Without it, the CDS contract should be declared invalid because it does not address the 3rd party. The 3rd party being John Q, who ends up eating it.

But, I like the idea of “Only hedgers can initiate transactions.” With that kind of policy, I’m on board with the insurable interest plan. You make a good point that it will limit derivatives markets to a size that reflects the markets in underlying assets.

Once the outstanding MTM was settled the existing contracts were torn up and replaced with new contracts issued at the price that was used as the valuation mark in the MTM settlement.

What *I* don’t understand is why AIG seemingly wasn’t required to put up collateral or subjected to similar marking-to-market requirements by Goldman. Or if they were, why these failed to alleviate the situation.

As a sidenote, from memory I think Goldman had the laxest credit rating limit on what they regarded as acceptable collateral of any of the houses we dealt with. Goldman should have been allowed to fail.

If the CDS markets had been run like casinos, we probably wouldn’t have had the meltdown of AIG etc.

As they were effectively pretending to be the house, AIG would have needed significant cash on hand as collateral for their CDS’s. Instead, they ended up playing the role of the underfunded patsy trying to win by playing all of the roulette wheels at once. As a result, the taxpayer was asked to step in and bail them out as well as all of the other gamblers on the floor instead of having the security guard detain them.

In the end, all we need to do is ask the markets to become at least as well regulated as a Nevada or New Jersey casino.

I just wrote a 40 page paper saying exactly the same thing. Shoot me an email if you would like to read it. So, there’s my bias disclosed.

One thing I have noticed time and time again for the proponents of naked CDS is that they never discuss what happens when the counterparty CANNOT pay. Aaron Unterman, a law professor, pointed out that this credit crisis was really a counterparty crisis, i.e., everything is hunky-dory until A.I.G. couldn’t pay up. If the counterparty cannot pay, netting does not eliminate CDS risk. If all naked CDS are private contracts, even if counter parties net on one contract, neither one knows what contracts the other holds with other counterparties. And since naked CDS are freely assignable, not even the counterparties know who they are transacting with half the time. And no one is required to keep enough capital reserves to pay these contracts. Hedge funds fold; banks go to the government with their hats in hand. If you want to gamble, fine; just don’t come to the government demanding taxpayer restitution for your crummy bets.

Not even bankruptcy will protect a counterparty – naked CDS are under a safe harbor provision, allowing a counterparty to loot the bankrupt counterparty. If you want to gamble, fine, but don’t come whining to the court system to enforce your crummy bets.

Less speculation would be a great thing. Zero speculation would actually hurt the markets

Second, I am not sure your suggestion would be practical to implement. I spent many years working at a sell side shops — and all of them co-mingled their proprietary trading (speculating) with their market making (hedging). All the trades cleared on the exchanged in a single firm account (or maybe by product). There was no way for the in-house accountants, much less the exchange or regulators, to know what trades were hedges and which were speculation.

Further, any large institution is going to have significant netting across traders and across desks. And in quite a few instances, there was netting between two books managed by the same trader (hedging different risks and/or speculating on different themes).

Lastly, your idea could increase costs for hedging in many cases. If your insurance company needs to do a rate lock (just as example) on $1 billion — its usually cheaper to do one big lock than 10 locks of $100 million. Right now, the sell side firm nets your lock against all the other risk on the desk, and hedges only the residual risk. If you separate hedging and speculating — then the entire risk would have to be hedged separately (probably at a higher cost to reflect higher liquidity risk)