Early October 2011, Dexia, a Belgian-French bank, collapsed for the second time in three years after their initial bail out in 2008, when they were rescued by the same governments that have propped the bank up again this month.

Interestingly enough, the bank passed the stress test with flying colours, and came in 12^th place out of 90 amongst the European banks stress tested in July of this year.

Dexia’s problems are largely the result of their business model, which relies heavily on short term borrowing in the capital markets to finance long term funding obligations. This is not a new issue, however. Northern Rock faced the exact same problem in 2007. From a capital adequacy perspective, both banks looked in remarkable good health a few months before their collapse.

The challenge with this funding strategy is that as soon as liquidity starts to dry up, it is no longer possible to raise the required amount of funds. General market sentiment due to the bursting of the sub-prime mortgage bubble in 2008 and uncertainty about the consequences of restructuring of Greek, Spanish and Portuguese sovereign debt on the Eurozone in 2011 made lenders cautious. This results in a significant reduction of the liquidity available in the market, and particularly impacts those that are heavily relying on the short term capital markets. I wouldn’t want to go as far as to say that this is unique to the current crisis though. Previous crises have been subject to exactly the same problem.

So, what does this say about the July 2011 stress test result?

It can certainly be assumed that the underlying model and/or its assumptions were, at a minimum, incomplete and perhaps even flawed. If this assumption is incorrect, it is difficult to explain why a bank that passes a stress test with flying colours needs to go cap-in-hand to the government not even six months later. Can it be resolved by purely adding some form of an ‘available liquidity’ indicator? Probably not since – as previous crises show – available liquidity in the market doesn’t change gradually over a period of time but instead tends to change rather abruptly. Perhaps there is some merit in also including asset and liability management strategies. Whilst also not easy to model either, the proportion of long term assets such as savings and deposit accounts or other sources of long term funds may be a significant factor in the probability of a bank being susceptible to volatility in the available liquidity in the market.

There is, however, one thing that truly puts Dexia in a class of its own. Just before it all went horribly wrong in 2008, one of Dexia’s wholly owned subsidiaries provided a loan of €1.5bn to two of its largest institutional shareholders to finance their stake in Dexia’s capital increase. Apparently that is, or at least was, not illegal in Belgium and, notwithstanding the regulator’s concerns, it did effectively raise Dexia’s capital adequacy to an acceptable level. Albeit artificially.

Dr Natalie Schoon, CFA