Dear all:

I found a challenge when I run a dynamic model on long panel data. I am hopeful that you could instruct me how to resolve my doubt.

I used a dynamic model on my long panel data (Temperature and Precipitation of German NUTS3 regions) which consist of 372 NUTS3 regions and 29 years. However, I am struggling to transform the dynamic model into Stata command where I want to include the following as system GMM command:
Array
where my dependent variable is the GDP per Capita, and it is explained by: Temperature and its squared terms; Precipitation and its squared terms; plus 1 lag of GDP per Capita.

my question is how to conduct this dynamic panel model correctly using xtabond2 command in Stata. Could you instruct me how to run a dynamic model on my long panel data correctly? How can I determine exogenous and endogenous variable in this dynamic panel model? How to determine which instrument variables should be used in the dynamic model? Can anyone instruct me how to run dynamic model correctly?

To easy to reproduce my problem, I would like to share my dofile, I am very hopeful that someone will help me out, this problem is challenging for me. Thanks in advance