The CDO Put

In a recent post, I off-handedly mentioned the "CDO Put." Basically, the CDO put, as I'm using it, refers to the fact that wider credit spreads result in making CDO creation easier. Thus a minor widening event will be met with increased CDO issuance, thus creating a back-stop to spreads.

Here's how it works. Let's say that the spread on BB-rated bank loans is LIBOR +250bps. So if I have a $100 million portfolio of these loans, I'm earning LIBOR +$2.5 million every year.

Now let's say that I can build a CDO of these securities with the following traches:

65% AAA (LIBOR +25)
8% AA (LIBOR +50)
8% A (LIBOR +80)
9% BBB (LIBOR +250)
10% Equity

That comes out to annual interest cost of LIBOR +$490,000. So we should have over $2 million left over to cover admin costs, cover defaults, and pay equity holders.

What happens if credit spreads widen? Well, we'd logically assume that the spread between AAA and BBB (or BB) bonds would widen. So let's say that the BB loan market is now +350bps. Let's say that the debt tranches on a new CDO deal widen as well:

65% AAA (LIBOR +35)
8% AA (LIBOR +80)
8% A (LIBOR +120)
9% BBB (LIBOR +350)
10% Equity

So now the annual interest cost is $700,000, a 43% increase. But now interest collected is $3.5 million. Now we have $2.2 million left over for costs, losses, and equity holders. So the net-net is that the CDO deal is more profitable as spreads wider.

If spreads in the loan market track spreads for similar-rated tranches in the CDO market, its a mathematical certainty that wider spreads will improve CDO economics, all else held equal. Thats because the CDO will always be issuing higher-rated debt than the collateral its buying. The basic arbitrage of a CDO is the spread between the lower-rated collateral and the higher-rated debt. Put another way, its the arbitrage of owning a portfolio of weak credits which collectively can garner a higher rating.

The truth is that CDO spreads are wider than generic bond spreads. I mean, a generic FRN with say AA rating and 8 years to maturity (a typical average life for a CDO tranche) would be in the 15-20bps range. But AA CDO tranches are much wider, usually in the 40-50bps area. AAA (and even AA) CDS spreads are usually less than 10bps. But the AAA tranche of CDOs of CDS is far wider, the most recent I saw was around 40bps.

So if investors lose confidence in CDO technology, spreads could widen in CDO tranches beyond where generic collateral spreads are going, destroying the CDO put. So far, the CDO market has held up beautifully in the face of the meltdown in sub-prime. CDO investors should be very encouraged. There shouldn't be any reason why CLOs should widen just because too many mortgage brokers were too aggressive with their credit standards.

Another thing to bear in mind is that the soundness of a CDO is not dependant in any way on the movement in credit spreads. By this I mean, if you buy a CDO today and spread for the collateral widen, and defaults don't actually increase, the cash flow of the CDO is unchanged. If you imagine that credit events, like LBOs, will cause spreads to widen, but actual defaults will remain relatively low, CDOs are probably better investments than traditional bonds.