i am a phd student working on Oil Price Shocks and asymmetries of industrial production for which i am using panel ARDL technique i have data for 40 countries divided into 4 group A=8 B= 22 C=10 D=1 based on oil imports number of variables i have are 5 .
for all groups i have to calculate model separately as in A groups oil imports are 0-10% B =10-20% C=20-30% D=30-40% of overall imporst
i am estimating asymmetric model has divided the oil shocks into positive and negative on basis of shin et al (2014)
i have to select optimal lag of the variable for which i am using code
forval i=1/8{
ardl lop lip lex opp opn if (cc==i)
matrix list e(lags)
di
}
which gives no observations
r(2000);
i have changed variable type to byte and then to float but of no benefit i am worried as i have investigated alot. also removed missing values.
will you please suggest solution of problem