I want to use
-Narayan and Popp (2010) unit root test with two endogenous breaks.
-LM unit root test of Lee and Strazicich (2003) which can accommodate two breaks in the intercept and the slope.
-KPSS stationarity test of Carrion-i-Silvestre and Sansó (2007) with two endogenous breaks in the intercept and trend.
Also, I want to use
-Panel Kwiatkowski–Phillips–Schmidt–Shin (KPSS) stationarity test with multiple breaks (Carrion-i-Silvestre, Barrio-Castro, and Lopez-Bazo 2005).
- LM unit root test with structural breaks (Im, Lee, and Tieslau 2005).
Please let me know how to use these tests in Eviews.
unit root with multiple structural breaks
Moderators: EViews Gareth, EViews Moderator
-
- EViews Developer
- Posts: 2672
- Joined: Wed Oct 15, 2008 9:17 am
Re: unit root with multiple structural breaks
None of the above tests are supported as in-built features. I will put them on the list for future development consideration.
Re: unit root with multiple structural breaks
Hello,
Could I please ask if in the unit root test I need to compare the estimated t-stat with all the critical values provided?
For example, in the results below do I consider my series as having a unit root at 1% and thus I need to test the first differences or as stationary (not having a unit root at 10%)? although I ve searched the forum for answers I couldnt find something relevant, so apologies if a post has skipped.
Lee Strazicich LM unit root test
Model: Crash (A)
Null hyphothesis : L_EL_EV has a unit root with break
Minimum test statistic (tau) -3.150229
Break point 2008M11
Selected lag 6
Test critical values 1% level -3.845292
5% level -3.280960
10% level -2.988276
Regression output
Variable Coefficient t statistic
S(t-1) -0.048952 -3.150229
Constant 0.001905 0.778143
B(t) -0.080214 -2.036415
Thank you very much
Could I please ask if in the unit root test I need to compare the estimated t-stat with all the critical values provided?
For example, in the results below do I consider my series as having a unit root at 1% and thus I need to test the first differences or as stationary (not having a unit root at 10%)? although I ve searched the forum for answers I couldnt find something relevant, so apologies if a post has skipped.
Lee Strazicich LM unit root test
Model: Crash (A)
Null hyphothesis : L_EL_EV has a unit root with break
Minimum test statistic (tau) -3.150229
Break point 2008M11
Selected lag 6
Test critical values 1% level -3.845292
5% level -3.280960
10% level -2.988276
Regression output
Variable Coefficient t statistic
S(t-1) -0.048952 -3.150229
Constant 0.001905 0.778143
B(t) -0.080214 -2.036415
Thank you very much
Who is online
Users browsing this forum: No registered users and 8 guests