Andreas Dombret: Now or later? Completing the European Banking Union

Lecture by Dr Andreas Dombret, Member of the Executive Board of the Deutsche Bundesbank, at the Graduate Institute Geneva, organised by the International Centre for Monetary and Banking Studies (ICMB), Geneva, 20 February 2018.

The views expressed in this speech are those of the speaker and not the view of the BIS.

Central bank speech  | 
28 February 2018
PDF version
 |  5 pages

1 Intro

Dear Charles Wyplosz,

ladies and gentlemen

Thank you very much for inviting me back to Geneva. When one has the chance to address an audience such as the one here at the Graduate Institute, one usually has a hard time deciding on the topic, because the academic curiosity gathered in this room will allow you to embark on almost any intellectual journey. But knowing that Charles has a special interest for the grand scheme of things in the euro area, and given my affiliation to European banking supervision, it was easy to pick the European banking union as a subject and the ongoing question: Should we complete it now or later?

2 The banking union and ongoing stability risks

When we are talking about the future of the banking union, we should not do so without looking at its past. So let us briefly jump back to the early summer of 2012. At that time, plans for the European banking union were still vague, but it quickly began to take shape, partly due to still very vivid memories.

In the run-up to the crisis, banks in the euro area had lost confidence in each other and refrained from lending. Depositors in countries that were hit by the crisis lost confidence in their domestic banking systems and withdrew deposits in significant volume. National supervisory authorities were themselves contributing to the crisis by ring-fencing their national banks in an attempt to block contagion. This, in turn, dried up the credit channels in the very countries that would have needed economic stabilisation the most.

One of the bold steps taken in Europe to prevent a repetition of such events was to create the banking union. And it was through great determination that the first pillar was put in place. Just two years after the idea started to gain traction, in November 2014, the common supervision over banks of the euro area - the Single Supervisory Mechanism - went fully operational. By then, supervisory staff totalling about 1,000 employees1 had already been deployed at the ECB; the balance sheets of all large euro area banks which would be under direct ECB supervision were subjected to a thorough assessment - and where capital gaps were discovered, strict plans on how to fill them had to be submitted. Let me add from my personal perspective that many supervisory processes also had to be reorganised in advance - given, for example, that the new regime requires that all supervisory decisions be passed by the Supervisory Board, which comprises the supervisory authorities from all 19 countries. So the first pillar was created with great vigour.

And at the beginning of 2016, the second pillar - the Single Resolution Mechanism - likewise became operational. By then, another major component of the new resolution regime for European banks, the Bank Recovery and Resolution Directive, or BRRD, had been implemented in all member states. Please bear in mind that the European Resolution Fund is still in the process of being filled. Its full capacity of 1 percent of covered deposits or roughly € 50-60 bn is planned to be achieved by 2023. Today, the volume stands at € 17.4 bn.

The overall impression is that the banking union started with a great deal of momentum and a lot of changes.

And this has caused some to ask: Why shouldn't we complete the banking union by erecting its third and final pillar with equal determination?

Indeed, thoughts on a third pillar of the banking union have been gaining more and more attention of late. One likely reason is that, despite many reform measures and efforts from all sides, we do not yet regard the euro area as being crisis-proof.

Consider, for example, the large piles of non-performing loans held in bank balance sheets in several member states. Although some banks have succeeded in reducing them, the overall stock of €950 bn across the entire European Union remains a significant threat to stability. What's more, some member states are still vulnerable due to high levels of public debt. At the same time, the strong connection between a country's government and banks domiciled there continues to exist. I am referring to the notorious sovereign-bank nexus which played an important role during the euro crisis.

As we cannot rule out that someday, another crisis will challenge the euro area, the big question is: What still needs to be done - and what can still be done?

It is in this context that the third pillar of the banking union is talked about, with arguments based on perceived justifications such as the idea that "we have already put in so much effort - now it's time for the third pillar of the banking union."

You are probably aware that the Bundesbank is among those institutions that have argued against a premature implementation of the third pillar in terms of a European deposit insurance scheme, or EDIS. Our position is that we should only consider this pillar once major reforms have been achieved, and risks reduced.

I want to take this opportunity today to elaborate on that position.

First, I believe it is important to stress the reasoning behind our position, and the fact that we are indeed very much interested in constructive ways to overcome euro area stability threats. This becomes even more important as debates on the further development of the euro area architecture frequently refer only to either "transfer union" fears or "European solidarity" hopes. But things are more complex than that. That's why I have decided to present my ideas on the further evolution of the banking union on the basis of two reproaches that I frequently encounter.

  • The first is: Do we not propose implementing reforms "later" to avoid having to push through reforms at all?
  • And the second is: Given that we are still not safe from a future euro area crisis, isn't this a sign that we're in a deadlock situation where completing the banking union "later" rather than sooner would only add to the fragility of the system?

3 What is meant by "later"?

First, let me explain why I believe that an immediate implementation of the third pillar of the banking union would be premature, and what "later" means in more precise terms.

First of all, I would like to take a glance at the core idea of deposit insurance. It is meant to secure deposits of retail investors and limit the likelihood of a bank run. This is an important goal. What is more, the mere presence of an incontestable deposit scheme may do good by preventing bank runs that are entirely self-generating. So, deposit insurance may have more to offer beyond being the entity disbursing private depositors. And in an increasingly integrated financial market across Europe, a European insurance scheme does make sense.

But, of course, we have to bear in mind the status quo in Europe. Through European legislation, we have already achieved harmonisation of national deposit guarantee schemes. National schemes are guaranteeing deposits up to €100,000 for private customers. Today, all euro area member countries have implemented the content of the relevant European directive at the national level.

One also should bear in mind the functionalities of national solutions, as they sometimes extend far beyond mere disbursement. In Germany, for example, savings banks and cooperative banks have their own schemes in place which are designed to prevent institutions from becoming insolvent in the first place.

So even the design of a potential European deposit insurance scheme would be a demanding task.

My point is, then, that we haven't got that far yet. Let me now, as promised, explain why concrete considerations on the third pillar have to wait until other and more eminent issues have been resolved.

Essentially, the argument is about looking beyond ideas that may seem promising in a narrow sense - like EDIS - and acknowledging that the entire architecture first has to be made ready for this change. One has to realise that the actual impact of a reform hinges on the entire system evolving. In other words: we have to take a system-wide view.

For the financial system, we have invested a great deal of time and effort in ensuring that market forces work not just in normal times but also in times of downturn and distress. This environment is conducive not only because of its disciplining effects on both debtors and creditors, giving them a clear and credible framework within which to act. It is also a key ingredient of the overarching architecture: because in a market environment, if individual action is not matched by individual liability, the rules can hardly be deemed sustainable.

These considerations need to be taken into account in connection with a European deposit insurance scheme. A new insurance can only be in line with market mechanisms if there is a level playing field from the very beginning.

Specifically, this means that before entering a phase of mutual insurance or risk sharing, we must ensure that the risks that are already known - let's think of them as legacy risks - are sufficiently contained. Otherwise, moral hazard risks are likely to emerge.

What are those risks where banks in the euro area are concerned?

Certainly, as we know today, the piles of non-performing loans need to be reduced significantly. Another risk to be levelled concerns the build-up of bail-in-able capital: the resolvability of banks will be heightened once institutions have met their individual target levels. Furthermore, national insolvency laws diverge to an extent that has a significant impact on risks to the national banking sector. These risks need to be addressed as well.

Finally, I would like to spend some time on another major threat that stems from a continuing sovereign-bank nexus - a situation that involves a real and well-known risk of fiscally overburdened states possibly infecting their own banking system and triggering a solvency crisis at banks. And this sovereign-bank nexus, in turn, affects the ex-ante likelihood of a European deposit insurance scheme becoming involved.

In order to improve the current situation, a crucial element has to do with current incentives contained in banking regulation. At present, sovereign bonds from member states of the euro area have a zero risk weight under the capital requirements regime. Essentially, this means that the economic risk attached to sovereign debt (which is larger than zero) is not reflected in capital requirements. Nor is there any upper limit for exposures to specific sovereigns. This preferential treatment is a factor that actually contributes to the sovereign-bank nexus and therefore needs to be altered.

Please bear in mind that given the risks that are currently not resolved in several areas, a third pillar is not only unjustified because of the heterogeneous levels of risk at national banks. It would also reduce incentives to address those national issues if a mutualised insurance scheme were in place.

So as you can see, the reasons for postponing EDIS are founded not on an ideology-based rejection, but very much on stability considerations.

4 The road to stability: are we locked in a stalemate?

This brings me to a second mindset I frequently encounter when discussing the future of the banking union. It concerns the idea of a trade-off between crisis mitigation and crisis prevention measures.

Essentially, some interpret the ongoing risks for the euro area as a sign that we are approaching a stalemate situation, where self-recovery potentials in the financial system will be exhausted - as will options for action. In this situation, the reasoning goes, would it not be better to compromise on short-term oriented "flexible" measures - however counterproductive in the medium and long term? And would this not be an argument for an immediate completion of the banking union?

The great value of systemic analyses is beyond doubt. But at the same time, I wonder whether a simple trade-off between short-term and long-term goals might not be the result of over-hasty generalisations.

Let me set these thoughts in the context of our main topic of European deposit insurance. My guess is that here, too, hasty calls for new policy agendas are partly the result of a too-narrow focus on this matter with regard to European financial stability.

First of all, as I mentioned before, national schemes with harmonised minimum requirements are already in place across Europe. Moreover, deposit insurance seems too-narrow a focus when looking at financial stability issues. For example, in four cases of ailing institutions in the euro area last year, we saw four entirely different outcomes.

And as far as criticism is concerned, the issue has not been about deposit insurance, but about an appropriate level of burden sharing. One lesson learned from last year is that we need to align national insolvency laws with the aims of the new European resolution regime.

This leads me to a more general point: the question whether we've run out of options and are approaching a stalemate regarding financial stability in the face of risks to the euro area financial sector. I have two points to make on that.

First, even though the first two pillars became operational some time ago, we're still in the middle of the regime shift. For example, looking at the new resolution regime, institutions as well as investors are learning to adapt to their new role as assigned by the BRRD. That is an ongoing process. Also, banks' individual targets for bail-in-able capital are just being laid down. Once these targets have been met by all banks, this will have a significant impact on the loss absorption capacity of failing banks and thus considerably strengthen their resolvability.

Second, there is definitely scope for action. And this is true of banks as well as of existing components of the banking union. For example, although the Single Resolution Board has been fully operating for over two years now, it is currently assessing resolution plans of large euro area banks - this is another milestone towards being able to act swiftly in case of an emergency.

But of course, banks themselves have room for manoeuvre. We saw this, for example, in the shape of last year's NPL reductions after SSM supervisors had pushed for more action. This is to say that the pressure exerted by supervisors already made a difference. That should encourage European supervisors to continue demanding rigorous action in this field. And such a demand is justified not only because it's part of risk management responsibilities in institutions, but also because resolute action creates better long-term prospects for the individual banks as well as for regional economic recovery.

5 Conclusion

Ladies and gentlemen

Expressed in one sentence, the main thrust of my talk would be that, for the further evolution of the euro area, we should remain sceptical about an inevitable deadlock, just as we should be wary of all-too-easy solutions.

This applies to the banking union as well. As regards the initial question on whether to complete the banking union now or later, my answer is: The third pillar, which sounds like an easy solution to some, is actually not a solution at all to the most pressing problems of the euro area. It may therefore be better to postpone the envisaged third pillar altogether - out of concern for stability.

It is possible to stand quite firmly on two legs for the time being. An incomplete banking union is in itself no reason to become uneasy, because we are not at point zero anymore, but right in the middle of a major paradigm shift, and the existing financial architecture has left us with a lot of room for manoeuvre.

My argument is that we are not stuck in a structural deadlock. Instead, we have to think of different paths as being available for the euro area. In that respect, our daily task is to make sure that we are on the most favourable path.

1 in Full Time Equivalents (FTEs).