60 years on: promoting European integration in the banking union

Speech by Ignazio Angeloni, Member of the Supervisory Board of the ECB, at the Università Bocconi, Milan, 24 March 2017

I am grateful to Bocconi, my alma mater, for inviting me to this ‘Executive Master in Finance’^[1]. Coming to Bocconi brings back fond memories. In its classrooms, including this one, I learned my first notions of economics. Within these walls I decided that my profession would involve economic policy – though, to be frank, back then I didn’t have a very clear idea of what this really meant. Even less did I suspect, at that time, that for much of my professional life the relevant geographical frontiers would not be the ones of my country, but those of Europe. In 1977, when I graduated here and left for the United States for my graduate studies, a single European currency was not envisaged; the Werner Plan had been shelved, most European currencies were floating and only a few of them were pegged to the Deutsche Mark in the so-called “snake”. The lira had undergone a dramatic devaluation only the year before.

Whence we come

Yet, European integration (at least on this side of the Iron Curtain) had been a very real thing for at least 20 years; specifically, since the signature of the Treaty of Rome, the 60th anniversary of which is being marked tomorrow. The Treaty of Rome was the beginning of a common European journey. Its central goal, peaceful coexistence with free movement of goods, people, services and capital, was achieved over many years, in steps: reducing tariffs, removing quotas, encouraging competition, liberalising capital, favouring the movement of workers and students across frontiers. Diplomacy, compromise, and all the other things that detractors of the European project today misleadingly call “bureaucracy”, became the order of the day in place of centuries of armed conflict. A promise of peace was made, and it was kept.

Many years afterwards, the Single European Act was the second main step in the process. In 1985, a Commission document identified major obstacles in the establishment of a true single market^[2]. Following up on that analysis, the Act stipulated that non-tariff barriers would be removed so that a genuine single market would be established by end-1992. The European passport for the provision of financial services, a single banking licence and the principle of “mutual recognition” (introduced by the European Court of Justice in 1979) were established at that time. They were the building blocks of what would eventually become the banking union.

The third major step, the monetary union, emerged in that period as a way to reinforce the Single Market. The idea of a single European currency was not new: aired for the first time by Gustav Stresemann in 1929, it had resurfaced after the war, without ever finding proper conditions or sufficient support for its implementation. In 1988, a new committee led by Jacques Delors was given the task of “studying and proposing concrete stages leading towards this (economic and monetary) union”. The committee’s blueprint drew heavily on the Werner Plan and, unlike that plan, Economic and Monetary Union indeed became reality, the main lines of the Delors plan becoming part of the Maastricht Treaty signed in 1992. In it, a deadline of 1998 was set for the creation of the single currency. Although the Treaty has since been amended by the Treaties of Amsterdam, Nice and Lisbon, Maastricht’s fundamental tenets have remained broadly unchanged.

The monetary union sowed the seeds for a fourth major step in European integration. In one of its articles, the Treaty stated that “the Council may, acting unanimously on a proposal from the Commission … , confer upon the ECB specific tasks concerning policies relating to the prudential supervision of credit institutions and other financial institutions with the exception of insurance undertakings”.^[3] That article struck a compromise between different visions, one arguing for an inherent link between monetary and banking policy, the other seeing them as different and in fact potentially conflicting. Especially with the benefit of hindsight, the underlying rationale for involving the ECB appears straightforward: for a single currency to work properly and monetary policy impulses to be transmitted smoothly in the whole area, banking markets cannot be segmented across national boundaries. In order to prevent this, a single supervisor, a single safety net, and a single set of banking rules must exist.

ECB Banking Supervision within the banking union

In spite of this logic, it took another 20 years and a major global and European financial crisis for a “banking union” to see the light. The circumstances which led to it were rather unique.

For a long time, it was felt that the mismatch between the domestic span of supervision and the cross-border nature of finance could effectively be handled through coordination among supervisory authorities. This line of thinking also downplayed the consequences of the discrepancy between the supranational scope of the ECB’s monetary policy and the national scope of the main channel of transmission of this policy, i.e. the banks.

The crisis exposed the failure of this line of thinking in two ways. First, it highlighted the existence of channels of international transmission, especially within the euro area, which could not be adequately controlled by a host of separate authorities, even supported by arrangements for cooperation and information exchange. Second, the interaction between national supervision and national fiscal policies – again, in spite of the coordination arrangements in place – created, in a number of weaker countries within the monetary union, tensions capable of putting both banking stability and sovereign solvency into question, ultimately threatening the financial integration of the euro area and even the cohesion of the single currency.

In June 2012, the euro area Heads of State or Government asked the Commission to present a proposal for a Single Supervisory Mechanism (SSM), or, as we often call it, European banking supervision. Though their statement did not explicitly refer to the banking union or to any other “pillars” of it, there was an implicit reference to the fact that supervision was a precondition for bringing to the European level other elements of the banking framework, notably in the area of crisis resolution. Those additional elements have only partly seen the light of day. A single resolution authority was established in 2015, and assumed full responsibilities in 2016. Conversely, a single deposit insurance, though having been recognised as an integral part of that construction^[4], does not exist yet, nor is there an agreed time frame for it.

ECB Banking Supervision was given by the European legislators two statutory goals: (i) the safety and soundness of credit institutions, and (ii) the stability of the financial system in the Union and in its Member States. Promoting financial integration, a related objective, is not indicated as a direct goal but is mentioned as an indirect one in the preamble of the Regulation establishing the SSM.^[5]

ECB Banking Supervision has operated for a little over two years now. The achievements in this short time are remarkable, starting from its resources and organisation. The ECB can count on a first-rate supervisory staff, recruited partly from the private sector and partly from the national supervisory authorities. A large number of staff members are organised in the Joint Supervisory Teams, or JSTs, each focused on the supervision of one banking group. The JSTs include ECB staff, who coordinate, and national supervisors, who have years of experience in overseeing those banks. A mix of nationalities ensures that any risk of bias within the JSTs is avoided. Most of the other ECB staff members are employed in horizontal supervisory functions – risk analysis, quality assurance, crisis management, supervisory policy, and so on – and are responsible for ensuring a level playing field across banks and countries. Again, the horizontal units are structured so as to contain a mix of expertise, backgrounds and nationalities.

At a higher level, the Supervisory Board (SB) has grown into a cohesive and effective decision maker. In addition to the Chair and the Vice Chair, the SB includes four representatives of the ECB; I have the privilege of being one of them. The other 19 members are nominated by the national authorities of the participating Member States and have one vote each. The members are statutorily independent and cannot seek or receive instructions from external parties (including, for the national members, their affiliated institution), and are bound to serve in the interest of the Union as a whole.

Before examining some of the more pressing current challenges, let me briefly mention two other major achievements of the first two years.

The first is the SREP (Supervisory Review and Evaluation Process), the methodology we use to measure the riskiness of banks and to set the supervisory capital requirements (Pillar 2) and the other prudential requirements. The SREP encompasses all main risk factors relevant for each bank, assembling them in a holistic judgement expressed in numerical scores. The underlying risk analysis is partly data-based and partly judgemental, and is built from three main information sources: the JST, the banks themselves and the results of stress tests. We developed this system anew, starting from international best practices, and used it for the first time in 2015. The SREP plays a decisive role in ensuring high-quality supervision, a level playing field as well as consistency and transparency of supervisory decisions.

The second achievement is the harmonisation of the regulatory options and discretions existing in European banking law, which we completed towards the end of 2016. With this project, all 19 national supervisory authorities have agreed on how to apply the margins of flexibility granted to them, prior to the establishment of the SSM, in a harmonised way.

Present challenges

Let me briefly outline three broad challenges^[6], which are among the most important ones facing banks and supervisors.

Challenges arising from economic developments

A first set of challenges arises from the interaction between the prevailing economic environment and bank business models. In a context where interest rate margins typically account for over half of the total income of large euro area banks, a prolonged period of low interest rates has evidently weighed down on bank profitability. The impact of low rates has been far from uniform across banks, however, as it depends on the underlying structure of assets and liabilities as well as on their ability to adapt to the new environment.

Concerning bank balance sheets, our analysis suggests that the ECB’s non-standard monetary policy measures have had a positive impact through various channels, and that this impact has tended to offset the decline in net interest income^[7]. Lower funding costs have not fully offset lower interest income, as deposit rates have tended to be “sticky” at low levels. However, banks have benefited from capital gains on sovereign bonds, from the increased volume of financial intermediation and improved credit quality and also, indirectly, from the improvement in the general economic outlook induced by the monetary easing.

As regards banks’ ability to adjust, evidence shows that some euro area banks have consistently outperformed their peers in terms of bank profitability in recent years^[8]. The most evident common trait among best performers is cost efficiency. Average cost-to-income ratio for large banks in the euro area is around 65%, but there is a very large discrepancy between best and worst performers. This suggests that there is still scope for improvement for many banks.

Looking forward, the prospect of a normalisation of the interest rate cycle will bring benefits but also some risks, depending on the timing and speed as well as the banks’ preparedness. Retail banks may look forward to higher revenues from traditional intermediation, but these effects might take time to work through. Temporarily, higher funding costs might accrue before banks are able to benefit on the asset side. Banks may also face a trade-off between higher market risk and lower credit risk.

In this area the ECB is currently engaged in two exercises. First, we are conducting a thematic review of banks’ business models and profitability drivers, also exploring the risks for banks’ business models emanating from fintech and non-bank competition. In this context, the ECB challenges and assesses the consistency of banks’ business models relative to banks’ stated objectives, but does not dictate or prescribe business models, the choice of which remains the responsibility of shareholders and managers. Second, the stress test to be conducted in 2017 on our supervised banks will focus on interest rate risk in the banking book, analysing how different interest rate shocks affect bank assets and liabilities valued at amortised cost. The results of the exercise will feed into our annual assessment of how much capital banks needs to hold (SREP 2017), specifically informing the assessment of so-called Pillar 2 requirements and Pillar 2 guidance.

Challenges arising from balance sheet fragilities

A second set of challenges for banks and supervisors stems from lingering balance sheet fragilities. They relate to both credit and market risk.

Concerning credit risk, the main problem, as is well known, is that of non-performing loans, or NPLs. NPLs, partly a legacy of the recession, are a major hurdle in post-crisis European banking. NPLs in the euro area are very large: in gross terms, they amounted to about €900 billion (third quarter of 2016), or 6.6% of total exposures; net of provisions, they amounted to 3.6% of total exposures. The problem is concentrated in a few countries,^[9] some of which are characterised by weaker economic growth and fiscal positions – hence, they are relatively more exposed to a negative loop between public budgets and banking fragility. There has been a modest improvement recently, with average NPL stocks falling by €66 billion in the year to the third quarter of 2016.

The ECB has approached this problem in steps, starting with a fact-finding phase, leading to a public consultation. As a result of that, this week we have published a guide to banks on how they should tackle NPLs, addressing key aspects regarding the strategy, governance and operations relating to bad loans resolution^[10]. The guidance promotes consistent recognition, provisioning and disclosure for NPLs, and also requests banks to define and implement policies to reduce NPLs in a quantitative manner. JSTs have already started to liaise with banks to this end. We are aware that the disposal of NPLs takes time, and that there may be a trade-off between speed of disposal and recovery values; but we are also convinced that delaying the clean-up of balance sheets unduly prolongs bank fragility and possibly hurts the economy. Research has shown this to be the case in a variety of countries.^[11]

As regards market risk, we conduct supervision from multiple angles. In the SREP, market risk is evaluated as part of the assessment of risks to capital, with risk levels and risk controls being analysed through a combination of quantitative measures and supervisory judgement. Market risks are also assessed in the stress test, specifically by measuring the sensitivity of banks’ fair value positions to changes in the general level and the structure of interest rates. For example, in the 2016 stress test the impact of market risks in the adverse scenario was substantive, slightly greater than that of credit risks. Moreover, during on-site inspections we cover the banks’ internal risk control arrangements and the correct classification of fair value assets and liabilities from an accounting perspective.

Recently, the so-called “Level 3 positions” – referring to assets and liabilities for which neither active market nor indirectly observed prices exist and therefore are priced using internal risk models – have attracted the attention of market analysts. Based on a sample of most euro area significant banks, Level 3 assets amount to roughly €160 billion, or 15% of CET1 capital at end-2016. This amount consists of roughly €50 billion of equity instruments, an equivalent amount of derivatives, plus about €30 billion of debt instruments and €30 billion of loans. Assets are matched by a roughly equivalent stock of liabilities. These exposures are relatively contained in the aggregate, amounting to roughly 1.1% of total exposures of euro area significant banks, but are more sizeable in some countries and banks. Since their prudent valuation is difficult due to the lack of liquidity, care should be taken when interpreting their potential risk impact. It is important to note that Level 3 assets cannot be deemed as “toxic” per se and should not be assimilated to non performing exposures; for instance, a mortgage loan would be probably a Level 3 asset when fair valued. The central issue for the supervisor is to understand if the pricing model that produces their fair value is reliable.

Level 2 positions refer to assets and liabilities that do not have an active market but whose prices are partly or indirectly observed; these positions are larger, with an order of magnitude of roughly 20 times that of the Level 3 assets and 1.5 times that of Level 1 assets. Again, the assets are matched by a similar stock of liabilities. While the volume is quite large, valuation in this case is easier, since an indirectly observed price in principle exists. Many of those positions consist of “plain vanilla” instruments.

A large part of Level 2 and 3 operations is undertaken by banks for financing purposes or to help clients hedge or diversify their risk

ECB Banking Supervision has launched a major project called “Targeted Review of Internal Models” (TRIM), which will help shedding more light, inter alia, on these sources of market risks. TRIM is a resource- intensive project involving a large number of on-site inspections focused on analysing internal risk models both for credit and market risk, each of which can take considerable time. While that project advances, more targeted analyses may contribute to a better understanding of fair value financial instruments at selected banks, and the risks involved. In addition, banking supervisors also acts through on-site inspections and targeted reviews of the valuation and risk control processes pertaining to fair valued instruments.

Challenges arising from regulation

A third set of challenges for banks and supervisors relates to ongoing regulatory changes. The legislative framework which underpins banking supervision consists of the Capital Requirements Regulation (CRR), directly applied to banks, and the Capital Requirements Directive (CRD IV). This legislation allows several elements of flexibility, available to supervisors or Member States. Moreover, since the directive needs to be transposed into national legislation, this opens the door to legislative differences between countries and prevents a truly level playing field within the banking union.

An important example is in the “fit and proper” assessments conducted by the supervisor on the appointment of managers and key function holders. The application of national laws revealed numerous divergences in supervisory practices across Member States, notwithstanding the minimum harmonisation afforded by CRD IV. The ECB has developed a policy stance to achieve a more harmonised and consistent fit and proper supervision under the currently applicable legal framework. At present, however, there is a limit to what this initiative can achieve, since not all countries have fully transposed the CRD IV itself, while in other cases the directive has been transposed in an uneven way. The ECB would like to see not only a full transposition of the CRD IV by all Member States, but also a further alignment of the transposition laws, so that “fit and proper” assessments can use a common yardstick.

The ongoing review of the CRR and CRD IV, which the Commission opened in November 2016 and which is the first review of European banking legislation after the launch of the banking union, offers an opportunity to address level playing field issues. The Commission proposals are now being analysed within the European Council, and subsequently will be examined by the European Parliament. The ECB has had an intensive dialogue with the Commission to ensure that the opportunities provided by the review to strengthen the supervisory action are taken. The ECB will publish, in due course, a formal opinion on the proposals.

Conclusions

Before concluding, I would like to briefly refer to a comment recently made, in Italy, about the even-handedness of ECB Banking Supervision. It was said that ECB Banking Supervision may be biased against certain bank business models relative to others or, even worse, may have deliberately been too severe to certain banks or have held a national bias.

Those who express such suspicions probably don’t have a good understanding of how supervision operates. There are three steps that supervisory decisions have to go through, amounting in fact to three different, complex and articulated processes, each of which contain precise safeguards against those risks.

The first step is at the analytical level. All bank-specific analyses are filtered at two levels, by the vertical teams specialising on individual banks (the JSTs) and by the horizontal structures I mentioned previously. A variety of expertise and viewpoints are brought in, and many people are involved. In the most difficult cases, the decision is supported also by an independent on-site inspection. The staff involved is mixed in respect of their expertise and their nationalities so as to ensure the necessary diversity. This organisation is provides an initial safeguard against all types of bias.

The second step is the decision-making level. Analyses and proposals are delivered to the Supervisory Board, which has the mixed composition (ECB and national) that I have already described. Each SB member, in turn, has some support staff (from the ECB or from the respective national authority) that has, among its responsibilities, that of challenging the ECB staff proposals. Preparatory meetings and consultations take place at all levels before the SB meetings, as a result of which a staff position can be adapted. Finally, proposals are thoroughly debated by the SB and are subject to simple majority-rule voting. Decisions prepared by the SB are passed to the ECB Governing Council which, under the Treaty, has been assigned the final decision-making power.

The third and final step consists of the accountability process. Few supervisory authorities in the world, if any, are as accountable as the SSM. Specific rules govern the way ECB Banking Supervision exercises its accountability towards the European Parliament, which includes lawmakers from all countries, as well as to the Council of the EU, on which the Member States are represented. The Parliament’s Economic and Monetary Affairs Committee regularly receives records of proceedings of the SB. Although bank-specific decisions are usually not discussed in these fora, it is implausible that any systematic bias in supervisory decisions goes unnoticed.

All this considered and based on my vantage point as inside observer, I firmly reject the idea that in the SSM as a whole there can be room for deliberately partial judgement or consciously biased decisions.

As in every human endeavour, of course, error is always possible. The best way for us – or any other authority – to contain such risk, besides striving for ever better standards, is always to act transparently and regard scrutiny and criticism positively.

Thank you for your attention.