Hi, I encounter this message while doing cross sectional analysis. . dfuller FDI___, noconstant regress lags(2) sample may not include multiple panels r(459); what should i do, Sir?
Well I think 8 lags will serve the purpose in general to specific modeling approach. It was just a demonstration. Better use 12/8/4 lags for quarterly data depending on size of data
Hiii...What if some variable wound be I(0) and some (1)?
How can we test cointegration?
Thanks
What if one of my variables is stationary at level? Thanks in advance!
Use ARDL, not JOhansen cointegration
The zooming is not clear
Hi, I encounter this message while doing cross sectional analysis.
. dfuller FDI___, noconstant regress lags(2)
sample may not include multiple panels
r(459);
what should i do, Sir?
Can you please elaborate little more
Should the Johansen Cointegration test by conducted on variables at levels or in their transformed state?
Level but log transformation can be applied
I mean log transformation before applying for test is ok
amazing
Glad you think so!
Why have you used 10 lags in the command varbasic
Well I think 8 lags will serve the purpose in general to specific modeling approach. It was just a demonstration. Better use 12/8/4 lags for quarterly data depending on size of data
Please share the dataset link
The very first command is data link
webuse lutkepohl (A built in STATA data)
webuse lutkepohl
Type this in command window