Ex-hedge quant here. Here's my semi-informed take. Quants didn't write most of these toxic derivatives, and they weren't conceived by computers, at least not a

Ex-hedge quant here. Here's my semi-informed take. Quants didn't write most of these toxic derivatives, and they weren't conceived by computers, at least not as far as I know. Though technological modeling was necessary to value many of these securities (Monte Carlo methods, etc.) I really doubt that any serious financial company would allow a computer to conceive of derivatives to any meaningful extent. But I could be wrong. These contracts, in many cases, were designed to be complex beyond comprehension, but AI methods don't help make them comprehensible to machines. I think of these 500-page derivatives contracts as similar to long software EULAs (which are intended to be so long and dry that 95% "TL;DR" and click "Accept", while even the diligent 5% miss details)... or to the packets in which health insurance companies detail exclusions. The things are intended to be virtually incomprehensible, making it easy for the writer to fool not only his customers, but also his bosses, by sweeping low probability risks (e.g. an "unprecedented" nationwide drop in real estate during a strong economy such as that which existed in mid-2007) under the rug. Quant funds were not hammered because (loosely speaking) the computers emergently "tricked" them. They got whacked because all the models were trained on the past, and 8/2007 - Pres. has been unprecedented. Also, the fact that a large number of seemingly independent agents (funds) have piled into the same few strategies means that an unwind from one will lead to losses for many.