Challenges and priorities for the ECB banking supervision
Remarks by Ignazio Angeloni, Member of the Supervisory Board of the European Central Bank,
Banking Union Conference, Banking & Payments Federation Ireland/Mazars,
Dublin, 27 November 2015
In its first year in operation ECB Banking Supervision exceeded expectations, but work is underway to bring the single supervisor of the euro area to ever higher standards. Mr Angeloni emphasises that harmonisation of the way supervision is conducted across the euro area is key. Otherwise there cannot be a level playing field which enables fair competition, a precondition for growth and prosperity. In this context the ECB Supervisory Board has identified around 120 “options and national discretions” in Union law which can be exercised by a supervisor, for which a single approach has been agreed upon. In a public consultation ECB Banking Supervision is currently seeking comments from stakeholders on two documents, a Guide and a Regulation, in which this agreement is spelled out. Mr Angeloni encourages the banking community to submit questions and suggestions during this consultation, which ends on 16 December, using the contacts on the website of ECB Banking Supervision.
Regarding day-to-day supervision, Mr Angeloni explains that business models and profitability will be key priorities in 2016. Another priority is credit risk, considering that a number of institutional credit segments exhibit a persistent high stock of non-performing loans (NPLs). Other objectives are the fostering of comparability and the quality of internal models as well as capital adequacy. Governance and risk propensity will remain a priority in 2016. These priority areas for the SSM will be further refined and will eventually be published.
Mr Angeloni underscores the importance of transparency in the way supervision is conducted. Investor decisions need to be supported by adequate information on the returns and risks involved. Some knowledge of supervisory requirements can help in this respect. To that end, an appropriate degree of disclosure on the annual supervisory exercise known as Supervisory Review and Evaluation Process (SREP) could be helpful. So far, the SSM has disclosed a broad description of a common methodology for the SREP in its Guide to Banking Supervision. Going forward, as the SREP methodology becomes more established, steps towards greater transparency will be possible.
It is a pleasure to be addressing this Banking Union Conference today, and I am grateful to Banking & Payments Federation Ireland for inviting me.
On 4 November, the ECB marked the first anniversary of the operational start of the Single Supervisory Mechanism (SSM) by launching a new high-level conference series, the ECB Forum on Banking Supervision. The proceedings of that first event, including webcasts of the various sessions, are available on the ECB Banking Supervision website . Part of the discussion centred on a first assessment of the SSM at the end of its first year. A common message emerged, namely that while the achievements so far have been impressive – indeed, beyond expectations – much work is still to be done to bring the single supervisor up to the highest standards. We, operating within that structure, share that conclusion. Our efforts are firmly focused on attaining those standards.
My remarks today will focus mainly on the present and the future, rather than on the past. But speaking in Ireland, a country that experienced a dramatic banking crisis only a short time ago, and considering the range of speakers today, I cannot but note how the European banking policy framework has changed since that crisis. We now have a single supervisor in the euro area, tasked with ensuring the soundness of financial institutions and a level playing field. We have a single resolution authority about to start work, supported by a single resolution fund. And we have a legal framework to handle crises, the BRRD, covering also the smaller institutions which remain under direct national supervision. None of these instruments was available in 2010 or 2011. The Irish situation had to be managed with a highly incomplete set of policy tools. After those reforms, and with the benefit of experience, the approach today would be different. I think the relevant lessons have been learnt; the rest is for history to judge.
I would now like to turn to the main focus of my address. Speaking from the perspective of – but not on behalf of – the Supervisory Board of the ECB, I would like to share a few thoughts on our current work and some challenges that we face. I will then discuss some priorities for the immediate future.
Achievements and challenges
One year ago, in November 2014, the SSM assumed its role as the supervisor of all major banks in the euro area. From June 2012, when the relevant decision was taken by political leaders, the setting-up of this new authority took 28 months. That time was used to prepare the legal framework, have it approved by the Council of the European Union and the European Parliament, recruit some 800 supervisors to work at the ECB, build the internal organisation, prepare a new supervisory model for the SSM, set up its decision-making bodies, and design the necessary IT and statistical support. This programme was led by the ECB and benefited from intense collaboration from all participating national supervisors.
In addition, a detailed balance sheet check was conducted on all banks under direct supervision, the so-called “comprehensive assessment”. Even before it had ended, this assessment led to several measures strengthening balance sheets – adopted autonomously by banks, but triggered by expectations of the assessment’s results. Our directly supervised banks increased their CET1 capital by more than €100 billion between mid-2013 (when the comprehensive assessment was announced) and mid-2015. The exercise ultimately identified 13 banks that still had a capital shortfall, by a total amount of €9.5 billion.
By providing a wealth of information on all significant banks, not just those whose balance sheets needed strengthening, the assessment provided a stimulus and a starting point for further supervisory actions, pursued over the last year. Many of these are still ongoing. I would like to comment on two of them specifically.
Harmonising the legal framework
One finding of the comprehensive assessment was the existence of major discrepancies in the quantity and quality of capital across the supervised banks, which were not related to the nature or extent of the underlying risks . Many of these differences, which impede fair competition and hamper a level playing field, are based on the manner in which the EU’s legal framework – composed of the Capital Requirements Directive (CRD IV) and the Capital Requirements Regulation (CRR) – is applied across the participating countries. Discrepancies originate partly from the transposition of the directive into national law and partly from diverging application of the many flexible provisions (the so-called “options” and “discretions”) contained in the legislation. “Option” means that there are alternative ways to apply a given provision; “discretion” means that the possibility exists to apply or not to apply a given provision. Before the SSM, those decisions were made by the national authorities. As new competent authority, SSM now needs a single coherent framework on how to apply those provisions. This harmonisation effort was formally supported by the Eurogroup in April this year.
The Supervisory Board identified around 120 options and discretions which can be exercised by the ECB, for which a single approach has been agreed. That package covers important areas such as capital, liquidity, large exposure requirements for cross-border groups, the phasing-out of capital components not included in the Basel framework and the prudential treatment of insurance participations, and others. In conducting this work, the guiding principle was to promote harmonisation with prudence, with appropriate gradualism considering the legitimate expectations generated by previous treatment by the national authorities. An impact assessment was conducted to measure the likely effect of the policy.
This gave rise to two legal instruments that are currently under consultation. The first is a regulation – a binding legal instrument – laying down the legal obligations for the significant banks of the SSM related to the prudential treatment of options and discretions that are of a general nature. The second document is a guide containing case-by-case provisions whose application is bank-specific. The guide is a non-legally binding instrument that provides guidance to supervisory teams on how to treat individual cases.
The consultation will end on 16 December. I encourage the banking community to submit questions and suggestions using the contacts provided on our website . We envisage that the package will enter into force around April 2016. From then on, the regulation and the guide will be directly applied to all banks under direct ECB supervision, with an immediate harmonising effect.
Some other options and discretions will require a follow-up, either because of further work to be done by the European Banking Authority (EBA) or the European Commission, or because more experience needs to be gained from the assessment of specific cases. In some cases, adjustments in national legislation may be necessary. This follow-up work is scheduled to start as soon as the current work is concluded in 2016.
Setting prudential requirements – the SREP
Supervisors, as we all know, regularly assess and measure the risks relevant for each bank. This core activity is called the Supervisory Review and Evaluation Process, or SREP, and it leads, on an annual basis, to the quantification of additional (i.e. Pillar 2) prudential requirements, which are applied on top of the minimum regulatory requirements (i.e. Pillar 1).
For the SSM this year, this process was a logical extension of the 2014 comprehensive assessment. Risk factors always evolve over time, so they need regular updating. Moreover, our exercise in 2014 centred on an asset quality review and a stress test, leaving out a variety of other risk factors, such as business models, funding structures, governance and risk controls. The SREP is, in principle, all-encompassing. This year, for the first time, the SSM conducted this process using a unified methodology, firmly rooted in EBA guidelines. This methodology combines quantitative and qualitative elements and treats all banks consistently, while accounting for different business models. The Supervisory Board repeatedly examined and refined the process in the course of the year.
Let me briefly explain how the process works.
The methodology follows banks’ internal strategies and decision-making processes and is structured along four components: business model assessment; internal governance and risk management; risks to capital; and risks to liquidity and funding. Each component is assessed both quantitatively and qualitatively. The quantitative assessment is based on a broad range of data covering own funds, financial reporting, large exposures, credit and operational risk, etc. The data, which are provided by the banks on the basis of harmonised standards, give rise to numerical risk scores.
The qualitative component involves supervisory judgements regarding things like risk controls, risk culture and governance. It is important to develop an articulate opinion on how these risks develop and on what impact they may have within each specific institution. This analysis requires internal knowledge and relies to some extent on the subjective judgement and experience of the supervisor. Within the SSM, the Joint Supervisory Teams, possessing this knowledge and experience, are responsible for providing the qualitative input. The judgement cannot be mechanical, but can and should be reasoned.
In order to ensure consistency across banks, certain criteria are followed. We refer to one of them as “constrained judgement”, meaning that the subjective element can influence the quantitative result only to a certain extent. Another key principle is proportionality, which means that the level of supervisory engagement (i.e. its frequency or intensity) should be linked to the complexity and systemic relevance of the bank’s activities.
Following this year’s SREP, the Pillar 2 requirements for the major institutions directly supervised by the ECB increased on average by 30 basis points relative to the previous year. Including also the phasing in of the macro-prudential buffers, the average increase totals around 50 basis points. This moderate increase is adequate from both a micro- and a macro-prudential perspective, allowing a gradual transition towards fully loaded Basel III requirements.
One aspect deserving attention is the communication strategy surrounding the SREP. Different considerations come into play here. On the one hand, enhancing public information about supervised entities, part of which is of a proprietary nature, can have undesired effects. One line of thought maintains that, in order to preserve an open and productive dialogue between the supervisory authority and the bank, all information exchanged should remain confidential. The possibility of proprietary information being released, especially to competitors, may discourage openness and deprive the supervisor of critical information. In some cases, supervisory judgements placed in the public domain without proper caution may increase the uncertainty surrounding individual (weaker) institutions, with risks also to financial stability.
These arguments carry weight but must be compared to the advantages of transparent communication. Ultimately, the SREP aims to ensure that banks have adequate prudential safeguards in relation to their level of risk. Investors’ decisions need to be supported by adequate information on the returns and risks involved. Some knowledge of supervisory requirements can help in this respect. First, it provides information on future capital plans, which in turn influence future returns on capital instruments. Second, it acts as a benchmark, indicating whether a bank is judged to be “safe and sound” by the supervisor. This helps with the calculation of risk. Thus, an appropriate degree of disclosure may enhance market confidence and encourage investment decisions, actually reducing uncertainty. This was, in fact, the rationale behind the very high degree of transparency that characterised the ECB’s comprehensive assessment in 2014.
When thinking about communication regarding the SREP, one should distinguish between the transparency of the process and the disclosure of the outcomes. The first has to do with public knowledge of the methodology employed. The second relates to public communication of the results. In particular, a transparent process helps to ensure that results are properly internalised by supervised banks and avoids misunderstandings. In the SSM, the Joint Supervisory Teams, which engage in a continuous dialogue with the banks, are the appropriate channel to convey this type of information. It should remain clear at all times that the SREP is not a mechanical process: a degree of reasoned discretion must always be preserved. That said, the dialogue between the Joint Supervisory Team and the bank limits the risk of “unpleasant surprises” at the end of the process, thus also facilitating proper public disclosure of the outcome. Such disclosure should, in any event, always be agreeable to the bank concerned .
Recent international experience is informative in this regard. While somewhat diverse, supervisory practices have evolved towards more transparency in recent years. The Comprehensive Capital Analysis and Review (CCAR) in the United States and the SREP processes in some non-euro area EU countries provide examples of this kind of development.
In the United States, the Federal Reserve System publishes its annual assessment of the capital planning processes and capital adequacy of the largest bank holding companies . The Federal Reserve does not publish its stress test methodology in order to avoid herding behaviour in risk model building and prevent activities being shifted to areas not captured by the stress testing models, but it does disclose its qualitative assessment of banks’ capital plans .
In Europe, CRD IV  requires supervisors to publish the general criteria and methodologies used in the SREP. The Prudential Regulation Authority (PRA) in the U.K. publishes the methodology for its so-called “Pillar 2A” capital requirements. The Danish FSA also publishes methodologies, including the benchmarks for assessing risk in several areas. Thus far, the SSM has provided a broad description of the common methodology for the SREP in its Guide to Banking Supervision. Going forward, as the SREP methodology becomes more established, steps towards greater methodological transparency will be possible.
The detail in the disclosure of the outcome differs across Member States. As regards Pillar 2 requirements, under CRD IV the decision to publish is left to the supervisory authorities. The Danish FSA publishes SREP capital add-ons for all banks, plus summaries of on-site supervisory examinations. Sweden’s Finansinspektionen has published plans for a detailed supervisory methodology and discloses SREP capital add‑ons. A generalised tendency towards greater transparency is underway also in this area. The United Kingdom has adopted a practice of allowing capital requirements to be disclosed following the notification of the competent authority. Following its introduction, a number of firms have voluntarily disclosed their capital requirements after notifying the PRA .
Further priorities for 2016
I would now like to turn to the main SSM work-lines planned for the coming year.
Business models and profitability will be a key priority during 2016. This last year we have developed a business model classification tool to conduct peer group analysis. We have also conducted a survey on banks’ profitability forecasts. Building on this work, next year we will conduct in-depth reviews of drivers of banks’ profitability at firm level and across business models. One area of interest is the question of how banks are coping with the protracted low interest rate environment and incoming new regulations.
Another priority is credit risk, because of the persistently high level of non-performing loans (NPLs). High NPLs dampen banks’ ability to lend and feed concerns regarding forbearance and under-provisioning. In this context, a taskforce has been established, led by a senior official from the Central Bank of Ireland, to develop a consistent approach to banks with high NPL levels. The SSM will also conduct work on excessive concentrations of risk, as well as on exposure to sovereigns and to commercial and residential real estate.
A major multi-year project presently ongoing aims at fostering comparability and quality of internal models. Given the large number and wide diversity of internal models and specific expertise required, this is a challenging task. A targeted review of internal models will be carried out to ensure that they comply with regulatory standards and in order to foster consistency across institutions.
Another important area is capital adequacy. Of particular relevance in this regard are banks’ ICAAP and, in this context, their internal stress testing capabilities, for the purpose of the EBA’s 2016 stress test and the SREP top-down stress tests. With a view to the implementation of the new TLAC and MREL requirements, the preparations of the banks for ‘gone concern’ scenarios will also be scrutinised, as well as the measures that they are undertaking in order to comply with the TLAC or MREL requirements.
Finally, banks’ governance and risk appetite will remain a priority in 2016. The financial crisis clearly showed that banks’ governing boards often lacked the full information needed to make good business decisions. This year the organisation and composition of banks’ management have been reviewed, focusing on the profiles of board members, so as to ensure that the relevant expertise is available at board level. A task force on behaviour and culture has been created. In 2016 we will build on the information obtained in this way to increase our scrutiny of this key aspect of banks’ corporate governance.
The priority areas for the SSM are being refined and will eventually be published.
Let me now conclude. The establishment of the SSM and its first year of existence have been intense, but rewarding. In a very limited period of time, numerous steps have been taken towards ensuring better-capitalised banks, high-quality supervision and a more consistent legal framework.
We know that many important issues still need to be addressed or explored further. The Supervisory Board and the staff of the SSM, in cooperation with the participating national authorities, are fully committed to enhancing this work and achieving ever more effective, efficient and transparent banking supervision.
Thank you for your attention.
I am grateful to Cécile Meys for her excellent drafting support, and to Korbinian Ibel, Thomas Jorgensen, François-Louis Michaud and Giuseppe Siani for helpful comments. The views expressed here are personal and should not be attributed to them or to the ECB.
See www.bankingsupervision.europa.eu/ecb/pub/pdf/aggregatereportonthecomprehensiveassessment201410.en.pdf, Section 1.2.
One argument against the detailed disclosure of methodologies is that it may facilitate “gaming” – i.e. elusive practices by banks. This can happen if banks let certain risky activities “migrate” to areas that are not adequately covered by risk indicators. On the other hand, banks’ awareness of the methodology would allow them to better manage their risk by assigning the appropriate capital cost to individual risk elements.
See Article 143(1)(c) of CRD IV.