Hello, everyone.
I am writing this post to ask about the result of the unit root test. If you can help me
it would be huge help for me.
I have panel data and it is about firm number as a x variable.
To test if my data has unit root or not, I used a command
xtunitroot fisher Firm num( as a X variable), pperron lags(1)
And I got result as below
In thus unit root test's result, according to one user in this forum, I can simply discard one of the result. For example, if I have large T, small n data, I can discard the value of Pm Statistics because Pm statistics is normally used for large T and Large N data.
What about my case? Only Inverse normal Statistics indicates that my data has unit root.
1) Can I assert that my data has unit root? 2) In what cases Inverse normal Statistics can be used to prove that the data has unit root? (For instance, if I have large N, small T data, can Inverse normal Statistics be used?
Thank you in advance.
Have a good day.
I am writing this post to ask about the result of the unit root test. If you can help me
it would be huge help for me.
I have panel data and it is about firm number as a x variable.
To test if my data has unit root or not, I used a command
xtunitroot fisher Firm num( as a X variable), pperron lags(1)
And I got result as below
Statistic p-value | ||||||
------------------------------------------------------------------------------ | ||||||
Inverse chi-squared(52) P 119.5297 0.0000 | ||||||
Inverse normal Z -0.5196 0.3017 | ||||||
Inverse logit t(134) L* -2.6177 0.0049 | ||||||
Modified inv. chi-squared Pm 6.6218 0.0000 |
What about my case? Only Inverse normal Statistics indicates that my data has unit root.
1) Can I assert that my data has unit root? 2) In what cases Inverse normal Statistics can be used to prove that the data has unit root? (For instance, if I have large N, small T data, can Inverse normal Statistics be used?
Thank you in advance.
Have a good day.