mathbabe

What’s the current deal on credit default swap data? Is the Dodd-Frank bill going to force any CDS pricing to be publicly available?

A bit of background: a credit default swap is something like insurance you pay in case the underlying bond is defaulted on (but not exactly, see here), so it’s relatively easy to infer the default probability from its price, as long as you have a good estimate of the “recovery rate,” which is the amount the bond pays out even though it’s defaulted. This rate can vary widely, and people sometimes lose sight of how sensitive everything is to that assumed number.

Here’s the thing. I am super into the idea of an open source ratings model (see this post and this post on open source ratings models, as well as this post on open models in general), and I think having CDS data as input to the model might vastly improve it over just using quarterly filings and stock market data.

Right now the standard ratings models don’t use CDS data, but I think that’s because they’re just really old. I’d guess that some combination of the old ratings model and the new CDS market would be great for an open source ratings model. And it’s true that CDS coverage isn’t perfect (i.e. there are not liquid CDS markets on everything you’d want ratings for) but on the other hand, for what it does, the market is super timely and people really watch it (sovereign debt is a great example of this).

As of a year ago all of this data was essentially owned and monopolized by Markit, which is made up of a bunch of CDS brokers. So even if I had the money to pay for the data, for licensing reasons I wouldn’t be able to make the data open source, which sucks. I know that there’s been talk about making this data publicly available, but I’ve been so involved with stuff like the Volcker Rule, I just haven’t kept up with the current CDS transparency rules. I mean, if we aren’t going to remove the CDS market or regulate it, at the very least we should be using it. Please tell me if you know.