|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
st: re: hadrilm vs xtfisher
<>
Carlos said
I have a question regarding two tests for panel unit root :
When is it ok to rely on the hadrilm test (null hypothesis: all panels
have stationary time series) as opposed to t
he xtfisher (Maddala-Wu) (null: all panels have nonstationary time
series)?
Or is it htat the hardrilm is not a test for unit roots while the
xtfisher is?
These tests, hadrilm vs. xtfisher, stand in the same relation as the
KPSS test (findit kpss) and Dickey-Fuller type tests (help dfuller,
help dfgls). The D-F null, for a single timeseries, is that of I(1),
reject implies I(0). The KPSS test has a null of I(0), reject implies
I(1) (or I > 0, strictly speaking). Panel unit tests may be set up the
same way, where you assume all series are I(1), all series are I(0),
or somewhere in between (e.g. at least one of the series is not I(1)).
Kit Baum | Boston College Economics & DIW Berlin | http://ideas.repec.org/e/pba1.html
An Introduction to Stata Programming
| http://www.stata-press.com/books/isp.html
An Introduction to Modern Econometrics Using Stata | http://www.stata-press.com/books/imeus.html
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/