"Everyone thought spreading shitty risks across MBS, CDOs, CDS, CDS^2 mitigated the risk, but it turns out i did not." I must quibble. Yes it is true that eve

"Everyone thought spreading shitty risks across MBS, CDOs, CDS, CDS^2 mitigated the risk, but it turns out i did not." I must quibble. Yes it is true that everyone thought spreading mitigated risk. In reality, it actually did mitigate the risk, just not to the extent expected by many. Indeed, a colleague of mine was desperately putting together a "recession" model for CDOs that included suitable co-variances lacking in the standard models (e.g., if foreclosure rates rise in an area past a certain threshold, house prices in general will be depressed ). There was no market for such a product because it introduced significant non-linearity in the valuation models. Nonetheless, clearly the folks at magnetar understood what was going on.