European Economy

Banks, Regulation, and the Real Sector

Limits on State-Funded Bailouts in the EU Bank Resolution Regime

by

In the post-crisis environment, the new European policy orthodoxy insists on avoiding state-funded bailouts of banks in distress under all but the most exacting circumstances. This is reflected in the two distinct but interrelated sets of norms governing bank resolution actions: The Commission’s norms on state aids in the banking sector as reflected in the Banking Communication of July 2013; and the new special resolution regime for credit institutions and investment firms adopted in May 2014 in the form of the Bank Recovery and Resolution Directive. The paper discusses the anti-bailout objective of the two frameworks, the way in which this is reflected in their operative provisions, and the degree to which the latter result in a truly binding regime, or admit exceptions and variations. It is shown that the overall effect of the provisions is to render outright bailouts almost impossible. Even when an intervention is permitted, this may take place only in prescribed forms and at a late stage within the resolution system’s financing cascade, which insists on substantial bail-in of ailing banks’ private claimholders, amounting to at least 8% of total liabilities, as a prior condition. The only exception is precautionary recapitalization; but this applies only to solvent institutions and cannot cover past losses. It may be wondered, however, whether a policy of strict insistence on bail-in in all cases of undercapitalization is wise. The problem has recently come to a head due to the troubles of the Italian banking system, with its huge pile of bad assets and numerous weak banks. The Italian banking system has a sufficient volume of bail-inable junior debt, thus making bail-in technically feasible. But at what cost?

1. From bailouts to bail-in: resolution without public financing as a key policy objective

In the aftermath of the Lehman Brothers debacle and the ensuing sudden freezing of interbank markets, the initial European response consisted in massive and across-the-board programmes of financial assistance to the banking systems of the Member States, in the form of blanket state guarantees, capital injections and, in many instances, the nationalizations of failing banks. The state support came with certain conditions, but these were not particularly stringent. Indeed, the European Commission’s DG Competition, in its role of final arbiter under the Treaty’s state-aid framework (TFEU, Arts. 107-109), gave its seal of approval to pre-notified national measures with flexibility and in record time.It did not take long, however, for the grave fiscal implications of the bailout packages to be felt. During the crisis, euro area governments utilized all the aforementioned forms of assistance. In the period 2008-14, total gross fiscal support to the financial sector amounted to 8% of the region’s GDP. By end 2014, amounts equivalent to 3.3% of GDP had been recovered; on the other hand, guarantees amounting to 2.7% of GDP remained outstanding and further potential losses could arise from impaired assets transferred from the banks to state-controlled asset management vehicles. All in all, the support resulted in a deterioration of the region’s overall budget balance of 1.8% of GDP and an increase of public debt by 4.8% of GDP. Critically, the scale and fiscal impact of the support diverged greatly across Member States (ECB, 2015a).Containing the exposure thus became a pressing policy priority. Combined with public indignation at the enormous size of the assistance offered to banks at a time of general economic hardship, this precipitated a global policy shift towards the principle that no banks should be considered “too big to fail” (“TBTF”) and that, more generally, the costs of failure should be primarily borne by banks’ own stakeholders (shareholders and creditors), with state financing becoming available only sparingly and as a complement to such burden-sharing by the stakeholders. The principle is reflected in the FSB principles for the resolution of global systemically important banks (FSB, 2014), as well in recent national legislation, both in the US and in European jurisdictions, setting up special legal frameworks for bank resolution.

Many aspects of the new resolution regimes draw their inspiration from earlier American bank insolvency law and practice. This, however, is not a case of simple legal transplantation or policy transfer. The scale of the recent troubles has led to a reprioritization of the objectives of bank crisis management, with fundamental implications for the overall policy approach and technical tools of resolution. Before the crisis, discussions of bank resolution and/or insolvency law revolved around certain special characteristics of banking institutions, which render the general system of insolvency proceedings inappropriate for the handling of bank failures, thus calling for a differentiated system of norms and techniques. The adoption of a special resolution regime for banks had been proposed by international standard-setters largely on this basis (BCBS, 2002; IMF and World Bank, 2009). In contrast, the recently adopted resolution laws are not merely intended to provide the legal preconditions for the orderly and expeditious implementation of bank resolution decisions, but further seek to determine their substantive content and final outcomes, by erecting legal barriers to the traditional tendency towards state-funded bailouts and the TBTF argument typically used to justify them.

Beyond their potentially ruinous fiscal consequences, bailouts create expectations regarding future state responses to financial troubles. Through this channel, the subsidization of bank stakeholders’ risk-taking by means of the externalization and absorption by the taxpaying public of the costs of insolvency exerts a very powerful distortive effect on ex ante incentives, entrenching moral hazard. This constitutes a major source of imbalances, risk and fragility in the banking system.

To relegate bailouts to history, the new approach to resolution entails the imposition of very strict limits to the financing of rescue operations with public funds and emphasizes the novel concept of “bail-in” (Hadjiemmanuil, 2015). This involves the absorption of past losses and/or the costs of recapitalization of weak banks by their own stakeholders, namely, their shareholders and creditors (junior creditors and subordinated bondholders and, potentially, senior bondholders and large depositors too, but not their smaller depositors). The move from bailouts to bail-in is supposed to mark a drastic departure from past practice, leading to the elimination of TBTF and the formation of a safe and stable financial sector (Huertas, 2014).

In the European context, two additional factors provide reasons to banish state-based bailouts. The first is the EU’s constitutive concern for competitive equality between business undertakings, and hence between the economies of the Member States, in the internal market, as reflected in the provisions of the TFEU on state aids, referred to above (TFEU, Arts. 107-109). These create a strong presumption against state aids in any sector, including banking (Commission Decision 95/547/EC, Crédit Lyonnais, OJ 1995 L 308/92). The second, which applies with particular force to the economies of the euro area, is the need to break the bank-sovereign “diabolic loop”, whereby the troubles of a national banking system can cause a fiscal crisis, and vice versa. More precisely, an attempt to support ailing banking systems with ample fiscal resources undermines the debt sustainability of fiscally weak Member States. This can precipitate or abet fiscal crises. At the same time, it raises doubts about the continuing availability of fiscal support for domestic banks. As a result, the latter’s access to cross-border wholesale funding is hampered, thus further increasing their funding needs and reliance on the state. Ireland and Spain provide conspicuous examples; but the mechanism has been in operation in all countries affected by the euro crisis (see also Navaretti, Calzolari and Pozzolo, 2016).

As a result, while at the height of the crisis the European institutions took a rather permissive stance on the issue of state support for banks, subsequently the European position became remarkably prohibitive. The current approach is incorporated in two distinct but interrelated sets of norms: on the one hand, the Commission’s norms on state aids in the banking sector as finally crystallized in the relevant communication of July 2013 (European Commission, 2013, or “Banking Communication”), which, compared to its initial policies of 2008-09, reflects a considerable hardening of the policy stance; and on the other hand, the new special resolution regime for credit institutions and investment firms proposed in 2012 and finally adopted in May 2014 in the form of the Bank Recovery and Resolution Directive (Directive 2014/59/EU, 2014 OJ L173/190, or “BRRD”). While the two regimes are adopted under different legal bases (TFEU, Art. 107 and TFEU, Art. 114, respectively) and have different legal force (“soft law” as against fully binding legal norms), they operate in tandem to erect barriers to further state-funded bailouts. In both cases, the European norms permit the provision of governmental assistance to ailing banks only in exceptional circumstances. They also insist on the principle of extensive burden-sharing by (certain) stakeholders (or bail-in) as a prerequisite (see Micossi, Bruzzone and Cassella, 2016).

The new legal situation raises for national decision-makers important new questions: How are the objectives of the two regimes reflected in their operative provisions? How effective and/or conclusive are the resulting constraints on national decision-making? In particular, what room do they leave for exceptions and variations? From another perspective, as the new approach has not yet proven itself in practice, one may ask, whether the overall policy is truly sound, and what, if any, are the alternatives. In the following paragraphs, an attempt will be made to sketch initial answers.

2. Forms of public financial assistance to ailing banks

A state may provide financial assistance to ailing banks in a number of ways. The response will depend on whether the problem is perceived to lie in the banks’ temporary illiquidity, due to their inability to hold on to their deposit base and/or refinance themselves in the interbank markets, or a more fundamental issue of actual or impending insolvency, as a result of operating losses and/or the deterioration of their asset portfolios. While the two aspects are closely interlinked and the situation will often be ambiguous, the authorities’ interpretation will determine the form of the intervention.In the former case, assistance will not necessarily take the form of a direct governmental intervention or rely on fiscal resources, because the central bank provides a ready alternative. Indeed, beyond its general refinancing interventions, which are classified as monetary operations and seek to accommodate the liquidity needs of the banking system as a whole, the central bank may also provide lending of last resort to banks on an individual basis. Both forms of central bank intervention have been very much in evidence in the euro area in recent years. The ECB has pushed back the limits of its monetary toolbox by launching a variety of novel programmes supporting the banking sector with refinancing of enormous proportions and for unusually long periods, while several national central banks have been allowed to extend further assistance to their domestic banks under the rubric of “emergency liquidity assistance” (“ELA”) (see Gortsos, 2015). Especially in the early phases of the crisis (2008–09), the governments of Member States have also engaged in direct interventions in support of the liquidity of their domestic banking systems, including through the provision of state guarantees to central banks for their refinancing exposures, the direct extension to banks of loans or temporary financing through special instruments, and the provision of guarantees of banks’ new liabilities.Where a bank is deemed to be undercapitalized or insolvent, the ECB or the national central banks will not provide fresh capital. A government, however, may be willing to restore the bank’s capital position with fiscal resources. This can be achieved through the injection of fresh capital (recapitalization), including in the context of the bank’s full nationalization. Alternatively, the government may improve the bank’s capital position (and simultaneously its liquidity) by purchasing its impaired assets at above-market prices or by providing guarantees over existing or, more likely, new assets. Tax privileges (such as deferred tax credits and deferred tax assets) may also be used to absorb or offset losses on assets and thus restore banks’ capital position, albeit not immediately. Finally, governments may provide inducements and sweeteners for “private” solutions, whereby non-state investors either purchase or recapitalize the ailing banks or buy portfolios of problem assets.

In all cases, the extension by national governments of financial support to ailing banks raises the question of compatibility with the Treaty norms on state aids. The same consideration applies when liquidity or capital support is provided by an emanation of the state, including its central bank, by state-owned banks or other enterprises, and even by institutions that, while being financed by the private sector, are under the state’s effective control. This consideration can bring within the scope of the provisions on state aids interventions by deposit insurance schemes and resolution funds, which raise their resources by levying contributions on banking institutions.

3. Public financial assistance to banks under the European state-aid framework

To be found compatible with the Treaty, state-aid measures which distort or threaten to distort competition must fall within one of the exceptions of TFEU, Art. 107(2)-(3). With regard to banks and banking systems, the potentially applicable exception will be that of Art. 107(3)(b), whereby state aid may be considered to be compatible with the internal market if it is necessary in order “to remedy a serious disturbance in the economy of a Member State”. Any national measure purporting to extend state support to individual enterprises or whole sectors on this ground will have to be notified to the Commission and approved by the latter. As already mentioned, since the beginning of the global financial crisis, the Commission has approved a great many national programmes for the support of individual banks and/or national banking systems. It has also set out its policy stance in a series of seven communications, which establish general criteria for the approval of state aids to the financial sector during the crisis. The Commission’s framework was explicitly designed as a temporary response to the crisis. Nonetheless, it continues to apply in revised form, on the ground that, even though the crisis has abated, “[t]he stress in financial markets and the risk of wider negative spill-over effects persist” and state interventions may still be needed to stabilize the banking sector (European Commission, 2013, para. 4 and 6).The Commission is thus willing to approve national support measures for reasons of financial stability. It should be noted, however, that the Commission’s understanding of the demands of financial stability cut both ways. On the one hand, financial stability as an overarching objective may justify a distressed bank’s or banking system’s access to state aid. On the other hand, the exact same objective requires that the state aid take place only at a late stage in a very strict financing cascade, be limited to the minimum necessary and be preceded by an appropriate contribution to the restructuring costs by the bank and its stakeholders out of their own resources (para. 8 and 15-20).In the initial phases of the crisis, the Commission required no more than a minimum degree of burden-sharing (that is, the absorption of past losses with available capital and the payment of an adequate remuneration to the state for its financing). In contrast, it did not demand any contribution by the banks’ creditors, due to a fear that this might precipitate runs and further destabilize the financial system. Now, however, the framework insists on full burden-sharing prior to the extension of any kind of restructuring aid (para. 16-19).

In situations where the problem is one of solvency and the state aid is aimed at covering an identified capital shortfall through recapitalizations or impaired asset measures (para. 28-55), the Commission demands, firstly, the utilization by the banks concerned of any feasible capital raising measures that might enable them to recover their viability or reduce to a minimum the necessary external support. Relevant capital raising measures include rights issues, voluntary conversions of subordinated debt instruments into equity, liability management exercises, sales of assets and portfolios, securitization of non-core portfolios, and employee earnings restrictions (para. 35-39). The Commission requires, secondly, burden-sharing by the shareholders, hybrid capital holders and subordinated creditors of the banks concerned, who must contribute to the maximum extent possible to reducing the capital shortfall by way of the write-down or conversion into equity of their respective claims, in reverse order of priority (para. 40-46). The Commission does not insist on a contribution from senior debtholders (such as senior bondholders and depositors); but following the coming into mandatory effect of the BRRD’s provisions on bail-in at the beginning of 2016, senior debtholders, including uninsured depositors, may now be brought within the burden-sharing cascade by virtue of these provisions (BRRD, Art. 43). It should be noted that any contribution by a deposit guarantee scheme to the costs of bank restructuring may also constitute state aid, on the basis that, while the scheme is funded with funds collected from the private sector, the use of the funds is imputable to the state (European Commission, 2013, para. 63). This will be a common occurrence in resolution proceedings pursuant to the BRRD, where the relevant deposit guarantee scheme will often be required to contribute to the financing of resolution as a least-cost alternative to the making of direct payments to covered depositors. The same consideration applies to interventions by a resolution fund (para. 64).

The Commission is also adamant that Member States must notify and seek its approval for financial measures such as guarantees on liabilities and liquidity support measures in support of solvent but illiquid banks (para. 56-61). With regard to this category of state aids, however, the Commission’s requirements do not focus on burden-sharing, but on the need to avoid open-ended interventions, which may generate undue fiscal risks and distortions of competition. In particular, any guarantees may only be granted for new issues of senior debt, thus excluding subordinated instruments; the debt instruments must be of short or medium duration; the state must receive adequate remuneration of its liquidity support, thus providing a disincentive for the banks’ continuing reliance on it; if the total liquidity support that a bank receives from the state exceeds 5% of its total liabilities or the sum of € 500 million, a restructuring plan must be submitted to the Commission; and if any bank causes the state guarantees to be called upon, a restructuring or wind-down plan must be submitted to the Commission (para. 59). Significantly, the banking sector’s refinancing by means of a central bank’s “ordinary” monetary operations, such as open market operations and standing facilities, is exempt from state aid controls. In contrast, the provision of ELA or any other form of individualized refinancing will fall within the concept of state aid, unless (a) it is given to a temporarily illiquid but solvent institution, (b) is fully secured by collateral, (c) is subject to a penal rate of interest, and (d) is extended at the national central bank’s discretion and is not backed up by a guarantee of the state (para. 62).

The Banking Communication has been challenged before the ECJ, but without success (Case C-526/14, Kotnik, judgment of 19 July 2016). The court found that the Treaty does not preclude burden-sharing by shareholders and holders of subordinated rights as a condition for the Commission’s approval of state aid to ailing banks. It further rejected the argument that burden-sharing violates the protection of legitimate expectations or the right of property, at least as long as the measures for converting hybrid capital and subordinate debt or writing down their principal do not exceed what is necessary to overcome the capital short-fall of the banks concerned. On the other hand, the court emphasised the non-binding legal nature of the communication. This is an instrument setting out the criteria used by the Commission when exercising its discretion under the Treaty provisions on state aids. Their publication sets a limit on the Commission’s discretion, which may not to depart from them without good justification, but does not impose independent obligations on the Member States. The latter retain the right to notify to the Commission state aid programmes incompatible with the Banking Communication, which the Commission is under a duty to examine, also taking into account any exceptional circumstances invoked by the Member States (Kotnik, para. 39-45).

The ECJ’s judgment points to the inherently malleable nature of the Commission’s framework. Even if the Banking Communication were a binding legal instrument, however, the situation would not be much different, since it contains numerous provisos, which could be used in appropriate cases to justify the provision of state assistance without extensive burden-sharing, potentially in forms equivalent to old-style bailouts. In particular, the Commission declares its readiness to take account of the macroeconomic environment, the specificities of the banks and Member State concerned, the presence of system-wide weaknesses in the domestic financial sector, the contribution of the sovereign crisis in the banks’ troubles, the feasibility of proposed burden-sharing measures, etc. (Banking Communication, para. 9-11). More directly, the Commission leaves open the possibility of an exception to the requirements of burden-sharing, including in the case of banks that fail to meet the minimum regulatory capital requirements, “where implementing such measures would endanger financial stability or lead to disproportionate results” (para. 45).

It is thus clear that the Banking Communication, for all its robust language and strong preference for burden-sharing through bail-in over state funding, is framed in terms sufficiently flexible for enabling the approval of almost every conceivable solution by way of “exception”. What must be doubted, however, is the actual willingness of the Commission to soften its stance. At present, all indications suggest that, even in the face of a simmering crisis with potentially highly detrimental consequences, such as that affecting the Italian banking sector, the Commission remains unperturbed and unwilling to budge. With the final entry into full effect of the BRRD’s provisions on burden-sharing on 1 January 2016, the Commission has found additional reasons for doing so.

4. The no-bailout objective in the resolution framework of the BRRD

The no-bailout objective is unambiguously set out as a tenet of European law both in the preamble and in the operative part of the BRRD, alongside the objectives of orderly and cost-effective resolution, the continuation of critical functions of failed banks, the preservation of systemic stability and the avoidance of contagion, and the protection of depositors and clients’ assets (BRRD, rec. (1), (5) and (45) and Art. 31(2)). The system is specifically intended “to obviate the need for bailouts using taxpayers’ money to the greatest extent possible” (BRRD, rec. (1)).At a rhetorical level, at least, the BRRD insists that, to avoid moral hazard, any ailing institution should be preferably restored to soundness with private resources at a pre-resolution stage; otherwise, it should exit the market and be placed in liquidation (rec. (45)-(46)). On this theory, resolution as a quasi-insolvency process aimed at the restructuring of the bank with external assistance is permissible only as an “exception”, justified by the fact that liquidation under normal insolvency proceedings may in certain, but definitely not all, cases threaten financial stability, disrupt the provision of financial services to the real economy, and impede the protection of depositors. These considerations of public interest may justify the continuation of all or part of the bank’s activities within the framework of resolution (rec. (45)). Even then, however, the resolution should be based on private sources of financing and avoid in all, but the most dramatic, circumstances access to the public purse. In practice, of course, a finding that resolution is necessary may turn out to be the rule rather than the exception (see Hadjiemmanuil, 2014); and the operative provisions of the BRRD, despite the forbidding language, contain ample room for discretion and ad hoc interpretations for such a development to be possible. Nonetheless, there can be no doubt that the directive’s norms are specifically designed to discourage direct access to state-funded bailouts, relying instead on bail-in, followed by external financing raised through levies on the banking industry, in the form of the pre-funded deposit guarantee schemes operating pursuant to the Deposit Guarantee Schemes Directive (Directive 2014/49/EU, OJ 2014 L173/149) and the BRRD’s own “financing arrangements” or resolution funds (BRRD, Arts. 99-107).To attain its mixed objectives, the BRRD specifies a common administrative and procedural model for bank-failure-related decision-making for all Member States, common triggers and conditions for the activation of resolution actions, and a set of general restructuring techniques, or “resolution tools” (BRRD, Art. 37). The resolution tools are designed to enable the failed bank’s survival and restoration to solvency under new ownership or, at least, the avoidance of piecemeal liquidation of the existing legal entity and the continuation of whole or part of its business activities through the transferal of portfolios of assets and liabilities to a successor legal entity. Three distinct tools serve this purpose: the sale of business tool (BRRD, Art. 38-39), the bridge institution tool (BRRD, Art. 40-41), and the asset separation tool (BRRD, Art. 42).

Evidently, the continuation of a failed bank’s operations and, if the bank is kept alive as a legal entity, the successful completion of its restructuring will typically depend on filling a funding gap. New financing may be needed either to provide the liquidity which is indispensable for operational continuity rests and/or to bring the bank back to acceptable levels of capitalization. The BRRD contains very detailed provisions on resolution financing, in an attempt to define a prescriptive financing cascade, consistent with its no-bailout policy (see Hadjiemmanuil, 2016). To maintain an ailing bank as a going concern, it will be necessary to absorb past losses and recapitalize the institution to the point where it meets the continuing requirements for authorization. In a nutshell, if new willing investors are not forthcoming, the BRRD seeks to achieve this result initially through the mandatory write down of the bank’s regulatory capital instruments or, when these are in debt or hybrid forms, their conversion into equity (Arts. 59-62). This may happen either at the pre-resolution stage or in the context of resolution (Art. 59(1)). Within resolution, beyond the aforementioned write down of regulatory capital instruments, the bank’s restructuring is financed by converting into capital or writing down the claims of non-exempted (“eligible”) liability holders by way of bail-in. This simultaneously reduces liabilities and increases the bank’s capital resources, thus pushing it back to solvency. Bail-in may be implemented in a structured manner as the fourth and final resolution tool (Arts. 43-55). The expectation is that the bail-in tool will be used as a matter of course when the funding gap cannot be covered by writing down the capital instruments. Bail-in may be complemented by a contribution by the relevant deposit guarantee scheme. The latter’s participation to such open-bank resolution financing will, however, be limited to the amounts that it would be required to pay out to covered depositors, if the bank undergoing resolution had been would up under normal insolvency proceedings (Art. 109). If the contributions of private parties are not enough, the appropriate national resolution fund or, for the Member States of the Banking Union, the SRM’s Single Resolution Fund, can contribute to the financing of resolution. This, however, will only be possible after a contribution of no less than 8% of total liabilities, including own funds, has been made by stakeholders by way of bail-in. In addition, the resolution funds’ intervention is limited to medium term-financing; and it cannot exceed 5% of total liabilities (BRRD, Arts. 44(4)-(6), (8); and Regulation No 806/2014, OJ 2014 L225/1, Art. 27(6)-(7)).

Recapitalization with public funds (whether national or pan-European) may be considered only if, after all the aforementioned sources of resolution financing have been exhausted (either because they were depleted, or because the limits on their contribution were reached), a bank remains undercapitalized, but its continuation as a going concern appears imperative for reasons of systemic stability (BRRD, Arts. 37(10)(a))­. While the resolution authorities have discretion to select the most appropriate method of resolution and to apply any of the resolution tools set out, the BRRD does not afford discretion as to the application of the burden-sharing cascade. This is also true of resolution actions under the Banking Union’s SRM. Accordingly, assuming that the legal prescriptions will be applied faithfully, including at times of actual crisis or alleged distress, the cascade shifts the bulk of the burden from the taxpayer to the banks themselves, along with their investors and creditors.

5. Need for state support under the BRRD

A bank may be the recipient of financial assistance from the state without being insolvent or even illiquid. In the BRRD’s scheme, however, there is an almost necessary link between the need for state aid and financial collapse. For this reason, the need for state support is turned into a trigger for a bank’s placement in resolution or even liquidation.Under the BRRD, an institution may be placed in resolution if its supervisor (or the relevant resolution authority) determines that it is “failing or likely to fail” (BRRD, Art. 32(1)-(2); EBA, 2015). Whether this is the case, is established by reference to four alternative triggers. Alongside the two classic tests of insolvency (namely, balance-sheet insolvency and inability to repay debts and other liabilities as they fall due), the triggers include: a breach of regulatory requirements sufficiently serious to justify the withdrawal of the bank’s authorisation; and the bank’s need for “extraordinary public financial support” (BRRD, Art. 32(4)). This description, however, is used in the BRRD to encompass any state aid given to banks for the purpose of preserving or restoring their viability, liquidity or solvency, as well as equivalent forms of financial support extended by public bodies operating at the supranational level, such as the euro area’s Single Resolution Fund and the ESM (Art. 2(1), point (28)).

If the supervisor has determined, based on any of the four triggers, that a bank is failing or likely to fail, the bank will be placed in resolution, provided that two further conditions are satisfied: that there is no reasonable prospect that it will be restored to health with private sector measures or supervisory actions, such as early intervention measures or the write down or conversion of capital instruments; and that use of the resolution tools is “necessary in the public interest” (Art. 31(1)). Interestingly, in this context the public interest is equated with the achievement of one or more of the objectives of the BRRD’s resolution regime (Art. 32(5)), which, in turn, specifically include the protection of public funds “by minimising reliance on extraordinary public financial support” (Art. 31(2)). A failing or likely to fail bank that does not meet these further conditions, must be wound up under normal insolvency proceedings.

The BRRD contains three exceptions to the rule that the need for extraordinary public financial support establishes that the bank is failing or likely to fail. Specifically, resolution will not be triggered if, in order to remedy serious disturbance in the national economy and preserve financial stability, the state provides support in one of the following forms:

(a) guarantees to back liquidity facilities provided by the central bank;

(b) guarantees of newly issued liabilities; and

(c) injections of own funds or purchases of capital instruments at prices and on terms that do not confer an advantage upon the credit institution (so-called “precautionary recapitalization”).

To fall within the exception, in all three cases the support must meet certain conditions (Art. 32(4), second para.):

  • The support must be confined to solvent institutions.
  • It must be of a precautionary and temporary nature.
  • It must be proportionate to the consequences of the economic disturbance providing its justification; and, last but not least,
  • It must not be used to offset losses that the recipient banks have already incurred or are likely to incur in the near future.

Two further conditions apply in the case of precautionary recapitalization:

  • The support must be extended at prices and on terms that do not confer an advantage upon the institution (Art. 32(4)(d)(iii)).
  • The support must be limited to injections necessary to address capital shortfalls identified based on supranational or national stress tests, asset quality reviews or equivalent exercises conducted by the ECB, the EBA or national authorities (BRRD, Art. 32(4), third para.; EBA, 2014).

Especially with regard to precautionary recapitalization, these conditions are highly restrictive and limit very considerably the applicability and usefulness of the exception

6. Government financial stabilization tools

The BRRD allows a Member State to contribute to the recapitalization of a bank which has been placed in resolution, but this may only occur in specific ways and subject to strict conditions under the rubric of “government financial stabilization tools” (“GFSTs”) (BRRD, Arts. 37(10) and 56-58). The concept includes two more specific forms of recapitalization, namely: the “public equity support tool” (Art. 57), which involves injections of capital by the state in exchange for equity and other instruments included in the calculation of own funds pursuant to the Capital Requirements Regulation (Regulation No 575/2013, OJ 2013 L176/1); and the “temporary public ownership tool”, which amounts to full nationalization of the bank (BRRD, Art. 58).GFSTs form part of the resolution process as an alternative financing source for the implementation of the resolution tools selected, at least in theory, by the resolution authority (Arts. 37(10) and 56(1)). Moreover, they are only applicable when the resolution seeks to preserve the bank as a going concern. They are, accordingly, incompatible with the transfer of the bank’s operations to a new entity, since this would lead to the old entity’s dissolution (but not with the transfer of the bank’s ownership to a new owner, including a bridge bank, which preserves the old legal entity). A government retains full discretion on whether to participate in resolution by way of a GFST, because a Member State may not be forced to finance resolution with fiscal resources (rec. (76)). Furthermore, the implementation of the GFST takes place in the hands of the government itself (most probably through its Ministry of Finance), albeit in close cooperation with the resolution authority (Art. 56(1)).The national government’s discretion, however, only goes in the direction of refusing assistance. In contrast, a government eager to bail out an ailing bank may find it impossible or unappealing to rely on the BRRD’s provisions on GFSTs for this purpose, due to the strict and inflexible conditions.

GFSTs have been included in the resolution framework in order to address the “very extraordinary situation of a systemic crisis” (Art. 37(10)). But even in this context of severe and widespread distress, they are only available as a “last resort” – that is, in principle when other resolution solutions, including bail-in, have been “assessed and exploited to the maximum extent practicable whilst maintaining financial stability” (rec. (8) and Arts. 37(10) and 56(3)). This determination, however, is not left to the competent ministry or government alone, but also involves the resolution authority and, possibly, the competent supervisory authority and the central bank too (Art. 56(3)-(4)). This is particularly important in the case of Member States participating in the euro area’s Banking Union, whose “significant” credit institutions are supervised by the ECB and resolved under the control of the SRB, thus bringing within the picture supranational decision-makers.

In addition, utilization of the GFSTs is only permissible after a bank’s own stakeholders have contributed in the absorption of losses and the recapitalisation effort through the write-down or conversion of capital instruments and bail-inable liabilities to an amount equaling at least 8% of the total liabilities, including own funds (Art. 37(10)(a)). In this sense, a government cannot use GFSTs as a substitute for bail-in, but only as a complement to it.

Last but not least, the resort to GFTSs is subject to the Commission’s prior approval under the state aid framework (Art. 37(10)(b)). In this context, in addition to other parameters, the Commission will assess independently both whether the proposed intervention complies with the condition of prior burden-sharing of 8% and the existence of a situation of systemic crisis (rec. (57)). This decreases significantly the likelihood that the intervention will be allowed to take place, as well as the discretion of the national government as to its form and content

7. Range of options regarding the provision of public support under the BRRD

Leaving aside the contributions made by resolution funds to the financing of resolution in the context of bail-in, the overall structure of the provisions opens up the following alternatives with regard to the permissible types of public financial support to banks in a state of distress and the conditions for their provision:

  • A central bank may extend liquidity support on an individualized basis by way of ELA or in a functionally equivalent form. Unless the conditions for exemption set out in the Banking Communication (para. 62) are met, this would fall within the concept of state aid and, consequently, within the definition of extraordinary public financial support in the BRRD (Art. 2(1)(28). However, the assumption is that this type of assistance is permissible in principle, and may not trigger resolution. Indeed, the BRRD includes a separate definition of ELA, in a manner that could provide an argument to the effect that the concept does not overlap with that of extraordinary public financial support, but sits alongside it (Art. 2(1)(29). Moreover, the preamble mentions that resort to ELA does not demonstrate per se that a bank is, or will be in the near future, unable to pay its liabilities as they fall due (rec. (41)); and one assumes that, since the provision of state guarantees in support of ELA is exempt, the same applies to the ELA itself.
  • A government may also provide liquidity support. For the recipient banks not to come within the definition of “failing or likely to fail”, however, this would have to take one of the two forms mentioned explicitly in the BRRD (guarantees to central banks and guarantees of new liabilities) and meet the conditions mentioned above (Art 32(4)(d)(i)-(ii)). This form of assistance counts as extraordinary public financial support, and will accordingly trigger a requirement to write down or convert capital instruments to any necessary extent (Art. 59(3)(e)). In practice, this will not be a serious problem, since the recipient banks are supposed to be solvent and there is no capital gap to be filled.
  • Conceivably, governmental liquidity support and/or support relating to the banks’ credit provision or impaired assets may also be extended in situations where the banks are undercapitalized. In this case, even though the institutions may retain positive net worth, there will fall within the definition of “failing or likely to fail”. As long, however, as their viability appears likely to be restored with private measures or through supervisory actions within a reasonable timeframe, the other conditions for resolution will not be met (Art. 32(1)(b)). In this case, too, the support will trigger the requirement to write down or convert capital instruments.
  • The government may seek to inject capital in an undercapitalized bank without thereby triggering the resolution process. This would appear possible only by way of precautionary recapitalization pursuant to the exception mentioned above (Art. 32(4)(d)(iii)). In any other case, the bank would be failing or likely to fail and self-evidently unable to correct the situation with private sector measures or recovery actions (otherwise, it would not have tried to gain access to public funding), thus requiring resolution, if not liquidation. As already noted, the conditions for recapitalization under this heading are very restrictive. Where, however, the conditions are satisfied, the bank can be recapitalized with public funds without first resorting to write down or conversion of existing capital instruments (Art. 59(3)(e)).
  • The government may seek to contribute to the recapitalization of a failed bank within the resolution process. As we have seen, the BRRD specifically provides for this possibility (Arts. 37(10) and 56-58). Once more, however, the conditions are tough and may render this possibility irrelevant or unappealing in the eyes of the national government.
  • In the Banking Union, it is also possible for the ESM to extend supranational public assistance directly to systemically important banks that are unable to meet their capital requirements, by activating its “direct recapitalization instrument (“DRI”), which can utilize resources of up to €60 billion. The DRI may be used if the relevant national government is incapable to undertake the recapitalization of domestic systemic banks at its own account and risk without thereby significantly endangering its fiscal sustainability (ESM, 2014). The instrument is not available for precautionary recapitalizations, but only for resolution-related ones; it can be used only in the very special case where the provision by the ESM of a loan to the government to enable the implementation of GFSTs at the national level (“indirect recapitalization instrument”) (ESM Treaty, Art. 15) is not advisable; and even then, it requires burden-sharing by the national government. Under present circumstances, resort to the DRI would appear to be highly unlikely, due to its very specific conditions as well as the need for unanimous approval at the level of the ESM’s Board of Governors.

It should be noted that, whether they take place within or outside the resolution process, all these forms of state aid require notification and approval from the Commission (Arts. 32(4), second para., and 34(3)).

It can be easily seen that the overall effect of the provisions is to render outright bailouts, without extensive private burden-sharing, almost impossible. Even when an intervention is permitted, this may take place only as a GFST at a late stage within the resolution system’s financing cascade, after substantial bail-in of banks’ private claimholders, amounting to at least 8% of total liabilities. Moreover, the forms of permissible interventions are limited to specified types of capital injections. In this manner, the BRRD effectively relegates public support to a supporting role within the resolution framework’s financing cascade and sets exceptionally strict limits to pre-resolution interventions.

The only exception to burden-sharing is precautionary recapitalization. Even in this case, the Commission’s approval could, conceivably, be contingent on capital raising measures, including voluntary, or even mandatory, write down or conversion of capital instruments and bail-in of junior liabilities, as required by the general policies set out in the Banking Communication. Admittedly, the exception of para. 45 of that instrument, whereby burden-sharing could be excluded if it would “endanger financial stability or lead to disproportionate results”, could be used to avoid this result. Some authors consider that a flexible interpretation of the provisions, always in the light of TFEU, Art. 107(3)(b), which allows state aids when necessary to remedy a serious disturbance in a Member State’s economy, could justify extensive precautionary recapitalization of weak banks without bail-in if in a particular Member State the banking system is extensively undercapitalized and the private sources of capital cannot remedy the situation (Micossi, Bruzzone and Cassella, 2016). Nonetheless, while the BRRD links specifically the option of precautionary recapitalization to the presence of economy-wide disturbances and systemic problems, its specific terms make it largely inappropriate for the most vulnerable banks. The exacting conditions for precautionary recapitalization (requirement of solvency, provision of the support at competitive market prices, inability to use the support to offset losses, dependence on the existence and outcome of a stress test or similar exercise, often specified at the supranational level) suggest that it cannot be used for the restoration of banks that are already weakened by losses or carry substantial amounts of NPLs. It is precisely these banks, however, that are likely to initiate contagion. In this sense, precautionary recapitalization cannot be credibly relied upon for the repair of a distressed national banking system as a whole, but only for the creation of a second line of defence in favour of stronger banks in the system. This may contain the troubles, but not prevent them altogether.

The BRRD contains numerous ambiguous provisions, which confer very substantial discretion on the supervisory and resolution authorities with regard to determining the point of “failure”, as well as to the allocation of resolution costs. Indeed, the supervisory and macroprudential authorities have considerable room for varying the capital requirements of credit institutions. The time allowed to a weak bank to return to full solvency is also discretionary. The same applies to the requirements and timeframe for repairs to portfolios and the management of NPLs. The triggers for resolution depend on supervisory judgements. Within resolution, there is wide room for the resolution authorities, always subject to the Commission’s approval, to allow discretionary exemptions from bail-in or partial bail-in, thus pushing the resolution costs further down the BRRD’s financing cascade (Art. 43(3)-(12)). All these discretionary judgements –which in the Banking Union typically involve supranational decision-makers– may serve to alleviate the pressure on national banking systems; but they may also aggravate it! The one area, where the room for discretion is very limited, is precisely that of public support

8. Economic and political limits of the no-bailout policy

It may be wondered, whether a policy of strict insistence on bail-in in all cases where a bank’s weakness is due to accumulated losses, is wise. Resolution, as distinct from liquidation, is supposed to be justified on grounds of systemic stability. In contrast, in conditions of economic distress and system-wide banking weakness bail-in as the preferred and essentially mandatory resolution tool can aggravate the situation (De Grauwe, 2013; Persaud, 2014).The bail-in tool is designed to achieve the restructuring of individual insolvent banks, thus preventing the knock-on effects of outright collapse, without resorting to the taxpayer. The direct transmission of losses, however, is not the sole form of contagion, nor the most important one. Contagion may also occur through market reactions to a particular incident and the way in which this was handled by the authorities. In a weak environment and in a context of widespread distress, bail-in with regard to one bank may lead other banks’ claimholders to reappraise their position, precipitating an across-the-board flight to quality. This will increase the cost of refinancing, potentially to prohibitive levels, precisely at the point when the banking sector as a whole is striving to raise additional funds through the issuance of capital or debt instruments. The existence of tight regulatory deadlines for related improvements will make things worse.The problem has recently come to a head due to the troubles of the Italian banking system, with its huge pile of bad assets and numerous weak banks. The Italian banking system has a sufficient volume of bail-inable junior debt, thus making bail-in technically feasible. But at what cost? Given the circumstances, the alternative of a public rescue intervention can be credibly supported both on economic and political grounds, since the European project can hardly afford another major crisis or the disaffection of one more country.

Many remain unconcerned. In a recent public intervention, a group of prominent European economists maintain that, with regard to the euro area’s financial sector, the fundamental architecture to ensure resiliency is already in place, thanks to Banking Union’s centralization of regulation and supervision and the bail-in-based SRM (Resiliency Authors, 2016). In their view, the present priority is “to make sure that the rules in place can be enforced. Italy provides two cases in point. First, non-performing loans have steadily increased and are carried on the books at prices substantially above market prices. Second, the Italian government has proven very reluctant to apply the bail-in rules. The credibility of the rules is at stake. Either they have to be applied, or credibly modified.” Even though the question of a “credible modification” of existing resolution norms is left open, there is little doubt that here the emphasis is placed on strict enforcement of the bail-in principle. This would, in fact, appear to be the preponderant view amongst academics and policy-makers, with the exception of those living in the Member States most directly affected by banking troubles.

 

A critical assumption behind the dominant view is that bail-in, not only eliminates moral hazard, but also ensures equality of treatment across the Union. If, however, the practical effects of bail-in are not truly symmetrical, its automatic application, without regard to the specific circumstances, would appear to be unreasonable and disproportionate. A common approach applied to fundamentally different situations and at the back of very dissimilar prior paths, is likely to increase divergences and inequalities across countries. The bail-in norms of the BRRD have come in force at a moment when the monetary and banking landscape of the EU, and the euro area in particular, is deeply fragmented, as a result of the sovereign debt crisis. Certain economies are hampered with acute macroeconomic problems and increased costs of credit, which fuel the generation of NPLs and curtail the domestic banks profitability and access to new funds. In other economies, like Germany, the banking system has been restored to health with ample public assistance in the immediate preceding period (Binder, 2016); and even now, large segments of the sector (namely, the public savings banks and the cooperative banks) can rely on IPS-style arrangements to avoid resolution. Moreover, the lack of a credible and neutral single fiscal backstop at the supranational level means that the financial risks faced by investors are not the same in all cases. The Commission’s intransigence in the Italian case can further widen the wedge between the national banking systems’ financing conditions. This, despite the fact that the establishment of a common backstop and equal monetary and financial conditions constituted the Banking Union’s raison d’être in the first place.

A strict insistence on bail-in may be counterproductive even on the resolution framework’s own terms. Cleansing banking systems from the huge pile of legacy NPLs is an urgent priority, since it is a prerequisite for the normalization of the credit intermediation function, especially in the problem economies of the euro area’s periphery. The question is, whether the domestic banking systems can be relied upon under all conceivable conditions to absorb losses and simultaneously restore their capital ratios to currently prescribed levels without public funding. In the case of Italy, divesting up front and at current prices the stock of NPLs and passing the full cost of losses to banks’ primarily domestic stakeholders could bring the banking sector to ruin and cause irreparable damage to the economy. Such a policy would push large swaths of the banking system into simultaneous resolution. However, the estimated capital shortfall may be exaggerated. To start with, the fact that NPLs appear in banks’ balance sheets at above current market prices (say, 30 to 35% as against 20%) may not be a function of misrepresentation of the portfolios’ true run-off value, but on the fact that secondary markets for distressed debt remain incomplete and there is limited experience with relevant transactions. Moreover, expected recovery rates depend on the available mechanisms of judicial debt enforcement. In this regard, it should be taken into account that the more expeditious and creditor-friendly procedural rules recently introduced in countries with an abysmal record with regard to the judicial enforcement of claims, such as Italy or Greece, have not yet be given the time to work. More generally, the pricing of NPLs critically depends on the state of the economy, the rate of supply of NPLs to the secondary market and the international appetite for assets in the relevant economies (which is, in turn, affected by the possibility of a banking crisis). An uncompromising “liquidationist” approach is almost certain to have negative repercussions on all these fronts, potentially fueling a debt-deflation spiral, especially in economies within the euro area, which are deprived of country-specific monetary policy tools.

In these circumstances, a pre-resolution action plan, involving the management of impaired assets and recapitalization of weak banks with state aid but without extensive burden-sharing, may constitute the most credible and reasonable response. Such an approach should be accompanied by an appropriate programme, with clear objectives, milestones and responsibilities, for the restructuring and modernization of the Italian banking sector within a reasonable timeframe. This could be achieved in the form of Commission-mandated conditions for the programme’s approval. An approach of this type could yield much better results than resolution – especially since the latter is likely to increase the surviving banks’ cost of funding, but also to require significant public funding, either in the form of GFTSs or, in the peculiar Italian case, where large part of the banks’ subordinated debt is held by domestic non-professional individuals, by way of compensation payments to the latter. In this case, then, resolution and its financing cascade may be both suboptimal as a means of delivering systemic stability and unable to fully protect taxpayers.

Admittedly, this may not be legally simple. While the Commission has the discretion to depart from the general policies of the Banking Communication, the BRRD presents a more significant hurdle. As we have seen, precautionary recapitalization is unavailable to cover past or expected losses, cannot be applied to undercapitalized banks (which should be the first to receive restructuring assistance) and may only take the form of an injection of own funds or purchase of capital instruments at realistic asset prices, but not that of NPL-related transactions, especially if these include an element of subsidization. The separation and absorption of bad or doubtful assets is envisaged exclusively within the resolution framework.

To avoid these strictures and alleviate the problem, the Italian government has encouraged the formation of Atlante, a private-sector backstop fund without fiscal support, which has already been put in use to bail out two small banks and to purchase portfolios of NPLs. This type of intervention does not amount to state aid, nor does it trigger resolution. However, the sums involved are not sufficient to meet the needs of a large bank, much less of the whole banking system. If, however, the state had participated in its financing (either directly or even through a state-owned bank or commercial entity), this would cease to be a private venture, with the consequence that its interventions would trigger resolution and burden-sharing (unless, of course, the distressed assets were purchased at no more than their market value). Accordingly, this approach can provide partial relief, but not a comprehensive and lasting solution to the wider problem. At the end of the day, in the face of the present predicament, the best option remains a judicious application of the provisions on bail-in and state aids that would allow the implementation of a balanced plan of state-supported recapitalization and impaired asset measures, with minimal burden-sharing, but with meaningful restructuring commitments on Italy’s part.

References

BCBS (2002). Supervisory Guidance on Dealing with Weak Banks: Report of the Task Force on Dealing with Weak Banks. March. Available at: www.bis.org/publ/bcbs88.pdf.

Binder, J.-H. (2016). Systemkrisenbewältigung durch Bankenabwicklung? Börsen-Zeitung, 12 July, 4.

De Grauwe, P. (2013). The New Bail-In Doctrine: A Recipe for Banking Crises and Depression in the Eurozone. CEPS Commentary, CEPS, Brussels, 5 April.

EBA (2014). Guidelines on the types of tests, reviews or exercises that may lead to support measures under Article 32(4)(d)(iii) of the Bank Recovery and Resolution Directive. EBA/GL/2014/09, 22 September.

EBA (2015). Guidelines on the interpretation of the different circumstances when an institution shall be considered as failing or likely to fail under Article 32(6) of Directive 2014/59/EU. EBA/GL/2015/07, 6 August.

ECB (2015a). The Fiscal Impact of Financial Sector Support During the Crisis. ECB Economic Bulletin No 6/2015, 74-87.

ECB (2015b). ECB Legal Conference 2015: From Monetary Union to Banking Union, on the Way to Capital Markets Union, New Opportunities for European Integration. European Central Bank, Frankfurt am Main. Available at: www.ecb.europa.eu/pub/pdf/other/frommonetaryuniontobankingunion201512.en.pdf.

ESM (2014). Guideline on Financial Assistance for the Direct Recapitalisation of Institutions. 8 December.

European Commission (2013). Communication from the Commission on the application, from 1 August 2013, of State aid rules to support measures in favour of banks in the context of the financial crisis (“Banking Communication”), OJ 2013 C216/1.

FSB (2014). Key Attributes of Effective Resolution Regimes for Financial Institutions. 15 October. Available at: www.financialstabilityboard.org/wp-content/uploads/r_141015.pdf.

Gortsos, C.V. (2015). Last-Resort Lending to Solvent Credit Institutions in the Euro Area: A Detailed Presentation of the Emergency Liquidity Assistance (ELA) Mechanism. In: ECB (2015b), 53–76.

Hadjiemmanuil, C. (2014). Special Resolution Regimes for Banking Institutions: Objectives and Limitations, in: Ringe, W.-G., and Huber, P.M. (Eds.), Legal Challenges in the Global Financial Crisis: Bail-Outs, the Euro and Regulation, 209-235. Hart Publishing, Oxford & Portland, Oregon.

Hadjiemmanuil, C. (2015). Bank Stakeholders’ Mandatory Contribution to Resolution Financing: Principle and Ambiguities of Bail-In. In: ECB (2015b), 225–248.

Hadjiemmanuil, C. (2016). Bank Resolution Financing in the Banking Union, in Binder J.-H., and Singh D. (Eds.), Bank Resolution: The European Regime, 177-209. Oxford University Press, Oxford.

Huertas, T.F. (2014). Safe to Fail: How Resolution Will Revolutionise Banking. Palgrave Macmillan, Basingstoke, Hampshire.

IMF and World Bank (2009). An Overview of the Legal, Institutional, and Regulatory Framework for Bank Insolvency. April 17.

Micossi, S., Bruzzone, G., and Cassella, M. (2016). Fine-Tuning the Use of Bail-in to Promote a Stronger EU Financial System. CEPS Special Report No 136, CEPS, Brussels, April.

Navaretti, G.B., Calzolari, G., and Pozzolo, A.F. (2016). Diabolic Loop or Incomplete Union? Sovereign and Banking Risk. European Economy: Banks, Regulation, and the Real Sector No 2016.1, 9-28.

Persaud, A.D. (2014). Why Bail-In Securities Are Fool’s Gold. Policy Brief No PB14-23, Peterson Institute for International Economics, Washington, D.C., November.

Resiliency Authors (Baldwin, R., et al) (2016). Making the Eurozone More Resilient: What Is Needed Now and What Can Wait? VoxEU.org, 25 June. Available at: http://voxeu.org/article/making-eurozone-more-resilient-what-needed-now-and-what-can-wait.