Monday, January 09, 2012

Get your money out of PIMCO....

....because its CEO, the ubiquitous Mohamed El-Erian is a nincompoop!

Check out this gibberish in today's WSJ:

"Fat tails"—the technical term for the extremes of an outcome distribution—are risks for any global system that loses its anchors. Economies and markets function differently, companies and households feel unsettled, and policy measures become less effective.

Oh my. Where to begin.

First, "fat tails" is not a "technical term". The technical term is excess kurtosis. Fat tails is the colloquial, layman's term.

Second, fat tails is decidedly NOT a term for "the extremes of an outcome distribution"! The normal distribution is an outcome distribution. It does not have "fat tails". In fact it is the lack of fat tails in the normal distribution that lead so many models to go astray

How can this dude spew nonsense like this and get away with it? He's failing Stats 101.  It must be the 'stache.

Finally, the second sentence is even weirder than the first. I cannot make out what he is saying. Is he trying to say that recent events have changed the shape of the "outcome distribution"? Or that when we realized the outcome distribution had fatter tails than was previously thought, people changed their behaviors? The second interpretation at least makes some sense.


John Thacker said...

"Heavy tailed" is a technical term in probability (where I have my PhD.) "Fat tails," however, is not in my experience.

Heavy tailed distributions have been studied for a long time and have a precise mathematical definition.

Actually, I know a fair number of probabilists who consider "excess kurtosis" to be a statisician's term, another way of saying a layman's term, and would prefer referring to the 4th moment or the 4th standardized moment. They are as skeptical of "kurtosis" as a term as you are of fat tailed.

codeandculture said...

I read the statement and just kind of mentally plugged in "where a lot of density is in" before "the extremes of an outcome distribution." I didn't even notice the elision until you unpacked it. We all use short-hand and I tend to give people the benefit of the doubt on it.

Angus said...

C&C: I know what you mean, but alleged economic experts should not be "using shorthand" (which is a very generous way to put it) in op-eds

matto said... provides you the most competitive quality training material like TestKing, IT Exams King, BrainDumps, pass4sure, ExamCollection, RealExams, ActualTests, question, answers of all IT certifications, exams, tests on cheap prices with 100% money back guarantee in case of failure in exams Such as Cisco, CompTIA, Oracle, Microsoft etc.plz Contact Us: ITSOLEXPERT Limited, Fern hills Business Centre, Foerster Chambers, Todd Street, Bury, Manchester, United Kingdom. Post Code: BL9 5BJ

Anonymous said...

I think your criticism of El Erian is very petty and juvenile.
Fine, the use of the term "fat tails" is not in the lexicon of mathematical statistics. But we all know what he means by this; at least those of us with any common sense.
The term "black swan" is equally un-mathematical, but pretty much anybody knows what is meant by that phrase when discussing financial markets.

Of course, Benoit Mandelbrot's book "The Misbehavior of the Markets - A Fractal View of Financial Turbulence" really is the go to source in explaining how markets mis-behave, and he shows quite clearly that returns in financial markets do not follow a normal distribution and that financial markets exhibit many of the attributes of chaotic systems.
It is this reason - the chaotic behavior of markets - that renders their modeling using conventional statistics or deterministic mathematics - simply dead wrong.

Simon Spero said...

"Fat tails" appears to be used in the financial sector, and appears to be a term of art in that domain, and does appear to roughly gloss as "leptokurtic".

Alexo said...

When I see one of his pieces... I pass.

El Erian is in a continuous writing, elegantly playing with "macro" concepts, the man is not afraid of abstractions.

The problem is that all together the resulting stuff is not very actionable (at least to me).

There is a puzzle: how can he conciliate such a heavy prose production with his executive heavy duties?

Anonymous said...

It reads like a Nicholas Nassim Taleb random word generator!

But seriously, it is nonsense.