Chapter 9. A Banking Union for the Euro Area

Petya Koeva Brooks, and Mahmood Pradhan
Published Date:
October 2015
  • ShareShare
Show Summary Details
Rishi Goyal, Petya Koeva Brooks, Mahmood Pradhan, Thierry Tressel, Giovanni Dell’Ariccia, Ross Leckow and Ceyla pazarbasioglu 

A banking union—a single supervisory and regulatory framework, resolution mechanism, and safety net—for the euro area is the logical conclusion of the idea that integrated banking systems require integrated prudential oversight. The case for a banking union for the euro area is both immediate and longer term. Moving responsibility for potential financial support and bank supervision to a shared level can reduce fragmentation of financial markets, stem deposit flight, and weaken the vicious circle of rising sovereign and bank borrowing costs. In steady state, a single framework should bring a uniformly high standard of confidence and oversight, reduce national distortions, and mitigate the buildup of concentrated risk that compromises systemic stability.


The fiscal and financial crisis in the euro area has exposed critical gaps in the architecture of stability in the region. In the years preceding the crisis, large capital flows within the euro area fueled the buildup of sovereign and private sector imbalances. The subsequent deterioration of balance sheets and reversal of flows has forced very sharp economic contractions and financial market fragmentation (Figure 9.1). Borrowing costs of sovereigns and national private sectors have diverged widely and persistently, cuts in monetary policy rates have had limited or no effect in several economies, and adverse sovereign-bank-real-economy dynamics have been prevalent across the region (Figures 9.2 and 9.3). The monetary union, in short, is malfunctioning.

Figure 9.1Euro Area: Financial Market Integration and Fragmentation

(BIS cross-border bank claims; percent of reporting country’s GDP)

Sources: Bank for International Settlements (BIS); and IMF staff calculations.

Note: Data labels in the figure use International Organization for Standardization (ISO) country codes.

Figure 9.2Euro Area: Diverging Funding Costs

Source: European Central Bank.

Note: ECB = European Central Bank; NFC = nonfinancial corporations.

Figure 9.3Euro Area: Sovereign-Bank Loops

(Credit default swap spreads; basis points weighted by gross debt)

Source: Bloomberg L.P.

Note: OMT = Outright Monetary Transactions.

Important measures—for near-term crisis management and longer-term architecture—have been undertaken. Adjustment programs are being implemented and progress is being made to unwind fiscal and external imbalances that developed over several years. Regional firewalls—the European Financial Stability Facility and the European Stability Mechanism (ESM)—have been created and strengthened to smooth adjustment. The framework for fiscal and economic governance has been enhanced through the “Six-Pack” and the Treaty on Stability, Coordination, and Governance. The European Central Bank (ECB) has provided substantial liquidity to banks, stepped in to address market strains through government bond purchases, and announced its framework for Outright Monetary Transactions. The Eurosystem has recycled part of the capital flight from stressed economies to the core through the TARGET2 payments balances.

In this comprehensive policy response, the role of a banking union for the euro area is twofold. As part of crisis management, it can reduce fragmentation of European banking markets. Direct bank recapitalization by the ESM can help restore the health of bank balance sheets and remove tail risks and potential contingent liabilities affecting sovereigns under stress. A precondition for direct recapitalization of banks by the ESM is the creation of an effective single supervisory mechanism (SSM), called for by euro area leaders in June 2012. In steady state, an integrated architecture for financial stability in the euro area would bring a uniformly high standard of enforcement, remove national distortions, and mitigate the buildup of risk concentrations that compromise systemic stability. Moving responsibility for potential financial support—and the associated banking supervision—to a shared level would reduce financial fragmentation and weaken the vicious circle experienced by many countries of rising sovereign and bank borrowing costs.

The European Commission presented a plan on September 12, 2012, on the elements of a new SSM that could begin operating in 2013. It called for adoption by end-2012 of European Union (EU) legislative proposals establishing a harmonized regulatory setup, harmonized national resolution regimes for credit institutions, and standards across national deposit insurance programs. On December 13–14, 2012, the European Council agreed that the SSM would come into operation in March 2014 or one year after the SSM legislation enters into force, whichever is later. Once the SSM legislation is adopted, ESM direct recapitalization could occur, with the ECB supervising the bank in need of assistance. The Council noted that adoption of a harmonized regulatory setup (the Capital Requirements Regulation/Capital Requirements Directive IV, or CRR/CRD IV) is “of the utmost priority,” and called for the adoption of the draft directive for bank recovery and resolution and for harmonization of deposit guarantee programs by June 2013. It affirmed that a single resolution mechanism with adequate powers and tools is required. This mechanism will be based on financial sector contributions and backstop arrangements that recoup taxpayer support over the medium term. The Commission made a proposal for such a mechanism in 2013, and a final agreement was reached at the end of 2014.

Although agreement has been reached on the SSM, different views on the modalities of the other elements of a banking union remain. These differences partly reflect concerns about the potential mutualization of liabilities and asymmetric cost sharing across members, as well as on the desirability of separating near-term crisis resolution, such as needed fiscal solutions and backstops for bank recapitalization, from longer-term architecture issues. They also reflect the complexities and difficulties of setting up a banking union in a relatively compressed period.

This chapter elaborates the case for, and the design of, a banking union for the euro area. This chapter discusses the benefits and costs of a banking union, presents a steady state view of the banking union, elaborates transition issues, and briefly discusses broader EU issues. The background technical notes analyze in depth the various elements of the banking union: a single supervisory framework, a single resolution mechanism and common safety net, and urgent issues related to ESM direct recapitalization of banks.

How Would a Banking Union Help?

Although finance spans the euro area in a dense network of cross-border banks and obligations, the current architecture of stability is based largely on national supervision, national resolution, and national safety nets. This architecture—outlined in Box 9.1—has several implications, including the following:

  • Bank-sovereign-real-economy links—Absent the ability to control local interest rate conditions, the existing architecture strengthens the link between a country’s banking and real sectors and the health of its public finances. In good times, banks may grow to overwhelm national supervisory capacity. In bad times, they may overwhelm national fiscal resources. Similarly, if a sovereign’s finances are sound, its backstop for its banks is credible. But if they are weak, then its banks are more vulnerable and will face higher funding costs. As a result, private borrowing costs rise with the sovereign’s, imparting procyclicality (costs rise as conditions deteriorate and capital flows out), impairing the transmission of monetary policy (because rate cuts have limited or no effect), and amplifying fragmentation of financial markets and volatility.
  • Skewed incentives—National authorities may unduly favor the national banking system and economy, regardless of outward spillovers, which lie beyond their mandates. In good times, they may not be stringent or capable enough to limit the buildup of excesses. In bad times, they may encourage the reduction of their banks’ cross-border activities, exacerbating financial fragmentation. Delays in resolving stresses would only increase the eventual cost. And because a bank’s distress may have adverse cross-border externalities, other countries may have no choice but to support those economies whose banking systems run into trouble.

Box 9.1.European Union: Existing Framework for Financial Stability

Regulation and supervision

Regulatory framework. Most financial rules in Europe originate from European Union (EU) directives and regulations. A minimum standard has been set in harmonized legislation. In May 2012, the European Council approved the “Danish compromise” legislative package—the Capital Requirements Regulation (CRR) and Capital Requirement Directive (CRD IV)—which transposed Basel III in the EU and harmonized prudential rules that would apply to all banks in the EU. National options and discretion would be the exception, although flexibility is envisaged for financial stability risks that differ across jurisdictions and institutions. National authorities may impose “systemic risk buffers” without the European Commission’s preapproval up to a limit, and above the limit with prior approval. They may temporarily impose stricter requirements, for example, in relation to risk weights for certain sectors, large exposure limits, and liquidity requirements. They also retain the flexibility to impose stricter requirements on individual institutions through Pillar 2 reviews.

National supervision. Banking supervision in the EU is the prerogative of national authorities. Although EU directives have set minimum internationally agreed-on standards, supervisory handbooks and approaches vary across member states. The CRR/CRD IV aims to strengthen elements of supervision (such as supervisory planning, on-site inspections, and more robust and intrusive supervisory assessments) and harmonize sanctions.

Cross-border cooperation. Given the high degree of financial integration in the EU, home-host cooperation is essential for effective supervision. Since 2010, guidance has been provided on the supervision of cross-border banks, and supervisory colleges provide a forum for discussions, but it is not clear that these are altogether effective. The EU directive on bank resolution and recovery clarifies nonbinding home-host relations and responsibilities in colleges, in particular in relation to the provision of intragroup liquidity provision.


National regimes. Many EU countries have relied on general corporate insolvency proceedings to deal with bank failures, an approach that has resulted in complex and lengthy windups or, more commonly, nationalization, with significant costs for the economy. Although some national frameworks have been strengthened through special resolution powers to facilitate the quick resolution of failing banks, they remain largely untested and may not, in all cases, be fully in line with international best practice (for example, the Financial Stability Board Key Attributes).

Harmonization. The Commission’s draft directive on bank recovery and resolution seeks to ensure that national authorities have strong preventive powers, including in relation to recovery planning; early intervention powers (for example, power to impose capital-raising and conservation measures, restrictions on activities, and implementation of recovery plans); and resolution tools (for example, the possibility to set up bridge banks, perform asset separations, override shareholders’ rights, replace management, divest nonessential businesses, or write down or “bail in” debt).

Member states are also required to set up resolution funds, built over 10 years and prefunded to a target ratio of 1 percent of total liabilities excluding own funds. The draft directive proposes after-the-fact levies; access to a credit line from the central bank; and borrowing arrangements among funds across countries, subject to safeguards.

Cross-border resolution. The draft directive also offers principles for early intervention and resolution of cross-border banks, such as liquidity provision within groups and the creation of resolution colleges to develop nonbinding, crisis-planning mechanisms. The absence of binding, burden-sharing agreements put in place in advance of any crisis, however, leaves unresolved the issue of coordination. It is not clear, therefore, that least-cost resolution would be achieved quickly and effectively. Nevertheless, implementation of the directive would set the stage for a more integrated EU resolution framework.

Safety nets

Lender of last resort (LOLR). Banks can place deposits and refinance eligible assets with the Eurosystem and, if collateral constraints bind, resort to emergency liquidity assistance from their national central banks. Although the ECB’s Governing Council has authority to ensure that LOLR activities by national central banks do not interfere with common monetary policy, losses arising on emergency liquidity assistance remain the responsibility of the national central banks, exacerbating sovereign-bank linkages.

Deposit insurance. Existing deposit insurance programs are national, with varying coverage limits, contributions, and fund sizes. Most programs are underfunded. The EU Directive on Deposit Guarantee Schemes has set minimum standards on coverage (€100,000 per depositor per bank) and the payout period. The Commission has proposed harmonizing national programs (for example, introducing common standards on financing and setting a target fund size of 1.5 percent of eligible deposits) and clarifying responsibilities (for example, improving insurance payments for cross-border banks), with the possibility of borrowing arrangements across national programs and with adequate safeguards.

The European System of Financial Supervisors (ESFS)

ESFS. Following the 2009 de Larosiere Report recommendations, elements of an EU supervisory structure were established in 2011. Although regulation remains the prerogative of the Commission, technical standards on sectoral microprudential regulation are tasked to the European Supervisory Authorities—the European Banking Authority (EBA), the European Securities and Markets Authority, and the European Insurance and Occupational Pensions Authority—and macroprudential oversight to the European Systemic Risk Board (ESRB). As EU agencies, these agencies have limited powers (including because of fiscal safeguards) and resources, with ultimate decisions remaining at the national level.

EBA. The EBA is a cooperative body for EU bank supervisors. It is tasked with issuing technical standards in regulatory and supervisory areas (subject to fiscal safeguards). It can organize and conduct peer reviews of competent authorities, including issuing recommendations and identifying best practices, to strengthen consistency in supervisory outcomes, promote supervisory convergence, address breaches of EU law, limit scope for regulatory arbitrage, foster a level playing field, and support consumer protection. It coordinates and ensures consistency of EU-wide stress tests.

ESRB. The ESRB’s role includes establishing macroprudential frameworks and ensuring effective coordination and internalization of cross-border spillovers. Its main instrument is the issuance of nonbinding risk warnings and recommendations through a “comply or explain” mechanism.

A single regulatory and supervisory framework would help contain systemic risks and curb the moral hazard that accompany common backstops and safety nets; a single resolution mechanism with adequate backstops would isolate and address pockets of weakness; and a common safety net would help prevent retail deposit runs that could overwhelm the capacity of any one economy:

  • Regulation and supervision—Regulation involves rules to prescribe what banks must or may not do, while supervision verifies and enforces such rules and adds broad discretionary powers to control undue risk-taking and ensure adequate capitalization. Both seek to complement market discipline imposed by bank creditors and shareholders. The SSM would facilitate a systemic approach to supervision of risk management across all countries and help identify and prevent the buildup of excessive risk concentrations. It would enforce regulations consistently across the banking union, reduce national distortions, and better address cross-border issues and fragmentation. It could be less subject to capture by local interests (if not to broader ones); for example, Agarwal and others (2012) show that, in the United States, federal regulators are significantly less lenient than state regulators (although the United States also has federal backstops in place).
  • Resolution and safety nets—An effective resolution mechanism would facilitate timely intervention to address weak banks and prevent contagion across the system. A single resolution authority would support market discipline and should minimize the costs of failing individual banks, although cost minimization needs to be considered at the system level for systemically important institutions. Together with a common safety net that comprises deposit insurance (to provide certainty to retail depositors) and a lender of last resort (LOLR) (for emergency liquidity), it would enhance the capacity to cope with shocks that may overwhelm any individual economy. A credible single resolution framework and safety net would address coordination and burden-sharing problems related to cross-border failures and internalize associated externalities. By moving responsibility for potential financial support from the national to the supranational level, they would decouple banks’ prospects from those of sovereigns with weak finances, and protect individual sovereigns from banking sector weaknesses. They would also limit the potential burden on taxpayers, including by “bailing in” creditors as necessary.

In these ways, a banking union would narrow gaps in the design of the monetary union, enhance confidence, and strengthen the basis for financial stability, sound credit, and sustainable growth.

Would a banking union have prevented this crisis? Arguably, it would not have halted the sovereign debt crisis in some countries. But a well-functioning banking union could have substantially weakened, if not broken, the adverse sovereign-bank-growth spirals, maintained depositor confidence, and attenuated the liquidity and funding freezes that followed. The ECB’s rate cuts would more likely have fed through to lower borrowing costs for the private sector. A strong banking union would also have limited the concentrated exposures of banks to certain risks. For example, euro area—wide supervisors would arguably not have allowed size, structure, and concentration risks to grow as they did in countries such as Cyprus, Ireland, and Spain, or general banking weaknesses to have accumulated in some other places. However, as experience in the United States and other economies suggests, supervision would have had to strive toward a high standard. Merely reorganizing supervisory structures would not have addressed the buildup of systemic risk or the too-big-to-fail problem.

Differences of views on cost sharing for resolution and deposit insurance and the legacy problems of the crisis could delay progress. But the cost of delaying resolution is likely to be far higher for everyone. Some member states remain exposed to potentially large costs, including through claims accumulated via Eurosystem exposures (Figure 9.4). To restore stability, agreement is needed on burden sharing for losses that occur in the future. These costs could be made more manageable by raising resources from the financial sector.

Figure 9.4Euro Area: Central Bank’s TARGET2 Balances

Source: Central banks of the Eurosystem.

  • Resolution funds—Resolution and deposit insurance funds are meant to cover individual bank failures, not systemic crises. As such, prefunded programs (in the steady state) could be fairly small. Funds could be raised over time from the financial sector to reach the target size and should be risk based (see the section “What Should the Banking Union Look Like?”). Together with a least-cost resolution mechanism and common backstops for systemic crises, the funds should be sufficient to ensure that bank failures are dealt with in an orderly fashion.
  • Backstops—In systemic crises, resolution and deposit insurance funds would likely be inadequate. Therefore, the assurance of fiscal and monetary backstops—the availability of common public resources for resolving banks and restoring confidence in the financial sector—is required to halt disorderly dynamics, such as deposit flight. A credible joint backstop, with governance safeguards, could substantially weaken bank-sovereign links and prevent inefficient cross-border resolutions along national lines. What is the fiscal cost, defined as bank recapitalization by the sovereign and other types of support?
    • ° Gross versus net costs—It is important to distinguish gross from net fiscal costs. Net costs are invariably much smaller because recovery of asset values over time allows some of the cost of up-front support to be recouped. According to Laeven and Valencia (2012), the gross fiscal cost of a median systemic crisis is 7 percent of GDP. For the current crisis, gross costs have thus far ranged from about 4 percent of GDP in Spain to about 40 percent of GDP in Ireland. Many countries have resorted to recapitalization programs, as well as guarantees of existing or new bank liabilities (for example, blanket guarantees of all liabilities or of deposits, and guarantees of unsecured debt) or of the value of assets to forestall the possibility of a run, but these measures have reinforced sovereign-bank links. However, the net costs are expected to decline as economies stabilize.
    • ° Credibility—The existence of common backstops does not mean that they will necessarily be tapped. Ideally, credibility would be sufficient to deter bank runs and capital flight; common backstops can prevent self-fulfilling panics that might occur if a national program is not credible. Actual costs to taxpayers could be relatively small, particularly if the need for subsequent sovereign bailouts is reduced. That said, when solvency issues arise, resources would need to be disbursed.
  • On balance—In the steady state, the costs and benefits across member states would be symmetric (relative to the size of the respective banking systems). All would benefit from a better-functioning financial system, and insurance would be available for everyone (given that it is not clear where future problems may arise). In the interim, resolving the current crisis will require recapitalizing banks and ensuring credible common backstops. Absent progress toward a well-functioning resolution mechanism and credible backstops, sovereign-bank spirals would be prolonged, exacerbating deleveraging, worsening the recession, and causing negative spillovers. Sovereign bailouts or the failure of a systemically important institution early in the new supervisory regime would raise costs and adversely affect the credibility of the new regime.

What Should the Banking Union Look Like?

The challenge for European policymakers is to halt the crisis while ensuring that actions dovetail seamlessly into the future steady state. Progress is required on all elements, and the governance of the banking union must provide the right incentives and promote timely decision making, lest national interests prevail and effectiveness is compromised.

  • A single supervisory mechanism (SSM) without a common resolution and safety net framework will do little to break the vicious circle between banks and sovereigns and stabilize the euro area. In particular, lack of a credible resolution framework would hamper the effectiveness of the SSM, and impede timely decision making by leaving national authorities to deal with the fiscal consequences of others’ supervisory decisions.
  • Bank recapitalization as well as resolution and deposit insurance mechanisms would lack credibility without the assurance of fiscal backstops and burden-sharing arrangements.
  • Conversely, common safety nets and backstops without effective supervision and resolution would break sovereign-bank links, but risk distorting incentives, reinforcing tendencies for regulatory forbearance, and shifting losses to the euro area level. Effective control must accompany, or precede, risk or burden sharing.

A banking union is necessary for the euro area, given the financial fragmentation, stresses, and deposit flight from one part of the currency union to another. An EU banking union, while desirable, raises more complex issues, not least the interaction of multiple central banks. Such interaction has consequences for the LOLR function and the relationship between monetary and macroprudential policies. Potentially different access to backstops or safety nets, such as the ESM that is currently available only to euro area members, adds to the complexities. It is, therefore, prudent to proceed first with a euro area banking union, albeit with an option to “opt in” for non–euro area EU members and with adequate governance safeguards for those who wish to stay out (see Section “Dealing with the Outs”).

Should all banks be covered, regardless of size, complexity, and cross-border reach? Or given potential administrative and resource limitations, should the approach be focused on systemically important banks and those that urgently need or may potentially require recapitalization? As experience has shown, systemically important banks require tailored solutions for oversight and resolution. Equally, however, larger numbers of small banks with correlated exposures (for example, Spanish cajas) can threaten systemic stability, especially when policy buffers are low. The ultimate goal should therefore be to supervise all banks, not just systemically important or vulnerable ones, because doing so would allow for complete and evenhanded treatment and limit the scope for regulatory arbitrage or capture. A phasing in of coverage, a risk-based use of scarce supervisory resources, and an appropriate level of decentralization would, in practice, need to be considered, with greater delegation initially to national authorities, and subsequent adjustment as capacity at the center is built.

This section presents considerations for a banking union in the longer term. Near-term transition and crisis management issues are elaborated in the next section. Box 9.2 briefly describes and analyzes the Commission’s proposal and EU Council agreement for an SSM and road map to a banking union.

Single Supervisory Mechanism

An SSM offers many potential advantages. It would facilitate a more systemic approach to tracking the buildup of risk concentrations and contribute to achieving comprehensive macroprudential oversight of the euro area. It would allow the coordination of supervisory actions across countries, and ensure consistent application of prudential norms. It would foster convergence of best practices across members, partly alleviate concerns of regulatory capture at the local level, and promote integration of the single market for financial services. In concrete terms, higher standards of supervision in place before the crisis might have meant a swifter identification of an unsustainable buildup of risk (for example, in Ireland or Spain) and a more timely and effective intervention to diffuse such risk (for example, by applying higher capital buffers or restricting excessive concentrations).

An SSM that involves the ECB has merit. The ECB has access to supervisory skill sets given that many Eurosystem national central banks are also national supervisors. Involving the ECB would give it access to supervisory information in support of its monetary policy and LOLR duties. However, housing both banking oversight and monetary policy under one roof could potentially lead to difficult trade-offs between the two, for example, when monetary policy decisions affect bank solvency, or when the need to safeguard financial stability may call for liquidity provision to insolvent banks. As a creditor, the ECB may also face conflicts of interest when, as a supervisor, it is required to withdraw a license and trigger resolution, resulting in losses to bank claimants. These potential trade-offs call for appropriate checks and balances, such as transparency in the decisions made and implemented by the supervisory board.

Designing a system involving many countries is complex. With more than 6,000 banks in the euro area, an appropriate division of labor would be needed between the center and the national agencies, because no single new body could supervise all banks with full effect. The ECB would need to rely partly on the competencies and resources at the national level, with clarity on the allocation of tasks and powers as well as strong oversight and accountability to ensure compatibility of incentives and contain risks of slippages. The ECB must also be adequately resourced to ensure that it has the capacity to perform key and strategic tasks while it is able to supervise systemically important banks and those that require, or may potentially require, public support. In this regard, the challenge of developing the requisite competence at the ECB and building credibility in supervision should not be underestimated.

Box 9.2.The European Commission Proposal and the Council of the European Union Agreement

European Union (EU) agreement. The September 12, 2012, Commission proposal and December 13–14, 2012, Council agreement covered the design of a single supervisory mechanism (SSM), the passage of three pieces of draft EU legislation, and the role of the European Banking Authority (EBA). According to the agreement, the European Central Bank (ECB) would start carrying out supervisory tasks in March 2014 or one year after the SSM legislation takes effect, whichever is later. Banks receiving or requesting public financial assistance would be targeted first; at the European Stability Mechanism’s request and as a precondition for direct recapitalization, the ECB could begin directly supervising these banks, regardless of the starting date of the SSM. The Council further called for adoption of the Capital Requirements Regulation/Capital Requirements Directive IV, or CRR/CRD IV, as of the utmost priority, and for adoption by June 2013 of draft EU legislation harmonizing and strengthening national resolution regimes and deposit guarantee programs. The powers of the EBA were confirmed as the regulatory and supervisory standard setter and the mediator of cross-border supervision and resolution issues arising in the EU.

Supervision. The agreement specified a clear mandate for bank safety and soundness to the ECB, and its accountability to the European Parliament and the Eurogroup. The ECB would directly supervise banks accounting for about 80 percent of euro area banking assets, including banks with more than €30 billion in assets or 20 percent of national GDP, or if otherwise deemed systemically important (for example, given cross-border reach). At least the three largest banks in each member state would be directly supervised, with the ECB retaining the power to bring any bank under its supervision if deemed necessary by the supervisory board. At the ECB, a supervisory board and a steering committee are being created to prepare and implement decisions and give voice to non–euro area members that opt in (because the Governing Council remains ultimately responsible).

Operational details. The agreement appropriately conferred broad investigatory and supervisory powers on the ECB, which is responsible for the effective and consistent functioning of the SSM (although national authorities are also responsible for the banks remaining under their direct supervision). The challenges of effective implementation must not be underestimated. Operational arrangements now need to be specified, which must make incentives compatible between national authorities and the ECB, ensure an adequate division of labor, and provide for appropriate information sharing within the SSM to underpin effective supervisory decision making. The ECB has adopted a detailed framework for the practical modalities of supervisory cooperation within the SSM. The governance setup is complex, but must seek to promote, not hinder, timely decision making. Moreover, to be effective, it is essential to appropriately and urgently staff the ECB. The Council agreement provided both national authorities and the ECB with powers to make use of macroprudential instruments specified in the relevant EU directives, although accountability must be clarified. In practice, close cooperation will be critical to ensure coherence and effectiveness of measures. Resolution and safety nets. A single resolution mechanism is an indispensable supplement to effective supervision, ideally centered on a single resolution authority, deposit insurance, and common backstops. The Council recognized the importance of a single resolution mechanism with adequate powers and tools. This mechanism is to be based on financial sector contributions and backstop arrangements that recoup taxpayer support in the medium term. The Commission made a proposal in this regard in in 2013. Though the immediate priority is to strengthen and harmonize national regimes, funds raised over time from financial institutions could cover individual small to medium-sized bank failures. Common backstops are essential to handle systemic failures. A time-bound road map to common safety nets—needed for depositor confidence and to break sovereign-bank links—would limit the risks of an incomplete banking union.

Guidance on the design of an effective mechanism is provided in the Basel Core Principles (the “Core Principles for Effective Supervision”). According to these principles, a number of preconditions and prerequisites must be met at the euro area level.

  • Preconditions for sound banking include (1) the implementation of coherent and sustainable macroeconomic policies, (2) a clear framework for financial stability policy, (3) an effective crisis management and resolution framework to deal with bank failures and minimize disruptions, (4) an adequate safety net to deal with a crisis of confidence while minimizing distortions, (5) a well-developed public infrastructure, and (6) effective market discipline.

Clearly, some of these preconditions are beyond reach in the immediate future, but they are essential for the effectiveness of the new system in the longer term.

  • Prerequisites to establish a sound basis for the SSM include (1) operational independence of the SSM; (2) clear objectives and mandates; (3) legal protection of supervisors; (4) transparent processes, sound governance, and adequate resources; and (5) accountability. The Commission proposal and EU Council agreement by and large meet these prerequisites. Clarity is required on resources, and aspects of the governance mechanism and legal robustness, for example, delegation, may require strengthening.

Supervisors at the national and supranational levels must have clear responsibilities and the powers necessary to perform their tasks. Having formal responsibility but no real enforcement power (as could occur at the ECB level) would carry serious risks, whereas having the power but no clear responsibility or accountability (as could occur at the national level) could lead to misaligned incentives and distorted outcomes. A monitoring mechanism thus becomes important. The ECB should have the ability to conduct joint inspections and peer reviews and establish cross-country teams of supervisors with an ECB-appointed head (drawing lessons from EBA’s experience with peer review could be helpful). Along with operational independence, accountability, and governance, joint inspection teams would provide safeguards to further align national interests with the interests of the center.

Although the euro area has more than 6,000 banks (a similar order of magnitude as in the United States; see Figure 9.5), the 150 largest banks account for about 80 percent of banking system assets. Thus, some degree of delegation is necessary. Full centralization is neither practical nor desirable because supervisory knowledge and resources remain at the national level. Full decentralization in which the center merely validates decisions is not desirable either, particularly when common resources are at stake (for example, ESM direct recapitalization of banks or future common backstops). The goal should be to create a coherent and consistent supervisory mechanism with adequate information flow and final significant decisions made at the center. To ensure incentives are compatible, the degree of delegation should be clarified. The extent of delegation would depend on the ECB’s supervisory classification of risks for each bank, and factors such as the importance of local knowledge and know-how, the systemic dimension of banks and tasks, and the amount of discretion required in decision making. For example, institutions with systemic implications should be subject to more intrusive supervision from the center, as should functions that are more subject to discretion, capture by the industry, or political influence. Consolidated supervision of financial groups would involve interagency coordination to oversee nonbank financial institutions as well.

Figure 9.5Banks under Supervision

Sources: European Central Bank; FED; FDIC; and NCUA.

To facilitate identification of, and action on systemic risks, including information sharing and home-host coordination, and internalization of cross-border externalities, some shift in macroprudential mandates and tools—away from member states and toward the ECB—should be considered. The ECB should be given binding powers to be able to use macroprudential instruments if it deems doing so to be necessary. Given the ECB’s established expertise on financial stability, the setup costs should be limited. The framework should involve national authorities and be sufficiently flexible to tailor solutions to local conditions. Therefore, national authorities should also be able to make use of macroprudential tools within the parameters and guidelines set by the ECB, and mechanisms are needed to resolve conflicts of interest that may arise between national authorities and the ECB. The ECB should also coordinate with the ESRB.

The regulatory framework would largely build on existing and upcoming regulations designed at the EU level, in close coordination with non–euro area EU members.

Single Resolution Mechanism

A single authority as presumptive receiver of failed banks can facilitate timely resolution, including of banks that operate across borders. Such an authority provides a mechanism for internalizing home-host concerns and reaching agreement on resolution and burden sharing. It can thus help to avoid the protracted and costly resolutions that occurred, for instance, in the cases of Fortis and Dexia (Box 9.3). A single authority is also necessary for aligning incentives for least-cost resolution—because a common backstop in the context of a decentralized mechanism would provide mal-incentives to shift residual losses from national taxpayers to those in the euro area. Pooling bank resolution capacity in a single body would achieve economies of scale, avoid incoherence and duplication, and accumulate expertise for the preparation and implementation of recovery and resolution plans, in particular for systemically important institutions. At the same time, there are important complexities. If a common approach to resolution is based on national legislation, significant resources may be needed at the national level to deal with different legal regimes. A single resolution authority, the preferred approach, would be based on supranational legislation, but any treaty change would require time.

Emerging best practices are laid out in the “Key Attributes of Effective Resolution Regimes for Financial Institutions,” according to which the resolution authority should seek to resolve financial institutions without disrupting financial stability (Financial Stability Board 2011). It should minimize costs to taxpayers; protect insured depositors; and ensure that shareholders and unsecured, uninsured creditors absorb losses. There are prerequisites and preconditions of effective resolution:

  • Preconditions for effective bank resolution include (1) a well-established framework for financial stability; (2) an effective system of supervision, regulation, and oversight of financial institutions; (3) safety nets and protection programs; (4) a robust accounting, auditing, and disclosure regime; and (5) a well-developed legal framework.
  • Prerequisites for a strong authority include (1) operational independence consistent with statutory responsibilities; (2) transparent processes; (3) legal protection; (4) sound governance and rigorous evaluation and accountability mechanisms; and (5) adequate resources.

Box 9.3.Resolutions of Fortis and Dexia

Fortis. In the aftermath of the Lehman Brothers collapse, an agreement was reached on September 28, 2008, to save the troubled Fortis group with taxpayer support from Belgium, Luxembourg, and the Netherlands. The agreement, however, fell apart soon thereafter as liquidity pressures mounted. Subsequently, the Belgian parent company sold the shares of the Dutch parts of Fortis to the Netherlands government, while Belgium and Luxembourg sought a common solution for their parts of Fortis, eventually agreeing to sell the banking arm to BNP Paribas. This breakup along national lines was a setback to financial integration in the Benelux and was likely more costly than a first-best joint solution for the group.

Dexia. Dexia failed in 2011 after losing access to wholesale funding and facing increased collateral demands on interest rate derivatives. The resulting breakup was segmented along national lines. On October 10, it was announced that the Belgian operations would be purchased by the Belgian government; foreign subsidiaries in Canada, Italy, Luxembourg, Spain, and Turkey would be put up for sale; and parts of the French operations would be purchased by two French public sector banks. The remaining troubled assets, including a €95 billion bond portfolio, would remain in a bank runoff (Dexia SA) that would receive funding guarantees of up to €85 billion provided severally but not jointly by the governments of Belgium, France, and Luxembourg and recapitalization of €5.5 billion. At end-2012, the Commission approved the resolution plan for Dexia Group.

A single authority would need a mandate, alongside the supervisor, to develop resolution and recovery plans well ahead of, and to intervene before, insolvency using well-defined quantitative and qualitative triggers. The authority would need strong powers to take early intervention measures (for example, to require capital conservation measures or restrictions on activities), restructure banks’ assets and liabilities (for example, apply a “bail-in tool” to subordinated and senior unsecured creditors, transfer assets and liabilities—”purchase and assumption”—to a sound acquirer, and separate bad assets by setting up an asset management vehicle), override shareholder rights (subject to them being no worse off), establish bridge banks to maintain essential financial services, and close banks.

Because resolution involves sensitive choices about the distribution of losses, clear burden-sharing mechanisms would need to be established in advance to realize least-cost resolution. At the same time, a systemic risk exception is needed (Box 9.4).

  • Hierarchy—Respecting the hierarchy of creditor claims, the resolution authority should be able to haircut or extinguish unsecured liabilities, starting with equity and potentially extending to senior unsecured debt, according to a clear creditor priority list. This ability would reduce uncertainty in the capital structure and any eventual resort to taxpayer funds. Given their explicit taxpayer backing, insured depositors would need to be given clear priority among unsecured bank liabilities to maximize recovery of deposit payouts from failed banks in resolution. Depositor preference provisions should be included in EU legislation, possibly in the draft recovery and resolution directive.
  • Contributions—Contributions from the industry—held in a fund—should be used first to finance resolution. Insofar as private sector contributions and loss allocation across uninsured and unsecured claimants might be insufficient in a systemic crisis, a common backstop would need to be tapped. Contributions could follow specific rules established in advance, for example, based on the ECB or ESM capital keys.

The resolution fund would build resources over time through levies on the industry. It would be prefunded through premiums. These premiums should be risk based, which would help capture national specificities in risk parameters. Use of funds could be complemented by arrangements to recoup net losses through levies on the industry after any crisis. In the event a systemically important institution is under stress, common backstops, including temporary funding support from the ECB, with safeguards, would ensure adequate liquidity. The EU draft directive on bank recovery and resolution states explicitly that, under the new EU regime, national resolution authorities will have the possibility to borrow from the central bank. The resolution mechanism should specify clear exit strategies that maximize the value of participations acquired and prohibit national preference.

Box 9.4.A Systemic Risk Exception for Europe

Europe’s bank-sovereign conundrum is intertwined with the issue of systemic risk. In many countries, national authorities are supporting banks that would otherwise fail, including those considered too big to fail. Fiscal authorities have plugged negative equity, topped up capital, provided guarantees, and set up state-owned asset management companies or asset protection programs—in some cases hurting sovereign credit standing in the process. National central banks, in turn, have extended vast volumes of emergency liquidity assistance to plug structural funding gaps in the wake of funding runs—financing this lender-of-last-resort liquidity through TARGET2.

Systemic risk creates legitimate concerns. Policymakers perceive, often correctly, that bank closure followed by least-cost resolution will trigger contagion, bank runs, and asset fire sales—including of government securities—threatening damaging wealth effects, loss of market access, and worse. In such situations, bailouts may present the lowest-cost option. However, in some countries with particularly oversized banking systems relative to their domestic tax bases—for example, Ireland—loss of market access by both banks and the sovereign has proved unavoidable.

The U.S. system, with its statutory “systemic risk exception,” may be instructive. In the United States, the default option for failed banks is resolution at least cost to the Deposit Insurance Fund. However, the statutes also include a systemic risk exception to the least-cost resolution requirement. Given evident moral hazard, the law sets a high bar to invoke the exception: a “three keys” approach requiring support by two-thirds of the Federal Deposit Insurance Corporation (FDIC) and Federal Reserve Boards and concurrence by the secretary of the treasury after consultation with the president. The exception was invoked four times in 2008:Q4, once to justify the FDIC’s blanket guarantee, and three times to rescue failing U.S. systemically important financial institutions—with each of the latter three operations combining fiscal support and large Federal Reserve funding commitments.

Europe, too, needs procedures to analyze the possibility of systemic risk in bank failure. A systemic risk exception for the euro area would lend needed clarity and credibility to the bank resolution process. Formal vetting procedures would limit moral hazard and protect resolution funds (aided by indicative criteria giving shape to unacceptable systemic risk). Recognition will be needed, however, that because of financial fragmentation the case for intervention can be made not only for a few European systemically important financial institutions, but also for “domestically systemic” banks. Thus, systemic risk determinations should include a concern for local credit conditions even for banks whose failure would have little direct impact at the euro area level.

A systemic risk exception for the euro area would be a unique construct. The triggering procedure would need to balance national and collective concerns, and recognize that systemic bank resolutions typically require both fiscal and central bank backstops. Findings of unacceptable systemic risk could, for instance, require two-thirds majorities of the national central bank and European Stability Mechanism boards (the latter later replaced by the single resolution authority) with concurrence by the ECB’s Governing Council. This or some similar arrangement would help ensure proper and credible vetting of bank failures, sorting banks into those warranting least-cost windup and those requiring support.

The resolution authority should manage the resolution and deposit insurance fund. Given the complex fiscal decisions involved, and the need for checks and balances, the resolution authority should be set up independently of, but on par with, the ECB supervisory mechanism. Coordination and information sharing between the two, with due attention to incentives compatibility, is essential. For instance, consideration could be given to the creation of a committee comprising the head of the supervisory function of the ECB and the chairman of the resolution authority. Alternatively, the ECB head of supervision could serve on the board of the resolution authority, together with national representatives and representatives of other EU bodies.

Common Safety Nets

Once uniform prudential oversight across euro area countries is effectively in place, it becomes feasible to contemplate the pooling of risk. By pooling risk, common safety nets—common deposit insurance and centralized LOLR—will not only help countries avoid disruptions that may overwhelm their individual capacities but also form a key pillar in the incentives compatibility of the banking union. If a weak sovereign is perceived not to be able to honor its safety net obligations, losses of confidence can quickly follow, triggering capital flight and deposit outflows. A pooled mechanism with credible backstops would more effectively protect confidence and diversify risks across banks. But without common safety nets and backstops, the banking union would remain an incomplete and risky construct that fails to delink the funding costs of weak sovereigns from that of their banks. It would also risk jeopardizing the credibility of the ECB and the SSM by leaving the system vulnerable to financial fragility.

The resolution and deposit insurance fund could be relatively small and cover some individual bank failures, with fiscal and central bank backing to be used in the event of a systemic crisis. The fund could, in practice, cover both resolution and deposit insurance; if the ranking of claimants is clear and adequate, the objectives of the two functions would be aligned. Although there is no well-established good practice, the typical target size of resolution and deposit insurance funds could range from about 1–2 percent of total liabilities (excluding equity) in large systems (as in the Commission proposal, or as in the United States) to 4–5 percent in smaller systems, where the aim is to cover two or three midsized banks and four to six small banks. The target size also varies with the quality of the institutional environment and resolution regime, including the presence of prompt corrective actions and early structured intervention mechanisms.

The idea of creating a common resolution fund, administered by the single resolution authority, while harmonizing deposit insurance programs but allowing them to remain at the national level has also been discussed. Such a model would enhance the effectiveness of the SSM while providing common financing for resolution, although without common backstops its impact would be limited. Under this model, it will be essential that national deposit insurance funds are available to contribute to resolution, up to the amount available for payout. Even so, the disadvantages would be less efficient risk pooling, which would not effectively decouple sovereigns and banks; complexities in cost allocation and implementation for cross-border failures, requiring close coordination between national deposit guarantee programs and the single resolution authority; and duplication of costs and administrative resources because the same banks would be assessed for both funds.

The LOLR makes liquidity support available to solvent yet illiquid banks. Centralizing all LOLR functions at the ECB would, in the steady state, eliminate bank-sovereign linkages present in the current emergency liquidity assistance (ELA) program (see Box 9.1). Centralization would require changes to the ECB’s collateral policy because by definition euro area banks that tap ELA cannot access Eurosystem liquidity owing to collateral constraints. Until such time as all banks are brought under the ECB’s supervisory oversight, ELA would be sourced through both the ECB (for banks brought under its purview) and national central banks (for banks that remain under national supervision, albeit with adjustments made to the national ELA limits).

How Do We Get There?

A firm plan of key deliverables toward a full banking union should anchor execution. In normal times, a gradual sequencing might start with first harmonizing rules governing national regulation, supervision, resolution, and safety nets, a process that may take some years as EU directives are fully adopted at the national level. This step would be followed by gradual development of new common institutions. Eventually, once there is agreement on burden sharing, the process would culminate in a full banking union, with an SSM, a single resolution authority, a common resolution and deposit insurance fund, and common backstops. After all, a decade elapsed between the European Council decision to realize the monetary union in 1989 and the launch of the euro in 1999.

But times are not normal. Critical to addressing the crisis is the repair of the financial sector, including through bank recapitalization that weakens the damaging bank-sovereign dynamic and creates more favorable conditions for stabilization. In the context of private and public sector deleveraging, raising resources domestically to recapitalize banks is challenging (impossible in some jurisdictions). At the same time, closing domestic systemically important banks continues to pose a risk of uncontrollable consequences. Shared support for recapitalization would facilitate financial and economic stabilization at the national level, and thus for the monetary union as a whole, although it raises questions about burden sharing and moral hazard.

In June 2012, euro area leaders affirmed, “it is imperative to break the vicious circle between banks and sovereigns.… When an effective single supervisory mechanism is established, involving the ECB, for banks in the euro area the ESM could, following a regular decision, have the possibility to recapitalize banks directly” (Council of the European Union 2012). The Council agreed on December 13–14, 2012, that, at the request of the ESM, the ECB may start directly supervising a bank as a precondition for direct recapitalization, once the SSM regulation takes effect (that is, even before March 2014). The Council also called for agreement on an operational framework for direct recapitalization, including the definition of legacy assets, by June 2013.

To be clear, the core purpose of ESM recapitalization of domestically systemically important banks undergoing restructuring must be to remove the residual risk from the balance sheet of a sovereign whose finances are already strained. Unviable, nonsystemic banks should be wound down at least cost, and systemically important banks should be resuscitated by shareholders, creditors, the sovereign, and the ESM as the quintessential patient, deep-pocket investor. By delinking the sovereign from future unexpected losses on bank balance sheets, ESM direct recapitalization would remove future tail risks from the sovereign balance sheet; ensuring that the banks have an owner of unquestioned financial strength would improve bank funding conditions. Thus, the ESM would attack the sovereign-bank link from both sides. In all cases, ESM involvement should be conditional on a determination of systemic risk, which could be as basic as a finding that the bank is too large for the sovereign alone to wind up, given the state of public finances. A robust mechanism for the systemic risk determination will be critical (Box 9.4).

Sequencing the key steps toward a banking union while implementing crisis resolution measures raises at least four issues: (1) what is essential for “effective” supervisory control? (2) how should steps toward a full banking union be phased in? (3) how should burden sharing in ESM direct recapitalization of banks be accomplished in practice? and (4) what are the risks? Each of these issues is taken up in turn.

Effective Supervision

How should the SSM be judged as having become effective?

  • Basel Core Principles—One approach would be to apply the Basel Core Principles. Full supervisory capacity would need to be set up, and the various prerequisites and preconditions met, including establishing an adequate safety net and crisis management and resolution frameworks. The approach demands completeness, and could take years to achieve. In effect, the present crisis resolution efforts should be viewed as stepping stones to the longer-term task of building a robust architecture for financial stability. The Basel Core Principles are therefore the internationally recognized standard for use in the new steady state.
  • Pragmatic approach—A more modest and pragmatic approach for the near term was adopted by the EU Council at its December 2012 meeting: as soon as the SSM regulation is adopted, the ECB could, at the request of the ESM, start to directly supervise a bank, as a precondition to it receiving ESM direct recapitalization. If pursued, this approach opens the door to the critical task of repairing weak, systemically important banks, while ensuring that the ECB has the powers and tools needed to supervise these banks. It is thus critical to swiftly secure legislative agreement and adoption, establish a single rule book, and begin preparation both for supervision at the ECB and for ESM direct recapitalization (including finding common ground on remaining issues in the first half of 2013).

Sequencing Banking Union

An agreement on burden sharing and common backstops at the outset would both pave the way to a full banking union and facilitate current crisis management. It would set the stage for a single resolution authority that would coordinate corrective actions with the SSM. Absent centralized resolution, the SSM would need to coordinate with multiple national resolution authorities, especially for banks operating across borders. As a result, conflicts over the distribution of losses could arise because the fiscal consequences of coordinated supervision remain national. The incentives would be to shift the costs of resolution, delaying action and jeopardizing the achievement of least-cost resolution. Therefore, sequencing in which the SSM is set up alongside an agreement on burden sharing and common backstops would ensure that decision making at all levels of the banking union is incentives compatible. Following that, a common resolution fund and deposit guarantee program could be set up to provide a risk-sharing mechanism.

The elements above—an SSM, single resolution mechanism with common backstops, and common safety nets—are necessary for a successful banking union. Missing elements would result in an incoherent banking union and, at worst, an architecture that is inferior to the current national-based one. Therefore, ideally, progress would be made on each of the elements. Given the need to resolve outstanding differences of views on the details and timing, however, it may not be possible to make progress on all the elements now. This lack of simultaneous progress could result in a different sequencing than ideally warranted. In any event, it is crucial that all the elements of a successful banking union are included eventually. A well-defined timetable at the outset would remove uncertainty, bolster confidence in the political willingness to build a robust financial stability architecture, and anchor execution. A possible approach could be as follows:

  • Harmonized legislation—As noted, adoption of the single rule book—CRR/CRD IV, the recovery and resolution directive, and the deposit insurance directive—needs to proceed urgently. Agreed-on drafts of these pieces of EU legislation should be submitted to the EU Parliament as soon as possible, with approval during the first half of 2013 and adoption in national legislations starting in the course of 2013.
  • Implementing the SSM—A phased rollout of the SSM could seek to make it “effective” for troubled systemically important banks over the course of 2014. The emphasis should be on establishing a strong SSM in which the ECB has formal powers, the decision-making processes, and the capability to perform essential supervisory tasks in an intrusive delegated monitoring model. The ECB must be able to request and receive all necessary information, conduct off-site diligence, field on-site inspections, and pursue further action on any bank in the euro area. A well-functioning information and evaluation infrastructure must be established quickly so that the ECB can serve as a central supervisor.
    • ° The ECB would need to put in place adequate resources and organizational capacity to commence selected supervisory tasks, which will be a complex and demanding exercise. Based on information to be provided by national authorities on their banks’ supervisory histories and risk profiles, the ECB could then start off-site stocktaking of the banks under its supervision, to prioritize institutions in need of deeper diagnostics based on risk.
    • ° Once essential capacity is in place, the ECB could start, say by end of 2014, actively supervising banks receiving state aid, with an appropriate degree of delegation to national authorities but with the key decisions made at the center. Reaching agreement before this date on ESM direct recapitalization would allow for direct recapitalization to occur soon thereafter.
    • ° With the largest 150 banks accounting for about 80 percent of banking system assets in the euro area, the supervision of most banks could remain the responsibility of national authorities, although the ECB should have the prerogative to step in wherever needed (as indeed provided for by the EU Council agreement). This would allow an initial focus on the banks at the heart of the current crisis, such as the Spanish cajas, where correlated exposures fueled systemic risks. In all these cases, appropriate delegation to national authorities would be essential, with accountability and governance arrangements, but with key decisions made at the center. Delegating also provides flexibility, particularly during the transition when the ECB builds resources, but caution is needed to not lock in an imperfect delegation practice in the SSM. Adequate early intervention powers for the ECB provide incentives for cooperation, and are essential for the effectiveness of the SSM.
  • Implementing single resolution and common safety nets—
    • ° It is essential that the EC commit to a firm timeline for implementing a single resolution mechanism, including burden-sharing arrangements. The Commission is to present a proposal in the course of 2013.
    • ° When agreement on adequate resolution (and deposit insurance) funding and backstops is in place, the single resolution authority could begin operating. Meanwhile, resolutions would be handled by the national authorities under strengthened regimes (and, as needed, support from the sovereign with borrowing from the ESM).

The ECB-centered SSM would have powers of early intervention. It would work with national resolution authorities to resolve or restructure weak institutions, until a single resolution authority with common backstops is established. To facilitate the process, there may be merit to establishing a temporary body or rapidly creating an EU agency tasked with the coordination of bank crisis management and resolution among national authorities and the ECB.

Steps should also be taken toward common safety nets. A reinsurance program, for instance, could be created from national deposit guarantee programs, funded at the euro area level through industry levies and contributions from member states. It would pool risk and weaken sovereign-bank links. Agreement beforehand on the shares of national and supranational funding in depositor payouts would limit moral hazard. Over time, the fund would build administrative capacity, and could be a step toward a permanent euro area program and resolution fund.

Finally, although working under the existing treaty framework is the swiftest way to begin, gradually strengthening the legal framework would minimize implementation and litigation risks (Box 9.5).

ESM Direct Recapitalization

Mobilizing the ESM direct bank recapitalization tool in a forceful and timely manner is critical to developing a path out of the current crisis, and would complement other measures such as the ECB’s Outright Monetary Transactions. Recapitalization of frail, domestically systemically important banks in the euro area, including some migration to the ESM of existing public support to such banks, can help break the vicious circle between banks and sovereigns, reduce financial fragmentation, repair monetary transmission, prepare for banking union, and thus help complete the economic and monetary union. To be sure, failing nonsystemic banks should be resolved at least-cost to national resolution funds and taxpayers. Equally, systemically important banks benefiting from ESM support will need effective supervision and reform to be returned to full viability and private ownership, with state aid rules mandating formal restructuring plans. In some cases, the sovereign itself may need an adjustment program, providing an enabling environment for asset price recovery.

Box 9.5.Legal Considerations

Legal basis of the banking union. Under Article 127(6) of the Treaty, the European Central Bank (ECB) is able to take on specific supervisory tasks without treaty change, upon a unanimous decision of the European Council and after consultation with the European Parliament and the ECB. The EU Council agreement vests in the ECB exclusive authority for a wide range of supervisory tasks. Article 127(6) provides a legal basis, and has been interpreted expansively to establish the single supervisory mechanism (SSM). The draft SSM regulation carefully attempts to specify the ECB’s remit, but litigation risks may, in principle, not be excluded—any financial institution confronted with a supervisory decision by the ECB could bring a case before the European Court of Justice (ECJ) on grounds of lack of competence. In the medium term, providing explicit legal underpinning for financial stability arrangements in the Treaty would further strengthen their legal soundness. This would allow financial stability to be anchored as a key objective under the Treaty and for the roles and powers of all the safety net players to be defined, including a full-fledged resolution authority with a common backstop.

Shared competences and responsibilities. Under Article 127(6) and the regulation that is based upon it, supervision will remain a shared competence between the ECB and member states. The ECB will be responsible for certain supervisory measures, while member states retain their powers with respect to any aspect that is not covered by the draft SSM Regulation (for example, anti-money laundering, consumer protection, and some macroprudential tools). For the tasks conferred to it, the ECB would make the final decisions vis-à-vis “significant” banks, while the national competent authorities (NCAs) would assist the ECB with the preparation and implementation of such decisions, pursuant to the ECB’s instructions. For other banks, NCAs will formally make supervisory decisions, but still under ECB instructions. The ECB will be responsible for the effective and consistent functioning of the SSM, and both the ECB and the NCAs will be subject to a duty of cooperation. The overall division of tasks and responsibilities will need to be clarified to remove any remaining uncertainties as to who, as a legal matter, will be ultimately accountable for supervisory decisions.

ECB governance. The Governing Council is the ultimate decision-making body of the ECB, as enshrined in the Treaty, including for any supervisory tasks conferred upon the ECB under Article 127(6). Several challenges may arise from this setup. First, because the supervisory board can only prepare the supervisory decisions to be made by the Governing Council, it is only the latter that will formally be responsible for supervision, in addition to monetary policy. Therefore, the separation of monetary and supervisory responsibilities can only be implemented at an operational level because the legal mandate of the ECB, pursued by the Governing Council as the ultimate decision-making body, remains unaltered. Second, the existence of multiple layers of governance arrangements, coupled with the impossibility of delegating decisions to the supervisory board, may create a burdensome process; legal risks may arise from the need to align daily supervision practices with the legal requirements dictated under the Treaty and the SSM regulation. Last, because non–euro area SSM participants cannot be represented on the Governing Council of the euro area, taking part in the SSM decision-making process would require alternative arrangements for their voices to be heard, such as through the supervisory board, with mediation channels to resolve differences.

EU banking laws. The EU’s banking laws have significant weaknesses in both form and substance. On form, the current approach based on directives implemented in national laws must swiftly be replaced by a directly applicable single rule book. On substance, current weaknesses in EU banking law will also have to be remedied, for example, weak fit-and-proper criteria and the absence of restrictions on related-party lending.

Legal actions at the national level. Whether the EU regulation conferring supervisory tasks on the ECB also requires legal changes at the national level is unclear. However, amendments to national legislation seem to be inevitable, to provide the legal clarity that will ensure the smooth functioning of the SSM. Absent a single, directly applicable rule book, such amendments to domestic legislation may also be necessary to improve national supervisory regimes.

The mobilization of the ESM direct recapitalization tool should ensure that frail, domestically systemically important banks have adequate capital, access to funding at reasonable cost, and positive profits—in short, a viable business model. To this end, asset valuations are critical, as are the roles of shareholders, creditors, and the domestic sovereign in bearing costs.

  • In principle, there would be significant advantages to breaking the vicious bank-sovereign circle if all the capital needed to ensure a systemically important bank was adequately capitalized was ultimately provided by a central fiscal authority. This would especially be the case if the scenario were to play out in a small jurisdiction, and even more so if it also had to internalize spillovers to others (that might result, for example, if external creditors did not share in losses, for fear of triggering wider problems). More generally, pooling risk would provide protection to all in advance of any crisis, because any country could, in theory, find itself in a similar position in the future.
  • In practice, although the treaty establishing the ESM provides for the possibility of losses, such losses are not expected in its financial operations, including bank recapitalization. As a bank investor, the expectation is that the ESM must be careful to take balanced risk positions. It likely could not provide capital that a patient investor would not expect to recover over time. Thus, capital needed to bring a systemically important bank out of insolvency (that is, to bring it from negative to nonnegative equity) would in the first instance need to be provided by shareholders and creditors, and then by the national government, with any remaining shortfall covered by the ESM. Fortunately, there are unlikely to be large, insolvent banks currently in most economies.
  • A balanced approach would prudently internalize the benefits of ESM capital support by looking ahead over a time horizon sufficiently long to realize the benefits. As a patient, deep-pocket investor, the ESM should take a long-term perspective in its investment decisions, recognizing that gross up-front crisis outlays tend to dwarf ultimate costs net of recoveries or capital gains and, in many instances, generate positive financial returns.
  • Asset valuation—The implications for asset valuation, which determines the size of recapitalization needs as well as investors’ upside or downside risks, are twofold. First, asset values should be neither too high (which would imply mutualization through the back door) nor too low (in which case, the private sector could simply buy the assets, and there would be limited benefit to having an official investor). Second, because the ESM is a patient investor willing to give the banks the necessary time to restructure, assets should be priced at values that give due consideration to the positive effect of recapitalization on asset values. This effect includes not just the direct positive effect of recapitalization (including more favorable funding costs) and recovery, but also the removal of tail risk events (see next bullet point).
  • Risk sharing—As a patient, deep-pocket investor, the ESM provides assurance to creditors that, in the event of a negative surprise, potential future capital needs can be met. In other words, although the ESM would not take on expected losses, it would shoulder the risk of unexpected future losses. This approach is in line with efficient risk sharing, wherein the patient investor bears the residual risk. In this regard, it should be noted that, conditional upon the ESM standing ready to take material losses in a downside scenario, the ESM would be unlikely to actually incur those losses, because the investment would minimize the risk of the adverse scenario occurring.
  • No first loss guarantees—ESM investments should not benefit from loss protection provided by the sovereign. Such approaches would preserve sovereign-bank links, undermining the purpose of ESM direct recapitalization. But there should be safeguards for the ESM (for example, built into the sales contract) against domestic policies that could directly harm the viability or profitability of the recipient banks (for example, onerous taxes imposed after the crisis or stiff resolution levies).
  • Exit strategy—Incentives should be provided for an early ESM exit and private investor entry. The timing would be built around the EU-approved restructuring plans. Mandatory sunset clauses should be avoided because they could affect negotiating power ahead of the deadline.
  • Adequate resources—Direct equity injections into banks could absorb significant amounts of ESM capital. It would be important to ensure that the ESM has adequate capital to not only allay any investor concerns about ESM credit quality, and thereby limit any rating implications, but also play its potential role as a common backstop for bank recapitalization.

This term has been very controversial, reflecting concerns that creditor countries could be expected to put capital into unviable banks. This is not what is being suggested above. Rather, losses on impaired “legacy” assets should be recognized through up-front provisioning and proper (long-term, postcrisis) valuation. It is not recommended that all impaired assets be segregated from the bank before ESM direct recapitalization and placed into recovery vehicles ultimately backed by the national taxpayer; such an approach would greatly reduce the effectiveness of the tool in addressing bank-sovereign links. Rather, bank health should be restored with shareholders, including the sovereign, bearing the expected loss of past excesses by being subjected to an independent valuation exercise consistent with the shared commitment to restore full viability after the restructuring period.

To further support balance sheet clean up, certain classes of legacy assets could be transferred to asset runoff vehicles such as asset management companies (AMCs) under ESM ownership. Expected losses would remain with the sovereign, given the terms of the foregoing recapitalization. But to limit further contingent fiscal liabilities and harness efficiencies, consideration could be given to allowing the ESM to set up and own AMCs. Possible roles for the ECB in supporting AMC operations could also be considered (although concerns regarding the prohibition on monetary financing may also be raised). ECB funding, if possible under its statute, would help smooth over time the warehousing and disposal of hard-to-value and hard-to-sell assets. An alternative would be for the ECB to support AMC operations indirectly by accepting ESM-guaranteed AMC bonds issued to banks in Eurosystem refinancing operations.

Risks in Transition

The move toward a banking union must not stop prematurely with an SSM. If it does, although the benefits of coordinated supervision would accrue, the costs derived from coordinating resolution across national authorities—with limited incentives for least-cost and rapid action, and no common backstops—could undermine effectiveness. Therefore, critical design aspects must not be deferred far into the future. Agreement on burden sharing and ESM direct recapitalization must also not be delayed, lest the costs of the crisis keep mounting. For example, the lack of common backstops could lead to a slower pace of restructuring or resolving problem banks. Policy paralysis or backsliding in the current environment could derail confidence and the recovery. Still, progress under constraints may be better than no progress at all, as long as all the essential elements of the banking union come together in time.

Pragmatism would need to govern decisions related to SSM resources. Mobilizing the necessary resources, building the requisite capacity, and putting in place effective and incentives-compatible structures will inevitably involve iteration. In the meantime, supervisory drift could occur. To contain these risks, clarity is essential on the responsibilities and accountability of the various supervisory authorities. The ECB must move swiftly to put in place cross-country teams for the supervision of the most systemically important or fragile banks. It will also need to set up the capacity to interact frequently and effectively with national resolution authorities.

Conflicts of interest could arise between the ECB’s monetary policy function and its supervisory responsibilities—for example, when monetary policy decisions affect bank solvency, or when the need to safeguard financial stability may call for liquidity provision to insolvent banks. These risks would be exacerbated by the lack of a robust resolution (and early intervention) regime and safety nets. Besides completing the architecture, consideration could be given to strengthening the governance of the decision-making process and accountability of supervision at the same level as the central banking functions. The Council agreement on the SSM strengthened the governance arrangements relative to the Commission proposal. The supervisory board will draft decisions that will be deemed adopted unless the Governing Council objects within 10 days in normal times and 2 days in stressful times. In practice, it will be important to balance the representation of national interests and public officials from the ECB in the governance structure of the SSM.

A banking union could change the very structure of finance. Further consolidation of banks could occur and new banks could enter (for example, as country-level rules give way to uniform union-wide ones), and interconnections may strengthen across the banking union. More banks could become too systemically important to fail at the union level—the banking union does not solve the too-big-to-fail problem—and the associated moral hazard might not be resolved (the SSM may help contain incentives to grow excessively, including, for example, through capital surcharges for systemically important financial institutions). These and other issues would raise new challenges for financial stability, requiring renewed vigilance and decisive action against identified risks. In this respect, some of the recommendations of the EU High-Level Expert Group on structural bank reforms (the Liikanen report) could prove useful.

Dealing with the “Outs”

This chapter argues that a banking union is necessary for the euro area. What is the benefit to non–euro area EU members of a euro area banking union? It should be recognized that, by enhancing stability and removing financial market fragmentation, a well-functioning euro area banking union generates positive spillovers and enhances the functioning of the EU single market for financial services. Therefore, other EU members have a legitimate interest in ensuring that the new system is set up properly. A single euro area supervisory mechanism can also solve coordination problems related to the supervision of cross-border banks.

But a euro area banking union raises a number of issues for the “outs”: (1) how can the interests of the “outs” be protected in decisions made by the SSM, for example, in the EBA and supervisory colleges in which the SSM would be represented by the ECB, on the integrity of the single market, and on issues such as the fiscal consequences of decisions on banks with cross-border operations; and (2) how can the interests of those who wish to join the banking union but keep their own currency be represented, given that they would not have a voice in the ECB Governing Council? Some also want the banking union to be made more attractive to join, for instance, by facilitating access to backstops and safety nets, albeit with commensurate contributions.

Strengthening the EBA provides an avenue not only for protecting the interests of the “outs” but also for coordinating action. In this regard, the Commission proposal and EU Council agreement confirming the role of the EBA as the mediator of cross-border supervision and resolution issues and the regulatory and supervisory standard setter in the EU is helpful. Non–euro area EU members should retain an adequate voice within the EBA. The Council agreement modified voting procedures within the EBA board with double majority voting to balance the interests of the “outs.” It will be important that the EBA be an effective and credible force in the single financial market, including limiting concerns about regulatory arbitrage. Likewise, the ESRB’s role as the main macroprudential oversight body in the EU would need to be strengthened further, and it should cooperate closely with the ECB, once the ECB takes on greater macroprudential responsibilities.

The Council agreement seeks to allow non–euro area EU countries to opt in, through representation and procedures on the supervisory board (since these members cannot be represented on the ECB’s Governing Council). As noted, draft decisions prepared by the supervisory board are deemed adopted unless the Governing Council objects within 10 days in normal times or 2 days in stressful ones. A mediation panel and a steering committee would also be created. These structures seek to aid decision making and resolve disagreements, and to reinforce cooperation between the ECB and national authorities. However, it will also be important to ensure that the complexity of the setup does not undermine effective and prompt supervisory decision making.

Over time, some EU countries may want to be part of the banking union even if they do not join the euro area. A strong banking union that offers risk sharing (while avoiding the mutualization of legacy issues) and ensures least-cost bank resolution could be an attractive proposition. Moving supervision to the ECB could improve supervisory quality in some countries, reduce compliance costs for cross-border banks, limit scope for regulatory arbitrage, eliminate host-home coordination issues, and increase the congruence between the market for financial services and the underlying prudential framework. Once in place, a single resolution authority and common safety nets, with backstops, would provide further risk-sharing benefits. However, there are also drawbacks and complications, including the interaction of multiple central banks (with implications for the LOLR function and the conduct of macroprudential policies), difficulties in ensuring adequate participation of the “opt-ins” in SSM decisions, a loss of sovereignty, and potentially less flexibility to deal with country specificities. These costs are likely to be less for those whose currencies are pegged to the euro, have high levels of foreign currency liabilities, or have a sizable presence of euro area banks in their financial systems. If these members adopt the euro at the same time as they join the banking union, the benefit would likely outweigh the cost, just as it does for current euro area members.


The crisis has forced a reconsideration of the underlying institutions needed to sustain the euro as a common currency. The extension of the discussion on common fiscal frameworks and institutions to the banking sector is an important move. Banking union is obviously not a panacea, but it can be pivotal in fighting the current crisis by breaking the vicious circle between sovereign and bank costs and by fixing the broken transmission mechanism from ECB policy rates to final borrowing and lending rates across the full span of the euro area. A common supervisor, resolution mechanism, and safety net would also lay the foundation for long-term stability and reverse the fragmentation into subzones of greater or lesser confidence.

Many issues must be tackled, but critical design aspects must not be deferred, and strong euro area—wide bank supervision and safety net measures must be implemented quickly—not least because the cost of dwindling confidence accumulated silently in the massive payments imbalances brought on by the flight of deposits and capital across the euro area, which has been reflected in the rapid rise in the ECB’s TARGET2 balances. These potential costs can be reversed and minimized by early and credible action on banking union. Although speed is important, reformers will need to be mindful that wrong sequencing and a piecemeal approach could actually worsen outcomes.


    AgarwalSumitDavidLuccaAmitSeru and FrancescoTrebbi.2012. “Inconsistent Regulators: Evidence from Banking.Working Paper 17736 National Bureau of Economic ResearchCambridge, Massachusetts.

    • Search Google Scholar
    • Export Citation

    Council of the European Union. 2012. “Euro Area Summit Statement.June29.

    European Commission. 2012. “EU High-Level Expert Group on Reforming the Structure of the EU Banking Sector.Chaired by Erkki LiikanenBrusselsOctober2012.

    • Search Google Scholar
    • Export Citation

    Financial Stability Board. 2011. “Key Attributes of Effective Resolution Regimes for Financial Institutions.Basel.

    LaevenLuc and FabianValencia.2012. “Systemic Banking Crises Database: An Update.IMF Working Paper No. 12/163International Monetary FundWashington.

    • Search Google Scholar
    • Export Citation

This chapter is based on “A Banking Union for the Euro Area,” IMF Staff Discussion Note 13/01, 2013.

    Other Resources Citing This Publication