Tuesday, October 28, 2014

A Rather Unsophisticated Analysis of the ECB Stress Tests

The results of the stress test have been released and we now know that all but 25 of the 130 or so largest banks have passed it. Certainly all the really 'big players' have passed it. Everything fine?

Yes and not really. Clearly, any detailed examination/audit of the largest banks is always a good thing. However, the markets will never trust a bank only because the ECB says that it can be trusted. There is much more to the secret of what makes up confidence.

How safe, to take just one example, is Deutsche Bank? Safe enough to pass the ECB stress test, for sure. But safe enough to withstand a major confidence crisis?

At the last reporting date, 30 Jun 14, Deutsche reported total assets on balance sheet of 1.665 BEUR (that is one-thousand-six-hundred-sixty-five billion). A bank is as good as the loans which it has on its books? Once upon a time, perhaps, but no longer today. Deutsche's loans amounted to 388 BEUR; that's only 23% of total assets.

Quite strikingly, Deutsche reported 'total financial assets at fair value' of 927 BEUR. Those were composed of 'trading assets' (211 BEUR), 'positive market values from derivatives' (485 BEUR), 'designated financial assets' (176 BEUR) and 'financial assets available for sale' (55 BEUR). Altogether, these financial assets represented more than half, 56% to be exact, of total assets of Deutsche.

What are those financial assets? Well, typically there is not a customer on the other side but a 'counterparty' instead. Typically, those counterparties are publicly rated, publicly rated by the same rating agencies which gave sub-prime paper a triple-A. And, typically, many of these assets are not traded in a public market where one could derive market prices. Instead, their value is often derived from a formula, be it Black-Scholes or whoever else. The reported value of those assets is as good as the formulae applied. When the formulae no longer work, the value of those assets can no longer be determined (as LTCM found out in 1998).

The equity ratios calculated by the ECB are not based on notional amounts but rather on imputed amounts. The equity requirement is a function of the risk-weighting of the assets (more risk requires more equity and vice versa). The point is that the risk-weighting is done by the banks themselves using very, very complex formulae. Any such system is always prone to manipulation. As a result, one must also look at the equity ratios based on the notional amounts in the financial statements.

Deutsche financed its total assets of 1.665 BEUR with equity of 68 BEUR, i. e. equity accounted for 4% of total financing. That is about twice the equity ratio which Deutsche reported only a couple of years ago; truly a significant improvement. It translates into a leverage of 24:1. Compared to the roughly 50:1 leverage which Deutsche had had, this looks very good. However, such leverage levels are still in the category of hedge funds. One can still argue that Deutsche is a hedge fund with an associated commercial bank.

The ECB examined each bank on its own merits. What was not examined, as far as I know, was the interdependency among the banks. Deutsche probably has its risk financial assets hedged with corresponding financial liabilities but on both sides they have counterparties. If a counterparty fails, Deutsche loses its risk protection.

Looking at the results overall, one gets the impression that the ECB examination had one very positive impact: it made the bankers nervous. Bankers should always be a bit nervous. Nervous about making bad loans; nervous about buying the wrong securities; nervous about getting caught at doing untoward things if they were to do untoward things. Most of the banks had taken some rather significant measures in the last year to improve their reported financial strength. Some banks have even announced to change and improve conduct and culture. Whether those are only temporary phenomena owing the the ECB examination or whether longer-term change will indeed result remains to be seen.


  1. Very nice post.

    When Greece had "credit event" (cds) many were saying that net payout was only 4 bil $, the probability of technical default at 90% possible and that cds were calculated in trading book and recorder marked to market.
    But if the formulae becomes null in an "unexpected event" for a company then what? Watching notonal ratios is indicative Klaus as you say.
    From this view also i understand why the Swiss are thinking about gold, with pros& cons.

    It seems fake construction all these.


    Why ECB or EBA dosn't make a reference for banks (whetever bank name is in europe at least) that only those are deposits loans in balance sheet and all the other are more cloudy in order people to know more?
    Simply, this is a hedge fund and bank this is a bank.

  2. ECB's vicious cycle mechanism