Dear STATA users,

I used STATA15 to run cointegration tests for my strongly balanced panel data of 400 observations ("xtcointtest"). I received different results for cointegrations tests of KAO, PEDRONI and WESTERLUND. Could anyone please help me to say whether I could conclude that I have cointegration between variables or not.

Please help to review test results as below:

KAO: cointegration in all 5 statistics

PEDRONI: inconclusive (Modified Phillips-Perron t said no cointegration while Phillips-Perron t and Augmented Dickey-Fuller t said cointegration). Also what are different between three statistics?
Statistic p-value
------------------------------------------------------------------------------
Modified Phillips-Perron t 0.7082 0.2394
Phillips-Perron t -2.3174 0.0102
Augmented Dickey-Fuller t -2.9049 0.0018


WESTERLUND: "cointegration" in some panels (no option) and "no cointegration" in all panels (option: allpanels)


In addition, I also have questions of processing steps afterwards. In case no cointegration, what should I do next? In case of cointegration, I understand that I need to estimate the long and short-term dynamics between variables. It seemed that I should use command "xtpmg" to estimate. Could anyone guide me more about it?

Thank you so much and I look forward to hearing from you!

All the best,
Anh Nguyen