0
$\begingroup$

This is the autocorrelation case.

$y_{t}=X_{t}B+u_{t}$

where $u_{t}=\rho u_{t-1}+e_{t},$ $e_{t}$ is iid

From this autocorrelated disturbances,

I might be able to say

$y_{t}=\gamma y_{t-1}+w_{t}$, i.e., ordinal AR(1) model

Is this right?

$\endgroup$
0
$\begingroup$

Yes ....

If you have

yt=XtB+ut where ut=ρut−1+et, et is iid

and you clear fractions then you have a model of the form

y(t)= γy(t−1) + X_{t}-[1-γ]X_{t-1} + e_t$, and $e_t$ is iid.

where $w_t= X_{t}-[1-γ]X_{t-1} + e_t$, and $e_t$ is iid.

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ This is being automatically flagged as low quality, probably because it is so short. Could you expand it a little? $\endgroup$ – gung - Reinstate Monica Dec 3 '15 at 14:43
  • $\begingroup$ This does not look like an AR(1) model to me; should it? $\endgroup$ – Richard Hardy Dec 3 '15 at 18:07
  • $\begingroup$ No ....it should not $\endgroup$ – IrishStat Dec 3 '15 at 21:00
  • $\begingroup$ So why is your answer a "yes"? $\endgroup$ – Richard Hardy Dec 3 '15 at 22:20
  • $\begingroup$ YES if you consider w(t) to be not iid $\endgroup$ – IrishStat Dec 3 '15 at 23:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.