14 January 2012

EBA to Postpone Stress Test for Banks: What’s next?

In 13 January the news spread all over the financial media: “The European Banking Authority (EBA) will this year postpone the annual stress test for banks usually published in July, Handelsblatt reported, citing a spokeswoman for the EBA.” (That’s how the financial media works: someone tells something that sounds good for making headlines, and hundreds of financial websites parrot it, adding little if anything to the original message or even distracting from it.)

Q: Why?
A: Reportedly, there is no point to run a new Stress Test until the banks haven’t completed their re-capitalisation tasks based on the conclusions of the 2011 EU Capital Exercise.

Q: Is the data being collected and published?
A: No, there will be no data collection.

Q: When will the regular stress testing exercise take place?
A: Well, maybe later this year, but maybe not at all in 2012.

(Source of the answers above: Handelsblatt, 12 January 2012)

Q: What’s next?

Brief “history” of the EU bank stress testing

I guess I’m not seriously mistaken when saying that the main motive behind the introduction of such centrally coordinated stress tests was to calm everyone down: all systemically important banks can withstand some pretty dire economic scenarios, and only a little additional capital might be needed for smaller banks. What came out in Europe from the regular stress testing exercise is a complete failure: Irish banking system collapsed barely two months after its largest banks had passed the 2010 stress test, Dexia failed after comfortably passing the 2011 stress test, there has been quite a panic around Commerzbank and some of the French banks etc. The intention was to stabilize the situation, the effect was rather destabilising.

Generally, the explanation to the failure has been found in the scenarios, which were “not stressful enough” as they didn’t include sovereign default. (Guess, what is the main criticism also about the U.S. bank stress 2012? According to CNNMoney: “The scenarios they are testing don't even include what's top of mind for investors right now: Defaults in the euro zone.”) The other major articulated problem with the European stress tests was that (although in line with the accounting rules), the sovereign debt exposures in the banking book were not marked to market, this despite that the banks actually had to take a 21% loss for Greek bonds with maturities before or in 2020 already in Q2 results of 2011 (thus for a report date which was before the stress test results were published on July 15).

Whether the problem was in the scenario, in the methodology or somewhere else (actually there were flaws everywhere), the European banking system just was / still is undercapitalised. Early autumn last year the situation in Europe worsened rapidly (which should not have been a surprise). On Tuesday October 11th, the then outgoing European Central Bank (ECB) president Jean-Claude Trichet warned that the European financial crisis had become "systemic" and demanded swift action, which was a loud call to re-capitalise the banking system. This time, the EBA did not calculate the additional capital need based on a stress test, but based on the latest available actual data and a target Core Tier 1 capital ratio of 9%. The exercise was called “2011 EU Capital Exercise”. Differently from the previous stress tests, the sovereign debt exposures in the banking book were now marked to market. The total capital shortfall of 114.7bn Euros was announced on 8 December, with the deadline for banks to build the buffers by end of June 2012. By that, cutting back lending was not foreseen as an option.

EBA and ECB (almost) admit that the EU bank stress testing exercise was a failed regulatory initiative (and not the only one).

Andrea Enria, the head of the EBA, said to Spiegel on 12 December 2011 that: “The problem is that everybody seems to look at the exercise as a sort of beauty contest: Are German banks better than Italian banks or the British ones?”

Andrea Enria doesn’t add (but probably has in mind) that the criterion, the capital ratio, based on which the banks are being compared may be quite a misleading measure... (More on this: 2011 EU-Wide Bank Stress Test: Scenario May Not Be Main Issue)

The fresh president of the ECB, Mario Draghi, told during his first press conference of 2012 about the topic. “The exercise has turned out to be procyclical,” he said. “Governments should have had capital in place if there was a need for capital. And the governments didn’t have it. The EFSF should have been in place, because that would have had a positive impact on markets, and it was not. [...] I’m confident that the same exercise is not going to be repeated without the conditions I mentioned before.” (Source: Financial Times, 12 January 2012; accessible for registered readers only)

Mario Draghi doesn’t add (but seems like has in mind) that the European governments, with or without the EFSF, do not still have enough capital in place for a real stress test. He doesn’t ask (although he could ask), what the European Systemic Risk Board is doing when allowing the European Banking Authority to conduct something that is procyclical?

What the 2012 stress test should be?

More precisely: should there be a stress test at all?

Whatever the scenario, it would either be perceived as “not stressful enough” (as always), or (when including sovereign defaults and similar) result in the conclusions that no one in the system dares to spell out. In other words, a stress test would anyway not have the desired stabilising effect. So, why to waste resources?

Instead, I’d suggest publishing the start data, and a base case forecast based on the common methodology. Everyone could then add as much stress as he/she deems necessary, and no one had reason to complain about the scenario. Ok, it’s not that easy in practice. Yet it would provide clearly more transparency, good basis for more constructive discussions in the future and hopefully less market panic, if further write offs of the troubling governments’ debts become necessary.


  1. This comment has been removed by the author.

    1. Stress test is just an instrument for restoring confidence. If done in wrong time it could have the opposite effect. The best way to get the picture is to perform it by yourself. Just change the parameters and see the most vulnerable banks.

    2. Precisely: stress test is just an instrument... And at some point I started to feel (while knowing what a huge work is behind a carefully done stress test) that too many resources are being devoted for it. Indeed, everyone seems to be stress testing. Paradoxically, far too often many don’t even understand the pass/fail criterion. I fully agree that if you want to get a picture, then the best way is to do it by yourself. The only thing is that one cannot be expert in every area and one does not have time to do everything. That's why we are paying to the analysts, right?