Remember that the model is
APt = bp1Et APt+1 + bp1 APt —1 + bp2Xt + Zpt, which can be rewritten as
n = jEtnt+1 + SxH + vpt.
The model is usually estimated by means of instrumental variables, using the ‘errors in variables’ method (evm)—where expected values are replaced by
actual values and the expectational errors:
nt = Y^t+i + 5xt + vpt – YVt+i – (A.22)
The implications of estimating the model by means of the ‘errors in variables’ method is to induce moving average errors. Following Blake (1991), this can be readily seen using the expectational errors as follows.
1. Lead (A.15) one period and subtract the expectation to find the RE error:
3. Finally, re-express in terms of original variables, again using Apt = n + aApt-i:
Apt – aiApt-i = (— ) (Apt+i – aiApt) + ( ) xt + ( f— ) £pt
W bpia2) biia. p
where we have exploited the two well-known relationships between the roots:
So even though the original model has white noise errors, the estimated model will have first-order moving average errors.