Remember Barclays’ monoline exposure?
Here’s an update, in a very indirect way, courtesy of Standard & Poor’s.
The below chart is the rating agency’s repossession rate forecasts for the UK, published on Monday.
The numbers are something of a problem for UK banks — and as, we’ll see, for some of the monolines which insure their mortgages.
For instance, S&P gives the example that if repossessions were to total 2 per cent over three years, under the agency’s worst-case scenario, and losses are about 1 per cent, the implied loss given default rate, a measure of the magnitude of likely loss under Basel II banking regulation, would be about 50 per cent.
That’s close to the LGD rate Barclays cited in relation to monolines in its last interim management statement. The bank uses the bond insurers to hedge its credit exposures on structured things like RMBS, CDOs and CLOs, racking up some $27bn of protection by March 2009, and taking hedge gains accordingly.
In this it is not alone. Other banks like RBS and Deutsche, also insure their products via the monolines.
However, Barclays’ monoline exposure proved to be something of an idiosynchratic concern to some analysts earlier this year, since monolines were being junked left and right and were generally teetering on the edge of oblivion. The bank itself maintained that the likelihood of being hit by both a default on the underlying asset and on the monoline itself was very low, thus it could keep accounting charges relatively low on its monoline exposure while simultaneously reaping-in hedge gains.
And here’s where S&P’s commentary gets interesting:
How could the LGD be so large for a well-secured activity? During the same period, property prices fell sharply, by about 15% across the U.K., but much more in some areas, for example, London and the Southeast (over 30%). It is likely that losses were concentrated in this region.
Moreover, high loan-to-value (LTV) lending was common in the late 1980s, with the median LTV ratio above 85% for much of the late 1980s and remaining remarkably high into the 1990s. LTV ratios were especially high for first-time buyers, who at the time received around half of all mortgage advances. This compounded the house price shock. Thus, an average LGD of 50% is conceivable to us, given the combination of 90% LTV at origination, a 25% fall in property values, a 15%-20% forced sale discount, and 5% costs, plus arrears roll-up, which could be as high as 15% given the mortgage rates at the time.
We think there would, however, have been some offset from mortgage insurance, which effectively passed some of the losses to the insurance industry. In addition, tax benefits on mortgages and other government support likely helped reduce losses.
Of course — the point is that while monolines might be absorbing some of those losses, there are banks, like Barclays, which have exposure on the loss-absorbing monolines. That suggests that Barclays’ joint probability defense is not really washing. If banks are losing out on non-performing loans and some of those losses are being shifted to monolines, some of that default risk should eventually arrive straight back at a bank like Barclays via its monoline exposure.*
But then, we suspected that already.
* Caveat accounting rules. There was some suggestion earlier this year that the relaxation of mark-to-market accounting rules by FASB and perhaps the IASB could allow banks to reconsider how they treat their monoline exposure.
Related links:
Barclays monoline risk - FT Alphaville
The on-going structured finance mess at Barclays – FT Alphaville