ECONOMICS:An attempt to quantify banking weakness in Ireland aims to provide a capital buffer to withstand the future
BANKS HOLD capital to guard against uncertainty and to reassure depositors and creditors that unexpected losses will not impair their ability to lend or to protect savings placed with them.
Regulators subject banks to “unlikely but plausible” scenarios designed to determine whether an institution has enough wealth – capital – to weather the impact of adverse developments. Stress tests are meant to find weak spots early and to guide preventive actions by bank directors and regulators.
The current Irish stress tests are somewhat different. They seek to quantify weaknesses already apparent in the banks, capitalise them to cover this and then add another layer of surplus capital to cover unforeseen eventualities. Changes in asset prices and default rates are often linked to a negative economic scenario in which, among other things, unemployment climbs and economic growth plummets. The objective is to provide a capital buffer that is robust to a range of economic outcomes.
Stress tests do not have a happy history. The International Monetary Fund has used them extensively over the past decade. The Irish Central Bank also used them, employing both a top-down and a bottom-up approach, neither of which worked.
The stress tests did not avoid problems with Depfa, Hypo, Northern Rock, Lehman, Anglo Irish, Irish Nationwide, etc, but this has not curtailed their use. On the contrary, they have become more popular. Ireland is on its second or third prudential capital assessment review (Pcar), the rather grandiose title now used instead of “stress test”.
Recapitalising the banks may be even more important than arguing over a reduction in interest rates that will save at most €450 million a year for a 7½-year period.
The debt service bill reflects two factors – the amount of debt and the rate of interest on it. At this stage, the interest rate may be the lesser problem. For example, a €10 billion additional recapitalisation requirement as a result of an unduly severe test would increase the annual debt service bill by more than the potential savings from a one percentage point reduction in the interest rate on bailout funds. It would also add 6 per cent to the debt burden, another nail in the coffin of solvency.
Not all stress tests failed. In May 2009, the US tested its top 19 banks. Ten were judged to need additional capital totalling $75 billion which was quickly provided – market confidence was restored and the tests were deemed a major success.
The EU experience is quite different. Of the 90 banks tested in July 2010, only seven failed and the additional capital required was €3.5 billion, well below the €30 to €80 billion the market expected.
If the EU tests were dead in the water from the outset, the Irish experience put the tin hat on it. Both AIB and Bank of Ireland passed the tests in July but were found to need additional capital just two months later. This shocking outcome became the defining feature of the 2010 EU stress tests and provides the background against which the 2011 tests will be judged.
Quite why this happened is unclear as the deteriorating National Asset Management Agency (Nama) haircuts must have been known to the Central Bank by July when the EU tests were held.
In terms of deviations from a central or baseline scenario, under the latest test equity prices are assumed to be 15 per cent lower, the dollar to be 11 per cent weaker, interest rates to be 1.25 percentage points higher and bond yields 0.75 of a percentage point higher.
The GDP shock is a four-percentage point deviation from baseline compared to three points last year when the economic outlook was less favourable. Lenders will also assume a loss of 19.8 per cent on Portuguese, 19.1 per cent on Irish, 17.1 per cent on Greek and 14.6 per cent on Spanish bonds held in their trading books.
The penalty for procrastination and insufficient disclosure is that the tests now have to be stricter. The 2010 tests were intended to be internal but were published in a frantic attempt to convince markets Greece, Ireland and others were not about to collapse.
Big banks are expected to use their own internal models – this is generally believed to confer an advantage. As far as I know, none of the covered Irish banks has its own model.
Not all EU countries are happy to publish the stress test results. Germany, which has significant bank problems, has long argued against this, with the full support of outgoing Bundesbank president Axel Weber. Germany has also led opposition to the use of a stricter definition of capital.
The result is that, at this late stage and with the Irish tests nearly complete, we still do not know what definition the EU will use and what the threshold will be.
Moreover, Weber was recently quoted as saying that the tests will not change anything, the implication being that Germany will get its way and that, irrespective of the severity of the macro assumptions, the required ratios will be such that the German banks will fare no worse than last year.
Contrast this with the Irish approach which sets out to be conservative, according to the Central Bank. Last summer, EU banks were required to have a 6 per cent Tier 1 capital ratio versus a stricter 8 per cent core Tier 1 here. By November, with the IMF in town, the Irish target ratio had been upped to 12 per cent, a 50 per cent increase, and this is likely to be the starting point for the current Pcar. The EU-wide target ratio was to be announced at end-February but is still awaited.
There is now a real possibility that Irish banks will be required to maintain a 12 per cent ratio with a strict definition of capital while the rest of the EU is tested against a 7 or 8 per cent ratio and a laxer capital definition.
Clearly, robust action will be required to restore Irish banks to normality. The question is this: is now the time to overcapitalise them when most of the banking system is in public ownership and the State unable to afford the cost that this would give rise to?
