16 November 2014

Notes on ECB’s Comprehensive Bank Assessment (Part 7)

The results of the ECB’s Comprehensive Bank Assessment were published on Sunday, October 26, precisely as “sources” had told to Reuters a while before the date was officially confirmed by the ECB (an intended leak by the ECB’s staff?). As I had expected (recall Part 6 of this series), the final conclusion concerning banks’ capital shortfalls was rather disappointing for those who had hoped to see “blood in the water”: a bare €25 billion which, when taking into account banks’ capital measures after the 31 December 2013 assessment cut-off date, is further reduced to just €9.5 billion.

On the other hand, this ought to be a good news: rigorous review conducted by the ECB and banking supervisors has reportedly confirmed that the European banking system as a whole is almost sufficiently capitalized and after a long period of deleveraging can once again start facilitating more lending in Europe. More lending, in turn, would presumably help economic growth. (Note the words in italic. The font is intentionally different.)

Specifically, that’s what Mr. Vítor Constâncio, Vice-President of the ECB, has reported to have said in the 26 October press release:
“This unprecedented in-depth review of the largest banks’ positions will boost public confidence in the banking sector. By identifying problems and risks, it will help repair balance sheets and make the banks more resilient and robust. This should facilitate more lending in Europe, which will help economic growth.”
This very much sounds like an empty promise (read on for why I think so). Does this guy actually believe what he is talking about?

The key questions to be asked are:
  1. Are healthy banks healthy indeed?
  2. Is simply covering the small identified capital shortfall enough to render all the 130 banks sufficiently capitalized? (€25 billion may sound a big number, but when compared to €22 trillion that these banks have in assets it’s almost nothing.)
  3. Will banks start lending more now after the conclusion of the assessment AND will this actually help the economic growth?
In its result presentations, the ECB directs us to answering “yes” to the first two of the above questions by stressing strictness of the exercise as well as measures taken by the banks in preparation to the assessment and during the process:
  • Title of slide no. 2 of the Final results press conference presentation: “The Comprehensive Assessment was an exercise of unprecedented scope and depth”
  • Title of slide no. 7 of the same presentation: “Significant balance-sheet strengthening since July 2013” (The slide shows that from July 2013 till August 2014, the banks had strengthened their balance sheets in the amount of €203 billion, incl. €60 billion as gross equity issuance, €32 billion as CoCos issuance, €44 billion as a result of internal capital generation, and €67 billion via asset sales and other measures.)
Instead of solely focusing to the “key figures” of €25 billion capital shortfall, banks having to adjust their asset values by €48 billion as a result of the AQR, additional €136 billion found in non-performing exposures and €263 billion capital depletion over the three-year horizon of the exercise under the adverse scenario, we should put things into perspective. Here I’d quote a report by the ESRB (the European Systemic Risk Board) titled “Reports of the Advisory Scientific Committee. Is Europe Overbanked?” (June 2014):

“Has banking grown too much in Europe? […] To tackle the question, we take an approach similar to that of a doctor treating a patient who seems overweight. […] According to all indicators, our patient is abnormally heavy.”

Well said… The European banking system is like an abnormally heavy patient. As follows, I’ll discuss the Comprehensive Assessment results in the light of this paper. This pretty much also provides an answer to Question #3 above. 


A few things are quite obvious without going into details:
  • The problem of over indebted non-financial sector in Europe has not declined considerably although some of the private debts have been transferred to the public debts over the course of past few years. Clearly, the conclusions of the Comprehensive Assessment do not reflect the need of significant reduction in private sector debt before sustainable economic growth path can be restored. We are talking about trillions of euros; the aggregate impact to the banks that participated in the exercise is further discussed below alongside with the published data.
  • At this stage, the ECB does not seem to be willing to resolve any of the so-called TBTFs (Too-Big-To-Fail banks). The choice would be difficult too because all TBTFs seem to be more or less equally shaky. This doesn’t mean that smaller banks are better; Europe just is over banked alltogether and the solution has to be systemic.
  • Bottom-up Comprehensive Assessment results are not consistent with the top-down analyses by the ESRB (as discussed in the above referred ESRB’s paper). Other way round, there is a crying discrepancy. Apparently, system-wide perspective is not adequately taken into account in the Comprehensive Assessment exercise.
  • In order to create a false sense of confidence, in its conclusions the ECB is heavily relying on the so-called risk-weighted exposures and complex risk weightings of individual banks while not even having validated the banks’ IRB models let alone questioning Basel II / III risk weighting formulas.
  • What the assessment results are signaling is not that the “healthy” banks are healthy but that the ECB is standing behind them. Regulatory forbearance once again – what else? A debtor’s problem is ought to be a creditor’s problem; it does not seem to be like that any longer; debtors may have problems, but creditors apparently don’t. The patient is overweight and so what until he/she is a paying customer? (I’m referring to the euro area banking system as the patient and to the financing of banking supervision.)

Comprehensive Assessment data

Let’s now move on to the thousands or even tens of thousands of data points that have been published by the ECB and the EBA in conclusion of the Comprehensive Assessment exercise and the 2014 EU-wide bank stress test. You can find and download this data as well as the aggregate reports from the following web addresses.
(The referred links work as at the time of publication of this post, but they have been changed recently and may be changed in the future).

At this point, I should note that an analyst’s life hasn’t been made too easy even given the fancy analysis tools. I wanted to find quite a few figures which I consider as necessary to start with, but run quickly into the difficulties of extracting the input data:
  • Loans-To-Assets and Loan-to-Deposit ratios for each bank to get an idea of the banks’ business models (in what extent a bank is a retail bank and in what extent it’s something else, and what’s the bank’s financing model); the amounts of loans and deposits were missing.
  • AQR-adjusted amount of non-performing loans and the ratios of Non-performing loans / Gross loans for non-financial private sector and public sector which, among others, would enable a bold top-down estimate of whether the AQR adjustment has been appropriate. The amounts of gross loans and non-performing loans were missing. NPE (non-performing exposure) ratios were given but neither the amount of non-performing exposures nor the total exposure amount used for the calculation.
  • Banks’ leverage ratios after the stress test to see how many banks would fall below the 3% threshold. I could not find the necessary inputs – only some approximations seem possible.
What “unprecedented transparency” are we talking about if the most basic raw data is simply not there? (I’m referring to the “unprecedented transparency” that is being highlighted as one of the unique features of the assessment exercise.)

Furthermore, Basel Committee and the EBA know damn well that the banks’ internal model-based approaches for the calculation of risk-weighted assets are opaque. (See for example “Reducing excessive variability in banks’ regulatory capital ratios. A report to the G20” (November 2014) by the Basel Committee on Banking Supervision, or publications on the EBA’s website in section “Review on the consistency of Risk Weighted Assets”) Why the heck is much of the published data risk-weighted or given with respect to CET1 Capital ratio which is calculated based on risk-weighted assets?

Out of the thousands or tens of thousands published data points little is actually useful – and the story that this data is telling, is not comforting.


A 1.3 trillion EUR hole remained “unnoticed”


The following rough top-down calculation reveals that even after the AQR-adjustment, the amounts of non-performing exposures (NPEs) must be strongly underestimated.

The practical issue is that when “zombie” companies and -households are artificially made looking like alive (e.g. by rolling over their debts), from bottom-up it can be argued that exposures are still performing. Who and how should to take the write-downs? If households are over indebted then how could their debt burden be reduced if most of the freshly “printed money” remains stuck in the financial system (as illustrated by the very low inflation rate despite of the ultra easy monetary policies ever since the start of Great Recession) and private individuals cannot undo their past reckless borrowings due to rigid bankruptcy laws?

That’s the problem with the debt burdened companies and households: instead of investments and consumption, their income goes to interest payments and debt repayments. Therefore from certain point on, contribution of new credit to the economic growth is observed to be nil or negative. Numerous research works suggest that optimal level of credit to the private sector is at around 80% of GDP, see the above referred ESRB’s report for details.

Credit statistics by the Bank of International Settlements (BIS) reveal that credits to euro area private sector (households and non-financial companies) amount to 15.8 trillion euros. That’s way above the desirable level of 7.7 trillion euros (80% of euro area GDP). Out of the total 15.8 trillion, 9.7 trillions is in the form of bank credit. When assuming that the private sector debt burden has to be taken back to the level of 90% of GDP (ok, let’s say we are satisfied with a bit less than the optimal economic growth) and the proportion of banks vs other creditors (share of bank credit: ca. 61%) is to be retained, total euro area bank credit to the non-financial private sector should be lower by EUR 4.4 trillion than it actually is.

The Comprehensive Assessment covered ca. 82% of total banking assets in the euro area; thus, a bold estimate is that the size of the covered banks’ euro area private non-financial sector loan book in gross terms is close to 8 trillion and their share in the debt reduction amounts to ca. 3.6 trillion euros (82% times EUR 4.4 trillion).

Yeah, from the ECB’s published materials we see that the AQR-adjusted loan loss reserves amounted to a bare EUR 405 billion as of 31 December 2013 – that is less than 12% of the total debt reduction need! Before AQR-adjustments loan loss provisions were even less, ca. 362 billion. So even if the AQR adjustment led to an increase of provisions by EUR 43 billion, it’s only a bit more than a drop in the ocean. An additional 474 billion euros not covered with the provisions was classified as NPEs, thus the total amount of credit exposures that was considered as problematic, was 879 billion euros (again, after the AQR adjustment).  

The banks’ total CET1 Capital is about 1 trillion euros. Assuming a provision coverage ratio of 46% of NPEs (which it currently after the AQR adjustment is, all banks taken together) and classifying all the 3.6 trillion euros (the euro area private non-financial sector debt reduction need) as non-performing, we get that the total amount of loan loss provisions should be ca. 1.7 trillion euros, i.e. nearly 300 billion more than the amount of capital and current loan loss reserves taken together (1.7-1-0.4). Then the banks would have to restore their capital of 1 trillion euros, revealing a hole of 1.3 trillion euros. 

The described calculations are illustrated in the Figure below. Note that so far we haven’t even discussed public sector debts and credits to other financial firms. Nor have we considered credits outside the euro area. So the 1.3 trillion euros is really just “a hole” that has remained “unnoticed”.  

(Click to enlarge)

Does this mean that we are going to see a series of bank failures in the near future? Not necessarily – far of it unless things get out of control and a “race to the bottom” is triggered (a possible but still unlikely scenario). The ECB now at least tries to close the identified gap (the ECB’s staff just has to have seen it) by using other means such as starting with an extensive QE – the Quantitative Easing or Money Printing, whatever name one might prefer.  


Once again, the assessment is demonstrating big banks’ ability to game regulation.


What else could one conclude when looking at the figures below?

The bigger the bank in terms of total assets the lower its risk exposure (risk-weighted assets) relative to non risk-weighted exposure (the "Leverage Exposure"):

(Click to enlarge)

The bigger the bank the lower its leverage ratio, calculated as CET1 Capital divided by non risk-weighted exposure (“Leverage Exposure”):

(Click to enlarge)

The higher the leverage (the lower the leverage ratio) the smaller the AQR adjustment to the CET1 Capital ratio (recall the previous figure: the bigger the bank, the lower the leverage ratio; relationships are not functional or even statistically reliable, but they are still observable):

(Click to enlarge)

For the top 20 banks in the assessment, regulatory capital ratio has little relation if any with the leverage ratio (the one bank that stands out as an outlier with the leverage ratio of 8.6% is Merrill Lynch International Bank Limited, i.e. a bank not headquartered in Europe):

(Click to enlarge)

One can argue that bigger banks are more sophisticated and less risky, but this argument alone cannot possibly be enough to justify the above observations.



I could go on, but I’ll stop here. I think I have made my point: the conclusions of the ECB’s Comprehensive Bank Assessment are driven by other considerations than the actual findings could possibly have revealed.

1 comment:

  1. I like this post to read and is a very useful information. Instead of solely focusing to the “key figures” of €25 billion capital shortfall, banks having to adjust their asset values by €48 billion as a result of the AQR. If you want essay related topic you can go for custom essay writing service will get better result.

    ReplyDelete