Fourth time lucky
How big is the capital problem of euro area banks? The European Banking Authority – the pan-European banking regulator – will in 2014 try to gauge the balance-sheet strength of banks in the monetary union for the fourth time in as many years. Unlike the past three attempts, this one may finally address one of the biggest problems – the impenetrable fog surrounding European banks’ bad debt data.
Before the EBA’s stress tests, the European Central Bank will in the first quarter of 2014 conduct something it calls an “asset quality review”. For the exercise to have any meaning, it will need to rely on a common method for assessing non-performing loans throughout the bloc. The ECB’s findings will then serve as a base for the EBA’s test.
One just has to look at consolidated banking data published by the ECB to see why the current bad loans framework needs a complete overhaul. The scale of the loans and debt securities of each member country’s banks are disclosed, as well as what each national regulator considers bad loans, and the percentage of that exposure that is already covered by provisions.
In theory, that should allow investors to undertake their own quick and dirty stress tests. Say that the bad loans on each state’s disclosed assets spike by 50 percent from current levels, and that banks can recover 50 percent of their losses. Then compare that to their existing provisions, and work out what the shortfall is.
On that basis, Spain and Greece, which have already built up buffers against painful recessions, score relatively well: provisions would cover 84 percent and 64 percent of the potential losses. Each country would only need to find an extra 12 billion euros for its banking sector. Germany and Italy, on the other hand, look more troubled: under the same scenario, they would have to find 36 billion euros and 131 billion euros, respectively, to shore up their banks.
The problem is that these figures are almost certainly wrong. One reason is that they don’t factor in future profit and existing capital buffers, which today are three times higher than in 2008. Another is that they are based on a crude average loss rate, rather than more granular individual estimates for different assets in each country.
But the real problem is the bad loans data itself. The euro zone’s seventeen different regulators have seventeen different interpretations of when a borrower has defaulted. That not only makes the stock of provisions suspect. It also means stresses applied to the non-performing loans base will be inaccurate, because the base itself is misleading.
The discrepancy between Germany and Italy is telling. On the face of it, Germany appears to have much better banking assets: only 1.7 percent of its 5.6 trillion euros of total debt instruments and loans was deemed non-performing as of 2012. Italy looks much worse: 10.9 percent of its 3.4 trillion euros of debt and loan assets was considered non-performing.
Yet using this data means comparing apples and pears. The Italian regulator wants banks to disclose all loans with terms that have been eased to help the borrower, and these “restructured loans” must be treated as impaired. In the German system, on the other hand, restructured loans aren’t disclosed, let alone treated as impaired. Italian banks must also class all loans from a borrower that missed payments as in default. That’s not the case for German banks.
What’s more, the Bank of Italy requires its charges to report bad loans net of any collateral, whereas German banks don’t have to. Were Italian banks allowed to reduce their NPLs in the same way, the average coverage on their impaired loans would leap from 41 percent to 89 percent, according to Deutsche Bank.
Harmonisation could therefore mean banks with strict regulators, like Italy, see their bad loans revised down. Countries with looser rules like Germany may have to adopt a more comprehensive definition of bad loans. If the harmonised criteria are adopted ahead of the EBA’s stress test, the volume of capital needed by each national banking system will be markedly different from what the current ECB data implies.
It may be for precisely that reason that some officials have started briefing that a standard definition of bad loans may not be a requirement. As it happens, that would protect German banks from the bad news that they need to raise a serious amount of capital. That would of course deprive the EBA test of any credibility, and get the mooted European banking union off to a rotten start. The ECB would be well advised to tackle that political challenge head on.