ARDL Bounds cointegration test for panel data

[Curious86]

Dear Experts,

I am new to this forum and would like to learn from your expertise.

I am working on a panel dataset with N = 39 and T = 14 (an unbalanced panel). For further analysis, I need to perform the cointegration test to examine the presence of a long-run relationship among the variables. I have the following questions about this.

i. When I select the Fisher (combined Johansen) option under the panel cointegration test, the error message is "Invalid or duplicate specification or insufficient data for estimation". What could be the possible reason and is there any solution to this issue?

ii. For the Bounds test for cointegration in panel data, I have watched a video on YouTube (https://www.youtube.com/watch?v=KbmWdvov9Sc), suggesting that it can be performed in Eviews 13 version. However, I am curious to know if this functionality is available in the EViews 12 Student Lite version as well. If not, does the Eviews 12 University version have the scope to perform the Bounds test for panel data?

I would appreciate your valuable suggestions.

Thanks

[EViews Mirza]

Hello there!

The most likely reason for seeing the "insufficient data for estimation" error is the fact that you have T = 14. Note that each lag specification will effectively sacrifice that many observations from each cross-section. Another possibility is that not all cross-sections actually have even 14 observations due to NAs.

As for the PMG bounds test, I don't believe it's available in EViews 12. It was a feature we added in EViews 13 and later.