Page 1 of 1

Estimateing a Model with AR errors

Posted: Thu Oct 14, 2010 10:04 pm
by jag_211085
Hi,

How do I estimate u[t] = rho[1]*u[t-1] + rho[2]*u[t-2] +rho[3]*u[t-3] +rho[4]*u[t-4] where u[t] is the residual series from a corresponding time series equation.

Is it just to put AR(1)..AR(4) in the original time series equation?

Thanks.

Re: Estimateing a Model with AR errors

Posted: Fri Oct 15, 2010 3:52 am
by trubador
By corresponding/original time series equation if you mean something like: Yt = b0 + b1*Xt + Ut, then the answer is "Yes". You can simply estimate it as follows:

Code: Select all

Y c X ar(1) ar(2) ar(3) ar(4)