proc sql;
create table esgprice37 as
select a.* , b.*
from CRSPnew a
inner join esgtest4 b
on a.cusip = b.cusip8 and b.DATADATE13 - a.DATE >=0
having (b.DATADATE37 - a.DATE)=min(b.DATADATE37 - a.DATE);
quit;
oh... my bad.
I need the closest 'next' day.
I will provide examples.
By the way, using the above code only provided outcomes that two dates are exactly matched (datadate37=date).
For example, when the DATADATE37 is 2012/03/28, my code does not match this date with 2012/03/31 from CRSP Monthly(I use monthly data now for convenience. the data I attach is also monthly data).
Rather, my code deletes all kinds of this unmatched dates and output the exactly matched dates only... (DATADATE37 = date).
The intention of the code below is not to delete unmatched data. But my outcome is not like that.
having (b.DATADATE37 - a.DATE)=min(b.DATADATE37 - a.DATE);
Please have a look on my data files and try the first above code and please look through the outcome. The output data missed almost half of the original observations. I referred to other posts in the community and the WRDS's article. BUT theire codes are the same as mine I used.
