Very straightforward explanation. what if i have a panel data of daily basis of 4 years for 6 variables and 65 cross sections. Cross sections dependence is present, all are I(0), and there is an issue of endogeneity. Would FMOLS and DOLS be suitable ? Thanx
@@Dr.Munshi-Naser Thank you for your reply. But as data is large i can assume it is normally distributed and use DOLS right? then i can use both and compare on basis of testing normality of errors and RMSE? would you recommend this?
Thanks dr. Naser for this amazing vedio. I have a panel data N=7 and T=22 years, my DV and most of Iv's are cross sectional dependent, using 2nd gen unit root test they are stationary at level and 1st deference, and heterogenous slops. In your opinion which method should I adopt in order to build a linear model, please?
thank you very much for this amazing explanation , I have a problem with my xtcointreg command that I use to apply fmols and dols models However, after installing xtcointreg and cointreg packages, I still receive an error message saying that the command is unrecognized or a function error? How can I resolve this issue? Thanks a lot in advance
Thank you verry much for this video. i understand better with this method. but i have a question. I've done tests (sargan test, abond test, unfamiliarity test) to choose between the SYSGMM and FDGMM models as the best model. but the results obtained in both models do not meet the Sargan test. so which model should i choose? is it ok for me to use the FEM model while still including the variable dependent lag