Bank stress tests are nothing more than tick-the-box reviews of yesterday's mistakes

PUBLISHED : Thursday, 15 July, 2010, 12:00am
UPDATED : Thursday, 15 July, 2010, 12:00am

European officials have said banks must try to raise money themselves before seeking state support if tests by regulators reveal vulnerabilities.

'It is firstly up to the banks themselves,' Dutch Finance Minister Jan Kees de Jager said in Brussels after a meeting with euro-area counterparts ... 'They will get a certain period to refinance themselves in the market, but the countries will immediately announce that there is a certain backstop.'

Bloomberg, Brussels, July 14

Let's get the Eurocrat jargon straightened up here. 'Vulnerabilities' means yesterday's mistakes and 'tests' means something called stress tests, which are tick-the-box reviews for discovering yesterday's mistakes.

These stress tests are hugely favoured by central bankers after mistakes have been made and there is trouble in the banking system. Stress testing is meant to soothe the public with assurances that the exact same mistakes will not be made again.

But as the exact same mistakes are rarely made again anyway, I have my doubts about the efficacy of stress testing. It just constitutes another form of crystal ball gazing and that ball is just as opaque as it always is when people ask it what the future holds.

Take the experience of banks in the United States as subprime mortgage defaults began to turn into a crisis three years ago. American banks had been doing their stress tests but very few of these tests picked up that the collateralised debt obligations underpinning their balance sheets were actually junk assets.

The ratings agencies said they were investment grade and it was this assessment that went into the stress tests. Yesterday's mistakes at the time had not taken into account that ratings agencies could make such howlers in their ratings.

It won't happen again. Stress tests will now include a provision for proper testing of collateralised debt obligations but it hardly matters.

The next time that the US suffers a banking crisis the culprit will be something else and the stress tests will miss it again.

The particular form of natural selection that we call a market at work pretty much guarantees that things will be so.

Markets exploit anomalies and bad practices that stress tests have not yet identified are anomalies. If you rely on mechanistic processes such as stress testing for credit analysis you will get such anomalies.

And you cannot stop stress testing from being mechanistic, certainly not if you are trying to assess 91 different banks on a comparable basis as the Eurocrats want to do. This exercise can only be done with tick-the-box methods.

Doing it in any worthwhile fashion is where banking crosses the line from science to art. The best way is probably to have grizzled seniors around, people who have seen it all, including the consequences of bad decisions they have themselves made.

These people are employed to sit back and to watch the business, saying little, hearing much and lunching with others of their kind for business gossip every day. Occasionally they run their noses over a deal proposed to their bank and say: 'No, not this one.'

It is an invaluable service. Experience is much better than the tick box at spotting trouble. The difficulty, however, is that seniors like these are only to be found today in family banks and partnerships. You can take it for granted that this does not describe any of the 91 banks on the Eurocrat list.

There is another good way, although a savage one, of ensuring healthy balance sheets in the banking system. It is called the execution block. If a bank destroys its own balance sheet through ill-considered or irresponsible credit practices, the knife falls and that is the end of that bank. The shareholders lose all and depositors get only cents in the dollar back.

Ouch! Yes, that would hurt. The depositors in particular would scream very loudly. They have come to the view that their investment in a bank deposit is a super-investment that carries no risk at all. Perhaps it is not entirely considered so in Asia yet but it certainly is so in Europe.

In this unusual financial arrangement the state guarantees the liabilities of the bank, officially or unofficially but either way unquestionably, while the private managers of the bank still control the assets.

It is an invitation to trouble. The managers are encouraged this way to take on ever riskier business because, if the risks pay off, they win, while, if the risks fail, the state loses.

But in Europe politicians fall over themselves at present to adopt this sure-fail model. 'The countries will immediately announce that there is a certain backstop,' says the Dutch finance minister of the refinancing that some of the 91 banks may be asked to undertake.

Get yourself a new finance minister, cousins. This one is bent on sending you to the poorhouse.