I am analysing a panel data set and have found my independent variables to be stationary and my dependent variable to be non-stationary but when first differences my dependent variable became stationary. so i am now running cointegration tests. I firstly used the Kao test and it suggested cointegration on all outputs in the test but the augmented dickey fuller section. I then tried to do the Pedroni test but where the values should be stata is outputting only a full stop “.” So I am unsure what this means or what I have done wrong? Has anyone experienced this please?
Also, I read in other threads that these were the best first two steps before carrying out the Hausman test, Breusch-Pagan Lagrange multiplier test, tests for heteroskedaticity, endogeneity, cross sectional dependence.
Please can someone advise me on the optimal order of these tests and whether I should be adjusting for these things in between tests at all? I chose to adjust the dependent variable by first difference t before i asked stata to carry out cointegration tests.
Any help will be greatly appreciated
Also, I read in other threads that these were the best first two steps before carrying out the Hausman test, Breusch-Pagan Lagrange multiplier test, tests for heteroskedaticity, endogeneity, cross sectional dependence.
Please can someone advise me on the optimal order of these tests and whether I should be adjusting for these things in between tests at all? I chose to adjust the dependent variable by first difference t before i asked stata to carry out cointegration tests.
Any help will be greatly appreciated
Comment