Using monthly series to forecast quarterly data

[miorinnovo]

Hi all,

I'd like to try to forecast a quarterly series using a monthly data set. Essentially I'd like to have one explanatory variable being the monthly series with one month left in the quarter, another being the same series but with two months left in the quarter, etc. Can anyone tell me how I can set this up?

thanks

[miorinnovo]

I could do this by automatically deleting the third month in every quarter and replacing them with the fourth month, and another series deleting months 2 and 3 and replacing them with the first month. Does anyone know how to do that?

[trubador]

I am really not sure if I understand it correctly. But, I'll give it a try:
Suppose, the quarterly and monthly variables are called y_q and x_m, respectively. And let's assume that they are both stored in the same workfile with page names "quarterly" and "monthly". Since the forecasting exercise will take place in the quarterly frequency, you should import the monthly variable:

Code: Select all

pageselect quarterly 'select the quarterly page
copy(c=f) monthly\x_m x_q 'copy and paste the first values of the monthly variable and rename
equation eq.ls y_q c x_q x_q(-1) 'you do not have to generate the second variable