European banks: Let’s talk bank resolution

One of the lessons that were learned from the financial crisis, that started 13 years ago, was that the banking sector has to have a better capacity to absorb losses when needed. In addition, it should be possible to put banks either in resolution, if there is a public interest to do so or wind them down, without causing wide-ranging havoc in the financial markets and the banking system.

The European Crisis Management and Deposit Insurance (CMDI) framework were created to set rules for handling failing banks while protecting depositors across the Union. The package is based on the Bank Recovery and Resolution Directive (BRRD), Single Resolution Mechanism Regulation (SRMR) and the Deposit Guarantee Scheme Directive (DGSD). The framework is set to be reviewed with new proposals expected in 2022.

The Single Resolution Mechanism (SRM), including the Single Resolution Board (SRB) and national resolution authorities, were created to ensure an orderly resolution of banks that are failing or likely to fail while having a minimum impact on the real economy and public finances. Building sufficient loss-absorbing capacity allows banks to both absorb losses and also to be recapitalised in a resolution, in case of need. The minimum requirements for own funds and eligible liabilities (MREL) are set for banks for this purpose with the interim requirements to be met by 1 January 2022. Banks are well-positioned for the interim requirements and moving towards meeting their final targets for 2024.

The bank resolution framework provides tools that can be used before and at the point when a bank is considered to be failing or likely to fail. Measures that can be taken prior to this point include early intervention measures and preventive measures. Once a bank is considered to be failing or likely to fail, the bank can either be put into resolution or liquidated, depending on whether there is a public interest for a resolution of the bank.

Prior to any resolution action, the capital instruments of the bank have to be written down. As outlined in the BRRD, resolution tools include the sale of the business, the creation of a bridge institution, asset separation and the bail-in tools.

Effective bank resolution clearly requires resources, which is why banks are now required to build loss-absorbing capacity on top of their capital buffers. The target is that resolution financing arrangements can be accessed only after private resources have been tapped. For a contribution from the Single Resolution Fund, losses corresponding to 8% of the bank’s total liabilities and own funds would first have to be borne by shareholders, holders of capital instruments and other eligible liabilities. In some circumstances, a deposit guarantee scheme (DGS) may be tapped to reach the required 8%. DGS intervention after a bail-in of liabilities is subject to the least-cost test, where the DGS contribution in resolution has to be less costly than reimbursing covered deposits in a payout event.

The hierarchy of claims is based on national insolvency laws that may differ between countries

The hierarchy of claims is based on national insolvency laws that may differ between countries. This hierarchy should also be respected in a resolution (no-creditor-worse-off principle). The regulatory capital instruments, CET1 capital, Additional Tier 1 capital and Tier 2 capital are the first to absorb losses. These would be followed by other subordinated items and non-preferred senior unsecured debt. This ranking is shared across countries.

What comes thereafter, differs between countries. While preferred senior unsecured debt, ranks straight after non-preferred senior unsecured across the Union, items ranking pari passu to preferred senior debt, vary. In most EU countries non-preferred deposits rank alongside preferred senior unsecured, such as in Belgium, France, Germany and the Netherlands, among others. Instead, some countries have introduced a depositor preference. For example, in Bulgaria, Croatia, Cyprus, Greece, Italy, Portugal and Slovenia non-preferred deposits, rank ahead of preferred senior debt in the hierarchy. This means that preferred senior debtholders may be on the hook for losses prior to all depositors in these countries in a resolution. In general EU countries have deposit ranking on three different levels, with covered deposits having the safest status followed by preferred deposits and non-preferred deposits.

No-creditor-worse-off principle is a guiding principle in resolution

Debtholders should not be worse off in resolution than in insolvency. This no-creditor-worse-off system is a guiding principle when it comes to the bail-in tool. While MREL requires banks to hold a minimum level of buffers that are bail-in-able, it is good to note that in resolution, the scope for a bail-in may be wider than the MREL buffers. The scope for a bail-in may extend to all liabilities, subject to the creditor hierarchy based on the national insolvency law unless the liabilities are specifically excluded from a bail-in. The BRRD excludes from a bail-in among others: covered deposits, secured liabilities such as covered bonds, client assets and liabilities with an original maturity that is shorter than 7 days.

For loss absorption, an item ranking alongside an excluded liability may pose problems with the no-creditor-worse-off principle. To this end, the amount of excluded liabilities should not be too high. The EBA has assessed that the mandatory exclusions from a bail-in for items ranking between non-preferred senior and non-preferred deposits, amount to 5.5% of the respective liability class, giving some insight into the matter. Items facing a mandatory exclusion from a bail-in would be slightly higher for large and small banks as compared with medium-sized banks.

The Single Resolution Board (SRB) makes detailed plans for any potential future bank problems and draws resolution plans for larger, complex banks. These plans identify critical functions and any impediments to resolvability and present the preferred resolution strategy and tools. The aim is to ensure the continuity of these critical functions, avoid significant adverse effects on financial stability such as contagion, protecting public funds, covered depositors and client funds.

MREL targets build on resolution plans

The resolution approach differs between different banks. It can be based on a single or a multiple points of entry approach. The point of entry here refers to where in the bank’s organisation, the resolution is in practice conducted. This depends especially on the bank’s legal structure and perhaps on its geographical reach.

With a single point of entry the resolution is done via one entity, usually the parent entity of the bank. The parent entity may be an operating parent entity or a holding company. In this case, losses from any subsidiaries are transferred to the parent entity, and the parent entity’s buffers (and investors) absorb losses when needed. In this case, external MREL resources are issued from the parent entity to third party investors. This is by far the most common approach for larger European banks.

The multiple point of entry approach is used for banks with complex structures that have substantial exposures via independently-run subsidiaries in several different countries. If one or some of the resolution entities run into trouble, only those entities may be put into resolution. In a multiple point of entry approach, external MREL resources are in practice gathered at all resolution entry points. This approach is used among others by certain global, Spanish and Austrian banks that exhibit a more complex structure with substantial independent subsidiaries.

Resolution authorities utilise resolution plans when setting MREL targets.

MREL requirements are designed to be set so that after the usage of the bail-in tool, the bank’s capital position would be high enough to continue functioning with sufficient market confidence. The recapitalisation capacity should be in a format that is long enough and can be credibly written down or converted into equity as outlined in the CRR.

The SRB sets MREL requirements based on risk-weighted assets and against leverage exposure measures. They consist of a loss absorption amount (LAA) and a recapitalisation amount (RCA). The LAA and RCA are both based on the Pillar 1 (8%) and Pillar 2 requirements (bank-specific). The combined buffer requirement, that has to be met by CET1 capital, is added on top. The leverage-based LAA and RCA requirements are set in line with the leverage ratio requirement. The amounts can be adjusted, for example, either upwards by the SRB, to include a market confidence charge, or downwards to reflect a selected resolution strategy, such as using a transfer tool.

MREL requirements are bank specific

So-called Pillar 1 banks include global systematic banks and other large banks that have assets above €100bn or that may pose a systemic risk. They have to meet a non-adjustable subordination requirement. Depending on the bank, the subordination requirement may be set at 8% of total liabilities and own funds (but the level will be capped at a maximum of 27% of RWA for top tier banks), based on RWA or based on LRE. Global systematic banks should meet subordination requirements of 18% of RWA (with a 3.5% exemption in certain cases) or 6.75% of LRE as of 1 January 2022. Other Pillar 1 banks should build subordination levels of 13.5% of RWA or 5% LRE. Non-Pillar 1 banks may face a subordination requirement to avoid a breach of the no-creditor-worse-off principle.

Subordination requirements can be met by own funds and eligible liabilities that are subordinated to all claims arising from excluded liabilities. G-SIBs may be allowed to utilise the 3.5% senior add-on for TLAC requirements.

The SRB has communicated to banks their MREL requirements in line with the BRRD2 framework. The interim requirements have to be met by 1 January 2022 and the full MREL requirements by 1 January 2024.

The average MREL targets by country are shown in the chart above. In addition to the differences in risk density, the size of the banks (ie, the number of the Pillar 1 banks) and the no-creditor-worse-off principle, also have an impact on setting the subordination requirements. Banks in Germany, Belgium and the Netherlands have to meet the highest subordinated MREL targets both as compared to RWA and as a proportion of their total MREL requirement. The subordinated requirement for these banks is clearly above 80% of their total requirement.

The lowest subordination requirements have been set for banks in Southern Europe. Banks in Greece and Cyprus are not subject to subordination targets. It is noteworthy that countries that have the lowest proportional subordination requirements of their total requirement include those countries that have also introduced a depositor preference.

Banks have made substantial progress in building their MREL buffers in recent years. The stock of MREL eligible liabilities and own funds has increased to €2,208bn in 1Q21. The stock has increased by €118bn or 6% since 2019. The amount of subordinated MREL has increased by an even faster 8% since 2019, reflecting the action taken by banks to meet their interim MREL requirements by 2022.

Shortfalls of the interim targets as of 1 January 2022, were very small (€238m or €5.4bn including CBR) for the whole system and driven only by the subordinated component as of 1Q21.

Excluding the combined buffer requirement, the overall MREL shortfall was €23.6bn according to the SRB in 1Q21 when comparing the MREL capacity against the final MREL targets. Furthermore, including the combined buffer requirement, the total shortfall would increase to €42bn for the whole system.

Banks especially in Greece, but also in Croatia, Cyprus and Portugal, have work to do in terms of MREL buildup

€16bn or 38% of the total shortfall is for Greek banks, corresponding to almost 10% against their RWA, as shown in the chart below. Greek banks have until the end of 2025 to meet the requirements. Following Greece, in Croatia, Cyprus and Portugal the MREL shortfalls are between 4-8% against RWA, meaning that banks in these countries also still have some work to do in terms of buildings their buffers. As banks with the largest gaps relative to RWA have on average lower issuer ratings, we consider a benign market sentiment is especially important for the further build-up of their MREL buffers.

The shortfall, excluding Greece, was €26bn, with the largest gaps in Italy, Spain and Portugal. The MREL shortfalls in Italy and Spain however are limited when comparing them to the total RWA. Banks in Germany, Austria, France and the Netherlands have very limited MREL shortfalls.

When including the combined buffer requirement, half of the resolution entities were listed by the SRB as reporting a shortfall against their MREL targets. Banks that have a shortfall especially include smaller entities with 72.4% of the total reported by non-Pillar 1 banks.

In our view, gathering MREL resources may take more time for some smaller entities. Some smaller banks may have traditionally relied more on CET1 capital in their capital structure and deposits in their funding mix. This could have resulted in these banks being less active in financial markets and investors not being familiar with the name. In addition, their smaller size may result in a smaller targeted issue, which means the bonds may not be included in the main bond indices and will therefore attract less attention from investors.

The CMDI resolution framework currently guiding the treatment of bank failures may be subject to change, with the first notes set to be played in the course of 2022. The review aims to increase the framework’s efficiency, proportionality, and overall coherence to manage bank crises in the EU, irrespective of the banks’ size and business model, and to enhance the level of depositor protection according to the EBA. The European Commission launched a targeted consultation on the review of the CMDI framework in January 2021. A Commission proposal for the review is expected in the second half of 2022.

A Commission proposal for the CMDI review is expected in the second half of 2022

The Commission consulted on promoting further harmonisation of the creditor hierarchy in bank insolvency and in particular on depositor preference across the union. The CMDI framework does not provide harmonisation of the hierarchy of claims across the union, which means that in some member states, non-eligible deposits rank pari passu with ordinary unsecured claims, while in some other member states all deposits have a preferred status. Furthermore, the deposit guarantee schemes are national. The treatment of depositors beyond the deposit guarantee (€100k) and the functioning of national DGS, differ across countries. The consultation raises the question of whether the framework has managed to shield public funds from bank failures, especially in the case of smaller and mid-sized banks.

The EBA published its opinion on the review in October 2021. The report concentrates especially on the bank funding sources required to handle a bank failure in either a resolution or insolvency. The sources include having sufficient loss-absorbing instruments and access to resolution financing arrangements. In its report, the EBA analyses the implications of introducing a depositor preference across the Union. The analysis concentrates on 368 resolution entities and a total number of 862 bank institutions, covering 63%-74% of the EU domestic bank assets.

According to the EBA, 187 of the total 368 banks have a resolution strategy, while 181 have a liquidation strategy. The choice between the two is driven by the public interest assessment (PIA). The SRB revised its approach to the PIA in May 2021 to account for system-wide events, likely broadening the universe of banks facing a resolution instead of liquidation. A wider application of the PIA would, in practice, mean that a wider range of entities would need to raise MREL liabilities pushing up MREL supply.

The EBA compares the ability of banks to reach an 8% threshold to tap resolution financing arrangements with their buffers as of end-2019 and with the current hierarchy of claims. This baseline is then compared with introducing a depositor preference across the Union.

The study finds that 272 banks out of 368 would, in the current regulatory framework and non-stressed level of loss-absorbing capacity, have sufficient bail-in capacity to reach the 8% TLOF threshold that is required to access resolution financing arrangements, without touching any deposits. 96 banks would incur losses on some types of deposits, impacting €18bn of (mainly non-preferred) deposits. The chart below shows that the largest losses on deposits would be borne by banks in France, Sweden, Spain and Germany. Deposit losses would be borne especially by medium-sized banks that have a relatively high reliance on deposits as a funding source. The EBA data suggests that 74% of small banks have either a high or medium-high reliance on deposits. The ratio for medium-sized banks is 48% while for large banks it is only 27%.

An EBA study finds that 74% of banks have sufficient bail-in capacity to reach the 8% TLOF threshold without touching deposits

Furthermore, of the 96 banks, 81 would see their non-preferred deposits being hit, while eight banks would need to bear losses on their preferred deposits as well. Two banks would see losses on all types of deposits (including covered deposits). Five banks would not meet the 8% threshold even if exposing all deposits, and interestingly of these, two were large banks.

The EBA also assesses the ability to tap DGS in the case of bank problems. Of the 91 banks that would need to share losses with deposits to meet the 8% threshold, only three could access a DGS and only two out of them could draw funds that would be sufficient to reach the 8% threshold.

In the case of a depletion of CET1 capital both for the combined buffer requirement and Pillar 2 requirement in the run-up to the resolution, there would be a change in the picture such that only 60 banks would meet the 8% threshold without exposing deposits. Only a 75% depletion of buffers would result in 122 banks reaching the threshold, while 198 banks would see deposits being hit. Also for the different CET1 depletion scenarios, the number of banks that could access a DGS and that could obtain an intervention to reach 8% TLOF is limited.

The EBA studies the impact of an introduction of a depositor preference that would change the creditor hierarchy. It would remove non-preferred deposits from the layer that also includes preferred senior unsecured, structured debt and derivatives. This layer would therefore become thinner in countries that have so far placed all these at the same level. This could be reflected as a higher loss given default for preferred senior unsecured in these countries, other things being equal, which should also be reflected as higher cost of preferred senior funding for these banks. Depositor preference would aim to provide more protection for deposits as they would move up in the hierarchy of claims. The impact on each type of deposit would depend on the details of the deposit preference (single-tier, 2-tier or 3-tier approach).

EBA notes that a deposit preference would increase the number of banks reaching the 8% TLOF threshold without jeopardising deposits

Assuming a deposit preference over other ordinary unsecured claims would according to the EBA significantly increase the number of institutions that could reach the 8% TLOF threshold without jeopardising deposits from 272 banks to 317 banks. The burden-sharing on deposits to reach the 8% TLOF threshold would decrease from €18.3bn to €6.4bn for all different deposit preference options. The EBA notes that here the loss exposure of covered deposits should be seen together with a potential for more extensive usage of DGS funding in resolution. The number of banks that could access a DGS would increase to 41 banks, assuming a single-tier deposit preference. This would result in a maximum amount of available DGS funds across the different deposit preference options due to the highest-burden on covered deposits, the main driver for the DGS contribution.

According to the EBA, the analysis is very sensitive to the recovery rate assumption, with lower recovery leading to a higher DGS usage frequency. The EBA notes that banks that would not obtain a high enough contribution from a DGS are concentrated in one member state. More frequent involvement of DGS would however result in higher costs for the industry and clients according to the EBA, because of the required payment of contributions needed to increase the DGS back to the targeted level. In general, the number of banks accessing DGS would potentially be higher with a single-tier depositor preference than with a three-tier depositor preference.

2022 will see advances in making European banks more crisis-proof. In the second half of 2022, the European Commission is expected to publish the proposal for the review of the crisis management framework. Bank resolution planning and the MREL, builds on this framework.

According to the EBA analysis, an introduction of a depositor preference would increase the number of banks that could reach the 8% threshold to access resolution financing without putting deposits at risk, compared with the current system. As depositor preference is not currently in use in most EU countries, the change could potentially result in a higher cost of issuing preferred senior unsecured debt for many banks, while the status of deposits could become safer. It is possible that an introduction of a depositor preference system could also result in lower subordinated MREL requirements, as seems to be the case for Southern European banks. In the longer term, this could even result in some shift between non-preferred senior and preferred senior issuance for the MREL.

If the review resulted in a larger set of banks being put under the resolution umbrella, further smaller entities would be required to build MREL buffers. Broader access to DGS funds could instead potentially result in less public-sector involvement, but might also result in a higher burden for the banking sector, in the form of higher contributions to the system.

In addition to the potential proposal for the CMDI review, in 2022 banks also face the first extensive wave of binding loss absorption requirements in the form of interim MREL requirements, which begin in January. The final requirements will be binding in 2024. Banks are well-positioned to meet the 2022 interim requirements but are expected to have a €42bn shortfall for the final targets for 2024.

On a country level, we consider that Greek banks especially lag behind in terms of building loss-absorption buffers. While Greek banks have a longer time to meet the requirements, the shortfall is substantial for the sector and meeting the targets, is in our view, highly reliant on the market prospects remaining supportive.