European banking stress tests — pour encourager les autres?

“In this country, it is customary to kill an admiral from time to time, to encourage the others”, as Voltaire wrote, satirising the execution of Admiral John Byng for a naval defeat that wasn’t really his fault . Similarly, with the results of the European Central Bank (ECB)’s stress tests coming up (technically, the Comprehensive Assessment of the ECB, jointly publishing its own Asset Quality Review (AQR), and the stress test carried out by the European Banking Authority (EBA — sorry about all these acronyms by the way), but I don’t think anyone’s unclear about who is in the driving seat here), it’s worth spending a couple of minutes thinking about the question — how bad do a bank’s stress test results need to be before we get into CEO-resignation territory?

As with the apocryphal Royal Navy, the purpose of such an exercise is not really much to do with justice, fairness, or any real systematic process of rewarding success and penalising failure. It’s more that we want a good display of cadavers swinging from yard-arms, just to concentrate the minds of the survivors and reinforce the notion that running your bank too aggressively in terms of its capital requirements can be a personally wealth-destroying decision. It matters, because the Euroland stress tests are going to be published in a format which appears to have been driven more by a desire for comprehensiveness rather than immediate legibility, and it will be surprisingly difficult to tell, at first glance, whether a bank has “really” passed or failed. So here’s a quick guide on “how to read the stress test results”.

As far as I can tell from the template that the ECB has given us, there are two quick and easy rules you can follow to save time — and time will be of the essence, since the results are coming out at noon CET on Sunday 26 October, giving about 20 hours until the market opens to get your head round 12,000 data points. First, concentrate your attention on the ECB disclosures at their website— the EBA will be publishing detailed results of their own stress tests, but the main conclusions are going to be reproduced on the ECB site. And second, read the back page first.

The reason I say “read the back page first” is that the ECB and EBA had a bit of a touchy problem of presentation to address, going into these stress tests. On the one hand, they want to establish a bit of credibility for themselves. The ECB is brand new to the business of bank supervision, while the EBA has a five year track record now (including that of its predecessor body, the Conference of European Bank Supervisors), one which basically consists of institutional credibility painstakingly built up through grinding out competent nuts-and-bolts regulatory work, then squandered in badly constructed and over-optimistic stress test exercises. And credibility, in the context of banking stress tests, is determined not by the details of the assumptions — there are a zillion and one ways of fudging those — but by the list of failures, in terms of both its length and the importance and size of the institutions on it. So from this point of view, the people carrying out the stress tests want a lot of failures, and they want it to include some high profile household names, ideally including some French and/or German banks to show that the whole exercise isn’t just a stitch-up aimed at bullying peripheral Euroland states.

On the other hand, though, you don’t necessarily want to blow the whole system up and call it credibility. It’s been decisively established that any bank which fails the stress test is going to be classified as a “failing institution” and have to recapitalise itself. Lots of European banks can’t do this in the market, either because they don’t have a stock market quote, or because you just can’t get a rights issue away quickly enough. So the burden falls on the respective governments. Which raises two problems — first, it’s not obvious that the peripheral states themselves have enough money to recap a load of banks without affecting their own credit, and second, the European Commission’s competition authorities have made it clear that they are more or less finished with signing off on the state aid approval for bank bailouts, unless it can be demonstrated that private sector creditors have taken a big hit. And the “private sector creditors” of a lot of European banks include a) other banks, and b) retail investors, who in many cases have been quite extravagantly mis-sold bank bonds on the basis of implicit promises of state support. Having rescued the Euro from a self-fulfilling political and financial panic in 2012, I suspect that the ECB is not really enthusiastic about triggering another one. This, obviously, would tend to militate in favour of a short list of fails, comprising only relatively marginal institutions that could be wound up without too much trouble.

How do you square this circle? Basically, by doing the test two ways, and being a little bit smart about how you publish the results. On the front cover of the results for each bank as published by the ECB, we will get the results of the Asset Quality Review. This test was carried out on the basis of the year-end 2013 balance sheet. Which means that, although it is a somewhat more lenient test (it only amounts to a re-audit of the recognition policies for bad assets, and doesn’t contain a forward-looking element), it doesn’t give the banks any credit for the capital that they’ve raised in 2014. Which is a lot of money — EUR30bn has been raised simply from shareholders, with about as much again coming from balance sheet reduction and retained earnings. So my guess is that the results on the front of the book will generate quite a large and exciting list of failures.

At the back of its publication, the ECB is going to put a summary of the results of the EBA stress test (more detail will be on the EBA website, but in my experience the additional detail doesn’t really add much in the near term, although it’s useful to have on file). The EBA will take into account the results of the ECB’s re-audit, and then apply them to a set of “economic scenarios” and a whole lot of black box analysis, in principle giving a more rigorous set of tests. However, the pass/fail mark is, naturally, lower in the stress scenario than in the ECB’s base case, and the stress test will give credit for quite a lot of capital measures carried out year-to-date. So it will turn out, I think, that a lot of banks will fail on the front cover, but pass at the back of the book — this would happen, for example, if a bank was made aware early in the year that it was at risk, and decided to do something about it.

I think I can see the thinking behind this way of presenting the results. The Euroland supervisors are hoping that the headline news will be made by the front pages of the reports, so they will be able to have it both ways — a big headline in the Financial Times and the Wall Street Journal saying that their test was credible because it failed so many big names, while at the same time tipping the wink to market analysts that most of the “failures” were not really failures at all, and that nearly all of the required recapitalisations have already happened. To be honest, I find this communication strategy rather clever.

Which raises the question though — which CEOs get the bullet? In my view, anyone running a bank which fails at the back of the book (ie, the EBA stress test) ought to go. A bank that has not been able to get itself sufficiently under control to pass a test which has been going on for a year, and not taken corrective action, hasn’t been managed properly. For the “front of book” failures, I’m less sure, and it will depend. If a bank has failed due to massive variance between the books they were reporting to the market, and the books that they had to report after the supervisor took a close look at them — well, I don’t think shareholders are likely to be particularly forgiving of that, and the only way I could see a CEO keeping his job in that situation would be if he had the presence of mind to blame his CFO, quickly (this is occasionally known as the “Deputy Heads Must Roll” principle). If a bank failed the AQR simply through not having enough capital, though, and if they had taken measures in the subsequent year to get those capital levels up to the right amount, then I personally would be tempted to say that this was fair enough; I doubt many chief executives would have the brass neck to claim a bonus for doing so, but it’s not in itself evidence of a mis-managed bank.

But I would not quibble unduly with someone — a regulator, a group of shareholders, even “the weight of public opinion” — who wanted to take a harsher approach and insist that any bank which appeared anywhere on the list of failures had to have a top level resignation. Because after all, we’re not really talking about a normal performance measurement situation here. We’re talking about an exercise which is vitally needed to build back the credibility of the Euroland banking system, which in turn is a vital component of any wider macro recovery. And for a purpose as important as that, it doesn’t seem totally unreasonable to me that some top executives might have to be asked to make the same sort of sacrifice as Admiral Byng.