Chinese University of Hong Kong

Department of Economics

Econ5490 Corporate Finance

Lecture 1: NPV Rule and Other Capital Budgeting Methods

1. A Summary of the Net Present Value Rule

In finance, the net present value (NPV) is defined as the sum of the present values (PVs) of the

individual incoming and outgoing cash flows over a period of time. Alternatively, incoming and

outgoing cash flows are described as benefit and cost cash flows, respectively. Benefit or cost

cash flows of nominally equal value over a time series are affected by the discount factor

resulting in different real value cash flows that makes future cash flows less valuable over time.

The concept of time value of money dictates that time has an impact on the value of cash flows.

If for example there exists a time series of identical cash flows, the cash flow in the present is the

most valuable, with each future cash flow becoming less valuable than the previous cash flow.

This decrease occurs because the discount factor represents the expected rate of return of each

cash flow in a different investment with identical risk, with each decreasing period, the present

value of a subsequent cash flow decreases. The NPV of an investment is determined by

calculating the present value (PV) of the total benefits and costs which is achieved by

discounting the future value of each cash flow. NPV is a useful tool to determine whether a

project or investment will result in a net profit or a loss because of its simplicity. A positive NPV

1
results in profit, while a negative NPV results in a loss. In a theoretical situation of unlimited

capital budgeting, a company should pursue every investment with a positive NPV. However, in

practical terms a company's capital constraints limit investments to projects with the highest

NPV whose cost cash flows do not exceed the company's capital.

In the case when all future cash flows are incoming (such as coupons and principal of a bond)

and the only outflow of cash is the purchase price, the NPV is simply the PV of future cash flows

minus the purchase price (which is its own PV). NPV is a central tool in discounted cash flow

(DCF) analysis and is a standard method for using the time value of money to appraise long-term

projects. Used for capital budgeting and widely used throughout economics, finance, and

accounting, it measures the excess or shortfall of cash flows, in present value terms, above the

cost of funds.

NPV can be described as the “difference amount” between the sums of discounted: cash inflows

and cash outflows. It compares the present value of money today to the present value of money in

the future, taking inflation and returns into account.

The NPV of a sequence of cash flows takes as input the cash flows and a discount rate or

discount curve and outputs a price; the converse process in DCF analysis — taking a sequence of

cash flows and a price as input and inferring as output a discount rate (the discount rate which

would yield the given price as NPV) — is called the yield and is more widely used in bond

trading.

2
Formula

Each cash inflow/outflow is discounted back to its present value (PV). Then they are summed.

Therefore NPV is the sum of all terms,

where

– the time of the cash flow

– the discount rate (the rate of return that could be earned on an investment in the financial

markets with similar risk.); the opportunity cost of capital

– the net cash flow i.e. cash inflow – cash outflow, at time t. For educational purposes, is

commonly placed to the left of the sum to emphasize its role as (minus) the investment.

Any cash flow within 12 months will not be discounted for NPV purpose, nevertheless the usual

initial investments during the first year R0 are summed up a negative cash flow.

Given the (period, cash flow) pairs ( , ) where is the total number of periods, the net

present value is given by:

Discount Rate

3
The rate used to discount future cash flows to the present value is a key variable of this process.

A firm's weighted average cost of capital (WACC) (after tax) is often used, but many people

believe that it is appropriate to use higher discount rates to adjust for risk, opportunity cost, or

other factors. A variable discount rate with higher rates applied to cash flows occurring further

along the time span might be used to reflect the yield curve premium for long-term debt.

Another approach to choosing the discount rate factor is to decide the rate which the capital

needed for the project could return if invested in an alternative venture. If, for example, the

capital required for Project A can earn 5% elsewhere, use this discount rate in the NPV

calculation to allow a direct comparison to be made between Project A and the alternative.

Related to this concept is to use the firm's reinvestment rate. Reinvestment rate can be defined as

the rate of return for the firm's investments on average. When analyzing projects in a capital

constrained environment, it may be appropriate to use the reinvestment rate rather than the firm's

weighted average cost of capital as the discount factor. It reflects opportunity cost of investment,

rather than the possibly lower cost of capital.

An NPV calculated using variable discount rates (if they are known for the duration of the

investment) may better reflect the situation than one calculated from a constant discount rate for

the entire investment duration.

For some professional investors, their investment funds are committed to target a specified rate

of return. In such cases, that rate of return should be selected as the discount rate for the NPV

4
calculation. In this way, a direct comparison can be made between the profitability of the project

and the desired rate of return.

To some extent, the selection of the discount rate is dependent on the use to which it will be put.

If the intent is simply to determine whether a project will add value to the company, using the

firm's weighted average cost of capital may be appropriate. If trying to decide between alternative

investments in order to maximize the value of the firm, the corporate reinvestment rate would

probably be a better choice.

Using variable rates over time, or discounting "guaranteed" cash flows differently from "at risk"

cash flows, may be a superior methodology but is seldom used in practice. Using the discount

rate to adjust for risk is often difficult to do in practice (especially internationally) and is difficult

to do well.

Use in Investment Decision Making

NPV is an indicator of how much value an investment or project adds to the firm. With a

particular project, if is a positive value, the project is in the status of positive cash inflow in

the time of t. If is a negative value, the project is in the status of discounted cash outflow in

the time of t. Appropriately risked projects with a positive NPV could be accepted. This does not

necessarily mean that they should be undertaken since NPV at the cost of capital may not account

for opportunity cost, i.e., comparison with other available investments. In financial theory, if

5
there is a choice between two mutually exclusive alternatives, the one yielding the higher NPV

should be selected.

Interpretation as integral transform

The time-discrete formula of the net present value

can also be written in a continuous variation

where

r(t) is the rate of flowing cash given in money per time, and r(t) = 0 when the investment is over.

Net present value can be regarded as Laplace, respectively, Z-transformed cash flow with the

integral operator including the complex numbers which resembles the interest rate i from the real

number space or more precisely s= ln(1+i).

6
From this follow simplifications known from cybernetics, control theory and system dynamics.

Imaginary parts of the complex number s describe the oscillating behaviour whereas real parts

are responsible for representing the effect of compound interest.

2. Comparison of NPV Rule with Other Capital Budgeting Methods

Motivation

Consider a project with the following cash flows:

C0 C1 C2 C3

-925 +1000 +1400 -1500

Initial Clean-up
Investment Costs

Internal Rate of Return = 4.62%

Opportunity Cost of Capital = 10%

Should we accept this project?

Introduction

NPV is a measure of wealth created for shareholders so maximizing NPV is like

maximizing the wealth of shareholders. We had a simple model, but you can show that in much

more complex environments the same rule still holds. The NPV rule always works.

In addition to the NPV rule, we also had the rate of return rule. In that case they were

equivalent, but in general which should we use? Why go to the bother of calculating NPV when

we could find out whether to invest or not by comparing rate of return with opportunity cost? As

7
the motivation example illustrates it's not always easy to get the correct answer with the rate of

return rule.

In this Lecture we are going to argue for the superiority of the NPV rule over the rate of

return rule or, in its more general form, the internal rate of return (IRR) rule and also over various

other traditional methods of project appraisal.

We will start with why it is better than the IRR. We shall argue that they are equivalent

provided the IRR rule is properly applied, but it is much more difficult to apply the IRR rule, and

hence we should use the NPV rule.

In the one period case we considered before, the one period rate of return R was given by

1+R= Payoff => 1 + R = C1 => C0 + C1 =0

__________ ___ ______
Investment -C0 1+R

In the more general case with more than one period it is not entirely clear what is meant by the

rate of return, so instead we use the more general concept of IRR.

What is the IRR? The IRR extends the last form given above for the rate of return. It is

the rate you discount at such that the discounted cash flow (DCF) is zero. The DCF is a function

of the rate, i, and the cash flows:

C1 C2 CT
DCF(i)  C 0    ... 
1  i (1  i) 2
(1  i) T

The IRR is the value of i such that DCF = 0, i.e.:

C1 C2 CT
DCF( IRR )  C 0    ...  0
1  IRR (1  IRR ) 2
(1  IRR ) T

8
Finding the IRR

The easiest way to find IRR is a calculator. One question that may occur to you is how is

the calculator finding it? To see this, it is helpful to consider what is going on graphically. We

can calculate the DCF at a number of values of the discount rate, join them up and then read off

the value of the discount rate at which DCF = 0. This is the IRR. For example, suppose we have

the following cash flows:

Example 1

C0 C1

-1 +1.1

In this case we have the following values of DCF for various values of i:

i DCF(i)

0 -1 + 1.1 = 0.1

0.05 -1 + (1.1/1.05) = 0.0476

0.1 -1 + (1.1/1.1) = 0

0.15 -1 + (1.1/1.15) = -0.0435

9
Hence the IRR in this case is 10% since this is the value of i at which DCF is 0.

When you use a calculator what it is effectively doing is the same thing. It is finding

where DCF cuts the axis.

IRR Rules Now that we know how to extend the definition of rate of return to many periods

how do we extend our simple rate of return rule from Lecture 2? An obvious extension would be:

Internal Rate of Return Rule (1):

Accept a project provided IRR is above the opportunity cost of capital.

In some cases this rule is perfectly correct, but we have to be extremely careful because in

other cases it has to be adapted in order to lead to the same decisions as the NPV rule.

Cases where IRR Rule (1) is the same as the NPV Rule

Let's consider a case like our example where the DCF is a downward sloping function of i,

the discount rate.

10
When does the IRR rule (1) say that we should accept projects? It says we should accept when

the IRR is above the opportunity cost of capital. For example, if the IRR is 10% and the

opportunity cost of capital is 5% we should accept the project. In other words, we should accept

at points to the left of the IRR. We should reject when the IRR is below the opportunity cost of

capital. For example, if the IRR is 10% and the opportunity cost of capital is 15% we should

reject the project. In terms of the diagram we should reject to the right of the IRR.

When does the NPV rule say that we should accept projects? The NPV of a project is just

the DCF where the rate used to discount is the opportunity cost of capital. Keeping with the

same example, when the opportunity cost of capital is 5% it can be seen from the diagram that

the NPV is positive. Hence we should accept. Similarly, we should accept whenever the

opportunity cost of capital is such that NPV is positive. When the opportunity cost of capital is

15% the NPV is negative and we should reject the project. Similarly, we should reject whenever

NPV is negative.

In this case the IRR rule (1) is the same as the NPV rule. In fact it can be seen from the

diagram that whenever the DCF of a project is a downward sloping function of i, the NPV and

11
the IRR rule (1) give the same result.

When Must IRR Rule (1) Be Adapted to Correspond to the NPV Rule

Now that we have seen that the two rules are the same when DCF is a downward sloping

function of i, when do you think they will differ? Suppose DCF is an upward sloping function of

i, the discount rate:

The IRR rule (1) again says we should accept the project to the left of the IRR. However,

you can see that in this case the NPV is negative so that the NPV rule would say that we should

reject the project. To the right of IRR the IRR rule (1) would say that we should reject but NPV

is positive so that the NPV rule would say we should accept. We know from Lecture 2 that the

12
NPV rule is always right so in this case IRR rule (1) must be wrong. What is happening

intuitively here? To see this, let's look at a very simple example where DCF is upward sloping.

Example 2

C0 C1

+1 -1.1

Notice that this is the same as our original example except that the signs are reversed. Hence the

numerical values of DCF at various values of i will be the same but the signs will be reversed.

i DCF

0 +1 - 1.10 = -0.1

0.05 +1 - (1.1/1.05) = -0.0476

0.10 +1 - (1.1/1.1) = 0

0.15 +1 - (1.1/1.15) = 0.0435

In this case we have the type of relationship depicted above. What is happening? You receive 1

now and pay back 1.1 next period; in other words, you are "borrowing" and the IRR is like a

borrowing rate. Should you try to borrow at a high rate or at a low rate? At a low rate. If you

can borrow at less than the opportunity cost of capital then clearly you can make a profit (i.e.

NPV > 0) and this is what you should do--but this is what the NPV rule is saying.

Consider our original example where DCF is downward sloping:

13
 C0 C1

-1 +1.1

In this case you are laying out $1 now in return for a payment of $1.5 next period. It is

like "lending" and the IRR is like the lending rate. You want to lend at as high a rate of interest

as possible. If the rate you can lend at is greater than the opportunity cost of capital, you can

make a profit so that NPV > 0. Again you can see that the NPV rule is intuitively right.

IRR rule (1) needs to be adapted to take account of whether you are "lending" or

"borrowing". This gives IRR rule (2).

IRR Rule (2)

(i) If DCF is a downward sloping function of i, accept the project provided the opportunity cost

of capital is less than the IRR.

(ii) If DCF is an upward sloping function of i, the project should be accepted if the IRR is less

than the opportunity cost of capital.

The IRR rule (2) will then be equivalent to the NPV rule. In order to use IRR rule (2) we

have to know whether a project is a "lending" or "borrowing" project. In what types of situation

is it likely to be the case that NPV is a declining function of the discount rate i?

C1 C2 CT
DCF  C 0    ... 
1  i (1  i) 2
(1  i) T

where T is the terminal date of the project.

14
 There is one obvious case where this is downward sloping (which is what we need for the

two rules to be equivalent). This is where the first cash flow is an outlay on the investment

project so C0 < 0 and the cash flows in all the subsequent periods are positive so C1, C2, ... CT >0.

As i goes up, the positive terms in the DCF expression get smaller since they are divided by

(1+i)t term whereas the negative C0 term stays the same. Hence DCF gets smaller as i gets bigger

and the expression is downward sloping. In this case the project is guaranteed to be a lending

project. However, if C1 or C2 or any Ct up to CT < 0, then the possibility arises that DCF is

upward sloping and the project is a borrowing one. The reason is that there are now negative

terms that become smaller in magnitude as i rises so DCF may go up. Thus it is not always that

easy to discover which case you are dealing with. For example, consider the following situation

that we looked at in the motivation example.

C0 C1 C2 C3

-925 +1000 +1400 -1500

Initial Clean-up
Investment Costs

Such a profile of cash flows may arise, for example, with strip mining. What is the relationship

between DCF and i in this case?

i = 0%: DCF = -925 + 1000 + 1400 - 1500 = -25 < 0

i = 10%: DCF = -925 + 1000 + 1400 - 1500 = 14 > 0

1.1 1.12 1.13
Hence we have something like this:

15
We then know it's a borrowing situation, and we should apply part (ii) of rule (2).

To summarize, if the first cash flow is negative and all the subsequent ones are positive

then you can use IRR rule (1). However, if there are any negative cash flows after time 0 you

have to use rule (2) and check whether it's a lending or borrowing project. This takes a fairly

long time to do.

With the NPV rule we don't have to determine the situation we are in--we just calculate

NPV. What NPV is measuring in terms of money today is the amount you are better off if you

use the 925 to do the project compared to just putting the 925 in the bank. Consider this way of

thinking for our motivation example. Let’s start by considering what you end up with at date 3

first for just putting the money in the bank and second putting it in the project and taking the cash

flows from that and putting them in the bank.

t= 0 1 2 3

16
(i) Put the 925 in the bank 9251.13 = 1231

(ii) Put the 925 in project

Project cash flow at date 1 1000 10001.12 = 1210

Project cash flow at date 2 1400 14001.1 = 1540
Project cash flow at date 3 -1500
Total at date 3 from project 1250

Wealth created by project relative to bank in date 3 money = 1250 – 1231 = 19

Wealth created by project relative to bank in date 0 money = 19/1.13 = 14 = NPV

The great advantage of NPV is it does this calculation automatically for you. It tells you

immediately how much wealth is created for shareholders by a project.

Multiple roots

The "lending"-"borrowing" problem is a disadvantage for the IRR rule. Moreover, it is

not the only one. There is also the problem of multiple roots that we come to next.

Let us continue with our motivation example

i = 30% DCF = -925 + 1000 + 1400 - 1500 = -10 < 0

1.3 1.32 1.33

We thus have the following situation

17
 In this case there are two values of i which are such that DCF = 0. Which is the internal

rate of return? We can devise a rule to deal with this type of case, but it is complicated.

IRR Rule (3)

(a) If there is just one IRR use IRR rule (2).

(b) Given two IRR: IRR1, and IRR2, there are two possible situations.

(i) If at IRR1, DCF is upward sloping and at IRR2 DCF is downward sloping (i.e. the curve is

n-shaped), then we should accept the project if IRR1 < Opportunity cost of capital < IRR2.

Otherwise, we should reject it.

18
(ii) If at IRR1, DCF is downward sloping and at IRR2, DCF is upward sloping (i.e. the curve

is u-shaped), we should accept if the opportunity cost of capital < IRR1 or > IRR2.

Otherwise we should reject it.

DCF

This is obviously an immensely complicated rule to apply. In fact, the problem is even more

serious than this example indicates since in general we may have many more than two values at

which DCF = 0, especially if the project is long-lived and there are many alterations between +ve

and -ve cash flows. For each of these possibilities we will have to develop a version of our IRR

19
rule.

It is simpler just to calculate NPV!

Mutually Exclusive Projects

Our problems with IRR are not at an end yet. Investment decisions are often concerned

with mutually exclusive projects, that is, ones in which you are choosing among different ways of

doing the same thing. In other words, in some circumstances you can only do one project, you

can't do all of them. An example would be replacing a machine on a production line; there may

be several alternative machines you could potentially use but you can only use one of them to

replace the old one. It would seem that the appropriate rule in this type of situation is:

"Naive" IRR Rule:

Choose the mutually exclusive project with the highest IRR.

Consider projects A and B depicted below, which are mutually exclusive.

In this case there is no problem--B is clearly superior to A. It always has a higher NPV and its

20
IRR is higher so that the naive rule works. When will the naive rule fail to work? Consider the

following situation:

Which is superior, A or B? In terms of NPV, B is superior when the opportunity cost of

capital is above i* and inferior when it is below i*.

In terms of the naive IRR rule, however, B would appear to be superior regardless of i.

What is going on here intuitively? To see this let's consider an example.

Example

Project C0 C1 IRR NPV at 10% NPV at 20%

A -1,000 +1,200 20 91 0

B -10 +30 200 17 15

It can be shown that in this example i* = 18.18.

21
If you graph A and B for NPV against i for this example you get the type of situation shown

above with 10% < i* = 18.18%.

What is happening is a question of relative versus absolute returns. A has a low relative

return but a high absolute return; B has a high relative return but a low absolute return. Given

they're mutually exclusive, you can't do both. What's important is which gives you the highest

absolute return since that's the amount that your budget constraint is pushed out or in other words

the money that actually goes in your pocket. You may be better off investing a large amount at a

low return than a small amount at a high return.

It is possible to fix up the IRR rule by looking at the IRR on incremental cash flows. This

is rather painful and so we will not go through it here. It is much simpler just to use NPV.

One important point to note is that the analysis above assumes perfect capital markets.

This implies that it is possible for the firm to borrow as much as it would like to. If capital

markets are imperfect and the firm is limited in the amount it can borrow there is said to be

“capital rationing.” The way that capital budgeting decisions should be made in this case is

considered further below.

IRR and the Term Structure of Interest Rates

So far we have been assuming that there is just one opportunity cost of capital which is the

same for every period. However, as we discussed in Lecture 3 this may not be the case. The

yield curve may not be flat.

22
 In this case the opportunity cost of capital for different periods differs, and we must

calculate NPV using the formula

C1 C2
NPV  C 0    ...
1  r1 (1  r2 ) 2

We discount C1 at the opportunity cost of capital for 1 year, C2 at the opportunity cost of capital

for 2 years, and so on.

Now if we use the IRR method, which opportunity cost of capital do we compare it to?

In a situation where the yield curve is not approximately flat we have to compare the project IRR

with the expected yield to maturity offered by a traded security that

(1) is equivalent in risk to the project;

(2) offers the same time pattern of cash flows.

This is obviously a difficult task. Again the NPV rule is much simpler. Next we

23
consider the advantages and disadvantages of a number of other investment criteria that are used

by firms.

Other Investment Criteria: Profitability Index and Payback

Profitability Index

Profitability index (PI) = NPV of project

Initial investment

Naive Criterion: Accept project if PI > 0.

It can be seen that only projects with a positive NPV are accepted, so it is again like the

NPV rule. However, as with the IRR, the profitability index ranks mutually exclusive projects

in a different way from NPV. The problem, as with IRR, is that they are both relative measures

of profitability, whereas NPV is an absolute measure, and it is the latter which is relevant here

since this determines the amount the budget constraint of shareholders can be shifted outward.

Consider the example we used with IRR and mutually exclusive projects again.

Project C0 C1 NPV at 10% PI

A -1,000 +1,200 91 0.091

B -10 +30 17 1.7

Here we should choose A since this pushes out the budget constraints of shareholders the

farthest. A naive profitability rule which chose the project with the highest PI would give an

incorrect answer; it would say choose B. However, we can again adapt the PI method to deal

with this by looking at the PI of the incremental investment similarly to IRR. Again,

appropriately adapted, PI, like IRR, is not incorrect but NPV is easier to use.

24
Payback

The payback rule requires that the initial outlay on any project should be recoverable

within some specified cutoff period. The payback period is calculated by counting the number of

years it takes before forecasted cash flows equal the initial investment.

Example

Project C0 C1 C2 C3 Payback NPV @ 10%

A -300 +300 0 0 1 year -27

B -300 +100 +100 +500 3 years 252

With the payback method we will accept A if the cutoff period is 1 year or over and B if it

is 3 years or over. With NPV, however, we would reject A and accept B. Thus payback can give

incorrect answers to the capital budgeting problem, unlike IRR and PI which were just more

difficult to use.

Why doesn't payback give the same results as NPV? There are three main problems with it.

(i) The first difference is clearly that with payback no account is taken of the difference in the

value of $'s at different dates. To try and deal with this we could use the discounted payback

method which takes into account the time value of money but this still suffers from the other two

defects.

(ii) The second defect is that the cutoff date is fairly arbitrary. If you use the same cutoff,

regardless of the project, then you reject too many long lived projects and accept too many short

lived ones.

There are certain special cases where we can calculate a formula for the optimal cut off

25
period which corresponds to maximizing NPV, e.g., if flows are spread evenly across the life of

the project, but in general this is not possible.

(iii) The third defect is that it ignores all flows after the cutoff date so there is a bias against long

lived projects.

Does payback have any advantages? One advantage of payback is its ease of calculation

compared to NPV. With short time horizons where things are constantly being replaced and the

projects are small (e.g., office supplies and equipment) it may be worthwhile to use the payback

method because of its ease of computation. In the majority of cases we shouldn't use it though

since, unlike IRR and PI, it gives incorrect answers.

There are various other criteria which are used for investment decisions. Many are

accounting based such as the book rate of return. These other methods are either difficult to use

or do not correspond to creating wealth for shareholders. I won't go through the deficiencies here

but you can read about them. NPV is best because it is relatively easy to use and directly

measures what we are interested in namely how much wealth is created for shareholders.

Capital Rationing

We have argued above that one of the problems with IRR and PI is that they are relative

measures rather than absolute measures. We suggested it is the absolute amount of money

received that shareholders are interested in so it can be better to take a large investment with a

low rate of return than a small investment with a high rate of return. Implicit in this argument is

the assumption that the firm can invest as much as it wants to. Capital rationing occurs when the

firm has limitations on the capital it can invest which prevent it from undertaking all investment

26
opportunities with positive net present value that are available to it. We next consider how

investment decisions should be made in this case.

There are a number of reasons why firms may operate with capital rationing.

Management may believe a rapid expansion will overtax the organization and use a capital

constraint as a proxy in restricting this growth. This is an example of soft rationing, where the

constraint is internally applied.

Because of problems of fraud, people may be reluctant to give any particular firm too

much money. This is an example of hard rationing, which is externally imposed by the market.

In these types of situations, how should the firm make its investment decisions? It is first

necessary to check that the assumption of perfect capital markets for shareholders is not violated.

Unless this is true, maximizing NPV will not remain a valid objective for the firm. Assuming

there are perfect capital markets for shareholders, the firm should choose its investment, I, to

maximize its NPV subject to the constraint that I is not more than I* where I* is the maximum

amount of capital that can be used.

We want to maximize the NPV of the I* we are allowed to invest and get the "biggest bang

for our buck". We need to select the projects that will maximize NPV per dollar. One natural

intuition is that we should rank by some relative measure such as PI or IRR and accept projects

until we have exhausted the capital constraint. As we shall see, this method sometimes works

and sometimes does not.

Consider what happens if you have two projects, A and B, which require equal investments

IA and IB, respectively, and which have 1+PI's as follows. Plot 1+Profitability Index (1+PI)

27
against Investment (I). Note that

NPV ( PV  I) PV
1  PI  1   1 
I I I

1+PI NPV

NPV

A B

I I
A B
I
*
I

In interpreting this diagram it is helpful to remember that each bar represents a project:

Total Area in each bar = (1+PI) * I

= (PV/I) * I

= PV

Also Unshaded Area = I x 1

=I

Therefore Shaded Area = PV - I

= NPV

Hence the objective of firms is to choose the projects which maximize the total shaded area.

28
In the diagram above, if the firm is constrained to I*, in this situation it should clearly do A.

Ranking by PI and accepting until the constraint is satisfied works in this case.

If there are three projects A, B and C as below which all require the same investment what

should the firm do?

1+PI

A B C

I I IC I
A B

I *

Clearly it should do A and B. Again the rule of ranking by PI and accepting until the

constraint is exhausted works. Is it always the case that the rule works? No! If the projects

require different investments the rule will not work.

1+PI

In this case it should

1
do A and C.

A B C
29

I I I I
A B C
*
I
 It's not even clear that it will always be best to do the most profitable project.

1+PI

A C B

I I I I
A C B
I *

C produces more NPV than A. You can't do both A and C, so you do C. You can still do B

and add more NPV.

Thus ranking by PI and choosing the projects with the highest return until you exhaust the

capital constraint can often give incorrect solutions. It is only if all investments are the same that

it definitely works. Essentially what we should do if investments are unequal is try all possible

combinations of projects that satisfy the capital constraint and choose the combination with the

highest NPV.

With 3 projects this is not too difficult, but as the number increases, the number of possible

combinations becomes large, and it may involve a substantial amount of computation. A

technique that does this is integer programming. This is more sophisticated than simple trial and

error: it chooses its combinations in a systematic way to increase NPV on each new combination.

30
When it gets to a point where no more combinations increase NPV, it stops. In situations where

it is not worth applying such sophisticated techniques because of the time involved ranking by PI

or IRR and taking the best projects until you exhaust I* may on occasions be the best technique.

31