Amongst all the criticisms of the European stress tests, there’s one glaring omission.
From the Committee of European Banking Supervisors’ summary report:
Government support measures received by institutions in the sample as of end 2009 have been taken into account and subject to specific analysis (see Section 4.5 of the report). The actual maturity of most government support measures and instruments goes beyond the time horizon of the exercise.
There’s something immediately incomprehensible about a stress test sparked by sovereign worries that ends up including sovereign support in its assessment — seemingly across the board, and without differentiation between those sovereign states. Such support measures may include capital injections, guarantees of assets, liabilities or funding, plus government and central bank-provided liquidity.
The stress test exercise takes into account two of these — pre-July 2010 capital increases through equity or hybrid instruments provided by governments, and government guarantees of bank assets.
According to the CEBS, 20 banks in the 91-bank sample “benefited” from asset guarantees. Meanwhile 34 banks benefited from €170bn of aggregate capital increases, making approximately 14 per cent of the total Tier 1 own funds of the banks in the sample. In sum then, including government support measures helped boost the aggregate stressed Tier 1 capital ratio by 1.2 percentage points.
The CEBS calls this a “fraction” of the total end-of-2011 stressed Tier 1 capital ratio of 9.2 per cent, which is true. But the differences between banks here look to be important. As the CEBS points out, the support measures added anywhere between 0.1 and 11.1 percentage points to banks’ stressed Tier 1 ratios. Government support may have made all the difference to some financials.
The CEBS says these government support measures were included in its stress test capital calculations since the overwhelming majority have “a useful life extending beyond the horizon of the [two year] exercise.” This might be consistent with the European Union’s comments that no sovereign will be allowed to fail, but as previously mentioned, it just doesn’t seem quite right.
For a start the banks with significant government support — perhaps with the exception of Germany and the Landesbanks — are often located in some of the weaker European countries. Think Greek banks, which have raised about €9.4bn via government injections. Or the Spanish cajas and injections from the FROB restructuring fund. Or the Irish banks and their government-induced capital.
On a different note, the CEBS’s attitude towards support is rather at odds with some recent capital market developments. In other words some of the ‘support’ included by the CEBS has already died.
The rating agencies, for instance, began stripping implicit government support from their assessments of banks’ hybrid bonds late last year after the European Commission basically turned on hybrid capital. The banks had to scramble to replace the subordinated stuff with new funding.
It’s now widely-accepted that old-fashioned hybrid bonds could be on their way out of banks’ capital structures, but in the CEBS’s stress tests the remaining hybrid stuff is still all good.
It still counts towards stress test Tier 1.
Just one instance of, as Wolfgang Münchau notes in Monday’s FT, the test parameters being rather cynically calibrated to achieve the desired result.
FT Alphaville:
Credit distortion de crise – FT Alphaville
Notch ‘em up, and notch ‘em down - FT Alphaville
Credit rating cliff risk – FT Alphaville