Hallo everyone!
I am using the fmols and dols command for my study. I have T 33 and N 20 unbalanced panel data, with heteroskedecasity, slope heterogeneity, no cross sectional depedendence, unit roots stationary at first differnece and co-integration (Westerlund).
I get significant results when I run fmols, ccr and xtmg. But when I run dols I get this error: estimates post: matrix has missing values.
I have made sure to remove all missing observations and I still get this error.
I am running a simple fmols and dols code:
xtcointreg dep, indep indep indep, est(dols).
My dependant variable is gini (all logged transformatios). I've used both disposable and pre-tax gini and get the same error for dols.
I have checked the Stata forum and I am fairly new to this, so I'm reaching out here.
Please let me know if you have any other questions I can answer to help with this. Thanks!
My first question would be if it is a balanced panel, because if it is not it could be the cause. The second suggestion is to check the stationarity of the variables, perhaps in difference if you can estimate your model.
To solve the "estimates post matrix has missing values" issue in DOLS (Dynamic Ordinary Least Squares) estimation, try the following:
Once these are addressed, re-run the DOLS estimation.
To solve the "estimates post matrix has missing values" issue in DOLS estimation:
Re-run the DOLS estimation after addressing these issues.
- Badges
- Science topic
- Similar topics
- Correction