Some of the largest banks have been in touch to complain about the recalculations of Europe’s stress test data that I published last week. Not a big surprise that they would. But I’ve accepted one of their arguments. Projections made by the European Banking Authority (EBA) for risk-weighted assets in the final year (2016) of the adverse scenario should be used, not latest available figures (from the Bankscope database).
My inclination is to use Bankscope data wherever possible rather than EBA data (to minimize effects on data of various forms of ‘regulatory massage’) . But for this calculation, it makes sense to use the EBA projections, to ensure as much as possible the internal consistency of the exercise.* weighted assets in 2016.
Two other changes have been made. To complement the 3 % threshold with a higher threshold, I’ve adopted a 5 % threshold (instead of the 7 % threshold in the first set of calculations). The latter remains relevant as a (fairly conservative) benchmark for what scholars believe would be appropriate, but the former is relevant for another reason. The ECB has argued that this time the European exercise was as strict as its equivalent in the United States, in terms of the severity of the adverse scenario.** But even if that had been the case, the argument still would have been flawed. In US stress-tests a leverage ratio is explicitly included, as one of several thresholds that banks are required to pass, contrary to what’s been the case in Europe, and this is the main determinant of the severity of stress tests.
Direct comparisons of US and European leverage ratios are complicated by different accounting practices. But the US is now close to internationally agreed definitions (Basel 3). The required minimum leverage ratio in US stress tests is 4 %, but even stricter for their biggest banks, which are now required to meet a 5 % threshold. In brief, the 4 and 5 % thresholds allow us to assess the claim made by Europe’s banking supervision institutions that its stress tests are now as tough as those of the US.***
Last but not least, I’ve expanded the set of banks covered in the calculations, from the original 11 to 32 of the largest banks in the Eurozone.
The results of the new calculations are given in Table 1 (stressed leverage ratios) and Table 2 (capital shortfalls) below. The main results are as follows:
- 18 out of the 32 banks fail on the criterion of a 3 % leverage ratio (measured as ‘fully loaded common equity Tier 1 capital’ to total assets) ****
- The aggregate capital shortfall is 80 bn euros (relative to the 3 % equity funding threshold), several multiples of the capital shortfall identified by the ECB for all 130 banks
- 29 banks fail the 4 % threshold for equity funding. If Europe’s stress tests had really been as tough as their US equivalents, only 3 of their 32 largest banks would have passed them, in other words.
- If Europe’s banking supervision authorities really wanted to aspire to US standards with respect to regulatory requirements to ensure bank solvency, they would need to demand of these banks to increase their equity funding by between 200 and 360 bn euros (depending on how many of the banks would be required to meet the 5 % threshold).
- The resulting capital shortfall figure for this set of banks would likely be at least 10-fold larger than the capital shortfall identified by the ECB for all 130 banks (24 bn euros).
This data exposes any claim that Europe’s 2014 stress tests were as strict as their US equivalent as somewhat off the mark. A new post, coming up shortly, will analyze in detail how the ECB makes a case to the opposite effect, exposing how misguiding its argument is.
In the meantime, I recommend an excellent essay by Morris Goldstein, outlining the main shortcomings of Europe’s stress tests (in much the same spirit as you’ll find on this blog).
Table 1 Leverage ratio for 32 Eurozone banks, EBA adverse scenario, Y2016 (bn euros
|CET1*(%)||RWAs||Total assets||RWA to TA||Leverage ratio (%)|
|ING Bank NV||8,20||344,1||787,6||0,44||3,58|
|Groupe Credit Agricole||8,60||621,4||1706,3||0,36||3,13|
|Société Générale SA||7,10||377,1||1235,3||0,31||2,17|
|Banco Santander SA||7,30||563,9||1115,6||0,51||3,69|
|Banco Bilbao Vizcaya Argentaria, SA||8,20||381,3||582,6||0,65||5,37|
|Intesa Sanpaolo SpA||7,80||310,0||626,3||0,49||3,86|
|Banca Monte dei Paschi di Siena||-3,50||76,0||199,1||0,38||-1,34|
|Banco Popolare Società Cooperativa||3,60||55,6||126,0||0,44||1,59|
|Unione di Banche Italiane SCpA||7,90||64,5||124,2||0,52||4,10|
|Caja de Ahorros y Pensiones de Barcelona||7,50||176,3||351,3||0,50||3,76|
|Banco Financiero y de Ahorros SA||8,60||99,3||269,2||0,37||3,17|
|Banco de Sabadell, SA||7,80||79,9||163,4||0,49||3,81|
|Crédit Mutuel Group||12,80||252,3||658,6||0,38||4,90|
|Erste Group Bank AG||6,80||109,6||199,9||0,55||3,73|
|Raiffeisen Zentralbank Österreich AG||3,90||92,9||147,3||0,63||2,46|
|KBC Group NV||6,30||102,7||241,3||0,43||2,68|
|National Bank of Greece SA||-7,30||57,9||110,9||0,52||-3,81|
|Piraeus Bank SA||-1,50||59,7||92,0||0,65||-0,97|
|Bank of Ireland||2,90||54,4||132,1||0,41||1,19|
|Allied Irish Banks, Plc||-3,60||66,7||117,7||0,57||-2,04|
|Coöperatieve Centrale Raiff.-Boerenleen||7,10||244,6||674,1||0,36||2,58|
|Caixa Geral de Depósitos SA||4,90||65,4||113,0||0,58||2,84|
|Banco Comercial Português SA||-0,30||45,3||82,0||0,55||-0,17|
Note: CET1* is “fully loaded Common Equity Tier 1 Capital” (here in per cent of RWAs), cf EBAs stress test report (Annex 1, Table 5). RWA: risk-weighted assets (data from EBAs bank-specific stress test results). TA: total assets.
Table 2 Capital shortfalls for 32 Eurozone banks, EBA adverse scenario, Y2016 (bn euros)
|Total assets||CET1*||Capital shortfall, 3 % threshold||Capital shortfall, 4 % threshold||Capital shortfall, 5 % threshold|
|ING Bank NV||787,6||28,22||3,29||11,17|
|Groupe Credit Agricole||1706,3||53,44||14,81||31,88|
|Société Générale SA||1235,3||26,77||10,29||22,64||34,99|
|Banco Santander SA||1115,6||41,17||3,46||14,61|
|Banco Bilbao Vizcaya Argentaria, SA||582,6||31,27|
|Intesa Sanpaolo SpA||626,3||24,18||0,87||7,14|
|Banca Monte dei Paschi di Siena SpA||199,1||-2,66||8,63||10,62||12,61|
|Banco Popolare Società Cooperativa||126||2||1,78||3,04||4,3|
|Unione di Banche Italiane SCpA||124,2||5,1||1,12|
|Caja de Ahorros y Pensiones de Barcelona||351,3||13,22||0,83||4,34|
|Banco Financiero y de Ahorros SA||269,2||8,54||2,23||4,92|
|Banco de Sabadell, SA||163,4||6,23||0,3||1,94|
|Crédit Mutuel Group||658,6||32,29||0,64|
|Erste Group Bank AG||199,9||7,45||0,54||2,54|
|Raiffeisen Zentralbank Österreich AG||147,3||3,62||0,8||2,27||3,74|
|KBC Group NV||241,3||6,47||0,77||3,18||5,6|
|National Bank of Greece SA||110,9||-4,23||7,56||8,67||9,78|
|Piraeus Bank SA||92||-0,9||3,66||4,58||5,5|
|Bank of Ireland||132,1||1,58||2,39||3,71||5,03|
|Allied Irish Banks, Plc||117,7||-2,4||5,93||7,11||8,29|
|Coöperatieve Centrale Raiff.Boerenleen||674,1||17,37||2,86||9,6||16,34|
|Caixa Geral de Depósitos SA||113||3,21||0,18||1,31||2,44|
|Banco Comercial Português SA||82||-0,14||2,6||3,42||4,24|
Note: CET1* is “fully loaded Common Equity Tier 1 Capital” (here in bn euros), cf EBAs stress test report. RWA: risk-weighted assets (data from EBAs bank-specific stress test results). TA: total assets.
Jakob Vestergaard, GEG Watch, 29 November 2014.
Photo by Liam Levitz
* In terms of what is the better estimate of equity funding shortfalls of Europe’s banks, I have little doubt that the first set of data by far defeats this new one, but that’s another matter.
** “In terms of overall severity, the magnitude of the cumulative shocks in the EBA stress-test exercise is broadly in line with those of the 2014 USA Federal Reserve’s CCAR adverse scenarios. Overall, in terms of peak level deviations from baseline GDP, the EBA stress test adverse scenario lies roughly between the CCAR ‘adverse’ scenario and the CCAR ‘severely adverse’ scenario” (ECB 2014: 38).
*** For further information on US stress tests, see the Fed’s Comprehensive Capital Analysis and Review (CCAR, March 2014). For the 4 % threshold for leverage; see the table on ‘required minimum capital ratios for advanced approaches’ (p. 13).
**** When Europe eventually adopts a leverage ratio it is likely that it will be measured as Tier 1 capital relative to total assets, hence allowing so-called Additional Tier 1 capital to be included (as opposed to a purer measure of only Common Equity Tier 1 capital). This is the direction the deliberations of the Basel Committee have taken recently, although the committee also say that it “will continue to collect data… to track the impact of using either Common Equity Tier 1 (CET1) or total regulatory capital as the capital measure for the leverage ratio” (BIS 2014:2).