Ignazio Visco: Asset quality and the new framework for managing banking crises in Europe

Keynote speech by Mr Ignazio Visco, Governor of the Bank of Italy, at the "Workshop on Stability of the Banking System", organised by the European University Institute (EUI) on the occasion of the annual conference "The State of the Union", Florence, 5 May 2016.

Central bank speech  | 
10 May 2016
PDF version
 |  4 pages

The sharp adjustment recorded in a number of financial market segments over the last few months somewhat reflects both worldwide and EU factors. Rising uncertainty on world growth prospects, the commodity price drops, persistent low levels of inflation and geopolitical risks may all have played a role. Italian and other European banks were hit hard, due also to concerns on asset quality and regulatory uncertainty, the latter in part, if not exclusively, related to the implementation of the Bank Recovery and Resolution Directive (BRRD), the new European bank resolution framework fully in place since the start of this year.

I will focus today on these two issues: asset quality and the new framework for managing banking crises in Europe.

Asset quality

Concerns about asset quality seem to focus on the valuation of the still large amount of non- performing loans (NPLs) in banks' banking books and of Level 2/3 assets in trading books, whose "fair value" is at least partly determined by banks' internal models and, then, regarded as relatively opaque by the markets.

In the case of Italian banks, market concerns about asset quality are to be taken seriously and should not be casually dismissed, even if there are good reasons to believe that they are somewhat overstated.

First, at 18 per cent of total loans on a gross basis as of December 2015, Italian banks' NPLs - whose value, net of provisions, is around €200bn - are certainly sizeable. It should however be observed that banks have on average a high level of guarantees against NPLs; real estate collateral, in particular, amounts to approximately €160bn. Bad loans (or "sofferenze", i.e. those to debtors in de jure or de facto state of insolvency, whose current net value amounts to €87bn) are more than fully covered by real estate collateral (€85bn) and other personal guarantees (€37bn). Provisions for NPLs have steadily increased since mid-2012; at 45 per cent, coverage ratios are now in line with the average for the main EU banks (44 per cent); they are at about 60 per cent for bad loans.

Second, the deterioration in credit quality, a legacy of the deep and prolonged recession, has been recently abating. In the fourth quarter of 2015 the flow of new NPLs fell, as a ratio to outstanding loans, to the lowest level since 2008 (3.3 per cent, compared with a peak of 6.0 per cent at the end of 2013). The rate of new bad loans is projected to decline further this year. The NPL ratio is finally peaking; a slight decline was actually observed in Q4-2015 for each of the top 5 banking groups.

Third, in recent months there has been a perception in the market (and in the press) that the Supervisory Board of the SSM (the single supervisor in the euro area) is pushing banks to rapidly dispose of their NPLs. Such perception might also have been among the drivers behind the significant fall of euro-area (and Italian in particular) bank stock prices. However, the perception is wrong, as the President of the ECB and the Chair of the Supervisory Board have recently, publicly, clarified. Indeed, an accurate and effective supervisory policy on NPLs must assess the situation at each specific bank. A policy that, instead, pushes banks across the board to rapidly dispose of NPLs in the market would result in a massive transfer of wealth from banks to private (speculative) investors that base their choices upon very high risk premiums and related returns.

I am convinced that dealing with systemic NPL issues in Italy necessarily requires a comprehensive strategy: improving banks' operational environment to manage NPLs; removing inefficiencies in the judicial and extra-judicial procedures that make the "time to recovery" longer than in the rest of Europe; kick-starting the NPL market, which is currently very thin and hence plays little role in resolving the NPL issue. On their side, supervisory authorities (both national and European) for some time have been strongly pushing banks to further improve their on-balance sheet management of NPLs.

In the course of last year it became clear that the establishment of a "bad bank" (or, more properly, an "asset management company" dedicated to the system-wide disposal of NPLs) with public support and/or participation would have been in contrast with the interpretation now given by the European Commission to rules on state aid. I had long been advocating the need for measures meant not to alleviate the difficulties of some specific bank but rather designed to reduce the impact of the deep and protracted recession that has hit the Italian economy since the global financial crisis: a solution that had been importantly adopted in other countries, evidently under a different regime and interpretation of the role of state aid rules.

As a consequence, last January the Italian Ministry of Economy and Finance has reached an agreement with the European Commission on a government guarantee scheme for senior tranches of securitised bad loans (the so called GACS). These guarantees comply with state aid rules as they are set up so not to be in contrast with what could be considered as market prices for NPLs, even if in Italy a deep and well established secondary market for NPLs does not exist. Therefore, the agreement marks a step toward the establishment of such a market, and also helps by ending the uncertainty accumulated over previous months.

Other measures recently adopted by the Italian authorities are also meant to address the root causes of the high stock of NPLs. Those on fiscal deductibility of provisioning have removed a penalising tax treatment of Italian banks. Last summer a first set of measures on bankruptcy and foreclosure proceedings were adopted to speed up credit recovery. The delays and widespread obstacles to court and out-of-court settlement of disputes are in fact a very important reason for the very high stock of NPLs in banks' balance sheets: a reduction of two years in credit recovery times could even lead to halving the level of bad debts as a share of the total. Feedback from market operators indicates that some effects are beginning to be seen.

In the last weeks, a private alternative investment fund (Atlante, managed by an independent Italian asset management company) has been first set up and then formally launched, with a twofold purpose. First, it is meant to support rights issues of Italian banks to reduce the risk that in the light of the current difficult market environment such issues would fail, with potential spillover effects to other Italian banks and ultimately to the entire economy (and there has already been a case where substantial support has been provided). Second, it could invest in bad loans (mainly junior and possibly also mezzanine tranches of securitised bad loans) at prices that may benefit from the organisation, expertise and servicing available within the banking system.

The fund has raised resources (more than €4bn) from Italian and foreign investors (banks, insurance companies and other institutional investors). Even if the resources available so far are somewhat limited, the setting up of the fund might well help to unlock the market for NPLs, therefore contributing - from its very start - to the solution of the NPL issue, a solution, however, that cannot but take a non-negligible amount of time. Being a private/market operation, it will be compliant with the current European regulation and consistent with the competition framework.

An important contribution to solving the NPL issue should also come from the new measures to speed up the recovery procedures just announced by the Government, which I very much welcome and on which a thorough evaluation of its likely quantitative impact is presently under way.

The new European framework for managing banking crises

Let me now turn to the challenges we are currently facing, in Italy like in other European countries, in dealing with weak banks, as a result of the financial and economic crisis.

The legislation on banking crises has two potentially conflicting objectives: maintaining financial stability - which leads to interventions in support of troubled banks to avoid systemic repercussions - and preventing banks from behaving opportunistically in the expectation of public intervention. Following the global financial crisis, the prevailing position at an international level has leaned towards the latter objective, far more so than in the past. This change in direction has certainly been influenced by the massive public interventions for rescuing banks that have weighed heavily on the state finances of many countries, in some cases jeopardising their fiscal sustainability.

In the European legislation there have been rapid, sweeping changes. In 2013 a Communication from the DG Competition of the European Commission provided for the immediate application of a new burden-sharing scheme which, in the event of a bank crisis, would impose losses not only on shares but also on subordinated bonds as a prerequisite for public intervention. In 2014 the BRRD, approved by the Council and the European Parliament, extended that scheme, starting from this year, to include also ordinary bonds and deposits of over €100,000 (i.e., the bail-in).

I have recently emphasised, also in the light of the (bad) experience we have gained in Italy with the resolution of four small banks last November, the importance of reconsidering the adequacy of the new legislative framework for managing banking crises. Just as the crisis has impaired the quality of bank credit, the legislative response on crisis management has engendered uncertainty about investing in bank liabilities. An instrument - the bail-in - devised to reduce the impact of a crisis must not create the premises to make one more likely: if this is the case, its design and/or its implementation must be rethought. We must strike the right balance: indeed, investors who have been hit by a bail-in will find no comfort in the fact that they have been protected as taxpayers. And we should not rule out the possibility of temporary public support in the event of systemic bank crises, when the use of a bail-in is not sufficient to achieve the resolution objectives but instead risks compromising financial stability.

I have not been alone in advocating such reconsideration. The need to assess the degree of flexibility of the BRRD during the review of the directive scheduled to take place by June 2018 was recalled, among others, by the IMF in its Global Financial Stability Report published a few weeks ago. It highlighted the necessity of applying the new rules (including those on state aid) with flexibility and caution during the changeover to the new regime, when public intervention is no longer admissible but the banks have not yet put in place sufficient buffers to absorb losses without undesired effects on systemic stability. A very recent CEPS paper by Micossi, Bruzzone and Cassella is a welcome contribution on these issues.

Indeed, in March 2013, during the BRRD negotiations at the EU Council of Ministers, the Bank of Italy and the Ministry of Economy and Finance presented a technical "non-paper" to all the other country delegations in order to illustrate the reasons for our preference for contractual (or targeted) bail-ins, to be applied only to newly-issued securities (i.e. not applied retroactively to securities issued under a different regulatory regime) that contain a specific clause in the contract recognising the power of the authorities to write down or convert the securities if the conditions for starting a resolution procedure were met.

We were - and remain - convinced that this approach would have given the authorities a credible tool to be used to resolve a bank crisis effectively, with no undesired effects on financial stability and at no extra cost to the taxpayer. On the contrary, according to the approach that has prevailed, a wide range of liabilities are subject to bail-ins, with very few exceptions, large corporate sight deposits included and a rather short learning time provided to creditors - especially retail ones. All this, in a world of asymmetric information, can indeed be a source of serious liquidity risk and financial instability.

The specific nature of the banking sector and the objective of financial stability should also be treated in a more structured way in the European Commission's approach to competition and state aid. Adopted in 2013 - and based on the unfortunately premature conviction that the difficulties of European banks had by then been dealt with - the current approach greatly limits government intervention to support banks that are fundamentally viable in order to remedy market failures or negative externalities that are not originated by managerial shortcomings.

Indeed, according to the BRRD, any public intervention that counts as state aid (and is not among the very few exceptions allowed by the directive) automatically triggers the resolution procedure. This, in turn, determines an overlap and, as a result, the prevalence of competition and state aid policy objectives over the objective of safeguarding financial stability. In other words, sufficient account is not taken of the fact that financial stability is of vital importance to the real economy.

*      *      *

To conclude, in the financial sector there are certainly risks of moral hazard as well as abuses and frauds, which the authorities, including banking supervisors, must prevent as effectively as possible. But prevention is not always possible and in periods of poor macroeconomic conditions risks of malpractice inevitably rise. This is why we need an overall resilient system, with banks well capitalised, better managed and more transparent, costumers with higher levels of financial education and authorities with severe sanctioning powers. This notwithstanding, we have learned that market failures are not infrequent, and information asymmetries and expectations' coordination problems may trigger sudden and possibly long waves of euphoria or panic.

It is by now clear that markets are not always self-equilibrating and when they fail the consequences may be rather serious. On the other hand, markets should operate as much as possible without their force being bridled. There is no foolproof formula for striking this balance. Clear rules, impartially applied, are essential, but so is the ability to make sound and timely decisions in the light of circumstances. In the case of banks, this calls for the possibility to recur to public backstops, and in a Union like ours also a supranational one, in the presence of systemic risks and risks of contagion. This is something that still differentiates, contrary to a new apparently established belief, banks from enterprises operating in different industrial sectors.

To repeat something that I said a year ago, before the start of the current debate on the consequences of the new regulatory set-up, substantial observance of the rules on market protection and competition remain indispensable. However, in assessing the public role in the prevention and resolution of crises, and not only financial crises, I believe that greater consideration should be given to the characteristics that distinguish policies designed to activate market mechanisms from state aid that distorts competition.