Page 1 of 1

adjustment parametr "Phi" ECM

Posted: Wed Jun 16, 2010 9:11 am
by danreg87
my academic project for operation research is to make a Engle-Granger model.

i've got problem with adjustment parametr "phi" ( by ECM and VECM )( i dont know is it the official name for it - my PhD uses it so let me use it even it's wrong in English )

general ECM:
d(logYt) = alfa d(logXt) - Phi* Z(t-1)

when phi is beetween -1 and 0 the interpretation is obvious - it shows as how much instability ( in % ) from period (t-1) will be remove in period t.

But what when PHI is 0,45 - my PhD told me it's not a mistake, but the interpretation will be diffrent and to be frank i've no idea how should it be.

Please help ;)