The quest for a more efficient bank resolution regime in the European banking union

1. Introduction 

Hello, and thank you for the invitation to speak here today. As an economist in a room full of lawyers, I am definitely playing an away game. I hope the crowd will be friendly. 

Let me start by congratulating Unidroit on its centenary. It is a little older than the Bank for International Settlements (BIS), which must wait four years to reach this landmark. But both have remarkable longevity.

Both the BIS and Unidroit were established in the period after the Great War. It was a time when the world was seeing the first efforts to build an international order designed to help maintain peace and stability by providing the means to resolve disputes before they erupted into bloodshed. Both have survived the tumult of the past century with its devastating international conflicts, tectonic geopolitical upheavals and economic crises. The fact that these two international institutions are still thriving and working towards common solutions – whether in international law or global monetary and financial stability – is a testament to the value and necessity of international cooperation.

My knowledge of Unidroit was cursory when, in 2021, Professor Ignacio Tirado invited the BIS’s Financial Stability Institute (FSI) to collaborate in producing a legislative guide on liquidation regimes for banks. We accepted willingly and joined Unidroit in what proved to be a demanding but rewarding experience. 

Over the years, the FSI has conducted several projects on bank insolvency regimes. Two facts that consistently emerged from our studies explain our interest in the Unidroit initiative. The first is the diversity of regimes across jurisdictions. This is clearly a problem where a bank failure has an international dimension but may also result in sub-optimal outcomes at the national level if the domestic framework is not adapted to the specificities of banks. This links to the second fact: many of those regimes are unable to preserve value and avoid costs for taxpayers because they lack appropriate tools for managing the risks presented by bank failures.

The Unidroit Legislative Guide on Bank Liquidation aims to address this by supporting jurisdictions in designing effective liquidation frameworks that recognise the special nature of banks. Over time, it should also promote alignment and reduce detrimental disparity. The Guide was adopted last summer and, to mark the achievement, the FSI hosted an event with Unidroit in Basel. We now use it regularly in our training and outreach activities.

I will devote the rest of my remarks precisely to bank failure regimes. Since this centenary event is in Europe, my focus will be on the European Union, or more concretely on how the current crisis management framework within the European banking union could be improved. 

2. The functioning of the Single Resolution Mechanism

The banking union project – composed of the Single Supervisory Mechanism (SSM), the Single Resolution Mechanism (SRM) and the still pending European deposit insurance scheme (EDIS) – represents a significant step towards further integration and better crisis prevention and crisis management. 

The SRM in particular has made, in just over a decade, remarkable progress. From a standing start, the project has put in place the arrangements and capabilities to make resolution a credible option. All significant banks have mature resolution plans and achieve their targets for minimum requirements for own funds and eligible liabilities (MREL), meaning they should have sufficient loss-absorbing capacity to support their resolution. That should reduce the need for external funding from resolution funds or public backstops. However, despite the framework’s successes, its efficiency is undermined by several features that derive from the unique institutional structure in Europe. 

By “efficiency”, I mean a framework’s ability to achieve its objectives without unnecessary complexity, uncertainty and costs. The concept differs from effectiveness – the ability to produce the intended outcomes. 

Inefficient frameworks generally result in unnecessary burdens and costs for authorities and firms. The SRM is no exception. These shortcomings take on particular relevance when, following the Draghi and Letta reports,¹  policymakers agree on the need to foster competitiveness by increasing the efficiency of EU regulation.

Moreover, the defects I will discuss ultimately limit the ability of the resolution “pillar” to support the objectives of the banking union. Those objectives are clear – to denationalise banks’ risks and to promote a more integrated European banking market. Neither of those goals has been fully achieved. There has been little progress in integrating the banking industry, and the bank-sovereign nexus has not been fully eliminated because, once other options are exhausted, public backstop funding for crisis management is still national.  

In the rest of this speech, I will outline three features specific to the SRM that undermine its efficiency and discuss how each is linked the European institutional framework. I will then briefly compare the SRM with other frameworks, before exploring the measures that would be needed to address those issues.

3. Features of the current banking union resolution framework

I will focus on the three features that are most consequential for the efficiency of the SRM. 

The first relates to procedural complexity. Resolution in the banking union involves multiple authorities. The Single Resolution Board, or SRB, is responsible for all significant banks. However, it cannot act alone. When a bank is failing and rapid intervention is required, the adoption and implementation of an SRB resolution scheme involve the ECB, the European Commission, possibly the Council and relevant national resolution authorities – in addition to the SRB itself. 

This is largely due to constitutional limitations. Since the SRM is an agency and does not have an explicit basis in the EU Treaties, legal principles on delegated powers apply. Although this doctrine is evolving – many of you will be familiar with the details of the Meroni case law²  – limits remain on the exercise of discretionary powers by the SRB. This explains the involvement of the Commission and other authorities in a complex decision-making process in circumstances in which time is of the essence. 

The second feature relates to the multi-layered nature of the framework. Resolution is governed by a complex combination of European and national laws. This characteristic of the EU framework is not unique to resolution, but it creates material complexity and inefficiencies in this context. 

The SRM coexists with national resolution regimes. In business as usual, resolution planning for significant banks involves both the SRB and national resolution authorities. Loss absorbency requirements – MREL – are calibrated by the SRB under European rules but imposed on banks under national rules. In the event of a resolution, the SRB resolution scheme must be executed by the relevant national authority using its powers under the local resolution regime.  

Moreover, resolution frameworks are underpinned by a patchwork of unharmonised national insolvency regimes. Among other things, this significantly complicates the safeguards for creditors, since the maximum losses they can suffer in resolution are determined by the losses they would have sustained in a liquidation – a yardstick that varies depending on the applicable national insolvency regime.  

The third feature relates to funding. This is arguably the area where defects are most consequential. Funding for resolution can come from three main sources: (i) the bank’s creditors and shareholders, through internal loss-absorbing capacity; (ii) external funding arrangements, such as deposit guarantee schemes and resolution funds; and (iii) taxpayers, through public backstops. 

In the banking union, access to external funding arrangements is subject to conditions, and the amount available is capped. For example, the Single Resolution Fund – the SRF – may be used only after shareholders and creditors of the failing bank have absorbed losses equivalent to at least 8% of the bank’s total liabilities and own funds. Once this hurdle is cleared, the funds available from the SRF cannot exceed 5% of the value of the bank’s total liabilities. In addition, there is no provision in the SRM regulation for extraordinary direct government support in resolution. 

The corollary of these limitations on external funding is an emphasis on banks’ loss-absorbing capacity as the primary source of resolution funding. This explains the stringent and detailed MREL framework that is costly for banks and complex to implement. 

This approach to funding is rooted in an aversion to sharing the costs of a bank failure that has also hampered the completion of the banking union. It is the reason why EDIS has not yet been agreed to, why the firepower of the SRF is ultimately limited and why, as a result of all that, bank risks are not yet fully denationalised. 

4. Are other jurisdictions more efficient?

I have outlined three areas in which the inefficiencies of the SRM are linked to institutional features and complexities inherent in the banking union: the involvement of multiple authorities, the coexistence of EU and national rules and the constraints on mutualised funding. How does the SRM measure up against other frameworks in terms of efficiency? If we compare it with two jurisdictions with developed banking markets – the United Kingdom and the United States – the SRM performs worse in many respects. I will give a few brief examples.³

The number of authorities involved in resolution and the complexity of the decision-making process in the SRM is sui generis. There is nothing comparable in these other jurisdictions where a designated resolution authority – the Bank of England or the Federal Deposit Insurance Corporation – has the necessary powers to act. In those jurisdictions, decision-making involves other bodies only in exceptional circumstances, such as where extraordinary funding is required. 

Regarding the interaction between resolution and insolvency, the SRM does not score well either. Unlike resolution, insolvency regimes are entirely national with only limited harmonisation. This is a clear source of complexity that does not exist in other jurisdictions. For example, in the United States all bank failures are dealt with under a single framework. 

On funding, the UK and US arrangements are more flexible. Both frameworks contemplate the possible need for extraordinary funding and provide a procedure for authorising its use – for example, the systemic risk exception in the United States. In parallel, the rules governing resolution-related loss absorbency are materially less complex.

5. Measures to improve efficiency 

In short, the SRM is more complex than similar frameworks in comparable jurisdictions. It is characterised by inefficiencies that largely result from institutional features specific to the EU and banking union frameworks, the constitutional constraints on the SRB and hardwired political agreement on sensitive issues such as sharing the financial burden. Any meaningful improvement requires those features to be addressed. 

I am not saying that is easy. Indeed, removing the constitutional constraints on the SRB would require a treaty change. This means that the involvement of multiple parties in resolution will remain for the foreseeable future. Nevertheless, a lot could be achieved through legislation to mitigate the other two areas of inefficiencies.

The interaction of national and EU regimes could be simplified by expanding the scope of the resolution framework relative to insolvency. Recent reforms to the EU crisis management and deposit insurance framework – the crisis management and deposit insurance framework (CMDI) – have taken a step in that direction by modifying the public interest assessment so that more failing banks meet the threshold for resolution.

However, the CMDI does not fully solve the issue. Since there are fewer constraints on the provision of external funding (including public funds) under national insolvency proceedings than under resolution, a failing bank may still – even if it considered systemic – be wound up rather than resolved. This is because more funds are available in liquidation to deal with the impacts of the failure.⁴

Action is also needed to further harmonise national bank insolvency regimes. As long as differences in creditor hierarchies and rights in liquidation remain, the application of the “no creditor worse off” (NCWO) safeguard will remain disproportionately complex for cross-border banks. Similarly, differences in the tools available under national liquidation regimes may further complicate the NCWO assessment. Those differences may also result in varying outcomes to the public interest assessment that determines whether a bank is to be resolved or liquidated. The assessment requires the SRB to compare outcomes in resolution and insolvency, and the tools available under the national regime will affect the latter. Fragmented insolvency regimes are also a brake on cross-border banking activities. 

Legislation can address this, although it would be technically challenging. However, the Unidroit Guide provides a good blueprint. 

Turning to funding, the reliance on banks’ loss-absorbing capacity as the primary source of resolution funding has resulted in an overly complex methodology that seeks to deliver precision in calibrating the amount of MREL each bank will need to support its resolution. That precision is probably illusory, given that the circumstances of failure are unknown. Legislation could simplify the MREL regime and reduce its stringency in appropriate cases.

However, changes to MREL would be conditional on action to make external funding arrangements more flexible. This means relaxing constraints on the use of the SRF and providing for public financial support in exceptional cases, when other tools prove unable to achieve the resolution objectives. 

The link between MREL levels and external funding is demonstrated by recent UK reforms, where the creation of a dedicated funding arrangement to support resolution transfers was coupled with a radical simplification of MREL for banks with a preferred transfer strategy.

The CMDI has taken a step in this direction by increasing the amount of funding available from the deposit insurer for a transfer transaction and facilitating access to the SRF. However, it falls short of a full solution for several reasons. For example, the facility for funding transfers is subject to complex conditions and only fully available to banks below an asset threshold of €80 billion. Moreover, the SRB cannot take account of any prospective external funding when setting MREL. Therefore, the CMDI would in principle have no impact on the stringency and complexity of MREL arrangements. 

Ideally, a full solution would include a common depositor protection arrangement that can support resolution actions within the banking union. Although the creation of a common deposit insurance scheme is not directly mentioned in its recent communication,⁵ it is positive that the European Commission envisages a proposal “to review and simplify the structure of the deposit framework to better align the responsibilities of and financing of crisis management and deposit insurance measures…”. Of course, this issue will remain subject to difficult political negotiations. 

The reforms I have suggested need to change the limited burden-sharing that underpins the current framework. This requires political will and real commitment, not only to increase efficiency but, beyond that, to deliver on the core objectives of the banking union. Indeed, it seems unlikely that without a complete banking union and a well functioning SRM, supported by sufficient mutualised external funding, the objective of denationalising banks’ risk could ever be met. Moreover, the same actions are a precondition for the removal of the national ring-fencing measures that currently hinder the integration of the European banking industry. 

There are some easy fixes – streamlining information requirements or resolution plans, for example – that can be achieved through delegated rulemaking. We have already seen work by the European Banking Authority and SRB in this direction. However, such initiatives, while good administrative practice, are little more than tinkering around the edges. They do nothing to address the structural drivers of inefficiency that I have discussed. 

6. To conclude

I will briefly conclude. The resolution framework of the banking union is complex, and complexity has costs. It increases the compliance burden for firms and authorities and renders the framework less efficient.

Much of that complexity arises from institutional characteristics that are unique to Europe. These characteristics derive from insufficient transfer of power and cost mutualisation and the remaining divergences between member states’ legal regimes. Institutional structure, complexity and efficiency are linked, so any meaningful reform that addresses the inefficiencies of the framework needs to consider those features – the layering of European and national rules and insufficient sharing of the risk and costs of bank failures. As I have argued, those reforms are also essential to deliver the denationalisation of banks’ risks and market integration, which are the objectives of the banking union.

In other words, there are no shortcuts. However, there is a huge incentive to get it right in terms of both efficiency and effectiveness. The legislative process which is likely to be launched after the Commission’s recent communication will be a good opportunity to fix this.