Dynamic Panel threshold model

[Gabriel Temesgen]

I am writing a paper applying a dynamic panel threshold model. however, I am not sure if I apply correctly the STATA codes of Seo et al. (2019). I am using 40 countries and 20 years dataset.
xthenreg pb ldebt cab_gdp , endogenous( outputgap gvar)
where pb refers to primary balance GDP ratio is the dependent variable, ldebt refers to lagged debt to GDP ratio the threshold variable; cab_gdp refers to current account balance as exogenous variable and output gap and government expenditure gap as a set of control endogenous variables. I put the result I found here
My questions are:
1. I read a paper putting results in lower regime, upper regime and overall. I can't divide the regimes to lower and upper regimes. and don't know as well how to run the post estimation codes.
2. How can I run a Post-estimation tests? I snapshot the table from the article ( (10.1002/mde.3262).
3. Can I check if debt square and debt cubic threshold exists?
4. Can I calculate the threshold for every country?
I kindly ask you to drop me the STATA codes for the questions above.
Thanks

[Rayhana Hasnaoui]

Dear Gabriel, I'm also interested in the Dynamic threshold panel data. If you got any news about the post estimations commands, please, share it with me. Kindest Regards

[Khethang Mokoena]

Hi, Guys i am also using panel threshold estimation of Kremer et al, (2013) i have been following the example/guideline on Stata. I am using stata 18 version. But when i run the model i get the following error
(nquantiles() must be less than or equal to number of observations plus one). I estimated the model
xtendothresdpd SMI L.SMI GDP INF EXR MSY FDI, thresv(EXP) stub(enr) pivar(EXP) dgmmiv(SMI) twostep vce(robust) sig(0.05). Unfortunately, i have liited knowledge of Seo model.


r(198);