04.01.2013 Views

Springer - Read

Springer - Read

Springer - Read

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

6.3 Unit Roots in Time Series Models 193<br />

Example 6.2.2 The lake data<br />

6.3 Unit Roots in Time Series Models<br />

then reoptimize, we obtain the model,<br />

1 − .270B − .224B 5 − .149B 8 + .099B 11 + .353B 12 Xt Zt,<br />

with {Zt} ∼WN(0, 0.0138) and AICC value −172.49.<br />

In order to check more general ARMA(p, q) models, select the option Model><br />

Estimation>Autofit and specify the minimum and maximum values of p and<br />

q to be zero and 15, respectively. (The sample ACF and PACF suggest that these<br />

limits should be more than adequate to include the minimum AICC model.) In a<br />

few minutes (depending on the speed of your computer) the program selects an<br />

ARMA(1,12) model with AICC value −172.74, which is slightly better than the<br />

subset AR(12) model just found. Inspection of the estimated standard deviations of<br />

the MA coefficients at lags 1, 3, 4, 6, 7, 9, and 11 suggests setting them equal to zero<br />

and reestimating the values of the remaining coefficients. If we do this by clicking on<br />

the Constrain optimization button in the Maximum Likelihood Estimation<br />

dialog box, setting the required coefficients to zero and then reoptimizing, we obtain<br />

the model,<br />

(1 − .286B)Xt 1 + .127B 2 + .183B 5 + .177B 8 + .181B 10 − .554B 12 Zt,<br />

with {Zt} ∼WN(0, 0.0120) and AICC value −184.09.<br />

The subset ARMA(1,12) model easily passes all the goodness of fit tests in the<br />

Statistics>Residual Analysis option. In view of this and its small AICC value,<br />

we accept it as a plausible model for the transformed red wine series.<br />

Let {Yt,t 1,...,99} denote the lake data of Example 1.3.5. We have seen already<br />

in Example 5.2.5 that the ITSM option Model>Estimation>Autofit gives<br />

the minimum-AICC model<br />

Xt − 0.7446Xt−1 Zt + 0.3213Zt−1, {Zt} ∼WN(0, 0.4750),<br />

for the mean-corrected series Xt Yt − 9.0041. The corresponding AICC value is<br />

212.77. Since the model passes all the goodness of fit tests, we accept it as a reasonable<br />

model for the data.<br />

The unit root problem in time series arises when either the autoregressive or movingaverage<br />

polynomial of an ARMA model has a root on or near the unit circle. A<br />

unit root in either of these polynomials has important implications for modeling.<br />

For example, a root near 1 of the autoregressive polynomial suggests that the data<br />

should be differenced before fitting an ARMA model, whereas a root near 1 of

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!