However my concern is that financial markets tend to concentrate risks with those who are most willing to take risks. More sophisticated instruments allow those risk-takers to get farther out on a limb at far higher leverage. And yes, I know the requirement for any company who is issuing massive derivatives to maintain quality credit ratings. But the move that I am talking about can happen in ways that traditional quality ratings find hard to measure, and the shift to a very risky exposure is very hard to monitor until things blow up.

The result is that risks are borne by people who can't handle them when they blow up. So parts of the system that thought they were guaranteed to not take the risk, get hit. Or (like Long-Term Capital) the amounts at stake become so large that they simple Are Not Allowed To Fail, and the government steps in. (And, of course, when risk-takers take risks on the bet that government will rescue their sorry asses, bigger and better risks get cheerfully taken.)

This is, of course, not even taking into account the opportunities for outright fraud that you get when scammers (to use the cliched example, eg at Enron) use the complexity you can get with custom derivative instruments to keep people from understanding the house of cards they are erecting.

So yes, if derivatives were used properly, I understand all of the theory for why they should be good things. However I still am not convinced that on balance the more obscure ones are a good thing.

Cheers,
Ben