From going concern to gone concern: The supervision-resolution continuum after a decade of the Single Resolution Mechanism

Speech by Pedro Machado, Member of the Supervisory Board of the ECB, at the EUI Bank Resolution Academy

Florence, 22 September 2026

I’m delighted to be taking part in this edition of the Bank Resolution Academy^[1].

Let me begin by thanking the Florence School of Banking and Finance for the invitation, and for their continued investment in something too easily taken for granted: building a community of practitioners with a common understanding of supervision and resolution.

Today, I can offer you a point of view, shaped by my own particular experiences in this field. I spent five years at the Single Resolution Board, responsible for resolution planning and decision-making, and – since last year – I’ve served as a member of European banking supervision’s Supervisory Board. I have, so to speak, crossed a bridge between the two pillars of banking union.

And my personal trajectory shapes the argument I will put forward this morning.

We often use the metaphor of pillars to describe banking union: the Single Supervisory Mechanism (SSM) as the first and the Single Resolution Mechanism (SRM) as the second – with the common deposit insurance scheme as the missing third pillar. The solid architectural imagery is reassuring, but analytically misleading.

Pillars stand apart; they carry a load in parallel. Yet, a decade of banking union experience has taught me that supervision and resolution are not parallel structures, rather they are points along a single continuum. This continuum runs from going-concern oversight, through recovery and early intervention, to determining whether a bank is failing or likely to fail, and finally to resolution or orderly market exit.

The credibility of the whole depends on the integrity of each segment and, above all, on the strength of the joints between them.

Why the continuum matters

Why insist on this framing?

Because we must at all times be clear on the fundamentals: the interconnection between supervision and resolution is critical to preserving financial stability.

Bank failures do not happen out of the blue: they stem from a process, which can range from slow-burn deterioration to a rapidly evolving collapse. A bank rarely fails on a Friday evening without leaving a trail of warning signs in its wake.

Asset quality worsens, funding becomes more expensive, governance weaknesses compound and confidence – the true currency of banking – erodes, sometimes slowly, sometimes with brutal speed. In Hemingway’s novel The Sun Also Rises, a character is asked how he went bankrupt. His answer: “Two ways. Gradually and then suddenly”.^[2]

The question for public authorities is not simply how to act at the point of failure – we must also ask how the actions taken at each earlier stage determine the options that are available at the end. This has a very practical implication for the supervisory side. Many of the prudential requirements we impose at the going concern stage – on capital, on liquidity, on governance, on data capabilities – could be relevant for a bank’s resolvability.

Conversely, the core elements of resolution planning – the calibration of MREL, the identification of the preferred and variant resolution strategies, the operationalisation of resolution tools, the identification of critical functions – feed back into the broader supervisory assessment on the solvency of the bank and the sustainability of its business model post-resolution.

Recovery plans and resolution plans are drafted by different actors under different legal regimes – by the banks themselves and by the resolution authorities, respectively – but they describe the same bank, the same balance sheet and the same potential trajectory of distress. Of course, the actions undertaken for ongoing supervision differ from those targeting crisis management, and if the latter must be implemented, the bank will necessarily emerge in a different shape and with a different balance sheet. And we must always keep in mind: if the two do not connect, the framework has a gap precisely where it can least afford one.

The first decade of banking union has taught us that this interconnection cannot be decreed; it must be painstakingly built. It rests on information sharing that is timely and granular, not merely formal. It rests on joint crisis simulations and dry runs, which do for institutions what fire drills do for buildings. And it rests on something softer but no less essential: mutual trust between supervisory and resolution authorities; mutual trust ensures that the escalation of a case from one stage of the continuum to the next becomes a coordinated handover rather than a siloed transfer.

And none of this rests on goodwill alone. The Memorandum of Understanding between the ECB and the SRB is the cornerstone of this relationship, organising the exchange of information both in business-as-usual and in crisis conditions.^[3] It is tangible proof that to talk of the continuum is not merely paying lip service to an abstract ideal; it is instead a clear recognition that the core principle of cooperation underpins all our operations.

It is telling, too, that the IMF’s 2025 Financial Sector Assessment Program for the euro area singled out precisely this integration between supervision and resolution as a strength of the European framework. But it also rightly reminded us that maintaining it requires constant effort on data quality, information sharing and the alignment of incentives.^[4]

The continuum tested: what the cases tell us

I will now further build the case for a continuum between supervision and resolution by examining two resolution cases that marked the SRM’s first decade of crisis experience.

The case of Banco Popular from June 2017 is usually told as a resolution story: the first application of the European regime to a significant institution, executed overnight, with shareholders and subordinated creditors absorbing losses, depositors fully protected and no taxpayer money involved.^[5] All true. But look at it through the lens of the continuum and a different feature stands out: the decisive act that set everything in motion was a supervisory one – the ECB’s determination that the bank was failing or likely to fail.^[6]

The failing or likely to fail assessment is the hinge on which the entire framework turns, the point at which going concern gives way to gone concern. That hinge worked in 2017 because supervision and resolution had prepared together, exchanged information intensively in the weeks before and understood each other’s constraints. The speed of action we rightly celebrate was not improvised; it was the product of an invisible continuum.

The case of Sberbank Europe from 2022 tells a distinct yet recognisably similar story. A sudden geopolitical shock – Russia’s invasion of Ukraine – translated to an acute deposit run across a cross-border group within days. The response resulted in different outcomes within a single banking group: the Austrian parent was orderly wound up under national proceedings, while the Croatian and Slovenian subsidiaries were resolved through transfers to local acquirers.^[7]

That differentiation was not a given: it followed from the public interest assessment, conducted entity by entity. Where an entity performed critical functions whose continuity mattered for the local economy – as the subsidiaries did in Croatia and Slovenia – resolution was warranted; where it did not – as with the Austrian parent, whose failure could be absorbed by ordinary wound up proceedings – resolution was not pursued. What made that differentiation possible in a matter of days was, once again, the machinery of the continuum: coordinated failing or likely to fail assessments, and close cooperation between the ECB, the SRB and national authorities across multiple jurisdictions. The continuum operated effectively not only in process but in outcome. Contagion was contained, financial stability in the affected Member States was preserved.

Beyond these two cases, the continuum operates silently every day. A credible end point to the continuum – the demonstrated willingness and ability to resolve – changes behaviour at the beginning of it. Banks that no longer expect public support internalise the cost of failure, build loss-absorbing capacity, and take resolvability seriously as a management responsibility rather than a compliance exercise. Market discipline is itself a product of the continuum’s credibility.

The law as connective tissue

There is another sense in which supervision and resolution form a continuum, and it is one particularly close to my heart as a lawyer: they are governed by a common legal framework.

Both supervisory and resolution decisions involve complex economic and technical assessments, often taken under severe time pressure with far-reaching consequences for shareholders, creditors and other stakeholders. Over the past decade, the EU courts have developed a substantial body of case-law – much of it arising from resolution litigation – that clarifies how such decisions are to be reviewed.

The Courts recognise a margin of discretion where complex assessments are involved; but that discretion is not unfettered. It remains subject to review for manifest error, misuse of powers, procedural soundness, proportionality^[8] and respect for fundamental rights, including the property and business freedoms protected by the EU Charter^[9].

Nowhere is this common legal framework more visible than in the determination that a bank is failing or likely to fail. The SRM Regulation choreographs the entry into resolution as a joint procedure: the ECB, as supervisor, makes the failing or likely to fail assessment and submits to the SRB for its own assessment; the SRB then determines whether there is any reasonable prospect that alternative private-sector or supervisory measures would prevent the failure, and whether resolution action is necessary in the public interest. The Court of Justice made the legal position clear in the ABLV case. It held that the ECB’s failing or likely to fail assessment is a preparatory act that produces binding legal effects only through the resolution decision that follows and can therefore only be reviewed within the action challenging that decision.^[10]

A key finding of the judgment is that the transition from going concern to gone concern is not treated as a handover between two separate authorities, but as a single procedure involving the relevant authorities

The SRB – and, where the procedure so requires, the European Commission – must therefore assess the resolution conditions and take responsibility for the resulting decision. Judicial review of that decision therefore extends to all findings made throughout the process, including the assessment that the bank was failing or likely to fail, as the General Court’s subsequent judgment on the merits in ABLV confirms.^[11]

For its part, the ECB has consistently defended its assessments, both publicly and before the EU courts, including in the ABLV, PNB Banka and Sberbank proceedings.

But the process does not end once the resolution scheme is adopted. Its implementation brings the supervisor directly back into the picture. The purchaser of the business must be assessed and, where relevant, receive approval for its qualifying holding. A bridge institution must be authorised. The licence of the residual entity must, in due course, be withdrawn. And the bank that emerges from resolution, whether recapitalised through bail-in or reshaped through a transfer, returns to going-concern supervision, typically under enhanced monitoring and with a restructuring plan yet to be implemented.

In other words, resolution is not an exit from supervision. It remains part of the same continuum

The process therefore comes full circle. Cooperation between the two authorities is just as essential on the day after resolution as it is in the days leading up to it.

Effective crisis management depends on a sound legal framework. Authorities that understand the limits of lawful action can move fast precisely because they are not moving blindly. When markets see that resolution actions can withstand judicial scrutiny, they view the framework as credible. Legal soundness, in short, is not an obstacle to effective action. It is what makes effective action possible.

CMDI reform and the remaining gaps

Let me now turn to the progress that has recently been made. In April the reformed Crisis Management and Deposit Insurance (CMDI) framework was published in the Official Journal of the European Union.^[12] I will not attempt to cover every aspect of the reform this morning. What matters for present purposes is that the reform focuses on exactly the same part of the process as my argument: at the points where supervision, early intervention and resolution meet along the continuum.

The package revisits early intervention measures, resolution conditions and financing arrangements. In particular, it strengthens the early intervention framework and gives the ECB a stronger role at that stage. The relevant powers are now anchored directly in the SRM Regulation, removing the legal uncertainty associated with the previous regime, which depended on national transposition. The escalation path from supervisory measures towards resolution, including the cooperation and information-sharing requirements between the ECB and the SRB, has also been made clearer and more predictable.

The space between supervision and resolution, where clarity matters most, is precisely where the legislator has sought to strengthen the framework. Its effectiveness will now depend on implementation and on all the authorities involved using these tools as parts of a single framework, rather than as instruments serving separate mandates.

But a continuum is only as strong as its weakest segment, and we should be clear about where the remaining gaps lie.

The first is liquidity in resolution. Loss absorption is now well provided for, with banks having made significant progress in meeting their MREL requirements. Although it happened out of the spotlight, it is one of the decade’s key achievements. But even a well-capitalised bank emerging from a resolution weekend must be able to fund itself on Monday morning. And that requires market confidence to return quickly.

Banking union still lacks a mechanism to provide liquidity in resolution, supported by a common fiscal backstop capable of responding to systemic stress. What has been agreed upon for the European Stability Mechanism’s common backstop to the Single Resolution Fund, once in force, will mark an important step forward.^[13] But it will not by itself address liquidity needs in a situation of extreme stress.

This, too, is a continuum problem that cannot be confined to a single stage of the process. Liquidity pressures build up while a bank remains a going concern, crystallise at the point of failure and ultimately determine its viability in the days after resolution. No single authority is responsible for every stage of that process, which is exactly why it calls for a joint approach.

The second concerns the harmonisation of the crisis-management framework itself.

Differences in national insolvency regimes complicate the counterfactual underpinning the no creditor worse off principle, i.e. what creditors would have received if a bank had gone through insolvency rather than resolution. They also mean that outcomes can depend, in part, on where a bank happens to fail. The CMDI reform represents an important step forward in this area, but further alignment of national insolvency and crisis-management rules remains necessary if resolution outcomes are to be consistent and predictable across the banking union.

The third, of course, is the missing pillar of banking union: a common deposit insurance scheme. Its absence continues to create differences in depositor protection, makes genuinely cross-border banking groups harder to build, and means banks and investors cannot price resolution risk consistently across the Single Market.

As long as this remains the case, the banking union will continue to reproduce some of the fragmentation it was created to overcome. And in the current debate on the competitiveness of the European banking sector, completing the banking union is not an optional extra. It is an essential part of the answer.

The same point is reflected in the Commission’s July Communication on the competitiveness of the banking sector and the Single Market in banking. It identifies fragmentation along national lines as a major constraint on the sector’s ability to finance Europe’s investment needs. And it points to completing the banking union, including its deposit insurance pillar, as an essential condition for a genuine Single Market in banking.^[14]

A banking sector whose safety net stops at the border cannot fully deliver the scale that the Single Market makes possible.

To these three, let me add a fourth and more recent challenge: non-financial causes of failure. Our framework was written with financial failure in mind, namely losses eroding capital and runs draining liquidity. Yet recent experience has shown that a bank can also be brought to the brink of failure by non-financial factors, such as a catastrophic IT or cyber incident, the sudden imposition of sanctions or the impact of war.

The failing or likely to fail assessment is sufficiently flexible to capture failures arising from such events. But much of the crisis-management toolkit – from valuation and bail-in mechanics to funding in resolution – was designed with balance sheet stress in mind rather than operational collapse.

Authorities have already begun discussing the issue, including with the Commission services, and it is increasingly clear that both the European legislator and, at the global level, the Financial Stability Board will need to consider it more systematically. For today, I simply want to highlight this as an area where the crisis-management framework may need to evolve over the coming decade.

A final word to practitioners

Before I conclude, let me say a few words to you, the participants in this Academy. Many of you work in supervision; many in resolution; some of you, like me, will move from one to the other over the course of your careers. My message is that these are not two professions living in splendid isolation. They are two parts of the same function, performed at different stages of a bank’s life. A supervisor who understands how a bail-in or a transfer tool is executed will be better equipped to supervise a bank as a going concern. A resolution planner who understands how a business model deteriorates will develop better strategies for a bank in a gone-concern situation.

Bringing these two perspectives together helps us strike the right balance between discretion and accountability, and between speed and legal certainty. This, ultimately, is what gives the framework its credibility.

The banking union’s first decade has established the operational credibility of the crisis-management framework and firmly embedded it in the European Union’s legal order. Its second decade will be shaped by our ability to complete the framework, particularly in areas such as liquidity, harmonisation and deposit insurance. Equally important will be our ability to adapt it to a financial system increasingly determined by geopolitical fragmentation, digital transformation and the growing role of non-bank intermediation.

That work will depend on people who understand how different parts of the framework connect, from going concern to gone concern and back again.

But completing the continuum is not only a task for the legislator; it is also a task we can pursue within our existing mandates. By bringing supervisory and resolution planning closer together where requirements overlap, streamlining approval processes and reducing turnaround times, the ECB and the SRB are actively seeking to simplify the demands that the two frameworks place on banks. This includes faster joint processing with the SRB of applications to reduce own funds instruments and, since July, of applications for the early redemption of MREL instruments.

These may not be changes that make headlines, but they are an important part of making the framework more effective in practice.