The SSM sails past the starting line: Seeking high-quality supervision and level playing field

Remarks by Ignazio Angeloni, Member of the Supervisory Board of the European Central Bank,
International conference on “Start of European Banking Union: Perspectives and Challenges” London School of Economics and Political Science,
London, 10 November 2014

Introduction ^[1]

It is a pleasure to be here today and to speak at this conference. For me it was a special honour to receive the invitation directly from Charles Goodhart, who was a constant source of inspiration and learning throughout my career.

A week ago, the Single Supervisory Mechanism (SSM) assumed its duties as banking supervisor for 18 European countries. Some observers have likened this date, 4 November 2014, to another one, 4 January 1999, the day the euro came into being and the ECB became operational. But the two circumstances are different. The single currency was 30 years in the making, with long-term goals – supporting European integration by enhancing monetary stability – and was launched in favourable economic conditions. The single supervision was conceived and implemented in a very short time (28 months to be precise), amid financial instability and very real danger of a disintegration of the euro. For this reason, the project carries very high expectations, which go hand in hand with equally high responsibility. The ECB is expected – perhaps I should say required – to succeed quickly, or else fail without appeal. I am reminded of the words Dante says to himself when he is about to undertake his voyage: “Qui si parrà la tua nobilitate”, which means: “Here will appear what you are worth”.

The launch of the SSM followed a period of intense preparation in which I had the honour of being involved. As with a sailing vessel passing the starting line is not a beginning but a step in a process, preceded by rigging out, training and planning, so it was for the SSM. The ECB and the national supervisors started preparing in July 2012, right after the political decision. The ECB worked closely with the Commission in the summer of 2012 when the latter prepared, in record time, a draft Regulation. The EU Council and the European Parliament introduced several changes in the draft before its adoption on November 2013. This year, the ECB has adapted its internal rules to accommodate the new function, recruited several hundred supervisors and staff in related areas (legal, statistics, IT …), created four supervisory departments and a secretarial unit within its organisational structure, and lastly – but definitely not least – conducted the check on bank balance sheets that goes by the name of “comprehensive assessment”.

Describing even concisely this preparatory work would take more time than I have today. Moreover, I am convinced that to meet any challenge one needs to focus on the future and not be distracted by thoughts of past achievements or failures. Accordingly, my focus will be on what comes next, on the goals the SSM should set itself in the initial stages and on the pitfalls to be guarded against. I clarify up front that, although I am always inspired by exchanges with other Supervisory Board members, the ideas presented here are personal and should not be attributed to them.

To begin, I propose that, with some simplification, we encompass the main immediate goals the SSM faces under two broad headings: (1) the quality of supervision, (2) the level playing field. The rest of my remarks will elaborate on this and be organised as follows. First, I will briefly examine the first and simpler notion, the “quality of supervision”. I will then move to the second and more complex question of how to interpret and how to achieve the “level playing field” in a multi-country supervisory system like the SSM. Next, as an illustration, I will use two examples in areas that are today at the centre of supervisory debates. Finally, I will conclude with some considerations regarding the overall level of banking risk.

The quality of supervision

The SSM Regulation ^[2] puts a good deal of emphasis on ensuring a high quality of supervision, an indication that the legislator was concerned not just about breaking the link between banks and sovereigns ^[3], but also about improving supervisory practices. The ECB should ensure that “credit institutions are subject to supervision of the highest quality, unfettered by other, non-prudential considerations” ^[4]. In order to ensure “high standards of supervision”, a concept I regard as equivalent to quality, the ECB can use all instruments provided by Union and national law, also exercising the options that Union law assigns Member States ^[5]. For the same purpose, the ECB can decide to extend direct supervision to less significant institutions, potentially bringing under its direct competence the whole banking sector ^[6].

How can we make this quest for “quality” more concrete?

As a minimum, the transition to the new system should not in itself entail a worsening of supervisory attention, due for example to a loss of relevant information or experience. This problem is more serious than it appears. The national supervisors (called, in our jargon, the “national competent authorities” or NCAs) have accumulated a wealth of knowledge on their supervised entities, often intrinsically linked to the individuals exercising the function (examiners, off-site analysts, etc.) and not easily transmissible. To include these elements, the SSM structure is built around “Joint Supervisory Teams” – groups of ECB and NCA supervisors – one for each significant bank, acting under ECB coordination. Those teams, supported by mixed on-site teams, will conduct the annual supervisory cycle and provide analysis and information to the Supervisory Board, which will in turn prepare decisions on individual banks ^[7].

Second, the new authority will adhere to the highest standards of independence, transparency and accountability. The SSM Regulation is very clear in placing the new supervisor at the forefront of international best practices in all three areas. The European Parliament will exercise close scrutiny. As a supranational authority overseen by another one, the ECB will be in a good position to distance itself from any national concerns, constraint and pressures.

Thirdly, high supervisory quality today means absorbing the main lessons of the recent crisis. This means also integrating traditional micro-supervision with a macro-prudential dimension ^[8]. The SSM Regulation covers the macro-prudential function explicitly, assigning to the ECB a shared responsibility with the Member States. The national authorities have the power to set the macro-prudential instruments foreseen in EU legislation (CRD IV and CRR) after consulting the ECB, and the ECB can adjust these instruments by making them more restrictive, after consulting the national authorities. The ECB is preparing modalities to make this policy interplay operational, establishing forms of dialogue and consultation with the Member States.

The level playing field

The second “frontier of performance” I mentioned is the level playing field. Again, there are several references to it in the SSM Regulation. The statutory goal of the new institution is defined in Article 1, as follows:

“… contributing to the safety and soundness of credit institutions and the stability of the financial system within the Union and each Member State, with full regard and duty of care for the unity and integrity of the internal market based on equal treatment of credit institutions…”.(my emphasis)

Prudential supervision should be “implemented in a coherent and effective manner” and the single rulebook should be applied “in the same manner to credit institutions in all Member States concerned” ^[9]. Level playing field consideration should also influence the choice of banks that are considered “significant” and hence supervised directly by the ECB ^[10].

This concept is even more difficult to define than the previous one. Banks covered by the SSM differ widely in business models, asset and funding structures, sector and geographical exposure, client mix, internal organisation and governance, and so on. Countries vary in economic conditions and legal arrangements. How can one take these differences meaningfully into account and yet practice “equal treatment”? What is the meaning of “level playing field”, say, between a retailer serving small and medium enterprises and individual depositors, and a large player with global reach and large market exposures? In many cases, banks so diverse will never compete against one other. Whichever the “field” that is supposed to be levelled, banks may never “play” in it at the same time.

There are, of course, regulatory requirements applying to all, which the supervisor is supposed to monitor and enforce. They consist, first and foremost, in capital requirements, including rules to define admissible capital instruments and to calculate assets weighted according to risk. Other rules are coming soon in other areas, like liquidity buffers and funding structures. In addition, the leverage ratio will need to be disclosed as of 1 January 2015 and the SSM already takes into account the leverage ratio as a supervisory pillar 2 measure. As to a legally binding standard (pillar 1) the Commission will propose a legislative proposal at the end of 2016, applicable from 1 January 2018 onwards. Without underplaying the importance of the legal regulatory basis, however, these provisions in themselves will not guarantee a prudential level playing field. No matter how detailed (and many people think that they are already too detailed and complex) they can never cover the full range of risk drivers. They are minimal criteria, above which supervisory judgement can and should be exercised.

On what basis can this judgement be made? Here the logic of the banking union helps us. To neutralise the linkage between banks and sovereigns, the European leaders decided to establish, with some sequencing and gradualism, a common safety net and, in parallel, a central supervisory authority. The two elements guarantee incentive-compatibility of the whole. In a single country, taxpayer-funded safety nets generate excess risk-taking by banks, which needs to be corrected by delegating risk control to an independent agent, the banking supervisor ^[11]. The multi-country dimension adds another layer to this moral hazard problem. A single supervision without common backstops is unsustainable, because national taxpayers are exposed to the mistakes of others. But the opposite – common backstops without single supervision – is not viable either, because it creates an incentive for supervisory laxity at the national level. In the banking union, single supervision and common backstops are complements that balance and justify each other.

Based on this reasoning, I am tempted to suggest that a basic criterion for defining the level playing field in our context be identified as a condition in which banks transmit, at the margin, the same amount of risk to taxpayers, regardless of the jurisdiction they are located in and of all their other characteristics. It is easy to see why this criterion is necessary. If somewhere in the system (say, in a given country) the risk borne by taxpayers is higher than elsewhere, banks receive an implicit public subsidy that distorts the competitive playing field. If the safety net is mutualised (as will be the case of the Single Resolution Fund that I will mention in a minute), the risk spills over onto other countries. The same criterion can be applied to other specific categories of banks. Differences can be tolerated in particular circumstances, for example for financial stability reasons or because certain bank exposures are recognised as having priority for economic or other reasons ^[12]. Defining those circumstances should not be a responsibility of the banking supervisor, except where there are risks to financial stability – remember that the SSM mandate, defined by Article 1 of the SSM Regulation, includes systemic stability. Although necessary, this criterion is not sufficient, because it covers only competitive distortions generated by the public sector. Others can arise from the structure of banking markets. Detecting all instances of competitive distortions falls under the competence of the competition authority, which in the EU resides in the Commission.

Measuring the taxpayer impact of bank exposures and risk profiles is a very difficult task. State-aid needs not entail much or any cost for public budgets over time if it is repaid within a reasonable time and yields adequate return. In all circumstances, however, there is a risk ex ante. A look at the data on state support to banks during the crisis, provided by Eurostat and the Commission (DG-Competition) ^[13] is sufficient to convince us of the huge dimension of the problem. If one considers the period 2008-12 (a special period, but a relevant one), the total amount of aid provided to the financial sector in the euro area, in the form of recapitalisation and asset relief, amounted to €456 billion, or 4.8% of euro area-wide GDP. The distribution of this amount was very uneven: among the large countries, at one extreme we find Germany (€144 billion, or 5.5% of national GDP) and Spain (€88 billion; 8.1%), at the other France (€26 billion; 1.3%) and Italy (€6 billion; 0.4%) ^[14].

If one looks at the impact on budgets deficits, a better approximation to taxpayer cost, government interventions are equally sizeable. On average, over the same 5-year period the government deficit increased in Germany by €41 billion, or 1.6% of GDP, and in Spain by €40 billion, or 3.7%. In France and Italy, the impact even switches sign: budget balances actually improved as a result of bank support operations.

By way of comparison, in the United Kingdom the total amount of aid in the form of recapitalisation and asset relief in that period was €123 billion, or 6.6% of GDP; the public deficit increased by €8 billion, 0.4%.

Banking is an inherently risky activity. Since it provides a positive externality to the economy, it cannot and should not involve a zero risk or zero cost to society. I shall return to this in my conclusions. What is important to note here, before I move on, is that the size and distribution of such costs depends not only on bank risks but also, heavily, on the crisis management framework – state aid rules, resolution regime and funding. This is an area where dramatic changes have taken place in Europe in recent times:

1. The state-aid framework has become stricter since 2013: all junior debt should be burned or converted into capital before any public aid is granted.

2. The bail-in provisions of the Bank Recovery and Resolution Directive (BRRD), in effect from 1 January 2016, are even stricter, involving senior debt as well. They envisage the following creditor hierarchy: subordinated debt, senior creditors and uninsured depositors, of which small depositors come last. Insured deposits (under €100,000) are never bailed in.

3. The Single Resolution Fund (SRF) of the Single Resolution Mechanism will start operating in 2016. National funds established under the BRRD will gradually be pooled together, thereby mutualising the fund over a number of years. 60% will be mutualised in the first two years.

My discussion of level playing field stops here. I will now discuss two examples that may help clarify the concept further. The examples are the following:

1. National options and discretions in bank regulation;

2. The prudential treatment of sovereign exposures.

National options and discretions

National options or discretions arise in European legislation (the Capital Requirements Directive IV, or CRD IV, and the Capital Requirements Regulation, or CRR) when Member States are allowed to choose between alternatives prescribed by the legislation itself, or when they can decide whether to follow or not certain provisions. Options and discretions are either general or individual. National supervisors should disclose cases in which general options and discretions are exercised, including the way in which they are exercised. The European Banking Authority (EBA) monitors national practices across the European Union and publishes summary tables on its website ^[15].

Altogether, there are 103 such options and discretions in the CRD IV and CRR, covering a variety of subjects: the definition and level of own funds, the treatment of sovereign exposures, the definition of public sectors, large exposures, liquidity regimes, various types of waivers to prudential requirements, etc. Importantly, neither the Directive nor the Regulation discusses the rationale of such provisions or binds Member States to converge; discretion is full and unconstrained within the boundaries specified by legislation ^[16].

Specifically, the European framework contains options and discretions for the transition of own funds to the fully phased-in Basel III definition and the exemptions from limits to large exposures. National practices diverge in the way they remove the so-called “prudential filters” from their capital frameworks: while international standards provide that unrealised gains and losses arising from the fair value measure of assets and liabilities be accounted for directly in CET1 capital, CRD IV/CRR allows for gradualism (“filters”) to smooth out the impact of the new regime. Some national authorities have decided to phase out their filters progressively, whereas others have already removed them fully. Although temporary (the minimum ratios will converge to the Basel standard in 2018), the discrepancies are substantial and have affected the results of the comprehensive assessment as well ^[17].

In addition, some options have a more permanent effect; for example, national authorities may permanently decide to allow banking conglomerates not to deduct certain holdings in insurance subsidiaries from their regulatory capital, using the FiCoD ^[18] rules, while the general rule is to either deduct such holdings or consolidate the relevant subsidiaries for capital requirement purposes.

The practice of allowing these margins of flexibility is by no means new and is not of European origin. The Basel Accord provided for such margins since the beginning. Basel I established the general principle that national authorities should always be free to apply more stringent criteria than those set by the international standard. This principle survived all successive versions of the Basel Accord and is also at the basis of the CRD IV/CCR framework ^[19].

We should be aware that this type of flexibility is not a free good. While reasonable at first sight, because it meets the will of more risk-averse constituencies and apparently tilts the framework towards prudence, in fact there are risks. By allowing individual nations to depart from the common rule in the direction of strictness, this type of discretion tends to weaken the support for stricter criteria equal for all. The consequence is a bias towards laxity in the minimum standards, to the detriment of financial stability (at least locally) and the level playing field. Basel II broadened the scope for such forms of discretion by making the framework more decentralised and responsive to specific national market characteristics as perceived by the supervisors ^[20]. Basel III inherited this trend: the latest version of the Accord contains 97 options and discretions to be applied by national competent authorities, mainly regarding the internal ratings-based approach for credit risk (33) and the standardised approach for credit risk (24). At present, 76 of these 97 options are under review by thematic working groups within the Basel Committee.

It should be clearly recognised that inconsistent application of national discretions, especially when leading to cross-border discrepancies in the level and the quality of capital, increases the potential reliance of banks on external support in certain constituencies relative to others. If unqualified and unconstrained, such latitude conflicts with the logic and the proper functioning of the banking union. For these reasons, reviewing it and agreeing on common approaches should be a priority for the new supervisor. Joint work with the EBA will be essential. The Single Rule Book needs to be developed further to deliver binding technical standards and guidelines and harmonised criteria for the application of options and discretions.

Let me mention some ways that can help us move in the right direction.

First, if national options and discretions are not to be eliminated altogether, at least a principle and practice should be established that recourse to them within the SSM area should be openly justified in the light of clear, pre-specified criteria. Transparency alone – simply relying on market forces to provide discipline – will not solve the problem. Capital measures need to be understandable, implementable and certain, and not dependent on national implementations or interpretations. As national regulators integrate international and European standards into their own rule books, they should be aware that deviations not justified by genuine risk-based specificities will eventually require compensating adjustments to the framework, leading to new distortions.

For supervisors to spell out the criteria and to provide justifications for national discretion, and for markets to assess them, we need a better understanding of the underlying phenomena. From this viewpoint, the virtual absence of research on the determinants and the impact of national options and discretions is disappointing. More analysis is needed. It would be important, for example, to understand what have been, in the recent experience, the prevailing motives behind the use of national discretion (competitive advantage, forbearance, simple inertia or other factors) and how they have affected bank behaviour, banks and bank risks ^[21].

Second, the exercise within the SSM of national discretions and options foreseen in CRD IV/CRR should be actively coordinated and guided, not only monitored. As mentioned already, in accordance with the SSM Regulation the exercise of national options and discretions is, since last week, entrusted to the ECB, which inherits this function from the NCAs. A difficulty may arise in cases where such discretions have been translated into national laws, as the SSM cannot formally override them ^[22]. As a non-lawyer, I find it surprising that national law can contradict in substance, if not in form, a European code like the SSM Regulation, without being overridden by it. Be that as it may, the SSM can contribute to the design of a harmonised legal framework and use its prudential powers to mitigate any remaining asymmetries. At present, further work on national options and discretions is taking place at the Basel level as well, to reduce them in number and materiality.

The treatment of sovereign exposures

My second example concerns the prudential treatment of sovereign holdings.

Traditionally, bank regulators and rating agencies have considered these exposures, if expressed in domestic currency, as free of any credit risk in nominal value, based on the idea that sovereign states can always honour their obligations by creating money. This assumption evidently presupposed the absence of an independent central bank. The euro crisis, which occurred in an area where the central bank is not only independent but also supranational, has shown beyond doubt that the risk-free assumption is widely and increasingly remote from reality.

In spite of this, prudential regulators still treat sovereign exposures as largely risk-free. In the standardised approach, exposures funded or denominated in the domestic currency receive a 0% weight, which means no capital is required to cover their risks. For institutions relying on internal ratings, there is no regulatory floor to default probabilities, which implies risk-weights of around 2% to 3% and capital requirement ratios of close to zero. Even large banks applying internal ratings are granted an exemption by EU legislation, allowing them to hold government bonds under the standardised approach without capital charge for sovereign risk. According to the 2011 EBA stress test report, only 36 out of the 90 participating banks applied their own internal model to sovereign risk, a smaller ratio than for other regulatory portfolios. If this wasn’t enough, sovereign issues benefit also through other channels, for example regarding the treatment of derivatives, concentration risks and liquidity requirements ^[23]. National discretions allow supervisors to treat certain public entities as equivalent to the central government.

The methodologies of the ECB’s comprehensive assessment gave sovereign exposures a more proper prudential treatment, without deviating from EU law. In the asset quality review, fair value adjustments were made on portfolios incorrectly classified as held-to-maturity. The stress test allowed for rating downgrades with related provisioning in the adverse scenario, as well as fair value adjustments in line with market conditions on available-for-sale holdings.

In general, however, the preferential treatment allowed to sovereign issuance, especially if combined with a pronounced home bias in sovereign holdings ^[24], opens the way to important deviations from a level playing field. If the safety net is partially mutualised, as in the SSM area going forward, the distribution of taxpayer risk across borders is distorted: countries where sovereigns are more risky transmit, via bank domestic sovereign portfolios with insufficient capital coverage, negative risk externalities to the rest of the area. Holdings of risky sovereigns are also incentivised in this way. This prudential treatment conflicts with the balanced functioning of the banking union.

This problem is increasingly recognised as such, and some gradualism is probably necessary anyway, but at present progress is exceedingly slow. The Basel Committee is starting new work on the subject. Going forward, the introduction of a leverage ratio, where sovereign exposures are included in full, should mitigate the problem somewhat: a leverage ratio requirement of, say, 3% is equivalent to a 35% risk weight at a capital ratio of 8.5%. From a supervisory perspective, in the meantime, the experience of the comprehensive assessment demonstrates that stress testing methods can help increase the risk sensitivity of sovereign exposures.

Conclusions

28 months after the political decision was made, the SSM structures have been established and the preparations are complete. We should now look ahead, setting future goals and standards.

Establishing a level playing field across borders is, in many respects, the novel challenge for the SSM. The international dimension of the problem requires new concepts and methods, for which I have proposed some ideas and examples. The rest resides, I would say, in the realm of high-quality supervision: a more traditional endeavour, though still a challenging one.

One last note before I conclude. Ensuring that taxpayer risks are balanced still does not answer a more fundamental question, namely: How large, overall, should those risks be? Banking – or indeed the whole of financial intermediation – poses risks for society that cannot and should not be eliminated fully. But how large should they be? While I am not even trying to approach this question now, I feel that the distributional issues involved here are such as to move us from the realm of delegated authority to that of politics, where supervisors and central banks should not be in charge. It is desirable that reflection should start with public opinion, eventually involving political decision-makers also at European level. Everywhere in the world, the crisis has made people aware of the importance of this issue. Supervisors and central banks can help the debate by providing correct technical information.

Thank you for your attention.