EBA: Triggers, calls & UK law

Do AT1 trigger levels still make sense? What factors determine whether banks can call AT1s? And how could Brexit affect UK law issuance? Delphine Reymondon, head of the liquidity, leverage, loss absorbency and capital unit, European Banking Authority (EBA), shared her thoughts on these current questions with Bank+Insurance Hybrid Capital and Crédit Agricole CIB.

EBA corporate pic web

Bank+Insurance Hybrid Capital: The absolute level of triggers on AT1, particularly the 5.125%, appear a bit low compared with reality, with the risk that they are increasingly perceived as gone concern rather than going concern instruments. Are supervisors contemplating any changes? And does this mean there is a greater need to give more clarification on the point of non-viability (PONV)?

Delphine Reymondon, EBA: First, we should probably recall how we originally fixed these levels, in particular the 5.125% trigger. Back in 2011, just after the crisis, the global framework was less developed and less complex than it is today. When we had to fix this metric we had only two buffers in place, the capital conservation buffer and the countercyclical buffer, the level of the requirements was lower than what it is today. At that time, to define this 5.125%, we simply used the level of the capital conservation buffer as the level which, if breached by a bank, 100% constraints on distributions would be triggered.

It was the result of a compromise and already at that time there were some voices saying that it was probably too low. This debate as to whether we should review the level of triggers has been reactivated, because we now have a more complex framework. We have more buffers and higher requirements and we are in a resolution world — back in 2011 this was not the case, or at least the resolution aspects were not defined as they are today.

In my personal view there are perhaps some merits but also some drawbacks in reassessing or reviewing the level of the triggers now.

Firstly, I do not think there will be an easy answer, because maybe some would even say that 7% may not be enough, and then what would be the right level?

This would then also reopen the capital definitions. It would also reopen the discussion about buffers, capital stacks, and all the different interactions. So it would not be so easy to do.

In addition, in my view we should keep some stability in the regulatory framework for capital, so I would personally not be in favour of this. Regulators are working on the interaction between going and gone concern frameworks. However, at the moment, we do not favour a change in the capital framework since we believe that the current one has proven to be an effective regime. We reaffirmed this recently in public communications with EU co-legislators in the context of the CRR review.

One change that has been introduced — which also came from a recommendation from our side — relates to the point of non-viability, which has now been included in the CRR2. It was missing in the first CRR because when Basel finalised its press release on the point of non-viability, CRR1 was almost finalised. That said, we will not change the level of the triggers for now.

BIHC: We know the regulator has to give its approval for any call. What we understand less is the process for this approval — we assume it’s a combination of economics and capital planning. Could you give us more detailed guidance? In particular there is the impression that the economics is coming more to the fore.

Reymondon, EBA: Indeed, from the EBA side, this issue of the exercise of forthcoming calls is something we will be looking at. We communicated on this aspect in our last AT1 monitoring report.

Delphine Reymondon EBA web

To be more precise, as regulators, we are not the ones to say yes or no to exercising a call — it is more the job of the supervisors. But the EBA did publish, back in 2013, some Regulatory Technical Standards to give indications on what criteria supervisors should consider when assessing an application from an issuer to exercise a call, for example the global situation of the bank, its profitability, its capital planning, and the margin compared to the capital requirements. It is a case-by-case basis assessment, you judge the particular situation of the bank and you check if the bank intends to exchange or replace the instrument. You also need to keep in mind the global market perspective in a way, because it is also important to know if, in case of need, the bank will be able to re-issue at short notice and under which conditions. These are more or less the aspects which are looked at.

We should also bear in mind that there is always a little bit of pressure, also from the supervisory side, because if you refuse the exercise of the call then you may put a stigma on a particular bank, so this is not an easy decision for the supervisor to make.

BIHC: Will the supervisor be able to clarify if it actually rejected a bank’s request to call an instrument? If an issuer doesn’t call a bond they may blame the regulator and ask investors not to punish them.

Reymondon, EBA: If I were still in a supervisory position then I would probably be quite reluctant for a bank to communicate that it is because of my supervisory decision that a call has not been exercised. But I believe that the supervisor should have a dialogue with the bank sufficiently in advance to see if views are aligned. I would expect that there would at least be some convergence in views on the opportunity to call or not the instrument. For banks that are in troubled situations, which have low margins compared to requirements, I believe that investors would probably already make some assumptions that the call may not be exercised. Thus I do not envisage a lot of situations where there would be a big surprise. In addition, if the bank does not want to call for economic reasons, in this case it can be explained without referring to a potential refusal from the supervisor.

BIHC: I understand that in the new CRR2/CRD5 draft there may be new provisions on bail-in acknowledgement for bonds that have not been issued under EU law, and today the market clearly has a new focus on this because of uncertainty regarding the future treatment of UK law AT1. Are there any discussions regarding potential grandfathering of those bonds? We heard rumours about a six year grandfathering period.

Reymondon, EBA: There are several aspects to this question. Indeed, first of all there are — as everyone knows —some political discussions and no one yet knows what the outcome of those will be.

A different aspect is that the requirements of the BRRD, in particular Article 55 statutory bail-in clauses, are already there, so this is not something new. By the way, EBA previously developed a standard clause for bail-in clauses for issuances in third countries. As we said before, the new aspect which will be introduced is the PONV as an eligibility criterion for capital instruments.

Then indeed the question with Brexit is how issuances under UK law will be treated. We issued an opinion a couple of months ago saying two things. The first was that if you want absolute legal certainty on the effectiveness of the clauses or the powers for the resolution authority, the best way would be to issue from an EU27 institution’s perspective under EU27 law and under UK law for UK institutions. Otherwise, the solution would possibly be to have a contractual clause. What we have seen in some issuances is some substitution/variation clauses to potentially change the governing law. The issue, which already exists, is that the resolution authorities theoretically have the power to discount some of the capital issuances from MREL if either there is no binding agreement with the third country or they believe that the contractual clause could not be triggered in an effective manner. So there would still be a need for banks to demonstrate to the resolution authorities that this clause would be effective in resolution, and this is of course more difficult to prove. I guess you could have a lot of legal opinions there, but legal certainty would for sure be less than if you issue under EU27 law.

In terms of grandfathering, what I believe is that for all the clauses linked to Article 55, there should be a priori no grandfathering, because it was already in BRRD1. What will be grandfathered a priori is the introduction of new criteria. Even if the grandfathering provisions are still being discussed, the PONV clause would very likely be included within a six year grandfathering period — it is something we recommended as EBA.