Countdown to the start of the SSM
Speech by Ignazio Angeloni, Member of the Supervisory Board of the ECB,
Conference on “El comienzo de la Supervisión Bancaria Europea” - Organised by KPMG and CUNEF – Collegio Universitario de Estudios Financieros,
Madrid, 31 October 2014
It is a pleasure to be in Madrid today and to participate in this conference. I am grateful to the organisers and to Antonio Sáinz de Vicuña in particular, for inviting me.
In four days, the Single Supervisory Mechanism (SSM) will be operational. This concludes a preparatory cycle that has spanned 28 months, starting with the announcement by EU leaders on 29 June 2012, and with several further milestones along the way, including: the adoption of the draft SSM Regulation by the EU Council (13 December 2012); the launch of the comprehensive assessment of the banks (23 October 2013); and the entry into force of the SSM Regulation (3 November 2013).
What I find remarkable about this new ECB tour de force is the fact that, unlike previous ones (including the launch of the euro), this one required much broader participation by the entire euro area banking sector. All banks involved in the assessment actively participated in the process. Bank managers and experts, many of which were previously unfamiliar with the ECB, engaged in intensive contacts with us. We held meetings, analysed common problems, looked for solutions and planned future work together. All this took place in an intense but constructive atmosphere. Personal contacts and professional development went side by side. All this is also, I think, European integration.
Normally I don’t like to indulge in thinking about the past, but rather prefer to concentrate on future challenges, but today it is worth making an exception. I will therefore share some remarks on the set-up phase that was just completed, including, first and foremost, the comprehensive assessment and its results, published last Sunday, 26 October 2014. After this, I will say a few words about the state of play as regards preparation within the ECB.
The comprehensive assessment, covering 130 of the largest euro area banks, representing 82% of total banking assets in the euro area, was unprecedented in both scale and scope. The fact that the exercise consisted of two components, a forward-looking part through the stress test and a point-in-time review of the value of assets (the asset quality review, or AQR), made it unique not only in Europe, but globally.
From the comprehensive assessment we obtained much more than just a dry list of banks with a capital shortfall; we obtained an unparalleled in-depth comparative insight into the participating banks. We now realise that starting our responsibility as supervisors would have been impossible without the comprehensive assessment. Completing it in one year – something many observers did not believe we could do – was an accomplishment in itself, the result of the professionalism, hard work and commitment of all the people involved at the ECB, the national supervisors (often called the national competent authorities, or NCAs), as well as the European Banking Authority (EBA) and the entire banking sector. To illustrate the vastness of the exercise, let me mention some numbers: more than 6,000 people were involved (part time or full time); over 800 bank portfolios and more than 119,000 debtors were individually reviewed, after having been selected using risk-based statistical criteria; 170,000 collateral items were re-valued; 765 collective provisioning challenger models were built and 5,000 level 3 provisions were re-valued.
Even before its conclusion the exercise already produced meaningful results. Since the exercise was announced in July 2013, to prepare for its findings, participating banks had already undertaken measures to strengthen their balance sheets, raising more than EUR 200 billion in total. This amount included issuance of equity and CoCos, retained earnings, asset sales, one-off items and additional provisioning. In particular, in the first nine months of 2014, 54 banks undertook capital actions, raising a total of EUR 57.1 billion, almost exclusively from the issuance of Common Equity Tier 1 (CET1) eligible instruments. The conversion of hybrid instruments and raising of AT1 only occurred to a limited extent (see slide 2) Due to the repayment or buy-back of capital instruments, there was an offsetting impact of EUR 16.6 billion. Altogether, the actions by the banks in that period led to a net capital increase of EUR 40.5 billion. This amount equates to a median increase in banks’ pre-comprehensive assessment CET1 ratios of 1.0 percentage point.
Asset quality review
The AQR resulted in adjustments of the bank balance sheets for an aggregate value of EUR 47.5 billion, on assets as of 31 December 2013.  Across the SSM participating countries, the mean AQR CET1 net change ranged from a reduction of 0.2% to as much as a reduction of 2.9% of starting risk weighted assets (RWAs). These adjustments originated primarily from specific provisions on non-retail exposures, collective provisioning assessment and impacts from the fair value exposures review.
Importantly, non-performing exposures were re-estimated, following a harmonised and simplified definition provided by the EBA. As this definition is more conservative than the average of banks’ internal definitions, the application of the simplified approach led to a EUR 54.6 billion increase in NPE stock to EUR 797.7 billion (see slide 3). The subsequent individual review of the credit files led to an additional increase in NPEs of EUR 81.3 billion, resulting in a total increase of EUR 135.9 billion to EUR 879.1 billion. This change represents an 18.3% adjustment to the initial NPE stock across all banks. The impact of the application of the EBA’s simplified approach and the credit file review of the stock of NPEs varied amongst debtor geographies, with overall increases ranging from 7% to 116%. I want to emphasise that the lack of homogeneity in the definition of NPEs across jurisdictions and supervisory authorities in the euro area was a major concern among investors and observers last year, when we started the exercise. Having implemented a new harmonised definition is, in itself, a major step forward.
The collateral valuation has also led to substantive adjustments. The majority of collateral which had not undergone a third-party valuation within the last year was re-valued for all debtors selected in the sampling. Collateral items amounting to EUR 367 billion were investigated and adjusted downwards by EUR 39 billion, representing a decrease of approximately 10% compared with previous bank-internal valuations. Collateral prices have, on average, fallen since 2005 and the average haircut applied on top of the indexation also increased. The reduction in collateral value had the biggest impact on commercial real estate properties and land, both in absolute and relative terms.
The second component of the exercise was the stress test. In comparison with previous stress tests, this one was different for three reasons. First, the adverse scenario was more stringent. The difference between adverse and baseline GDP growth rates for the European Union was around 5.1 percentage points at the end of the second year of the 2014 stress test exercise, compared to between 3 to 4 percentages in the previous three EU-wide exercises. Second, the horizon was longer, three years as opposed to two. Finally, and crucially, the AQR findings were integrated into the banks’ stress test results. This contributed to improving the quality and reliability of the results. Since, for reasons of timing, the AQR and stress test had to be conducted in large part simultaneously, a linkage between the two was necessary; this is what we call the “join-up”. In concrete terms, this means that adjustments to the probability of impairment and loss given impairment parameters were used not only within the AQR itself, but also to adjust stress test credit risk parameters. In the end, the aggregate impact of the stress test, including join-up but not the AQR, in term of percentage changes between year-end 2013 and year-end 2016 in the average CET1 ratio of participating banks was an increase of 0.2 percentage points under the baseline and a decrease of 3.0 percentage points under the adverse scenario. The absolute capital impact under the adverse scenario amounted to EUR 262.7 billion (slide 4). Overall, the capital impact over the three-year stress test horizon amounted to EUR 43 billion under the baseline scenario and to a reduction of EUR 181 billion under the adverse scenario. In terms of capital ratios, the overall impact for the system was a decline of 3.4 % in the adverse stress test scenario relative to the end-2013 values, of which 0.4 % attributable to the AQR and 3.0 % attributable to the stress test and join-up (slide 5). In terms of median impact on the participating banks, the capital impact consisted of a decrease in the CET1 ratio by 4.0 percentage points, from 12.4% to 8.3% in 2016 (slide 6). The impact was well distributed across countries; the main impact was observed in France, followed by Italy and Germany (slide 7). The impact on Spain was much lower, and mainly attributable to the adverse stress test scenario. This is logical, considering that Spain had already undergone an AQR in 2012.
Although not fully comparable, the median projected CET1 ratio reduction in the Comprehensive Capital Analysis and Review (CCAR) carried out in the United States in 2014 was 2.9% . The impact was 3.9% in the AQR and stress test carried out in Spain in 2012  and 2.1% in the EBA stress test carried out in 2011 .
Overall, the comprehensive assessment identified a capital shortfall of EUR 24.6 billion across 25 participating banks, at the date of 31 December 2013. After comparing the projected solvency ratios with the thresholds defined for the exercise (8% of CET1 in the baseline scenario and AQR, 5.5% of CET1 in the adverse scenario), the impact was concentrated more than proportionately on some countries, largely as a result of the initial conditions of the banks in question (slide 8). The capital operations completed in the first nine months of this year that I referred to earlier have reduced the number of banks with a capital shortfall to 13, and the overall shortfall to EUR 9.5 billion.
The next two slides show the CET1 impact of the AQR and the stress test, respectively, for sub-categories of banks according to the value of the initial CET1 ratio. The charts allow seeing if the capital impact of the exercise across different banks was somehow related to their initial condition. This appears not to be the case for the AQR (slide 9), whereas for the stress test there seems to be a positive correlation (slide 10): the better the initial condition of the bank, measured by the capital ratio, the stronger the downward impact of the stress test on the corresponding bank. This evidence warrants further analysis before drawing definitive conclusions. One possible interpretation is that banks with lower initial capital ratio may have used the flexibility allowed by the bottom-up decentralised stress test methodology to mitigate (“optimize”, from their viewpoint) the impact of the stressed scenario.
Follow-up to the comprehensive assessment
Following the publication of the comprehensive assessment, those banks for which we have identified a capital shortfall will have to submit capital plans by 10 November 2014. The capital plans will be examined and approved by the ECB. They are expected to be covered within six months for those identified in the AQR and the baseline scenario of the stress test, and within nine months for those identified in the adverse scenario of the stress test (this is the most frequent case). In this context, it is important to note that not only banks in which capital shortfalls have been identified may choose to raise capital going forward. A number of banks that did not have a capital shortfall have seen their capital buffer significantly depleted and may decide, over time, to also take action in order to strengthen their position again.
Let me mention before concluding on this point that, in addition to acquiring a host of quantitative information on capital levels, a lot of qualitative information was collected during the course of the assessment as well. We will use all this information as a basis for our normal supervisory work, which will start on 4 November. In addition, the 2014 Supervisory Review and Evaluation Process conducted by the national authorities will embed the results of the comprehensive assessment, leading to an overall capital adequacy assessment.
SSM preparations – state of play
Let me now spend a few words on the state of play of the SSM preparations.
As regards SSM governance structures, the SSM is fully operational. Since January this year, the Supervisory Board, composed of a Chair, a Vice-Chair and three ECB representatives plus one representative from each national competent authority, has been meeting twice a month. In total, there are now 23 members, all of whom are bound by the SSM Regulation to pursue the interests of the EU as a whole. The Supervisory Board has met 19 times so far. In addition, the meetings of the Board are prepared by a Steering Committee, whose participation is limited to eight members of the Supervisory Board, notably the Chair, Vice-Chair, the permanent ECB representative on the Supervisory Board and five of the NCA representatives on the Supervisory Board. The national participation follows a rotation scheme.
In addition, the Administrative Board of Review, the body that will carry out internal administrative reviews of the decisions taken by the ECB, has been established. It has five regular and two alternate members, all chosen among prominent academics and former policy-makers with expertise in legal and supervisory matters.
In addition, the ECB Regulation on the establishment of the Mediation Panel was formally adopted in June and its members, one per Member State, chosen from among the members of the Governing Council and the Supervisory Board, have been appointed. In accordance with the principle of separation between monetary policy and supervision, enshrined in the SSM Regulation, the Mediation Panel will aim to resolve differences of views in the event of an objection by the Governing Council to a draft decision prepared by the Supervisory Board.
As regards institutions that will be directly supervised by the ECB, the Governing Council adopted the decisions determining the significance of 120 institutions on the basis of the non-objection procedure. These institutions have been notified by the ECB. The lists of significant and less significant banks were published on the ECB´s website on 4 September 2014.
With respect to staffing, a major effort has been made to recruit a large amount of professionals, from the national authorities as well as from the market. This effort is on track. All appointments followed an open and transparent procedure. The ECB has received more than 20,000 applications and a total of 62 recruitment campaigns were held. Close to 900 staff members have been recruited out of the approximate total of 1000 budgeted positions. Importantly, from the viewpoint of starting the conduct of supervision on 4 November, all Joint Supervisory Teams (JSTs) are operational and adequately staffed. At this stage, all JST coordinators have been selected. In preparation for their new tasks, the JSTs have already been very active, in some cases also participating in phases of the comprehensive assessment, as well as in meetings with supervisory colleges.
As regards the framework through which supervision will be conducted, a Supervisory Manual has been finalised. This is an internal document that describes the processes and methodology for the supervision of credit institutions, as well as the procedures for cooperation within the SSM and with authorities outside the SSM. The manual conforms to the single rulebook developed by the EBA. The manual is a living document and will be updated regularly. The part of the Manual that is relevant to outsiders was published last month in the form of a “Guide to banking supervision”. This Guide explains the overall functioning of the SSM in a user-friendly format.
The ECB is also working on developing harmonised data templates based on Common Reporting (COREP) and Financial Reporting (FINREP). All banks will report supervisory financial information. However, in line with the proportionality principle, less significant supervised entities will be subject to reduced reporting requirements and will be given more time for implementation.
I should not omit, before closing, the monumental work conducted by the legal services throughout the process. This led to a number of important outputs, especially during the preparation of the SSM Regulation and the related debates in the Council and the European Parliament. ECB and NCA lawyers have joined forces to prepare a Framework Regulation (the set of secondary rules that will govern the functioning of the SSM), and to amend the internal rules of the ECB to accommodate the new function.
Finally, last but not least, the set-up of supervision in the SSM entailed a number of challenges in the area of information technology as well.
I will summarise my conclusion in just three words: we are ready.
The ECB supervisory staff and the Supervisory Board are focused on their future responsibilities. We are aware that there will be risks, that there will be challenges; in many respects, the real challenges start now. And be sure: there will also be occasional mistakes, because no human pursuit is infallible. But the steps already made have been made well. The Latin poet Horatius wrote, “Dimidium facti, qui coepit, habet”, which means, “Well begun is half done”. We are still very far from half-of-the-road, I would say, but nonetheless we should feel strengthened, though not complacent, by what has already been accomplished.
Thank you for your attention.
I am grateful to Cécile Meys for her contribution.
It is important to note that this number only represents the adjustments made to available CET1 and does not represent the capital shortfall, due to excess capital being held by banks.
Dodd Frank Act Stress Test severely adverse scenario.
Due to low number of participating banks this figure is a weighted average.
Exercise occurred on the basis of a lower initial capital base than the current exercise.