27 November 2013

Notes on ECB’s Comprehensive Bank Assessment (Part 2)

In his opening speech at the European Banking Congress “The future of Europe” (22 November 2013), Mr. Mario Draghi, the President of the ECB, said about the ongoing ECB’s comprehensive bank assessment:
“It is clear that there needs to be much more confidence in banks within and across countries that are joining the SSM. This is the objective of the ECB’s comprehensive assessment.”
Then he briefly explained how the assessment is ought to boost confidence. The keyword was/is “transparency”: giving all parties more transparency.

As was discussed in Part 1 of this series, more transparency about the banks’ balance sheets is badly needed indeed. So this sounds good as far as it goes. But… wait a minute: is it going to be real transparency for the investors and public in general, or is it rather that the outcome of the exercise is decided already and the key issue for the ECB is to build a credible process around it? (The pre-decided outcome would be something in style: all systemically important banks are fundamentally sound; the banking system as a whole just needs x billion euros of additional capital where “x” is a number that sounds big enough but is still manageable with limited private sector involvement and not too much public money.)

As also implied in Part 1, the amount of fresh capital that is actually needed for plugging the holes in the banks’ balance sheets is fairly big, measured in trillions rather than in billions of euros. I’d even say: none of the systemically important European banks would survive a fair stress test without external support. So in my notes, I’m first of all trying to follow how the credible process is being built – or the illusion of it. I also highlight open issues and “devils in details” should I observe them.

Just to remind, the assessment will consist of three elements:
1) a supervisory risk assessment to review, quantitatively and qualitatively, key risks, including liquidity, leverage and funding;
2) an asset quality review (AQR), i.e. reviewing the quality of banks’ assets, including the adequacy of asset and collateral valuation and related provisions; and
3) a stress test to examine the resilience of banks’ balance sheet to stress scenarios.

The figure below presents the timetable of the exercise as communicated so far.

(Click to enlarge)

General notes

Whatever else it might be, the ECB’s comprehensive assessment is definitely:
* Extensive in scope;
* Heavily technical exercise;
* Nothing conclusive but work in progress;
* Huge communication effort and –experiment;
Let me explain.

Concerning the scope, the ECB is going to get into its disposal an unprecedented amount of detailed banking data: from ca. 130 credit institutions in 18 eurozone Member States (current Member States plus Latvia which will be joining the euro starting from 1 January 2014), covering ca. 85% of euro area bank assets. It’s not only about the aggregated stats on pre-defined data templates. Much of this data will most probably be on customer and exposure level. While it can easily be argued that collecting this data is absolutely necessary for meeting the objectives of the exercise, it also raises serious questions about data protection and concentrating enormous powers into the hands of the ECB officials in the era when banking is increasingly becoming an information business rather than being mainly money business. It even doesn’t matter so much if they will be able to identify each bank customer or not; based on the information gained from the Big Data, one can make reasonable predictions about you even if knowing little about you personally.

“Heavily technical exercise”… That’s the wording used by Mr. Ignazio Angeloni, head of the ECB’s financial stability division, in the 23 October press briefing. What it means is first of all a lot of work with the data templates and making data more or less comparable across banks and –jurisdictions. Different data recording and reporting practices, different accounting principles, different supervisory practices… it all needs to be dealt with somehow. The EBA is preparing and publishing new technical standards which, in the extent possible, are ought to be followed in the on-going bank assessment. In some cases, the implementation of these new definitions requires rather large scale developments in the banks’ IT-systems.

Another technical as well as methodical aspect is how exactly the collected data is going to be analyzed. Yeah, as discussed below, there is even no reliable regulatory basis for the assessments, let alone uniform methodologies and clearly defined principles for judging if a bank is solvent or insolvent.

Clearly, while consolidating the different supervisory practices and comparing banks’ data, bank supervisors will learn a lot. However, given the technical challenges, all the exercise should be viewed as work in progress rather than expected to deliver something final.

In terms of communication, the European authorities very much seem to have taken the route of over-promising and under-delivering.

Consider these statements by the European Commission, for example:

“The Council IS OF THE OPINION that the new supervisory set-up within the ECB provides the necessary assurance to the market in terms of rigour of the comprehensive assessment.” 
[Comment: An optimistic statement that assumes a great deal of market ignorance about the current state of affairs.]

The above paragraph continues:
“The involvement of an independent third party and the use of harmonised definitions, including for Non Performing Loans (NPLs) and forbearance as well as resulting adjustments of Risk Weighted Asset (RWA) calculations will further support it.”
[Comment: Well, if they are referring to Oliver Wyman as the “independent third party” – which they most probably are – then I’m not so sure, see the reason below. Furthermore, the practical implementation of the referred harmonized definitions will not be that harmonized at all; also discussed below. Last but not least, adjustments to RWA do not involve assessment of banks’ internal risk classification models as we can find out from the 23 October Note to Comprehensive Assessment.]

“In this regard, the Council WELCOMES the communication by the ECB of 23 October, which clearly sets out the key features of the comprehensive assessment.”
[Comment: Well, as illustrated below, the communication wasn’t that clear at all – especially when it comes to concluding based on the assessment results.]

Will they succeed in convincing banks’ stakeholders and public in general? That’s the experiment. For the starters, press is invited to co-operate (read: tell to the public what it is instructed to tell):
“It is of crucial importance not only how we present the material, but also how the material is reported. That is why we are offering our cooperation to you and asking for your cooperation in return. The material needs to be reported correctly, swiftly, thoroughly and in full to the relevant audiences.”

There are a few more general issues to be pointed out.

One of these is that while consulting both, financial supervisors and banks, the selected external consultant Oliver Wyman in a way plays the roles of both, “poacher” and “gamekeeper”. They may be the best professionals in the field, but there is certainly a conflict of interests involved.

Another is that as banks go through the process they might be assumed to de-risk and de-leverage which via clogged lending channel is likely to have adverse effects to the real economy in Europe. Even though regulators are aware of it and monitoring the developments, they do not seem to have other choices but to make compromises in the rigorousness of the assessments and/or introduce the European version of the Troubled Asset Relief Program (TARP), both “solutions” of which would cause further market distortions and encourage banks behaving irresponsibly also in the future. Alternatively, the decision to resolve certain banks would have to come as surprise to the bank stakeholders so that shareholders and unsecured creditors of such banks could not “escape” losses. In a way the later idea is appealing as it enables to write off more bad debts and thus reduce overall debt overhang; the downside is that consequences of such shock therapy would be hard to predict.

Notes about the supervisory risk assessment

As of now, not much can be said about this “Pillar”: the ECB and the national competent authorities (NCAs) still have to develop a new risk assessment system which is partly going to be used in the bank assessment exercise. It is promised to be a scoring system that draws on the best practices of the existing supervisors at the national level. Hmm… existing supervisory practices clearly haven’t been good enough…

Supposedly, the pass/fail criteria will more or less be based on the Basel III regulatory framework. The additional elements will be incorporated in CRD IV only in the beginning of 2014; that’s when more clarity can be expected.

Notes about the asset quality review (AQR)

Today concerns about the banks’ balance sheets mostly relate to uncertainty surrounding (i) the extent of the use of forbearance, potentially aiming at, or in practice leading to, delaying loss recognition and masking asset quality deterioration, and (ii) the consistency of asset quality assessment across the EU, particularly regarding the line drawn in the different jurisdictions between performing and non-performing categories. Thus, AQR is the core of the whole exercise: no supervisory risk assessment or stress test has point until the current actual situation is known.

The basis of the exercise is the EBA Recommendation on the conduct of asset quality reviews. The technical details are specified in the EBA’s document “Implementing Technical Standard (ITS) on Supervisory Reporting on forbearance and non-performing exposures” (EBA/ITS/2013/03). Mainly it’s about introducing two new unified definitions for the exposures of questionable quality, this in addition to the existing accounting and regulatory definitions of impairment and default: non-performing and forborne exposures.

I wonder what the outcome of the AQR might be; a re-stated balance sheet for each participating bank, perhaps?

Anyway, public consultation regarding the draft technical standards provides some good insights about the issues related to conducting the AQR. For example, some respondent said that in their current reporting systems it is impossible to distinguish refinancing stemming from financial difficulties from other refinancing. That’s rather telling about the quality of the data that goes into the AQR...

Furthermore, as according to the ITS the definitions of forbearance and non-performing exposures will not replace the definitions of impaired or defaulted assets, or be used as an input in the computation of incurred losses, risk-weights and regulatory capital amounts, it remains unclear in what extent and how the insights gained during the AQR will be reflected in the assessment result.

In the 23 October Note to Comprehensive Assessment, the exercise is told to be broad in scope and cover:
* Sovereigns and institutional (including interbank), corporate and retail exposures;
* Both the banking book and the trading book;
* On-balance sheet and off-balance sheet exposures.
It means looking specifically at individual asset items, individual exposures: are they allocated correctly, is the risk on that particular exposure calculated correctly etc.

That sounds comprehensive, but of course, supervisory resources are limited and they don’t have the capabilities to look in detail into the billions of items for all banks. So the review is communicated to be risk-based and concentrating on those elements of individual banks’ balance sheets that are believed to be most risky or non-transparent (subject to minimum coverage criteria). Sampling procedures will be applied.

As a further limitation, full assessment of banks’ internal models used for the calculation of risk-weighted assets will not take place within the time frame of the exercise:
“The thing that we thought that we would not be in a position to do in the time frame, is to go back and re-check and validate the models which were used to calculate those presumably not fully correct risk weights.” (Ignazio Angeloni, Q&A session in 23 October press briefing)

So it’s obvious that the denominator of the capital adequacy ratios (i.e. risk-weighted assets) can still be gamed... In other words, at this stage I’d not really trust any conclusions based on the regulatory capital ratios. I’d still look at the simple leverage ratios, calculated based on the restated balance sheet figures.

This paragraph from 24 October Euromoney article is pointing to a major issue with the AQR, namely time pressure:
“In Washington in October, the coffee-break chatter among European bank regulators mainly concerned fears that the time pressure on the ECB to conduct its review of Europe’s largest banks before becoming their lead regulator next year might preclude it from doing a thorough job.”

Yeah, the ECB, the EBA and NCAs have a full year, but the ECB is still in the process of hiring and/or training staff. Indeed, when reading the transcript of the ECB’s 23 October press briefing, I get the impression that even if quantity is not a problem, finding sufficiently qualified resources is definitely a challenge.

The takeaway: the outcome of the AQR can possibly be preliminary at best.

Notes about the stress test

I’ll comment on this “pillar” of the comprehensive assessment when the details on the stress test, the methodology and the scenarios to be used will be published. It is expected to happen in the beginning of 2014.

Notes about overall conclusion and disclosure of assessment results

No intermediate results are planned to be published. The overall conclusion that will come at the end, will involve a great deal of subjective judgment of the supervisors. There will be insights collected during the exercise as well as certain criteria such as capital threshold that sound objective (tough can be manipulated), but at the end the decision of which banks will “pass” and which banks will “fail” the test, will be a judgment and not mechanical. It may also be that all participating banks are going to be required to take certain measures and/or accept sort of TARP money.

In the above referred press briefing, Mr. Ignazio Angeloni’s answer to the rightful question: “How do you ensure that you are doing the same for every bank then?” was that “this is the normal life of the supervisor, right?”
Is the judgment going to be fair or driven by some sort of other motives, we don’t know.


I think, the short answer to the question: “Is it going to be real transparency for the investors and public in general?” is that the details of the exercise are far from being clear for an outside observer. The usefulness of the assessment to the bank outsiders will very much depend on the data (data, not interpretations) that will be published at the end. More specific communication in this regard would make the whole thing much less obscure.

The main quality assurance to the stakeholders is perhaps that the ECB puts its own reputation at stake by supervising the whole exercise and as the lead regulator from November 2014. It more or less gives the ECB’s guarantee to the banks that will pass the test and/or take the prescribed follow-up actions; at the same time, it doesn’t mean an end to the practice of keeping zombies alive on the expense of public in general (either via direct or hidden costs and taxes). The fact that the ECB will have much more data and information at its disposal only means that it will have more power, for good or bad – depending on people in charge today and in the future. 

1 comment:

  1. please keep sharing of knowledges with us.Thanks a lot for your great posting.
    Instant Annuity Rates