We're the ones who will pay if stress tests are ultimately wrong
IN a statement he may come to regret, former European parliament president Pat Cox yesterday grandly descri- bed stress tests on Irish banks as "the most important judgment call on numbers that has ever been made in the history of the State".
That kind of grandiose claim is part of a curious assumption in Europe that the results will end all debate and conjecture about the health of Ireland's banking system. They won't.
As we have seen from previous stress tests on the Irish banks, modelling for loan losses in financial institutions is a notoriously inexact science. Virtually no regulator who carries out such exercises ever gets them exactly right.
It's not just regulators who find their models don't work as they expected -- bank executives themselves are rarely good judges of how their own balance sheets will perform.
For instance, in 2008 Anglo Irish Bank thought a puny €500m would be sufficient to get it through the first three years of the property crash.
And it's not just Irish banks and regulators that get it wrong.
Regulators in Europe produced stress test figures last year that hugely lacked authority. On that occasion, the Committee of European Banking Supervisors certainly didn't find that AIB and Bank of Ireland were critically under-capitalised -- but now we are told they are.
Considering how uncertain the science is, it is puzzling that the Europeans are placing their trust in this piece of work.
Even Financial Regulator Matthew Elderfield and Central Bank Governor Patrick Honohan have made it clear how uncertain the exercise actually is: "It is not an economic forecast: it employs hypothetical scenarios,'' they pointed out a few weeks ago.
We are told this will finally "draw a line in the sand'' on the Irish banks and their catastrophic loan losses.
This is supposed to give the EU Commission, the ECB and the IMF the kind of certainty they want to potentially tweak the €85bn rescue programme and clean up the mess that is the Irish banking system.
But, of course, a stress test is only as good as the economic assumptions underpinning it. And those types of assumptions -- such as house prices, inflation, GDP growth, domestic demand and exports -- are notoriously hard to predict. Property prices, even when they stabilise, are arguably the hardest of all indicators to anticipate.
Economists tend to get it wrong when house prices are in decline and they tend to get it wrong again when prices are recovering. What economists crucially miss are what are called "turning points", those tipping points in a market when the curve starts to move upward permanently.
We have no idea when the turning point will come for the Irish property market and, as a result, the stress test assumptions being used are inherently unstable. In that context, they may not draw this mythical line everyone, including Europe, is hoping for.
TO get an idea of just how inexact stress testing is, consider this: Under the "baseline" or optimistic scenarios being used by Mr Elderfield and Mr Honohan, the economy will grow this year by 0.9pc; under the adverse or pessimistic scenario, it will shrink by 1.6pc. This is a significant margin of error and highlights how little economists really know about our economic prospects.
If Ireland ends up on the wrong side of this margin of error, it could be very expensive. If the stress test assumptions are too benign, the banks will need more money in the future yet again and the entire system may become completely nationalised. If the tests are too severe, we could end up with over-capitalised banks.
In other words, public money unnecessarily channelled into the banks could have been deployed more productively elsewhere, or not spent at all. Over-capitalisation would also markedly increase the scale of the national debt.
While the European authorities and the IMF are increasingly driving the stress test process from a distance, it is Irish taxpayers who will pay the price if they are badly wrong.