BMA4106 Investment and asset management

Lecturer 2 : Equations of Value and the Internal

Rate of Return
Notes Prepared

Gikonyo Kiguta
Mount Kenya University
 Contents

1 Equations of Value and the Internal Rate of Return . . . . . . . . . . . . . . . 5

1.1 Introduction 5
1.1.1 Internal rate of return . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2 The comparison of two investment projects 9
1.3 Different interest rates for lending and borrowing. 12
1.4 Discounted payback period. 12
1.5 Measurement of investment performance 15
1.5.1 Money-weighted rate of return . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.5.2 Time-weighted rate of return . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
1.5.3 Linked internal rate of return . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Articles 21
Books 21
 1. Equations of Value and the Internal Rate of Retur

1.1 Introduction
In economics and accountancy the yield per annum is often referred to as the “internal rate of
return” (IRR) or the “yield to redemption”. The latter term is frequently used when dealing with
fixed interest securities, for which the “running” (or “flat”) yield is also considered.

1.1.1 Internal rate of return

The internal rate of return for an investment project is the effective rate of interest that equates the
present value of income and outgo, ie it makes the net present value of the cashflows equal to zero.

If all the payments for the project were transacted through a bank account that earned inter-
est at the same rate as the internal rate of return, the net proceeds at the end of the project (ie the
accumulated profit) would be zero. A higher internal rate of return indicates a more “profitable”
project.

Example
Consider the following project

Project R
Project R delegates all the development work to outside companies. The estimated cashflows for
Project R are (where brackets indicate expenditure):

Beginning of Year 1(150, 000) (contractors’ fees)

Beginning of Year 2(250, 000) (contractors’ fees)
Beginning of Year 3(250, 000) (contractors’ fees)
End of Year 3 1, 000, 000 (sales)

Find the internal rate of return for Project R.

6 Chapter 1. Equations of Value and the Internal Rate of Return

Solution
We need to find the interest rate i that satisfies the equation of value:

−150 − 250v − 250v2 + 1, 000v3 = 0

We already know that:

At 20% : −150 − 250v − 250v2 + 1, 000v3 = 46.8

since the outgo precedes the income, the value of i must be greater than this. So we need to try a
higher rate:

At 25% : −150 − 250v − 250v2 + 1, 000v3 = 2.00

We can approximate i by extrapolating (linearly) using these two values:

0 − 46.8
i ≈ 20% + × (25% − 20%) = 25.2%
2.00 − 46.8

Question
Consider the following project

Project S Project S carries out all the development work in-house by purchasing the necessary
equipment and using the company’s own staff. The estimated cashflows for Project S are:

Beginning of Year 1(325, 000) (new equipment)

Throughout Year 1 (75, 000) (staff costs)
Throughout Year 2 (90, 000) (staff costs)
Throughout Year 3 (120, 000) (staff costs)
End of Year 3 £ 1,000,000 (sales)
The staff costs can be assumed to be paid uniformly throughout the year.

Find the internal rate of return for Project S and hence determine which project is more favourable
using this criterion.

Example
An investment of 100 returns 70 at the end of year 1 and 70 at the end of year 2. What is the internal
rate of return, r∗ ?
1.1 Introduction 7

Solution.
The sequence of cash-flows is a = (−100, 70, 70). So r∗ satisfies

70 70
−100 + ∗
+ =0
1+r (1 + r∗ )2

Let x = 1/ (1 + r∗ ) . Hence 70x2 + 70x − 100 = 0. We need the root with x > 0 and so x = 0.796
and r∗ = (1/x) − 1 = 0.257. The internal rate of return is 25.7%

Three payments leads to a cubic, and so on. The solution for r∗ will often have to be obtained
numerically, but this is usually easy as g(r) is monotonic.

Numerically finding the IRR.

Starting values for a numerical search are often required. Here are four possibilities:

- (a) Approximate 1/(1 + r) j = (1 + r)− j by 1 − jr. Then the equation

c1 c2 cn
c0 = + 2
+···+
1 + r (1 + r) (1 + r)n
∑nj=1 c j −c0
gives the following approximation r0 for r∗ : c0 = ∑nj=1 c j − r0 ∑nj=1 jc j or r0 = ∑nj=1 jc j

- (b) Convert to c0 (1 + r)n = c1 (1 + r)n−1 + c2 (1 + r)n−2 + · · · + cn . Then approximate (1 + r) j

by 1 + jr

- (c) Assume all future cash-flows equal the average cash-flow. This means that we approxi-
mate

Time 0 1 2 ... n−1 n

Cash flow −c0 c1 c2 ... cn−1 cn
by

Time 0 1 2 ... n−1 n

Cash flow −c0 x x ... x x
where x = ∑nj=1 ci /n.

This leads to
   
x 1 1 1
c0 = 1− = xa n ,r where a n ,r = 1−
r (1 + r)n r (1 + r)n

and values of a n ,r can be found in actuarial tables. There is further discussion of a n ,r in the next
chapter.

- (d) Suppose the cash flow has the following form:

Time 0 1 2 ... n−1 n

Cash flow −c0 x x ... x cn
There is an interest payment of x for each of the n years plus a final payment of cn − x. Hence the
8 Chapter 1. Equations of Value and the Internal Rate of Return

capital gain is cn − x − c0 and this capital gain can be approximated by a payment of (cn − x − c0 ) /n
every year. So the approximate rate of return is

x + (cn − x − c0 ) /n
i=
c0

Assuming there is a positive capital gain, we will have x/c0 < i < [x + (cn − x − c0 ) /n] /c0 . This
will usually be the best method for cash flows of this form.

Example Consider the following cash-flow sequence

Time 0 1 2 3 4 5 6 7 8 9 10
Cash flow -50 4 4 4 4 4 4 4 4 4 59
Use method (d) described above to find an initial approximation to the internal rate of return,
r∗ .

Solution.
The return per year is 4 + (55 − 50)/10 = 4.5. Hence r∗ ≈ 4.5/50 = 0.09. In fact, r∗ = 0.0866868.

Example
(a) Consider the cash-flow sequence a = (−87, 25, −40, 60, 60) at times 0,1,2,3 and 4. Use the first
3 methods described above to find initial approximations to the internal rate of return, r∗ .
(b) Use the first 3 methods to find the yield on project B in example (2.1c) on page 22 .

Solution.
(a) The first method gives r0 = ∑ c j / ∑ jc j = 18/365 = 0.049. The second method gives r0 =
18/293 = 0.061. For the third method, we have x = 105/4 = 26.25. So we want r with
a 4 ,r = 87/26.25 = 3.312 From actuarial tables, a 4 ,0.07 = 3.387 and a 4 ,0.08 = 3.312; so take
r = 0.075. The actual value of r∗ is 0.056.

(b) The first method gives r0 = ∑ c j / ∑ jc j = 53/1016 = 0.052. The second method gives 53/592 =
0.0895. For the third method, we have x = 148/8 = 18.5. So we want r with a 8 ,r = 95/18.5 =
190/37 = 5.135. From actuarial tables, a 8 ,0.11 = 5.146 and a 8 ,0.115 = 5.0455, so take r = 0.11

The actual value is r = 0.06832. The third method gives a poor approximation because the mean of
18.5 is a poor summary of the sequence (6,6,6,6,6,6,6,106)

Comparison with NPV.

If the interest rate is i, then NPV (i) is the present value of the cash-flows of the project. The
project is profitable if NPV (i) > 0

Suppose r∗ , the yield or IRR exists, and NPV (i) > 0 for i < r∗ and NPV (i) < 0 for i > r∗ . Then
the project is profitable if and only if i < r∗ .

4.5 Problems with the use of the IRR. Consider the cash-flow sequence a = (c0 , c1 , . . . , cn ) . Let
x∗ = 1/ (1 + r∗ ) . Then x∗ is the solution for x of c0 + c1 x + c2 x2 + · · · + cn xn = 0 This equation has
n roots for x. Any root for x which satisfies x > 0 leads to a solution for r with r > −1. There may
be more than one root with r > −1. Hence the internal rate of return cannot be defined. Even if
there is a unique root, it is possible that g(r) is not monotone. It then follows that the property that
 1.2 The comparison of two investment projects 9

NPV is positive for interest rates i on one side of r∗ and negative for interest rates on the other side
of r∗ may not hold. So again it is not sensible to define the IRR.

The main disadvantages of using the IRR are:

• Some cash-flow sequences do not have an IRR: there may be no or multiple solutions for r∗ .
• The IRR measure does not use the value of the currently available interest rates. It is solely an
internal measure—that is why it is called the internal rate of return. In particular, if an investment
accumulates funds for a final payment, then the IRR calculation assumes that the sinking fund earns
interest at the internal rate of return. Similarly, if a loan is required to finance a project, then the
IRR calculation assumes the interest rate on the loan is at the internal rate of return.
• NPV and IRR may lead to different conclusions—an example is given in the next section.
The NPV method is generally preferred.

Example
Consider the following sequence of cash-flows:

Time 0 1 2 3
Cash flow -32 128 -166 70
Find the internal rate of return, r∗ .

Solution.
The equation c0 + c1 x + c2 x2 + · · · + cn xn = 0 gives −32 + 128x − 166x2
+ 70x3 = 0. We can easily
see that x = 1 is a solution. Hence we have 0 = (x − 1) 70x2 − 96x + 32 = (x − 1)(7x − 4)(10x −
8). Hence x = 1, 4/7, or 4/5. Hence r∗ = 0, 0.75, or 0.25. There is no unique value for the IRR.
Also NPV (r) < 0 for r ∈ (0, 0.25) and NPV (r) > 0 for r ∈ (0.25, 0.75)

Example Find the IRR for the following cash-flow.

Time 0 1 2
Cash flow 2 -4 3
Solution.
We have 2 − 4x + 3x2 = 0 and this has no real roots.

1.2 The comparison of two investment projects

Suppose now that an investor is comparing the merits of two possible investments or business
ventures, which we call projects A and B respectively. We assume that the borrowing powers of the
investor are not limited.

There are therefore no restrictions on how much the investor can borrow.

Let NPVA (i) and NPVB (i) denote the respective net present value functions and let iA and iB
denote the yields (which we shall assume to exist). It might be thought that the investor should
always select the project with the higher yield, but this is not invariably the best policy. A better
criterion to use is the profit at time T (the date when the later of the two projects ends) or, equiva-
lently, the net present value, calculated at the rate of interest i1 at which the investor may lend or
borrow money. This is because A is the more profitable venture if:

NPVA (i1 ) > NPVB (i1 )

10 Chapter 1. Equations of Value and the Internal Rate of Return

The fact that iA > iB may not imply that NPVA (i1 ) > NPVB (i1 ) is illustrated in the following
diagram. Although iA is larger than iB , the NPV (i) functions "crossover" at i0 . It follows that
NPVB (i1 ) > NPVA (i1 ) for any i1 < i0 , where i0 is the cross-over rate. There may even be more than
one cross-over point, in which case the range of interest rates for which project A is more profitable
than project B is more complicated.

The following graph show the yields for the two projects

Example Consider the following two cash-flow sequences.

Time 0 1 2 3
Project A Cash flow, a -80 96 1 5
Project B Cash flow, b -80 10 10 90
(a) Find rA∗ , the IRR for project A and rB∗ , the IRR for project B and show that rA∗ > rB∗ . This
suggests that project A is preferable to project B.

(b) Show that NPV(b) > NPV(a) if the interest rate is r = 0.04.

(c) Treating NPVi (a) and NPVi (b) as functions of the interest rate i, plot their graphs for i ∈ (0, 1)

Solution.
(a) For project A, solve −80 + 96x + x2 + 5x3 = 0. This gives 20 + x + x2 (−4 + 5x) = 0. Hence


1.2 The comparison of two investment projects 11

x∗ = 4/5 and
rA∗ = 1/x∗ − 1 = 1/4 For project B, solve −80 + 10x + 10x2 + 90x3 = 0. This gives
10 1 + x + x2 (−8 + 9x) = 0. Hence x∗ = 8/9 and rB∗ = 1/x∗ − 1 = 1/8

(b) If r = 0.04 = 1/25 then NPV(a) = 17.677 and NPV(b) = 18.87 and project B is preferable to
project A!!

(c) This is shown in figure below. The value of i where the two graphs cross is called the cross-over
rate.

Example
An investor is considering whether to invest in either or both of the following loans:
Loan X : For a purchase price of 10, 000, the investor will receive 1, 000 per annum payable
quarterly in arrears for 15 years.

Loan Y: For a purchase price of 11, 000, the investor will receive an income of 605 per annum,
payable annually in arrears for 18 years, and a return of his outlay at the end of this period.

The investor may lend or borrow money at 4% per annum. Would you advise the investor to
invest in either loan, and, if so, which would be the more profitable?

Solution

We first consider loan X :

(4)
NPVX (i) = −10, 000 + 1, 000a 15

(4)
and the yield is found by solving the equation NPVX (i) = 0, or a 15 = 10, which gives iX ≈ 5.88%
 12 Chapter 1. Equations of Value and the Internal Rate of Return
(4)
This is easily checked by calculating a 15 @5.88%

(4) 1 − v15 1 − 1.0588−15

a 15 = =
= 10.00 as required
i(4) 4 1.05881/4 − 1

For loan Y we have:

NPVY (i) = −11, 000 + 605a 18 + 11, 000v18
and the yield (ie the solution of NPVY (i) = 0 ) is iY = 5.5%.

1.3 Different interest rates for lending and borrowing.

Usually, an investor has to pay a higher rate of interest on borrowings (iB ) than he obtains on
investments (iI ) . The size of the differential depends on his credit worthiness, size of the amounts,
time of loan, etc. The concepts of NPV and IRR are then not meaningful.

1.4 Discounted payback period.

For many projects, there is an initial cash outflow followed by a sequence of cash inflows. Let A(t)
denote the accumulated value of the investment at time t; then
Z t
A(t) = ∑ ct j (1 + iB )t−t j + 0
ρ(s) (1 + iB )t−s ds
t j ≤t

Typically A(t) will be initially negative and then become positive. Let t1 denote the smallest value
of t with A(t) ≥ 0. This time t1 is called the discounted payback period or DPP− it is the smallest
time such that the accumulated value of the project becomes positive. Equivalently, the DPP is
the smallest time t such that the NPV of payments up to time t is positive. The value of the DPP
depends on iB but not on iI .

If iB = 0 then the quantity corresponding to t1 is called the payback period. Thus the payback
period is the time when the net cash flow (incomings less outgoings) is positive. It is similar to the
DPP− but it ignores discounting. This means that it ignores the effect of interest. It is therefore an
inferior measure and should rarely be used.

Suppose the project ends at time k.

- If A(k) < 0 (or equivalently NPViB (a, ρ) < 0 ) then the project is not profitable and the DPP does
not exist.
- If the project is profitable and DPP = t1 then the accumulated profit when the project terminates
k−t1 k−t j
+ t1 ρ(s) (1 + iI )t−s ds
Rk
at time k is A (t1 ) (1 + iI ) + ∑k≥t j >t1 ct j (1 + iI )

Example Consider the following cash flow:

Time 0 1 ··· n−1 n

Cash flow −C x ··· x x
This corresponds to an investment of C which leads to annual payments of x for n years.
(a) Find an expression for the discounted payback period, t1 .
(b) If t1 ≤ n, find an expression for the accumulated profit at time n in terms of t1 . Assume an
annual interest rate of i1 .
1.4 Discounted payback period. 13

Solution.
Now A(0) = −C and A(1) = C (1 + i1 ) + x. Also
h i
A(t) = −C (1 + i1 )t + x (1 + i1 )t−1 + · · · + (1 + i1 ) + 1 for t ∈ {1, 2, . . . , n}

For t ≥ n + 1
h i
A(t) = −C (1 + i1 )t + x (1 + i1 )t−1 + · · · + (1 + i1 )t−n+1 + (1 + i1 )t−n

Then t1 is the smallest integer t with A(t) ≥ 0 provided such a t1 exists. (b) If t1 ≤ n, then the
accumulated profit at time n is
h i
A (t1 ) (1 + i1 )n−t1 + x (1 + i1 )n−t1 −1 + · · · + (1 + i1 ) + 1

The DPP and payback period are measures of the time it takes for a project to become profitable.
However they do not show how large or small the profit is. The DPP is especially useful if capital
is scarce. A project which has a smaller DPP than another may be regarded as preferable because
the capital is released earlier and available for use elsewhere. However, the other project which has
a longer DPP may have a higher NPV! This can occur if the other project has a late large cash inflow.

If the interest rate assumed in the calculation of the NPV is correct, then the project with the
NPV is more profitable. It may be sensible to use a higher interest rate for discounting projects
which are “longer” in order to allow for the uncertainty in forecasting future interest rates. This
leads on to the idea of sensitivity analysis. Because some of the quantities used in the calculation
of the NPV are estimates, it is sensible to vary these quantities one at a time to see how the value of
the NPV varies with the assumptions.

Questions

1. A computer manufacturer is to develop a new chip to be produced from 1 January 2008

until 31 December 2020 . Development begins on 1 January 2006. The cost of development
comprises 9 million payable on 1 January 2006 and 12 million payable continuously during 2007 .
From 1 January 2008 the chip will be ready for production and it is assumed that income will be
received half yearly in arrear at a rate of 5 million per annum.
(i) Calculate the discounted payback period at an effective rate of interest of 9% per annum.
(ii) Without doing any further calculations, explain whether the discounted payback period would be
greater than, less than or equal to that given in part (i) if the effective interest rate were substantially
greater than 9% per annum. (Institute/Faculty of Actuaries Examinations, April 2005) [6 + 2 = 8]

2. A motor manufacturer is to develop a new car model to be produced from 1 January 2002
for 6 years until 31 December 2007. The development cost will be 33 million, of which 18 mil-
lion will be incurred on 1 January 2000 , 10 million on 1 July 2000 and 5 million on 1 January 2001 .

The production cost of each car is assumed to be incurred at the beginning of the calendar year of
production and will be 9, 000 during 2002. The sale price of each car is assumed to be received
at the end of the calendar year of production. Both the production costs and the sale prices are
assumed to increase by 5% on each 1 January, the first increase occurring on 1 January 2003. It is
also assumed that 5,000 cars will be produced each year and that all will be sold.
14 Chapter 1. Equations of Value and the Internal Rate of Return

The sale price of each car produced in 2002 is 12, 100.

(i) Calculate the discounted payback period at an effective rate of interest of 9% per annum.
(ii) Without doing any further calculations, explain whether the discounted payback period would
be greater than, equal to, or less than the period calculated in (i) if the effective rate of interest were
substantially less than 9% per annum. (Institute/Faculty of Actuaries Examinations, April 2000 )
[9 + 3 = 12 ]

3. An investment project gives rise to the following cash flows. At the beginning of each of
the first three years, 180, 000 will be invested in the project. From the beginning of the first year
until the end of the the twenty-fifth year, net revenue will be received continuously. The intial rate
of payment of net revenue will begin at 25, 000 per annum. The rate of payment is assumed to grow
continuously at a rate of 6% per annum effective.

(i) Calculate the net present value of the project at an effective rate of interest of 7% per an-
num.
(ii) Calculate the discounted payback period of the project at an effective rate of interest of 7% per
annum.
(iii) Calculate the annual effective rate of growth of net revenue which would be required if the
project is to have a zero net present value at an effective rate of interest of 7% per annum. (Insti-
tute/Faculty of Actuaries Examinations, Sept

4. (i) An investor is deciding to invest in a project. Explain why the discounted payback pe-
riod is a poorer decision criterion than net present value assuming the investor is not short of capital.

An investor is considering two projects A and B. Project A involves the investment of 1 mil-
lion at the outset. The only income to be received will be a payment of 3.5 million after ten years.
Project B also involves the investment of 1 million at the outset. Income will be received from
this project continuously. In the first year the rate of payment will be 0.08 million, in the second
year 0.09 million, in the third year 0.10 million, with the rate increasing by 0.01 million each year
thereafter until the tenth year, after the end of which no further income will be received.

(ii) Calculate the net present value of both investment projects at a rate of interest of 4% per
annum effective.
(iii) Show that the discounted payback period of project A is after that of project B (no further
calculation is necessary).
(iv) In the light of your answer to (i) above, explain which project is the more desirable to an
investor with unlimited capital, and why. (Institute/Faculty of Actuaries Examinations, April
2002) [2 + 10 + 3 + 2 = 17]

5. A car manufacturer is to develop a new model to be produced from 1 January 2016 for
six years until 31 December 2021. The development costs will be 19 million on 1 January 2014, 9
million on 1 July 2014 and 5 million on 1 January 2015 . It is assumed that 6,000 cars will
be produced each year from 2016 onwards and that all will be sold. The production cost per
car will be 9, 500 during 2016 and will increase by 4% each year with the first increase occur-
ring in 2017 . All production costs are assumed to be incurred at the beginning of each calendar year.

The sale price of each car will be 12, 600 during 2016 and will also increase by 4% each year with
the first increase occurring in 2017 . All revenue from sales is assumed to be received at the end of
each calendar year.
 1.5 Measurement of investment performance 15

(i) Calculate the discounted payback period at an effective rate of interest of 9% per annum.
(ii) Without doing any further calculations, explain whether the discounted payback period would
be greater than, equal to, or less than the period calculated in part (i) if the effective rate of interest
were substantially less than 9% per annum. (Institute/Faculty of Actuaries Examinations, April
2013) [9 + 2 = 11]

1.5 Measurement of investment performance

It is often necessary to be able to measure the investment performance of a fund (for example a
pension fund, or the funds of an insurance company) over a period.

This is important for those responsible for investment funds, eg the trustees of a pension fund, to
monitor how the fund is performing, ie to find out the rate of return the fund is achieving, and to
compare this with the performance of other funds.

In this section we will look at three measures of investment performance.

In most cases our calculations will be based on the market value of the fund, ie how much
the fund’s assets would realise if they were sold on the day in question. It is helpful to realise that
the value of the fund will go up or down as a result of changes in the following components:

Income generated This includes interest payments, dividends and

by the fund rental payments "earned" by the fund’s assets.
Changes in market The price investors are prepared to pay for assets in
value - capital the fund will vary from day to day, in the same
gains/losses way that the price of apples may vary over time.
This includes all "extra" money paid into the fund
that was not generated by the fund itself.
"New money"
Withdrawals from the fund correspond to negative
new money.

In investment performance calculations it is important to distinguish between money generated by

the fund (the first two items in the table) and "new money". Otherwise there is the danger of double
counting or under counting. The cashflows in this section all refer to new money only.

1.5.1 Money-weighted rate of return

One measure of the performance is the yield earned on the fund over the period. The yield earned
on the fund is also called the “money-weighted rate of return” (MWRR).

The money-weighted rate of return is the interest rate satisfying the equation of value incor-
porating the initial and final fund values and the intermediate net cashflows.

Normally we discount each of the cashflows in the equation of value, however here we will
accumulate the initial value of the fund and the cashflows and equate it to the final value of the fund.
Note that the equation of value used in calculating the MWRR only takes account of new money.
Any cashflows generated by the fund itself must be ignored.
16 Chapter 1. Equations of Value and the Internal Rate of Return

For example, consider a fund with value F0 at time 0, with net cashflows ctk at times t1 ,t2 , . . . ,tn
and fund value FT at time T ≥ tn , then the equation of value, equating values at time T, is:

F0 (1 + i)T +Ct1 (1 + i)T −t1 +Ct2 (1 + i)T −t2 + · · · +Ctn (1 + i)T −tn = FT

where i is the effective annual rate of interest earned by the fund in the interval [0, T ]

In this equation of value the left-hand side is the value at time T of the fund at the start of
the period plus or minus all the cashflows received or paid out in the interval.

Example

The market value of a small pension fund’s assets was 2.7m on 1 January 2006 and £3.1m on 31
December 2006. During 2006 the only cashflows were:

- bank interest and dividends totalling 125, 000 received on 30 June

- a cash payment of 100, 000 received on 1 August when a block of shares was sold
- a lump sum retirement benefit of 75, 000 paid on 1 May - a contribution of 50, 000 paid by the
company on 31 December.

Show that the money-weighted rate of return is 16.0%

Solution

Only the last two payments represent new money. So the equation of value (working in £000s) is:

2, 700(1 + i) − 75(1 + i)8/12 + 50 = 3, 100

Evaluating the LHS at interest rates either side of 16.0%, we find:

i = 15.95% ⇒ LHS = 3, 097.9 < 3, 100

i = 16.05% ⇒ LHS = 3, 100.5 > 3, 100

So the MWRR lies in the range (15.95%, 16.05%) ie it is 16.0% (to the nearest 0.1% ).

Example
The value of a fund’s assets was 10m on 1 December 2005 and 11m on 31 December 2006. The
cashflows during this period were:

- 10, 000 received on 1 March

- 50, 000 paid out on 1 June
- 75, 000 received on 31 October

Calculate the money-weighted rate of return.

Solution
The equation of value (working in 000 s) is:

13 10 7 2
10, 000(1 + i) 12 + 10(1 + i) 12 − 50(1 + i) 12 + 75(1 + i) 12 = 11, 000
 1.5 Measurement of investment performance 17

As a first guess we could use a first-order binomial expansion, replacing (1 + i)n with (1 + ni)
13 10 7 2





10, 000 1 + 12 i + 10 1 + 12 i − 50 1 + 12 i + 75 1 + 12 i ' 11, 000

⇒ 10, 000 + 10 − 50 + 75 + 130,000+100−350+150

12 i ' 11, 000

129, 900
⇒ i ' 965 ⇒ i ' 8.9%
12
Evaluating the LHS of the equation of value at interest rates close to 9%

i = 9% ⇒ LHS = 11, 012.8

i = 8% ⇒ LHS = 10, 903.8

Interpolating between these two values, we get:

i−9 11, 000 − 11, 012.8
= ⇒ i = 8.9%
8 − 9 10, 903 · 8 − 11, 012 · 8
So the MWRR is approximately 8.9%.

1.5.2 Time-weighted rate of return

The rationale here is to calculate the “growth factors” reflecting the change in the value of the fund
between the times of consecutive cashflows (ie during periods when “nothing happened”), then to
combine these growth factors to come up with an overall rate of return for the whole period.

The time-weighted rate of return is found from the product of the growth factors between consecu-
tive cashflows.

Define F0 , FT , and ctk as above, and let c0 be the cashflow (if any) at time t = 0; in addition
let Ftk − be the amount of the fund just before the cashflow due at time tk , so that the amount of the
fund just after the receipt of the net cashflow due at time tk is Ftk − + ctk . Then the "Time-Weighted
Rate of Return" (TWRR) is i per annum, where:
Ft1 − Ft2 − Ft3 − FT
(1 + i)T = ···
F0 +C0 Ft1 − +Ct1 Ft2 − +Ct2 Ftn − +Ctn
Each factor on the right hand side gives the proportionate increase in the fund between cashflows.
The product of these factors gives the notional accumulation factor for a single investment of 1 at
time t = 0, invested until time T

Again, note that the cashflows in the formula for calculating the TWRR only include those relating
to new money. Any cashflows generated by the fund itself must be taken into account in the figures
for the fund value.

Using the TWRR eliminates the effect of the cashflow amounts and timing, and therefore gives a
fairer basis for assessing the investment performance for the fund.

Example
Calculate the TWRR for the fund in the example on page 27 given the extra information that the
fund value (including all accrued interest and capital gains) was 3.0m on 30 April 2006 .

Solution
18 Chapter 1. Equations of Value and the Internal Rate of Return

The progress of the fund (again working in 000 s) was as follows:

1 January to 30 April Fund value increased from 2, 700 to 3, 000
1 May Cashflow of −75
1 May to 30 December Fund value increased from 2, 925 to 3, 050
31 December Cashflow of +50, taking fund value to 3, 100

So, during the period from 1 January to 30 April, there were no cashflows and the fund value grew
by a factor of:
3, 000
= 1.111
2, 700
During the period from 1 May to 30 December, the fund value grew by a factor of:

3, 100 − 50
= 1.043
3, 000 − 75

So the growth factor for the whole year is 1.111 × 1.043 = 1.159 and the TWRR for 2006 is 15.9%

Example
The table below shows the progress of a lottery winner’s investment portfolio for the 2006 calendar
year:

Fund value Cashflow

1 January 75, 000
31 March 90, 000
1 April Won Lottery. Winnings invested: 2, 700, 000
30 September 2, 600, 000
1 October Funds withdrawn to pay off gambling debt: 2, 500, 000
31 December 125, 000
Calculate (to the nearest %) the MWRR and the TWRR for 2006 for this portfolio and com-
ment on your answers.

Answer
The MWRR is the value of i for which:

75(1 + i) + 2, 700(1 + i)9/12 − 2, 500(1 + i)3/12 = 125

Establishing a first guess:

9 3



75(1 + i) + 2, 700 1 + 12 i − 2, 500 1 + 12 i = 125

⇒ 75 + 2, 700 − 2, 500 + 900+24,300−7,500

12 i = 125

17,700
⇒ 12 i = −150 ⇒ i = −10.2%

Trying some values close to this first guess:

i = −10% ⇒ LHS = 127.35

i = −11% ⇒ LHS = 112.57

 1.5 Measurement of investment performance 19

Interpolating between these values:

i − (−10) 125 − 127.35

= ⇒ i = −10.2%
(−11) − (−10) 112.57 − 127.35
So the MWRR is approximately −10% The TWRR is given by:
90 2,600 125
1 + i = 75 × 90+2,700 × 2,600−2,500 = 1.200 × 0.932 × 1.250 = 1.398
⇒ i = 39.8%

So the TWRR is +40% (to the nearest %).

Looking at the component factors in the TWRR, we see that the strong positive returns in the first
and third subperiods outstripped the smaller percentage loss in the middle subperiod, leading to a
positive TWRR for the year as a whole.

The MWRR, however, puts almost all the weight on the middle subperiod when the fund size was
largest. This was the period when the lottery winner was rich and the returns on the fund were
strongly negative. It is the performance for this period that leads to the negative MWRR for the
year as a whole.

1.5.3 Linked internal rate of return

Linked rates of return provide a way of combining rates of return for successive subperiods to
obtain an approximate rate of return over a longer period. The linked (internal) rate of return is
found from the product of the yields for the subperiods.

If the rate of return on a fund is measured over a series of intervals. (0,t1 ) , (t1 ,t2 ) , (t2 ,t3 ) , . . . , (tn−1 ,tn ) ,
such that the annual effective rate of interest earned by a fund in the interval (tr−1 ,tr ) is ir (where i1
is the annual rate earned in (0,t1 )) then the "Linked Internal Rate of Return" is i per annum, where:

(1 + i)tn = (1 + i1 )t1 (1 + i2 )t2 −t1 (1 + i3 )t3 −t2 . . . (1 + in )tn −tn−1

The linked internal rate of return will be equal to the TWRR if the subintervals (tr−1 ,tr ) are the
same in each calculation. In practice, the yields ir may be calculated by approximate methods, and
then, if the subintervals used are sufficiently short, the linked internal rate of return will be close to
the TWRR.

Notice that here, the interest rates are annual rates. So for quarterly subintervals over one year, the
linked internal rate of return would be:
1 1 1 1
(1 + i) = (1 + i1 ) 4 (1 + i2 ) 4 (1 + i3 ) 4 (1 + i4 ) 4

However, if we calculated jk to be the quarterly effective return in quarter k for one year, then the
linked internal rate of return would be given by:

(1 + i) = (1 + j1 ) (1 + j2 ) (1 + j3 ) (1 + j4 )

The calculation is similar to the calculation used for the TWRR, except that the subperiods no
longer correspond to the intervals between cashflows, but correspond to "convenient" times in the
calendar year.
20 Chapter 1. Equations of Value and the Internal Rate of Return

A common situation is for an investment manager to have estimates of the quarterly returns
of a fund available from computer printouts, which can then be combined to calculate linked
rates of return for longer periods. The quarterly rates of return in this situation may be calculated
"exactly" using MWRRs, or they may be calculated on an approximate basis.

Example

A life office operates a Far East fund which achieved quarterly money-weighted rates of return of
4.1%, 2.8%, 1.7% and 2.1% during the four quarters of 2004 and halfyearly rates of 2.5% and 3.8%
during the two halves of 2005. Calculate the linked rate of return for period under consideration.
Solution The growth factor for the two-year period is:

1.041 × 1.028 × 1.017 × 1.021 × 1.025 × 1.038 = 1.182257

√
So the annual linked rate of return for the period is 1.182257 − 1 = 8.73%.

Example
Another life office, which also operates a Far East fund, achieved rates of return which under-
performed the first life office by 1/2% per quarter in each of the first two quarters of 2004,
outperformed it by 1% per quarter in each of the last two quarters of 2004 and equalled the perfor-
mance in 2005. Calculate the linked rate of return for the rival life office over the same period.

Solution
The growth factor for the period for the rival life office is:

1.036 × 1.023 × 1.027 × 1.031 × 1.025 × 1.038 = 1.1939

√
So the linked rate of return for the period is 1.1939 − 1 = 9.27%.
 Bibliography

Articles
Books