Sign tests for unit root and change in persistence
The behaviour of unit root tests defined in terms of sign transform is here analysed. The sign transform allows us to gain robustness to non-normality. Contrarily to the standard unit root tests, our results show that these tests are very reliable in the presence of breaks, provided there is no change in the persistence of a series. In case of change in persistence, the paper proposes to compare sign-based test functions separately computed before and after the change point. A statistically significant difference would signal the presence of change in persistence. An example and a Monte Carlo study conclude the analysis.
Year of publication: |
2014
|
---|---|
Authors: | Furno, Marilena |
Published in: |
International Journal of Computational Economics and Econometrics. - Inderscience Enterprises Ltd, ISSN 1757-1170. - Vol. 4.2014, 3/4, p. 269-287
|
Publisher: |
Inderscience Enterprises Ltd |
Saved in:
Saved in favorites
Similar items by person
-
LM tests in the presence of non-normal error distributions
Furno, Marilena, (2000)
-
Estimation of a small macro-model under the assumption of contaminated distributions
Furno, Marilena, (1991)
-
Robust methods in econometrics : a review essay
Furno, Marilena, (1989)
- More ...