In other words, it would be a good thing if risk were concentrated in a few, identifiable, "single points of failure" because then the failures could be contained.

That's the theory. But the risks would not be "contained".

What you get is the likes of Shell, BP, and Gazprom plus hedge funds - not to mention the investment banks' proprietary teams who ride on the backs of the others - all outside the box.

And if you look at the capitalisation of the typical clearing house it's not even a pimple on the arse of the risks they run.

A clearing house is an almost precise analogy of a credit institution. They keep a cushion of capital (equity plus margin) to cover defaults.

The trouble is that I would bet my bottom dollar that their risk management is inadequate in addressing the "black swans" eg a London Tin Crisis, or a Metallgesellschaft (where NYMEX was dead lucky that the Germans bailed them out) and IMHO the bigger they get, the bigger the disaster when it comes: which, like an earthquake in an earthquake zone, it inevitably will.

Simply unitising risks into "n'ths" allows it to be pooled among and backed by all of the end users, with default pools held by a "Custodian".

The risk continues to be managed by the same people as service providers who are currently doing so as intermediaries.

"The future is already here -- it's just not very evenly distributed" William Gibson