Panel CIPS Unit Root Test STATA — Transcript

Full transcript

1. 0:00in this video I'm going to be examining
2. 0:03the stationarity of a variable using the
3. 0:06cips unit truth test DC IPS initial test
4. 0:09is used in the presence or when there is
5. 0:13cross-sectional dependence in a model
6. 0:15and the command is the XTC IPS is active
7. 0:19cips for this video I'll be using the
8. 0:21log of GDP as an example uh so I'll be
9. 0:25testing the stationarity of the log of
10. 0:26GDP lgdp I'll click a comma and include
11. 0:31the max the max lag
12. 0:33and the BG lag
13. 0:37so I'm including the max lag to B2 and
14. 0:40the BG lag to be also also be two the
15. 0:43mag slag specifies the maximum number of
16. 0:45logs to be included in the model while
17. 0:48the BG lag establishes the serial
18. 0:50correlation order I click enter
19. 0:54now we can see the coefficients of the
20. 0:57cips test which is uh in absolute value
21. 1:00it is 2.558 and is greater than all the
22. 1:05conventional levels of statistical
23. 1:06significance so these mean that our
24. 1:09variable is stationary in levels you
25. 1:12could also include a Time Trend in the
26. 1:15model
27. 1:17so the results is still similar we could
28. 1:20see that the coefficient value of the
29. 1:21cips test is greater than the
30. 1:24coefficient value the critical values
31. 1:26are ten percent five percent and one
32. 1:28percent indicating that our variable is
33. 1:30stationary in in levels in case our
34. 1:34variable is not stationary levels we
35. 1:35could actually
36. 1:37re-estimate the log of GDP the
37. 1:39stationary term log of GDP in Phase
38. 1:42Difference by including a d dots
39. 1:47as before we can see that the
40. 1:49coefficients of the CRPS test is greater
41. 1:52than conventional levels of statistical
42. 1:54significance the critical values of the
43. 1:57conventional levels of statistical
44. 1:58significance you could also include the
45. 2:01time Trend in the analysis and phase
46. 2:03difference
47. 2:05we could also see similar conclusions
48. 2:08where the corruption of the cips test is
49. 2:10greater than the critical values at 10 5
50. 2:13and 1 so this this variable is
51. 2:18stationary levels and as well also
52. 2:20stationary after first difference