Thank you, this does it. I had to add the following statement: data want; set want; if ste_2_year=ste_1_year then ste_2_year=.; if ste_3_year=ste_2_year then ste_3_year=.; if ste_4_year=ste_3_year then ste_4_year=.; if ste_5_year=ste_4_year then ste_5_year=.; if ste_1_year=. then ste_2_year=.; if ste_2_year=. then ste_3_year=.; if ste_3_year=. then ste_4_year=.; if ste_4_year=. then ste_5_year=.; run; This keeps data that doesn't have the required length of times from returning identical values for 1 year and 2 year sd's. ... View more

I also usually add a unique statement after the select to get just one line for company, year, and standard deviation. ... View more

I added some sample data to the original post. ... View more

Reeza, Thank you for the response. I indeed have access to the expand procedure. I am not sure how to use it in this case. I don't really want to collapse my existing time series. I want to be able to group by more than one year. The problem I encounter is that my multi-year periods overlap. For example 2001 is part of the two year sequence of 2000-2001 and 2001-2002. ... View more