Bending and Breaking the Single Resolution Mechanism: The Case of Italy

This article examines the political economy and law of bank resolution in the case of Italy – spe- cifically, its treatment of three failing banks – the Monte dei Paschi, Veneto and Vicenza banks – which were resolved in 2016–17. These three cases stand out for the relatively large degree of dis- cretion exercised by the national resolution and state aid authorities, ultimately with the permission of their European counterparts. This article examines the motivations of Italian authorities in lobbying the Commission for leeway in applying the bank recovery and resolution directive and analyses the intricacies of the legal framework to underline the extent of discretion exercised by policy-makers. It concludes that the discretion visible in these three cases is not (entirely) contained within EU law, and that bending the law or turning a blind eye to infractions was key to understanding the EU approach to Italy.


Introduction
This article examines the question why bank resolution in Italyspecifically its treatment of three failing banks that were resolved in 2016-17 -Monte dei Paschi (MPS), and the Veneto and Vicenza popular banks (Veneto banks)deviated so strongly from what European rules on bank resolution that went into effect on 1 January 2016 demanded. The banks were rescued or merged into other banks without the full bail-in of bondholders, as foreseen in the bank recovery and resolution directive (BRRD) and the single bail-in resolution mechanism regulation (SRMR). These cases stand out for the large degree of discretion exercised by Italian and European authorities responsible for resolution and state aid. What factors drove their behaviour? And can the result be correctly depicted as bending or breaking EU law? This article examines both the intricacies of the discretion exercised by policymakers and the legal framework. It concludes that economic nationalism rooted in the Italian financial system largely explains the drive of the Italian government to push European authorities for special treatment in the resolution of Italian banks, that the discretion visible in these three cases is not (entirely) contained within EU law, and that the Commission bent the law and broke it by turning a blind eye to infractions in Italy. The centrality of individuals (who are simultaneously domestic savers, investors in the bank, and voters) to the stability of Italian banks explains both the Italian government's drive to protect investors and the willingness of the Commission and the Single Resolution Board (SRB) to accommodate the Italian government's demands. Both Commission and SRB appear to have been motivated by a desire to ensure local financial stability in the absence of domestic or EU funding, while the Italian government sought to assuage the concerns of a broad-based class of domestic investors with considerable electoral firepower.
We argue that while Italian state intervention was not overtly nationalist in these cases, its actions permitted the banks to uphold the intricate network and links that are a core characteristic of the Italian financial system. These links also make those banks unattractive to outside investors, and make it politically and economically impossible to apply European law to them. Weak interbank links are compensated by the banks' reliance on their customers as a source of stable funding and on state treasuries as a source of safe assets. The latter, in turn, increases state incentives to prevent EU law from undermining the role that banks play in supporting the Italian Treasury. These features, coupled with a secular decline in the profitability of traditional Italian banking, lead to strong economic and political pressure on the Italian state to push at the European level for more and more leeway to keep its financial system intact. This pressure has found a receptive ear at both the Commission and the SRB in the absence of alternatives that would preserve the stability of the Italian financial system. We chose the Italian case as a key instance of a member state breaking and bending the application of the BRRD to unpack the political and economic motivations from both the Italian and EU sides.

I. Analytical Framework
To provide empirical grounds for our argument that the Italian variety of financial capitalism drives its economic nationalism and pushback against EU law, this article first looks into economic nationalism, its links to the political economy of the Italian financial system and its impact on the Italian government's compliance with the BRRD. It then looks at the legal framework, including EU legal principles, powers, discretions and limitations (as applied to EU institutions, and their implications for the competent national authorities), and determines the extent to which Italy and EU institutions bent or broke the initial legal framework. The importance of dual national-EU competence for bank resolution and the existence of loopholes and gaps in the multi-level legal framework that can be exploited are explored. This provides a test of how strongly national discretion is built into the original frameworks and what its limits are intended to be, to assess whether and how EU rules have been bent and broken.

II. Economic Nationalism and Varieties of Financial Systems
There are three competing and overlapping understandings of the relationship between state, interest groups and behaviour at the national level that impact on Italy's behaviour towards its legal obligations under the BRRD and the SRM. The first is that Italian authorities engage on principle in economic nationalism, understood as protecting and promoting the national ownership (and market share) of companies (Helleiner and Pickel, 2005;Pickel, 2003), at the expense of others and international obligations, and even in the face of high opportunity costs (Epstein, 2017;Epstein and Rhodes, 2014;Gros, 2013). 1 Protecting ownership on national grounds can be seen in the Bank of Italy's blocking of foreign takeovers until 2005 (for example, in blocking the takeover by Antonveneta by ABN Amro under Governor Antonio Fazio). It may extend to nationalization as well to prevent foreign ownership. If economic nationalism prevails, governments are willing to break EU rules to ensure national ownership and control of banks.
A second explanation, liberal economic nationalism, is that governments will protect and promote national firms, but only within legal limits. States exploit loopholes, interpretations and discretionary powers to do this (Helleiner, 2002). Examples are rescue measures, including orchestrated mergers, subsidies, hard ownership cores and special voting rights and restrictions that benefit insiders and undermine foreign ownership or control (Goyer and Valdivielso del Real, 2014). After Fazio's departure, foreign ownership was in principle possible, but subsequent governments promoted the consolidation of Italian control of the economy through mergers and acquisitions that encouraged the takeover of smaller, weaker banks by their larger commercial counterparts (Bulfone, 2016;Deeg 2012;Deeg and Donnelly, 2016;Quaglia and Royo, 2015). Given the BRRD's restrictions on state aid, liberal economic nationalist (LEN) initiatives could lead to the government negotiating new and broader interpretations of resolution rules with EU authorities (bending the law and developing it further, together with EU authorities). Negotiations of extraordinary legal instruments such as precautionary recapitalization fall into this category, as discussed below. The key to this process is that the innovations are not unilateral deviations on the part of government.
We take a third approach, borrowed from comparative political economy, that governments protect and promote key stakeholders that are essential players in national varieties of financial capitalism to ensure the continuity of financial services and stakeholders' political support (Grittersová, 2014). Even where a government is not inherently nationalist it will find itself politically coerced to act that way, even if by doing so it breaks EU law. This parallels expectations of non-compliance where EU and domestic demands clash (Börzel, 2002;Falkner et al., 2004).
We work here with the dynamic interpretation of varieties of capitalismrather than the static one, and focus on the organization of finance in particular. The classic static framework expects countries to exhibit the stable characteristics of either highly coordinated, planned economies reliant on deposit-funded banking for finance; highly liberalized economies reliant on more volatile capital markets (Hall and Soskice, 2001;Wade, 2004;Zysman, 1983) or, in the Mediterranean, bank-financed economies with poor coordination that are compensated for by an interventionist state (Amable, 2003;Molina and Rhodes, 2007). A dynamic approach allows for incremental innovation by companies and governments under stress (Crouch 2005, Hall andThelen, 2008). Streeck and Thelen (2005) postulate that economies liberalize incrementally as the opportunity costs of retaining the status quo increase.
Applied to banking, some banks in coordinated and Mediterranean market economies respond to declining profitability by liberalizing their sources of capital and their asset portfolios to achieve greater returns (Hardie and Howarth, 2013;Pagoulatos and Quaglia, 2013). Rather than increase bank resilience in the long run, however, Hardie and Howarth (2013) show that the shift from deposits to market instruments as a source of cash, and from loans to market assets as a source of income, renders traditional banks fragile by exposing their balance sheets to market volatility. It further undermines key stakeholders holding together the relevant variety of financial capitalism. We argue that where fragility increases, government incentives to support banks with interventions rise. We hypothesize further that where the affected actors are voters, the likelihood of intervening increases with the size of the electorate affected, even in violation of EU rules. Following Quaglia and Royo (2015), there is evidence that this transition to fragility in Italian banking is significant.
Our contribution makes a more robust link between the methodological nationalism described in the financial capitalism literature and the motivations behind economic nationalism. These are driven by efforts to protect and promote the local variety of financial capitalism even at times of stress. The latter is against expectations that government will drop protection when it is too costly, or when legal obligations demand it. The political costs of alienating stakeholders outweigh economic costs and international obligations, but also respecting EU legal obligations.

III. The Italian Variety of Financial Capitalism and Economic Nationalism
We follow up on this set of expectations and postulate that bank behaviour drove economic nationalism in Italy 10 years after the great financial crisis, resulting in a willingness to bend and even break EU rules. Italian banks have responded to shortfalls in income and capital by relying more heavily on turning depositors into bondholders on the liabilities side of the balance sheet, as well as on state treasuries and corporate and household loans, than they did in the past. This has trapped Italian governments into having to choose between accepting and applying EU law on state aid and resolutions, and breaking the rules, if bending them will not suffice to save Italian banks.
In concrete terms, we expect governments faced with a choice between honouring legal obligations to refrain from state aid and other forms of intervention and forbearance, and ensuring the continuity of the existing local financial system, to choose the latter. The alternative, allowing local banks to fail and (perhaps foreign) competitors to take their place, will prove unacceptable. Under such conditions, governments will seek to bend and even break their EU obligationswhatever it takesto preserve national financial institutions and arrangements in ways that are incompatible with foreign ownershipwithout explicit motivations about nationality of financial institutions' owners. This would mean potential non-compliance on issues of state ownership and aid.
Our starting point for the analysis of the Italian variety of financial capitalism begins with the increasing marketization of Italian banks along the lines provided by Hardie and Howarth (2013). Like Pagoulatos and Quaglia (2013) and Quaglia and Royo (2015) we take this point further by investigating the specific nature of market-based banking in Italy in more detail. We review below the behaviour of the Italian banks, and of the Renzi government in responding to those challenges.

IV. Banks, the Economy and Marketization
The Italian banking system has been transformed under pressure. Since joining the euro, the Italian economy has stagnated, meaning that traditional loans to business have turned Shawn Donnelly and Ioannis G. Asimakopoulos increasingly into non-performing loans (NPLs), which has required banks to find better assets and raise new capital. This capital shortfall is significantly higher than the EU average in Italy, with up to 30 per cent of assets impaired (European Banking Authority, 2017;European Systemic Risk Board, 2017). The NPL stock raised the attractiveness to banks of state treasuries as a market asset category. But this safe haven strategy still leaves banks vulnerable to periodic financial market speculation on Italian treasuries (Merler and Pisani-Ferry, 2012). This trend continues today, as Italian banks increase holdings of Italian treasuries during periods of financial stress (Allen, 2018). 2 On its own, however, this was insufficient to stabilize banks.
Increased EU capital standards, but especially the advent of the Single Supervisory Mechanism in 2014, generated even more demand for cash in the Italian banking sector (Merler and Wolff, 2013). Small banks in particular were hit by downward economic trends (Deeg and Donnelly, 2016) due to their weak capacity to refinance on capital markets. In the case of Veneto and Vicenza, their status as cooperatives precluded such a step. Other Italian commercial banks found that major family blockholders could veto increases in capital that might otherwise give banks increased financial security, to avoid a reduction in the value of their existing shares and their voting power (Bulfone, 2016). This effect was seen in the collapse of Carige bank in early 2019 (Mandala and Za, 2018). 3 Taking this option for recapitalization off the table further increases the incentive for banks to seek other sources of funding. Italian banks marketized by selling subordinated debt certificates to their own depositors, who received higher interest rates than a deposit account and would prove more patient than international investors.
Banks therefore resorted to subordinate debt bought by depositors to raise cash, which are reduced in value or converted to shares (bailed-in) in the event of a bank failure. Using depositors as bondholders, who would be bailed-in under EU law in the event of an insolvency, saved Italian financial capitalism for a time. This practice had existed before (Pistelli, 1999) but was expanded. What the depositors did not price in was that in the event of a need to recapitalize the bank, the BRRD would require them to lose all or most of their investment, that is, their savings. Instead of enjoying protection from bankruptcy through the deposit insurance system, they acquired the status of bondholders and voters who would be wiped out by the application of EU law. Market liberalization thus happened, but not to outsiders.
This practice changed the Italian financial capitalism by adding junior creditors as a powerful new force, given their crucial role in holding the system together. Given high NPL levels, poor levels of trust in Italian banks and the unwillingness of foreign investors to step in, Italian banks now rely on two domestic groups to keep them afloat, with the support of the Italian government and central bank, at the cost of remaining perennially unattractive to outside investors. The domestic creditors can be divided into two categoriessenior and junior creditors. Senior creditors are more likely to be institutional (other banks and financial companies), preferring the safety of investment over returns. This group will not be first in line to lose their assets during a bail-in, thanks to their contractual senior status. But not all senior creditors are equally able to withstand a cut in their income and investments. Other Italian banks and insurance companies, given their own high levels of non-performing assets, would find themselves more vulnerable to financial contagion in the event of a bail-in than counterparts elsewhere in the eurozone. This is something that both Italian and European authorities are aware of in the context of a resolution.
The second group, junior creditors, are more likely to be retail investors (more specifically, depositors who shift their savings into junior debt), seeking a better interest rate than their savings accounts provide. This group is the special feature of the Italian financial sector that keeps it afloat, at least until the banks in question are unable to meet their financial obligations. As a stopgap measure, the expansion of subordinated debt was a market-based innovation developed to keep financial institutions (and their major family blockholders and senior creditors) afloat in the short term after the onset of the eurozone crisis that then become a lasting feature of the Italian financial system. While keeping distressed Italian banks temporarily afloat, this feature increased the political resistance to resolving insolvent banks in accordance with EU law and proved unattractive to international investors, leaving a state bailout or nationalization as the only political option. The Italian government then pursued bailout as an alternative, which the SRB approved on the basis of the banks' insignificance and then the Commission approved on the basis of protecting local financial stability (European Commission, 2017a). This required bending and breaking EU law, discussed below. Interested parties in Italian financial capitalism benefit?
The increased junior bondholding was not enough to stabilize banks as the economy stagnated further, NPLs grew and state treasuries became more fragile. A liberal economic nationalist approach to resolving bank problems can be seen in the Renzi government's promotion in 2015 of the Atlante fund (a private equity vehicle funded by consortia of Italian banks, which would pool together to provide cash to support Italian banks in distress, particularly by buying NPLs), and in April 2016 the Atlas fund, with the same purpose, before that government's demise in December 2016.
A willingness to involve foreign investors along LEN lines can also be seen. In the case of MPS, the Renzi government supported the bank's 2016 attempt to attract non-Italian (American) investors as a means of averting closure. MPS enlisted JP Morgan to support an international investor search (Aloisi et al., 2016), 4 but struggled with the fate of depositor bondholders (Il Fatto Quotidiano, 2016) 5 before the government stepped in and, with the European Commission's support, negotiated a precautionary recapitalization for the bank in 2017. This problem is exacerbated by weak collective action by Italian bank consortia, refusing to buy NPLs from distressed banks as a means of making the latter more attractive to outside investors at the behest of the Italian government. This happened when the Atlas fund refused to buy bad loans from MPS, which in turn led a Qatari investor to back out of a major planned investment (Arnold and Sanderson, 2016). 6 In this context, cash injections from the private sector, in conformity with the BRRD, remained unavailable. This meant that the government found itself having to choose between accepting BRRD demands for a bail-in and championing the cause of Italian depositors turned subordinated bondholders, as discussed below (Donnelly, 2018a;Hale, 2016). The search for foreign investors, however unsuccessful, is at odds with the premise that the government would block attempts at foreign ownership. But given the banks' financial difficulties, it is difficult to imagine that a capital injection from a new investor would take place without a further bail-in of existing Italian creditors. Given the Italian government's further steps to protect depositor bondholders, it is unlikely this would have happened. The Renzi government's willingness to distance itself from outright nationalism therefore appears to have been fruitless due to foreign reluctance to invest in Italian banking, the political difficulty of bailing-in depositor bondholders and the subsequent rise of domestic political pressure. These features of Italian financial capitalism led it to the aggressive pursuit of more leeway in bailing them out, even against the intent of the BRRD.
This practice of junior bondholding keeps the Italian variety of financial capitalism largely intact and reinforces it by continuing to protect blockholders and senior creditors from outside competition. Allowing it to function required a decision by the Italian government to subsidize the system under national bankruptcy law, by the Commission to allow it and by the SRB to open the door for the Commission to approve the use of national rather than EU procedures. Under the BRRD, a bail-in of 8 per cent of assets, starting with junior bondholders, would have wiped out many of the retail depositors who are also voters. The Commission's approval of an Italian government decision to reclassify those bonds as senior (to protect them), or to compensate them directly for their losses due to bank "mis-selling" (although the practice goes back decades) underlines the vicious cycle in which the politics of keeping the Italian banking system afloat drags down efforts to clean up its financial problems.
In sum, we can identify changes to the Italian variety of financial capitalism, which is becoming increasingly fragile due to the increased importance of market-based banking but is held together by the interests of stakeholders. Under stress, this generates pressure for government to act in ways that are effectively nationalist, by virtue of its incentive to break EU law, not just bend it, and which also make foreign investment unattractive. The European Commission has created room for the Italian government to do this in the context of ensuring local financial stability.

V. Legal Drivers: Powers and Discretionary Room in EU Law
Above, we argued that the persistence of an Italian variety of financial capitalism, which is becoming increasingly fragile due to the increased importance of market-based banking, but is held together by the intense interests of insider and outsider stakeholders, generates pressure for the government to act, nationally and on the European stage, in ways that are effectively nationalist, by virtue of the incentive to break EU law, not just bend it, and in ways that make foreign investment unattractive.
It is remarkable that EU institutions have in part accommodated these pressures in the short term in light of the consequences of imposing widespread economic costs on Italian voters. Faced with choosing between vigorous law enforcement and the accommodation of national differences to ensure financial stability, the European Commission (2017a) elected the latter in the cases of Veneto and Vicenza, rather than reduce the value of bonds held by Italian households.
Bending and Breaking the Single Resolution Mechanism: The Case of Italy * Bank resolution is arguably a sui generis legal construction since it is an administrative procedure with insolvency-like effects, without the stakeholders having a say on the final resolution outcome. Following initiatives of international fora and under the pressure of the eurozone sovereign bank debt loop (Howarth and Quaglia, 2016), EU member states agreed on a harmonized bank resolution framework for all EU member statesthe BRRD, and on a binding substantive and procedural framework for all eurozone member statesthe SRMR.
However, this shift of competences from the national to the European level for various reasons did not come easily. Firstly, banking was perceived before the crisis as an area strictly under national control. The fact that banks tend to hold a large amount of sovereign debt of their home state reinforced this perception. Secondly, bank insolvency was considered to be a purely national competence, given its implications for investors and the real economy. That said, national insolvency remains a disincentive to apply EU resolutions. Thirdly, member states did not want the European Central Bank (ECB) or the Commission to be bank resolution authorities (Zavvos and Kaltsouni, 2014). That is why the initial proposal to increase Commission power was watered down and most competences were moved to the SRB, an EU agency. Fourthly, moral hazard concerns were raised as the European resolution would meanto a certain extentusing European resolution funds. Judging from the current resolution financing arrangements, nations hesitated to agree to that (Asimakopoulos, 2018).
Given this plethora ofsometimes conflictingnational and supranational interests, EU bank resolution is enforced by a multilevel governance system (Asimakopoulos, 2019). The ECB is the first to signal that a bank is failing or likely to fail. Then the bank is liquidated under national insolvency laws, unless the SRB decides to resolve the bank at the European level on the grounds of public interest. If the SRB rules that such public interest exists, then the Commission can challenge the ruling within 12 hours and let the Council decide afterwards. In other words, at least three institutions are involved in triggering resolution: the ECB, the SRB and the Commission. National resolution authorities then carry out the resolution in accordance with the SRB's decisions. In principle, national resolution authorities enjoy limited discretion and the SRB can overrule them in case of non-compliance (SRMR, Article 29). A bank may also try to invoke the precautionary recapitalization clause in the SRMR that allows a solvent bank to be recapitalized according to state aid rules without entering resolution (SRMR, Article 18[4]).
Judging from the experience of MPS, which made use of the precautionary recapitalization clause, and the Veneto banks that were liquidated under Italian laws, we focus here on powers and discretionary room when it comes to triggering bank resolution. Therefore, our analysis concentrates on the use of precautionary recapitalization over resolution (MPS) and on the choice between liquidation and resolution (Veneto banks).

Precautionary Recapitalization Versus Resolution
To understand the rationale of precautionary recapitalization one needs to understand how things worked before the BRRD and SRMR were put in place. EU bailouts took the form of state aid provided by member states to their banks. The legal basis for the bailouts during the crisis and therefafter is article 107(3)(b) TFEU. It deems state aid to be compatible with the internal market if it is to "to remedy a serious disturbance in the economy of a Member State". In order to specify the conditions applying to bank bailouts under 107(3)(b) TFEU, the Commission's 2013 banking communication requires a certain level of burden-sharing to take place before the Commission approves the state aid. The burden-sharing required by the Communication is significantly less than under the bail-in rules of the BRRD and SRMR that entered into force in 2016. The former requires a bail-in of shareholders and junior creditors, while the latter extends it to senior creditors as well. Eight per cent of the bank's liabilities must be bailed in before the bank under resolution can receive direct capital injections by the state. Therefore, state aid burdensharing is looser than the BRRD and SRMR bail-in.
That said, how does one assess a serious disturbance in the economy of a member state under state aid law, and how does the Commission conduct this public interest assessment? According to article 107(3)(b) TFEU there needs to be a market failure serious enough to affect at least the entire national economy ('serious disturbance') followed by a positive public interest assessment that reflects the fundamental principle of proportionality that needs to be met at all times (Binder, 2017). A "serious disturbance" is the criterion used in this assessment, with at least a national disturbance being the rule of thumb for accepting or rejecting the existence of public interest. No further clarification has been provided as to the specifics of this assessment. The Commission merely states that 'the global financial crisis can create a serious disturbance in the economy of a Member State and that measures supporting banks are apt to remedy that disturbance' (European Commission, 2015, recital 106). Moreover, the Commission takes into account the aid recipient's past economic difficulties in its reasoning (European Commission, 2016, recitals 26, 27) making no explicit connection to seriousness or to the specific implications for the current disturbance. This provides great latitude. That said, even 10 years after the financial crisis, the definition of a serious disturbance has been almost exclusively developed in the context of the financial crisis. In addition, even small banks that fall outside the scope of the SRMR have benefited from using 107(3)(b), thus increasing the legal uncertainty over the use of state aid in the banking sector (European Commission, 2017b, recital 13). In other words, given the ad hoc nature of those state aid decisions and the fact that they do not constitute any sort of legal precedent, the only way to ensure agreement over what a serious disturbance means is through the Commission's Communications.
Then came the BRRD and SRMR, which requires an 8 per cent creditor contribution (both junior and senior) to bank losses before extraordinary governmental support can be provided. However, precautionary recapitalization allows state aid to be provided alongside the aforementioned burden-sharing without the bank entering resolution. Pursuant to article 18(4)(d), SRMR extraordinary public financial support can be provided in order to remedy a serious disturbance in the economy of a member state and preserve financial stability. Such measures need to be precautionary and temporary, proportionate to the aimed outcome and applied to solvent banks only. The wording is identical to that of article 107(3)(b) TFEU, the legal basis for state aid provision to ailing banks.
Enforcing precautionary recapitalization is challenging for two reasons. Firstly, the decision-making process is mostly private, and secondly, the limited burden-sharing is by itself a motive for banks, national resolution authorities (and, depending on the case, the Commission, the ECB and the SRB as well) to favour it over a full bail-in or resolution. Starting from the decision-making process, it is up to the ECB and the SRB to decide whether the conditions of precautionary recapitalization are being met or not. However, the ECB and the SRB take these decisions based on data that only they have. Any public disclosure of the data is prohibited due to financial stability concerns. Therefore, doubts have been raised as to whether MPS, a bank with limited connections to other financial institutions and mostly relying on standard deposit taking and lending, could fit the definitions of 18(4)(d) SRMR and 107(3)(b) TFEU (Götz et al., 2017). The second issue, that of the different burden-sharing requirement under resolution and precautionary recapitalization, generates the wrong incentives primarily to member states and banks, which will seek to make use of this favourable clause before entering resolution. It is not a coincidence that before the BRRD was enacted many countries rushed to rescue their banks (Italy didn't), and that Italy requested the use of that provision in the case of the Veneto banks as well.
That leads us to the role of the Commission, which is responsible for approving state aid, as the third entity involved in decision-making after the ECB and the SRB. In the MPS case, state aid was provided as part of a restructuring process. The Commission decision focused on two elements; first and foremost, on the potential long-term viability of MPS and, to a lesser extent, on its market power. How the two relate to each other remained unanswered, given the relatively low market shares of MPS (Tanninen, 2018). If there were solid reasons to provide state aid, they were definitely not included in the Commission's reasoning.
It is therefore impossible to identify the legal reasoning behind the decision to permit precautionary recapitalization, which provides the necessary discretionary room to all interested and affected partiesboth national and supranationalto interpret (or bend) the law in their favour.

Liquidation Versus Resolution
After the ECB makes its 'failing or likely to fail' assessment, the SRB decides whether a bank should be liquidated under national insolvency proceedings or resolved at the European level on public interest grounds. So far, the SRB has established the existence of a public interest in the case of Banco Popular in Spain, and has rejected it in the case of the Veneto banks and ABLV Luxembourg. The credibility and transparency of the public interest assessment remain vague. As the outcome of those cases has shown thus far, the Veneto banks were liquidated while receiving liquidation aid, and the Luxembourg Commercial Court challenged the SRB's assessment and ordered the ABLV to be restructured rather than liquidated.
The Commission appears as the SRB's equal in triggering resolution, as it has to cosign the resolution scheme. However, these initial cases show that this legal arrangement is mainly intended to mitigate legitimacy concerns linked to having an EU agency taking such life and death decisions without the oversight of the Commission (Chamon, 2014). However, the time limits under which the Commission andpotentiallythe Council must operate, are unquestionably strict. In the Banco Popular case, the time span between the SRB's proposal and the Commission's endorsement was merely 77 minutes (Crump, 2017). We may therefore safely conclude that the Commission either did not carry out a comprehensive assessment within the allotted time frame, or had already promoted its own views while sitting as an observer on the SRB's executive board. However, even if negotiations between the SRB and the Commission (and the member states concerned) take place behind closed doorswhich must be indeed the casethis should be expressly stated in the SRMR.
Nevertheless, if the SRB decides that there is no public interest in resolving a bank, then the default scenario of national insolvency is to be applied and the Commission (and the Council) have no say in it. In the Veneto banks case, the institutions were at odds: the SRB found no public interest but the Commission immediately afterwards established there was sufficient public interest to justify the provision of state aid to the Veneto banks under resolution (liquidation aid). The Commission and SRB determined the public interest in different ways. Meanwhile, the ABLV Luxembourg case has taught us that the ECB or the SRB's assessments are not always substantiated on quantitative data.
The SRB has so far used the criteria of 'critical functions' and the 'significant adverse effect on the financial system' to assess the existence of public interest. A function is considered critical when its sudden disruption is likely to have a material and negative effect on third parties. It can therefore overlap with the objective of financial stability (De Groen, 2017, Merler, 2017. The SRB has the discretion to make an ad hoc assessment based on these objectives, which may lead to controversial assessments of the public interest, as shown above. The comparison between Banco Popular in Spain and the Veneto banks in Italy is insightful in this regard (Monti, 2017), 8 as in almost identical notifications the SRB concluded with different assessments of the public interest. Moreover, and following the SRB's negative assessment in the Veneto banks case, the Commission effectively made a different interpretation by later approving that state aid could be offered to these banks in liquidation (European Commission, 2017c). The statement of the Intesa Sanpaolo CEO that if it weren't for Intesa, financial stability concerns would have arisen, seems to had added to the aforementioned controversy (Aloisi and Scherer, 2017). 9 Looking deeper into the Veneto banks case, the SRB stated that resolution was not in the public interest as: (1) the banking functions performed by the banks were not critical, as they were provided to a limited number of third parties and were capable of being replaced in an acceptable and timely manner, (2) a potential failure would be unlikely to cause significant adverse effects on financial stability due to low interconnectedness, and (3) normal Italian insolvency proceedings would achieve the objectives of resolution just as well as resolution itself. The main features of the final liquidation scheme include a partial transfer of the businesses of the banks to Intesa Sanpaolo, the entry of the banks into insolvency, and a cash contribution of €4.785 billion by the Italian government to Intesa Sanpaolo to help it absorb the losses from the acquisition. State intervention was different, however, for MPS and the Veneto banks. In the case of MPS it took the form of a capital injection, while in the case of the Veneto banks its contribution was a direct cash payment to Intesa Sanpaolo. In the first case there is still hope for repayment, perhaps at a higher value, while in the latter the cash will not be repaid and there is no possible advantage.
The Veneto banks case looks controversial when we consider the alternatives available within the EU resolution framework (Champsaur, 2017). The first is a replication of the MPS solution, precautionary recapitalization. The SRB determined that precautionary recapitalization would not constitute an adequate measure on its own, as additional private funding would be required for the government to avoid covering possible future losses. The second alternative would be to apply a proper resolution according to the EU framework, similar to the Banco Popular case. The difference here has to do with the buyer. Santander agreed to buy Banco Popular without making a distinction between good and bad assetsit acquired everything for €1. On the other hand, Intesa Sanpaolo required full reimbursement for the prospective losses on its tier 1 capital before proceeding with the acquisition. Another solution would have been to apply a full bail-inup to 8 per cent of liabilitiestogether with the sale of assets to Intesa, packaged with public support in the form of government financial stabilization tool. However, this solution would have meant bailing-in senior creditors as well as junior bondholders. A fourth solution would have been to try and exclude the bail-in of retail investors while applying it to the rest of the creditors (SRMR, Article 18[3]). However, that option involved a high risk of litigation due to the preferential treatment of certain creditors. The fifth option would have been for the Italian banks to enter into insolvency proceedings according to Italian law, provided that the SRB declared that the resolution of the two banks was against the public interest. Insolvency defuses the political resistance of both groups.
Two days after the SRB's decision, the Italian government adopted the Italian Decree, according to which extraordinary and urgent measures could be taken on the treatment of bank failures, subject to approval by the Commission under the state aid rules. The provisions of this Decree provided for the forced transfer by the insolvency administrator of the banks' businesses to the higher bidderexcluding subordinated debt and outstanding litigation claims, combined with the provision of public support by the Italian government to the higher bidder in the form of a €4.785 billion direct cash payment, as well as ancillary guarantees of up to €11.2 billion. The excluded liabilities and the NPLs were transferred to a risk management company. The subordinated bonds issued by the Veneto banks were not transferred to Intesa, which, however, paid €60 million restitution to the small depositors that held such bonds, with another €140 million being paid, if needed, by a dedicated fund managed by the Italian interbank deposit protection fund and funded by the banking sector. On the same day the Commission confirmed the Decree's compatibility with the EU state aid rules on the grounds of the extraordinary necessity and urgency to adopt measures allowing the orderly exit of the two banks while avoiding significant economic disturbance in the region. This solution was essentially a resolution masquerading as mandatory administrative liquidation, which allowed the provision of public support without fulfilling the precondition of the 8 per cent bail-in of senior and junior creditors. Put simply, the EU and Italian authorities agreed on liquidation with the provision of state aid.
Even though the national character of insolvency laws made this outcome possible, its legality is questionable. Prior to the BRRD and SRMR, governments used the only tool available to them at the time: state aid rules, coupled with the obligation of some form of burden-sharing. However, after the BRRD and SRMR were fully put into force in 2016, liquidation and public support have required a full 8 per cent bail-in. Otherwise such schemes raise significant legality concerns. The reason that they are legally controversial is not just that they stand in contrast to the rationale of the EU resolution framework, which is to prohibit public support unless a full bail-in is applied (including senior creditors, if necessary), but also in contrast to the 'no creditor worse off' principle (NCWO) that applies in bank resolution. Given the questionable legitimacy of the SRB as the EU agency competent to bail-in creditors, the NCWO principle stands out as the ultimate defence that permits effective legal protection against the decisions of the SRB. The NCWO principle requires that shareholders and creditors are treated no worse under resolution than under insolvency (BRRD, Article 74). Laws such as the Italian Decree result in changing the timing of 'normal insolvency proceedings' from the moment at which a resolution action is taken to a future post-resolution moment (Champsaur, 2017). At that point, post-resolution asset prices will have already plunged close to zero, thus making the NCWO principle and, in general, legal protection ineffective. Moreover, the fact that this Decree is an ad hoc legislative measure and the fact that it replicates EU resolution ruleswith the exception of the bail-in of senior creditorsmean that it does not qualify as a normal insolvency proceeding for the purposes of the BRRD and SRMR. Even the fact that the Italian Decree refers to the exact amount of cash needed, so that Intesa could acquire the Veneto banks (indeed, the name Intesa is not mentioned in the Decree) confirms that this decree was an effort to apply resolution tools without bailing-in the creditors, who were primarily retail investors.
Overall, from a legal perspective, the BRRD and SRMR offer ample room for discretion. The latter does not necessarily derive from the wording of the rules but rather from the very nature of resolution as highly complex and technical matter. Moreover, certain legal loopholes, mainly related to the inconsistencies between resolution and state aid rules, as well as the ones related to the SRB's public interest assessments, have allowed for some stretching of the enforcement of the rules. All this is combined with the national character of insolvency laws. This demonstrates the Italian government's willingness to circumvent the intent of European resolution law.

Discussion and Conclusion
This article has dissected two aspects of the approach of the Italian government and EU institutions to dealing with the MPS, Veneto and Vicenza banks, with a view to determining what these cases tell us about state motivations and the strength of EU institutions in applying the law on bank resolutions. The first section viewed decisions through the lens of Italian financial capitalism, in which junior creditors ensure the financial stability of banks and generate tremendous political incentives for Italian authorities to bend and break EU law. This allows them to protect and bail out stakeholders, particularly depositor bondholders. This further makes Italian banks unattractive to foreign investors. Outwardly, this has the same appearance and effect as economic nationalism. The Commission and the SRB's leniency toward Italy has facilitated this behaviour so far. The second section examined the legal structure of the BRRD and SRMR and the procedures for managing insolvent banks, and demonstrated that the adjustments demanded and made to the application of the BRRD bent and even broke the letter of the law.