Not sure what the 2006 data has to do with this. but...

Name the cell in column B adjacent to 1/1/06 as BOY_2006, and adjacent to
1/1/07 as BOY_2007. Ditto for 1/1/08. (BOY means Beginning Of Year)

Now, your YTD sum for any month in 2008 will be

=sum(offset(boy_2008,0,0,month(Ax),1))
where x in the Ax is the row in which the current month's date appears in
column A.

if you want to divide by the earlier year's YTD, then the formula above
divided by

sum(offset(boy_2007,0,0,month(Ax),1))

"Marena" wrote:

I have a series of monthly dates formatted as 01/01/2006, 02/01/2006, etc. to
most recent data point in column A. In column B I have data. In column C I
would like to calculate a rolling YTD number that is the sum of the YTD data
from this year (based on the latest data point/date that is populated in
column B) and divide this by the sum of the previous year's similar YTD sum.
(i.e. if the latest data point is for 03/01/2008, the calculation would be
(03/01/08 data +02/01/08 data + 01/01/08 data)/ (03/01/07 data +02/01/07 data
+01/01/07 data). I have been working with OFFSET and COUNTA, but I am
definitely not there. Any help would be greatly appreciated.