Comments on: Toward Transparent Derivatives Trading

By: Nate Oman

Nate Oman — Mon, 18 May 2009 14:29:11 +0000

It seems to me that one of the biggest problems with counter-party risk is that in theory these risks were supposed to be disclosed in regular financial statements where the CDSs were to be carried as liabilities. The problem is that this gave to the accountants the insurmountable task of valuing liabilities in an opaque market. Just having people report their CDS contracts to the SEC in real time wouldn’t make that big of a difference. It seems to me that what we need is a thick and relatively commoditized market in these contracts, so that we can use the market to value the risk rather than kidding ourselves into believing that through some heroic act of calculation the accountants and actuaries can do it for us. At the end of the day, it seems to me that CDS contracts are very much like commodity futures. The best solution is to set up clearinghouses that deal in standard forms and start generating thick, transparent markets to value the CDSs, rather than relying on accountants or (worse yet) regulators.

By: anonprof

anonprof — Mon, 18 May 2009 10:04:01 +0000

Frank–Thanks for the Atwood quote–it may be the best one on the crisis yet.

By: Frank Pasquale

Frank Pasquale — Mon, 18 May 2009 02:04:22 +0000

I’ve thought a bit about these ideas in the context of health insurance, AJ. For example, if a doctor agrees to see all of his patients any time, is he effectively an insurer, an “entity bearing risk”? It seems very unlikely that more than a few people will get sick at once–but it also seems reckless for him not to at least make some contingency plans.

As for difficulty of calculating expectation values, it seems to me that the proposal here is designed to smoke out the worst offenders (like AIG). That is a “smell test” model, which reminds me of this reflection on “sight vs. the other senses:” “how has the metaphor of vision for knowledge colored our conceptions of knowledge”?

By: A.J. Sutter

A.J. Sutter — Mon, 18 May 2009 01:49:34 +0000

I’m a little confused as to what should be transparent. Maybe I’m thrown by Acharya & Engle’s phrase “actual risk” — what does this mean?

Is a plain vanilla insurance contract transparent about “actual risk”? Yes (kind of, aside from lawyerly gobbledygook), if that phrase means describing what insurer must pay to insured, and under what conditions precedent. But no, if that means disclosing what are the odds that insurer will have to pay.

It sounds like A&E are talking about transparency of risk in that second sense, though. And here there is a deeper problem than just trade secrecy: it’s that expectation values — which are obtained from mutiplying the magnitude of a payoff by the probability of the payoff occurring — are not reality. Transparency of the parties’ models isn’t transparency of “actual risk”. It’s simply transparency about parties’ expectations, which may be delusory.

Maybe the problem is that “transparency” and “actual risk” suggest that risks can be accurately quantified, calculated and written down. So how about a change of pardigm, away from the visual, “transparency” metaphor and more towards an olfactory one? Derivatives disclosures might not enable determination of “actual risk,” but in the aggregate they could alert regulators or the public that the whole mess stinks.