I have been using the Quicken Internal Rate of Return as a quick and
easy judge of investment performance for my portfolio. I actually
have recalculated it for one account in Excel in order to understand
the inner workings. There is an obvious skewing when you don't own a
security for more than one year but I wonder how other people analyze
their investment performance.
I'm not sure I understand the the statement about the "obvious
skewing" of the calculation in those cases where the security is owned
for less than a year. What do you mean by that?
As a test, I purchased TEST STOCK A on 1/1/2010 for $5/sh, commission
free. Then I informed Quicken that TEST STOCK A was worth $7.50 on
6/30/2010. Running an Investment Performance report for the stock for
the period 1/1/2010 to 6/30/2010 resulted in an IRR of 127.55%, which
is exactly correct per my own spreadsheet calculation. If I change
the period to 1/1/2010 to 12/30/2010, leaving the price at $7.50,
Quicken comes up with an IRR of 50.30% vs. my calculation of 50.334%
which seems pretty darn close and may come down to how each program
decides it's "close enough" to a NPV of $0 and stops its calculation.
It would seem that what you have done is verify the internal rate of
For IRR, it has an inherent assumption that you have held the stock
for one year. I think if you look at a stock that doubled over a year
vs a stock that you bought a month ago and it doubled, it will not
compare them accurately. I could be wrong. I will check later but I
was looking to see if anyone used a different method. You seem to be
using the same but in an excel spreadsheet. No?
I thought you were were stating that there was a problem with
Quicken's calculation, but what you're really facing is the inherent
limitation of the IRR calculation and how it's presented.
The IRR calculation in use by Quicken calculates a return that's
stated as an annual percentage rate, irrespective of the time the
stock is held, and that convention of stating returns as "annual" is
very common throughout the financial community. Quicken could just as
easily calculate a return stated as a monthly percentage rate or a
weekly percentage rate or any other period. It's simply that we are
accustomed to having rates stated as annual percentage rates as a way
to compare different financial products.
For example, a bank might offer a 28-day CD with a rate of 1.2 %
compounded daily, but it will also disclose an Annual Percentage Rate,
even though the CD isn't going to be around for a year. If a
different bank offered a 28-day CD with a different rate and a
different compounding period the APR (IRR) would give us an easy
metric to compare the two CD's.
The APR (IRR) calculation is less useful in the case of a 1-month CD
with an APR of 12% vs. a 1-year CD with an APR of 8%. IF we can roll
over that 1-month CD 11 times at the same APR THEN we'll come out
ahead of that 1-year CD with an APR of 8%, but will we be able to do
that? The IRR calculation can't answer that question.
The math behind the IRR calculation assumes that cash returns from an
investment are reinvested at the calculated IRR, and that assumption
might or might not be true. Say you have a stock that pays quarterly
dividends but the stock price is the same as it was a year ago and
today you calculate a 1-year return of 10%. You sure as heck
*haven't* invested those dividends at 10% in this interest rate
environment. On the other hand, say you have another stock that
doesn't pay any dividend but its 1-year return, all stock price
appreciation, is also 10%. Which is the superior performing stock?
The IRR calculation won't tell you that; it all comes down to how you
personally value those dividends. If those dividends are what kept
you in groceries for the last year, then that dividend-paying stock
would be more valuable to you, even though the two stocks have the
Going back to your "if you look at a stock that doubled over a year vs
a stock that you bought a month ago and it doubled, it will not
compare them accurately" statement, Quicken compares them perfectly
accurately (mathematically) *within the limitations and assumptions of
the IRR calculation.* There's just no one perfect metric when it
comes to these things.
Quicken's IRR calculation has been buggy for years. Unless they have
fixed it, don't make any decisions based on the numbers it reports.
One of the biggest problems was that you had to know the security
price on the first day of the period and also one day after the end of
the period. So January 1, 2009 and January 1, 2010 must be known.
Partial years required the same numbers. Also, many of the %-return
numbers Quicken reports use the same IRR calculation and thus ae
buggy, too. Caveat emptor.
To an extend the MIRR - modified IRR solves some of the issues with IRR by
allowing finding a rate of return that ensures that proceeds from an invest
ment is reinvested at a different rate than the rate at which costs are fin
MIRR itself has a limitation as it assumes that investment is held till its
maturity however that may not always be the case. The HRR - horizon rate o
f return solves the issues with both IRR and MIRR by allowing you to find t
he return up till a horizon that lies anywhere from commencement of investm
ent till its maturity thus