[Muhammad Ibrahim Shah]

JanDitzen I have another query. Lets say I have y as independent variable, 5 independent variables (x1,x2,x3,x4,x5). we test for a unit root in the panel data setting and get y,x2, and x3 as stationary at the first difference (I(1)). But x1,x4, x5 are I(0), i.e., stationary at level. so my command of CS-ARDL would look something like this:

xtdcce2 d.y, lr(l.d.y x1 d.x2 d.x3 x4 x5 ) lr_options(ardl) cr(d.y x1 d.x2 d.x3 x4 x5) cr_lags(0) reportconstant

Is this correct? Since my dependent variable, x2 and x3 are stationary at difference, I put d. infront of them whereas I use x1,x4 and x5 in level form since they are stationary at level?

[Niti Khandelwal]

Dear All,
I have a very basic question regarding the panel ARDL technique. I have a mix of I(0) and I(1) variables with the dependent variable being I(1). This justifies the use of panel ARDL technique. But, when I am estimating the model, should I include I(0) variables in levels or in first differences?