|
PureBytes Links
Trading Reference Links
|
For all of those MS'er's who are going to ask, what's a Hurst
Exponent. It's math concept from fractals and chaos theory. Here's
the simplest explanation I've seen, and I've added some translation
for you.
The Hurst Exponent is a measure of the smoothness of fractal time
series based on the asymptotic behaviour of the rescaled range of the
process. (Translation: it measures correlatons in a data series on
any time scale. Simpler Translation: it measures the fractal
dimension of a data series. Even Simpler Translation: it measures how
much fractals jump around--well sort of.)
The Hurst exponent, H, is defined as:
H:=log(R/S)/log(T)
where T is the duration of the sample of data, and R/S the
corresponding value of rescaled range.
Hurst generalized an equation valid for the Brownian motion in order
to include a broader class of time series. In fact, Einstein studied
the properties of the Brownian motion and found that the distance R
covered by a particle undergoing random collisions is directly
proportional to the square-root of time T:
R=k*T0.5
where k is a constant which depends on the time-series. The
generalization proposed by Hurst was:
R/S=k*TH
where H is the Hurst exponent.
If H=0.5, the behaviour of the time-series is similar to a random
walk;
if H<0.5, the time-series covers less "distance" than a random walk
(i.e., if the time-series increases, it is more probable that then it
will decrease, and vice-versa);
if H>0.5, the time-series covers more "distance" than a random walk
(if the time-series increases, it is more probable that it will
continue to increase).
Given a time series x(n), n=1,....N, H can be estimated by taking the
slope of (R/S) plotted vs. n in a log-log scale.
H is related to the fractal dimension D:
H=E+1-D
where E is the Euclidean dimension (E=0 for a point, 1 for a line, 2
for a surface). For one-dimensional signals, H=2-D
H is also related to the "1/f" spectral slope:
=2H+1
I think this has to be programmed in C++ and then imported into MS as
a dll. Erik Long used to have a MS product for this. He may be out of
business but you can call and find out. His product for MS was called
Fractal Finance.
He wrote an article in the May 2003 S&C called Making Sense of
Fractals. Some code may be in there that's useful to you, but I don't
think so.
Tetrahex
555 W. Madison Street
Chicago, Illinois 60661
Contact: Erik Long
Contact number: 312.775.7468
Some MS users have programmed a fractal noise measurement that
approximates the Hurst and have used that successfully.
Join this group and ask them for John Connors C++ code for Hurst, or
for an approximation of Hurst or if they know of any conversions.
This group really loves that stuff. Find Igor. He knows about
approximations in MS.
http://groups.yahoo.com/group/Behavioral-Finance/
I haven't really progressed past moving averages so this stuff is way
beyond me. I'm still trying to figure out cloning.
However, many, many people have gone crazy trying to program the
Hurst Exponent in MS, so Roy leave this one alone.
JO
--- In Metastockusers@xxxxxxxxxxxxxxx, karile <karile@xxxx> wrote:
>
> Hi,
>
> Can someone code the Hurst Exponent in Metastock ?
>
> Thanks for your help,
>
> Karile
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Rent DVDs Online - Over 14,500 titles.
No Late Fees & Free Shipping.
Try Netflix for FREE!
http://us.click.yahoo.com/Tq9otC/XP.FAA/3jkFAA/zMEolB/TM
---------------------------------------------------------------------~->
To unsubscribe from this group, send an email to:
Metastockusers-unsubscribe@xxxxxxxxxxx
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
|