EViews.com

HI everybody,

I have a large database (Xt) with a lot of monthly time series.
I have estimated various forecasting model with this database.

Currently, I would like to evaluate all of the models by transforming my database in a pseudo real time database (I would like to replicate the data availability situation that is faced in real time application of the models). All series are available from january 1981 to june 2006 (but for some variables, I 'dont have information for june 2006) and I would like to cut my sample for example by 10 months but I have to take into account that some individual series are not published with the same delays => I WOULD LIKE TO ELIMINATE the original observations from the original database (Xt) from the end of sample to the (t-h month , here 10 months)? Do you have some solutions?

Thanks a lot

I don't follow what you're asking.

I don't follow what you're asking.

I think he wants to do pseudo out of sample forecasting but the method he proposes to do so seems strange to me. Rather than messing around with the database you can just set up your equations to forecast recursively.