Recapitalisation is in the air. Falling investor confidence has contributed to a sharp decline in Italian banks' share values, in particular since the UK's 'Brexit' vote to leave the European Union. Total gross non-performing loans ("NPLs") in Italy have increased by approximately 160 per cent. since 2009 and now amount to approximately €360 billion – a fifth of Italy's GDP. Reports indicate that only half of this amount has been provisioned.

EBA stress test results are due to be published by the end of July and will provide guidance on the extent of the required recapitalisation. The Italian government is negotiating with European institutions for ways to inject public capital into the Italian banks. However, the European bank resolution framework and State aid rules limit the ability to use public funds to prop-up the capital of banks without losses being imposed on creditors. This creates a dilemma for the Italian government as €200 billion worth of senior and subordinated bonds is held by domestic retail investors, and looming in the near future is a politically charged referendum on constitutional reforms.

This alert seeks to present potential next steps that the Italian government or supervisory authorities may take or permit to facilitate the recapitalisation of Italian banks. Among others, we address the following topics:

  • What options are available within the European bank resolution framework and State aid rules to recapitalise Italy's struggling banks?
  • What impact will the high levels of domestic retail investors have on bank recapitalisation plans?

Brexit: a catalyst?

In terms of substance, little has changed since the British referendum last month. The challenges faced by Italian banks are entrenched and the NPL issue far predates the Brexit vote. However, the Brexit vote has stoked a fear that low economic growth will persist and interest rates will be 'lower for longer', to the detriment of bank profitability. Accordingly, this week the IMF lowered its forecast for economic growth in Italy to 1 per cent. for 2017 and just under 1 per cent. for 2016.

Whether or not Brexit has been the trigger, or even the excuse, for urgent discussions on how to recapitalise Italian banks, action is (and has been for several years) urgently needed.

Current landscape and the need for a government solution

It is widely expected that the stress test results will confirm that Italian banks need to reduce their NPL levels and repair their capital ratios. Italy has already taken some steps to address this. In February this year its government introduced a State guarantee scheme to promote the securitisation of NPLs (so-called "GACS"), while in April it orchestrated the creation of Atlante, a €4.25 billion back-stop fund sponsored by Cassa Depositi e Prestiti (a State-owned institution), banks, insurers and other institutional investors, such as the Italian banking foundations.

Atlante aims at stabilising Italy’s financial system by purchasing (i) newly issued shares of banks which do not meet their SREP capital requirements and (ii) NPL portfolios. According to the fund’s rules, investments in newly issued Italian bank shares may not exceed 70 per cent. of its gross asset value (see our February 2016 alert – "New measures to incentivise NPL market in Italy"). In late June, Italy was given the go-ahead by the European Commission to supply as much as €150 billion in government liquidity guarantees for its struggling banks until the end of the year.

The challenge ahead

Recent events indicate that private capital may not be available to resolve the recapitalisation of Italian banks. The resolution of four smaller Italian banks at the end of last year and the recent failed share issuances of Veneto Banca and Banca Popolare di Vicenza emphasise the challenge in finding sufficient private sector funds. Both issuances were effectively rescued by the Atlante fund purchasing shares worth €2.5 billion. It is widely acknowledged that Atlante has a huge burden to shoulder with relatively limited resources remaining, which will likely prove insufficient to support further significant recapitalisations.

It is possible of course that a solution may be found whereby the banks are recapitalised – at least just enough for the immediate future – without drawing on public funds. However, it seems most likely that State support will be needed if there is to be a serious attempt to restore long term confidence in the sector. Herein lies a paradox of the new European bank resolution legislation – if the price of public sector support is burden sharing (losses) by non-privileged bank creditors, this may itself cause the very systemic instability that the new rules are designed to prevent.

This raises the question: how can the Italian government support its banks within the framework of European legislation in a way that will not give rise to further instability in the financial system?

An Italian peculiarity: the retail dynamic

A central theme of the recently implemented European bank resolution legislation is that private sector creditors (principally, shareholders, senior and junior bondholders and depositors with deposits not protected by deposit guarantee schemes) must potentially contribute to the recapitalisation of a failing bank before State support can be granted. This means that there must be a 'bail-in' of creditors before a 'bail-out' by the taxpayer.

This fundamental concept is a major cause for concern in Italy, where high levels of retail holdings exist owing to a tax quirk which had the effect that, until 2012, investments in bonds issued by Italian banks were subject to a lower tax rate in comparison to bank deposits and similar savings instruments. Retail investors purchased senior and subordinated bonds over many years in search for yield and in lieu of bank deposits. The bond offerings were viewed by many savers as akin to investments in certificates of deposit. The investor demand was met by supply as Italian banks relied, if not over-relied, on the availability of retail savings. The senior and subordinated bond offering documents passed the review of Italian regulatory authorities.

According to the Bank of Italy, approximately 10 per cent. of the value of financial assets held by Italian households is represented by bail-inable liabilities of Italian banks (including bonds and deposits but excluding shares), which reportedly amounts to approximately €427 billion.

So will there be burden sharing as anticipated by the BRRD?

Imposing losses directly onto the savings of a large number of the Italian public could further weaken public confidence in the Italian banking sector in Italy. Many retail holders are, by nature, unsophisticated investors who claim not to have fully understood the risk that their savings could be wiped out to satisfy a new principle of European bank resolution law. In addition, the Italian government will be mindful to avoid repeating the negative public reaction to the resolution in 2015 of four smaller banks in which retail investors were not protected. The EU supervisory authorities may also be more inclined to reach a compromise with Italy in light of the strained politics in Europe following the 'Brexit' vote.

Nonetheless, it is hard to imagine that almost as soon as the BRRD and the Single Resolution Mechanism have been fully implemented, the European Commission would allow a flagrant disregard for its core principles upon its first major test case. Indeed, the European Commission rejected the Italian government's proposal in late June to suspend the resolution framework and State aid rules to allow a recapitalisation without bail-in. The challenge therefore remains to determine what flexibility the rules afford.

What might State support look like?

Default solution: Bail-in in accordance with BRRD

The default solution under the European bank resolution framework would be to put a failing bank into resolution prior to any public recapitalisation.

Under the BRRD this requires prior burden sharing by shareholders and other creditors of 8 per cent. of the bank's liabilities through bail-in or otherwise, before State funds may be used to recapitalise a bank. Equity and subordinated debt holders must be bailed-in first. Senior creditors and depositors must participate in the burden sharing if this is required to reach the 8 per cent. threshold. Only deposits that fall under the mandatory deposit protection scheme are exempted. By potentially requiring losses to be incurred by senior creditors and providing for the application of the burden sharing up to the 8 per cent. threshold, the BRRD rules go beyond the burden sharing requirements under EU State aid rules – where no such threshold is provided and burden sharing for senior creditors is not required, as described below.

Could retail creditors be exempted from a bail-in?

The framework does allow certain creditors to be protected from losses, in exceptional circumstances. Any such exclusion of a group of creditors must be strictly necessary and proportionate to avoid giving rise to widespread contagion which would severely disrupt the functioning of financial markets, including of financial market infrastructures, in a manner that could cause a serious disturbance to the economy of a Member State or of the European Union. It may be arguable to exclude retail depositors and retail bondholders from a bail-in on this basis. It may even be possible to justify that other bondholders may be excluded from bail-in.

This will ultimately be a judgement call by supervisory authorities in accordance with public administrative law. The reaction to the decisions by the Bank of Portugal at the end of last year following the resolution proceedings of Banco Espirito Santo shows how such decisions can hugely undermine confidence in the sector when they are perceived to be incompatible with the conventional legal norms or settled rules of the market.

It should be noted that any such decision to give privileged treatment to certain creditors, who would otherwise have been bailed-in, may give rise to a right of compensation for disadvantaged creditors under the no-creditor-worse-off principle enshrined in the BRRD. This principle guarantees that creditors are treated no worse than in a hypothetical insolvency of the bank. However, the source for compensating creditors affected by this principle should be other banks that contribute to the Italian national resolution fund.

Or could retail creditors be otherwise compensated for losses suffered in a bail-in?

Another possible approach to the problem posed by retail holders might be for the Italian authorities to create a default compensation scheme. Retail holders may be expected to evidence that the bonds sold to them were prima facie an unsuitable investment.

Such compensation, if provided by the Italian State, might also constitute State aid to the extent that the banks themselves were partly or wholly responsible for the sale of such bonds. However, retail holders could potentially be compensated immediately by the State, with banks contributing to the compensation costs over a longer time period that would be compatible with their recapitalisation plans. The European Commission has already signalled to Italy that it may be willing to allow such a compensation scheme under State aid rules. A similar investor compensation scheme is in the process of being established by the Italian State to reimburse the holders of subordinated bonds issued by the four banks that were put into resolution in 2015 and may in principle be extended through additional public funding to other retail bondholders in the event that other Italian banks are resolved.

An alternative to 'bail-in'? Precautionary recapitalisation under Article 32(4)(d) of the BRRD

Public funds may be injected into the banks without triggering a resolution via so-called precautionary recapitalisation. The BRRD permits capital injection from State funds, provided it is precautionary, follows a stress test, and specifically addresses capital shortfalls implied by an 'adverse scenario'. Precautionary public sector recapitalisation is not permitted for banks that are insolvent or considered to be likely to fail in the near future under the 'point of non-viability test' in the BRRD. Article 32(4)(d) of the BRRD also permits a public sector guarantee to be given in respect of access to central bank liquidity.

Banks will be notified of the results of their individual stress tests ahead of publication and the EBA and the ECB will be presently engaged in discussions with the Italian banks over the emerging results. A potential outcome could be some measure of State funded recapitalisation based on Article 32(4)(d) of the BRRD at least in respect of any shortfalls implied by a failure to meet the adverse stress scenario.

However any such support would be subject to prior approval under the EU State aid framework. Also, because State support must be limited to what is needed to address the hypothetical adverse stress test scenario, other capital may be needed to bolster a bank’s baseline capital ratio.

Precautionary recapitalisation: What burden-sharing is required by the State aid rules?

A precautionary recapitalisation must be in line with State aid requirements, which are found in the Banking Communication 2013 (the "Communication"). The Communication limits mandatory burden sharing to equity, hybrid capital, and subordinated bonds. On the face of it, this would include the retail holders of subordinated bonds of Italian banks. The Communication is not however prescriptive as to the quantum of required burden sharing but instead adopts the potentially more flexible formula that burden sharing should be 'adequate'.

This more flexible approach featured in the 2015 recapitalisations of certain Greek banks, most notably Piraeus Bank and the National Bank of Greece. Both banks were recapitalised through a combination of new private sector capital introduced via a consensual liability management exercise followed by State aid from the Hellenic Stability Fund which was granted as precautionary recapitalisation under Article 32(4)(d) of the BRRD. Approval in both cases was conditional upon participation by existing junior and senior creditors. These cases demonstrate that public sector support can be used in conjunction with private sector contributions in ways that avoid a bank being 'resolved' and creditors being forcibly bailed-in under the BRRD.

It is also notable that the Communication provides a more general exemption for burden sharing by subordinated bondholders. These may be exempted only in limited circumstances; if implementing such measures would endanger financial stability or lead to disproportionate results. This exception could extend to cases where the aid amount to be received is small in comparison to the bank's risk weighted assets and the capital shortfall has been reduced significantly in particular through capital raising measures such as rights issues, voluntary conversion of subordinated debt into equity, liability management exercises, capital generating sales of assets and portfolios, securitisations and other measures.

Watch this space

The extent of the Italian government's steps to recapitalise the banks – whether as a bold and widespread solution for the sector, or limited to satisfying only the immediate needs of certain banks – is yet to be seen. But it seems very likely that one way or another, it will involve a combination of mechanisms permitted under the umbrella of the European banking framework, whilst protecting (at least to a large extent) retail investors from direct losses.

A precautionary recapitalisation would seem to carry less risk of contagion or other possible unforeseeable consequences than a bail-in under a formal resolution of a bank. There are also useful examples of prior successful recapitalisations involving State support, in the shape of certain Greek bank recapitalisations last year. The question is: what is adequate burden sharing, in the unique context of Italian banks? That negotiation between the Italian government, the Bank of Italy and relevant European supervisory institutions, is well underway.