Sunday, July 12, 2009

Tail Volatility and 1987

There was an interesting paper on estimating the probability of the tails via option prices (see Backus, Chernov, and Martin here). Barro seems to have really revived the Rietz 'peso problem' model, by looking at a variety of large macro-economic shocks related to wars and revolutions in the twentieth century. There's an excellent summary of this work here.



One person noted that things could actually be pretty boring. That is, in 1987 the market fell so much on one day (23%), ever since then there has been a large premium for out-of-the-money options. The graph above is from Jackworth and Rubinstein (1994). This shows that basically the market did not anticipate the 1987 crash, but since then, has this event priced in.

The funny thing is how far back and how deep this literature on disasters is. Nassim Taleb and his acolytes (the Taleban) seem to think economists are totally wedded to the Gaussian distribution, which ignores fat tails and extreme events. That is, he sees the naive Black-Scholes model, and ignoring the volatility smile, infers that this means economists don't believe disasters matter. This is just a really ignorant and misleading statement, fun for dilettantes who love to lampoon experts via a caricature, but it is really a waste of time.

5 comments:

Lance said...

You might be interested in this EconTalk post with Justin Fox: http://www.econtalk.org/archives/2009/07/justin_fox_on_t.html

InVivoAnalytics.com said...

Congrats on your new book!

Teresa

AHWest said...

I'm so sick of Taleb's and Roubini's constant media promotion of themselves. I don't think either one specifically forecasted the kind of crisis that happened. But several of my colleagues now think that they have. I argue that saying "there will be an unexpected crisis sometime" is not a forecast.

I do know a few fellow equity analysts that were grumbling about insane credit terms in mortgages and ratings practices that would "take down the firm" at the major ratings agency that EF didn't work for. And a bunch of equity analysts chuckling at David Weiss in 2006 when he said that the US could avoid recession as long as the property market held up. But I don't recall anyone specifically making big market bets based upon this unspecific concern.

I'd say that CLSA's Chris Wood was the earliest outspoken forecaster of the collapse of securitized loans in general (that I heard), back in maybe early 2007. He also called for gold at $2000 which hasn't worked out yet.

I really like (the vast majority of) your book, just finished it.

Eric Falkenstein said...

I appreciate that AH. I don't expect all of it to work on any thinking person, so I'm glad you liked most of it. If you could put up a review at Amazon that would be really helpful!

Anonymous said...

i don't care for taleb. precisely for the reason you stated. he's an economist, not a psychologist. has he forgotten that?

but kahnemann's heuristics are very real to me.
he's a psychologist, not an economist.
the studies speak for themselves. you can't attack him. you can only attack the methods used and data collected in the studies.
or form your own alternative conclusions. and that's how science should be.