On Mar 28, 12:20 pm, "joeu2004" wrote:
On Mar 28, 7:38 am, wrote:

So here is what I'm looking at...
December - 100 - N/A
January - 200 - 100% change (1.0)
February - 300 - 50% change (.5)
March - 100 - 66% change (.66)

I can tell what the percentage change is from month to month, but I'm
not sure what the best way is to figure out the average percentage
change, overall. I can't add up the decimal versions of each
percentage difference and then divide by the number of months to get
some kind of overall average, right? In this case, that would turn
out to be an average variance of 72%.

Actually, 28%. The last change -- "66%" -- is really -67%.

The average of the percentages __might__ make sense in some contexts.
It answers the question: "what is the expected percentage change
month-to-month?". That is indeed the average of the percentages.

But I suspect you are more interested in the geometric mean -- that
is, the net change over time. In the example above, you went from 100
to 100 over 3 months. You expect a net change of 0%. Right?

One way to compute that is the following array formula (commit with
ctrl-shift-Enter, not Enter):

=geomean(1+B2:B4) - 1

Alternatively, the simple formula (not an array formula):

=(B4/B1)^(1/3) - 1

Be sure to format the cell as Percentage, if that is what you want to
see.

Thanks for that explanation. What operation does the ^ represent in
Excel?