"PJ Hooker" wrote:
Csmithers
Here are the monetary inputs.
1-Jan-12 (130,000)
31-Jan-12 6,685
29-Feb-12 13,133
31-Mar-12 19,262
30-Apr-12 24,986
31-May-12 30,386
30-Jun-12 35,549
31-Jul-12 40,478
31-Aug-12 45,168
30-Sep-12 49,682
31-Oct-12 54,075
30-Nov-12 58,357
31-Dec-12 62,535
[....]
Well you made a common error in selection of the date
for cash flow at time period t=0 thus your IRR and XIRR
numbers mismatched

The date for time period t = 0 would be 12/31/2011 rather
than 1/1/2012

There is no "common error" if those are the actual dates of the cash flows.

But perhaps the point you intended to make is: in using the IRR with those
cash flows, we must assume the first cash flow occurs on 31-Dec-2011 (i.e.
all transactions occur at regular intervals).

And that will contribute to differences between the results of the Excel
XIRR and IRR functions.

So for an apples-to-apples comparison, we should fudge the date of the first
cash flow in the XIRR parameters.



"PJ Hooker" wrote:
Once you make this correction the XIRR would equal 633.07%
which is the same as the annualized IRR you calculated

Excel XIRR returns about 632.84%, whereas Excel IRR returns about 18.0617%,
which is about 633.35% when annualized by compounding.

(I wonder if the difference between PJHooker's 633.07% and Excel XIRR's
632.84% is due to the fact that Excel XIRR always uses 365 days for a year,
whereas 2012 actually has 366 days. But when I make that substitution in my
own NPV formulation, I get about 636.85%, not 633.07%; and using 633.07%
results in a large error in the NPV.)

In any case, they still are not "the same" for the other reasons that I
provided, primarily differences in cash flow frequencies (the fact that each
monthly cash flows is not the same number of days apart). There is no way
to overcome that primary difference.

But I must admit: I am surprised by the difference that 1 day makes.