It wasn't supposed to be like this. EU authorities introduced resolution legislation and tightened state aid rules following the financial crisis to ensure creditors, not taxpayers, incur most of the costs of bank failures. Instead, certain governments have continued to support troubled banks, sometimes resorting to creative methods to adhere to the letter of the law, but perhaps not policymakers' original intent.

Continued bank bail-outs cast doubt over some European governments' commitment to their own reform program. S&P Global Ratings thinks that these actions have stretched the credibility of the resolution agenda, but not yet to breaking point. For as long as governmental and regulatory decisions remain unpredictable, market participants will inevitably lack confidence in the effectiveness of the resolution process, and the scope and timing of any government support.

We see several reasons why various European governments and regulators currently shy away from using their enhanced powers to exert market discipline over failing banks. Much of the sector lacks a sufficient layer of bail-inable debt, which is the mechanism to transfer losses to creditors. Some countries may also be nervous about using largely untested resolution powers while confidence in parts of the banking system remains fragile. Within the eurozone, a fragmented decision-making framework and inconsistent national insolvency laws provide scope for governments to assume losses in the interest of short-term financial stability.

We removed government support from our ratings on most Western European private sector commercial banks in 2015, reflecting the implementation of resolution powers. We believed this legislation was a sound basis for the creation of effective resolution regimes, and made government support less predictable than before. We maintain these views today. Some countries--most notably Switzerland and the U.K.--were strong advocates of resolution from the outset and have made significant progress in bringing it to life. For the remainder, we think government support may diminish further as their resolution frameworks mature, including procedural reforms to enhance consistency, and more banks become truly resolvable. This process will take several more years, and we see the 2024 deadline for eurozone banks to complete their bail-in buffers as an important milestone. If, on the other hand, some jurisdictions establish bail-outs as a normal and predictable response to bank failures, we could reintroduce government support to ratings on systemically important banks in those countries.

Governments Retain Options To Support Troubled Banks

Some EU governments have exploited every available option within the Bank Recovery and Resolution Directive (BRRD) and state aid rulebook to channel taxpayer funds to struggling banks (see chart 1). Although the legal pathways differed, the consistent outcome in these cases has been significant government support for depositors, senior bondholders, and other unsubordinated creditors.

In some cases, governments have supported banks using mechanisms purposely included in the BRRD and state aid framework to enable a degree of flexibility in a stress scenario. For example, state-funded recapitalizations of Germany's NordLB and Portugal's Caixa Geral de Depositos were permitted under state aid rules because the European Commission (EC) ruled that the funds were provided on market terms that private investors would accept. These two banks were already state-controlled and the capital injections were led by the existing public sector shareholders. The support for Italy's Monte dei Paschi and Banca Carige was provided under a clause in the BRRD that permits precautionary, temporary government intervention.

However, we question whether governments and regulators have taken a consistent stance in applying these sections of the BRRD and state aid frameworks. For example, it is debatable whether support for Monte dei Paschi and Carige was truly precautionary and temporary, or rather recognition of existing balance sheet stress. In the case of NordLB, the EC regarded the bank's business plan as achievable and calculated a 8.9% internal rate of return on invested capital in 2020-2024, consistent with the estimated 8%-10% cost of equity. In our view, this approach to evaluating state aid is relatively mechanistic and private investors faced with the same investment opportunity might give more weight to macroeconomic uncertainties and the execution risks inherent in a complex bank restructuring program.

In other cases of bank distress, governments discovered creative, unintended routes to provide support within the constraints of the BRRD and state aid rules. The primary examples to date are two Italian regional banks, Banca Popolare di Vicenza and Veneto Banca (see “To Fail Or Not To Fail: The Point Of Nonviability Is Unclear For European Banks,” published on May 31, 2017). A third, Banca Popolare di Bari, may join them if the EC approves the Italian government's rescue plan.

Popolare di Vicenza and Veneto Banca had been troubled for some time before the European Central Bank concluded in June 2017 that they were failing or likely to fail. The Single Resolution Board determined that the conditions for a resolution action were not fulfilled, and therefore handed responsibility to Italian authorities to liquidate the banks under the national insolvency regime. However, the Italian government committed taxpayer funds to facilitate the transfer of the banks' senior liabilities and performing assets to a white knight, Intesa Sanpaolo (see “Italian Bailouts Show EU Authorities Walk A Tightrope While Banks Transition Toward Bail-ins,” published on July 4, 2017).

Some governments struggle to quit their bail-out habit

At first glance, it is odd that various European governments persist in using taxpayers' money to save failing, mostly second-tier banks. The BRRD handed regulatory authorities a range of new tools to manage struggling banks, but they do not appear to have gained traction yet. Continued bail-outs risk perpetuating the so-called sovereign-bank "doom loop" if investors see bank recapitalizations as a long-term contingent liability of some or all governments.

We see four main, interconnected factors behind persistent government bail-outs:

1. Limited appetite to impose losses on certain creditors

Memories of the financial crisis still loom large across parts of Europe's financial system, and policymakers are keen to minimize the risk of renewed instability. As a result, some authorities appear wary about imposing losses on influential creditors through the application of their powers to resolve or liquidate failing banks. Italy has been the prime example of this trend. The influential creditors there include retail depositors (including uninsured balances), senior and subordinated retail bondholders, and unsubordinated corporate and institutional liabilities. Time will tell whether government support is merely a temporary stopgap to protect these liabilities while a larger number of firms build an appropriately-sized layer of subordinated MREL (minimum requirement for own funds and eligible liabilities).

In contrast, there are recent examples, mostly outside the eurozone, where small banks have been resolved or liquidated largely as BRRD legislators originally envisaged. For instance, in Denmark, where policymakers favor resolution over liquidation even for non-systemic institutions, two small banks have been resolved since the BRRD was enacted. The resolution authority imposed losses on subordinated and bail-inable senior creditors (including uninsured depositors), and transferred the banks' residual balance sheets to new owners. Denmark's deposit guarantee scheme (DGS) also contributed to these resolutions. In Poland, the January 2020 resolution of a small cooperative bank, Podkarpacki Bank Spóldzielczy w Sanoku (PBS Bank), involved the transfer of retail and small business customers' insured and uninsured deposits to a bridge bank. The resolution authority (which also manages Poland's DGS) chose to leave public sector and large corporate deposits behind in the failed entity, which entered insolvency. The authority favored resolution over a standard liquidation process to maximize public sector depositors' recoveries.

Across the region, regulatory authorities appear sanguine about bailing in banks' MREL-eligible subordinated debt held by institutional investors. Indeed, burden-sharing by holders of regulatory capital instruments is an established condition of state aid approval for government support measures. This suggests that authorities may be more amenable to resolution once MREL buffers are broader and deeper across the banking system. Recent experience shows that forcing losses on retail holders of subordinated debt is politically fraught, with compensation schemes refunding eligible investors in failed Italian and Spanish banks. National legislators have powers to address this concern by deterring retail ownership of MREL instruments through, for example, adjusting investor protection rules or raising the minimum investment size. In addition, institutions selling subordinated debt to retail investors will be subject to stricter suitability test requirements from December 2020 under the updated BRRD.

2. Many banks are not currently resolvable

When the BRRD became law in 2015-2016, EU banks were not actually resolvable in practice. In our view, most are still not. An essential precondition of resolvability is sufficient balance sheet capacity to absorb losses, restore solvency, and protect systemic liabilities. A number of banks have made good progress in building MREL, but many have yet to begin in earnest. The European Banking Authority disclosed in February 2020 that, of 222 banks covered by a resolution strategy other than liquidation, 117 had an MREL shortfall at year-end 2018 totaling €178 billion.

This picture is unlikely to change quickly (see “The Resolution Story For Europe's Banks: Life In The Halfway House,” published on July 18, 2019). Although we expect a broader range of banks to issue MREL this year, those in the eurozone have until January 2024 to complete the required buffer. Furthermore, eurozone banks with less than €100 billion in total assets are not systematically required to maintain MREL in excess of the minimum own funds requirement, which implies that some institutions in that cohort are earmarked for liquidation rather than resolution. EU laws also permit banks to include material senior preferred debt in their MREL buffers under certain conditions and if approved by resolution authorities. This debt ranks pari passu with certain operating liabilities and could reduce the chances of a successful resolution if it is bailed in, whether selectively (likely leading to legal challenges by bondholders under "no creditor worse off" safeguards) or in combination with equal-ranking liabilities (jeopardizing the stabilization of the bank). Some EU countries have amended their creditor hierarchies to reduce those pari passu claims by preferring uninsured deposits, but these measures do not fully isolate senior preferred bonds and increase inconsistencies across the region.

Other preconditions for resolvability include operational continuity and availability of funding. We see progress in these areas as uneven and generally less advanced than in MREL building (see “Increasing Disclosure Is Set To Shine More Light On Bank Resolvability,” published on March 18, 2019).

3. Fragmented mix of regional and national decision-making within the eurozone

Responsibility for supervision, resolution, and liquidation decisions in respect of eurozone banks is currently divided between regional and national authorities, which can interpret the same facts differently. In the cases of Popolare di Vicenza and Veneto Banca, for example, the Single Resolution Board (SRB) determined that resolution actions were not in the public interest because neither bank provided critical functions and their failure was not expected to have a significant adverse impact on financial stability. The Italian government seemingly reached the exact opposite conclusion from the SRB because it chose to provide support in the insolvency process to facilitate the balance sheet transfers to Intesa. The EC approved this measure under its state aid rules to avoid economic disturbance in the Veneto region.

The case of Latvia's ABLV also highlighted inconsistent outcomes under national insolvency legislation. After the European Central Bank (ECB) determined in February 2018 that ABLV was failing or likely to fail, the SRB determined that a resolution action was not in the public interest, and concluded that the bank should be wound up. There then followed a period of legal wrangling over whether ABLV and its Luxembourg subsidiary were in fact insolvent. Both entities ultimately agreed to enter liquidation processes to maximize recoveries to creditors. Latvia's DGS was activated to support insured depositors and there was no government intervention. Another small Latvian bank, PNB Banka, followed a similar path last year but the court initiated insolvency proceedings somewhat faster, just under a month after the ECB assessed the bank as failing or likely to fail.

Eurozone policymakers have proposed amending the BRRD, broadening the reach of the SRB's decision making authority, and harmonizing national insolvency laws to ensure a more consistent framework for handling failing banks. However, this could be a lengthy process and some national governments may oppose the further centralization of responsibilities.

We see greater clarity over the roles of resolution and orderly liquidation in countries such as Sweden, Switzerland, the U.K., and the U.S. In contrast to the eurozone, these countries benefit from largely unified legal systems and centralized decision making processes. They have been at the forefront of operationalizing resolution by pushing banks to build bail-in buffers and address barriers to resolvability.

We note the court judgment last year that the 2014 recapitalization of Italy's Banca Tercas by the country's DGS did not constitute state aid. This was on the grounds that the intervention was not orchestrated by the government and the DGS would have had to reimburse depositors later if the bank had defaulted. The EC has appealed this ruling. If the appeal fails, the Tercas precedent could establish a long-term method to rescue banks outside of resolution, without imposing losses on creditors. Elke König, chair of the SRB, recently called for greater clarity and consistency about when DGSs may recapitalize banks in place of resolution or orderly liquidation.

Less controversially, the BRRD cites mutual support structures provided by institutional protection schemes (IPSs) as private sector measures, which implies they would not be classified as state aid. IPSs ensure their member institutions, typically cooperative and savings banks, have the solvency and liquidity needed where necessary to avoid bankruptcy. Alongside the state contribution, the rescue of NordLB was partly funded by its IPS, the members of which are all public sector banks. The EC did not rule on whether the IPS' involvement represented state aid, which suggests to us that it does not differentiate between public and private sector IPSs. IPSs are particular to specific sectors including the German Landesbanken and Sparkassen, and therefore are not a relevant consideration for many European banks.

4. Resolution and wind down tools remain largely untested

Some governments may see bank bail-outs as lower risk than the largely-untested resolution and orderly liquidation tools. Bail-outs bring financial and possibly reputational consequences but there have been few adverse political repercussions for the governments that led recent bank rescues. This could embolden others to follow suit unless policymakers take steps to ensure more consistency in managing failing banks, with greater reliance on market discipline.

The most visible resolution action to date involved Spain's Banco Popular, but it was not a thorough test of the process because Santander acquired the entire bank following the write-down of Popular's capital instruments and raised the additional funds necessary to complete the recapitalization (see “Eurozone Bank Resolution Framework Passes The Banco Popular Test, To A Point,” published on June 19, 2017). There are currently no examples of a bank that has undergone a bail-in led, open bank resolution and been restored to health as a viable stand-alone entity. In theory, resolution should work for banks that have built suitable MREL buffers and addressed other barriers to resolvability. However, authorities face a challenging task in reassuring depositors and counterparties about the predictability of the process and the financial strength of the resolved bank.

There are various legal cases in progress related to the handling of the Popular resolution by the SRB and other parties. We think that litigation was largely inevitable when resolution powers were applied for the first time. This is because authorities' assessments of the failing bank are partly subjective and legal challenges enable investors to clarify the parameters of resolution outcomes. Once these cases are completed, we will assess whether they have implications for the future effectiveness of the resolution regime.

Likelihood of resolution should increase as banks address barriers to resolvability

Amid the continuing government bail-outs of failing banks, it is fair to question the credibility of the EU's resolution regime, particularly as it is applied within the eurozone banking union. The region implemented these reforms relatively quickly after the banking and sovereign debt crises. From the outset, member states' levels of commitment to resolution differed widely, and the framework incorporated various compromises. The BRRD framed resolution or orderly liquidation as the presumptive, but not mandatory, outcomes for failing banks. It is clear, certainly with the benefit of hindsight, that resolution powers cannot be effective unless banks are actually resolvable in practice. Banks' journeys towards resolvability are taking longer than we originally expected and, in many cases, still have a few more years to run. Once this process is more advanced, we think resolution should become a more demonstrably practical option. We think procedural changes by regulators and resolution authorities, both cross-regional and national, could benefit the consistency and predictability of resolution and liquidation outcomes.

For most systemically important EU and Swiss commercial banks, we removed rating uplift for extraordinary government support from our ratings during 2015, shortly before EU member states implemented the BRRD in national law. In some cases, we fully or partly replaced this with uplift in recognition of banks' additional loss-absorbing capacity (ALAC). We have subsequently included ALAC uplift in more of our ratings on systemically important European banks as they have built subordinated MREL buffers, and we expect the number to increase further (see chart 2).

Our 2015 rating actions reflected two conclusions we drew following the enactment of the BRRD: EU governments' support for private sector banks was uncertain and the region's resolution regime was effective. These were purposely forward-looking assessments that anticipated the evolution of resolution frameworks over subsequent years. Resolution is a relatively young, ambitious concept that fundamentally reshapes how banks are regulated, managed, organized, and resourced. The SRB became operational only in January 2015 and then started to draft resolution plans for the larger eurozone banks under its direct responsibility. Furthermore, in several countries, the national legislative changes that enabled banks to issue senior nonpreferred debt (a staple MREL instrument in most EU countries) were typically enacted only in 2017-2018.

We maintain the same two conclusions today regarding uncertain government support for private sector banks and the resolution regime's effectiveness. We have been surprised by governments' continued willingness to provide support and their ability to navigate the constraints introduced by the BRRD. Unlike Ireland, Portugal, and Spain, Italy did not institute a bank recapitalization program prior to the BRRD taking effect. Italian authorities appear to have little appetite to use resolution powers to force creditors to absorb losses arising from legacy problems, such as nonperforming loans (NPLs), particularly if retail creditors would take a hit. As structural weaknesses such as legacy NPLs are steadily addressed and banks increase their resolvability, we think authorities' propensity to take resolution actions may increase.

Resolvability is not binary and it will be difficult to determine that a bank is definitively resolvable. Aside from MREL issuance and restructuring of legal entities, there is currently little transparency around banks' progress in addressing barriers to resolvability. We are hopeful that authorities' plans to publish resolvability assessments will provide more detail in this area (see “Increasing Disclosure Is Set To Shine More Light On Bank Resolvability,” published on March 18, 2019).

Watching brief on government support and resolution regime effectiveness

We intend to keep our assessments of government supportiveness and resolution regime effectiveness under review as banks address resolvability and authorities work to make the framework more credible and consistent.

If reforms stall in a particular jurisdiction and we think the resolution regime is unlikely to be implemented as comprehensively as we originally envisaged, we could determine that the resolution process is not sufficiently effective, which would make all banks in that jurisdiction ineligible for ALAC uplift. If we identify material, unmitigated resolvability constraints specific to a particular bank, we could either raise the standard 5% and 8% thresholds that result in one and two notches, respectively, of ALAC rating uplift, or determine that it is not ineligible for ALAC uplift.

We could reintroduce government support uplift to certain bank ratings if bail-outs become a normal and predictable response. However, this seems an unlikely scenario for as long as the current EU resolution legislation remains on the statute book. Although Canada and Hong Kong have introduced effective resolution powers, we think their governments retain significant flexibility to intervene and we still see them as supportive of systemic banks. In contrast, the BRRD constrains member states' options and positions resolution or orderly liquidation as the standard outcomes.

Related Criteria

• Bank Rating Methodology And Assumptions: Additional Loss-Absorbing Capacity, April 27, 2015
• Banks: Rating Methodology And Assumptions, Nov. 9, 2011

Related Research

• The Resolution Story For Europe's Banks: Life In The Halfway House, July 18, 2019
• Ending Too Big To Fail: Different Journeys, Different Destinations, April 4, 2019
• Increasing Disclosure Is Set To Shine More Light On Bank Resolvability, March 16, 2019
• The Government's Intervention In Carige Is One Step Closer To Ending The Long-Running Story Of Italy's Troubled Banks, Jan. 10, 2019
• Italian Bailouts Show EU Authorities Walk A Tightrope While Banks Transition Toward Bail-ins, July 4, 2017
• Eurozone Bank Resolution Framework Passes The Banco Popular Test, To A Point, June 19, 2017
• To Fail Or Not To Fail: The Point Of Nonviability Is Unclear For European Banks, May 3, 2017