A happier union for Europe's banks

The EU's proposed bank resolution framework is a step in the right direction but does not go far enough – a banking union needs more backbone, including a single resolution authority.


In June the EU proposed two significant measures: banking union and a framework for bank resolution. The two should go hand in hand. But how?

A resolution regime is one of the cornerstones of any banking union – it determines what happens when banks go bankrupt. Equally necessary are prudential regulation and supervision (to reduce the probability of problems), and deposit guarantee scheme (to reduce the probability of bank runs causing bankruptcies), and liquidity assistance from the central bank (to reduce the chance solvent banks go under due to illiquidity).

The resolution and supervision lay at the heart of the trade-off made by the EU heads of state at the end of June; the European Stability Mechanism could support directly the recapitalisation of failed banks, when banks were henceforth subject to a 'single supervisory system'.

An effective bank resolution framework should have six components:

– A clearly defined closure rule;
– A resolution authority empowered to take prompt decisions about the bank in resolution;
– A set of effective resolution tools;
– Respect for the priority of claims;
– Arrangements to provide the failed bank with immediate liquidity; and
– A financing mechanism for bank resolution that limits moral hazard and contagion.

Although the proposed directive has each of these components, it is a minimum harmonisation proposal, not a proposal for the resolution regime of a banking union.

Take first the proposed directive’s closure rule. It empowers the bank’s supervisor to intervene when the bank has reached the point of non-viability and no supervisory action can prevent the bank’s failure. But who is the empowered bank supervisor, the national or the single European supervisor? For member states that join the banking union, it should be the single supervisor rather than national authorities – a step that would enhance objectivity with respect to the bank’s true condition limiting forbearance and negative spillovers between countries.

Second, the resolution authority must be empowered to take rapid decisions regarding the bank in resolution. This explains the public character of resolution authorities. The resolution authority's decisions should be subject to judicial review, but the court should not be able to delay resolution and any remedies that the plaintiff may win should be limited to monetary compensation. The proposed directive adheres to these precepts.

However, it is silent on who should be the resolution authority: the supervisor or a separate resolution authority? The later would limit potential forbearance of supervisors. Currently, the proposed directive envisages a decentralised resolution framework. For member states in the banking union there should be a single resolution authority, which would be better able to internalise more fully potential externalities in the decision to resolve a bank.

Third, the resolution authority must have adequate tools to resolve a bank. The directive harmonises, for the first time, those tools, which include the right to: sell the bank to a third party, transfer the bank’s deposits (along with matching assets) to a third party or to a newly created bridge bank, and to bail in (write down or convert to equity) some or all of the bank’s liabilities as well as to place the bank into liquidation. However, the scope of the directive is limited to credit institutions. These tools should also be applicable to the credit institution’s parent holding company and to its non-bank affiliates in the EU.

Fourth, the resolution regime should enable creditors to know ex ante how they would fare under resolution. Ideally, the resolution regime would respect the absolute priority of claims. At a minimum, creditors should be given the assurance that they will be no worse off than they would have fared under liquidation (the proposed directive does this). Shareholders should suffer first loss, and providers of non-core Tier 1 capital and Tier 2 capital should be subject to conversion or writedown prior to any losses being imposed on senior creditors. Finally, creditors in the same class should receive equal treatment. Here, the proposed directive departs from the priority because it proposes to carve out from bail-in senior debt with remaining maturity of less than one month. This would induce investors to shorten the maturity of their funding to banks (in contrast to liquidity regulation which seeks to induce banks to lengthen the maturity of their liabilities).

Far more problematic, however, is how the proposed directive treats uninsured deposits under bail in. Insured deposits would not suffer any loss – that would go to the deposit guarantee scheme. But uninsured depositors could be bailed in and suffer loss of principal as well as loss of access to their funds. That could disrupt payments and even the economy at large. These problems could be largely avoided, if the directive were amended to give deposits preference. Depositor preference also lowers the risk to the deposit guarantee scheme – a feature that could help lay the groundwork for a deposit guarantee scheme that would cover the entire banking union. Ideally, the directive should reinforce deposit preference by setting a minimum for the amount of liabilities subject to immediate bail in (non-core Tier 1 capital, Tier 2 capital and unsecured senior debt).

Fifth, a bank’s orderly resolution (as opposed to liquidation) will only be successful, if the bank in resolution has access to liquidity. Such liquidity support will enable the bank in resolution to avoid 'fire sales' of its assets. Moreover, it should be on a super-seniority status secured by the bank’s unencumbered assets including without limitation its investments in subsidiaries and affiliates. This is envisaged in the proposed directive. For such support to be sound, limits may be needed on the degree to which banks may encumber their assets prior to resolution. As to the source of such liquidity, the central bank is certainly a logical candidate.

A resolution fund is the final component of the resolution regime. Its scope should be limited to compensating the resolution authority for any losses that it incurs in resolving failed banks. The resolution regime determines the possible losses that the resolution authority could face and the possible claims on the resolution fund, whose main objectives, as established in the directive, are (a) to guarantee the assets or liabilities of the bank under resolution; (b) to fund the takeover of banks impaired assets, and (c) provide finance to the bank under resolution/bridge bank. Such claims would be relatively small, if (1) supervisors refrain from forbearance and put a troubled bank promptly into resolution (as they are more likely to do in a banking union under a single supervisor); (2) there is a requirement for the bank to have outstanding a minimum amount of liabilities subject to bail-in and (3) the bank in resolution has access to liquidity.

Consistent with the objective of limiting public costs of resolution, the directive proposes that resolution funds will be funded over time by financial entities subject to the resolution regime via levies to a level equal to 1 per cent of their insured deposits. Considering that the decision of resolving a bank as opposed to its liquidation is highly dependent on its systemic importance, there is a rationale for bank levies to be based on such criterion together with an ex ante assessment of the banks resolvability. Both are envisaged in the proposed directive. But the proposed directive stops short of establishing a single resolution fund. That will be required, at least for the banking union.


The proposed Crisis Management Directive goes a long way toward putting resolution on a sounder footing. It creates the basis for investors, not taxpayers, to bear the cost of bank failure. But these steps are not enough for banking union. The banking union should have a single resolution authority and single resolution fund along with a single supervisor. And, banks in the banking union should have depositor preference as well as a requirement that they issue a minimum amount of liabilities subject to bail in. That will assure that banks in the banking union are ‘safe to fail’.

The views expressed in this column are the authors’ and do not necessarily represent those of the Bank of Spain or Ernst & Young. Originally published on www.VoxEU.org. Reproduced with permission.

Related Articles