Forecast standart error

[irulishka]

Hello all EViews users
I have a big problem with using of uncertainty coefficient in forecast SE Calc in model y x ar(1).
I have a formula SE_forec_un(t)=(SE_forec^2+xt*covar*xt')^0.5, but this formula works correctly only in regression like y x.

May be you can help me and give me some information about my mistake in formula or advice me books or magazines, there I can found some information about whis...

Thanks in advance, Ira