Panel Regression versus Dynamic Models

[lal mohan kumar]

Dear all,
I am trying to replicate a study, where I am regressing the change in capital expenditures(DELCAPEX) is regressed on cash flows(CAFLOW) and Market to Book ratio(MB). I also want to add Lags of my dependent variable (here,L.DELCAPEX) and independent variable( L.CAFLOW). I ran the following fixed effects regression code in Stata.

xtreg DELCAPEX  CAFLOW MB L.DELCAPEX L.CAFLOW  i.year,fe vce(robust

I am having only 7 years of data. I don't know whether the above panel regression model is correct or not. Also
1. Can I include lagged values of dependent variable and independent variables in the model and estimate it via panel regression?
2. Does the no: of years(7 in my case) matter here?
Any help in this regard will highly helpful to me

[Carlo Lazzaro]

Ial:
if you plan to include lagged values of the regressand as predictors, you should switch to Dynamic Models.

[lal mohan kumar]

Thank you very much, Carlo
Hence in some cases, if lagged dependent variables are included as independent variables, then it is better to use a dynamic model .Am I correct to certain extent?

[Carlo Lazzaro]

Yes, you're.