The Trouble with Stress Tests Critics Question Soundness of Bank Check-Ups

The European Union has agreed to publish the results of stress tests on a number of the bloc's leading financial institutions. The problem, critics say, is that the one-size-fits-all scenarios could fail to identify the potential for some systemic risks.
Von Beat Balzli, Christian Reiermann und Wolfgang Reuter
Deutsche Bank CEO Josef Ackermann: A spring stress test performed on European banks included 22 banks representing around 60 percent of total assets in the European banking sector. German banks are reported to have had "decent results."

Deutsche Bank CEO Josef Ackermann: A spring stress test performed on European banks included 22 banks representing around 60 percent of total assets in the European banking sector. German banks are reported to have had "decent results."

Foto: A3796 Uwe Anspach/ dpa

In the end, the controversial stress tests for banks did justice to their name at the very highest levels of government, as they became a source of stress for European leaders meeting in Brussels last week.

On Thursday, the European Council, the European Union body comprised of the heads of government and state of the European Union, discussed the question of whether the results of the latest stress tests performed by the Committee of European Banking Supervisors (CEBS) in the spring should be published. Or perhaps it wasn't such a good idea, after all?

Germany apparently has nothing to fear from the results achieved to date. In addition to Deutsche Bank and Commerzbank, a state-owned bank was subjected to a stress test. The banks themselves are not commenting on the tests, but insiders say that they ended up with "decent results." The Spaniards are also confident of success, because their major banks did relatively well. Their ailing savings banks, on the other hand, did not participate in the test.

A total of 25 lenders were studied in the process, which involved two scenarios. The underlying premise of the basic scenario was moderate development in economic output until 2011. In the so-called adverse scenario, on the other hand, the analysts assumed that there would be a sharp decline in the gross domestic product. Both scenarios contained concrete values for Germany, Great Britain, the United States and the euro zone, and the banks then simulated the effects of these values.

CEBS assumed that the rating agencies would continue to drastically downgrade inventories of highly complex securities into which American subprime mortgages and corporate loans were packaged. The goal was to test how long the banks' equity levels would last in extreme cases. The lower the ratings of these toxic assets, the more capital the banks must hold in reserve as a safety cushion against losses.

'Not Particularly Meaningful'

The bank stress tests were first performed throughout the EU in the summer of last year. Some 22 banks, which represented about 60 percent of total assets in the European banking sector, took part in the study.

But some experts question the purpose of such tests. "The CEBS stress test is not particularly meaningful," says a senior risk manager at a major German bank. He argues that the test is performed on a "one size fits all" basis for all banks, even though there are sometimes drastic differences among their business models. Effective bank supervision, says the risk manager, would require more differentiated and significantly more stringent tests.

"Besides, the tests do not take the systemic concentration risk into account," says the risk expert. The current tests, he argues, are not capable of assessing the risk of domino effects. This would require a centralized system to record the banks' positions in the foreign currency, derivatives and stock markets.

Although European politicians specializing in financial matters recognize this risk, they believe that the most important criterion is comparability, which only a uniform test can provide. The results of such a test should also be publicized, as the politicians in Brussels agreed last Thursday. The United States, they argued, managed to significantly improve confidence in banks with a similar procedure more than a year ago.

Nevertheless, the politicians in Brussels decided that the test performed in the spring should not be published. The results, they argued, are simply too old and are not very meaningful, because the European financial system was on the brink of collapse only a few weeks after the tests were performed. The Greek crisis had come to a dramatic head, and the banks were starting to distrust each other again and were refusing to lend money.

The finance ministers must now agree on a new stress scenario. A larger number of banks would take part in the new test, although the exact number is still unclear. Publication of the results is expected in mid-July.

Translated from the German by Christopher Sultan