I discuss the determination of the Hurst exponent in the second half of the attached paper.

Thanks dear Steve! I can't seem to find the attachment. Could you upload it again, please?

http://www.actuaries.org/AFIR/Colloquia/Orlando/Craighead.pdf

I found another more directed reference than my old paper:

http://blogs.cfainstitute.org/investor/2013/01/30/rescaled-range-analysis-a-method-for-detecting-persistence-randomness-or-mean-reversion-in-financial-markets/

There appears to be several different examples if you Google "Rescaled Range Analysis"

Good Luck!

Dear Steven!

Oh, yes, indeed an older paper ;-) Thank you! I already came across the scond link you postet.

The true reason why I wrote was  the following:

I am already using 2 different scripts in maltab:

http://www.mathworks.com/matlabcentral/fileexchange/25414-rescaled-range-analysis

and

http://www.mathworks.com/matlabcentral/fileexchange/4325-rescaled-range-analysis

However, I am kind of not trusting both because e.g. when I create a browian noise with a H=0.5 a test input (matlab: bn=wfbm(0.5,1000)) , I get an H value of 1 as the output!

Also when integrating white noise, I instantly get H=1.

is this normal for R/S analyses??

Thanks fo much!

Just to follow up for the records:

The following code is a collection of different Hurst exponent estimation methods and works nicely!

http://www.mathworks.com/matlabcentral/fileexchange/19148-hurst-parameter-estimate