Sabine Lautenschläger: Making the comprehensive assessment a success
The views expressed in this speech are those of the speaker and not the view of the BIS.
Ms Pereda, Mr Zambeletti and Mr Albarracin,
Luis, (Governor Linde),
Ladies and gentlemen,
When preparing today's speech, I was confronted with quite a laundry list of topics I could talk about. Ranging from the details of the comprehensive assessment itself to what will happen once we know the outcome of this exercise; our new supervisory approach; the intricacies of some banking regulation and beyond. I would love to have time to address all of these issues today, but I will limit myself to two:
- first, how we can make the comprehensive assessment a success;
- second, some pertinent questions on the measures to deal with capital shortfalls.
General background and goals of the exercise
The Single Supervisory Mechanism (SSM) will take over supervision of the significant euro area banks in November. In advance, we are conducting a rigorous comprehensive assessment of banks' balance sheets that will ensure that our Joint Supervisory Teams, comprising supervisors from the ECB and the national competent authorities (NCAs), can start supervision with a clean slate. We can thus base supervision on a thorough evaluation of the condition of the individual institutions, not only regarding market and credit risks in individual banks, but also processes, policies and accounting specifications, and the sensitivities of core business fields - and all this with a higher degree of comparability thanks to the harmonisation of the methodologies and implementation of the assessment.
But the importance of this exercise extends well beyond the supervisory realm - one could say that the supervisory aspects are only a positive side effect. The objective of the comprehensive assessment is to restore trust in the euro area banking system by creating transparency on the condition of bank balance sheets and enforcing repair - where necessary. This is crucial for the macroeconomic environment of the monetary union, since credit supply to the real economy in Europe is strongly linked to banks' financing services and thus dependent, too, on banks' capital and funding conditions - although this is not the only prerequisite.
At the current juncture, banks in the euro area continue to suffer from general distrust concerning the quality of their assets. Many market participants believe that bank balance sheets may still carry substantial hidden losses, and the resulting uncertainty is directly reflected in - among other things - refinancing costs and stock market valuations. Even small deteriorations in a bank's performance can sometimes suffice to provoke speculation and negative market reactions.
This lack of confidence affects bank lending conditions, both local and cross-border, for firms in the real economy. Banks in the euro area need to respond to these market expectations and uncertainties. They increase their capital ratios by retaining their earnings, holding off lending and avoiding additional risks. Small and medium-sized firms in peripheral countries are particularly affected by the sustained shortage in credit supply.
Key factors for success
The comprehensive assessment should contribute substantially to breaking this negative causal chain. It is not the only factor needed for reviving the loan business, particularly in the peripheral countries. But it is one of the important ones. Restored confidence in the euro area banking sector should lead to lower funding costs for banks, which should be able to shift their focus back to their core business: lending to firms and households. In turn, this should lead to improvements in credit conditions for the real economy. For these effects to materialise, two elements are absolutely crucial.
First, we need to succeed in establishing transparency. In order to positively impact on investors' perceptions of euro area banks, we must ensure that the exercise is sufficiently thorough for a true understanding of the banks' condition. Any remaining uncertainty and doubts would be counterproductive.
Second, any weaknesses identified must be addressed in a swift and decisive manner. Findings such as a lack of provisioning or the mis-valuation of assets must lead to direct and effective corrective actions, which ensure that adequate capital levels are reached or maintained. In this context, the exercise also offers an opportunity to highlight the balance sheet repair measures that have already been taken during the last months. The expectation of the forthcoming AQR has already had positive effects on the banks involved. We estimate, based on public information, that since July last year banks have strengthened their balance sheets by an amount of €104 billion, not only through capital increases, but also collateral revaluations and higher provisions.
The realisation of these two objectives of transparency and repair should induce a significant change in the markets' perspective of the euro area banking sector. The comprehensive assessment thus offers a great opportunity, although seizing this opportunity may sometimes seem challenging - both for banks and for supervisors.
For the comprehensive assessment to be a success, we have to adhere to some critical factors.
Reasons for optimism
First, our exercise needs to be more comprehensive than a standard stress test. In order for the comprehensive assessment to enjoy credibility, we must prove that we had deep insight into specific business areas, portfolios and individual credits as well as into processes and working guidelines. We must check whether accounting standards and workflows have been adhered to and ascertain whether the files of credit institutions faithfully reflect the contents of their systems.
The comprehensive assessment is more comprehensive than any previous exercise - in terms of its scope and of the number of banks covered. We will examine in detail no less than 760 banking book portfolios. To reach our objective - to improve market confidence in euro area banks - it was essential to select more than 50% of banks' portfolio exposure. The exercise would not have been seen as credible if a smaller set of portfolios had been selected. We will review some 135,000 credit files. A total number of more than 6,000 supervisors and auditors are currently conducting the review on the ground. For this review, we have and will collect vast amounts of data. I am aware that fulfilling our data requests puts quite a burden on banks, particularly the smaller institutions. We know that some of those face tighter constraints on the number of staff they can dedicate to specific tasks relating to the asset quality review (AQR) and that they may generally be less used to regular supervisory exercises such as the stress tests conducted by the European Banking Authority (EBA). We did get some complaints which we take seriously.
But even if we pare down our data requests to the essentials, we will still request tremendous amounts of data.
By the way, let's not trick ourselves - if there were no complaints at all, I would wonder whether we were doing a good job! And we need the data to do this job correctly. We need to stretch banks' and our own resources to the limit, not only to drill down to the detail, but also to allow for comparability.
Second, the two parts of this assessment complement each other: the AQR will give us a point-in-time picture of the state of banks' balance sheets.
The AQR will ensure that the following stress test will be based on clean data, thus avoiding the weakness of the previous stress tests, in which the calculations were based on exposures valued only by the bank concerned. The use of common valuation standards will improve comparability, and the examination carried out by supervisors will add to the credibility of the exercise.
The stress test adds a forward-looking element. It will focus on the resilience of these balance sheets to certain shocks to the economic and financial system. Both elements will come together, as the results of the AQR will feed into the stress test. The comprehensive assessment is thus likely to be much more demanding than previous stress tests; this has the added advantage that those banks that pass deserve greater market confidence.
The linking of the AQR and the stress test obviously constitutes a significant challenge, not least because it has not been done in this way before. We, the ECB and NCAs are close to finalising the actual modalities of this linking. Before finalising and publishing the relevant part of the methodology, we will seek the assessment of banks as well as auditors, using all available input to ensure a smooth and technically-sound process.
The set-up of quality assurance is the third improvement offered by the comprehensive assessment. We have a strict quality assurance in place for the entire exercise. This is key for ensuring the integrity and comparability of the results and will further enhance their credibility.
Our quality assurance framework during the AQR is based on three layers or "lines of defence": NCA bank inspection teams, composed of supervisors and auditors, are responsible for validating the quality of their submissions. NCAs' technical assistance and quality assurance teams then conduct further checks and validations across banks in the respective jurisdiction. Finally, the ECB's central project management office and the ECB country teams review the quality of national submissions to the ECB, also carrying out cross-country checks and analyses.
The recent publication of the stress scenarios has focused attention on this final phase of the assessment.
For the stress test, we will have a double layer of quality checks: we will first conduct data consistency checks for the bottom-up stress test, checking both model inputs and assumptions, and then carry out a top-down stress test to cross-check the results of the bottom-up exercise. For effective quality assurance, we will need to request some additional data that is not captured in the mandatory EBA template.
To be successful, we must not be content with simply identifying banks' weaknesses. For the new European supervisor, for me, it will be of utmost importance to respond quickly, stringently and consistently when significant shortfalls at banks are revealed. Any banks still facing a capital shortfall will be asked to submit capital plans shortly after we announce the results of the comprehensive assessment. But I have to admit I expect banks with relevant capital shortfalls to have their solution ready before we publish the results. It is the only way to avoid unnecessary market uncertainties.
We will of course closely monitor the implementation of these plans. Any capital issuances to cover shortfalls will need to be carried out swiftly and will need to focus on capital instruments of the highest quality. I will come back to these two aspects later.
The execution of the AQR, its successful linking to the stress test, the quality assurance set-up and the stringent and swift responses of banks and the supervisor to identified capital shortfalls will still require significant efforts between now and late October.
Crucial aspects of the remaining process ahead of us - communication with banks and the public
But there is another critical success factor - effective and efficient communication with the banks, the public and the markets.
Let me explain briefly why I believe that, at this stage, it becomes increasingly important to have direct interaction between the supervisor and the banks. I am convinced that we need to verify the fact base used in our assessment. In my opinion, it would be difficult to conduct our data analyses in isolation in Frankfurt, and then communicate the facts to the banks when the results of the assessment are disclosed in October, without ever discussing matters at the technical level. This would fail to do justice to the complexity of the issues at stake, and would not be proper supervisory practice. For this reason, it would not only be appropriate, but also necessary, to have several points of interaction between now and October, whereby banks will be contacted by supervisors in order to be able to contradict partial and preliminary facts of the AQR and the stress test.
Apart from verifying findings, another major reason why I find interaction crucial, are aspects within the work blocks that may require immediate action by the banks because they have an impact on other areas of the assessment. Let me take the first work blocks, the processes, policies and accounting review (PP&A review) and the data integrity validation as examples. A possible finding in the PP&A review could be that a portfolio of held-to-maturity bonds needs to be reclassified as available for sale. In this case, the bank would need to calculate the impact of the change in classification on the Common Equity Tier 1 (CET1) ratio. Some of the findings from the data integrity validation may indicate the need for the bank to make adjustments to the data sets used for stress testing purposes.
A key aspect that should be mentioned here are market disclosure requirements. Of course, we are aware that they exist, and that in some cases, they could oblige a bank to publically disclose a certain finding right away.
While this may not necessarily be the case, it could happen. Here, it is crucial to be aware that issues which we as supervisors would highlight in such technical interactions are generally prudential rather than accounting judgements, and that they will only be preliminary fact findings, which will still need to be examined and assessed during the national quality assurance process as well as during the cross-country checks and analyses. We ultimately need final decisions on supervisory assessments, before any such outcome of a prudential exercise becomes binding. As you know, this will not be done until the end of the process.
This leads me to the broader issue of disclosure of the comprehensive assessment results. Since the initial announcement of the exercise, we have expressed our preference for a single communication at the end of the assessment, in which the final results will be disclosed to the public. As confirmed by the feedback received so far, we think this is very much in the interest of the participating banks.
That being said, we have also stated very clearly that in some individual cases, severe weaknesses may be discovered during the process - weaknesses for which corrective action cannot be postponed until October. Such cases could occur at any stage of an examination. The national supervisors, who are still the competent authorities for the relevant banks until the beginning of November, would need to impose the required action in liaison with the ECB.
In terms of public communication, our common goal should be for the information and messages on the comprehensive assessment that reach the markets to be as clear and precise as possible, minimising any risk of misinterpretation and speculation. We need to be very careful in designing the disclosure templates, as these will form the basis for transparency.
Here, we have to strike a balance between the objective of the comprehensive assessment to restore confidence in the market and thus the desire to publish a wide range of data on the one hand, and on the other, the banks' interest that they should not to be exposed in every detail to competitors. I am also convinced that we should publish these templates well in advance of the final disclosure. Market participants will then have time to digest the format, and will know exactly what type of information to expect on the day the results are released.
Addressing capital shortfalls
I would now like to turn to the issue of how to deal with any capital shortfalls identified in the comprehensive assessment.
It is clear that, first and foremost, banks will need to fill potential capital gaps via market-based solutions. In line with our previous communications, those can include retained earnings, new issuances of common equity, suitably strong contingent capital, and sales of selected assets at market prices.
As I see it, reductions in risk-weighted assets due to the validation and roll-out of appropriate internal models to additional portfolios should not be eligible to address a capital shortfall, unless these changes had already been planned and are under consideration by the competent authority - we are already used to this kind of constraint from the last EBA recapitalisation exercise.
In our latest press release on the comprehensive assessment issued at the end of April, we further elaborated on two aspects that I consider absolutely crucial.
First, new capital issued to cover shortfalls will need to be of the highest quality. This is key for ensuring that a solid and thorough strengthening of capital positions takes place as a result of the exercise. We thus put a strong focus on CET1 and limit the admissible use of convertible capital instruments. Any capital shortfall revealed by the AQR and the baseline scenario of the stress test may only be covered by CET1 capital. Additional Tier 1 instruments may be used to cover shortfalls revealed by the adverse stress test scenario, but their use is limited to a maximum volume of 1% of RWA. We further differentiate between the levels of conversion triggers on those instruments - the higher the trigger, the larger the amount of the instrument that may be used to cover a shortfall.
The second key aspect in respect of covering shortfalls is timing. The relevant recapitalisation actions will only be considered credible and effective if they are carried out swiftly, after the publication of results.
For this reason, we have also specified that the coverage of shortfalls is expected to take place within six months for shortfalls identified in the AQR or the baseline stress test scenario, and within nine months for those identified in the adverse stress test scenario. The six and nine-month periods will start from the release of the comprehensive assessment results in October 2014. We think these time frames strike a good balance between swiftness and feasibility. Of course we want to be ambitious, but after all, for a capital plan to be credible, its time frame also needs to remain realistic, and we definitely took this into account.
Today's last question is how we want to enforce the results of the comprehensive assessment. The SSM, the new European supervisor, will - in all likelihood - incorporate the outcome of the assessment into the yearly Pillar 2 decision. This will enable us to use the range of instruments related to Pillar 2.
As you know, these include quantitative measures, including restrictions to the distribution of dividends, limitation or even prohibition of bonus payments, prohibition of credit lending and limitations on opening up new business areas. In addition, Pillar 2 includes a number of qualitative measures (addressing management and reporting issues for example), internal controls and risk management practices. We will make use of the full Pillar 2 tool box as appropriate to address the specific situation of each institution.
Conclusion
Let me conclude by briefly summarizing some key messages: the comprehensive assessment needs to be a success, and we have catered for the necessary critical factors. First, our assessment is much more comprehensive than previous stress test exercises - not only in name!
Second, we have strict quality assurance in place for all phases of the assessment. Third, closely monitored corrective actions based on the outcome of the comprehensive assessment will make Europe's banking sector stronger and more solid.
I am aware that the process towards that goal may at times be challenging - both for banks and for supervisors. But I am sure that we can master that challenge to the benefit of all.