[Leonardo Guizzetti]

Yeah that's bizarre. It does seem as though the coefficients following the "incorrect" syntax is simply evaluated as 0 before being passed into nlcom, which is consistent with your results. The more subtle issue is that if one plugs in a local macro or scalar, no matter how it was arrived at, that number is treated as a scalar constant (mathematical sense).

[Kong Chantha]

Dear professor,
how about nlcom code for calculating the speed of adjustment coefficient (ADJ) in ARDL-ecm? Thank