EViews.com

This thread is about LSunit add-in that perfoms Lee-Strazicich LM unit root tests with one or two endogenous structural breaks

Hello,

Thank you for the add-in. Is there any equivalent of the "THIN" code in RATS to deal with big databases. Indeed, I have 4600 observations and it takes a really really long time to perform the test using lsunit add-in.

Thank you for your help

Thanks for the good point. I will try to include the thin option soon.

Many thanks for providing this useul add-in
I installed the add-in in my Eviews 9.5
When using the add-in with Model selection = Break and Break number = two (leaving all other settings o default values), I do obtain an Error Message stating "TEMP is not defined or is an illegal command."
Any idea on the reason for such an error?
Many thanks

Max

could you provide the data?

Dear all,

I have also encoutered the same problem when I tried to test for the presence of unit roots under the Break model, two breaks:

"TEMP is not defined or is an illegal command".

I would very much appreacite it if I could get some help.

Thanks!

The bug is fixed. Please update the add-in.

What Information Criteria is using this package? Is it AIC or BIC? How I can choose it to be BIC? Thanks.

There is no option to choose the information criteria. It uses t statistic to select lag length.

Dear all,
I am trying to run Lee Strazicich ur test but get "Error 72 in encrypted program" message. On the sample data provided with the add-in it works. What could be a problem and how can it be fixed?
Thank you!

Could you post the data?

Hello,

Could I please ask how is the maximum number of lags used determined? I understand that I can reduce them instead of 8 that is set by default in the LSunit add-in but my results change if I set it as max 8 or max 4.
Thank you very much for your time.

Yes, you can change the maximum number lags.

dakila wrote:Yes, you can change the maximum number lags.

Thank you very much for that Dakila! But how is the maximum number determined?
If I use 8, result supports unit root if I use 4, no unit root is found. But what is the actual way I choose if I ll use 8 or 4 as maximum taking into account that they give different results?

Thank you very much for your time!

You can't determine the maximum number of lags. You have to choose it according to some expert suggestions.