When Scott Patterson wrote about CDOs as though they were derivatives this
morning, I
slapped him a little: I thought it a silly mistake. But now I see that Tony
Jackson is doing
exactly the same thing in the FT, and I’m beginning to wonder whether there’s
something I’m missing. Here’s Jackson:
Contrast the Bear Stearns case, which triggered the latest mini-crisis. The
other banks that inherited the subprime derivatives in question have held
off selling them, precisely because they risk crystallising a much lower market
price – which would then apply across the board.
The "subprime derivatives in question" are, of course, CDOs. And
most of what goes into CDOs is not subprime derivatives at all, but rather loans
or mortgage-backed securities.
Now I do appreciate that some CDOs do play in the CDS market. (All these three-letter
acronyms can be confusing: a CDO is a collateralized debt obligation, which
basically just means pool of loans. A CDS, on the other hand, is a credit default
swap – a derivative contract based on whether or not a certain credit
goes into default. I know that they share their first two letters, but the two
abbreviations don’t have a single word in common.)
But here’s the thing: insofar as the problem with CDOs is their subprime exposure,
then their CDS exposure is not an issue. The CDS market is much less developed
in the world of subprime mortgages than it is in the world of corporate and
sovereign credits. CDSs on certain mortgage-backed securities do indeed trade,
and they’re generally more liquid than the underlying securities. But as I understand
it, these particular derivatives are trading vehicles, or dynamic hedges: they’re
not generally long-term investments which get issued by CDOs and then have to
be unexpectedly valued in a liquidation scenario.
If you look at the sheer quantity of subprime-backed bond issuance, it seems
clear that there’s more than enough actual paper to go around – if CDOs
wanted exposure to subprime mortgages, they tended to simply buy those subprime
mortgages. By contrast, much of the rise in CDS issuance is a consequence of
demand for credit meeting the reality of relatively little new bond issuance.
So CDOs start writing credit protection instead – a strategy which replicates
the risks of buying bonds, without having to actually source the bonds in question.
As a consequence, there are a lot of CDOs which have a lot of exposure to the
CDS market. There are also a lot of CDOs which have a lot of exposure to the
subprime market. I just don’t think they’re the same CDOs.