Deperately seeking rolling forecast help

[EViews Esther]

Please upload the program with your adjustment.

[Mo']

Thank you for the prompt response.

Attached is the program (supposedly) adjusted to carry out a 5-step ahead forecast also with the corresponding workfile.

Regards,

[EViews Esther]

Code: Select all

' 5-period-ahead forecast %5pers = @otod(@dtoo(%start)+!i+!window-1) 'start point %5pere = @otod(@dtoo(%start)+!i+!window+4) 'end point if {%end} < {%5pere} then 'check whether the forecast end point is greater than the workfile end point return endif

Please note that the string %pere should be renamed to be %5pere.

[Mo']

Thanks again.
I'd try it out as soon as I have access to Eviews7.

[EViews Glenn]

As the error message suggests, the following code snippet isn't valid

Code: Select all

if {%end} < {%5pere} then 'check whether the forecast end point is greater than the workfile end point return endif

since the keywods "@FIRST" and "@LAST" are only recognized in commands to set samples. What you'll have to do is to set up your window loop so that it knows the number of times to be calculated rather than trying to test for a condition inside the loop. There are a number of workfile functions which will tell you how many observations in the workfile, etc. that should be enough for you to find the correct number. I think some of the rolling examples that we provide use tehse functions.

[Mo']

Eviews Glenn,

I am sorry to disturb but I am a bit confused by your suggestion. Please could you expansiate as I have made the corrections suggested by Eviews Esther but I still end up getting the same error.

Regards,

Mo'

[wlai4]

Hi all,

i am very new to eviews and i am trying to do a rolling window one step ahead forecast using a arma23 model and then a garch11 model.

in the LOOP program im not too sure how to specify these equations. i see in the discussion that the forecast can be done by simply opening the equation in eviews. So does that mean that the line of code below allows you to forecast whichever model you specify. Further to this im not sure i understand what you mean by opening the equation and running the program

[%cmnd = _this.@command 'EDIT: get the equation specification]

Sorry, i a real beginner with eviews so some of the terminology is to really new to me and im not too sure what it means and it would be tremendous help if someone could clarify this for me.

[wlai4]

Hi,

i have played around with the program lines but i dont think i am getting the correct results. I have attached a file of the program i used to run the one step ahead rolling window forecast with eq.01 been a ARMA23 model. Basically my in sample is 1388 observations from 1/05/2007 6/22/2012 and my out sample is 50 observations from 6/25/2012 ending on 8/31/2012. (total observations is 1438)

Please advise where i have gone wrong which im sure there will be a few as this is my first time programming in eviews.

your help will be much appreciated.

thank you