The silver lining to the lenient stress test

By Felix Salmon
July 23, 2010
nothing in the way of bad news coming out of the European bank stress tests. Essentially, the tests could be tough or they could be lenient; and there could be significant failures or no significant failures. The outcome, in the end, was that the tests were lenient and that there were no significant failures.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

There’s a certain amount of relief in the markets that there’s absolutely nothing in the way of bad news coming out of the European bank stress tests. Essentially, the tests could be tough or they could be lenient; and there could be significant failures or no significant failures. The outcome, in the end, was that the tests were lenient and that there were no significant failures.

Clearly, a tough test with no significant failures would have been better news — but at the same time, a lenient test with significant failures would have been worse news. So we’re somewhere in the middle, and muddling along: in the CDS market, most banks have seen their spreads tighten modestly, by single-digit amounts, according to Markit.

The European stress tests are less credible than the US stress tests, for reasons laid out by Chris Whalen and many others. They’re a bit of a europudding, but they’re better than nothing. And it’s worth remembering that a lot of people had similar reactions to the US stress tests, too, immediately before and immediately after they came out — few people at the time reckoned that they would have much effect in terms of restoring confidence and credibility to the interbank market.

But they did help, a lot, mainly because they forced US banks to raise a lot of new capital. The European stress tests don’t do that, and therefore they look like a bit of a waste of time. Remember that the markets, like the rest of us, prefer to see actions to words, and the actions which followed the US stress tests undoubtedly strengthened the US banking system. By contrast, the European stress tests don’t seem to require any extra capital-raising whatsoever from any banks that anybody cares about. Only seven banks failed; three are state-owned already (Germany’s Hypo, Spain’s Cajasur, and Greece’s ATEBank), and the other four are tiny Spanish banks you’ve never heard of.

In any case, the only stress test that global markets really care about, in the European context, is a Greek default. And this stress test didn’t even attempt to model what might happen to European banks’ balance sheets in that event.

Was there any point in doing this test? I think so, yes. It got all the European national bank regulators talking to each other and cooperating more than they do normally, and thinking hard about important questions related to the solvency of the European banking system. The test itself was weak, to be sure. But the Committee of European Banking Supervisors has a lot of granular information now which it can try to use for the purposes of crisis prevention. Whether it will or not is unclear, as is whether it will succeed if it tries. But its very existence is probably a good thing. So even if the stress tests look like a joke, they might have something in the way of positive consequences.

11 comments

Comments are closed.