Pretesting For Multi-Step Ahead Forecasts with STAR Models

[wert12]

Dear Tom;

Here is the code below to replicate the Pretesting For Multi-Step Ahead Forecasts with STAR Models, International Journal of Forecasting 31(2), 2015. pp. 473-87. I downloaded it from http://www.time-series.net/time-series_papers. But I got

"Null Hypothesis : The Following Coefficients Are Zero
DY2 Lag(s) 12
DY3 Lag(s) 12
DY4 Lag(s) 12
F(3,325)= 1.07251 with Significance Level 0.36085382

## SX11. Identifier START is Not Recognizable. Incorrect Option Field or Parameter Order?
>>>>do cap_T = start,<<<<
If the name isn't mistyped, it's possible that you have a poorly formatted instruction
Common errors are
* a space before the ( in an option field
* a missing space before = in a SET or FRML
* a missing $ at the end of a long line which continues to the next"

How can I solve it?

[TomDoan]

Note that we have had nothing to do with writing those programs. There's something clearly very wrong with that---not only is START not defined, but even if it were defined

do cap_T = start, end
box(noprint,define=ar4,constant,diffs=1,ar=4) y start cap_T

doesn't make any sense, since that would be estimating a model with one data point when CAP_T=START. I think START is supposed to be the first entry in the forecast evaluation window, and the BOXJENK instruction is supposed to be estimated over * CAP_T rather than START CAP_T.