VoxEU Column EU institutions Financial Regulation and Banking

The European banking union at three: A toddler with tantrums

The recent resolutions of the Spanish Banco Popular and of two smaller Italian banks – Veneto Vanca and Banca Popolare di Vicenza – can be seen as a first important test for the banking union. This column assesses the progress made over the past three years. It argues that a ‘never bailout’ rule is inefficient, especially if referring to legacy problems; that a crisis should be resolved before a new regulatory framework is put in place; that to avoid national solutions, we need to go to a complete banking union; and finally, that the process will take some time, and new institutions and regulations are only a small step.

A little less than three years ago, the bank regulatory framework in the Eurozone seemed to have reached a decisive moment, with the completion of the Comprehensive Assessment and the establishment of the Single Supervisory Mechanism and Single Resolution Mechanism, and the implementation of the Bank Recovery and Resolution Directive (BRRD). Taxpayer-funded bailouts were supposed to be replaced with bail-in of equity and bondholders, bank supervision strengthened, and the deadly embrace between governments and banks – at the core the Eurozone crisis – loosened, if not cut. While this represented enormous progress compared to the situation in 2008 (with no bank resolution framework and little cooperation across borders), there were still significant criticisms, related both to rather complicated resolution mechanism and the lack of a Eurozone-wide deposit insurance scheme (e.g. Schoenmaker and Wolff 2015, Beck et al. 2017b). There were remaining doubts on the insolvency of several Italian banks that had passed the Comprehensive Assessment, including two of the banks recently sent into liquidation.

The events at the end of last month can be seen as a first important test for the banking union. First came the resolution of the Spanish Banco Popular, taken over by Santander while its equity and junior bondholders’ claims were wiped out. No taxpayer money was used; rather, Santander will finance the takeover with a capital raising of €7 billion in the market. Second came the resolution of two smaller Italian banks, Veneto Vanca and Banca Popolare di Vicenza, declared to be “failing or likely to fail” by the SSM on 23 June. As in the case of Banco Popular, equity and junior bondholders’ claims were wiped out. Unlike in the case of Banco Popular, the Italian government is exposed to losses of up to €17 billion: Intesa Sanpaolo took over the good assets of the two banks for a symbolic one euro and with a €5.2 billion government subsidy; the bad assets were put into a bad bank to be liquidated, backed by a €12 billion state guarantee. Against the letter and spirit of the BRRD, senior bondholders were made whole.

While the resolution of Banco Popular went relatively swiftly, the final decision on the two Veneto banks was a protracted affair, with long negotiations between the European Commission and the Italian government. As the BRRD does not allow the use of taxpayer money for bank resolution before senior bondholders are bailed-in, an exception had to be made. In this case, the Single Resolution Board decided that these two institutions do not constitute a financial stability risk and can therefore be resolved at the national level. Italian government, in turn, argued that bailing-in junior debtholders and liquidating the banks would inflict economic damage on the region, as this would have implied calling in loans, with the European Commission ultimately agreeing to government support for the bank liquidation process. Some observers pointed out that the government would have had to pay €10 billion anyway if it had bailed-in senior bondholders, as these bonds came with a government guarantee. This political manoeuvre to get around the spirit, if not the letter, of the new bank resolution framework and mindset in the Eurozone drew immediate criticism, especially in Germany.

Pass or fail?

Has the banking union already failed its high ambitions of ending bailouts? Has the Eurozone yet again written rules just to break them a few years later? As so often, the events of the past few weeks tell us as much about the shortcomings of the half-baked banking union, as about the ghosts of past crises still lurking in the shadows, the politics of banking regulation, and the high hurdles we face on our way to a Eurozone banking system replacing 19 national banking systems.

One strong argument of Italian policymakers was that the liquidation of the two institutions without any taxpayer support would have burdened the rest of the Italian banking system with the costs, which would have triggered further bank failures. This is an argument that points to the core problem of the Eurozone crisis: the resolution of even small banks is too much for one country (especially with sovereign debt levels as Italy) and that is where the Eurozone financial safety net should have come in. In the absence of a Eurozone wide deposit insurance scheme, there is strong pressure on the national government to step in. The counterargument is that these problems have been known for a long time and should have been addressed years ago, before the banking union came into existence. Not surprisingly (though not necessarily correctly), the Germans and Dutch do not want to pay for problems that should have been addressed much earlier.

In this column, I argue that:

  • A ‘never bailout’ rule is inefficient, especially if referring to legacy problems;
  • A crisis should be resolved before a new regulatory framework is put in place;
  • To avoid national solutions, we need to go to a complete banking union; and
  • This process will take a long time and new institutions and regulations are only a small step.

These arguments are partly confirmed by the differences in the Spanish and Italian experience, as the Spanish authorities have done a better job in cleaning up legacy problems in their banking system, including with the use of an asset management company (AMC). Of course, one can argue that the Spanish case was easier as other banks stood ready to take over the failing bank.

Never say never again

The bail-in rule under the BRRD was the reaction to over €500 billion in taxpayer money used during the Global and Eurozone Crises (Laeven and Valencia 2013). The promise was that the new regulatory and supervisory regime would prevent this from ever happening again. Recent resolutions, such as that of Banco Espirito Santo in Portugal in August 2014, suggested this new framework was here to stay, even if with negative but limited short-term costs for the real economy (Beck et al. 2017a). Nevertheless, as pointed out by many economists at the enactment of the BRRD, it has been naïve to claim that taxpayer money would never be used again. While ex ante optimal to never bail out, there are ex post situations where a (partial) bailout can be optimal (Beck 2011, Dewatripont 2014). This time inconsistency is ultimately behind a supposedly fixed rule of bail-in, to avoid moral hazard. However, it also creates unnecessary expectations that will have to be disappointed eventually. And this inflexibility has also been reflected in a construct on the European level that – when push came to shove – proved to be very flexible, at least politically.

Clean up the mess first

Aside from the inflexibility of a ‘never bailout’ rule, it also does not make sense to apply this rule before cleaning up the previous crisis. One of the key lessons of decades of crisis resolution in developed and emerging markets alike is that an early resolution (which implies recognition and allocation of losses early on) is critical for a swift recovery. This has been a hallmark of the post-Lehman 2008 crisis resolution in the US, and has been missed in Japan but also large parts of the Eurozone. Economists have pointed to the slow recovery in both Japan and the Eurozone as being partly due to the slow resolution of undercapitalised banks in several periphery countries (Hoshi and Kashyap 2015). Weak banks had strong incentives to evergreen non-performing loans, delaying the day of reckoning, but also slowing down economic recovery as assets were tied up in under- or non-performing firms – a vicious cycle that can only be cut with politically decisive action. And while Spain eventually took such decisive action during the 'caja crisis', Italy continued to delay to resolution of its weak banks, mainly for political reasons.

In summary, it would have been better to resolve these banks many years ago. Putting in place the banking union and the BRRD rule of ‘no bailout’ without taking into account the existing structure of liabilities and existing bank fragilities was a mistake.

How should the widespread bank fragility in Italy be addressed? Rather than institution-by-institution, as currently done, a country-wide bad bank would have been a better solution. While this has been used successfully in several countries during the crisis, including in Ireland and Spain, the Italian attempt at establishing one was never big enough and was limited to a recapitalisation fund (Atlante) rather than a true asset management company, as in other countries. Ideally, there would have been such an AMC at the end of the Comprehensive Assessment in 2014 for the whole Eurozone, to exploit scale economies and overcome national biases, but political frictions avoided establishing such an institution (Beck and Trebesch 2013).

National solutions still trump Eurozone solutions

If there is one important lesson coming out of both the Spanish and the Italian cases it is that nation states still trump European interests. In both cases the resolution was effectively done at the national level, even though under supervision and ultimate approval from European institutions. The case of Banco Popular is a bit less clear-cut and can be justified with Santander having easier access to the necessary information. This does not hold in the case of Intesa and its takeover of the good assets of the two small regional lenders. The takeover of good assets by Intesa Sanpaolo with state guarantee would have been difficult, if not impossible, to undertake with a non-Italian player, especially from a political perspective.

The conclusion from both cases (in the Spanish case more as demonstration effect) has been that the resolution process is still purely national, both for small and mid-sized banks. This is the consequence of not having a Eurozone-wide deposit insurance available, while this national bias is also the cause for political reluctance to move to completing the banking union with a deposit insurance scheme. Deeper than regulatory constraints, however, is that there is still a political bias for national banking that will take a long time to overcome.

Where next?

The bailout of senior creditors of the Italian banks has strengthened German resistance to completing the banking union with a Eurozone-wide deposit insurance scheme. Critics of such a move can point to Italian taxpayer money used in Italy as a precedent for German taxpayer money being used in the future for similar situations, through a joint deposit insurance scheme. This criticism, however, misses two points:

  • It was the move from the European to national level that allowed this, which would not be possible under a completely European structure, i.e. where resolution authority is completely delegated to the SRM.
  • It was the lack of cleaning up before putting in place the banking union that is at fault here.

The consequence should not be a turning away from Eurozone-wide solutions but strengthening them further, both by forcing the remaining bank fragilities in Italy to be addressed and by completing the banking union.1


Many of the themes have been brought up again and again and point to the same underlying political problem. So, to summarise:

  • The case of Banco Popular has shown that SSM and SRM can work well – quick intervention, quick resolution, no taxpayer money; one-nil for the banking union.
  • The case of Italian banks has shown that it is naïve to think that taxpayer money will never again be used – there are situations where this is necessary. In this specific case, it should have been done much earlier. No winners here, only losers: the banks and their equity and junior bondholders, the Italian government, and the banking union.
  • The case of the Italian banks has also shown that a crisis should be cleaned up first before imposing a new regulatory regime; clean the field before starting the game.
  • Finally, both cases show that we are still far away from a Eurozone-wide financial banking system, which complicates both the completion of a Eurozone-wide safety net and a complete cut of the bank-sovereign deadly embrace. It will take more than a common regulatory regime and supervision and resolution framework to get there.
  • Italy is a problem and will continue to be so for some time (due to continuous banking problems, sovereign indebtedness, and political uncertainty). As the past weeks have shown, the future of the Eurozone might very well go through Italy.

Ultimately, this recent experience shows that the same political frictions that delayed the proper solution of the Eurozone crisis have prevented an optimal financial safety net for the Eurozone. But these same political frictions are back at work. As much confidence as many observers have in the SSM and SRM and the new regulatory framework, it is these political frictions that we should be afraid of. Ultimately, it takes political leadership to overcome them and get us to more, not less, Europe.

Author's note: The authors would like to thank Andre Silva for comments and suggestions. 


Beck, T, S Da-Rocha-Lopes and A Silva (2017a), “Sharing the Pain? Credit Supply and Real Effects of Bank Bail-ins”, CEPR Discussion Paper No. 12058.

Beck, T, E Carletti and I Goldstein (2017b), “Financial Regulation in Europe: Foundations and Challenges”, in L Matyas and R Blundell et al. (eds), Economics without Borders: Economic Research for European Policy Challenges, Cambridge University Press, pp. 470-510.

Beck, T and C Trebesch (2013), “A Eurozone Bank Restructuring Agency”,, 18 November.

Beck, T (2011), “Bank resolution: a conceptual framework. Financial Regulation at the Crossroads: Implications for Supervision”, Institutional Design and Trade 12: 53.

Carmassi, J, S Dobkowitz, J Evrard, A Silva and M Wedow (2017), “Exposure of the European Deposit Insurance Scheme (EDIS) to Bank Failures and the Benefits of Risk-based Contributions”, ECB Macroprudential Bulletin 3: 21-33.

Dewatripont, M (2014), “European banking: Bailout, bail-in and state aid control”, International Journal of Industrial Organization 34: 37–43.

Draghi, M (2017), Letter to Mr Dimitrios Papadimoulis, Member of the European Parliament.

Hoshi, T and A Kashyap (2015), “Will the U.S. and Europe Avoid a Lost Decade? Lessons from Japan’s Postcrisis Experience”, IMF Economic Review 63: 110-63.

Laeven, L and F Valencia (2013), “Systemic banking crises database”, IMF Economic Review 61(2): 225-270.

Schoenmaker, D and G Wolff (2014), “Options for European deposit insurance”,, 30 October.


[1] Draghi (2017) recently reinforced that the ECB considers the European Deposit Insurance Scheme (EDIS) to be the necessary third pillar of the banking union that should be put in place as soon as possible. In response to the critics of common deposit protection, Carmassi et al. (2017) also show in the recent ECB Macroprudential Bulletin that a fully-funded Deposit Insurance Fund would be sufficient to cover pay-outs in a non-systemic banking crisis, and that there would be no unwarranted systematic cross-subsidization within the EDIS.

1,890 Reads