How to design the European banking union?

Marius Zoican 08 January 2015



The Global Financial Crisis ignited the debate around a common European bank regulator – a banking union for the Eurozone. Ferry and Wolff (2012) look at the fiscal implications of a joint European regulator and advocate a common fiscal backstop mechanism. Colliard (2014) studies the optimal architecture of the single supervision mechanism (SSM) – the supervisory pillar of the banking union – and argues there is a conflict of objectives between local and joint supervisors.

In Zoican and Górnicka (2014), we shift the debate to the single resolution mechanism (SRM), the second important pillar of the banking union. Under the resolution mechanism, the common regulator rather than the national authorities decides whether to bail out an insolvent bank. We argue that the mechanism generates tension between increased regulatory efficiency in responding to bank defaults, on the one hand, and weaker commitment to liquidate failed systemic institutions, on the other hand. If the latter effect dominates, internationally-oriented banks take on more risk, augmenting existing moral hazard problems. If banks hold opaque assets, such as structured derivatives, the negative impact of the single resolution mechanism is particularly large. A hybrid system in which the resolution mechanism coexists with the current national resolution mechanisms is arguably a superior arrangement, as it can both prevent insolvency contagion and keep risk taking incentives low.

A systemically safer Eurozone

Centralised bank default resolution has undeniable benefits. The IMF argues the joint regulator reduces insolvency contagion through an intervention policy that minimises costs for all member countries (IMF, 2013). If a multinational bank defaults, a centralised decision bypasses protracted negotiations between national authorities (see the Dexia, Fortis bailouts). Most importantly, the banking union is in the unique position to limit the spread of losses across European banks.

Cross-exposures of Eurozone banks are highly asymmetric; Banks in Germany, France, and the Netherlands have net loans of almost one trillion dollars to counterparts in Greece, Italy, Ireland, Portugal, or Spain.1 Cross-border links between banks create the scope for default contagion, as noted by Freixas et al. (2000). National authorities face a natural tension between domestic taxpayers and foreign creditors. If a highly indebted bank becomes insolvent, taxpayer money is needed to bail it out. A large part of the funds, however, is claimed by foreign institutions. Domestic authorities are understandably biased towards domestic taxpayers (see Allen et al. 2011). They are more likely to impose losses on international creditors through the liquidation of a heavily indebted bank. On the other hand, a banking union represents the interests of both ‘domestic’ and ‘foreign’ actors. To avoid contagion, it is more likely to be lenient with systemic international banks. Systemic contagion risks are contained and welfare improves.

‘Contagion insurance’ premium for the Eurozone core

Since banks in larger countries have more consistent international exposures, they benefit most from lower contagion risk. A banking union offers them a form of systemic risk insurance. Consequently, the contributions of the Eurozone core countries to the common resolution fund should include a premium. Banks in Germany, France, or the Netherlands will be able to invest in growing European markets while bearing a smaller contagion risk than under the existing mechanism.

‘Too-European-to-fail’ problem

While a lenient banking union minimises contagion, it can also engender risky behaviour at the bank level. Internationally indebted banks might become ‘too-European-to-fail’, i.e., their systemic status is reemphasised. A lesson from the past Crisis is that systemic banks are able to take on increasing amounts of risk while relying on implicit regulatory support. Is too-European-to-fail a major concern? I argue so! Modern banks hold large positions in opaque and complex assets, such as structured products and derivatives. For these asset classes, a marginal drop in risk management standards can have a very large negative impact on the returns and solvability of the bank.

Alternative design solution

A hybrid regulatory architecture solves the bank’s incentive problem. The banking union and the national regulators coexist. The banking union’s mandate, however, is limited to European-wide crises, e.g., triggered by an external shock such as the house market crash in the US. National crises are dealt with by national authorities who bear the full costs of insolvent bank interventions. Contagion risks are contained when it matters most, but internationally systemic banks sometimes face a tougher stance from national regulators. It can be the best of both worlds.


Allen, F, E Carletti, and A Gimber (2011), “The financial implications of a banking unit,” Banking Union for Europe: Risks and Challenges, ed. T Beck, (London, Centre for Economic Policy Research).

Colliard, J E (2014), “Monitoring the supervisors: Optimal regulatory architecture in a banking union”, Manuscript, European Central Bank

Ferry, J P, and G B Wolff (2012), “The fiscal implications of a banking union”, Bruegel Policy Brief 2012/02

Freixas, X, B M Parigi, and J-C Rochet (2000), “Systemic risk, interbank relations, and liquidity provision by the central bank”, Journal of Money, Credit and Banking 32, 611--638

IMF (2013), “A banking union for the euro area”, Staff Discussion Notes No. 13/2013.

Zoican, M A and Górnicka, L (2014), “Banking union optimal design under moral hazard”, Tinbergen Institute Discussion Paper 13-184/VI.


1 Source: Bank for International Settlements website and own calculations. See Figure 1.

Figure 1. Net and gross average international balances of banks from selected countries against GIIPS (Greece, Ireland, Italy, Portugal, and Spain) countries, between 2008:Q1 and 2013:Q1. The size of the marker is proportional to the total position.

Source: Bank for International Settlements, own calculations.



Assistant Professor in Financial Economics, Université Paris-Dauphine

Vox Talks