How to Break a CPDO

CPDOs, or Constant Proportion Debt Obligations, were briefly popular at the end of 2006 as a way of getting high returns on AAA-rated paper. At the time, many commentators (notably excluding myself) said that CPDOs were too good to be true, and that the ratings agencies who gave them a AAA rating were, in a word, wrong. Those commentators are feeling a bit smug right now, since CPDOs look very much as though they are getting close to breaking point.

CPDOs certainly seemed to be very hard to break. You could model a credit crunch, or a whole series of defaults, or massive spread widening, and the CPDOs really didn't mind all that much: when spreads go up, their value goes down, but their income goes up.

There was one other possible thing which could harm a CPDO, however, and it was a known issue right from the start. CPDOs "roll" every six months: they unwind the credit protection they wrote half a year ago, and write new protection on the current index. Because the current index is longer-dated than the old index, it normally trades wider than the old index, which gives the CPDO extra income. If the new index trades inside the old index, there can be a problem. This is what Citigroup said in a November report:

Change of composition is another downside factor that plays into the roll dynamics. By construction, “blown up” or downgraded credits are removed from the new basket and replaced with “average” spread names. As the result the new basket tends to be less risky and have less carry. For CPDOs that means that when spreads widen idiosyncratically the MTM losses on the old series may not be fully compensated for by the increased carry on the new series. For example, the GM and Ford downgrade has resulted in large MTM losses in CDX4. These names were subsequently removed from CDX5 and it was issued at a lower spread. Although clearly a factor, we found this “loss of carry” to have a relatively weak influence on CPDO final value unless happening under extreme conditions.

Well, guess what? It seems that this "loss of carry" is happening "under extreme conditions". "It will be painful," says Alea of the CPDO roll today: rather than seeing spreads on the new index about 7bp wide of spreads on the old index, the new index will actually be tighter than the old index by as much as 15bp. And that's the kind of thing which can really hurt a CPDO which is already suffering from capital losses.

The culprits are mainly companies which used to be investment-grade but which have been leveraged up in private-equity takeovers and are now rated as junk: Alliance Boots, Alltel, Boston Scientific, Expedia, and ResCap. According to a report by David Watts of CreditSights, these downgrades have happened at exactly the wrong time for CPDOs, and many of them now risk being unable to repay their principal in full at maturity. In some CPDOs, the chance of breakage might be as high as 20% or more, which is definitely not something which becomes an AAA-rated instrument.

Of course, the fact that CPDOs can break does not mean that they will break. And the good news, from the CPDOs' point of view, is that the credit crunch has made future LBO-driven downgrades much less likely, which in turn means that this roll problem is unlikely to be repeated.

In any case, it seems that the sequence of events needed to break a CPDO is now clear. First you need a lot of liquidity and very tight credit markets, which make junk-fueled LBOs attractive. Then, while the CPDO is still highly levered at the beginning of its life, you need a massive credit crunch, which sends spreads gapping outwards and which hits the asset value of the CPDO. When the roll comes – bingo, the CPDO becomes risky enough that the credit rating agencies, at this point, are going to have to embark on some multi-notch downgrades.

So was I wrong about CPDOs? In a word, yes. I did say that they could fall dramatically in asset value, and I even said that they could get downgraded. But when I said that, I was imagining a downgrade to maybe single-A, not a downgrade to junk, which now seems to be a serious possibility. It turns out that CPDOs were, in fact, more fragile than I thought. Although if anybody wants to sell me one at today's distressed prices, I'd still be a buyer.