Mj. Cannon et al., EVALUATING SCALED WINDOWED VARIANCE METHODS FOR ESTIMATING THE HURST COEFFICIENT OF TIME-SERIES, Physica. A, 241(3-4), 1997, pp. 606-626
Three-scaled windowed variance methods (standard, linear regression de
trended, and bridge detrended) for estimating the Hurst coefficient (H
I are evaluated. The Hurst coefficient, with 0 < H < 1, characterizes
self-similar decay in the time-series autocorrelation Function. The sc
aled windowed variance methods estimate H for fractional Brownian moti
on (fBm) signals which are cumulative sums of fractional Gaussian nois
e (fGn) signals. For all three methods both the bias and standard devi
ation of estimates are less than 0.05 for series having N greater than
or equal to 2(9) points. Estimates for short series (N < 2(8)) are un
reliable. To have a 0.95 probability of distinguishing between two sig
nals with true H differing by 0.1, more than 2(15) points are needed.
All three methods proved more reliable (based on bias and variance of
estimates) than Hurst's rescaled range analysis, periodogram analysis,
and autocorrelation analysis. and as reliable as dispersional analysi
s. The latter methods can only be applied to fGn or differences of fBm
. while the scaled windowed variance methods must be applied to fBm or
cumulative sums of fGn.