Heteroskedasticity Robust Panel Unit Root Tests
Author
Summary, in English
This paper proposes new unit root tests for panels where the errors may be not only serial and/or cross-correlated, but also unconditionally heteroskedastic. Despite their
generality, the test statistics are shown to be very simple to implement, requiring only minimal corrections and still the limiting distributions under the null hypothesis are completely free from nuisance parameters. Monte Carlo evidence is also provided to
suggest that the new tests perform well in small samples, also when compared to some of the existing tests.
generality, the test statistics are shown to be very simple to implement, requiring only minimal corrections and still the limiting distributions under the null hypothesis are completely free from nuisance parameters. Monte Carlo evidence is also provided to
suggest that the new tests perform well in small samples, also when compared to some of the existing tests.
Publishing year
2014
Language
English
Pages
112-135
Publication/Series
Journal of Business & Economic Statistics
Volume
32
Issue
1
Links
Document type
Journal article
Publisher
American Statistical Association
Topic
- Economics
Keywords
- Unit root test
- Panel data
- Unconditional heteroskedasticity
- GARCH
- Crosssection dependence
- Common factors
Status
Published
ISBN/ISSN/Other
- ISSN: 0735-0015