Hello everybody,

I'm trying to forecast values for the strategy of companies. I have a panel dataset for companies from 2000 - 2015, but I only need the forecasted values for 2005 - 2015. What I would like to do is take the periods 2000 - 2004, apply exponential smoothing and receive the forecasted value for 2005, then start over for the periods of 2001 - 2005 and forecast 2006, and so on. In the end I only need the last forecasted value for each of the time windows one after the other as one variable. Does anybody have any suggestions of how to do that?
I thought about using

tssmooth exponential double smooth_`var'_2005=`var' if inrange(year,2000,2005)

and do this for all periods, but I'm not sure if it would do what I expect it to do and if there is an option with less manual effort? The rolling command wouldn't really help in this case because it would always "overwrite" the end value I'm interested in, no?

Thank you so much for your help in advance