That sounds peachy, except that there's a difference between perception and reality, and a difference between mitigating risk and the creation of systemic risk. The notion that honest risk pricing can occur wrt deals doomed to failure is an absurdity, as well. A mortgage pool stuffed with no-doc teaser rate ARM's in a highly inflated market is just such an instrument, even though Li's gaussian cupola would probably tell us differently.
http://www.wired.com/techbiz/it/magazine/17-03/wp_quant?currentPage=all
At some point or another, we need to limit risk by systematically important institutions, and the obvious answer is to first understand what risks are necessary and what risks are not in a functional system & regulatory environment. We also need to eliminate the profit potential in conflict of interest dealings. I'd offer that synthetic OTC derivatives introduce unnecessary risk and potential conflicts of interest that we would better do without. They lack an important element- insurable interest. Neither party has an insurable interest in the underlying asset, because neither party owns that asset. It's pure gambling at very, very high stakes., often with other people's money. So when that blow up threshold is crossed, as it was in the collapse of the ownership society, total risk is much higher, and the whole system blows up, which is simply unacceptable.
The reason such products exist at all is because the middlemen, the bankers, obtain enormous transactional fees in separating the two parties, and often gamble themselves in the process. It's also fed by too much hot money chasing too few returns, rather than being invested in productive enterprise. Extreme concentrations of wealth and income create that situation, and run counter to the whole idea of stability in an axiomatic sort of way. When winning big satisfies only the lust of greed & power while losing big affects one's lifestyle not at all, then enormous risks will be taken, and calculations justifying them will be found. It's the nature of greed.