Frequency conversion and aggregation of incomplete series except in current year

[jramm]

Hi all,

I'm trying to aggregate monthly time series into quarterly and annual frequencies. Some monthly observations have missing values, which can be solved by using fetch(c=sn) or fetch(c=an), however, this creates a new problem which is that the annual value for 2020 get calculated (since a few months of data already exists in 2020). So, the question is, is there a way to aggregate incomplete series with the exception of the current year?

Thank you so much for the help,
Joachim

[EViews Gareth]

Not directly, but the obvious solution is to NA out the last year after the fetch

[jramm]

Thank you for the reply.

If I understand you correctly, you mean to NA 2020 after the fetch? I believe this might create a problem down the road since various time series will have complete monthly data for the entire 2020 at different times (i.e. some monthly series are released sooner than others). When series A have 12 months of data 2020 and and series B only have 11 months of data, the annual value for series A will still be NA even though I exists?

The data gets dumped from eviews into excel so I was thinking of creating two different workfiles in eviews, one with monthly data up until 2019 (w/ aggregation of incomplete series) and one workfile for monthly data from 2020 (w/ no aggregation of incomplete series), and then dump the data from these two workfiles into excel (and do the same for the quarterly aggregation and data). This "solution" is not ideal since I believe the quarterly workfiles would then have to be updated each quarter and the annual workfiles once per year? This sounds like quite a bit of work so any other suggestions would be greatly appreciated.

Thanks,
Joachim

[EViews Glenn]

Try the "propnas" option which produces NAs for partially observed periods.