SH,

If you would simply calculate the standard deviation for the last 2 years using dates, capture out to 2SD, and then project inside that curve for this year, you could probably tighten up your model. You do recall how to do that from Econ/Stats, don't you? wink