (raising similar issues as for trade finance transactions) and which are widely often used in domestic commercial transactions/projects between clients and contractors in third-countries (also confirmed via SWIFT messages) such as tender bonds, advance payment bonds, retention bonds, performance bonds or warranty bonds.
Q3. Do you agree that the categories of liabilities in the above table do not meet the definition of impracticability for the purpose of Article 55(6)a)?
- At page 9, in the box number 5, the RTS indicates that the liabilities that arise out of an existing agreement which the acquiring entity has no power to amend are not in the scope of the conditions of impracticability. The RTS states that “this situation has not been kept for the list of conditions since it is believed there will be a contract for the acquisition of the instrument (...)”. Please note that this approach is not correct as we are not referring to the bail-in recognition clause to be included in the assignment or adherence agreement but to the one to be included in the underlying agreement, to which the entity will become a party once the assignment or adherence agreement is executed (which is a different legal relationship that cannot be amended by the parties to the assignment or adherence agreement). For example:• An assignment agreement by means of which an existing lender transfers part of or all its position in a syndicated facility agreement to a new lender (the transfer agreement will be executed by the assignee, the assignor and the facility agent but not by the borrower and the other parties under the facility agreement) and so therefore the borrower (and the other parties to the syndicated facility agreement) cannot give recognition that they may be bailed-in, or
• An adherence letter to a non-disclosure agreement (NDA) will not be signed by the beneficiary of the underlying NDA (but by the original recipient under said document and the new recipient) and so therefore the counterparty cannot give recognition that they may be bailed-in.
- At page 9, in the box number 6, the RTS indicates that the liabilities that are contingent to a breach of contract are not in the scope of the conditions of impracticability.
However, according to our Members, liabilities arising out of a breach of contract are not to be excluded from the impracticability criteria for various reasons:
• This approach seems to be in contradiction with the intention of the EU legislator formulated in recital 26, par. 2, of BRRD2: “For example, under certain circumstances, it could be considered impracticable to include contractual recognition clauses in liability contracts in cases where it is illegal under the law of the third country for an institution or entity to include such clauses in agreements or instruments creating liabilities that are governed by the laws of that third country, where an institution or entity has no power at the individual level to amend the contractual terms as they are imposed by international protocols or are based on internationally agreed standard terms, or where the liability which would be subject to the contractual recognition requirement is contingent on a breach of contract or arises from guarantees, counter-guarantees or other instruments used in the context of trade finance operations.”
Indeed, we understand that contingent liabilities such as NDAs, Reps and Warranties etc. which cause a financial compensation to be required only in the case of a contract’ breach, are hereabove included. As consequence, for such liabilities (i.e. liabilities contingent to a breach of contract) no inclusion of a bail-in clause should be necessary.
• The potential liabilities are not quantifiable as such and usually are being settled by courts and tribunals without any control on the outcome.
• The nature of these liabilities is purely contingent, they usually do not even come to existence.
• Most contingent liabilities will be completely irrelevant for a bail-in and it would be unreasonable to burden the institution with the obligation to impose a contractual recognition clause if such clause will not be of any practical relevance. This holds particularly true in case of contingent liabilities with mirroring counterclaims: in many cases, the contingent liability vis-à-vis the counterparty is counterbalanced by a mirroring counterclaim against a client; any bail-in into the contingent liability will therefore directly reduce the corresponding counterclaim, resulting in a zero-sum game.
The EBF therefore would suggest a general waiver, given without any detailed notification to the resolution authorities.
- At page 8, in box number 1, the consultation paper indicates that “the request of the counterparty to renegotiate the contract and/or an increase in pricing or a refusal by the counterparty to agree to be bound by a contractual bail-in recognition clause” is not a reason for impracticability. Our Members do not agree with this statement. In the context of the service contracts that a bank negotiates, mutual agreement plays a role and banks will always try to negotiate in order to include the bail-in clause. If the impracticability is defined as not ‘legally or otherwise’ possible to include the conditions in the agreement, it is a fair observation that a final refusal by a contractor effectively means that the 'otherwise' condition cannot be fulfilled. After all, both parties must agree. Same reasoning if the contracting party only wants to include the bail-in clause if important clauses/guarantees are weakened to the disadvantage of the bank or if this is used to contract at higher prices. If there are no alternatives on the market for the bank, it is not possible to include the bail-in clause without substantially weakening its position.
- In terms of low value contract/liability under box 4 (materiality threshold), we signal (SEE ALSO RESPONSE TO Q10 and 11) that a materiality/value threshold would be an extremely useful instrument to reduce unnecessary burdens and allow institutions, as well as resolution authorities, to concentrate their efforts on the practically relevant liabilities/instruments. In the event of a bail-in, the resolution authorities will – in the interest of efficiency - necessarily focus their efforts on higher value liabilities without individually pursue liabilities of low value based on contractual recognition clauses which would have to be enforced against counterparties in third countries, simply because the cost will significantly outweigh the benefits. This, of course, would not preclude the authority from simply applying the bail-in to such low-value/smaller liabilities based on their resolution powers and without resorting to contractual recognition.
Q4. Do you consider that there is any condition of impracticability that has not been captured in the analysis?
Yes.First and foremost, and as already indicated above, it is important to underline that it will not be possible to identify all potential conditions of impracticability. It needs to be recognised that the impracticability can result from a combination of circumstances which cannot be easily categorised or reduced to specify predetermined types of conditions or cases.
In this context, it is necessary to take into account that the experience over the past years has clearly demonstrated the difficulties EU institutions have had with such clauses and the very considerable competitive disadvantages they face in the international markets because of this requirement (particularly in view of the still extremely broad scope which is much broader than that of any other jurisdiction). As already mentioned in our general remarks, a too restrictive application of the newly established possibility to waive the need to introduce contractual recognition clauses may force EU institutions to withdraw from certain markets or transactions with third country counterparties. We do not see how such an effect is supposed to be justified in the name of a formal improvement of the resolvability, especially since the absence of a contractual recognition clause does not preclude the authorities from applying the bail-in tool.
Against this background, we would suggest a much more flexible and open approach regarding the specification of conditions which indicate a legal or other impracticability.
This could be best achieved by setting out a list of non-exhaustive examples of such conditions. A similar solution would also, in our opinion, enable the resolution authorities - together with institutions - to better take into account developments and practical experiences made.
At the very least, we believe it is necessary to set out more clearly that the listed conditions do not preclude an institution to refer to other conditions, indicating such (de facto or de jure) other impracticability cases.
As regards the specific conditions suggested in the draft RTS, and as already indicated above, we believe that the following should also be considered as a condition indicating impracticability:
• Contingent liability.
• Low-value contract/liability. Examples are the so-called "click-wrap" agreements, (i.e. agreements that a bank enters into online and where a set of conditions must be accepted; for instance, the purchase of licenses of very limited value).
As already also indicated (Question 2), we strongly believe that it is essential to expressly clarify that liabilities/contracts which are not direct/primary contractual payment obligations and where the value/amount will regularly be undetermined, are – as such - not captured by the Art. 55 requirements. Anything else would require institutions to notify such liabilities/contracts clearly producing unnecessary burdens for both institutions and authorities, when receiving these notifications.
Q5. Do you agree with EBA’s approach for developing the draft ITS?
We refer to our responses on the Questions 6, 13 and 14.Q6. Do you consider reasonable 3 months for entry into force of the ITS, as allowing enough time to set-up the proper and adequate capabilities to notify with this ITS?
No.Depending on the specific circumstances and the composition of the portfolio of liabilities, three months is likely to be very challenging. With the current draft, our Members, especially those with trade-finance activities in third countries, would have to notify individually tens of thousands of contracts over the course of a year. In this context, it has to be taken into consideration that institutions are currently facing numerous challenges requiring far-reaching operative changes and adjustments to existing procedures and contractual arrangements, including the ongoing benchmark-replacement projects, the impending coming into force of Brexit and the parallel implementation projects in respect of Art. 71a BRRD. We therefore see the need for a longer implementation period, ideally combined with a more flexible/staggered approach allowing for a risk-based implementation by institutions. Of course, clarifications clearly limiting the scope of potentially affected contracts/liabilities (such as the suggested clarification concerning the exclusion of indirect /secondary payment obligations) would greatly help the implementation process.
It would also be much more practicable if the resolution authority would determine upfront exempt categories of liabilities – i.e. types of liabilities for which the inclusion of the contractual bail-in clause is deemed impracticable – which would eliminate the need to notify all contracts separately.
Q7. Do you agree with EBA’s proposed conditions of impracticability?
Art. 1 (1) (a)/(b) - Breach of the law / explicit and binding instructionAs already indicated in our response to Question 1, we believe that the two situations addressed under lit. (a) and (b) will not be of much practical relevance: the more probable and practically relevant situation will be that the imposition of contractual recognition clauses on counterparties is likely to conflict with local rules or practices such as investor (or consumer) protection rules, or mandatory legal principles concerning clauses which are deemed unfair/unilaterally imposed (without allowing for any meaningful negotiations) and attempt to extend extraterritorial effect to measures by foreign public authorities.
In many cases, it will therefore not be possible to determine with certainty whether the clauses can effectively be implemented or not. At best, it will be possible to conclude that there is a considerable likelihood of challenges or, conversely, that it is more likely than not that the clauses will be effective. However, there will never be any legal certainty in this regard.
One way to address this issue would be to take this situation into account, if not as a condition which directly implies unconditional impracticability, at least as a relevant criterion for the assessment that the non-inclusion of the clauses does not adversely affect the resolvability.
In lit. (a) the reference to “law” should also include any rule or regulation (whether or not having the force of law) issued by an authority or court of the third country and applicable to the liability.
Art. 1 (1) (c) - Liability arising out of instruments or agreements concluded in accordance with and governed by internationally standardised terms or protocols which the institution or entity is unable to amend
The condition addresses the highly pertinent situation that existing and established standard terms and practices in the relevant market effectively preclude institutions from including contractual recognition clauses into the applicable terms of the agreement. However, and in order to ensure that this condition does not become practically redundant, the additional condition that the institution was unable to amend these terms cannot be interpreted too rigidly and should not cover a purely theoretical possibility to amend terms: for example, in general transactions on the basis of established terms and practices, ICC rules for guarantees and letters of credit (but not limited to these!) allow for some level of individual amendments. However, the market participants are - in practical terms - not able to unilaterally impose the clauses. The inability to amend in practice such kind of instruments or agreements (whether or not governed by ICC rules) should be expressly recognised in this §, in compliance with Recital 26 of BRRD2.
Art. 1 (1) (d) - Liability governed by contractual terms to which the institution is bound pursuant to its membership of, or participation in, a non-Union body, including financial market infrastructures, and which the institution or entity is in practice unable to amend
The comments to lit. (c) apply correspondingly.
Art. 1 (1) (e) - Liability is owed either to a commercial or trade creditor and relates to goods or services that, while not critical, are used for daily operational functioning and where the institution or entity is in practice unable to amend the terms of the agreement concluded on standard terms
The condition is too rigid and restrictive: first, as already mentioned, it should be clarified that contracts/liabilities which are not direct/primary payment obligations are excluded as such: the contractual recognition requirement can only apply to agreements which provide for payment obligations and cannot apply to any secondary /indirect monetary obligations which may or may not arise out of, or in connection with, the relevant agreement and where the value/amount is not even known. Anything else would effectively mean that any legal or even non-contractual agreement, yet contract-like relations (depending on the applicable law), would be captured by Art. 55 BRRD. This would be a clear and unnecessary overreach and also result in considerable uncertainty.
In addition, and as in the above cases, the additional condition of an inability of the institution to amend the terms is particularly unnecessary in the context of these agreements/contracts: it simply cannot be expected that institutions need to demonstrate that they tried to include contractual recognition clauses in types of agreements where it is already unclear whether a liability will ever exist.
In accordance with Recital 26, we would suggest adding another condition: “f) the liability which would be subject to the contractual recognition requirement is contingent on a breach of contract” (SEE RESPONSE TO Q3).
Art.1 (2)
The requirements established under par. 2, determining when an institution can deemed to be unable to amend an agreement, are far too restrictive. As already pointed out in our comments on lit. (c) to (e) above, this additional condition is unhelpful and should be deleted: experience has shown that institutions face great difficulty in unilaterally imposing contractual recognition clauses on counterparties where this is not customary in the relevant market and for the relevant products. These difficulties are not limited to cases where the counterparties set the terms. Rather, it also applies to all cases where the terms are grounded on customs and market practices.
Besides, our Members propose to add to the RTS a further generic criteria (for instance, in the third paragraph in article 2) where the bank could demonstrate that the bail-in of an instrument has no impact/no benefit on its resolvability. In this case, the resolution authorities shall not require the inclusion of the bail-in clause. This condition would not put at risk the process of notification since the threshold provides itself, for liabilities below the threshold, a line for a resolvability assessment. Furthermore, the BRRD itself provides specific criteria to assess the impact on resolvability based on a 10% threshold within a liability class.
Q8. Can you provide examples of instruments or contracts for which it would be impracticable to include the contractual recognition which are not captured by the above proposed conditions?
As to the need to expressly/clearly exclude non-payment liabilities, see already immediately above as well as our response to Questions 3 and 4.Also, the draft RTS does not provide indications for situations where the RA would require the inclusion of such clause whereas it is operationally impossible (e.g. does the bank then need to terminate the contract or accept a compliance breach?).
Q9. Are the proposed conditions of impracticability clear and meeting their purpose?
We refer to our responses to Questions 7 and 8.Q10. Is the article providing the conditions for the Resolution Authority to require inclusion clear?
As to our central concerns regarding the scope of Art. 55 and the need for a more flexible approach, see already our response to Question 4.We understand the proposal regarding Art. 2 of the RTS on the cases where the resolution authorities have to require or may require institutions to include contractual recognition clauses – together with Art. 1 of the draft RTS ¬– to effectively provide for the following avenues for institutions to address the issue that they will not be able to impose these contractual recognition clauses on their counterparties:
• The draft RTS set out two different avenues regarding the permissibility of the non-inclusion of contractual recognition clauses:
o The first one is based on Art. 1 of the draft RTS, in the form of a positive determination of impracticability which automatically and irrefutably makes it permissible for institutions to omit the inclusion of contractual recognition clauses.
o The second one is based on Art. 2 of the draft RTS, setting out criteria under which the authorities either have to or can require an institution to include the contractual recognition clauses. Under this second approach, institutions are effectively allowed to omit the inclusion of the contractual recognition clauses until/unless the authorities require them to do so on the basis of the criteria set out in Art. 2 of the draft RTS.
This Art. 2 approach distinguishes between two sets of circumstances which effectively mean the following for institutions:
• Art. 2 (1), as currently drafted, would effectively preclude institutions from omitting contractual recognition clauses where the amount of the liabilities in question is equal or higher than EUR 20 million, or where the maturity is equal or higher than six months. This would effectively mean that a liability with a longer maturity than 6 months and regardless of the amount, or in an amount of EUR 20 million and regardless of the maturity, would always be subject to the contractual recognition requirement.
• Art. 2 (2) would effectively allow an institution to omit the contractual recognition clauses as regards liabilities below the thresholds of Art. 2 (1) of the draft RTS, except where the resolution authorities determine that the omission to include the clauses significantly affects the resolvability of the institution (regarding the thresholds approach in Art 2 (2) see also answer to Q11).
While we understand the approach, we believe that the resulting framework for institutions is far too restrictive and formalistic and will not resolve the very real and considerable problems institutions are facing.
By narrowing down the scope of liabilities which may effectively benefit from Art. 2 of the draft RTS (i.e. the possibility to omit contractual clauses to liabilities below EUR 20 million with a maturity of less than six months), a significant portion of other types of liabilities with longer or unclear maturities, which are irrelevant for a bail-in and/or where the failure to include contractual recognition clauses would not adversely affect resolvability in any meaningful way, will continue to be subject to the contractual recognition requirements.
A clearer focus on the impact on the resolvability is essential to grant the authorities and the institutions the necessary flexibility to ensure that the contractual recognition requirements under Art. 55 BRRD do not result in unreasonable burdens for institutions, authorities and counterparties, without actually improving the resolvability.
As already indicated above, this would have unreasonable consequences not only because of the resulting considerable burdens and competitive disadvantages for EU institutions in international markets, but also considering that impeding the access of EU institutions to essential international markets, or even requiring a withdrawal from such markets, will not improve the resolvability of institutions in any way.
We therefore believe that a more flexible and open approach which focusses on the actual impact on resolvability would be the only reasonable and appropriate approach.
One element could be – as already suggested in responses to earlier questions – a non-exhaustive list of conditions/examples of impracticability under Art. 1 of the draft RTS.
Q11. Do you agree with EBA’s proposal for the conditions for the resolution authority to require the inclusion of the contractual term?
We refer to our comments on Questions 10 and the general concerns we raised in the responses to Questions 4.Specifically referring to the thresholds (in terms of liability/contract amount or maturity) introduced under Art. 2 of the draft RTS, we would like to also signal the following:
• These thresholds create an unlevel-playing field among EU banks: indeed, large banks will be penalised as they have more high-value contracts than smaller banks.
• These thresholds are not part of the level 1 text and reduce unnecessarily the flexibility of judgement of the RA in their assessment of resolvability.
Therefore, they should be removed from the RTS in their current form, i.e. as a way to automatically trigger the inclusion of the bail-in term. Materiality thresholds (in terms of maturity and amounts) should be used as a way to waive the obligation to notify the RA for liabilities below them. Moreover, considering the various sizes and capital situations of the banks, the amount-based threshold should be increased for larger banks and maturity-based threshold should also be raised:
• In terms of amount, a more proportionate approach should be used with the use of higher thresholds (or a % of the balance-sheet for instance) for the largest banks.
• In terms of maturity, it seems quite unreasonable not to consider the very short duration contracts (non-renewable) as automatically outside the scope of the art.55 requirement, taking into account the time for the RA to assess whether the inclusion of the bail-in clause is deemed necessary or not (3 to 9 months).
The assessment of the maturity of renewable contracts should also be clarified. For instance, a threshold of 3 years would make much sense given the nature of the liabilities in the scope of the impracticability conditions currently considered.
In any case, these threshold conditions should be independent, and not cumulative conditions to trigger the flexibility of the RA described in the second paragraph of Article 2. The latter explains that the resolution authority may require the inclusion of the bail-in clause when none of the conditions of impracticability described in Article 1 are fulfilled and when “the agreement’s or instrument’s value creating the liability is less than EUR 20 million and the remaining maturity of that agreement or instrument is shorter than 6 months.
We would strongly recommend the EBA that in this paragraph (p.17 of the CP) both conditions are made non-cumulative, which means that the flexibility of the resolution authority would be possible either when the amount is less than EUR 20 million or the residual maturity is shorter than 6 months. This would also alleviate the burden of the very constraining notification system proposed in this consultation as already mentioned above.
We consequently think the reference to the fact that “the institution cannot notify contracts for impracticability based on their value or maturity” (page 10 of the draft RTS), should for operational reasons be omitted.
Finally, we consider unclear whether one document or every principal document in a financing transaction should include a bail-in clause. Some industry guidance suggests all security documentation should include it in addition to the principal facility agreement. This potentially means more cost, more legal time and more negotiation for no additional benefit.
Q12. What is the likely amount of the liabilities to be notified under article 55 BRRD, as average per liability and as expected maximum per liability? What is the expected average maturity of the liabilities to be notified under article 55 BRRD?
Considering the breath of the scope of liabilities captured by Art. 55, it is impossible to give any helpful indication.Q13. Do you agree with EBA’s proposal for the reasonable timeframe for the resolution authority to require the inclusion of the contractual term?
No.Experience has shown that the imposition of the contractual recognition clauses is very burdensome and time-consuming (for both parties). Rigid timetables will not be helpful in this context. We therefore strongly urge EBA to consider a significantly more flexible and risk-based approach.
In terms of implementation, we would welcome further clarifications on the timeframe for the resolution authorities to require the inclusion of the clause, specifically referring to the time given to the Authorities to take their decisions and the time given to the banks to comply with their request.
Besides, further information on grandfathering provisions for existing transactions would be appreciated. In this respect, we would like to be sure that new rules will only apply to the new transactions flow, and that for the stock of transactions banks will have at least one year for implementation after the entry into force of the RTS.
Finally, the answer of the IRT on a waiver request may take place after the sign-off of the contract. Should it be the case, the bank would be put in very awkward positions in front of its clients. A fast track process would be very much welcome to avoid such situations and ensure that contracts that need to be signed swiftly can also be processed by the resolution authorities at an equivalent speed. This would be especially relevant for banks’ trade finance activities which consist in a high volume of transactions in any given year, and which compete against US and Asian players which are not subject to the same rules.
Q14. How much time do you need to implement the technical specifications provided in this draft ITS?
As already indicated above, we believe that a three-months period is likely to be too demanding.The time necessary for the institutions to comply with these requirements depends also on the way in which the RAs will ask for the template submissions (if the submission is required through XBRL, it will probably take even more time to be implemented).
Q15. Do you consider the draft ITS comprehensive for submitting a notification of impracticability?
Notifications:The tables imply that institutions may only notify cases of impracticability which can be connected to any of the listed conditions. As already indicated, the conditions suggested in the draft are too narrow and/or of little practical relevance. It is therefore necessary to allow for a notification of other conditions and this needs to be reflected in the tables.
Individual ID for liabilities:
The proposal assumes that at least some liabilities will and can be notified on an individual basis. Considering the very broad scope of Art. 55, which captures a vast array of liabilities (even if one - as strongly urged – excludes agreements which do not provide for direct/primary payment obligations), many of them will need to be notified as it will be impossible for institutions to include contractual recognition clauses. It would be extremely difficult to identify all these liabilities individually. It is thus unlikely that institutions will make regular use of an individual notification but would prefer to rely on categories defined by the resolution authority as per Article 55(7) BRRD.
Material amendment:
The purpose of a notification of a material amendment is not clear: as long as the amendments do not change the liability in such a way that the original assessment no longer applies (primarily if the amount of the liability is changed and now breaches the thresholds because of the change), amendments will be irrelevant and a requirement to notify them would be clearly unreasonably burdensome. This should be clarified, i.e. by confirming that material amendments for the purposes of the notification requirements are only amendments which substantially increase the amount of the liability.
Legal opinions:
The proposal assumes that the institutions will, at least in some cases, obtain legal opinions confirming the impracticability. Such legal opinions are not a requirement set forth by BRRD2.
Legal opinions are, of course, an important instrument to analyse the effectiveness and enforceability of standard agreements and clauses used by institutions. However, in this context, the opinions would not be relevant. First, most of the conditions of impracticability are not of a legal nature and can thus not be addressed by a legal opinion. To the extent that the impediment is indeed a legal one (Art. 1 (1) (a) and (b)), this would, however, have to be a negative opinion (confirming the illegality/ineffectiveness of a clause).
Many of the reasons will be practical and matters of fact which arise either in the industry sector, or in the countries in which the borrower and financial institutions are operating. They are not legal matters which a law firm could opine on. If an opinion were needed, a law firm would be likely to specifically exclude these matters from the scope of its opinion and liability.
As already indicated above, we can expect that only in exceptional cases will a clause be inadmissible/illegal/prohibited as such. The more relevant case will probably be that the legal analysis results in an inconclusive assessment identifying existing legal risks and uncertainties.
Insolvency ranking:
This item makes sense for liabilities governed by third-country laws and issued by third-country subsidiaries. For those issued by branches, the ranking in the European parent’s creditor hierarchy should apply. We presume that a specific chart of insolvency ranking (0080) in the third countries will be communicated by the RA.
Q16. Do you consider the templates and instructions clear?
See our above comments on Question 15.Q17. Do you have any suggestions or proposals in relation to the draft ITS template and the instructions to fill it in?
“0070 LIABILITY TYPE : Institutions shall identify if the contract/instruments is related to one of the following: - Hedging - Interbank Deposit - Deposit with non-banking clients - Derivative - Derivatives - Hedging - Borrowing/Funding - Trade finance - Operational services which are not critical to the functioning of the entity – Others”.This sentence, read alongside the Recital 26 of BRRD2, which reads “it could be considered impracticable to include contractual recognition clauses in liability contracts […] where the liability which would be subject to the contractual recognition requirement is contingent on a breach of contract”, suggests only financial liabilities are considered by resolution authorities, which is absolutely logical regarding the objectives of article 55 BRRD, i.e. resolvability of the entity. Nevertheless, and despite the efforts made from the banking sector on that subject since 2016, and to the best of our knowledge, the term “liabilities” is not defined. It is therefore not possible for entities to decide to introduce the bail-in clause only in agreements containing financing liabilities (as opposed to any other kind of liabilities such as respect of representations and warranties for example). As communicated above, it is therefore necessary to exclude with certainty all non-financial liabilities, which would for instance enable entities to avoid painful and useless negotiations to include a bail-in clause in non-disclosure agreements (NDAs).
Q18. Do you find any specific piece of information required in the template as hard to provide or unclear how to fill in?
The following clarifications would be appreciated from our Members:• Does the estimated amount column imply that the volume of liabilities needs to be monitored to remain within the indicated amount? And how long the estimation is supposed to be valid?
• We presume that ‘Contracts/instruments creating new liabilities’ means: the contracts under which new liabilities may arise, and not the individual transactions or drawings in execution of those contracts.
• What about new contracts/guarantees? Should they first be reported before the bank can deliver the guarantee/conclude the contract? In practice, this will not work.
• Re. the final maturity date (0030): what in case of contracts/instruments valid until further notice (initially agreed or issued for an indefinite period of time)?
• We presume that a specific chart of insolvency ranking (0080) across the third countries will be communicated by the RA (see also Q15), to cover the case of subsidiary liabilities.
• Re. governing law (0120): for certain trade finance instruments (such as letters of credit) it is not market practice to insert a clause dealing with governing law.
• More information about the “legal opinion” (0160) that seems to be required would be welcome. It apparently overrides the level 1 text.
• Re. counterparty – name (0170): this is problematic, because this may lead to problems re. banking secrecy and the duty of confidentiality of a bank.
• Governing law: in the case of trade-finance liabilities (i.e. guarantees and standby letter of credit), many third-country contracts are based on ICC rules (e.g. URDG458, URDG758 or ISP98). Counterparties for such contracts have access to the ICC dispute resolution mechanism which rely on the ICC International Court of Arbitration. Therefore, banks’ IT systems record neither the underlying governing law nor the competent court.
• There is general lack of clarity on the time necessary for the institutions to comply with these requirements (see Q14), which depends also on the way the RA will ask for the templates’ submission (e.g. if the submission is required through XBRL, it would probably take even more time to be implemented).
Regarding Template N 01.02, it is highly preferable and much more practicable that the resolution authority will determine “exempt categories”.
Regarding Template N 02.00 (Liability Insolvency Classes), it is very burdensome to fill in this template (cf. “all outstanding amounts of liabilities governed by the law of a third-country”).
Q19. Do you agree with the draft Impact Assessment? Can you provide any numerical data to further inform the Impact Assessment?
Regarding the Impact Assessment (IA), it is mentioned that it should analyse “the potential related costs and benefits”. In our opinion, the burden of negotiation represented by the mandatory inclusion of a bail-in clause in almost any agreement creating any form of liability, and without considering its real impact on resolvability, has not been taken into account in the draft of the IA.In addition, no comparison to the cost of a proper equivalent system (between EU and non-EU resolution regimes) has been made. The balance between the gain in terms of legal security and operational burden/costs is obvious, for instance, on the securities and debts, where we welcome the article 55’s effect. On the contrary, for some liabilities, this is very cumbersome and its benefits deeply limited (contingent liabilities when depending on the breach of a contract, non-monetary liabilities, etc.).
Finally, no assessment regarding the absence of a level-playing field between EU entities, subject to this requirement, and non-EU entities has been made.
In our opinion, the proposed draft of IA is therefore missing some crucial points, which may have led to a more comprehensive assessment of impracticability.