You know, a certain FT reporter took a lot of shtick for a this article.
The gist — European sovereigns were increasingly turning to the kind of pre-crisis financial engineering to shift them out of crisis. The European Financial Stability Facility, you’ll remember, was often likened in principal to a giant Collateralised Debt Obligation, with its emphasis on credit enhancement.
Anyway, the ‘CDO’ comparison was enough to annoy EFSF chief executive Klaus Regling, who picked up his pen in February to coin a letter of refutation to the FT. “The essential difference between the EFSF and a CDO is that EFSF debt has no tranche structure. There is no seniority and all investors have exactly the same rights,” he wrote. But even if the EFSF itself reckons it’s not a CDO…
… The market seems to disagree.
Last week, the eurozone agreed to increase the size of the EFSF from a lending capacity of €225bn to a full €440bn, by upping the guarantees from participating EU countries. Now there will be two sets of bonds issued from the EFSF — those from the pre upsized facility, and ones from the enlarged EFSF.
And guess how analysts are comparing the two? By using CDO valuation metholodology.
Here’s Nomura’s European rates strategist Nikan Firoozye:
We are effectively analysing a super-senior tranche of a CDO. For a given EUR100mn bond, the old structure has a pool of EUR120mn in guarantees (of which EUR75mn are AAA and EUR45mn are AA and lower ratings), and the proceeds of the bond issue are invested in EUR75mn of loans (directly corresponding to AAA guarantees) and EUR25mn of AAA collateral to ensure the rating. The total pool is then EUR220mn and the super-senior principal is protected as long as defaults remain below EUR120mn.
Correspondingly we see the new structure has EUR165mn in guarantees, of which EUR102mn are AAA (just slightly more than the AAA guarantees and AAA cash buffer in the original). What offers far greater protection against losses is the EUR63mn of AA and lower rating guarantees together with the larger loan size of EUR100mn. With a total pool of EUR260mn, the super-senior principal will remain intact unless defaults exceed EUR160mn.
We value this super-senior tranche using a simple CDO evaluator, with a gamma copula (and gamma=200% to emulate Normal copulas) in Figure 2. We can then value each structure under various default correlations. In the example we consider the guarantee pool ex-Greece/Ireland/Portugal and consider a loan to Portugal as backing the bond issue. CDO models typically are applied to lower default correlations, but we can only assume relatively high correlations across eurozone sovereigns. We note that the minimum fair-value spread between the two is about 2bp in the 5yr and 6bp in 10yr, but that this decreases to flat in 5yr and 2-3bp in the 10yr when default correlations are reduced to as low as 80%. We note that fair value, according to the CDO valuation methods is between 30-40bp tighter than current spreads, but that given liquidity considerations and the overall complexity of the structure, we do not think that this is particularly relevant to levels of EFSF in general.
The details might seem irrelevent to you, but it’s worth remembering, as Firoozye notes — that this particular CDO is now at the heart of the eurozone. It also, incidentally, probably plays a big role in that other bit of European financial engineering, the Special Purpose Vehicle in the French proposal.
Related links:
Collection of French proposal research - The Long Room
Europe’s EFSF really is not saving anything - FT Alphaville