From: John Conover <john@email.johncon.com>
Subject: Re: Profit dreariness
Date: 21 Jul 1999 16:48:29 -0000
David Lloyd-Jones writes:
>
> "S. Hales" wrote:
>
> > Depends on the time scale. Over time a stock's price volatility could
> > approach the normal distribution but over shorter time scales exhibit fat
> > tails.
>
> But empirically the fat tails show up at all time scales, don't they?
>
> My impression was that people only us normal and log-normal distributions
> because they're easy to study, not because they represent the real world -- like
> the guy looking for his keys under the streetlight rather than where he lost
> them.
>
Hi David. Yea, the US exchanges exhibit leptokurtosis for at least the
last 27 years of daily data:
http://www.johncon.com/john/correspondence/981229233103.31169.html
The DJIA does the same for the century. Looks like about a 57%
persistence from one day to the next.
John
--
John Conover, john@email.johncon.com, http://www.johncon.com/