During the crisis, individual institutions such as Hypo Real Estate required public assistance of €100 billion or more.1 So how can a European Resolution Fund of only €55 billion possibly suffice for all banks in the Eurozone?
It could, provided the Fund is part of a well-designed architecture for regulation, supervision, and resolution, that makes banks not only less likely to fail but also safe to fail – meaning that they can be resolved without cost to the taxpayer and without significant disruption to financial markets or the economy at large.
This architecture rests on four pillars:
- Prudential regulation and supervision,
- ‘No forbearance’,
- Adequate ‘reserve capital’, and
- Provision of liquidity to the bank-in-resolution.
In concept, these four pillars can be found in EU legislation: the Capital Requirements Directive/Regulations (CRD IV), the Banking Recovery and Resolution Directive (BRRD), and the Single Resolution Mechanism (SRM). But much remains to be done. The EU Bank Regulator has to turn BRRD concepts into ‘blueprints’ (binding technical standards), the ECB has to establish a rigorous supervisory approach, and the EU has to finalise arrangements for the SRM. This column reflects on how they can do so in a manner that ensures that the Resolution Fund is large enough.
Regulation and supervision
CRD IV will require EU banks to hold higher capital and to bolster their liquidity. Via the Single Supervision Mechanism, the ECB will provide common and rigorous oversight of Eurozone banks.
Macroprudential supervision should reinforce these microprudential measures. Through the European Systemic Risk Board and similar bodies at the national level (e.g. the UK’s Financial Policy Committee) and at the global level (the Financial Stability Board), authorities are taking steps to identify and mitigate systemic risks. Together, these measures should reduce the likelihood that banks will fail, and reduce possible claims on the Resolution Fund.
Supervisors should push banks into resolution as soon as the bank reaches the point of non-viability. At this point the resolution authority assumes control over the failed institution. To maintain continuity of the failed bank’s critical economic functions, this handover should proceed rapidly and smoothly.
In finalising the Single Resolution Mechanism, policymakers would do well to ensure that the ECB as the single supervisor can initiate resolution promptly – as soon as the bank reaches the point of non-viability. Without such assurance, private fund providers can flee. That raises the cost to investors who remain, as well as to the lender of last resort and deposit guarantee schemes. It also potentially increases the claims on the Resolution Fund.
Adequate ‘reserve capital’
For resolution to work, the bank-in-resolution needs to be recapitalised immediately. This capital should come from investors, not taxpayers. This will be the case if the bank has sufficient ‘reserve capital’ (capital in addition to common equity that is immediately available to absorb losses) outstanding at the point at which the bank enters resolution.
The Banking Recovery and Resolution Directive lays the foundation for this. It requires that Additional Tier 1 and Tier 2 capital be subject to conversion or write-down at the point of non-viability. It also mandates the EU bank regulator to write a binding technical standard to establish a minimum requirement for eligible liabilities (MREL) for bail-in. Finally, the BRRD establishes a creditor hierarchy with certain ‘carve-outs’. This recognises the claim that secured creditors have to collateral pledged by the bank, grants insured deposits preference, and allocates first loss capital instruments (common equity first, then Additional Tier 1, and then Tier 2.)
Table 1. BRRD establishes a ‘waterfall’ for losses with ‘carve-outs’: Allocation of €100 loss under different assumptions regarding the position of senior debt
*Numbers are rounded
Granting insured deposits preference greatly reduces the likelihood that losses would be attributable under bail-in to such deposits, particularly if the authorities follow a no-forbearance policy. Accordingly, there is only a limited risk that resolution will result in a claim on the deposit guarantee fund – a factor that should be reflected in any risk-based premiums levied on banks.
The problem with the BRRD’s approach lies in the middle layer. This includes unsecured senior debt and various other liabilities, some of which are explicitly exempt from bail-in (carve-outs) – such as obligations to payment systems – and others which the resolution authority may decide at the point of resolution to exclude from bail-in based on financial stability grounds. Such an arrangement could result in liabilities not exempt from bail-in incurring a loss greater than they would have experienced under liquidation (see Table 1), giving rise to a potential claim under the ‘no creditor worse off’ condition. However, the BRRD leaves open how Member States would fulfil such a commitment, aside from setting limits of the extent to which the resolution fund might be used (up to 5% of total liabilities). As a result, liabilities within the middle layer are likely to gravitate toward categories exempt from bail-in, such as repos. As senior debt cannot migrate to the exempt category, it will become de facto subordinated to the other elements in the middle layer.
Accordingly, when it writes the technical standard for MREL, the EU Bank Regulator may wish to put senior unsecured debt explicitly into a mezzanine category (senior to capital instruments, but junior to customer liabilities). Alternatively, for banks with little access to wholesale debt markets, counting equity towards MREL may be particularly valuable. Structuring MREL in this fashion would significantly reduce potential claims on the resolution fund under the ‘no creditor worse off’ condition. More importantly, it would help ensure that bail-in does not reach customer liabilities. This will reduce the risk of such liabilities, promote continuity of critical economic functions, and avoid disruption to financial markets and the economy at large.
Providing liquidity to the bank-in-resolution
For resolution to succeed in ensuring continuity of critical economic functions, the bank-in-resolution must have access to adequate liquidity – especially in the first days and weeks following its entry into resolution and its over-the-weekend recapitalisation via bail-in.
The Resolution Fund can help the bank-in-resolution obtain liquidity. It can provide what amounts to a back-up guarantee to a liquidity provider against losses the provider might suffer if the bank-in-resolution failed to repay the facility, and the proceeds realised from the liquidation of any collateral pledged by the borrower were insufficient to repay the facility.2 The Resolution Fund credit enhancement would considerably facilitate the provision of liquidity to the bank in resolution, possibly opening the door to private entities providing such facilities.
The Resolution Fund should not provide solvency support to banks. Investors in the failed bank should bear such losses, not taxpayers. Broadly speaking, the BRRD adheres to this principle. However, the BRRD allows Member States to use the resolution fund where the losses cannot be passed to other creditors, provided (i) investors have incurred (via bail-in) losses amounting to 8% of the bank’s liabilities; and (ii) the funding provided by the Resolution Fund is limited to 5% of total liabilities including own funds.
Financing the resolution fund
The BRRD envisages ex ante funding to build Member States’ resolution funds up to the target level. To the extent that Member States have bank levies, these should be earmarked for the resolution fund and put in abeyance once the fund reaches its target level. To the extent that a bank is active in more than one Member State, contributions to the respective resolution funds should be coordinated (including the possibility of borrowing from one another), and there should be no duplication of levies on the same banking group. Levies should also be risk-based, with discounts available to those firms that have a high level of MREL relative to assets and/or risk-weighted assets.
In sum, resolution funds are a good example of how a lot can be done with relatively little. Indeed, by placing a lid on the Resolution Fund, the BRRD and SRM may in fact be reinforcing the need to develop a blueprint that will ensure that investors, not taxpayers, bear the cost of bank failures.
Disclaimer: The views expressed in this column are the authors’ and do not necessarily represent those of the Bank of Spain, the Eurosystem, or Ernst & Young. Any errors are the authors’.
Kane, E (2004), “Designing safety nets to fit country circumstances”, World Bank Research Working Paper 2453.
Eisenbeis, R and George Kaufman (2006), “Cross-border banking: Challenges for deposit insurance and financial stability in the European Union”, Federal Reserve Bank of Atlanta Working Paper 2006-15.
1 Total government assistance provided to Hypo Real Estate amounted to €134 billion, consisting of €9.8 billion in capital infusions and €124 billion in guarantees (each expressed in terms of maximum amount outstanding). See http://www.fmsa.de/export/sites/standard/downloads/20131231_Overview_Measures.pdf.
2 The liquidity provider will demand that the bank-in-resolution pledges any unencumbered asset the bank might have.