Problem in forecasting with dynamic model

[UmBert]

I have a problem in forecasting with my dynamic model (dyn_fin) in the workfile. I have estimated my model with the dataset from 1978q1 to 2014q2 and now I want to estimate the "virtual future" from 2014q3 to 2016q2 for the variable PCI. When I try to do this I get only one point of forecast due to the adjustment of the sample (static and dynamic forecast as well). I can't understand why there is an adjustment because I don't have any missing data and with another similar workfile I do not have this problem.

Am I making a mistake?
Can someone help me to find the way to get my forecast with the right sample?

[EViews Gareth]

You don't have data for all of your regressors over the forecast period.

[UmBert]

No, this is not true. My dataset is complete until 2016q2.

Look at the attached screenshot.

[EViews Gareth]

Looks at the regressors.

[UmBert]

Can you suggest me the path to check for this because I don't know where to go.

[EViews Gareth]

For each regressor that you specified in the equation, look at the series and check if it has data

[UmBert]

I have checked each regressor and for each series I have the data until 2016q2 so this cannot be the problem. There are not any missing data in the whole series.

[EViews Gareth]

I suggest you look harder.

2016-08-12_080841.png (89.89 KiB) Viewed 6791 times

[UmBert]

OK...but in order to have the the missing data, I have to make a regression. If I made the regression, the forecast would be useless. This restriction on EView for what concerns forecasting makes the application totally useless...if you have the data, why forecasting?

[startz]

This has nothing to do with EViews. It is a fundamental issue with forecasting. You either need a model in which you know the future value of the right hand side variables or you need to build an auxiliary model to forecast those right hand side variables.