Hi,
I am using a dynamic panel regression using xtabond2 where y=lagged y + control variables
I ran the regressions then I changed the y variable for robustness . The AR(2) test remained the same
Then for additional robustness I took a subsample where i removed teh 2007-2009 financial crisis years . The AR(2) test gave very different result.
I am wondering if it is normal to end up with the same AR(2) when I change the dependent variable or is it a sign that i did a programming mistake?
also is it normal that AR(2) drops from 70.3 to 12.9 if i omit three years from the sample?
I am using a dynamic panel regression using xtabond2 where y=lagged y + control variables
I ran the regressions then I changed the y variable for robustness . The AR(2) test remained the same
Then for additional robustness I took a subsample where i removed teh 2007-2009 financial crisis years . The AR(2) test gave very different result.
I am wondering if it is normal to end up with the same AR(2) when I change the dependent variable or is it a sign that i did a programming mistake?
also is it normal that AR(2) drops from 70.3 to 12.9 if i omit three years from the sample?
