AT1: Adapt Or Abolish?

Within its simplification proposals published in December 2025, the European Central Bank (ECB) proposed two alternatives to improving the going concern loss absorption capacity of the capital stack. The first alternative was to enhance the features of AT1 debt such that the product would better align with the requirements for going concern capital and be more similar to equity. The more drastic alternative would be to abolish the instruments altogether from bank capital stacks.

The ECB stopped short of giving any further details on its thoughts on AT1. Based on comments by Luis de Guindos, Vice President of the ECB, in January 2026, the Governing Council seems to have a preference for the less dramatic alternative, including tweaking the terms. Furthermore, he stressed already in December that any change to the AT1 product will not be done retroactively, but instead will be about the way forward. We present some potential alternatives below.

One of the major criticisms concentrates on whether AT1 instruments can be truly loss-absorbing before the bank even approaches the point of non-viability.

In the EU, resolution authorities have statutory powers to write down or convert AT1 debt independently of resolution action and without a breach of the set trigger levels. The usage of these powers is, however, complicated in practice by the requirement that resolution authorities have to exercise the write-down or conversion in accordance with the priority of claims under normal insolvency proceedings. These rank AT1 debtholders with a higher priority than CET1 capital. The Single Resolution Board and the ECB have confirmed their intention to respect the hierarchy of claims in crisis interventions: it should be the shareholders first that absorb losses and only after the full use of CET1 capital would AT1 instruments be written down. This stands to clearly limit the potential use cases of these powers.

In addition, while the EU AT1 instruments come with quantitative loss-absorption triggers, you could see the bulk of these triggers as more or less obsolete.

Bank failures are not always driven by an accounting loss and a resulting hit to capital buffers. Instead, it is rather the deposit outflows and stress in the funding and liquidity position that may become detrimental and fast.

Most euro-denominated AT1 instruments have a mechanical trigger set at 5.125% and some at 7% of the CET1 ratio. Breaching the trigger leads to a write-down of the principal or a conversion of the instrument to CET1 instruments. While bank fundamentals may be deteriorating at a fast pace on the liquidity or funding side, it is often the case that the capital ratios are still well above these mechanical trigger levels.

Furthermore, for most banks with notes with a 5.125% trigger, the aggregated level of the minimum Pillar 1 CET1 requirement and the Pillar 2 requirement is higher than the loss absorption trigger level. Adding the combined buffer requirements further increases the more normalised level of CET1 capital, and further distances this from the mechanical triggers.

In practice, we consider that a bank could end up in a resolution long before these mechanical trigger levels are reached, at which point the notes would not be absorbing losses on a going concern basis but rather on a gone concern basis.

The trigger levels come from EU (and Basel) rules. A trigger event occurs when the CET1 ratio falls below 5.125% or a level higher than this, as specified in the bond documentation. Revisiting these trigger levels is likely to be among the alternatives to be discussed in the case of enhancing AT1 features to make the debt better loss-absorbing. We consider that higher trigger levels would be, from the point of view of AT1 holders, a negative and, from the point of view of shareholders, a positive.

Even a complete removal of the mechanical triggers is likely to be considered. The ranking of shareholders and AT1 debtholders may, in some cases, cause some uncertainty. In case of a breach of a trigger, only AT1 absorbs losses, while in the case of statutory loss absorption, both shareholders and AT1 stand to absorb losses. This may be seen as providing uncertainty in times of stress with contrasting interests for the two layers. A conversion instead of a write-down structure could be seen as aligning the two interests a tad more. In the EU, the bulk of AT1 instruments have a write-down structure and only a smaller proportion are structured as convertibles.

For AT1 instruments to be eligible for Tier 1 capital, the bank must, at its own initiative, be able to cancel the coupons of the bonds at all times and for as long as they like under the EU rules. A coupon cancellation can't constitute an event of default, and there can be no obligation for the bank to pay these skipped coupons at a later stage. Also, the level of distributions can't be driven by the basis of the credit standing of the institution.

A coupon skip has been intended to be among the measures that banks can take to provide more flexibility, for example, in the case of deteriorating fundamentals. However, banks have been very reluctant to do so in the past and have continued to pay coupons even under more substantial duress.

A coupon skip comes with a strong negative stigma. A coupon cancellation can be seen as a clear signal of trouble by financial markets and potentially even as the bank approaching the point of non-viability. In some cases, the actual direct saving from a skipped coupon may also seem limited as compared with the potential adverse market reaction impacting all outstanding instruments and thereby, the funding costs of the bank.

Another factor that likely also contributes to banks being unwilling to skip coupons is related to their nature. AT1 coupons are non-cumulative. Any skipped coupons will not be paid at a later stage. This could be seen as problematic from the point of view of the creditor hierarchy. If a bank skips a dividend to its shareholders, once its situation improves later on, it can pay extra dividends. This is not the case for AT1 instruments.

Banks may also take a more cautious approach to lending in times of stress to reduce the risk of negative consequences for AT1 coupons. The capacity of banks to pay AT1 coupons gets restricted if the CET1 capital ratio no longer meets the combined buffer requirement. In this case, the bank can only distribute the so-called maximum distributable amount as dividends, AT1 coupons or variable remuneration. A review of whether the calculation and functioning of the MDA (and M-MDA) framework works as intended could be among the potential avenues forward.

In the past, the ECB has indicated support for strengthening the features of AT1 paper to reduce the stigma effect associated with cancelling coupon payments. One alternative it presented was to review the CRR2 definition of distributable items such that only profitable banks or banks with positive retained earnings could make AT1 coupon payments or dividend payments.

The ECB has not previously supported adding a power to introduce a system-wide restriction on distributions, even in exceptional circumstances, due to several potential drawbacks. During Covid-19, the ECB instead published a recommendation for banks to refrain from distributions due to the high uncertainty at the time. However, this was not extended to AT1 distributions.

Features, including more mechanical restrictions on coupons or coupon accumulation, could also be considered.

AT1 instruments are perpetual, i.e., they don't have a final maturity date. The notes, however, come with a call date (minimum of five years after issuance), when they can be redeemed. To get supervisory approval for a call, 1.) the notes either have to be replaced with own funds instruments of equal or higher quality or 2.) the institution has to demonstrate that its capital buffers following a call would remain in excess of the requirements, including the combined buffer requirement by a necessary margin.

A replacement has to be done at terms that are sustainable for the income capacity of the institution. This needs to include an assessment of the extent to which the replacement instruments would be more costly than those they would replace. The profitability of the institution should continue to be sound and not see any negative change after the replacement of the instruments at that date and for the foreseeable future.

Banks, however, have a preference to redeem AT1 instruments at the first opportunity, as this is seen as market practice and a failure to do so may impact secondary spreads. Also, an extension may raise questions about the reason behind the extension. Banks may prefer to call even though refinancing them could, strictly speaking, be seen as more costly than retaining the existing instruments. While more and more banks run with an economic call policy, there is quite some room still to move even within this strategy. Larger banks with more substantial AT1 layers may take more of a portfolio approach to AT1 refinancing instead of a bond-to-bond approach. Refinancing in a different currency also provides its own touch to the matter.

In current markets, replacement economics are not really an issue as most calls look economic in any case. This may, however, change once the secondary markets weaken, and consequently it gets more expensive to issue a new AT1 instrument. Getting supervisory approval for a call is also likely to get more difficult in a weaker macroeconomic environment, in the case that the bank's capital buffers get thinner.

Stricter rules regarding the possibility to call AT1 are likely one route that is being looked at to make AT1 instruments better fit for their purpose. The ECB already suggested in 2022 that the features could be improved by limiting the possibility to call only if replaced with a CET1 instrument or a cheaper AT1 instrument.

The more drastic alternative to addressing the said AT1 problems would be to completely remove non-CET1 instruments from the going concern capital stack. This was one of the suggestions floated around by the ECB in December 2025, likely coming from the German bank regulators.

The ECB suggested that the AT1 instruments could then either be fully or partially replaced with CET1 capital or eliminated without replacement in the going concern capital stack.

A complete removal of AT1 instruments from the going concern capital stack would leave CET1 capital as the only means for banks to absorb losses before the bank fails or is likely to fail. This would reduce complexity quite a bit. If banks were required to replace AT1 capital with CET1 capital (fully or partly), this alternative would, however, increase CET1 capital requirements, which would go against the wishes of the ECB regarding maintaining the“current resilience”. A higher CET1 requirement only for European banks would also place the sector at a disadvantage compared with their international peers, a potential drag on the availability of lending and the European economies. AT1 instruments could also be replaced by Tier 2 capital, although this would not really solve the underlying issue with loss absorption.

Removing AT1 without a replacement with CET1 capital would mean that the total going concern capital requirements would be lowered. The importance of CET1 capital in bank capital stacks would increase. The European rules would no longer be in compliance with the international Basel rules.

A full elimination of AT1 instruments would mean scarcity and a very high likelihood of a call for the outstanding instruments once they reach the first call date.

The layering of new capital and loss-absorbency requirements since the global financial crisis, on top of old ones, has resulted in a rather complex set of rules. While this piece has mainly concentrated on the impacts for AT1 instruments, potential changes to the instruments may come with a larger remodelling of the bank's capital and loss-absorption stack. The ECB has indicated that changes to the loss absorbency framework should be done without reducing the components on banks' balance sheets which can be used to absorb losses and recapitalise in case they fail.

One alternative floated would be to rely more on CET1 capital as the main source of capital to absorb any weakness, while loss-absorbing but non-CET1 capital could be used to meet requirements for resolution buffers on a gone concern basis. An FSB working paper, for example, suggested a single minimum going concern capital requirement based on CET1 capital with simplified CET1 buffers on top. Requirements for other capital instruments and loss-absorbing capacity would be replaced by a requirement for additional resources for loss-absorbing capacity that could be fully met with eligible debt. The breach of the single minimum requirement would lead to a resolution, while a breach of the resolution requirement would come first, with other types of penalties.

If the regulatory capital stack were changed to only CET1 capital, the role of the different subordinated and senior debt layers could also be seen as less evident. How many different gone concern loss-absorbing layers would a bank really need in the end? It is doubtful whether the answer would be four, which is more or less the current structure.

We consider that the potential changes to the bank capital structure and AT1 instruments could have substantial consequences for European banks. AT1 instruments have become an important part of bank capital structures. Abolishing or changing AT1 instruments could potentially impact some €140bn in AT1 instruments within 107 European banks based on the EBA transparency data.

France, Spain, Germany and the Netherlands rank at the top in AT1 volumes for European banks. While 60% of the total is issued by the 10 largest banks, AT1 instruments are relatively widely used, with 69 banks out of 107 utilising AT1 paper. In particular, larger banks with a more binding leverage requirement have been active in AT1 markets. For banks with AT1 outstanding, the average AT1 layer was closer to 2.1% of RWA, and 52 banks were running with an AT1 layer of at least 1.5%. Replacing all this with CET1 capital would come with a substantial cost to banks across the board.

The numbers compare with some further €235bn in Tier 2 instruments in these banks, which are a bit more split across the names, with the top 10 corresponding to some 48% of the total. These capital layers are on top of some €1.5tr in CET1 capital.

We see the ongoing discussion related to bank capital as part of a larger one. The entire bank loss absorption stack has become a rather complex creature with parallel capital and loss absorption requirements both on an RWA and leverage basis.

The European Commission is working to simplify bank rules. Our base case would be that Europe is reluctant to move too far away from international rules. Also, in a more international arena, there are signs of a desire to reconsider the composition of bank capital layers and, as such, the role of bank capital instruments in it. Any larger change would take years to complete.

We don't see a complete abolition of AT1 instruments as likely any time soon. A change would be complicated by the substantial size of the market, with a large share of EU banks using these instruments to top up their capital positions. A requirement to replace these instruments with CET1 capital would be costly, and place European banks at a disadvantage compared with their international peers. Not replacing them would mean that European rules would fall short of the Basel standards. A more radical change would make AT1 instruments more likely to get called, as banks would shift to a new regime.

We consider that the current rules, however, already allow for quite a bit stricter application than what we have seen in the past. Some tweaks to AT1 features could be forthcoming at least.

Quite some expectations have already been priced in, with the spreads for euro-denominated AT1 instruments this week reaching new lows since the beginning of 2014. We would continue to navigate this potentially changing market by having a preference for shorter calls with higher resets.

MENAFN04022026000222011065ID1110696296