Defragging the CDS Market

You know when you've had a computer for a long time, and the hard drive is full of crap, and some tech wizard comes up with a way of defragging it or something, and the idea is that suddenly it'll be much leaner and cleaner and faster? Well, the CDS market is a bit like that hard drive, it would seem, and two firms in the CDS arena, Markit and Creditex, are the tech wizards who have a bright plan to come in and clean it up.

The new compression approach improves on previous tear-up processes by delivering significantly better compression results while leaving market risk profiles unchanged. The process involves terminating existing trades and replacing them with a far fewer number of new "replacement trades" which have the same risk profile and cash flows as the initial portfolio, but with less capital exposure.

The idea is to reduce the nominal amount of CDS outstanding - currently $62 trillion or something equally impossible to grasp - by tearing up trades which offset one another and generally reducing the market to the smallest possible number of contracts on any given name. Counterparty risk won't be reduced a lot, but documentation risk will be. Consider for instance an anecdote in today's FT:

In one case the seller of credit protection recently discovered that the final agreement on insuring a portfolio of collateralised debt obligations had never been signed, either by it or a French bank which in this case was buying protection. Now, with the meltdown in that market, the seller has returned all the premium payments to the buyer and torn up the agreement, saying that because it was never signed, it has no legal obligation to pay up.

I wish that Henny Sender had more details here on both the buyer and the seller of protection; if I were the French bank in question, I wouldn't have too much compunction about outing the culprit. But this kind of thing is clearly a risk to the CDS market, and if hastily-drawn-up contracts get torn up as part of the new scheme, that would be an excellent outcome.

The question of course is whether this is all too little too late. The tear-up process won't even start until Q3, and even then will only be active on the minority of CDS contracts which are written on single names. The really hairy ones - the ones written on baskets and CDOs and other such creatures - won't be touched. That's a careful and responsible way of doing things. But it might need to be accelerated if things start getting really messy.