Rather than start directly with the weekend rumour of using the European Central Bank to “leverage” the EFSF…
… instead we’ll start what the ECB has done with the current Greek bond swap (these are answers to questions from the official bond swap website):
Which is to say, nothing. No appetite for sovereign losses from Greece. Despite the ECB being the country’s largest “private” owner of its debt. There’s no tendering of the ECB’s circa €49bn of relatively short-term Greek bonds in prospect and it will not take the swap’s circa 21 per cent write-down. So, given this rather important precedent, would the ECB take any sovereign loss arising from its bond purchases?
Would the EFSF, should it find itself partially in the ECB’s bond-buying position? Would German politicians allow it?
That’s our main question arising from levering up the EFSF, via the device of the Facility providing first-loss credit protection (an equity tranche) to the ECB buying sovereign debt. Basically, we reckon it assumes that:
- The first loss from a sovereign is comparable, in magnitude or nature, to the first loss from any other asset (such as the asset-backed securities targeted in similar past programmes such as TALF, for example)
- The first loss is acceptable for the EFSF to bear, and that the ECB would also be prepared to accept second losses.
Neither assumption holds water, we believe. It’s worth going through why, not because levering the EFSF is necessarily going to be part of future eurozone plans, but because it’s a sort of parable about how loss from sovereign default is a unique kind of financial loss.
There are other ways of ECB leverage, such as the EFSF’s obtaining a bank license and directly borrowing from the central bank, but its chief executive seems clear that this probably wouldn’t be legal. So we’re left for now with the “deal” that the Telegraph reported over the weekend:
The complex deal would see the EFSF provide a loss-bearing “equity” tranche of any bail-out fund and the ECB the rest in protected “debt”. If the EFSF bore the first 20pc of any loss, the fund’s warchest would effectively be bolstered to Eu2 trillion. If the EFSF bore the first 40pc of any loss, the fund would be able to deploy Eu1 trillion.
(There are related plans, for example the Bruegel “war game”.)
You could think of this as the EFSF and ECB creating a genuine CDO. (Risk retention! Tranches! Frightening default correlation!) A bit more prosaically, it bears a family resemblance to UK or US crisis vehicles, including the UK Treasury’s Asset Protection Scheme for banks or the structure of TALF. Very similar first-loss provisions. But! No one has designed first loss for eurozone sovereign debt before.
More prosaically still, the whole thing’s a shell game designed to allow the ECB to keep on buying government bonds. Which is arguably fine. It’s the one with the massive balance sheet. If you think Italy’s whole problem is a liquidity drought, it needs the ECB there as a market-maker of last resort. (This is partly why we don’t have much faith in non-ECB-based plans to lever the EFSF, for example using it to issue discounted bills directly to the market. This is going to run out pretty quickly for saving Italian debt and retaining the Facility’s AAA rating. Similarly, the IMF has only so much ammo for serving as leveraged institution.) The EFSF’s first-loss capital buffer is a fig-leaf, in this interpretation of it all being about the ECB balance sheet.
Which doesn’t stop it being an expensive fig-leaf!
This is where the first assumption we’d like to question comes in. The size of the first-loss protection – such as the EFSF insuring the first 20 per cent of losses – doesn’t reflect the likely size of loss from sovereign defaults, or the political nature of default. Greece is shaping up to be a classic example of this already, so we’ll use that. Greek bonds are now trading as recoveries more or less, and imply that investors will recoup 30 to 40 cents in the euro (Greece CDS specialised in trading recovery rates imply 25 cents). A lot of this reflects a prediction that Greece will retain assets that it could have allotted to debt recoveries for its own purposes, for example to recapitalise its banks, and that official lenders will pressure Greece to make them whole. The point is, sovereigns determine their own recoveries. There are other factors at work here, notably the correlation of deterioration in sovereign credit, as well.
Taking the sovereign which is really in question here — Italy — it’s extremely difficult to gauge likely recoveries on the same basis, considering how Italy entering default is still relatively a doomsday prospect. On the other hand, the ECB-EFSF vehicle would be loading up on Italian debt in such size that it’s worth considering. What isn’t wise is to pretend that EFSF credit protection will be a bazooka that won’t ever need firing.
And it is very worth considering given the weakness of the second assumption above — that the EFSF or the ECB would be ready to withstand first and second losses respectively. The Greek bond swap already gives everyone a neon-lit sign that the ECB will avoid taking losses, even after acquiring exposure on a pari passu basis to private holders. The swap is voluntary but it doesn’t bode very well for the future. Pushing the EFSF into first loss also contrasts with the clear unwillingness of some guarantors to take any loss, i.e. the demands of Finland and other states for collateral on EFSF loans. Given that all sovereign default losses are engineered in some way, “semi-official” buyers of “private” government debt could defect and press for lower losses when push comes to shove.
Or maybe they wouldn’t. The point is — this doesn’t look the kind of credit protection that could fix the crisis, because this is a sovereign crisis that has broken a set of “risk-free” assets forever, not a squall in the ABS markets. It’s going to take a bit more than that…