If my process has long range dependence (hurst exponent > 0.5 ) can it be concluded that it is stationary/non-stationary? How? Is there any correlation between Long range dependence and Stationarity?
Asked
Active
Viewed 70 times
1
kjetil b halvorsen
- 63,378
- 26
- 142
- 467
Vaib
- 71
- 5
-
1One way would be to use a Fourier series approach and apply a regular test for stationarity. – Digio Jul 28 '17 at 08:54
-
Sure, thanks. Please read the updated question once. – Vaib Jul 28 '17 at 09:00
-
Wide-sense stationary is assessed by a linear model so it has nothing to do with the length of the series anymore than a simple linear fit has to do with the number of rows in its training set. – Digio Jul 28 '17 at 09:16