Tuesday, December 02, 2008

Mortgage Pig vs Black Swan

RIP. Gilliard was a master of the snark element, Chris Lightfoot of the hacker element, Tanta of the policy element of great blogging. Together it's the blogging triathlon. (Yes, I agree this sounds a bit chilly.)

More broadly, folk like "Black Swan" Taleb have to deal with the fact that their great unpredictable validating event was predicted, years in advance, with considerable precision, by unofficial people who weren't paid six figure bonuses in any currency, as well as by economics professors who were paid considerably more. The CR team can claim to have predicted the housing crash not like economists - There is a significant risk of this happening, but we can't say when - but like meteorologists. Something like a crash has been observed building up in the North Atlantic, and it is coming closer at a rate of 5 knots. We expect it any time after X hundred hours.


Anonymous said...

Not so much unpredictable, as unpredicted. As I understand it, Kaleb's argument is that our great financial institutions' prognosticating is based on a simplistic "tomorrow's weather will be the same as today's" model; right most of the time, but utterly unable to cope when a volcano erupts or a tsunami strikes (metaphorically speaking).

If these spectacular crashes can, in fact, be predicted reliably, then that only strengthens his case that the financial models in use are rubbish.

guthrie said...

I havn't read anything by Kaleb, but postulate that what you might call a "stereotypical black swan" view would be attractive to many, ie "Its all too complex to predict therefore its not our fault it went wrong". Contradicting all the evidence that yes, it was all their fault and yes, it was all entirely predictable. Even I, with no formal training but an eye for history, could see that it was a bad idea, and predicted the housing market downturn, albeit nearly a year too early.

Anonymous said...

The models were rubbish, in that the higher-ups apparently told the modelers to refrain from making unprofitable assumptions about housing prices ever going down.

I've not seen anything which doesn't come down to the fact that the people with authority in a lot of companies were making fat cash for riding the bubble, and pumping it up further. Those bonus checks have long since cleared the banks, and won't be given back.


[1] 'Unprofitable' for the higher-ups, who were collecting multi-million $/year bonuses for profits which were only justified under bad assumption.

kostenloser Counter