forecasting from VAR with jagged edge

[EviewsUser1]

I have three monthly variables that are released with various lags -- one is available up through the current month, one is available up through the previous month and the third is available through two months back. I estimated a BVAR and want to forecast for the period covered by the partially available data -- but I want to use the actual data where it is available and forecast the values where ther actuals are not available. Will Eviews do this sort of forecast?

[EViews Gareth]

Yes.

Once you've made the model object from the VAR, you can click on the Scenarios button, then the Excludes tab and check the box that says "For all other endogenous - Exclude for periods where actuals exist". Then hit OK, click the Solve button and set the solution sample to be the period you want to forecast over (starting from the earliest period).