(*cough*)
>> They are also zero sum, BUT they do not take place within an
>> exchange. So they are not regulated and they are not transparent.
OTC derivatives are useful and important: unregulated transactions that are illiquid and irregular are an important part of the landscape. But: this particular market (credit default swaps, and most especially the synthetics) should never have been allowed to grow as big as it did without oversight and regulation.
I propose a simple rule: when a particular OTC instrument becomes "big enough" (TBD; too small and you risk stifling innovation and decreasing efficiency; too high and you have what we have today), it must be submitted to an exchange. The principal reasons for OTC contracts -- illiquid and irregular -- don't apply when you grow a big market. If CDSs had become listed -- and thus regularized, transparent, and position-limited -- in 2005 or so, we'd be in a very different world today.
Here's a paper from 2006 that suggests a way to do it; it's not like people weren't thinking about how to do it, it's just that no one was forcing the issue.
http://www.kellogg.northwestern.edu/research/fimrc/papers/jakola.pdf
---
Doug writes:
> But it's gone now, and the obligations remain.
Is it actually gone? We've had a lot of paper write-downs due to mark-to-market reporting requirements, but we haven't actually had that many defaults; and the big one that every feared -- Lehman -- had the settlement go rather smoothly. The corresponding decline in trust surrounding this unwinding that has led to massive liquidations has absolutely trashed nearly all of the markets; but was the CDS madness anything more than a trigger?
We have kind of a weird cycle ... all these contracts were sold, and someone asks: what would happen if they all defaulted and the insurance was priced too low? Holy crap, we'd all be bust! And so everyone rushed to correctly state their exposure, which triggered margin calls, which caused demand for liquidity, which caused the markets to drop, which cased further restatements and further margin calls. But in the end, only a few institutions have defaulted ... just as the models predicted.
That's what I don't get about articles like the above: they bring up the idea of "what it would cost to pay off all the mortgages?" as if that's the problem. It's not. Which of course is why people are tempted to ask the question. But it's not a real question.
/jordan