|
Activity Number:
|
146
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 30, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Business and Economics Statistics Section
|
| Abstract - #309391 |
|
Title:
|
A Power Study of Seasonal Unit Root Tests for Quarterly Data
|
|
Author(s):
|
Qianyi Zhang*+ and David Dickey and Sastry Pantula
|
|
Companies:
|
North Carolina State University and North Carolina State University and North Carolina State University
|
|
Address:
|
3520 Cum Laude Ct, Raleigh, NC, 27606,
|
|
Keywords:
|
seasonal unit root ; frequency ; power
|
|
Abstract:
|
There are two real unit roots (1 and -1) and a pair of conjugate unit roots (i and -i) for quarterly processes with seasonal unit roots at all frequencies. This paper reveals why the test of Dickey, Hasza and Fuller (DHF), which tests the null hypothesis that the characteristic equation has all 4 of these roots, has low power when the underlying processes have seasonal unit roots at some but not all of these frequencies. By focusing on the time series with seasonal unit roots only at zero and semi-annual frequency (1 and -1), this study proposes a test of the null hypothesis of seasonal unit roots at all frequencies against the alternative of seasonal unit roots of 1 and -1 only. Monte Carlo simulations are used to compare the performance of DHF, Kunst test and the proposed test.
|
