The increment ratio statistic
We introduce a new statistic written as a sum of certain ratios of second-order increments of partial sums process of observations, which we call the increment ratio (IR) statistic. The IR statistic can be used for testing nonparametric hypotheses for d-integrated () behavior of time series Xt, including short memory (d=0), (stationary) long-memory and unit roots (d=1). If Sn behaves asymptotically as an (integrated) fractional Brownian motion with parameter , the IR statistic converges to a monotone function [Lambda](d) of as both the sample size N and the window parameter m increase so that N/m-->[infinity]. For Gaussian observations Xt, we obtain a rate of decay of the bias EIR-[Lambda](d) and a central limit theorem , in the region . Graphs of the functions [Lambda](d) and [sigma](d) are included. A simulation study shows that the IR test for short memory (d=0) against stationary long-memory alternatives has good size and power properties and is robust against changes in mean, slowly varying trends and nonstationarities. We apply this statistic to sequences of squares of returns on financial assets and obtain a nuanced picture of the presence of long-memory in asset price volatility.
Year of publication: |
2008
|
---|---|
Authors: | Surgailis, Donatas ; Teyssière, Gilles ; Vaiciulis, Marijus |
Published in: |
Journal of Multivariate Analysis. - Elsevier, ISSN 0047-259X. - Vol. 99.2008, 3, p. 510-541
|
Publisher: |
Elsevier |
Saved in:
Online Resource
Saved in favorites
Similar items by person
-
Minimum distance lack-of-fit tests under long memory errors
Koul, Hira, (2015)
-
A two-sample test for comparison of long memory parameters
Lavancier, Frédéric, (2010)
-
Asymptotics of M-estimators in two-phase linear regression models
Koul, Hira L., (2003)
- More ...