You are on page 1of 6

Modified Internal Rate of Return Revisited < Thought | SumProduct are ... https://www.sumproduct.com/thought/modified-internal-rate-of-return-r...

Adding Value for the Times

 LOGIN (THOUGHT/MODIFIED-INTERNAL-RATE-OF-RETURN-REVISITED#) 

Modi ed Internal Rate of Return


Revisited

Home (home)  Thought (thought)  Modi ed Internal Rate of Return Revisited

Let’s Be Frank: Incensed by MIRR..?


In this article we return once more to the Modi ed Internal Rate of Return and respond to one of the most
common queries received. By Liam Bastick, Director with SumProduct Pty Ltd.

Query
Further to your previous articles on Internal Rate of Return (IRR) and Modi ed Internal Rate of Return (MIRR), I
recognise that MIRR may be a more appropriate measure of an investment’s attractiveness. However, the Excel
function MIRR(Cash¡ows,Annual_Finance_Rate,Annual_Reinvestment_Rate) only works where periods are of
equal length. How can I calculate the MIRR for cash¡ows relating to unequal period lengths?

Advice
If I had a dollar for every time this question was asked I might have some money in my wallet.

I think I could write a book on the problems with calculating IRRs in Excel and how MIRR might be a more valid
approach (see Irreverent IRR (thought/irreverent-irr) and Many Happy Returns: Modi ed Internal Rate of Return
(MIRR) (thought/many-happy-returns-modi ed-internal-rate-of-return-mirr)). I am unsure whether anyone would
read it though.

If you are not familiar with these issues, I would strongly recommend you read the aforementioned articles
before continuing.

To summarise, IRR is problematic for several reasons:

There can be more than one discount rate which makes the Net Present Values of the underlying cash
¡ows equal to zero;

IRR assumes a symmetry implicit between the reinvestment rate (i.e. the return surplus cash ¡ows may
accrue on deposit) and the interest rate (i.e. the rate charged for sums borrowed). This assumption only
holds true in Cloud Cuckoo Land; and

Excel’s functions IRR and XIRR can be limited; the former only works for cash¡ows reported in periods of
equal length, whereas the latter will not always calculate correctly given the logic of the underlying
algorithm. Goal Seek may often be a better approach to con rm IRRs in a spreadsheet.

The Modi ed Internal Rate of Return (MIRR) is de ned as follows:

where:

1 de 6 05/11/2016 18:48
Modified Internal Rate of Return Revisited < Thought | SumProduct are ... https://www.sumproduct.com/thought/modified-internal-rate-of-return-r...

NPV() is the Excel NPV function

rrate is the reinvestment rate

frate is the nance rate

values[positive] is the positive values in the array only

values[negative] is the negative values in the array only

n is the number of periods

This formula will always give the same value regardless of the number of changes of sign in the cash ¡ow. It
also takes into account the disparity between reinvestment and nance rates.

The attached Excel le ( leadmin/user_upload/SP-MIRR-Revised-Example2.xls) walks through how MIRR


works (essentially all negative sums are discounted to the start of the project at the nance rate, whilst all
surplus cash ¡ows are compounded to the nal period, gaining interest at the reinvestment rate).

Excel has a function that will compute the MIRR for cash¡ows occurring in periods of equal length:

( leadmin/_processed_/csm_Image-02-MIRR-Illustration_3b4f754d51.png)

MIRR Illustration

Using the formula, MIRR is arguably quicker to calculate than IRR, more objective (only one solution) and takes
into account the differing rates implicit in the cash ¡ows. The MIRR is usually lower than the IRR (as in our
example above, assuming the reinvestment rate will typically be lower than the nance rate).

The MIRR is often seen as a nancial measure of an investment’s attractiveness. It is used frequently in capital
budgeting to rank alternative investments of similar size (although this may not always be an appropriate
approach: NPV or NPV per $ invested [the so-called “bang for buck” key factor analysis] may be more suitable
metrics).

If negative cash¡ows are rolled up to the rst period and positive ¡ows are carried forward / reinvested to the
nal period (as explained above), IRR and MIRR will equate:

( leadmin/_processed_/csm_Image-03-Equating-IRR-and-MIRR_bf4b98b1d4.png)

Equating IRR and MIRR

Reverting to our reader’s question, how do you do this when the cash¡ow periods are not of equal duration?

The answer is surprisingly simple if the roll-back / roll-forward technique used in the Excel le is employed:

2 de 6 05/11/2016 18:48
Modified Internal Rate of Return Revisited < Thought | SumProduct are ... https://www.sumproduct.com/thought/modified-internal-rate-of-return-r...

Cash out¡ows are discounted back to the rst (minimum) date, discounted on a daily basis using the
nance / interest rate;and

Cash in¡ows are reinvested to the nal (maximum) date, compounded on a daily basis using the
reinvestment rate.

( leadmin/_processed_/csm_Image-04-MIRR-Illustration-2_8d6cae8c09.png)

Revised MIRR Illustration

3 de 6 05/11/2016 18:48
Modified Internal Rate of Return Revisited < Thought | SumProduct are ... https://www.sumproduct.com/thought/modified-internal-rate-of-return-r...

In this example (replicated in the attached Excel le ( leadmin/user_upload/SP-MIRR-Revised-Example.xls)), all


periods are of equal duration (365 days) so that the MIRR may be reconciled to the original periodic example
( rst illustration).

However, once you are satis ed these techniques work as intended, dates can quickly be modi ed, e.g.

( leadmin/_processed_/csm_Image-05-XMIRR-Illustration_2e5bab7be8.png)

So-called “XMIRR” Illustration

4 de 6 05/11/2016 18:48
Modified Internal Rate of Return Revisited < Thought | SumProduct are ... https://www.sumproduct.com/thought/modified-internal-rate-of-return-r...

If XNPV calculates NPVs on an aperiodic basis and XIRR calculates IRRs on an aperiodic basis, then it seems
reasonable to call this XMIRR!

In fact, in these last two illustrations four approaches have been compared:

1. XIRR: uses a methodology similar to the IRR approach of Irreverent IRR (thought/irreverent-irr) using the
aggregated cash ¡ows. With only two cash¡ows and only one change of sign, only one IRR is possible.
2. Compound Approach: Using an Excel function is not necessary if the aggregated cash ¡ows have been
computed. From the two cash ¡ows, dividing the aggregated in¡ows by the positive value of the aggregated
out¡ows gives the growth over the period of time. By using a simple compounding formula, this can be
readily converted to an equivalent annual rate (i.e. the MIRR).
3. Array Approach: This is a variant often found on the web. An array function is used (see Array of Light
(thought/array-of-light) for more information). This formula simply uses the original cash¡ows and performs
all computations in one go:

{=POWER((SUM(IF(Values>0,Values*(POWER(1+Annual_Reinvestment_Rate,
(MAX(Dates)-Dates)/Days_in_Year)),0)))/(SUM(IF(Values<0,Values
/(POWER(1+Annual_Finance_Rate,(Dates-MIN(Dates))/Days_in_Year)),0)))*-
1,Days_in_Year/(MAX(Dates)-MIN(Dates)))-1}

4. User De ned Function: Thanks to David Hager for this one – here a Function has been built using VBA
code that again only requires the initial cash¡ows and dates (macros must be enabled for this to work):

VBA Code for XMIRR User De ned Function

5 de 6 05/11/2016 18:48
Modified Internal Rate of Return Revisited < Thought | SumProduct are ... https://www.sumproduct.com/thought/modified-internal-rate-of-return-r...

The attached Excel le ( leadmin/user_upload/SP-MIRR-Revised-Example1.xls) contains this code in the VBA


module.

Personally, I would use the second approach. I think it makes sense to step out the computation so that others
may follow the logic and it leads to by far the simplest formula.

Word to the Wise


Do remember that if your objective is to maximise shareholder value then Net Present Value computations
should be undertaken ahead of MIRR calculations. If an investment utilises a scarce resource, then perhaps key
factor analysis (i.e. NPV divided by the scarce resource) may be a more appropriate decision tool.

Where MIRR computations are required for aperiodic durations, eagle-eyed readers may have spotted that the
attached Excel le assumes there are always 365 days in a year, which is not the case in reality. Before people
try to build computations to allow for this additional complexity, do bear in mind the likely accuracy of your
cash¡ow forecasts!

 First Name Last Name Email Address SIGN UP

Newsletter

CONTACT US TODAY (/CONTACT)

© 2009-2016 SumProduct Pty Ltd. | Registered Of ce (registered-of ce) | Sitemap (sitemap)

SumProduct are experts in Excel Training. Providing Financial Modelling, Strategic Data Modelling, Model
Auditing, Planning & Strategy and Training Courses.

Graphic Design by  Mixing Bowl (http://mixing-bowl.com.au)— Web Design & SEO by  Potent Web
(http://www.potent.com.au/)

6 de 6 05/11/2016 18:48

You might also like