To XIRR or Not To XIRR

[Jonathan]

What IRR is the investment community most familiar with? I want to
publish an effective annual rate for my investors. We give monthly
distributions at the end of the month. And from what I understand I
have at least two options:

1) XIRR()
2) ((1+IRR())^12)-1

Which one of these calculations (or other) is the most commonly
accepted.

Thank you,
Jonathan

 

[Fred Smith]

The choice is simple.

If the cash flows are regular (happen at the same interval, say monthly, for the
entire period), use IRR
If the cash flows are irregular (which almost always happens in the investment
business, as people put money in and take it out on their own schedule), use
XIRR.

 

[joeu2004]

What IRR is the investment community most familiar with?

The "investment community" is very large. What do you mean?

I want to publish an effective annual rate for my investors. We
give monthly distributions at the end of the month. And from
what I understand I have at least two options:
1) XIRR()
2) ((1+IRR())^12)-1

I presume you realize that for regularly-spaced cash flows, there
is very little difference. Frankly, I doubt that your investors would
notice the difference -- unless, of course, you tell them.

And there is a third option: 12*IRR(). I find this prevalent in
academic papers. Personally, I think it is wrong.

Which one of these calculations (or other) is the most commonly
accepted.

It really depends on what you mean by "commonly accepted".

Since XIRR() is available only in an Excel add-in (albeit it bundled
with Excel) -- and I don't know since when -- I think it is a safe bet
that for regularly-spaced cash flows, IRR() is the more "familiar"
and "commonly used" function, probably with monthly periods in
your case.

But whether the "investment community" compounds the periodic
IRR() or simply multiplies it by the number of periods, I cannot say.
Much of what I read simply multiplies :-(.

PS: Although XIRR() might seem more accurate, I think it
misrepresents the real daily compounded rate from an investment
perspective. I think that should only take trade days into account
-- on average, 252 per year and 21 per month. So arguably, IRR()
is the better compromise anyway, at least for regularly-spaced
cash flows.

 

[Jonathan]

Thanks guys,

I think I am going to go with option 2: (1+IRR()^12)-1, since our
distributions are consistantly monthly.

In much of my financial dealings to date I have too seen IRR()*12 --
but reading a handful of Fred's posts have justly convinced me that it
is simply wrong. Even though, I know I am going to have to teach
FIN101 to many of my investors ;-)

-- jonathan