Mortgages unlikely to be black hole for banks; SMEs more likely
It is one of the most expensive fishing expeditions ever undertaken. Armed with costly new nets Matthew Elderfield and Patrick Honohan (and their staff) are to once again trawl through the loan books of the banks hoping to land something, anything.
While their most recent trawl took place just a few months ago, this time the IMF are waiting in the wings, insisting this trip through the muddy swamp of Irish banking is "thorough'' and "credible''.
It would appear that unless Mssrs Elderfield and Honohan find something new and alarming, the exercise will be considered neither credible nor thorough. To find the correct starting point for the search is the first task.
In some respects that is obvious. Go to the area where the markets are most worried about and start digging. The digging takes place in the strange context where the Irish banks are to get immediate capital of €10bn, partly from the rest of us, to fill a hole nobody has actually identified.
The Irish banks have over €100bn of mortgage loans, with AIB sitting on €27bn of mortgages, Irish Life & Permanent sitting on €30bn and Bank of Ireland sitting on €28bn, with the rest spread between EBS and Irish Nationwide.
What the latest stress test and loan assessment needs to tell investors and the IMF/EU team is how much of this portfolio overall will be lost and who will be taking the losses. To make such a fine judgment call will require far more information than is currently in the public domain.
For example, 70pc of this €100bn mortgage book (€70bn) was lent out in the final years of the boom, where credit standards were at their most lax and house prices at their highest.
In that context the outside world needs to know (or would like to know) who has the biggest exposure to the mortgages originated in the years between 2004 and 2007 for instance?
It is this €70bn of late stage mortgage lending that is likely to contain the problem, if there is one, for the Irish banks.
Two thirds of this €70bn of pre-crash borrowing was given to borrowers putting 30pc or less of their own money into the deal, translating into €47bn of mortgages.
Taking this group out separately, Goldman Sachs estimate one third of these borrowers took out 100pc mortgages, meaning that €16bn worth of mortgage borrowers put not a cent into their house purchase.
This on its own doesn't mean this group are going to default, but it does reduce their chances to take out other loans on the strength of the equity in their home. It also exposes them to negative equity at an earlier stage and most crucially increases the scale of the exposure of the Irish banks to declines in house prices.
Unfortunately, nobody is able to model how much of these rather ugly looking mortgage books will ultimately default. Models from the UK property crash in the early 1990s are hardly illuminating or relevant to the scale of the Irish crash.
There are two considerations here. One is how many mortgages default, and what will it cost the bank when they do default.
On the first question the worst case scenario is for a default rate of 8pc, or 10pc for the very bearish analysts. (apologies here to Morgan Kelly, the UCD economist, who envisages something he describes as "mass defaults'' -- what this means precisely he has not made clear).
Crudely put, this produces losses of €10bn at worst, and then you have to ask what is the actual loss rate going to be -- the bank can still pursue the borrower (mortgages are fully recourse in Ireland) and they can also take the security, albeit at depressed prices.
Analysts are looking at actual losses (known as loss-given-default) of about 50pc (the legal costs of repossession and processing reduce the recovered amount), which leaves the Irish banks sitting on losses of about €5bn, a large figure and about 3.2pc of GDP.
But equally it is hardly catastrophic, and the figure gets further beaten down by the fact that the losses don't have to be taken into the profit and loss account in one year, and some provisions have already been taken against them by the banks.
SME loans, on the other hand, are of more concern. Already, say accountants Mazars, one third of these loans are overdue, a hugely higher rate than mortgages. That firm estimates that there are €32bn of SME loans out there, while Bloxham estimates SME and corporate lending together amounts to €92bn.
The reason SME loans are a bigger danger is they are often unsecured or badly secured. When an SME loan gets into trouble, the bank is left picking up the pieces by taking stock, a lease of some kind and not much else.
It is this loan book that is likely to contain the horrors the IMF and EU are worried about. If they do find a black hole, it might be helpful to tell the rest of us, the providers of the capital, how big it actually is.