Mathematics for Business and

Economics
Lecture Notes
Christian Anzur∗ Monroe Eskew† Arno Kimeswenger‡
Eirini Kouvela§
September 6, 2023
University of Applied Sciences Wiener Neustadt

∗
christian.anzur@fhwn.ac.at
†
monroe.eskew@univie.ac.at
‡
arno.kimeswenger@fhwn.ac.at
§
eirini.kouvela@fhwn.ac.at

1
Contents

1 Basics of Financial Mathematics 4

1.1 Percentages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.1.1 Average Percentages . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.2 Simple Interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.2.1 Day Count Conventions . . . . . . . . . . . . . . . . . . . . . . . 13
1.3 Compound Interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.4 Continuous Compounding . . . . . . . . . . . . . . . . . . . . . . . . . . 22
1.5 Switching Systems and the Nominal Interest Rate . . . . . . . . . . . . . 26
1.5.1 Switching periods . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
1.5.2 Switching periods and Systems . . . . . . . . . . . . . . . . . . . 28
1.6 The Cash Flow Diagram, the Equivalence Principle, the IRR and NPV . 33

2 Annuities and Loans 37

2.1 Present and Future Value of Annuities . . . . . . . . . . . . . . . . . . . 38
2.2 Annuities in advance and arrears . . . . . . . . . . . . . . . . . . . . . . 42
2.3 Interest Period ̸= Payment Period . . . . . . . . . . . . . . . . . . . . . . 44
2.3.1 Interest period smaller than annuity period . . . . . . . . . . . . . 45
2.3.2 Interest period bigger than annuity period . . . . . . . . . . . . . 47
2.4 Perpetuities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.5 Loans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

3 Economic Functions 57
3.1 Introduction to Functions . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.2 Linear Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.3 Non-linear Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.4 Simultaneous Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.5 Demand and Supply . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
3.6 Cost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
3.7 Revenue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
3.8 Profit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
3.9 Equilibrium and Break-even . . . . . . . . . . . . . . . . . . . . . . . . . 109

4 Calculus in Economic Applications 116

4.1 Calculus Basics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
4.2 Marginal Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
4.3 Economic Applications of Maxima and Minima . . . . . . . . . . . . . . 136
4.4 Point of Inflection, Progressive and Degressive Behaviour . . . . . . . . . 149
4.5 Elasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

2
Preface

These lecture notes cover the topics of the lecture “Mathematics for Business and Eco-
nomics” at University of Applied Sciences Wiener Neustadt. We recommend the book of
Sydsæter (Sydsæter et al., 2016) for more detailed explanations and the book of Bradley
(Bradley, 2013) for more examples. The contrast to the mentioned literature is that in
these lecture notes the notation and explanations are written in the manner they are
taught in the “DACH” region (German, Austria and Switzerland).

The first two chapters are about financial mathematics. Basic knowledge in this field is
necessary for every alumni of a business study and in our opinion for every citizen as
part of their general education. Since money is a key factor in economics, we cover the
math of money at the very first chapter of these lecture notes. In Chapter 1 we cover the
basis of financial mathematics. A solid understanding of percentages, different methods
of compounding and cash flows is crucial. Chapter 2 goes a little bit deeper and covers
systems of regular payments which appear very often in all kinds of finance and private
life, e.g. paying back a loan. The math itself is not difficult but the conventions which
are used in banking sometimes make our life a bit harder.

The next two chapters concern the applications of mathematics in economics: both
microeconomics (the economic behavior of consumers and companies) as well as macro-
economics (the bird’s perspective on economies). Chapter 3 is about the modelling of
relevant business parameters, like costs, profits or revenues, with the help of mathe-
matical functions. For instance, how can we find a proper price for a good or when do
companies break even? Chapter 4 uses the mathematical ideas of calculus to optimise
relevant figures, e.g. to make a maximum profit or to minimise average costs. Moreover,
this chapter discusses different types of growth, as well as the sensitivity of businesses
when they are exposed to changing circumstances.

To meet the requirements of modern real-life applications, Microsoft Excel is used. This
allows us to tackle tasks which are too complex to solve by hand.

We hope that these lecture notes are helpful for your study as well as for your related
tasks in your private and business life.

Your suggestions and comments on the text are always welcome and encouraged.

3
1 Basics of Financial Mathematics

You know, every now and then

I think you might like to hear something from us
Nice and easy
But there’s just one thing
You see we never ever do nothing
Nice, easy
We always do it nice and rough
So we’re gonna take the beginning of this song
And do it easy
But then we’re gonna do the finish rough
This is the way we do “Proud Mary”.1

1.1 Percentages

No other field of mathematics is in my opinion at the same time so simple and difficult.
Simple because it is just about division and multiplications of numbers - things you have
learned a long time ago - and difficult, because obviously many people have difficulties
with it and mix things up a lot. We are trying to avoid those common fallacies by
focusing on the very simplicity of percentages.

The sign %, which certainly all of you know, bears the logic already in it. It is a division
of two numbers, e.g. 4/5 which gives 0.8 as result and the two 0s within the symbol
remind us to read it as 100 times that number in terms of percentages, hence: 80%.

Let’s try that within an example:

Example 1. Let’s suggest you are aware that of the 9 million inhabitants of Austria,
5.3 million are at least partly vaccinated against COVID 19 (status from July 21). What
percentage is that? We just take the 5.3 million and set them into a relationship to all
inhabitants by building the division: 5.3/9 = 0.58̄. We can also say that the relation of
5.3 to 9 is the same as 0.58̄ to 1:
5.3 0.58̄
= .
9 1
Because 1 stands for the unit, for the whole. Furthermore, if we mention the unit not
as 1 but as 100 hundredths, the message as well as the fraction stays the same:
5.3 0.58̄ 58.8̄ 1
= = = 58.8̄ · = 58.8̄%.
9 1 100 100
1
(Turner, 1971)

4
In case you are interested in the number of people vaccinated (x), if the percentage was
just 40%, you just have to solve the following equation:
x
0.40 = ,
9
x = 0.40 · 9 = 3.6.

And in case you know that in a country are 23 million vaccinated people, which is 60%,
and you want to know how many inhabitants there are, i.e. what the unit is, you have
to solve the following equation:
23
0.60 = ,
x
23
x= = 38.3̄.
0.6
■

What we should have learned by now is that the sign % is just an abbreviation of the
word “hundredths” and the calculation is as easy as to divide two numbers. No formulas
needed. You’ve learned some? Forget them.

What is very important to be aware of is that a percentage is nothing else than a fraction,
hence:
numerator
fraction = .
denominator
Keeping that in mind helps you to avoid some of the most common errors with percent-
ages.

You certainly remember from school that adding or subtracting two fractions is not that
easy if the denominators are not the same, e.g.
4 3
+ =?
5 6
The result cannot be found by adding 4 + 3 or adding 5 + 6:
4 3 4+3 7
+ ̸= = = 0.63.
5 6 5+6 11
It’s a little bit more difficult than that:
4 3 4·6+3·5 39
+ = = = 1.3.
5 6 5·6 30
What has this to do with percentages? Let’s have a look at the next example:

5
Example 2. Let’s assume that 4 people out of 5 are vaccinated in group A (the rate is
80%) and 3 people out of 6 are vaccinated in group B (the rate is 50%). Compute the
overall rate for the union of both groups.

If we add up both rates we get

80% + 50% = 130%.

That is nonsense, firstly because it is larger than 100%, secondly because this means
4 3 4·6+3·5
+ = .
5 6 5·6
But this fraction is not what we want. Or can you interpret the right side?

So we want is to find the rate: 4 + 3 vaccinated out of 5 + 6

4+3 7
= = 0.63 ≈ 63, 64%.
5+6 11
■

Example 3. “During the economic crisis the number of sales dropped by 30%, but in
the last year the increase was 35%, so now we are 5% above pre-crisis sales”, announced
the CEO of a Dow Jones company.

Why is he wrong, and what is the correct level of sales compared to pre-crisis numbers?

Unfortunately the CEO missed the part of the lessons where calculations with fraction
were taught. He added up the −30 and the +35 and didn’t mention that the denominator
of the first fraction were the pre-crisis sales and the denominator of the 35 the post crisis
sales:

today = 1.35 · post = 1.35 · (0.70 · pre) = 0.945 · pre.

Today’s sales make up 94.5% of the pre-crisis sales, so despite the recovery period, it
still dropped by 5.5%. ■

The task to calculate 50% of “something” is to multiply this “something” with 0.50. Or in
case of 10%, multiply “something” with 0.10. If you are to increase “something” by 10%
you take that “something” and add the 10% times “something”, what leads to “something”
times 1.10. A common usage is also to speak of basis points or percentage points. In
the latter case you do not increase a value by the relative number which some percentage
implies (e.g. an increase of the unemployment rate of 2% from the initial value of 4%
would mean it is now 4% · 1.02 = 4.08%) but by the absolute value (e.g. if the democrats

6
had 25% in the elections last year and this year 4 percentage points more, it is meant
that they have 29% now and not 25% · 1.04 = 26%). Mainly in the finance context it is
often talked about basis points, which are just a hundredth of a percentage point (e.g.
an interest rate of 1% is increased by 10 basis points yields 1.1%). It is just as easy as
that; try it a couple of times. Table 1 should help you a bit.

English Mathematics
increase by 5% ·1.05
decrease by 2% ·0.98
decline by 95% ·0.05
increase by 105% ·2.05
decline to 95% ·0.95
7% increase by 5 percentage points 7% + 5% = 12%
7% increase by 5% 7% · 1.05 = 7.35%

Table 1: Phrases to formulas

If you calculate anything other than what can be found in the above table, it makes
sense to write it down as an equation. The following example should clarify.

Example 4. A car costs 43 000 € incl. VAT (value added tax, in Austria it is normally
20%). What is the net price (price excl. tax)?

If you knew the net price, you would have to multiply it by 1.2, and the result would be
43 000 €. That’s an equation that can be solved very easily:

netprice · 1.2 = 43 000,

43 000
netprice = ≈ 35 833.33 €.
1.2
A quite common error is to deduct the 20% from the price by multiplying with 0.8. But
firstly, ·0.8 ̸= ÷1.2, and secondly, remember that a percentage is always a fraction. In
our case the numerator is the tax, but the denominator is the net price:
VAT
= 20%.
netprice
■

Remark. As often as possible, you should establish equations to solve problems. That
creates clarity, and you are able to identify structures! ■

7
 Figure 1: The coffee man by Shannon Wheeler

1.1.1 Average Percentages

You certainly have already calculated averages of some numbers. The most common
process here is to add up all numbers and divide by their quantity. That measure is
called the arithmetic mean and plays a major role in many fields especially within
statistics. Here’s its formula:
n
x1 + x2 + x3 + . . . + xn 1 X
x̄ = = · xi .
n n i=1
P
The appearance of the sum sign shouldn’t bother you too much. It is just a very
practical short notation for the long sum and means “Add all x’es up”.

But as we discussed above, such a calculation can be done with lengths, weights, Euros,
and perhaps other units, but not with percentages, because they are fractions.

Remark. Percentages are fractions and normally they can’t be added up. Therefore
normally the arithmetic mean of percentages may not be calculated! ■

But how should we proceed to calculate an average with percentages? The following
example should help you to clarify it.

8
Example 5. The revenues of a company decreased by 10% in 2019, decreased by 5% in
2020 and increased 21% in 2021. What is the relative change of the revenue over all 3
years? What is the average change per year during that period?

Let’s start with the wrong answer: The relative change is −10% − 5% + 21% = 6% and
the average change therefore 6%/3 = 2%. That is wrong! In most cases the difference
to the right answer won’t be big, but that depends strongly on the numbers. It’s wrong
in any case.

How can we calculate the revenue after these three years, if we call the starting revenue
“rev”? We did that already (see example 3).

rev21 = rev · 0.90 · 0.95 · 1.21 = rev · 1.03455.

So the revenue at the end is 3.455% bigger than three years before. We can find the
average percentage p̃ by presuming that the change each year was the same:

rev21 = rev · (1 + p̃) · (1 + p̃) · (1 + p̃) = rev · (1 + p̃)3 .

Solving that equation leads to:

rev21 √
r
3
1 + p̃ = 3
= 1.03455 ≈ 1.01139.
rev
The average change over the three years is about 1.14%. ■

This method of calculating averages is called the geometric mean:

v
u n
√ uY
x̃ = x1 · x2 · x3 · · · xn = t
n n
xi .
i=1

Exercise 1. The price of a flat screen in Spain is 400 € without VAT (i.e. value added
tax), which is 16%. If you sell the screen in Austria, where VAT is 20%, to which amount
and what percentage does the price increase?

Solution: 480 €, +3.4% ■

Exercise 2. After the successful consultancy job you write the invoice of 150 000 €,
including 20% VAT. You have to pay the VAT to the tax office and moreover the income
tax, which is 38% of your income excluding VAT. How much is your net income and
what percentage of your gross invoice total is it?

Solution: 77 500 €, 51.67% ■

9
Exercise 3. A box containing 5 balls costs 8.50 $. If the balls are bought individually,
they cost 2.00 $ each. How much cheaper is it, in percentage terms, to buy the box as
opposed to buying 5 individual balls?

Solution: 15% ■

Exercise 4.

a) The net price of a certain product is 45.66 € and the sales tax is 20%. Calculate the
gross price of the product.

b) A certain gross price of a product is 87.65 € and the sales tax is 20%. Calculate the
net price of the product.

Solution:

a) 54.79 €

b) 73.04 €

Exercise 5. The price of gasoline (95 octane) was 6% more in January 2017 than in
January 2016. In 2018, it will be 20% more than in 2017. In January 2019, it will be
40% more than in 2016. The price of one litre was 1.05 € in January 2017.

a) Calculate the increase from 2018 to 2019 as a percentage of the price from 2008.

b) How much cheaper has gasoline been in 2016 than it will be in 2019?

c) Calculate the average increase per year as a percentage from 2016 over the next 3
years.

Solution:

a) 10.1%

b) 0.40 € (28.6% cheaper)

c) 11.9%

10
Exercise 6. The value of a certain stock on September 12 is 100 €. On September 13 it
goes up by 10%, on September 14 it goes down by 10%. Calculate the value of the stock
on both days.

Solution: 110 €, resp. 99 € ■

Exercise 7. In a basin of saltwater of 60 m3 , the amount of water sunk by means of

evaporation from 95.5% to 93%. How much saltwater remains in the basin? (Although
there is nothing special with this situation, the answer might surprise you)

Solution: 38.6 m3 ■

Exercise 8. The share of wages of the whole national income adds up to 70%. The rest
consists of income from capital and land. The prognosis is that the amount of wages will
rise by 10% in the coming year, whereas the income from capital and land just by 2%.
According to that prognosis how much will the whole national income rise?

Solution: 7.6% ■

Exercise 9. A quantity increases by 25% each year for 3 years. How much is the
combined percentage growth p over the three year period?

Solution: 95.31% ■

1.2 Simple Interest

Let’s think and talk about money. John D. Rockefeller thought that would pay off:

I know of nothing more despicable and pathetic

than a man who devotes all the hours of the waking day
to the making of money for money’s sake.2

Since we learned a lot about percentages, we are able to apply this concept to a broad
range of situations. If we apply it to money, we are talking about interest rates. Ev-
erything has it’s price, and so does money. The price of money is the interest. The
concept is as old as the concept of money itself and was at the center of many religious
and economic disputes over the centuries. If one part hands over money to another, that
latter has to pay a price for this loan, depending on demand and supply, the basis of the

2
(Lapham, 1988)[Note to Ch. 8.]

11
financial markets. Due to this price-finding concept, it is also possible that the price is
negative and the lender even has to pay interests to the debtor. That’s the case at the
moment (2014-2022) for cash lump sums at banks, but shouldn’t bother us too much as
it’s just a matter of “+” or “-” for us.

The initial sum of money we are talking about - often called the principal P0 - is crucial
for the interest I: Double money - double interests, and so on. But time has a linear
influence to the interest too: three times as long - three times as much interest. Therefore
the interest stands in direct proportion to the time and the principal. The coefficient of
that direct proportionality is the interest rate i:3

I = P0 · i · t.

The value after a certain time t consists of the Principal and the interest:

Pt = P0 + I = P0 + P0 · i · t = P0 · (1 + i · t). (1)

This formula is very simple indeed and the reason for the name of the subsection. What
is important when using this formula is to insert i and t in the same units. E.g. if i is
given for the period of one year (p.a. = per annum) you have to insert the time in the
unit years (for example 6 months as 0.5 years).

Let’s practice:

Example 6. Mr. Blumel puts a sum of 2 000 € into his bank account which pays him
1% p.a. because he is so important. He leaves the money in that account from 1.1. until
the 1.7. of the same year. What is the value of the sum at the end of this period?

We can use (1) directly:

P0.5 = 2 000 · (1 + 0.5 · 0.01) = 2 000 · 1.005 = 2 010.

How long would it take him to achieve 15 € interest for his principal of 2 000 €?

Answering that needs to rearrange (1):

Pt 2 015
(1 + i · t) = = = 1.0075,
P0 2 000
i · t = 0.0075,
0.0075
t= = 0.75.
0.01

3
With that notation we refer to the textbook of Bradley, which is UK-based and uses an i for the
interest rate. Internationally the letter r is more common.

12
The unit here is in years because we substituted 0.01 for i, which was an annual rate. If
we want to figure out what day that is, we have to figure out 0.75 · 365 = 273.75 days
after the 1.1. Obviously there is a further hiccup, which we will discuss in the following
section. ■

If we were interested in i or P0 again (1) has to be rearranged, but the task is analogous.

1.2.1 Day Count Conventions

Es könnte so einfach sein,

...
ist es aber nicht.4

In Example 6 we have discovered a devil in a detail: How many days are 0.75 years?
How many days are between 1st January and 1st of July? Is that really exactly half
a year, or a little bit more, or less? Does that depend on the occurrence of an leap
year? Even small differences play a major role if it’s about a huge sum of money. So
business counterparts agreed upon rules to count the days. But to our disgruntlement,
there is not only one agreement, but many. Table 2 gives you an overview5 . This is
also important when using a software package to make calculations with dates and time
periods. Usually there exist a couple of commands due to the variety you will see in
Table 2. In all exercises we use 30E/360, where the date of deposit does not count and
the date of withdrawal counts.

Name Description
30E/360 Every month has 30 days, one year has 360 days. The 31st does
not count.
30/360 Like 30E/360 but the 31st as period end date turns to the 30th ,
if the deposit date is the 31st (and the deposit date turns into
the 30th because of 30E/360).
actual/360 Differences of dates are calculated exactly, year’s length stays
360 days.
actual/365 What do you think?
actual/actual (ISMA) Interest days and years lengths are calculated exactly.
other methods No limit to people’s creativity.

Table 2: Day count rules

4
(Die Fantastischen Vier, 2007)
5
You can gain a very detailed overview at http://www.deltaquants.com/day-count-conventions.

13
Example 7. Let’s assume that the date of deposit is 1.5.2023 and the date of withdrawal
is 10.10.2023. How many days are relevant for compounding if we use 30E/360? If the
amount is 100 € and the interest rate is 10%, how much money do we finally have?

Let’s count the days:

(30 − 1) + 4 · 30 + 10 = 159
Finally we get the amount
 
0.1 · 159
100 · 1 + ≈ 104.42
360
■

Example 8. Let’s take a closer look on a very common situation: a supplier credit.
For paying a bill for something you’ve bought you have two options in many cases:
Paying immediately (or within a very near time frame) and therefor get some discount
or paying later (often four weeks after getting the bill) but the full amount of the bill.

Being more specific we suggest a bill of 1 200 € and a discount of 5% if you pay within
a week. The due date for payment should be in 6 weeks. What would be the equivalent
yearly interest rate of this offer?

The date of payment of the reduced sum (which is 1 200 · 0.95 = 1 140) is not exactly
given ("within a week"), therefore we suggest the latest possible payment (which is at
the very end of the first week). Because of this the time between the two hypothetical
payments is 5 weeks.

Here comes what many people do wrong in this case: You could think about how much
the 5% for 5 weeks are worth for a year which would be
360
5% · = 51.42 . . . %
5·7
per annum.

But the increase within the 5 weeks is not 5%, because the amount “growth” from 1 140
(95%) to 1 200 (100%) in this time, which is an increase of
100
= 1.0526 . . .
95
which corresponds to 5.26 . . . %. This is equivalent to
360
5.26 . . . % · = 54.13 . . .
5·7
i.e. ≈ 54.1% in terms of years. ■

14
Exercise 10. At which date does Mrs. Smith have to put 30 000 € in her bank account, if
the interest rate is 2% p.a., so that the bank account increases to 30 200 € on 10.10.2011
(use the 30E/360 day counting method and simple interest).

Solution: 120 days, 10.06.2011 ■

Exercise 11. For the ultimate Christmas present for his girlfriend, Thomas needs 560 €.
Which interest rate is needed, if he puts all his cash (532 €) in the bank account on April
17 in the same year? He has to buy the present on December 23 at the latest. (use
simple interest and 30E/360)

Solution: 7.70% ■

Exercise 12. Calculate the present value of 6 000 £ that is expected to be received in
three years’ time with simple interest of 7.5% per annum.

Solution: 4 897.96 £ ■

Exercise 13. The following incoming payments show up at a tax inspection: 25 000 €
on 19.01.2008, 140 000 € on 27.03.2008 and 19 000 € on a date that which is illegible,
and 60 000 € on 15.06.2008. On which date did the payment of the 19 000 € appear, if
on 30.06.2008 the money on the account (incl. interest at 4%) is 246 088.89 €? Use
simple interest and 30E/360 DCC.

Solution: 45 days, 15.05.08 ■

Exercise 14. Supplier credit: On the receipt of an electronic gadget, there is the follow-
ing note: “You have to pay the invoice total within four weeks after receiving the invoice,
if you pay within one week you can deduct 2.5% discount.” What is the equivalent yearly
interest rate (simple interest) for that arrangement?

Solution: 44.0% ■

Exercise 15. A machine costing 20 000 € depreciates by 800 € each year.

a) Calculate the value of the machine after 10 years. What is the respective total amount
of depreciation?

b) How long does it take until the machine is worth half of its original value? How long
until it is worth one quarter?

15
Solution:

a) 12 000 €, 8 000 €

b) 12.5 years, 18.75 years

1.3 Compound Interest

Houston, we’ve had a problem here.6

Simple interest is very easy to use and to calculate, but there’s a huge problem inherent
in it: it can be outsmarted: If two parties agree on 2% p.a., then we could easily calculate
the capital after a period of five years as 1+5·0.02 = 1.10 times the principal. The money
increases by 10% over 5 years. But if you take your money out after one year and deposit
it immediately afterwards, the value after two years would be 1.02 · 1.02. If you do that
every year over 5 years, your principal will increase by 2% over five consecutive years
which leads to 1.025 ≈ 1.10408, an overall increase of more than 10.4%. An agreement
that can be outsmarted isn’t a good one.

The easiest solution is to agree upon a formula that reflects the yearly withdraw and
pay-in even if one doesn’t really need to do that. This leads us directly to the following
formula of compound interest:

Pt = P0 · (1 + i)t . (2)

That method is completely different from the method of simple interest. There isn’t a
big difference in the beginning, the result after exactly one year is the same with both
methods:

P1 = P0 · (1 + 1 · i) = P0 · (1 + i)1 .

But the difference increases enormously after a longer period of time. The growth of the
capital now is exponential. Figure 2 illustrates the difference of both methods (calculated
with a principal of 100 € and an annual interest rate of 2%).

This extreme behaviour of compound interest is the topic of the famous story of Joseph’s
penny, mentioned 1772 by Richard Price:
6
(Swigert, 1970)

16
 Figure 2: Comparison of simple and compound interest

Money bearing compound interest increases at first slowly. But, the rate of
increase being continually accelerated, it becomes in some time so rapid, as
to mock all the powers of the imagination. – One penny, put out at our
Saviour’s birth to 5 per cent, compound interest, would, before this time,
have increased to a greater sum, than would be contained in a hundred and
fifty millions of earths, all solid gold. But if put out to simple interest, it
would, in the same time, have amounted to no more than seven shillings and
four pence half-penny.7

Try to verify that story by doing your own calculation or comparison! In Figure 3 you
can see the famous canvas Joseph and the Christ Child.

Now let us use this formula in order to make the four basic calculations with it:

Example 9. What is the value of a lump sum of 1 € invested with 5% p.a. for 200 years
(to make your inheritors happy)?

P200 = 1 · 1.05200 ≈ 17 292.58 €.

Feel the power of exponential growth! ■

Example 10. What is the value of a future gain of 1 million euros today, if it will
appear in 25 years with an interest of

a) 5% p.a.,
7
(Price, 1771)[Page 19]

17
Figure 3: Joseph with his famous son considering an investment.
Joseph and the Christ Child, painted by Guido Reni in 1640. Oil on canvas.
(Reni, 1640)

b) 0% p.a.?

We now have to rearrange (2) for the principal (often also called Present Value) by
simple division.

a) P0 = 1 000 000
1.0525
≈ 295 302.77 €

b) P0 = 1 000 000
1.0025
= 1 000 000 €

Example 11. What interest rate is necessary in order to double your money within 10
years?

We now have to solve (2) for i, the interest rate, which is the base of the exponent.
Therefore we have to take the 10th root:
√
r
10 2P0 10
(1 + i) = = 2 = 1.07177 . . .
P0
The interest rate is a little bit above 7%. ■

Example 12. How much time is needed in order to double your money at an interest
rate of 3.5%?

18
We now have to rearrange (2) for the time t, which is seen as the most difficult task of
these four, because the time is located in the exponent, and to solve it you need to apply
logarithms:

2P0 = P0 · 1.035t ,
2 = 1.035t ,
ln 2 = t · ln 1.035,
ln 2
t= = 20.148 . . . years.
ln 1.035
■

Remark. Don’t learn all the rearranged formulas by heart. You will forget them anyhow,
hopefully the earliest after the exam. Try to practice rearranging the formulas. Once
you’ve managed that, you can apply this skill in many situations and can spare your
valuable brain! ■

It is common to use simple interest in the year of deposit and in the year of withdrawal.
For all other years, the compounded method is used. This combination of simple and
compound interest is called mixed interest.

Example 13. Let’s assume that we deposit 1 000 € on 16.02.2015 and that the date of
withdrawal is 25.4.2018. Compute the amount we can withdraw. (interest rate: 2% p.a.,
day count rule: 30E/360, method: mixed interest)

First we have to count the relevant days for 2015 and 2018:

14 + 10 · 30 = 314, for 2015,

3 · 30 + 25 = 115, for 2018.

In 2016 and 2017 compound interest is used. So we end up with

   
0.02 · 314 2 0.02 · 115
1 000 · 1 + · 1.02 · 1 + ≈ 1 065.31 €
360 360

Exercise 16. You are given a savings account with 1 500 $. A further 500 $ is deposited
in the account at the end of each year for 3 years. Assuming an annual interest rate of
1.2% and yearly compounding, calculate the present and future value of the account.

Solution: FV ≈ 3 072.72 $, PV ≈ 2 964.70 $ ■

19
Exercise 17. A (very popular) old lady has 20 million dollars saved in her bank account
earning a fixed 2.5% interest (compound) per annum. Actuarial tables prognosticate that
her life expectancy from today is exactly 10 years.

a) How much will she be worth on the day she is expected to expire?

b) What if the actuaries are wrong and she manages to live not only 10, but 30 more
years?

Solution:

a) 25.6 million dollars

b) 41.95 million dollars

Exercise 18. Calculate the compound interest rate required for 1 000 € to grow to 2 000 €
in 10 years.

Solution: 7.18% ■

Exercise 19. Compare the average interest rate of the following offers:

a) 12% in the first and the second year

b) 11% the first year and 22% in the second year

c) as b) but vice versa.

Solution:

a) 12%

b) 16.37%

c) like b)

Exercise 20. A capital of 20 000 € is invested for two years with 4% per year, afterwards
2% p.a for three years, and at last 5 years with 6%. How much did the capital grow within
the whole investment period? At which average interest rate has the capital been invested?

20
Solution: 10 720.38 €, 4.39% ■

Exercise 21. A bank offers a savings product with increasing interest rates: 0.00% for
the first two years, 0.25% for the third year, 0.50% for the fourth year, and incredible
10.75% for the fifth year. Find the effective yearly interest rate. The answer is not
2.30%!

Solution: 2.22% ■

Exercise 22. How long should it take for an investment of 3 700 £ to grow to 5 000 £
when interest is compounded annually at 6.5%?

Solution: 4.78 years ■

Exercise 23. How long should it take for an investment of 56 000 £ to grow to 74 000 £
when interest is compounded annually at

a) 6% and

b) 3%?

Solution:

a) 4.78 years

b) 9.43 years

Exercise 24. How long should it take for an investment to double in value when interest
is compounded annually at

a) 6% and

b) 3%?

Solution:

a) 11.90 years

b) 23.45 years

21
 ■

Exercise 25. Let’s assume that we deposit 20 000 € on 28.09.2010 and that the date of
withdrawal is 2.8.2023. Compute the amount we can withdraw. (interest rate: 4% p.a.,
day count rule: 30E/360, method: mixed interest)

Solution: 92 days for 2010, 212 days for 2023 and therefore 33 109.94 € ■

1.4 Continuous Compounding

We just covered the most important method of interest calculation - the compound
interest. To keep calculation easy let’s assume that the interest rate is 10%. Therefore
after one year the starting capital P0 gets

P1 = P0 · (1 + 0.1) = P0 · 1.1

But there is still a big disadvantage in this method: If you agree with your counterpart
on this formula (and the fictive “hop-on/hop-off” of your money every year), one can
outsmart also this method by taking out the money earlier, let’s say after 6 months and
paying it into the account immediately afterwards - of course together with the interests
(which is the yearly interest halved). What happens? Let’s figure out the value after
one year again:
     2
0.1 0.1 0.1
P1 = P0 · 1 + · 1+ = P0 · 1 + = P0 · 1.1025 > P0 · 1.1
2 2 2

That worked great, so let’s do this quarterly and get an even better P1 :
 4
0.1
P1 = P0 · 1 + = P0 · 1.1038 . . . > P0 · 1.1025
4

The smaller the period gets, the larger P1 . The only stable method (which cannot be
improved anymore) would be with an infinitely small time period.

You could counter that this is just a thought experiment and could never be realized.
You’re right. But also in the compound method system does nobody wait until the 31st
of December at 23:59 for somebody in the bank who takes out his money and pays in
immediately. That’s also just a model.

Let’s figure out what happens mathematically. Making the interest period smaller and
smaller leads to always increasing interests over a year. Figure 4 shows this increase. It
is clear from the graph that the function is increasing, but to an always smaller part, so

22
Figure 4: Increase of capital after one year with increasing number of interest periods

that a certain value is never reached or exceeded. It is illustrated by the black horizontal
line - an asymptote to that function. If we take this value as capital after one year, the
system cannot be outsmarted any more. Such a system would be a stable and elegant
one. We just need a formula for it.

Let’s start by letting the periods get smaller and smaller and an arbitrary interest rate
i. At first it gets compounded once a year:
P1 = P0 · (1 + i)
next twice a year:
 2
i
P1 = P0 · 1 +
2
then maybe 12 times a year:
 12
i
P1 = P0 · 1 +
12
or let’s just say n times a year:
 n
i
P1 = P0 · 1 +
n
If we want to display that it gets compounded, we are not allowed to insert “∞” instead
of n, because “∞” isn’t a number - it’s more of a concept. Therefore we say that n tends
to ∞ more and more by using the mathematical symbol “lim”:
 n
i
P1 = P0 · lim 1 +
n→∞ n

23
and for arbitrary time t:
 n·t
i
Pt = P0 · lim 1 +
n→∞ n

The 18th century genius Leonhard Euler, see Figure 5, showed, that the above sequence
(if i = 1, P0 = 1 and t = 1) tends to the famous irrational Euler number e:
 n
1
lim 1 + = e = 2.7182818284590452 . . .
n→∞ n

Euler calculated 23 digits of that irrational number as early as 1748, see (Euler, 1748).

Figure 5: Leonhard Euler (1707-1783)

Bildnis des Leonhard Euler, painted by Jakob Emanuel Handmann in 1753.
(Handmann, 1753)

With this Euler number as a limit for the series, we come to the following equation of
continuous compounding (a little bit tricky, I know):
 n·t   n ·i·t
m= ni i 1 i
Pt = P0 · lim 1 + = P0 · lim 1 + n
n→∞ n n→∞
i
 m·i·t   m i·t
n
m= i 1 1
= P0 · lim 1 + = P0 · lim 1 + = P0 · ei·t
m→∞ m m→∞ m

therefore

Pt = P0 · ei∞ ·t (3)

24
We use i∞ instead of i to be able to distinguish between compound interest and con-
tinuous compounding. Although the concept of continuous compounding is somehow
abstract, it has many advantages (see section 1.5) and is broadly used in securities anal-
ysis and capital markets modelling (e.g. within the Black-Scholes Model). Let’s find out
how it behaves in the wild.

Example 14. A bank account shows 5 000 € today. What is the value after 10 years if
we use continuous compounding with i∞ = 2%?

We simply have to apply (3):

P10 = 5 000 · e0.02·10 ≈ 6 107.01 €
■

Example 15. How much time is needed in order to double your money with continuous
compounding i∞ = 3%?

Lets start with formula (3) and rearrange it:

2P0 = P0 · e0.03·t
2 = e0.03·t
ln 2 = 0.03 · t
ln 2
t= ≈ 23.1 years
0.03
So you can see that this is a little bit easier than with standard compounding. ■

Example 16. Which interest i∞ is necessary if we use continuous compounding, P0 =

100 € and P5 = 122 €?

Lets start with formula (3) again:

122 = 100 · ei∞ ·5
1.22 = ei∞ ·5
ln 1.22 = i∞ · 5
ln 1.22
i∞ = = 0.03977 . . .
5
So i∞ ≈ 3.98%. ■

Exercise 26. A lump sum of 5 000 € shall be compounded for ten years with an interest
rate of i = 4%. The interest method is continuous compounding. How much will the
capital be after 10 years? What would the equivalent yearly interest rate be?

Solution: 7 459.12 €, 4.08% ■

25
1.5 Switching Systems and the Nominal Interest Rate

If you are of the opinion that the financial mathematical world is a little bit complicated,
that isn’t because of mathematics - we love the simple, clear and elegant way - but
because of the way things evolved in the business world. We are now facing different
methods of compounding (simple, compound, continuous) and different methods of day
counting as well as different interest periods (years, quarters, days, . . . ). Because of that
it is very important for us to change from one system into the other, in order to make
comparisons and decide, for instance, which offer or contract would be more beneficial.

1.5.1 Switching periods

To compare compounded amounts in case different interest periods’ length were used,
we always compare the results after the same time period (e.g. one year). If we found an
interest rate of one period length which leads to the same result as if using another one
with another period length, we say the two interest rates are equivalent. In most cases
we compare a yearly interest rate i with a monthly one i12 , but that is not necessarily
always the case.

Let’s start with simple compounding when comparing an interest i for a longer period
and one im , which should be applied m times in the same period (see Figure 6).

Figure 6: long and short interest periods, the result should be the same

Comparison after one year yields:

P0 · (1 + i) = P0 · (1 + m · im )

26
hence,
i
im = or i = im · m (4)
m
Just multiply or divide - it’s as easy as that. Unfortunately simple interest isn’t used
everywhere (mostly in case of bonds or at banks for interest periods smaller than a
year).

The logic in the case of compound interest is the same, only the formulas are different.
Again comparison after one year yields to:

P0 · (1 + i) = P0 · (1 + im )m

hence,
√
i = (1 + im )m − 1 or im = m
1+i−1 (5)

And in the case of continuous compounding? That problem does not appear at all -
there are no different formulas for “continuous”. We like that.

Example 17. A government bond pays interests (called “Coupons”) every quarter of a
year. The coupon is calculated from the face value of the bond. It shall be 2% per quarter.
If you want to compare it with a yearly compounded investment, what would this yearly
interest rate be?Because the Coupons are paid within a year, simple interest has to be
applied.

Due to formula (4) it is very easy to get the equivalent yearly interest rate:

i = 2% · 4 = 8%

Example 18. A zero bond can be bought with 90% of it’s nominal value and after 10
years you receive the nominal value back. This is called “zero bond” because no coupons
are paid during the duration of this security.If you want to compare it with another
investment, what yearly interest rate would the zero bond be equivalent to?

At first we have to calculate the overall interest of this bond. Be aware that it is not
10% because 90 increases to 100, which is an increase of 11% (100 : 90 = 1.11 . . .). In
case of yearly interest rates it is common to apply compound interest, therefore formula
(5) is used:
√
i = 1.111 . . . − 1 = 0.01059 . . . ≈ 1.06% p.a.
10

27
1.5.2 Switching periods and Systems

Because we want to be the almighty superstars of finance - never ripped off by anybody
- we have to be able to change between different interest periods and different systems
at the same time.

The most important case of all of these combinations you could think of is to figure
out the equivalent yearly rate of compounding interest. It is called the APR (annual
percentage rate) and is very important for making correct comparisons of different
kinds of payments.

If the compounding is applied many times during a year (m-times, e.g. quarterly), then
you could use formula (5) to calculate the yearly rate as APR. But some people (or
institutions or companies) don’t do that; instead they use the wrong formula (4). As
said, the result is wrong and to indicate that, they don’t use the name APR, but call it
nominal yearly rate. Very weird. Our task will be to make up for that nonsense and
calculate the APR out of the nominal yearly rate. You can find that in the following
example.

Example 19. The text of an advertisement for a bond could be: “Our bond offers a
nominal rate of 6%, compounded quarterly. Please, please buy it.” Besides some mar-
keting issues, it can be easily found out what the quarterly interest rate is: Because they
used the wrong formula to get the 6% we can easily reverse that process by dividing by 4
and get 1.5% for the quarterly rate i4 . Because we know our stuff, we only have to apply
formula 5 to find out the APR:

1 + APR = 1.0154 = 1.06136 . . .

or together with the division by 4

 4
0.06
1 + APR = 1 + = 1.06136 . . .
4
or even with the −1 if you want
 4
0.06
APR = 1 + − 1 = 0.06136 . . . = 6.136 . . . %
4
■

So, using the nominal yearly interest rate with compounding m times during a year, we
end up with the following formula for the APR:
 m
inom
APR = 1 + −1 (6)
m

28
Example 20. What should the interest rate be in case of continuous compounding in
order to get the same capital after 3 years as with an yearly rate of 2% (compounded
interest). The starting capital is 200 €?

Solution one: If we use yearly compounding we’ll get:

P3 = 200 · 1.023 = 212.2416 €

If we use continuous compounding with an unknown interest rate i∞ , we come to the

equation

212.2416 = 200 · e3·i∞

212.2416
= e3·i∞ ln
200
212.2416
ln = 3 · i∞ ÷3
200
i∞ = 0.01980 . . . = 1.98 . . . %

For getting the same result, the continuous rate is a little bit smaller which illustrates
the fact that continuous compounding is most “effective”.

Solution two: In fact we are interested in i∞ which is equivalent to an APR of 2%. The
duration (3 years) and the capital (200 €) are not relevant at all:

1 + APR = 1.02 = ei∞

i∞ = ln 1.02 = 0.01980 . . .

so i∞ ≈ 1.98% ■

We have seen in the previous example that the APR is given by

APR = ei∞ − 1 (7)

in case of continuous compounding.

Exercise 27. What is the annual percentage rate p.a. if a capital gets compounded
monthly with an interest rate of 1.5%?

Solution: 19.56% ■

Exercise 28. A student has received a 33 333 $ loan from a wealthy aunt in order to
finance her 3-year college program. The terms are that the student repays her aunt in
total at the end of 3 years with

29
a) simple interest,

b) annual compounding,

c) continuous compounding

at a rate of 3% per year. Determine the interest that must be paid.

Solution:

a) 2 999.97 $

b) 3 090.87 $

c) 3 139.11 $

Exercise 29. After an extensive shopping tour you have 1 € left. You want to invest
wisely for the duration of 10 years. Your bank offers five choices:

a) 3.50% annual interest rate, compounded once at the end of the duration,

b) 3.25% annual nominal interest rate, compounded yearly,

c) 3.22% annual nominal interest rate, compounded monthly,

d) 3.21% annual nominal interest rate, compounded weekly,

e) 3.20% annual interest rate, compounded continuously.

Which offer do you choose?

Solution: Here the overall increase in percentage is given:

a) 35%

b) 37.7%

c) 37.93% (this is the best one)

d) 37.84%

30
e) 37.71%

Exercise 30. A bank offers the following three conditions:

a) 2.9% p.a. (annual nominal rate), compounded every 30 days

b) 2.95% p.a. (annual nominal rate), compounded every 60 days or

c) 3% p.a. (annual nominal rate), compounded every 120 days.

Which condition would you choose if you want to invest one single year? Which annual
rates should the two other conditions show to be equivalent to the best one?

Solution:

a) APR = 2.94%, must be 2.988% to be equivalent

b) APR = 2.99%, must be 2.99% to be equivalent

c) APR = 3.03% - that’s best

Exercise 31.

a) A loan of 50 000 €, which is due in 8 years, has to be replaced today. Which amount
has to be paid today (compounded every quarter with 3% per quarter)?

b) On 1.1.2014, the account balance shows 40 000 € and on 1.1.2017, it shows 50 000 €.
What was the interest rate if the money was compounded monthly?

c) When does a capital duplicate, if the interest rate is 6% semiannual?

Solution:

a) 19 416.85 €

b) 0.62% per month

c) 5.95 years, so 6

31
 ■

Exercise 32. Assume that you compare interest rates on deposits from different banks.
Bank Austria offers a yearly nominal rate of 2.9%, compounded semi-annually. At Raif-
feisen, you can deposit your money at a yearly nominal rate of 2.85%, compounded
monthly, and Erste Bank offers a yearly nominal rate of 2.8% with weekly compounding.
Calculate the effective interest rate (APR) for each offer. Which offer would you choose?

Solution: Bank Austria 2.921%, Raiffeisen 2.888%, Erste Bank 2.838% ■

Exercise 33.

a) What sum must be deposited today at a nominal rate of 1% per year, if the goal is
to have an amount of 10 000 €, 50 years from today? Calculate with (i) yearly, (ii)
quarterly, (iii) continuous compounding.

b) The sum of 123 456 789 € is deposited at an annual nominal rate of interest of 1.5%.
How long will it take for the value of the deposit to double if the interest is compounded
(i) annually, (ii) semi-annually, (iii) monthly, (iv) daily, (v) continuously?

Solution:

a) (i) 6080.388, (ii) 6069.09, (iii) 6065.31

b) (i) 46.55 years, (ii) 46.38 years, (iii) 46.23 years, (iv) 46.21 years, (v) 46.209 years

Exercise 34. There is the rule of 70 in economics, which says that you can calculate
the doubling time of a capital by dividing 70 by 100 times the interest rate.8

a) Calculate the doubling time with i = 5% and continuous compounding and compare
this result with the rule of 70.

b) Do the same with yearly compounding.

c) Try to derive a mathematical explanation for the rule of 70 with continuous com-
pounding.

d) Do the same with yearly compounding.

8
This exercise is for people with special interest in finance.

32
e) Sometimes you can also find the slightly adopted rule of 72. What could be the reason
for that?

1.6 The Cash Flow Diagram, the Equivalence Principle, the IRR
and NPV

If you compare different payments at different times, it is very helpful to draw a sketch.
In case of payments on a timeline, that sketch is called a Cash Flow Diagram. Figure
7 shows 3 different payments - K1 , K−2 and K3 -, which should be compared. In the
picture an arbitrary point “0” is chosen for the comparison. Therefore K−2 needs to
be compounded for 2 years; the other payments need to be discounted before the
comparison.

Figure 7: 3 payments to be compared

By the way:

Remark. Never, never, never ever compare payments or add them up or do anything
with them before you’ve transferred them to the same point on the timeline by either
compounding or discounting. ■

In most cases the comparison is done by figuring out, what the value of all payments is
worth at the date of the first payment, whose appearance is most often labelled as date
“0”. This value is called the “present value” 9 .

Sometimes a Cash Flow is made out of money earned and money spent. In this case we
have to mention the money spent as a negative number and than add up all discounted
payments (Figure 8 shows such a situation) to get the so-called net present value.
9
Be aware that the present value does not need to be “now” - this would indeed be very impractical
as “now” continuously changes.

33
Very common is another idea of comparison: What should the interest rate of a payment
series be (often an investment with an initial payment) in order to get a Net Present
Value of Zero? This interest rate is called internal rate of return or IRR and can be
calculated by solving the following equation (IC refers to incoming cash, OC to outgoing
Cash) after q (which is 1 + IRR):
X 1 X 1
0= ICi − OCi
qi qi

Unfortunately (or you might probably say: fortunately) this is an equation, which cannot
be solved analytically, but just numerically. Let us clarify with some examples:

Example 21. By spending 3 000 $ in one year, you can earn 1 000 $ every year for five
years beginning one year after your payment. Is this a good investment for you, if the
interest rate is 8% p.a.?

Some people (not you, I know) would sum up the 5 · 1000 = 5000 and compare with the
spending of the 3000 and would argue that that investment brings in 2 000 $ profit. But
as said within remark above: Nope!

At first it is a good idea to draw a sketch (Figure 8). We have to choose a time at which

Figure 8: Cash Flows on a timeline

we compare the payments. Most common would be the beginning of the timeline. We
would then have the present value of the spending and the earnings and then would be
able to do our comparison.

In this case it is easier to choose the time “1” since we can use the 3 000 $ directly and
compare with the other discounted values. Discounting the five earnings leads to
1 000 1 000 1 000 1 000 1 000
+ + + + ≈ 3 704.39
1.08 1.082 1.083 1.084 1.085

34
That number can be compared with the spending and lets you come to the conclusion that
the profit of that business will be 704.39 $ one year from now.

Another interesting question copes with the influence of the interest rate of that payment
scheme. But that’s for later. ■

Exercise 35. You receive the following offers for the sale of your old home:

a) 30 000 € on 31.12.17, 30 000 € on 31.12.20, 30 000 € on 31.12.23

b) 20 000 € on 31.12.17, 25 000 € on 31.12.18, 45 000 € on 31.12.23

Which offer is best for the seller (compound interest at 6%)?

Solution:

a) P VA = 76 337.39 €

b) P VB = 75 308.13 €

Exercise 36. At the end of his job contract on 31.12.2024, Jonathan will receive a
payment of 200 000 €. However, he wants three payments of equal amounts on 1.1.2018,
1.1.2019 and 1.1.2022 instead. Calculate these payments (5% p.a.).

Solution: 51 218.73 € ■

Exercise 37. Internal rate of return:

a) You give your trustworthy lecturer 10 000 €, and he promises to pay back 5 000 € after
one year and 5 500 € after two years. Calculate the internal rate of return IRR.

b) You give your trustworthy lecturer 10 000 €, and he promises to pay back 2 000 €
after one year, 2 500 € after two years, 3 000 € after three years, and 3 500 € after
four years.

Use Microsoft Excel for calculating IRR.

Solution:

a) mathrmIRR ≈ 3.26%

35
b) IRR ≈ 3.584%

Exercise 38. A project involves an initial outlay of 550 000 £. The expected cash flow
at the end of the next five years is given as: 108 300 £; 130 500 £; 170 900 £; 200 000 £;
160 000 £. Determine the IRR:

a) Graphically by plotting the NPV against i for i = 0.05; 0.07; 0.09; 0.11; 0.13; 0.15; 0.17.

b) By calculation using the MS Excel command “IRR”.

Solution: IRR = 11.34% and Figure 9 ■

Figure 9: Solution with Excel

Exercise 39. The net cash flow for two projects, A (fast food) and B (amusements), is
as in Table 3:

Begin of year 0 1 2 3 4 5
Project A −420 000 −5 000 122 000 130 000 148 000 150 000
Project B −95 000 −10 000 −120 000 200 000 110 000 −50 000

Table 3: Net cash flow

36
a) Use the net present value criterion to decide which project is the most profitable if a
discount rate of (i) 6% and (ii) 8% is used.

b) Calculate the IRR with Excel of each project. Which project would now be considered
more profitable?

Solution:
Project A: at 6%: NPV = 2 331.68; at 8%: NPV = −5 964.2; IRR = 7.56%
Project B: at 6%: NPV = 6 475.72; at 8%: NPV = −1 549.35; IRR = 7.6%
At 6% A is more profitable, at 8% B is more profitable, according to the IRR, B is more
profitable. ■

2 Annuities and Loans

An annuity refers to a stream of payments, called annuity payments (also install-

ments or PMT, that are of equal size and occur at regular intervals. The specific
dates on which these payments are made or are due are called annuity payment dates,
and the time periods between these dates are referred to as payment intervals. The
entire period during which the annuity payments are made is known as the duration
of the annuity. Typically, these payments are also subject to an interest rate i and the
corresponding factor q = 1 + i, as we have seen in previous chapters. Each unit of money
is therefore subject to interest as long as it is invested.

The following example shows the structure of an annuity:

Example 22. Mr. Miller arranges a payment plan with his car dealer for his new used
car. Annual payments of 2 000 € are agreed upon, which will be transferred on December
31st of each year. These payments started in the year 2020, and an interest rate was
negotiated for periods within the year: quarterly rates of i4 = 3%. Sketch the Cash Flow
Diagram.

In Figure 10 you can find the corresponding Cash Flow Diagram. ■

After reading the next section we are able to calculate the present and future value (or
any other in between) by just compounding and discounting the payments as we are
used to. If, and only if, the annuity period is equal to the interest period, things will be
a lot easier, as we will see next.

37
 Figure 10: Cash Flow Diagram of annuity

2.1 Present and Future Value of Annuities

As mentioned in the Cash Flow Diagram section, it is necessary to discount or compound

all payments to the same date in order to compare them. Because of the equivalence
principle, we could use any time we want, but two points out of the infinite possibilities
are more common than all the others: the beginning (Present Value) and the end (Future
Value) of the annuity.

The present value (also: principal value, or PV) of an annuity is the total value
at the beginning of the annuity period. This means that all payments are discounted
to the starting point according to the principle of equivalence and then added up. See
Figure 11 for a visual explanation.

The future value (also FV) of an annuity is the total value at the end of the annuity
period. This means that all payments are compounded to the end point according to
the principle of equivalence and then added up. See Figure 12 for a visual explanation.

Let us take a closer look at the Future Value now. It is quite hard work to compound
every payment and add them up at the end (OK, you’re right - MS Excel would help
us doing so), but we can simplify the process dramatically by at first factoring out the
payments PMT of the future value FV:
FV = PMT · (1 + q 1 + q 2 + · · · + q n−1 )
The term within the brackets is called a geometric sum, because every part of the sum

38
Figure 11: Cash Flow Diagram of the Present Value of an Annuity

Figure 12: Cash Flow Diagram of the Future Value of an Annuity

39
is the one on his left multiplied with the factor q. For such geometric sums a formula
exists we can be derived the following way:

Sn = 1 + q + q 2 + · · · + q n−2 + q n−1
Sn · q = q + q 2 + q 3 + · · · + q n−1 + q n = Sn − 1 + q n
Sn · (q − 1) = q n − 1
qn − 1
Sn =
q−1

as long as q ̸= 1 which is the case if and only if i = 0%. Using this relationship, we get
a very easy to handle and compact formula for the Future Value of an annuity:
qn − 1
FV = PMT · (8)
q−1

The present value could be found the same way (with the calculation based on the
geometric sum), but it is far easier just to take the formula for the FV (formula 8) and
discount it according to the scheme shown in Figure 13.

Figure 13: Relationship between Present Value and Future Value

Therefore we get:
1 qn − 1 1
PV = FV · = PMT · · (9)
qn q − 1 qn
The next two examples show how practical it is to use this formula.

Example 23. A seller of a rental property receives the following two offers:

40
a) 500 000 € immediate payment, with the remainder paid in 5 yearly installments of
250 000 € to be paid at the end of each respective year.

b) A lump sum payment of 1 700 000 € after three years.

Which of these offers is more favorable for the seller, assuming he can invest his money
at an annual interest rate of 6% p.a.?

To compare the two offers, we have to transform all payments to the same date. I choose
the beginning - the present value. At first we have to calculate q = 1 + i = 1.06 and
use (9) to calculate the present value of the first offer, and of course we have to add the
initial payment:
1.065 − 1 1
PV = 250 000 · · + 500 000 ≈ 1 533 090.95
1.06 − 1 1.065
To calculate the present value of the second offer, which consists of only one payment,
we just have to discount this payment to the present:
1
P V = 1 700 000 · ≈ 1 427 352.78
1.063
Now we can see that offer 1 is preferable - it pays out more. ■

Example 24. Over 10 years, an amount of 1 000 € is saved annually, with an annual
interest rate of 3.5%. What is the total saved after 10 years?

Solution: The payment is 1 000 €, and the duration is 10 years. The end amount is
sought, so we can use formula (8).
1.03510 − 1
FV = 1 000 · ≈ 11 731.39
1.035 − 1
■

Exercise 40. A child is given a savings account with 2 000 € . A further 500 € is
deposited in the account at the end of each year. Assuming an annual interest rate of
6%, calculate the value of the account (a) at the end of 10 years, (b) and the end of 20
years.

Solution: (a) 10 172.09 € , (b) 24 807.07 € ■

Exercise 41. An amount of 1 000 € is saved at the end of each year for more than 10
years. The yearly interest rate is 3.5% per annum. What amount does the account show
at the end of year 10?

Solution: 11 731.39 € ■

41
Exercise 42. Over 15 years an amount of 1 000 €is saved at the end of each year, the
yearly interest rate is 4% per annum. What is the present value?

Solution: 11 108.39 € ■

Exercise 43. An installment of 500 € is paid to you at the beginning of each year over 5
years. What should be the equivalent amount, if you want to receive one single payment
at the beginning? The interest is 2% per annum.

Solution: 2 403.86 € ■

Exercise 44. Ludwig bought a house for 480 000 €. Find the yearly payment if the loan
is to be amortized over 35 years at a rate of 8.5% per annum.

Solution: 43 290.90 € ■

2.2 Annuities in advance and arrears

In the previous chapters, it was assumed tacitly that the payments of an annuity are
made at the end of each period. So, at the beginning of the annuity (time 0) no payment
is made, but one is made at the end (time n). This is called an ordinary annuity (also
in arrears; German: “nachschüssige Rente”).

Of course, it is also possible to make the payments at the beginning of each period, in
which case no payment is made at the end of the annuity. This procedure is called an
annuity due (also in advance; German: “vorschüssige Rente”), and this changes the
way the annuity values are calculated.

However, every annuity due can be regarded as an ordinary annuity, and we can continue
to use the calculation methods we have learned so far. Figure 14 shows an example of
an annuity. If we consider the time periods from 1 to n + 1, the annuity is an annuity
due, but from 0 to n, it is an ordinary annuity. This graph also illustrates the extremely
important fact that an annuity should never be specified without its annuity period.
If you do so, you will have to do without the distinction between annuities due and
ordinary annuities.

Anyhow don’t get confused by it: you can also think about it the following way: If every
payment is paid one time period earlier (and the number of payment stays the same)
then the value of all these payments together at a given time should be more than the
value for an ordinary annuity. Because the time shift is one period, we get the value of

42
 Figure 14: Relationship between annuity payed in arrears and in advance

the annuity due by compounding the ordinary annuity once, hence:

FVadvance = FVarrears · q (10)
The same counts for the present value:
PVadvance = PVarrears · q (11)
Summing everything up, shows that you can switch form the PV to the FV by multi-
plying with q n , and you can switch between in arrears and in advance by multiplying
with q. So you get 4 formulas for the price of one. The easiest would be remembering
formula (8) and derive the others from it.

Often you can find 4 formulas in the literature which look really different from each other.
That is because they simplified the formulas mathematically. Normally simplifying is a
good idea, but in this case it hides the structure and obscures understanding.

Example 25. A payment of 500 € is made at the beginning of each year for five years.
What is the equivalent amount if one were to receive the total amount at the beginning
of the payment period (i = 2% per annum)?

This example is asking for the present value of an annuity due. The payment is 500 €,
and the number of payments n is 5. We can calculate both the future value of the annuity
due and its present value as follows:
1.025 − 1
FVadvance = 500 · · 1.02 = 2 654.06 . . .
1.02 − 1

43
and by discounting we get the present value:

PVadvance = 2 654.06 . . . · 1.02−5 ≈ 2 403.86

Exercise 45. An installment of 500 € is paid to you at the beginning of each year over 5
years. What should be the equivalent amount, if you want to receive one single payment
at the beginning? The interest is 2% per annum.

Solution: 2 403.86 € ■

Exercise 46. A famous plastic surgeon overhauls some parts of a famous Hollywood
star. But that didn’t work well. Hence, the indemnity trial ended with a duty of paying
1 000 000 €, to be paid on the 1st of January 2019. (a) The surgeon prefers to pay fixed
yearly installments, whereat the first on January 1st in 2022 and the last on January 1st
in 2030. How big are the installments if i = 6.5%? (b) How many installments should
he have to pay if he – instead – starts on January 1st in 2020 to pay 200 000 € (at an
interest of 6.5% per annum)?

Solution: (a) 170 403.73 € (b) 6.24 years ■

Exercise 47. Ester wants to make an annual deposit over a period of four years. She
gets two offers by her bank (a) 2.8% per annum on an advance payment, or (b) 3% on
a payment in arrears. Which offer is better?

Solution: (a) is the better offer ■

Mr. Green likes to buy a new car and needs a capital of 35 000 €. He plans on paying a
yearly rate in advance over a period of 12 years. Which rate does he have to save when
calculating with an interest of 2.4% per annum?

Solution: 2 491.62 €

2.3 Interest Period ̸= Payment Period

Almost unnoticed (indeed we have mentioned it at the beginning), the previous consid-
erations assumed that the annuity period is as long as the interest period. However,
this assumption does not apply even to the most prominent case, namely that of state
pension schemes. Basically, we have to distinguish between two more cases:

44
a) The interest period is smaller than the annuity period.

b) The interest period is greater than the annuity period.

To solve these problems, we choose a method of problem-solving that is frequently used

in mathematics: we reduce the new unsolved case to the already solved one, and can
thus continue to use all the formulas that have already been derived.

2.3.1 Interest period smaller than annuity period

If there are m interest periods within an annuity period (see Figure 15), we try to find
a way to adjust the interest period to the annuity period. This can be easily done by
calculating an “effective interest rate” within the annuity period, which is equivalent to
an interest rate with an interest period equal to the payment period. This approach is
often referred to as the ISMA10 method in this context.

Figure 15: Interest period smaller than annuity period

Assuming we have m interest periods within the time span of one annuity period, the
new interest iℓ , where ℓ stands for long, should lead to the same result as the m-fold
compounding with i.

(1 + i)m = (1 + iℓ )

10
ISMA stands for International Security Market Association, since 1995 it is the International Capital
Markets Association, see also http://www.icmagroup.org

45
or

iℓ = (1 + i)m − 1 (12)

or

qℓ = q m

With this Interest rate iℓ , we now can use (8) or (9) because the assumption that the
interest rate period equals the annuity period is fulfilled again.

Example 26. Let us consider an annuity due, which consists of 8 payments of 15 000 €
per year, always on January 1st , with the first payment on 1.1.2016. The interests are
added quarterly with an nominal interest rate of 8%. So the interest rate for a quarter
needs to be i4 = 0.08/4 = 0.02. For this annuity, we should calculate the present
value, the future value, and the value at the 1.10.2016. Firstly we have to calculate the
equivalent effective yearly rate by

qℓ = q44 = 1.024 = 1.0824 . . .

We then receive the future value by applying (8):

qℓ8 − 1
FVadvance = 15 000 · · qℓ = 174 226.027 . . .
qℓ − 1
The present value gets computed by discounting:
1
PVadvance = 174 226.027... · = 92 450.132 . . .
qℓ8

For finding the value at the mentioned specific day, we need to compound the present
value for three quarters, using of course the quarterly interest rate i4 resp. q4

V1.10.2016 = PVadvance · q43 = 92 450.132 . . . · 1.023 = 98 108.820 . . .

Exercise 48. A couple estimates that they can afford a mortgage payment of 500 € at
the end of each month. They can obtain a two-year mortgage at a nominal yearly interest
rate of i = 5%, compounded monthly. What is the largest mortgage they can afford?

Solution: 11 396.94 € ■

46
2.3.2 Interest period bigger than annuity period

Concerning this case, there are a couple of different methods to choose from, e.g. the
ISMA method, the US-method or the 360-days-Method. In this text, we are only cover-
ing the first one: ISMA. This method says that we have to calculate the (small period)
interest rate im out of the long-term interest rate so that the periods fit the annuity
period.
√
qm = m q
Respectively:
√
m
im = 1+i−1
The procedure is exactly the same as in (12), just the other way round. After having
transformed the interest rate, we can again use (8) or (9) to calculate present or future
value of the annuity.

Example 27. Calculate the future value of an annuity due, which consists of quarterly
payments of 150 € and runs for 6 years (i = 5% p.a.).

At first we have to calculate the equivalent quarterly interest rate:

√
4
q4 = 1.05 = 1.01227 . . .
or
i4 = 1.227 . . . %
Now we can use (8) and multiply it with q (owing to the fact that it is an annuity due
according to (10)). In doing so, we must not forget that we are in a system with quarters
and we have 24 of them (6 years).
q424 − 1
FV = 150 · q4 · ≈ 4 207.91
q4 − 1
■

Exercise 49. Mr. Smith decides to conclude a retirement savings plan on his 50th
birthday. It should help him increase his monthly income for 10 years, beginning on his
65th birthday. Therefore, he wants to pay 30 000 € yearly. The interest rate is 6.5%
during the saving period and 6% during the retirement period. Which capital is available
at the beginning of each year for Mr. Smith during his retirement?

Solution: 99 032.40 €

47
Exercise 50. A sum of 50 000 € should be on your account in five years. Calculate the
corresponding equal quarterly deposits if interest is 7.5% per annum and the payments
are done at the end of the period.

Solution: 2 094.05 € ■

Exercise 51. Steve signs a contract, requiring him to pay back € 1 000 after one year,
another € 1 000 after two years, and finally 5 000 € after five years. In turn he receives
100 € per month at the beginning of each month for five years. Assuming an interest
rate of 3% per annum, compounded monthly, calculate the present value of this contract.

Solution: -647.36 € ■

2.4 Perpetuities

The application of this supposed utopia will be illustrated by the following example.
If you regularly withdraw an amount from an account with the capital C0 (which, by
definition, is an annuity), the capital is reduced by PMT at each withdrawal time, while
the remaining amount grows by the factor q. If the amount of withdrawal precisely
matches the interest, theoretically, this process can be carried out indefinitely – the
capital remains unchanged. Because of that, the present value of the perpetuity can be
calculated as
PMT
C0 = PV∞ = (13)
i
Because of its simplicity, this relationship is also applied when large terms are agreed
(e.g. in the case of leases) or when endowments are involved (such as in the case of the
Nobel Prize).

Above formula can also be derived with the mathematical concept of the limits of a
sequence:
qn − 1 1
PV∞ = lim PMT · ·
n→∞ q−1 qn
qn − 1 1
= lim PMT · · n
n→∞ i q
1 − q1n PMT
= lim PMT · =
n→∞ i i
The difference between annuity due and ordinary annuity persists: if the payments are
in advance you only have to multiply with q.

48
Example 28. An amount of 500 000 € is available for a foundation. At what interest
rate must this amount be invested so that a literary prize of the amount of 10 000 € can
be awarded annually in arrears?

This example is a perpetual annuity where the cash value and the annual rate are known.
From this, one can then already calculate the interest rate:
10 000
i= = 0.02 = 2%
500 000
■

Exercise 52. “8 000 € monthly – your whole life!” That is the main prize of a well-
known lottery. The sum is promised to be paid at the beginning of each month. Mrs.
Bernd and her husband are proud winners. An employee of the lottery offers a single
payment of 1 300 000 € instead. Should they accept the offer, if they calculate with an
interest of 5% per annum and the assumption, that they will both live for another 30
years?

Solution: Value of the monthly payments: 1 515 425.76 € ■

Exercise 53. At the end of every year the best student of Business Consultancy should
be decorated. What foundation capital is necessary, for paying 20 000 € per year in
perpetuity, when the interest rate is 5% p.a.?

Solution: 400 000 € ■

Exercise 54. A lottery is offering a yearly life-long payment and has a starting capital
of 2 000 000 €. Calculating with an interest rate of 5% per annum, which yearly payment
can be promoted if it will be withdrawn (a) at the beginning of the year, or (b) at the end
of the year.

Solution: (a) 95 238.09 € (b) 100 000 € ■

2.5 Loans

It would not even be necessary to devote a separate chapter to the repayment of a

debt. All related questions can be solved with the help of the calculation of interest and
annuities; it’s just an application of the previous chapters. However, loans and credits
are so important in the economic world that it is worth taking a focused look at them.

49
So it is all the more reassuring to know that there is nothing mathematically new in
store for you in this chapter.

In by far the most situations, the repayment of a debt is being done in annuity repay-
ments of the always same amount. This could be done monthly or yearly, but in either
case, it is done periodically. For this reason, paying back a debt or loan can be handled
as an ordinary annuity. Concerning the notation, the present value of the annuity is
now called the amount of debt you take (from somebody - very often it is a financial
institution, e.g. a bank) and we use L0 for that.

This is why (9) also applies for loans:

qn − 1 1
L0 = PMT · · (14)
q − 1 qn
By manipulating this formula we are able to calculate the Payment:

q n · (q − 1)
PMT = L0 · (15)
qn − 1
as well as the number of payments needed to pay back the whole debt
PMT
ln PMT−L0 ·(q−1)
n= (16)
ln q
which can also be expressed with the help of the principal repayment of the first year
T1 :

ln PMT
T1
n=
ln q
These are the most relevant formulas to work with at the beginning of the loan-taking
process, where you consider if you can afford it, how long it takes to get rid of the loan
again, and so on.

In many cases, it is also important to figure out which amount of money you still owe
to the lender at a given time in the future. For instance, you can think about wiping
out the debt or transferring the loan to somebody else. The outstanding balance can be
calculated with the help of the following idea: If the duration of debt is n = 10 years
and you want to find the outstanding balance after t = 5 years, you could figure out the
value of your debt after t = 5 years if you had not paid back a cent: it would be L0 · q t .
From this sum you have to deduct the amount, which you did pay back. This amount
is the future value of an annuity after t = 5 years.

qt − 1
Lt = L0 · q t − PMT · (17)
q−1

50
This formula can be used to find the remaining debt at a given time t directly.

One other method to manage loans is to figure out the interest and the remaining debt
year by year in an iterative manner. The interest at a given time can be calculated out
of the remaining debt the time before:

It = Lt · (q − 1)

And the value which lowers the debt every year is just the residual value between the
(e.g. yearly) repayment and the interests:

Tt = PMT − It

With the two recursively defined values for It and Tt , we can derive a plan, which shows
the remaining debt, the repayment and the interest year by year (or more generally:
period by period). This is called a loan plan and of course it can perfectly be established
with the use of spreadsheet software (e.g. Microsoft Excel).

Let us try:

Example 29. Find a loan plan for a loan of L0 = 100 000 €, an interest rate of 4%,
and a duration of 10 years. The payments shall be done on a yearly basis.

We can use (15) to find the payment of 12 329.09 € (rounded to cents). In Table 4, the
ending balance for each year is calculated by subtracting the principal payments made in
that year from the beginning balance. The annual interest payment is calculated as the
beginning balance multiplied by the interest rate (4%). The annual principal payment is
calculated as the total payment minus the annual interest payment. The ending balance
of a given year becomes the beginning balance of the next. Please note that in year 11,
there should be the value 0, which does not appear because of the rounding of our total
yearly payment.

Please also note that in the case of constant total payments, the remaining debt after
50% of the time is not 50% of the principal repayment, but more (here: 54 887 €). Figure
16 shows the development of balance and interests in time.

Figure 17 shows that the progress of the remaining debt is not described by a line. It
does not fall below 50 000 € in year 5, but in year 6. The higher the interest rate is, the
longer it needs to reduce the debt. ■

Example 30. A loan of 200 000 € shall be paid off within 10 years. The yearly total
payments should stay the same, and the interest rate is 4% per year. Calculate the yearly
total payments.

51
Year Beginning Balance Annual Interest Annual Principal Ending Balance
1 100 000.00 4 000.00 8 329.09 91 670.91
2 91 670.91 3 666.84 8 662.25 83 008.66
3 83 008.66 3 320.35 9 008.74 73 999.72
4 73 999.72 2 960.00 9 369.09 64 630.83
5 64 630.83 2 585.23 9 743.86 54 886.97
6 54 886.97 2 195.48 10 133.61 44 753.36
7 44 753.36 1 790.13 10 538.96 34 214.40
8 34 214.40 1 368.58 10 960.51 23 253.89
9 23 253.89 930.16 11 398.93 11 854.96
10 11 854.96 474.20 11 854.89 0.07
11 0.07

Table 4: Loan plan

Figure 16: Principal Repayment and Interest over time

52
 Figure 17: remaining debt over time

How does the duration change, if one could just afford a yearly payment of 20 000 €?
Round down to full integers. If a residual remains in this case, it should be repaid
together with the last payment.

We can figure out the payment using (15).

1.04 − 1
PMT = 200 000 · 1.0410 · ≈ 24 658.19
1.0410 − 1
If the payment shall be 20 000 €, of course it needs longer to pay off the debt. (16) gives
20 000
ln 20 000−200 000·0.04
n= ≈ 13.024
ln 1.04
Because this is not an integer, a residual remains after 13 years. This residual can be
calculated using (17):

1.0413 − 1
200 000 · 1.0413 − 20 000 · ≈ 477.95
0.04
The last payment therefore needs to be 20 477.95 €. ■

Example 31. One wants to take a loan of 100 000 € with an interest rate of a gigantic
10% and total yearly payments of 12 000 €. The duration should be calculated. Round
this number down to full integers to get the “real” duration n. The remaining debt has
to be payed at the end of year n + 1.

53
Furthermore, the bank charges fees and deducts them right at the beginning of the loan.
Fees are spicy: 6 000 €. What’s the effective interest rate then?

At first we find the duration with

12 000
ln 12 000−100 000·0.1 ln 6
n= = = 18.799 . . .
ln 1.1 ln 1.1
So the duration is 18 years with an reduced (residual) payment in year 19. This last
payment is 9 681 € using (17):
1.118 − 1
L18 = 100 000 · 1.118 − 12 000 · =
1.1 − 1
To get the payment of year 19 we have to compound L18 one year:
PMT19 = L18 · 1.1 ≈ 9 681.82
In order to calculate the effective rate we can use the IRR command of MS Excel, which
gives us 10.96%. In order to execute the IRR command we can use Table 5 of incoming
and outgoing cash, noting we can subtract the 6 000 € fee from the initial payment: ■

Example 32. A loan of 100 000 € with a total annual payment of 10 000 € is repaid with
an interest of 1%. At what time is 20% of the loan liquidated?

In order to figure this out, we can use (17) with an remaining debt 80 000 € (80% of the
loan):
1.01t − 1
80 000 = 100 000 · 1.01t − 10 000 ·
0.01
80 000 = 100 000 · 1.01t − 1 000 000 · 1.01t − 1 000 000
900 000 · 1.01t = 920 000
92
1.01t =
90
92
t · ln 1.01 = ln = 2.208 . . .
90
So after around two years, nearly 20% of the loan is paid back. But we only cross over
this mark at the third payment. ■

Exercise 55. A loan of 100 000 € is to be repaid in annual payments over 10 years.
Assuming a fixed interest rate of 10% per annum, calculate (a) the amount of each
annual payment, (b) the total interest paid, (c) the amount of interest paid in the first
and second repayment.

Solution: (a) 16 274.54 € (b) 62 745.39 € (c) 10 000 € resp. 9 732.54 € ■

54
 Year Cashflow
0 94 000.00
1 −12 000.00
2 −12 000.00
3 −12 000.00
4 −12 000.00
5 −12 000.00
6 −12 000.00
7 −12 000.00
8 −12 000.00
9 −12 000.00
10 −12 000.00
11 −12 000.00
12 −12 000.00
13 −12 000.00
14 −12 000.00
15 −12 000.00
16 −12 000.00
17 −12 000.00
18 −12 000.00
19 −9 681.82

Table 5: Incoming and outgoing cash

55
Exercise 56. A debt of 200 000 € shall be repaid with a constant yearly payment within
10 years and an interest rate of 4% per annum.
(a) Calculate the installment.
(b) At which value does the maturity time change, if 20 000 € per year can be afforded
(round to integer years).
(c) Should there be a residual left (based on the assumption of an annual payment of
20 000 €, it gets paid together with the last payment. Calculate the last transaction.

Solution: (a) 24 658.18 € (b) 13.024 years (c) 20 477.95 € ■

Exercise 57. A debt of 800 000 € should be repaid within 21 years at an interest rate of
6.8% per annum. The first payment is due after one year since taking the debt. Figure
out:
(a) the installment
(b) the last row of the loan plan
(c) the residual debt after 5 years
(d) the time to maturity if the installments are either 80 000 € or 40 000 € per year (e)
how much is the final payment at an installment of 80 000 €, if it is paid together with the
last installment (f ) after which time is 60% of the debt amortized (installment 80 000 €).

Solution: (a) 72 648.68 € (b) 68 023.11 € - 4 625.57 € - 72 648.68 € - 68 023.11 € (c)

695 474.60 € (d) 80 000 : 17.31 years, 40 000 : never (e) 104 497.46 € (f ) 12.49 years

Exercise 58. Suppose 80 000 € must be paid back in 10 years in equal installments at
the end of each month, with interest at the nominal rate of 7% per year compounded
monthly.
(a) Find the monthly payment.
(b) Find the last row of the loan plan.
(c) Calculate the residual debt after 5 years.
(d) After which time is 50% amortized?

Solution: (a) 928.87 € (b) 923.48 € - 5.39 € - 928.87 € - 923.48 € (c) 46 909.68 € (d)
70.26 periods = months

56
3 Economic Functions

This section consists of two parts. In the first part (Section 3.1-3.4) we revise some
basic mathematical concepts such as functions (linear and non-linear) and simultaneous
equations. If you feel that it is been a while since you have worked with mathematics
and you want to freshen up your memory and your mathematical skills, then this is the
right section for you. If you already feel confident about your mathematical skills you
can skip it and move on to the second part.

In the second part of this section (Section 3.5-3.9) we discuss a number of mathemati-
cal applications in the field of Business and Economics, such as demand, supply, cost,
revenue and profit functions.

3.1 Introduction to Functions

A function can be thought of as a rule which operates on an input and produces an

output. This can be demonstrated as shown in Figure 18.

Figure 18: A general function

In order for a rule to be a function it must produce only a single output for any
given input. For instance, the function with the rule “double the input” is shown in
Figure 18. Note that with an input of 5 the function would produce an output of 10.
With a more general input, like x, the output will be 2x.

Normally we assign a letter to a function in order to label it. The function in Figure 19
has been given the symbol f .

The input to a function is called its argument. When we are given the argument of a
function we can obtain its output. For instance, given the function s(t) = 5t − 3 we can
calculate the value of the output when the argument equals to 1.

57
 Figure 19: f (x) = 2x

In this case we have:

s(1) = 5 · 1 − 3 = 5 − 3 = 2.

Exercise 59. State the rule of each of the following functions:

a) g(x) = 4x + 11

b) h(t) = 7t − 49

c) k(x) = 5
2−x

d) m(n) = n
3
+ 1
2

Solution: a) multiply the input by 4 and then add 11, b) multiply the input by 7 and
then subtract 49, c) subtract the input from 2, then find the reciprocal of the result and
multiply it by 5, d) divide the input by 3 and then add 12 . ■

Exercise 60. Given the function f (x) = 3x + 4 find:

a) f (5)

b) f (−2)

c) f 1


3

d) f − 23

Solution: a) f (5) = 19, b) f (−2) = −2, c) f ( 31 ) = 5, d) f (− 23 ) = 2 ■

58
Exercise 61. Given the function g(t) = 3t2 − 7 find:

a) g(3x)

b) g(x + 5)

c) g(6x − 4)

d) g(4t + 9)

Solution: a) 27x2 − 7, b) 3x2 + 30x + 68, c) 108x2 − 144x + 41, d) 48t2 + 216t + 236 ■

The graph of a function

There are several ways of representing functions. Until now we have seen the algebraic
form of functions. In this section we will discuss two more representations: the table of
values and the graph.

Let us consider the function g(x) = x3 . The output of this function can be obtained by
multiplying the input by 31 . By choosing different values for the input to this function,
we can calculate the corresponding outputs. In Table 6 we have done this for integer
values of x between -3 and 3. In this way we have built the table of values in the interval
[−3, 3] with a step size of 1 for g(x).

Input x −3 −2 −1 0 1 2 3
Output g(x) −1 − 23 − 13 0 1
3
2
3
1

Table 6: Table of values

Each pair of input and output values can be represented on a graph by a single point. The
input values are measured along the horizontal axis (x-axis) and the output values are
measured along the vertical axis (y-axis). We often write the function as y = g(x) = x3 .

We can now plot each pair of x and y values in the table as a single point shown as •
in Figure 20. The point is often labelled as (x, y). The values of x and y are called the
coordinates of the point. The joined points produce the required graph as shown above.
In this case the graph is a straight line which goes through the origin (0, 0).

Dependent and independent variables

Since x and y can have a number of different values they are variables. Here x is
called the independent variable and y is called the dependent variable. When we
choose a value of the independent variable x, the function rule enables us to calculate

59
 2

1.5

0.5

-3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 3 3.5

-0.5

-1

-1.5

-2

x
Figure 20: Graph of g(x) = 3

the corresponding value of the dependent variable y. The input to the function is the
independent variable and the output is the dependent variable.

The domain and range of a function

The domain of a function is the set of values that the independent variable takes. For
instance, the function

f (x) = 6x + 3, −2 ≤ x ≤ 10

has any value of x between −2 and 10 inclusive as its domain because it has been stated
like this. If the domain of a function is not explicitly stated then it is taken to be the
largest set possible. For example,

g(t) = t2

has domain R = (−∞, ∞) since g is defined for every value of t and the domain has not
been stated otherwise.

The range of a function is the set of all values taken by the function for a given domain
(the set of y values). For example, the image of f (x) is [−9, 63] and the image of g(t) is
[0, ∞).

Often the domain of the function is not given and we need to define it ourselves. In
that case we need to take into account at which values is the function undefined. These
values of the independent variable must be excluded from the domain. For instance for
the function f (x) = x1 we would have to exclude 0, since 10 is meaningless. In that case
the domain would be R \ {0} = (−∞, 0) ∪ (0, ∞).

Example 33. Consider the function given by f (x) = x2 + 2, −3 ≤ x ≤ 3.

60
a) State the domain of the function.

b) Draw up the table of values (with step size one) for this function.

c) Plot a graph of the function.

d) Deduce the range of the function by inspecting the graph.

Solution:

a) The domain of a function f (x) is the set of values that x is allowed to take. In this
case: [−3, 3].

b) Table 7 is the requested table of values:

x −3 −2 −1 0 1 2 3
f (x) 11 6 3 2 3 6 11

Table 7: Table of values f (x)

c) The graph of f (x) is shown in Figure 21.

12

11

10

-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7

Figure 21: f (x) = x2 + 2

d) The range of a function f (x) is the set of values that the function takes as x is varied.
We can deduce this information from the graph but also from the table of values. Here
the range of values is [2, 11].

61
 ■

Example 34. Explain why the values x = 2 and x = −3 must be excluded from the
domain of the function f (x) = (x−2)(x+3)
5
.

Solution: When x = 2 or x = −3 the denominator is 0, so f (x) is not defined here. ■

Exercise 62. Plot a graph of the following functions. In each case state the domain
and the range of the function.

a) g(t) = 3t + 2, −2 ≤ t < 5

b) m(n) = 2n2 + 8, −2 ≤ n ≤ 4

c) w(x) = 6 − x2

Solution: a) [−2, 5), [−4, 17), b) [−2, 4], [8, 40], c) R, R ■

Exercise 63. Solve Exercise 62 with Excel. ■

3.2 Linear Functions

Any function of the form

y = f (x) = ax + b

is called a linear function. The constant a is called the coefficient of x and b is referred
to as the constant term. The graph of a linear function is always a straight line.

The value of the constant term is also known as the vertical or y-axis intercept, since
it is the value of y where the line cuts the y-axis. The coefficient of x gives the slope or
gradient of the line.

Example 35. Plot graphs of the following functions y = x

2
− 3 and y = −2x + 3.

Solution: For each function we need to find two distinct points and join them with a
straight line. For instance, for y = x2 − 3 we choose x = 0 so that y = −3. As second
point, let x = 6 so that y = 0. The points (0, −3) and (6, 0) can be plotted and joined as
shown in Figure 22. We proceed similarly for the function y = −2x + 3. Let x = 0 so
that y = 3 and for the second point let x = 3 so that y = −3. ■

62
 5

-4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

-1

-2

-3

-4

-5

-6

Figure 22: Graph of y = 12 x − 3 and y = −2x + 3

In the above example we note that in the first line y = x2 −3 the gradient a has a positive
value, this produces a graph that slopes upwards from left to right. The larger the value
of a, the steeper is the graph. A negative value of a, like in y = −2x + 3, produces a
graph that slopes downwards from left to right. If a is zero the line is horizontal (parallel
to the x axis), that is its gradient is zero.

Measuring the slope of a line

If the formula of the linear function is not given we can calculate the slope of the line
from its plot. Figure 23 shows how the slope of a line is calculated. The line through

10

-12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7

-1

Figure 23: Measuring slope

the points A and B has a positive slope since it is rising from left to right. The slope is

63
measured as follows:
change in height 2 1 0.5
Slope = = = = = 0.5
change in distance 4 2 1
Here the height increases by 0.5 units when the horizontal distance increases by 1 unit.

On the other hand, the line through the points C and D has a negative slope since it is
falling from left to right. A decrease in height is indicated by a minus sign:
change in height −6 −1.2
Slope = = = = −1.2
change in distance 5 1
The height decreases by 1.2 units when the horizontal distance increases by 1 unit.

Exercise 64. State the vertical intercept of each of the following lines:

a) y = x
2
− 1
3

b) y = 1 − 3x

c) y = −5x

Solution: a) − 13 , b) 1, c) 0 ■

Exercise 65. State the gradients of the following lines? Which one has the steepest
gradient?

a) y = 17x+4
5

b) y = 9x − 2

c) y = x
3
+4

Solutions: a) − 17
5
, b) 9, c) 13 , (b) has the steepest gradient. ■

Exercise 66.

a) Calculate the gradient of the line joining (1, 0) and (15, −3).

b) Calculate the gradient of the line joining (10, −3) and (15, −3).

Solution: a) − 14
3
, b) 0 ■

64
Exercise 67. The equation of a line is given as 4x + 2y − 8 = 0.

a) Write down the equation of the line in the form y = ax + b.

b) Plot the line.

Solution: a) y = 4 − 2x ■

Inverse Function

In the functions that we have seen until now y always depends on x or y is a function
of x.

This can be denoted in several ways such as y = f (x), y = g(x), y = h(x), etc., where
the letters f , g and h are used to distinguish different formulae in x.

For example:

a) y = −3x + 11 or f (x) = −3x + 11

b) y = 2 − 5x or g(x) = 2 − 5x

When equations are used in applications sometimes we write y in terms of x, denoted

as y = f (x) (input: x, output: f (x)), but in other applications we require x in terms of
y, denoted as x = f −1 (y) (input: f (x), output: x). This latter function, x = f −1 (y), is
called the inverse function, see Figure 24.

Figure 24: Inverse Function f −1

Example 36. The function

y = 11x − 12

65
can be written after rearranging the terms and writing x as the variable on the left hand
side
y + 12
x=
11
y
x= + 1.09
11
This is the inverse function written symbolically as:
y
x = f −1 (y) = + 1.09
11
■

Exercise 68. Find the inverse of the following functions:

a) f (x) = 4x + 7

b) s(t) = 1
t+1

Solution: a) f −1 (x) = x−7

4
, b) s−1 (t) = 1−t
t
■

3.3 Non-linear Functions

Solving Quadratic Equations

A quadratic equation has the general form

ax2 + bx + c = 0

where a, b and c are constants. The solution of this equation may be found by using the
quadratic formula
√
−b ± b2 − 4ac
x= (18)
2a
The value of the term inside the square root is called discriminant (∆ = b2 − 4ac) and
gives rise to different types of roots.

• ∆ = b2 − 4ac > 0, then the term under the square root is positive leading to
two different solutions (real roots). The graph will cut the x-axis at two different
points as shown in Figure 25 (a).

66
 • ∆ = b2 − 4ac = 0, then the term under the square root is zero, leading to two
identical solutions (two identical real roots). The graph will touch the x-axis at
one point only as shown in Figure 25 (b).

• ∆ = b2 − 4ac < 0, then the term under the square root is negative, leading to two
different complex roots (this case is not in the scope of this course). The graph
will never cut or touch the x-axis as shown in Figure 25 (c).
q
20

18

16

14

12

10

-16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16

-2

-4

Figure 25: (a) real roots, (b) repeated roots, (c) complex roots

Example 37. Solve the following equations:

a) x2 + 6x + 5 = 0

b) x2 + 6x + 9 = 0

c) x2 + 6x + 10 = 0

Solution:

a) Firstly we identify the coefficients of the equation x2 + 6x + 5 = 0 which are a = 1,

b = 6, c = 5 and then we calculate the discriminant ∆ = 62 − 4 · 1 · 5 = 16 > 0. The

67
 discriminant is positive, hence we have two real roots:
√
−6 ± 16
x=
2·1
−6 + 4 −2
x1 = = = −1
2 2
−6 − 4 −10
x2 = = = −5
2 2

b) The coefficients of the equation x2 + 6x + 9 = 0 are a = 1, b = 6, c = 9. We calculate

the discriminant ∆ = 62 − 4 · 1 · 9 = 0 and since it equals to zero, we have two real
roots (repeated):
√
−6 ± 0
x=
2·1
−6 + 0 −6
x1 = = = −3
2 2
−6 − 0 −6
x2 = = = −3
2 2

c) The given equation x2 + 6x + 10 = 0 has a = 1, b = 6, c = 10 as coefficients. ∆ =

62 − 4 · 1 · 10 = −4, the discriminant is negative, hence we have two complex roots11 .
Although complex roots are not a part of this course, readers who are interested can
find the solutions here:
√
−6 ± −4
x=
2·1
−6 ± 2i
x=
2
−6 + 2i
x1 = = −3 + i
2
−6 − 2i
x2 = = −3 − i
2

Exercise 69. Solve the following equations:

a) x2 − 3x + 2 = 0

b) x2 − 5 = 4x
11
Complex roots are numbers that have an imaginary part to them. The imaginary part, i, is found
when taking the square root of a negative number.

68
c) 2x2 = 10x

d) x2 + 3 = 2x

e) 25x2 = 40x − 16

Solution: a) x1 = 2,√x2 = 1, b) x1 √
= 5, x2 = −1, c) x1 = 0, x2 = 5, d) there is no real
solution (x1 = 1 − i 2, x2 = 1 + i 2), e) x = − 54 repeated ■

Sketching a quadratic function f (x) = ax2 + bx + c

A quadratic function has the form f (x) = ax2 + bx + c, where a ̸= 0. Its graph is called
a parabola and has a curved shape.

• If a > 0, then the parabola opens up (Figure 26).

• If a < 0, then the parabola opens down (Figure 26).

Figure 26: f (x) = 12 x2 and g(x) = − 12 x2

One of the main points of a parabola is its vertex. It is the highest or the lowest point
on its graph. The coordinates of the vertex are:
  
−b −b
,f ,
2a 2a

69
for a derivation see Section 4.3.

x2
For instance, the vertex of f (x) = 2
is (0, 0) (see Figure 26):
−b 0
= =0
2a 2 · 12
 
−b 1
f = f (0) = · 0 = 0
2a 2

Example 38. Sketch the graphs of f (x) and g(x) and calculate their vertices.
f (x) = 2x2 + x − 5
g(x) = −x2 + x − 2
Solution: Initially we calculate a table of values for the given function in a reasonable
interval. In this case we will compute the values of f (x) in [−4, 4] with step size one.
The results are illustrated in Table 8.

x −4 −3 −2 −1 0 1 2 3 4
f (x) 27 13 3 −3 −5 −3 3 13 27

Table 8: Table of values f (x) = 2x2 + x − 5

Then we plot the graph, like in Figure 27.

110

100

90

80

70

60

50

40

30

20

10

-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9

-10

Figure 27: Graph of f (x) = 2x2 + x − 5

Lastly we compute the vertex:

−1 1
=−
2·2 4
−1 1
f( ) = f (− ) = −4.875
2·2 4

70
Vertex: (−0.25, −4.875).

We work similarly for g(x) = −x2 + x − 2. In Table 9 we see the values of the function
in the interval [−4, 4] with step size one.

x −4 −3 −2 −1 0 1 2 3 4
g(x) −22 −14 −8 −4 −2 −2 −4 −8 −14

Table 9: Table of values g(x) = −x2 + x − 2

Figure 28 illustrates the plot of the function.

-8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9

-10

-20

-30

-40

-50

-60

-70

-80

-90

-100

-110

Figure 28: Graph of g(x) = −x2 + x − 2

Accordingly we compute the vertex.

−1 1
=−
2 · (−1) 2
   
−1 1
g =g = −1.75
2 · (−1) 2

Vertex: (−0.50, −1.75).

Exercise 70. Plot the graphs of the following functions and calculate their vertices:

a) x2 − 3x + 2 = 0

b) −x2 − 4x + 5 = 0

Solution: a) (1.5, −0.25), b) (−2, 9) ■

71
Exercise 71. Solve Exercise 70 with Excel. ■

Cubic Functions

A cubic function has the form

f (x) = ax3 + bx2 + cx + d

where a ̸= 0 and a, b, c and d are constants. Setting f (x) = 0 produces a cubic equation
of the form:

ax3 + bx2 + cx + d = 0

There is, unfortunately, no easy general formula for solving a cubic equation, but ap-
proximate solutions may be found graphically by plotting the graph and finding the
points of intersection with the axes like in the example that follows.

Example 39. Plot the graphs of the following equations:

a) y = x3

b) y = 0.5x3 − 5x2 + 8.5x + 27

c) y = −0.5x3 − 5x2 + 8.5x + 27

Can you estimate from the graph:

• the roots,

• the minima and maxima?

Solution: For each of the functions we will firstly set up a table of values and then we
will plot the graph. Lastly we will comment on the roots and the minima and maxima.

a) In Table 10 we see the values of y = x3 in the interval [−2, 2]. As we notice in Figure
29 the function has only one root at x = 0 and it has no minimum or maximum
points.

b) In Table 11 we see the values of y = 0.5x3 − 5x2 + 8.5x + 27 in the interval [−8, 8]. As
we can notice in Figure 30 the function has only one root at x = −1.56 approximately,
a local maximum at x = 1, y = 31 and a local minimum at x = 5.5, y = 5.6.

72
 x −2 −1 0 1 2
y −8 −1 0 1 8

Table 10: Table of values y = x3

120

100

80

60

40

20

-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6

-20

-40

-60

-80

-100

-120

Figure 29: Graph of y = x3

x −8 −4 0 4 8
y −617 −119 27 13 31

Table 11: Table of values y = 0.5x3 − 5x2 + 8.5x + 27

100

80

60

40

20

-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11

-20

-40

-60

-80

Figure 30: Graph of y = 0.5x3 − 5x2 + 8.5x + 27

73
c) Table 12 displays the values of function y = −0.5x3 − 5x2 + 8.5x + 27 in the interval
[−8, 8]. The function has three roots, at x = −11, x = −1.7 and x = 2.8 approxi-
mately, a local minimum at x = −7, y = −107 and a local maximum x = 0.8, y = 30
(see Figure 31).

x −8 −4 0 4 8
y −105 −55 27 −51 −481

Table 12: Table of values y = −0.5x3 − 5x2 + 8.5x + 27

100

80

60

40

20

-13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8

-20

-40

-60

-80

-100

-120

Figure 31: Graph of y = −0.5x3 − 5x2 + 8.5x + 27

Try now to solve the example with Excel. ■

Remark. Cubic functions are curves, which may have:

• One root to three roots.

• No minimum/maximum or minima/maxima.

These properties are illustrated in Figures 29, 30 and 31. ■

Exponential Functions

Before introducing the exponential functions we would like to remind the reader what
an exponential expression is and a number of laws that can be used in order to simplify
this kind of expressions.

74
An exponent is another name for a power or index. Expressions involving exponents are
called exponential expressions. For instance, 23 and ab . In the exponential expression
ax , a is called the base and x is the exponent (Croft and Davison, 2019).

Exponential expressions can be simplified and manipulated using the following laws of
indices:

• am · an = am+n

am
• an
= am−n

• (am )n = amn

An exponential function has the form

y = ax
where a is a positive constant called the base and x is called the exponent or index or
power of the exponential function.

Hence y = 9.8x , y = 0.25x and y = 17x are all exponential functions. Note that, in an
exponential function, the independent variable appears as a power.

The most commonly used exponential function, often referred to as the natural expo-
nential function, is
y = ex
where the base e is the exponential constant whose value is e = 2.71828182 . . . which we
have already seen in section 1.4.

Example 40. Plot the graphs of the following exponential functions on the same dia-
gram:

• y = 2x

• y = (2)−x

Solution: The function y = 2x has base 2 and variable power x. The function y = 2−x
has base 2 and variable power −x. To plot the two functions, we calculate a table of
values (with step size one) such as in Table 13. The graphs of the two exponential
functions can be seen in Figure 32. From this diagram we note some properties of the
exponential function:

• As x becomes large and positive, 2x increases without bound.

75
 x −2.00 −1.00 0.00 1.00 2.00
2x 0.25 0.50 1.00 2.00 4.00
2−x 4.00 2.00 1.00 0.50 0.25

Table 13: Table of values y = 2x and y = 2−x

-5 -4 -3 -2 -1 0 1 2 3 4 5

Figure 32: Graphs of y = 2x and y = 2−x

76
 • As x becomes large and negative, 2x approaches 0.

• 2x is always positive.

The property that 2x increases as x increases is referred to as exponential growth.

On the other hand, we see that 2−x decreases as x increases. This is referred to as
exponential decay. ■

Exercise 72. Plot the graph of the function y = ex .

Solution: See Table 14 and Figure 33. ■

x −2.00 −1.00 0.00 1.00 2.00

ex 0.14 0.37 1.00 2.72 7.29

Table 14: Table of values y = ex

-5 -4 -3 -2 -1 0 1 2 3 4 5

Figure 33: Graph of y = ex

Exercise 73. Plot the graph of the function x = 3 + 2e−t . Which value does x approach
as t increases?

Solution: As t increases, x approaches 3. See Figure 34. ■

77
 5

0 1 2 3 4

Figure 34: Graph of x = 3 + 2e−t

Logarithmic Functions

Consider the statement

8 = 23

Here 2 is the base and 3 is the power. This may be expressed in an alternative way using
logarithm as

log2 8 = 3

We read this as logarithm of 8, to the base 2, equals 3.

Here you can see some more examples:

• 25 = 52 may be written as log5 25 = 2

• 81 = 34 may be written as log3 81 = 4

• 0.125 = 0.53 may be written as log0.5 0.125 = 3

Generally, if a is a positive constant with a ̸= 1, and

N = ax

then

loga N = x

78
The number a is called the base of the logarithm. Note that a is positive and a ̸= 1.
In practice, most logarithms are to the base 10 or e. Logarithms to base e are called
natural logarithms and are denoted by ln. Logarithms to base 10 are denoted simply by
lg.

Expressions using logarithms can be simplified by using the following laws:

• loga A + loga B = loga AB

loga A
• loga A − loga B = loga B

• loga An = n loga A

A logarithmic function has e.g. the form

y = lg x

and

y = ln x

for x > 0. In Table 15 we consider y = lg x and y = ln x for different x-values.

x 0.50 2.00 4.00 6.00 8.00 10.00

lg x −0.30 0.30 0.60 0.78 0.90 1.00
ln x −0.69 0.69 1.39 1.79 2.08 2.30

Table 15: Table of values y = lg x and y = ln x

The graphs of the two functions are shown in Figure 35.

-1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

-1

Figure 35: Graphs of y = logx and y = lnx

From the graphs illustrated above we note the following common properties:

79
 • As x increases, both lg x and ln x increase indefinitely.

• As x approaches 0, both lg x and ln x approach minus infinity.

• lg 1 = ln 1 = 0

• lg x and ln x are not defined when x is negative or zero. Thus the domain of the
logarithm functions is x > 0. The range of these functions is R.

These properties are also for arbitrary bases true.

Remark. The logarithmic function y = loga x (where a > 0 and a ̸= 1) is the inverse
of the exponential function x = ay . For instance, for the natural exponential function
holds

f (x) = ex
f −1 (x) = ln x

In Figure 36 you can see the plots of f (x) and its inverse function.
8

-12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12

-1

-2

-3

-4

-5

-6

-7

-8

Figure 36: Graphs of f (x) = ex and f −1 (x) = ln x

Solving equations involving logarithms and exponentials

The process of finding x, having been given 10x or ex , is known as taking logs. In the
following examples we discuss this technique for solution.

Example 41. Solve the following equations:

a) 10x = 17.4

80
b) ex = 5.25

c) 3 · (100.5x−2 ) = 96

d) lg x = 1.4

e) ln x = −0.5

f ) lg(x2 + 2) = 2.6

g) 1 750 = 753e0.03t

Solution:

a) We have 10x = 17.4, so taking lg s gives:

x = lg 17.4
x = 1.2405

b)

ex = 5.25
ln ex = ln 5.25
x · ln e = ln 5.25
x = ln 5.25 = 1.6582

c)

3 · (100.5x−2 ) = 96
96
100.5x−2 =
3
0.5x−2
10 = 32
0.5x − 2 = lg 32
0.5x = lg 32 + 2
lg 32 + 2
x= = 7.0103
0.5

d)

lg x = 1.4
x = 101.4 = 25.12

81
e)

ln x = −0.5
x = e−0.5 = 0.61

f)

lg(x2 + 2) = 2.6
x2 + 2 = 102.6
x2 = 102.6 − 2
√
x = 102.6 − 2 = ±19.90

g)

1 750 = 753 · e0.03t

1 750
= e0.03t
753
e0.03t = 2.234
0.03t = ln 2.234
ln 2.234
t= = 28.11
0.03

Exercise 74. Solve the following equations:

a) lg(x + 2) = 2.5

b) 2 · ln x − ln(x + 1) = 0

c) ex+5 = 1.56

d) 2x · 2x+1 = 7

e) 4 · lg(5x − 6) = −0.8

f ) 20 = 40 · (1 − e−x )

Solution: a) x = 314.228, b) x = −0.618 and x = 1.618, c) x = −4.555, d) x = 0.9037,

e) x = 1.3262, f ) x = 0.6931 ■

82
Rational Functions
f (x)
A rational function has the form g(x)
, f and g are polynomial functions of x and g(x) ̸=
0.

The simplest rational function is y = x1 . To plot the graph, we need to calculate a table
of values such as Table 16.

x −1.50 −1.00 −0.50 0.00 0.50 1.00 1.50

1
y= x
−0.67 −1.00 −2.00 ? 2.00 1.00 0.67
1
Table 16: Table of values y = x

We note that no y-value can be defined for x = 0, so we cannot plot a point there.

-5 -4 -3 -2 -1 0 1 2 3 4 5

-1

-2

-3

-4

-5

-6

1
Figure 37: Graph of y = x

Here the y-axis is a vertical asymptote and the x-axis is a horizontal asymptote. An
asymptote is a line which the curve approaches at a distance from the origin.

Example 42. Sketch the function y = 4

x−1

Solution: Firstly, we calculate the value of x that gives rise to division by zero. This
will determine the vertical asymptote. Division by zero occurs when the denominator is

83
zero. In that case:

x−1=0⇒x=1

Then we calculate the points of intersection with the y-axis, using the fact that x = 0 on
the y-axis:
4 4
y= ⇒y= = −4
x−1 0−1
In order to get an idea of the curvature as the graph approaches the vertical asymptote,
we calculate the coordinates of some points to its left and right, see Table 17 and Figure
38:

x 0.00 0.40 0.80 1.20 1.60

4
y= x−1
−4.00 −6.67 −20.00 20.00 6.67
4
Table 17: Table of values y = x−1

-12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 11 12

-1

-2

-3

-4

-5

-6

-7

-8

-9

4
Figure 38: Graph of y = x−1

Try to plot now the graph with Excel. ■

3x2 +x+1
Exercise 75. Sketch the function t(x) = x−1
. Provide also a table of values in a
reasonable interval.

Solution: The table of values is illustrated in Table 18 and the graph in Figure 39.

84
 x −7 −5 −3 −1 1 3 5 7
3x2 +x+1
t(x) = x−1
−17.6 −11.8 −6.3 −1.5 ? 15.5 20.3 25.8

3x2 +x+1
Table 18: Table of values t(x) = x−1

3x2 +x+1
Figure 39: Graph of t(x) = x−1

3.4 Simultaneous Equations

In this chapter we discuss different methods for solving simultaneous equations (graph-
ically and analytically).

Elimination

One way of solving simultaneous equations is by elimination. Elimination, as the name

implies, involves removing one of the unknowns. This method is outlined in the following
example.

Example 43. Solve the simultaneous equations:

3x + 5y = 31 (19)
2x + 3y = 20 (20)

Solution: Firstly, we need to modify each equation so that the coefficient of x is the same
in both equations. This can be achieved if equation (19) is multiplied by 2 and equation

85
(20) is multiplied by 3.
6x + 10y = 62
6x + 9y = 60
Now the unknown x can be removed or eliminated if the second equation is subtracted
from the first:
6x + 10y = 62
(−) 6x + 9y = 60
0x + 1y = 2
The result implies that y = 2. To find x we substitute the value found for y into either
of the given equations. For example, using equation (19):
3x + 5 · 2 = 31
3x = 21
x=7
Hence the solution of the given equations is x = 7 and y = 2. Note that you should
always check your solution by substituting back into both of the given equations. ■

Substitution

The next method that we will discuss is substitution. From either equation we derive
an expression for one variable in terms of the other. We then substitute this expression
into the other equation and solve.

Example 44. Solve the simultaneous equations:

2x + 3y = 12.5 (21)
−x + 2y = 6 (22)

Solution: We solve (22) for x,

x = 2y − 6
and then we substitute to (21):
2 · (2y − 6) + 3y = 12.5
4y − 12 + 3y = 12.5
7y = 24.5
24.5
y=
7
y = 3.5

86
Substitute y = 3.5 into (22) and solve for x:
x = 2 · (3.5) − 6 = 1
The solution of the two given equations is x = 1 and y = 3.5. ■

Graphical Solution

An alternative way of solving two simultaneous equations in two unknowns is to draw a

graph. The equations we have studied until now are linear. By plotting a line for each
equation and locating the point of intersection of the two lines we can find the solution
of the equations. If the two lines do not intersect then the equations have no solution.
If the two lines are identical, there are an infinite number of solutions.

Example 45. Solve the simultaneous equations:

4x + y = 9 (23)
−x + y = −1 (24)
Solution: We can rearrange both equations into the standard form of the equation of a
straight line:
y = 9 − 4x
y = −1 + x
By selecting two points on each line we can produce the graphs of the two equations as
shown in Figure 40. At the point where the two lines intersect the x and y coordinates
must satisfy both (23) and (24) and so the point of intersection represents the solution.
As we can see from the graph the point of intersection is (2, 1). The solution of the given
equations is therefore x = 2 and y = 1.

Try now to solve the same example with Excel.

Equations with infinitely many solutions

A set of equations has infinitely many solutions when there is an infinite number of sets
of values that satisfy all equations. Consider the following example.

Example 46. Solve the simultaneous equations:

2x + y = 8 (25)
4x + 2y = 16 (26)

87
 6

-4 -3 -2 -1 0 1 2 3 4 5 6 7 8

-1

-2

-3

-4

Figure 40: Graphs of y = 9 − 4x and y = −1 + x

Solution: We multiply (25) by 2 and we have

4x + 2y = 16
4x + 2y = 16
So the two equations are identical. There is only one equation in two unknowns. If x is
given any value, the corresponding y-value can be calculated.

For example, when

x = 0 ⇒ 2y = 16 ⇒ y = 8
x = 1 ⇒ 2y = 16 − 4 ⇒ y = 6
x = 2 ⇒ 2y = 16 − 8 ⇒ y = 4
and so on. There is an infinite number of (x, y) that satisfy both equations. ■

Remark. If we solve the simultaneous equations of the above example graphically, we

will see that the equations represent coincident lines, hence every point on one line is
also a point on the other line. Since a line has infinitely many points, there are infinitely
many solutions or points in common. The graphical solution is left as an exercise to the
reader. ■

Equations with no solution

It can occur that there is no set of values that satisfies both equations. Consider the
following example.

88
Example 47. Solve the simultaneous equations:

y =1+x (27)
y =2+x (28)

Solution: With the elimination method we find

0 = −1

This is not possible, therefore there is no solution. ■

Remark. If we attempt to solve the equations of the above example graphically, we will
see that the two lines will never meet since they are parallel and thus will never have a
point in common. Hence there is no solution. ■

Exercise 76. Solve the given simultaneous equations:

a) 7x + 11y = −24 and −9x + y = 46

b) 5x − y = 7 and 2x + y = 7

c) 2x − 2y = −2 and 5x + y = −9

d) 5x − 2y = 15 and 15x − 45 = 6y

e) 4x − 3y = 4 and 2y + 1.5x = 20

Solution: a) x = −5 and y = 1 b) x = 2 and y = 3 c) x = − 53 and y = − 23 d) infinitely

many solutions e) x = 5.44 and y = 5.92 ■

Exercise 77. Solve the simultaneous equations of Exercise 76 with Excel. ■

Remark. It is possible to have simultaneous equations with non-linear equations, but

solving them gets generally a bit more complicated. ■

3.5 Demand and Supply

The demand and supply decisions made by consumers, producers and the government
influence economic activity within an economy. In the following section we discuss the
mathematical tools needed for the analysis of these two.

89
The demand function

The term demand refers to the amount of some good or service consumers are willing
and are able to purchase at a price. There are several factors that influence the demand
for a good X, such as the income of the consumer, the price of substitute goods, the
level of advertising, etc. In basic economic analysis, all factors except the price of the
commodity are often held constant. The simplest model for the demand function is
written as

Q = f (P )

where Q is the quantity demanded for good X and P is the price of the good X.

Let us assume that the demand for good X is given by the linear equation Q = 200−2P .
This equation describes the law of demand: a rise in price of good X decreases the
quantity demanded of that good. Conversely, a fall in price will increase the quantity
demanded.

The demand function Q = 200 − 2P is illustrated in Figure 41 with Q plotted on the

vertical axis and P on the horizontal one.

200

180

160

140

120

100

80

60

40

20

-20 0 20 40 60 80 100 120 140

Figure 41: Demand function Q = 200 − 2P

In most applications in economics, however, Q is plotted on the horizontal axis and P

on the vertical one. In order to do so, we need to estimate the inverse demand function,

90
P = f −1 (Q) = g(Q). In this case we have

Q = 200 − 2P
2P = 200 − Q
P = 100 − 0.5Q

The inverse demand function is illustrated in Figure 42.

140

120

100

80

60

40

20

-20 0 20 40 60 80 100 120 140 160 180 200 220 240

-20

Figure 42: Demand function P = 100 − 0.5Q

In many cases the demand function P = g(Q) can be modelled by the simple linear
equation

P = a − bQ

where a and b are constants. Since this is the equation of the straight line the following
can be deduced:

• a > 0: the vertical intercept of the demand function is positive. When the quantity
demanded is zero, Q = 0, the price per unit is P = a. In other words, when the
price per unit P = a there is no demand for the good.

• (−b) < 0: the slope of the demand function is negative. The negative slope
indicates that price drops by b units for each unit increase in quantity.

Example 48. Let us consider the aforementioned demand function

P = 100 − 0.5Q.

91
a) State the slope and the vertical and horizontal intercepts.

b) What is the quantity demanded when P = 10.

Solution:

a) The vertical intercept is 100 (where the straight line cuts the y-axis). This means
that when the price P = 100 there is no demand for the good Q = 0. The slope
∆P
∆Q
= −0.5. This shows that the price drops by 0.5 units for each unit increase in the
quantity demanded.

We can calculate the horizontal intercept by substituting P = 0 into the equation:

P = 100 − 0.5Q
0 = 100 − 0.5Q
0.5Q = 100
100
Q= = 200
0.5
This means when the price is zero, so actually when the good is free, 200 units are
demanded.

b) to calculate the demanded quantity when P = 10 we just have to substitute P = 10

into the demand function.

P = 100 − 0.5Q
10 = 100 − 0.5Q
0.5Q = 100 − 10
90
Q= = 180
0.5

Example 49. A demand function is given by the equation

P = 3 000 − 1.5Q.

a) Plot this equation as a function P = f (Q) for reasonable values of Q.

b) State and give a verbal description of the slope and both intercepts.

c) What is the price when Q = 1 000?

92
d) What is the quantity demanded when P = 1 000?

e) Find an expression for the inverse demand function, i.e. the demand function in the
form Q = f −1 (P ) and plot its graph.

Solution:

a) We have Q > 0, since it does not make sense to have a negative quantity of goods or
services. In order to find the upper bound of the interval where the demand function
makes sense we set P = 0 and solve the equation for Q.

0 = 3 000 − 1.5Q
1.5Q = 3 000
3 000
Q= = 2 000
1.5
Now we can plot the demand function P = 3 000 − 1.5Q in 0 < Q ≤ 2 000 as shown
in Figure 43.

3000

2500

2000

1500

1000

500

0 500 1000 1500 2000 2500

Figure 43: Demand function P = 3 000 − 1.5Q

b) The slope is ∆Q
∆P
= −1.5. This means that there is a price decrease by 1.5 units for
each unit increase in quantity supplied.

The vertical intercept is (P = 3 000, Q = 0), this means that there will be no units
demanded at a price greater or equal to 3 000.

93
 The horizontal intercept is (P = 0, Q = 2 000), hence when the good is for free 2 000
units will be demanded.

c) In order to calculate the price when 1 000 units are demanded we just have to substitute
Q = 1 000 into the equation and solve for P

P = 3 000 − 1.5 · 1 000

P = 3 000 − 1 500
P = 1 500

d) In order to find the demanded quantity when the price is 1 000, we substitute P = 1 000
into the equation and then solve for Q

P = 3 000 − 1.5Q
1 000 = 3 000 − 1.5Q
1.5Q = 3 000 − 1 000
2 000
Q= ≈ 1 333
1.5

e) The equation of the demand function in the form Q = f −1 (P ) is derived as follows:

P = 3 000 − 1.5Q
1.5Q = 3 000 − P
3 000 − P 2
Q= = 2 000 − P
1.5 3
In Figure 44 is illustrated the graph of Q = 2 000 − 23 P .
2500

2000

1500

1000

500

0 500 1000 1500 2000 2500 3000

Figure 44: Demand function Q = 2 000 − 23 P

94
The supply function

When economists talk about supply, they mean the amount of some good or service a
producer is willing to supply at a price. There are several factors that influence the
quantity supplied of good X, such as the price of the good itself, the price of other
goods, the tax, etc. The simplest model for the supply function is written as

Q = f (P )

i.e., quantity supplied depends on price only, as long as the other variables upon which
supply depends remain constant.

Let us assume that the supply is given through the linear equation Q = 2P + 20. A
rise in price leads to an increase in the quantity supplied of that good or service. This
positive relationship between price and quantity supplied is called the law of supply.

In many cases the supply function P = h(Q) can be modelled by the simple linear
equation

P = c + dQ

where c and d are constants.

The graph of the supply function P = c + dQ is illustrated in Figure 45.

Figure 45: Supply function P = c + dQ

• c ≥ 0: the vertical intercept is positive

• d ≥ 0: the slope is positive. Price increases by d units for every unit increase in
quantity supplied.

95
Example 50. The supply function is given by

P = 10 + 0.5Q.

a) State the slope and the vertical and horizontal intercepts.

b) Plot the supply function in the interval [0, 100].

c) Estimate the value of Q when P = 15 from the graph. Confirm your answer from the
equation.

d) Write the equation of the supply function in the form Q = h(P ). Plot the graph of Q
in terms of P .

Solution:

a) The vertical intercept is c = 10. This means that the firm will supply no units at a
price less or equal to 10.

The slope is ∆Q
∆P
= 0.5. This means that there is a price increase by 0.5 units for each
unit increase in quantity supplied.

We can calculate the horizontal intercept by substituting P = 0 into the equation of

the supply function and then solving for the corresponding value of Q

P = 10 + 0.5Q
0 = 10 + 0.5Q
0.5Q = −10
−10
Q= = −20
0.5
Hence the horizontal intercept is (Q = −20, P = 0). Note here that this result
economically does not make sense, since we got as result a negative number of produced
units.

b) In order to plot the graph of the supply function over the given range we can use the
intercepts calculated in (a) and estimate a table of values (see Table 19):

Q 0 20 40 60 80 100
P 10 20 30 40 50 60

Table 19: Supply schedule

96
 70

60

50

40

30

20

10

-10 0 10 20 30 40 50 60 70 80 90 100 110

Figure 46: Supply function P = 10 + 0.5Q

c) From the graph in Figure 46 when P = 15, Q = 10. We confirm that through the
equation by substituting P = 15 and solving for Q

P = 10 + 0.5Q
15 = 10 + 0.5Q
0.5Q = 5
5
Q= = 10
0.5

d) The equation of the supply function in the form of Q = h(P ) is derived as follows:

P = 10 + 0.5Q
0.5Q = P − 10
P − 10
Q= = 2P − 20
0.5
The graph Q = h(P ) is given in Figure 47

Exercise 78. The demand and supply functions for hotel accommodation per room dur-
ing the low season are given by the equations:

Pd = 96 − 0.8Q
Ps = 40 + 0.4Q

where Q is the number of bookings and P is the price in Euro.

97
 40

30

20

10

0 10 20 30 40 50 60 70 80 90 100 110

-10

-20

P
Figure 47: Supply function Q = 2
− 20

a) Plot the demand and supply functions on the same diagram. Check your graphs with
Excel.

b) For each function:

i) Calculate the value of Q when P = 0.

ii) Calculate the price when Q = 0.

iii) How many bookings will be demanded and supplied when the price is 60 Euro per
room?

a) Figure 48 displays the two functions:

140

120

100

80

60

40

20

0 20 40 60 80 100 120 140 160 180 200

Figure 48: Pd = 96 − 0.8Q & Ps = 40 + 0.4Q

b) i) Qd = 120, Qs = −100, ii) Pd = 96, Ps = 40, iii) Qd = 45, Qs = 50

98
 ■

Exercise 79. It is known that the number of concert tickets demanded is 300 € when
the ticket price is 50 € and 200 € when the price is 60 €.

a) Assuming that demand is a linear function of price, determine the equation of the
demand function and plot the graph of the function.

b) Use the equation to calculate the change in demand when the price

a) increases by 5 €

b) decreases by 7 €.

Solution: a) Q = −10P + 800 is the demand function and in Figure 49 you can see its
graph.

900

800

700

600

500

400

300

200

100

-10 0 10 20 30 40 50 60 70 80 90

Figure 49: Q = −10P + 800

b) −50 and +70 ■

3.6 Cost

Cost is the amount of money a company needs to produce the items they are selling.
The total cost of producing a good normally consists of:

99
 • Fixed costs, FC: costs that are fixed irrespective of the level of output, e.g., rent
on premises.

• Variable costs, VC: costs which vary with the level of output, e.g., raw materials,
labour, electricity, etc.

Hence, total cost (TC) is the sum of fixed and variable costs:
TC = FC + VC
These costs can often be modelled by a linear cost function, such as
TC = 40 + 2Q
where the fixed costs equal to 40 (the vertical intercept) and the variable costs are given
by 2Q, where 2 is the slope of the line.

Example 51. Using a mobile phone costs 5 € per month, and an additional 0.05 € per
minute.

a) What is the cost for one month if the phone is used for a total of x minutes? Write
down the total cost function.

b) Graph the total cost function.

c) What are the smallest and largest number of minutes you can use the phone in a
month if the monthly telephone bill is to be between 10 € and 20 €?

Solution:

a) FC = 5 Euro and VC = 0.05x. With TC = FC + VC, the total cost incurred in using
the mobile phone is
TC(x) = 5 + 0.05x

b) The graph of the total cost function is illustrated in Figure 50.

c) The monthly telephone bill lies between 10 and 20 Euro, therefore, we have:
10 ≤ TC(x) ≤ 20
10 ≤ 5 + 0.05x ≤ 20
10 − 5 ≤ 0.05x ≤ 20 − 5
5≤ 0.05x ≤ 15
5 20
≤ x ≤
0.05 0.05
100 ≤ x ≤ 300

100
 100

90

80

70

60

50

40

30

20

10

0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500

Figure 50: TC(x) = 5 + 0.05x

The mobile phone can be used at least 100 and no more than 300 minutes.

3.7 Revenue

A company receives revenue when it sells output. The total revenue, TR, is calculated
as the price of the good, P , multiplied by the number of units sold, Q
TR = P · Q
If the price of each unit sold is the same, then total revenue can be modelled as a straight
line. For instance, if P = 20 then the total revenue function is linear
TR = 20 · Q
This line goes through the origin (0, 0) and the slope equals to 20.

Example 52. Suppose that crispy chicken box is sold for 2.90 € irrespective of the num-
ber of units sold.

a) Write down the equation of the total revenue function.

b) Graph the total revenue function.

Solution:

101
a) Total revenue is price multiplied by the number of units sold:

TR = 2.90Q

b) Total revenue is represented graphically by a straight line, which goes through the
origin (0, 0) (intercept = 0) and its slope is 2.90. The graph is illustrated in Figure
51.

30

25

20

15

10

0 1 2 3 4 5 6 7 8 9 10

Figure 51: TR = 2.90Q

3.8 Profit

Profit is the total revenue received less the total cost. We denote profit with the Greek
letter π, since the letter P is reserved to represent price. Hence, profit is given by the
formula

π = TR − TC

A company is said to “break even” when total revenue is the same as total cost and
hence profit is zero. At break-even,

π=0
TR = TC

When:

i) TR > TC the company makes a profit

ii) TR < TC the company makes a loss.

102
Example 53. A local newspaper has fixed production costs of 70 € per edition and vari-
able printing and distribution costs of 0.40 € per copy. The newspaper sells for 0.50 €
per copy.

a) Write down the equation for i) total revenue, ii) total costs, iii) profit.

b) What profit (or loss) results from the sale of 500 copies?

c) Calculate the number of newspapers that must be sold in order to break even.

d) Graph the profit function. From the graph estimate the break-even quantity.

Solution:

a) i) TR = 0.50Q,
ii) TC = 0.40Q + 70,
iii)

π = TR − TC
π = 0.50Q − (0.40Q + 70)
π = 0.10Q − 70

b) We can find the profit (or loss) that results from selling 500 copies by plugging 500
into the profit function. We get

π(500) = 0.10 · 500 − 70 = −20

This means that if 500 newspapers are sold, it will result in a loss of 20 €.

c) To find the break even quantity, we can set profit equal to zero and solve for Q

π=0
0.10Q − 70 = 0
70
Q=
0.10
Q = 700

In order to break-even 700 newspapers must be sold.

d) Profit is represented graphically by a straight line π = 0.10Q − 70 in Figure 52.

103
 400

300

200

100

-1500 -1000 -500 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000

-100

-200

Figure 52: π = 0.10Q − 70

Example 54. A company is thinking about entering the sports climbing market and
wants to sell innovative belay devices. The estimated variable costs (labour, materials,
marketing) per unit amount to 20 €, the fixed costs (machinery, loans, management)
amount to 270 000 €. Selling price will be 30 € per piece. Thorough market research
shows that at this price the company is able to sell 25 000 units altogether over the life
of the project.

a) Write down the equation for i) total revenue and ii) total costs.

b) Plot these functions on the same diagram. Check your graphs with Excel.

c) Express the (natural) domain and the range in set notation.

d) Write down the equation for profit.

e) Calculate the number of items that must be sold in order to break even (break-even
point).

f ) Estimate the total profit when all planned units can actually be sold.

g) What is your advice for the company?

Solution:

a) i) TR = 30x, ii) TC = 270 000 + 20x

b) The TR and TC graphs are shown in Figure 53.

c) TR : (0, ∞) → (0, ∞), TC : (0, ∞) → (270 000, ∞)

104
 2000000

1800000

1600000

1400000

1200000

1000000

800000

600000

400000

200000

0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000 55000 60000

Figure 53: TR = 30x and TC = 270 000 + 20x

d) π = TR(x) − TC(x) = 30x − (270 000 + 20x) = 10x − 270 000

e) In order to calculate the break-even point we set the TC and TR equal:

TC(x) = TR(x)
270 000 + 20x = 30x
30x − 20x = 270 000
10x = 270 000
x = 27 000

f ) π(25 000) = 10 · 25 000 − 270 000 = −20 000

g) Produce more!

Example 55. Find the total revenue function TR(Q) = P (Q) · Q for a producer facing
the following linear demand function:
P (Q) = 100 − 2Q,
where P stands for the price of a certain product and Q for the quantity of products sold.
Graph the function for 0 ≤ Q < 55 and interpret your findings economically.

Solution: In order to calculate the total revenue we substitute P (Q) into the equation
TR(Q) = P (Q) · Q
TR(Q) = (100 − 2Q) · Q
TR(Q) = 100Q − 2Q2

105
The total revenue is given by a quadratic function. In order to plot its curve, we solve
the quadratic equation and calculate the roots.

100Q − 2Q2 = 0
2Q · (50 − Q) = 0

Hence, Q = 0 or Q = 50. Additionally we calculate the vertex of the parabola

b 100
− =− = 25
2a −4
The graph of total revenue is illustrated in the following diagram (Figure 54):

1400

1200

1000

800

600

400

200

0 5 10 15 20 25 30 35 40 45 50 55

Figure 54: TR = 100Q − 2Q2

When 25 units are produced and sold total revenue reaches its highest point. ■

Exercise 80. A monopolist defines the price in relation to the quantity as

p(x) = 51 − 2x.

His average costs are

x 105
AC(x) = +9+ .
3 x
Calculate:

a) the profit function π(x),

b) the break-even points and

106
c) the function that represents the quantity as function of the price.

Note that: AC(x) = TC(x)

x

Solution: a) π(x) = − 37 x2 + 42x − 105, b) between 3 and 15, c) x(p) = 51−p

2
■

Exercise 81. Question b) from Exercise 80 can be solved with two different ways with
the use of Excel. Try them out! ■

Exercise 82. A public transportation company sells a single ticket for 2 € and carries
500 000 persons a day. Market research shows that if the price increases 0.70 € then
10 000 people less would buy a ticket.

a) Find the function of a(x) which shows the number of carried persons as a function of
the price x. A linear relationship is assumed.

b) Display the function of the daily revenues subject to the price (r(x)).

c) Find the price which maximizes the daily revenues. Check your result with Excel.

Solution: a) a(x) = −1 000

7
000
x + 785 714.28, b) r(x) = −1 000 000 2
7
x + 785 714.28x, c)
x = 2.75 and a(2.75) = 392 857.1 ■

Exercise 83. The total revenue for an accounting journal is given by the equation

TR = −Q2 + 1 800Q − 44 375

where Q is the number of journals sold.

a) How many journals are sold when TR is zero?

b) How many journals are sold when TR is 765 625 €?

Solution: a) Q1 = 1 775 and Q2 = 25, b) 900 ■

Exercise 84. A firm charges a fixed price of 80 € for each product sold. The firm has
a total cost function

TC = Q3 − 136Q

a) Write down the total revenue and the profit function.

107
b) Determine the break-even point(s). Check your results with Excel.
√
Solution: a) TR = 80Q and π = −Q3 + 216Q, b) Q1 = 0, Q2,3 = ± 216 ■

Exercise 85. An average cost function is given by the equation

AC = 1 200 − 128Q + 4Q2

a) Find the values of Q at which the average cost is zero. Explain why this AC function
is always positive.

b) Sketch the average cost function over the range Q = 0 to Q = 35.

c) From the graph write down the minimum value of the average cost and the value of
Q at which this minimum occurs.

Solution: a) ∆ = (−128)2 − 4 · 4 · 1 200 = −2 816 < 0 hence, there are no real solutions.
a = 4 > 0 → parabola upwards, b) The average cost function is illustrated in Figure 55.
c) AC(16) = 176 ■
1600

1400

1200

1000

800

600

400

200

0 5 10 15 20 25 30 35 40

Figure 55: AC = 1 200 − 128Q + 4Q2

Exercise 86. Solve Exercise 85 with Excel. ■

Exercise 87. Given the following total revenue and total cost functions:

TR(Q) = 550Q − 3Q2 ,

TC(Q) = 85Q + 9 500.

108
a) Express profit π as a function of Q,

b) Determine the break-even point and the region of positive profit by finding the roots
of π(Q),

c) Find the value Qmax where the profit is maximised and calculate the maximum profit.

Check your results with Excel.

Solution: a) π(Q) = −3Q2 + 465Q − 9 500, b) Q1 = 130.78 and Q2 = 24.21, c) π(77.5) =

8 518.75 ■

Exercise 88. A firm’s profit π has been growing consistently over time from 10.8 million
in 2010 to 12.7 million in 2013.

a) Express profits as a natural exponential function of time, i.e. π(t) = π0 · ert , and find
the annual rate of growth r.

b) What is the expected profit in 2014?

c) When will the profit double?

Solution: a) r = 0.0540, b) π(4) = 13.405 Mio., c) double in 2023. ■

3.9 Equilibrium and Break-even

With the use of the method of simultaneous equations we can investigate the equilibrium
conditions in various markets; for instance, the goods, labour and money markets. In
the following we discuss some applications from the goods market.

Goods market equilibrium (market equilibrium) occurs when the quantity demanded
(Qd ) by consumers and the quantity supplied (Qs ) by producers of a good or service
are equal. Equivalently, market equilibrium occurs when the price that a consumer is
willing to pay (Pd ) is equal to the price that a producer is willing to accept (Ps ).

Example 56. The demand and supply functions for a haircut are given by:

Pd = 50 − 3Q
Ps = 14 + 1.5Q

where P is the price of a haircut and Q is the number of haircuts.

109
a) Calculate the equilibrium price and quantity.

b) Calculate the level of excess supply (i.e. Qs − Qd ) when P = 38.

c) Calculate the level of excess demand (i.e. Qd − Qs ) when P = 20.

Solution:

a) Market equilibrium occurs when Qd = Qs and Pd = Ps . Since the functions are

written in the form P = f (Q) we equate prices, thereby reducing the system to an
equation in Q only; hence, solve for Q:

Pd = Ps
50 − 3Q = 14 + 1.5Q
50 − 14 = 3Q + 1.5Q
36 = 4.5Q
36
Q= =8
4.5
The equilibrium quantity is Q = 8. Now we solve for the equilibrium price by substi-
tuting in either Pd or Ps :

Pd = 50 − 3 · 8 = 26

b) Level of excess supply when P = 38. We substitute the price in the demand function

38 = 50 − 3Qd
3Qd = 50 − 38
3Qd = 12
Qd = 4

and then we work accordingly with the supply function

38 = 14 + 1.5Qs
1.5Qs = 38 − 14
24
Qs = = 16
1.5
Hence, Qs − Qd = 16 − 4 = 12. The level of excess supply is 12 haircuts.

c) Level of excess demand when P = 20. Firstly we substitute the given value in the

110
 demand function

20 = 50 − 3Qd
3Qd = 50 − 20
3Qd = 30
Qd = 10

and then in the supply function

20 = 14 + 1.5Qs
1.5Qs = 20 − 14
1.5Qs = 6
Qs = 4

Hence, Qd − Qs = 10 − 4 = 6. The level of excess demand is 6 haircuts.

Example 57. The demand and supply functions for golf lessons at Greens Club are are
given by the equations:

Pd = 200 − 5Q
Ps = 92 + 4Q

a) Calculate the equilibrium price and quantity algebraically and graphically. Check your
graphical solution with Excel.

b) The government imposes a tax of 9 € per lesson:

i) Write down the equation of the supply function adjusted for tax, hence plot it on
the same diagram as before. Check your solution with Excel.

ii) Calculate the equilibrium price and quantity when the tax is imposed.

iii) Outline the distribution of the tax, that is, how much of the tax is paid by the
customer and the club (supplier).

Solution:

111
a) In order to estimate the equilibrium we equate prices, Pd = Ps and then solve for Q:

Pd = Ps
200 − 5Q = 92 + 4Q
200 − 92 = 5Q + 4Q
108 = 9Q
108
Q= = 12
9
The equilibrium quantity is Q = 12. Now we solve for the equilibrium price by
substituting in either Pd or Ps :

Pd = 200 − 5 · 12 = 140

The equilibrium point is (Q = 12, P = 140).

For the graphical solution we plot the curves of the two functions, Pd (Q) and Ps (Q),
on the same diagram (see Figure 56).

220

200

180

160

140

120

100

80

60

40

20

-2 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

Figure 56: Pd = 200 − 5Q and Ps = 92 + 4Q

The solution of the equation system is the point of intersection of the two lines, i.e.
the equilibrium point (12, 140).

b) i) When 9 € of tax is applied then the new supply equation is

Psnew = 101 + 4Q

We plot Psnew = 101 + 4Q on the same diagram (see Figure 57).

112
 220

200

180

160

140

120

100

80

60

40

20

-2 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

Figure 57: Pd = 200 − 5Q, Ps = 92 + 4Q and Psnew = 101 + 4Q

ii) In order to calculate the equilibrium price and quantity after the tax is imposed,
we equate prices

Pd = Psnew
200 − 5Q = 101 + 4Q
200 − 101 = 5Q + 4Q
99 = 9Q
99
Q= = 11
9
For the equilibrium price we substitute in Psnew and we have:

Psnew = 101 + 4 · 11 = 145

Therefore, the new equilibrium point is (11, 145).

iii) By comparing the two equilibrium points (12, 140) and (11, 145) we can deduce
the tax distribution. The customer from 140 € pays now 145 € that means 5 €
increase for the customer. For the seller is the new price without tax 136 €
(145 − 9 = 136) and in comparison to the previous price (140 €) we have a 4 €
decrease.

Exercise 89. The demand and supply functions for two complementary products X and
Y , pitching wedges and putters (for pitch and putt), respectively, are given as:

QdX = 190 − 2PX − 2PY

QdY = 240 − 2PX − 4PY

113
and

QsX = −10 + 2PX

QsY = −40 + PY

Find the equilibrium price and quantity for each good.

Solution: Pitchers, X: PX = 27.5, QX = 45. Putters, Y : PY = 45, QY = 5 ■

Exercise 90. The demand and supply functions for a spa holiday package are

Qd = 81 − 0.05P
Qs = −24 + 0.025P

a) Calculate the equilibrium price and quantity, algebraically and graphically.

b) Graph the supply and demand function, showing the equilibrium.

c) The government imposes a tax of 120 € on each spa holiday.

i) Write down the equation of the supply function adjusted for tax, hence graph it
on the same diagram as before.

ii) Calculate the equilibrium price and quantity when the tax is imposed.

iii) Outline the distribution of the tax, that is, calculate the tax paid by the consumer
and by the travel agent.

Check your graphical solutions with Excel.

Solution: a) P = 1 400, Q = 1 620 − 20Q, b) Pd = 1 620 − 20Q, Ps = 960 + 40Q, c)

i) Psnew = 1 080 + 40Q, ii) Q = 9, P = 1 440, iii) Consumer pays 40 €, company pays
80 €. The graphical solutions are illustrated in Figure 58. ■

114
3500

3000

2500

2000

1500

1000

500

0 5 10 15 20 25 30 35 40

Figure 58: Pd , Ps and Psnew

115
4 Calculus in Economic Applications

The interested reader can find more about this topic in (Bradley, 2013)[Section 6]. Many
explanations and examples are taken from this book.

In this section we will introduce derivatives of functions. The corresponding math-

ematical field is called (differential) calculus. The most prominent application of
derivatives is finding maxima and minima. E.g. if the profit function, depending on
the quantity, is given, it can be advantageous to know the quantity such that the profit
is maximal, see Figure 59.

Figure 59: Profit function π depending on quantity Q

This is only one of many important applications of derivatives. Due to the fact that
derivatives have something to do with change, we will find a relation to the economic
concept of marginal functions. We will be able to distinguish between a so-called
progressive and degressive behaviour and last but not least we will discuss elasticity,
again concepts which are used in the field of economics.

All these topics have something to do with approximating complicated functions by

simple functions - the so called tangent, see Figure 60.

Differential calculus was mainly introduced by Gottfried Wilhelm Leibniz (German

mathematician and philosopher 1646-1716) and Sir Isaac Newton (English mathemati-
cian, physicist, etc. 1642-172712 ), see Figure 61 and 62.13

12
During his lifetime two calendars were used: the Julian (“Old Style”) calendar and the Gregorian
(“New Style”). In Newton’s case 1727 is the year of death in “New Style”. 1726 is the year of death
in “Old Style”.
13
Pictures were taken from www.wikipedia.com.

116
Figure 60: Approximation of function f by tangent g in (2, 0.4)

Figure 61: Leibniz Figure 62: Newton

117
In this section we are dealing with economic functions which describe e.g. the total
revenue, the profit or the total costs. All these functions depend on the quantity. Later
we also deal with functions which describe the demand, which depends on the price.
Of course we can use the notation f (x) for all functions, but it is more common and
self-explanatory to use Q instead of x for the quantity and P instead of x for the price.
If we focus on the total revenue, we use the notation TR(Q). For the profit function
we use π(Q) and for the total cost function TC(Q). If we consider demand functions,
we use the notation Q(P ), as the demand is nothing else then the demanded quantity.
Of course there will be some other economic functions, but the idea stays the same.
In contrast to that we will use the notation f (x) in most cases (sometimes also g(x),
a(x), b(x), etc.) if there is no application behind or if we discuss theory for general
functions.

4.1 Calculus Basics

Starting point is a geometric problem: Let’s consider a function f which maps real
valued x to real valued y. We use the notation y = f (x). Let x0 be arbitrary but fixed
and y0 = f (x0 ) the corresponding function value. We now want to find a straight line g
which satisfies the following two properties:

• g(x0 ) = y0 , i.e. g passes the point (x0 , y0 ),

• g approximates the function f for x close to x0 best. I.e. g(x) ≈ f (x) if x ≈ x0 .

In Figure 63 we can see the function f (green) and some straight lines (blue, red,
orange), which satisfy the first property. These straight lines - the secants - have a
second intersection point (x1 , y1 ), with y1 = f (x1 ). But none of the secants satisfies the

Figure 63: Secants of f (green) left and corresponding zoom right

second property. If we have a look at Figure 63, we find out that the closer x1 gets to
x0 , the better the approximation (at least in a small area around x0 ). In Figure 63 the
red secant approximates f better than the blue and the orange one.

118
To get a better understanding of the secants we notice, that a secant is a straight line
and therefore there holds
g(x) = k · x + d,
where k is the slope of g and d the intercept. The secant g has two intersection points
with f , i.e. there holds
g(x0 ) = f (x0 ) and g(x1 ) = f (x1 ).
Summing up we have to solve two equations:
k · x0 + d = f (x0 ) (29)
k · x1 + d = f (x1 ). (30)
Subtracting (29) from (30) we obtain
k · x1 − k · x0 = f (x1 ) − f (x0 ),
or analogous
k · (x1 − x0 ) = f (x1 ) − f (x0 ).
As x0 ̸= x1 we can divide by (x1 − x0 ) and get
f (x1 ) − f (x0 )
k= . (31)
x1 − x0
With the knowledge of k it is easy to get the missing intercept d. We use (29), get
d = f (x0 ) − k · x0 (32)
and know everything about the secant.

Example 58. Lets consider the function f (x) = 0.1x2 and the x-values x0 = 2 and
x1 = 3. We want to find the corresponding secant.

Using (31) we get:

0.1 · 32 − 0.1 · 22 0.5
k= = = 0.5.
3−2 1
Using (32) we get:
d = 0.1 · 22 − 0.5 · 2 = −0.6.
The secant is given by
g(x) = 0.5x − 0.6.
In Figure 64 we can verify the result. ■

119
 Figure 64: Secant

As x1 gets closer and closer to x0 we get better and better approximations by secants of
f in an area around x0 and in the limit case we get the slope:14
f (x1 ) − f (x0 )
k = lim .
x1 →x0 x1 − x0
This formula is very important and we call this the differential quotient. We use the
notation
f (x1 ) − f (x0 )
f ′ (x0 ) = lim , (33)
x1 →x0 x 1 − x0
so we get

k = f ′ (x0 ). (34)

f ′ (x0 ) is called the first derivative of f in x0 . Next it is easy to get the missing
intercept d. We use (29) and get

d = f (x0 ) − f ′ (x0 ) · x0 .

As k can be (hopefully) calculated by using (34), the straight line is well defined and we
get the tangent

g(x) = f ′ (x0 ) · x + f (x0 ) − f ′ (x0 ) · x0

and hence

g(x) = f (x0 ) + f ′ (x0 ) · (x − x0 ). (35)

Remark. In advance it is not clear if the limit in (33) exists. However, for most
functions this works. ■
14
lim means that x1 tends to x0 , lim is an abbreviation for limit
x1 →x0

120
Remark. Sometimes x1 is written in the form x1 = x + ∆x, where ∆x is defined by
∆x = x1 − x0 . Then (33) modifies to
f (x0 + ∆x) − f (x0 )
f ′ (x0 ) = lim . (36)
∆x→0 ∆x
■

In the whole section x0 itself was arbitrary, hence we can vary x0 and rename it by x.
Therefore we get the function f ′ (x) and call this the first derivative of f .

As the first derivative of f is a function, it is possible to define the first derivative of f ′ (x)
and call this the second derivative of f (x). We use the notation f ′′ (x) for the second
derivative. The same argument can be used to define the third, forth derivative and
so on.

Going back to our tangent-problem, the remaining question is how to get the value of
f ′ (x). The answer which is not satisfactory is to use (33) and compute it by hand. For
standard functions however the differential quotient (33) is well known and we can use
Table 20 to get the first derivatives.

f (x) f ′ (x)
c 0
x 1
x2 2·x
.. ..
. .
xn n · xn−1 , with n ∈ {2, 3, . . .}
ex ex
1
ln(x) x
, for x > 0

Table 20: Derivatives of standard functions

Remark. In Table 20 we claimed that n ∈ {2, 3, . . .}. We can generalise to n ∈ R but

then we have to consider the following:

• If n = 0 or n = 1 the formula is not really nice because it says:

′
x0 = 0 · x−1 for n = 0,
′
x1 = 1 · x0 for n = 1.

For this two cases it is easier to remember 1′ = 0 and x′ = 1 (as mentioned in

Table 20) than worrying about 00 or 00 .

121
 • If n ∈ {−1, −2, . . .} we have to make sure that x ̸= 0 (division by zero otherwise).
E.g. for n = −1 we get

′
x−1 = −x−2 .

On the left hand side we have x−1 = 1

x
and on the right hand side −x−2 = −1
x2
. In
both cases we have to exclude x = 0.

• If n ∈ R \ Z and n > 1 we have to make sure that x ≥ 0 (negative radicand

otherwise). E.g. for n = 1.5 we get

′
x1.5 = 1.5 · x0.5 .
√ √
On the left hand side we have x1.5 = x3 and on the right hand side x0.5 = x.
In both cases we have to exclude negative x-values.

• If n ∈ R \ Z and n < 1 we have to make sure that x > 0 (negative radicand or

division by zero otherwise). E.g. for n = 0.5 we get

′
x0.5 = 0.5 · x−0.5 .
√
On the left hand side we have x0.5 = x and on the right hand side 0.5·x−0.5 = √
0.5
x
.
In both cases we have to exclude negative x-values and for the right part we have
to exclude x = 0 too.

Remark. For interested reader we want to use (36) to get the derivative of the function
f (x) = x2 by hand:

′ f (x0 + ∆x) − f (x0 ) (x0 + ∆x)2 − x20

f (x0 ) = lim = lim
∆x→0 ∆x ∆x→0 ∆x
2 2 2
x0 + 2x0 ∆x + ∆x − x0 ∆x · (2x0 + ∆x)
= lim = lim
∆x→0 ∆x ∆x→0 ∆x
= lim (2x0 + ∆x) = 2x0 .
∆x→0

As x0 was arbitrary, we have proven the important rule

′
x2 = 2x.

We will learn some additional convenient rules, but before that we have a look at some
examples:

122
Example 59. Let’s consider the function f (x) = x3 . We want to find the tangent of f
in the point (2, f (2)).

We can use (35) to get the tangent. In our situation there holds x0 = 2, y0 = f (x0 ) =
f (2) = 23 = 8 and from Table 20 we know that f ′ (x) = 3 · x2 . Therefore we get
f ′ (x0 ) = f ′ (2) = 3 · 22 = 12. Summing up, the tangent is given by

g(x) = 8 + 12 · (x − 2) = 12x − 16.

In Figure 65 we can see the correctness of the calculation. Please note the different
scaling for the x- and the y-axis. ■

Figure 65: Tangent of f

Example 60. Compute the first and the second derivative of

a) f (x) = x4 ,

b) g(x) = ex ,

c) h(x) = ln x, for x > 0.

Looking at Table 20 we get

a) f ′ (x) = 4x3 and f ′′ (x) = 12x2 ,

b) g ′ (x) = ex and g ′′ (x) = ex (this is nice),

c) h′ (g) = 1
x
= x−1 and h′′ (x) = −1 · x−2 = −1
x2
.

123
So far we can only handle functions from Table 20 but we have no rules to differentiate
e.g. 5x. It is time to extend our pool of function with known derivatives.

The first rule is very easy and is called sum rule: Let f1 and f2 be two functions, then

[f1 (x) + f2 (x)]′ = f1′ (x) + f2′ (x). (37)

The second rule is as easy as the first one: Let f be a function and c a constant factor,
then

[c · f (x)]′ = c · f ′ (x). (38)

Example 61. Compute the derivatives of the following functions:

a) a(x) = 8 + 2x + 3x5 ,

b) b(x) = 3x + ex ,

c) c(x) = ln(x) + 5, for x > 0.

Applying both rules we get

′ ′
a) a′ (x) = 8′ + (2x)′ + (3x5 ) = 8′ + 2x′ + 3 · (x5 ) = 0 + 2 · 1 + 3 · 5 · x4 = 2 + 15x4 .
Maybe this was a little bit too slow. We will skip all this steps in future.

b) b′ (x) = 3 + ex ,

c) c′ (x) = 1
x

So far so good, but still we do not know how the derivatives of e.g. x · ex (product),
2x 2
1+x2
(fraction) or e2+2x (composition) looks like. The product rule helps us in the
first case: Let f1 and f2 be two functions, then

[f1 (x) · f2 (x)]′ = f1′ (x) · f2 (x) + f1 (x) · f2′ (x). (39)

Example 62. Compute the derivatives of the following functions:

a) a(x) = (2 + 4x + 2x4 ) · ex ,

124
b) b(x) = (4 + 2x4 ) · (2 + 2x2 ).

Using the product rule (and of course the sum rule and the constant factor rule too) we
get

a) The derivative of 2 + 4x + 2x4 is 4 + 8x3 and the derivative of ex is again ex so we

get

a′ (x) = 4 + 8x3 · ex + 2 + 4x + 2x4 · ex

= 6 + 4x + 8x3 + 2x4 · ex .

b) The derivative of 4 + 2x4 is 8x3 and the derivative of 2 + 2x2 is 4x, so we get

b′ (x) = 8x3 · 2 + 2x2 + 4 + 2x4 · 4x

= 16x3 + 16x5 + 16x + 8x5

= 16x + 16x3 + 24x5 .

This function can be derived without the product rule as well: Expanding b(x) we get

b(x) = 8 + 8x2 + 4x4 + 4x6 .

So we can use the sum rule and the constant factor rule and get:

b′ (x) = 16x + 16x3 + 24x5 ,

which is the same result.

Next we have a look at the quotient rule: Let f1 and f2 be two functions, then we
get
′
f ′ (x) · f2 (x) − f1 (x) · f2′ (x)

f1 (x)
= 1 . (40)
f2 (x) [f2 (x)]2

Example 63. Compute the derivatives of the following functions:

8+2x+2x4
a) a(x) = 1+x2
,

2·ex
b) b(x) = 2x+4x2
.

Here we use the quotient rule together with the sum rule and constant factor rule.

125
a) The derivative of 8 + 2x + 2x4 is 2 + 8x3 and the derivative of 1 + x2 is 2x, so we get
(2 + 8x3 ) · (1 + x2 ) − (8 + 2x + 2x4 ) · 2x
a′ (x) =
(1 + x2 )2
2 + 8x3 + 2x2 + 8x5 − (16x + 4x2 + 4x5 )
=
1 + 2x2 + x4
2 − 16x − 2x + 8x3 + 4x5
2
= .
1 + 2x2 + x4

b) The derivative of 2 · ex is 2 · ex and the derivative of 2x + 4x2 is 2 + 8x. So we get

2 · ex · (2x + 4x2 ) − 2 · ex · (2 + 8x)
b′ (x) =
(2x + 4x2 )2
ex · (4x + 8x2 − 4 − 16x)
=
4x2 + 16x3 + 16x4
e · (−4 − 12x + 8x2 )
x
= .
4x2 + 16x3 + 16x4

The last rule we want to introduce is called chain rule: Let f1 and f2 be two functions
and let a third function f be of the form f (x) = f1 (f2 (x)) (this is called composition)
then the derivative is given by
f ′ (x) = f1′ (f2 (x)) · f2′ (x). (41)
f1′ is called the outer derivative and f2′ (x) is called the inner derivative.

Example 64. Compute the derivatives of the following functions:

a) a(x) = e2x ,

b) b(x) = ex+x ,
2

c) c(x) = ln (1 + x2 ).

Additional to the chain rule we use the sum rule as well as the constant factor rule (as
always).

a) Lets define a1 (z) = ez and a2 (x) = 2x, then we get a(x) = a1 (a2 (x)). The derivative
of a1 is given by ez and the derivative of a2 is given by 2. Using the chain rule we
get:
a′ (x) = e2x · 2 = 2 · e2x .

126
b) Lets define b1 (z) = ez and b2 (x) = x+x2 , then we get b(x) = b1 (b2 (x)). The derivative
of b1 is given by ez and the derivative of b2 is given by 1 + 2x. Using the chain rule
we get:
2
b′ (x) = ex+x · (1 + 2x) .

c) Lets define c1 (z) = ln(z) and c2 (x) = 1 + x2 , then we get c(x) = c1 (c2 (x)). The
derivative of c1 is given by x1 and the derivative of c2 is given by 2x. Using the chain
rule we get:
1 2x
c′ (x) = 2
· 2x = .
1+x 1 + x2

Example 65. Compute the derivative of

f (x) = 2x .

In Table 20 there is no rule for ax . But if we modify f we get15

f (x) = 2x = ex·ln(2) .

Therefore we can use the chain rule and get f ′ (x) = ex·ln(2) · ln(2) = 2x · ln(2). ■

It is also possible to combine all five rules. So we have a large enough toolbox to handle
derivatives.

Example 66. Compute the derivatives of the following functions:

x·e2x
a) a(x) = 1+x2
,
 
b) b(x) = ln 2 + e1+x+2x .
2

Additionally, find the tangents for x0 = 0.

15
We use the rule ax = ex·ln(a) for a > 0.

127
a) Using all five rules we get
(1 · e2x + x · e2x · 2) · (1 + x2 ) − x · e2x · 2x
a′ (x) =
(1 + x2 )2
e2x · [(1 + 2x) · (1 + x2 ) − 2x2 ]
=
1 + 2x2 + x4
e · (1 + 2x − x2 + 2x3 )
2x
= .
1 + 2x2 + x4
For the tangent we compute: a(0) = 0, a′ (0) = 1 and therefore the tangent is given
by g(x) = x, see Figure 66.

b) Using all but the quotient rule we get

1 2
b′ (x) = · e1+x+2x · (1 + 4x) .
2+ e1+x+2x2
For the tangent we compute: b(0) = 2 + e = 1.551 . . ., b′ (0) = 2+e e
= .576 . . . and
therefore the tangent is given by g(x) = 1.551 . . . + 0.576 . . . · x, see Figure 67.

Figure 66: Tangent of a Figure 67: Tangent of b

Example 67. Compute the tangent for

f (x) = x3 − 10x2 + x + 100

and x0 = 2. What is the absolute and relative error if we consider x = 3 and if we use
the tangent instead of f ?

f (2) = 70 and f ′ (x) = 3x2 − 20x + 1, so f ′ (2) = −27. The tangent is therefore given by

g(x) = f (2) + f ′ (2) · (x − 2) = 70 − 27 · (x − 2) = 124 − 27x.

128
If we consider x = 3 we get f (3) = 40 and g(3) = 43. The absolute error is therefore
given by |g(3) − f (3)| = |43 − 40| = |3| = 3 and the relative error is
g(3) − f (3) 3
= = 0.075 ≈ 7.5%.
f (3) 40
See Figure 68 to check the correctness of the tangent. ■

Figure 68: Tangent

Exercise 91. Lets consider the function f (x) = x2 − x3 and the x-values x0 = −2 and
x1 = 2. Find the corresponding secant.

Solution: g(x) = 4 − 4x ■

Exercise 92. Compute the first 3 derivatives of the following functions:

a) a(x) = 2 + 4x2 + 2x5 ,

b) b(x) = x + x2 ,

c) c(x) = 42x

Solution:

a) a′ (x) = 8x + 10x4 , a′′ (x) = 8 + 40x3 , a′′′ (x) = 120x2 ,

b) b′ (x) = 1 − 2
x2
, b′′ (x) = 4
x3
, b′′′ (x) = −12
x4
,

c) c′ (x) = 42x · 2 ln(4), c′′ (x) = 42 x · 4 ln2 (4), c′′′ (x) = 42x · 8 ln3 (4).

129
Exercise 93. Compute the first and the second derivatives of the following functions:

x2 −4
a) a(x) = x−2
,

(2x−3)2
b) b(x) = (x−1)2
,

e2x
c) c(x) = x

Solution:

a) a′ (x) = 1, a′′ (x) = 0 (use standard methods first, in a second try simplify the function
before calculating the derivatives)

b) b′ (x) = 4x−6
(x−1)3
, b′′ (x) = 14−8x
(x−1)4
,

e2x ·(2x−1) 2e2x ·(2x2 −2x+1)

c) c′ (x) = x2
, c′′ (x) = x3
.

Exercise 94. Find the tangents of the following functions for

a) a(x) = 2x2 + 4 with x0 = 1,

b) b(x) = ln x with x0 = 1,

c) c(x) = ln(x + 1) with x0 = 0,

d) d(x) = 1
x
with x0 = 1.

Solution:

a) a(1) = 6, a′ (x) = 4x, a′ (1) = 4, ga (x) = 1 + 4x,

b) b(1) = 0, b′ (x) = x1 , b′ (1) = 1, gb (x) = x − 1

c) c(0) = 0, c′ (x) = 1
x+1
, c′ (0) = 1, gc (x) = x

d) d(1) = 1, d′ (x) = −1
x2
, d′ (1) = −1, gd (x) = 2 − x.

130
Exercise 95. Compute the tangent for

f (x) = −2x3 + 20x2 − 2x − 200

and x0 = 2. What is the absolute and relative error if we consider x = 3 and if we use
the tangent instead of f ?

Solution: g(x) = −248 + 54x, 6, 8% ■

4.2 Marginal Functions

As already mentioned we are interested in economic functions like the total revenue
TR(Q), the profit function π(Q) or the total cost function TC(Q). To be able to handle
all functions at once, we use f (x) instead. So let’s assume that an economic function
f (x) is given. We are now interested in the following task:

Let’s assume that x settled at a fixed value x0 with corresponding function

value f (x0 ). What increase/decrease of f can be expected if x0 increases by
1?

In Section 4.1 we have seen that functions f (x) can be approximated by tangents. To do
so we fix x0 and compute the tangent by using (35). Therefore we get the approximation
(for x close to x0 ):

f (x) ≈ f (x0 ) + f ′ (x0 ) · (x − x0 ).

If we are interested in the behaviour of f for x = x0 +1 we get x−x0 = 1 and therefore

f (x0 + 1) ≈ f (x0 ) + f ′ (x0 ).

Hence we see, that the difference of the function values between the “new” case x0 + 1
and the “old” case x0 can be approximated by using the derivative:

∆f = f (x0 + 1) − f (x0 ) ≈ f ′ (x0 ).

In economy the first derivative is called marginal function. So if we consider total

cost function TC(Q), we use the terminology marginal costs MC(Q) for TC(Q)′ , if
we consider the total revenue function TR(Q), we use marginal revenue MR(Q) for
TR′ (Q), etc.

We are not restricted to increase x only by 1. We also get an approximation of the

change ∆f if we increase x by ∆x:

f (x + ∆x) ≈ f (x0 ) + f ′ (x0 ) · ∆x

131
and therefore

∆f = f (x + ∆x) − f (x0 ) ≈ f ′ (x0 ) · ∆x.

But of course ∆x has to be “small” to get a good approximation. Values ∆x which are
said to be small depend on the function f .

Remark. This is the first time we derived theory by using the notation f (x) and use a
different notation for the applications. So don’t be confused! ■

Example 68. Given the total cost functions TC depending on the quantity Q:

a) TC(Q) = 10 + 4Q,

b) TC(Q) = 120Q − 8Q2 + 31 Q3 .

First compute the change in TC exactly and second approximate this calculation if Q
increases

• from 15 to 16,

• from 20 to 21,

• from 15 to 20,

• from 15 to 14.

Solution:

a) We compute the exact results first:

• ∆Q = 1, TC(15) = 70 and TC(16) = 74, therefore ∆TC = 4,

• ∆Q = 1, TC(20) = 90 and TC(21) = 94, therefore ∆TC = 4,

• ∆Q = 5, TC(15) = 70 and TC(20) = 90, therefore ∆TC = 20,

• ∆Q = −1, TC(15) = 70 and TC(14) = 66, therefore ∆TC = −4.

Next we use the marginal costs MC(Q) = TC′ (Q) = 4, so we get MC(15) = MC(20) =
4 and:

• ∆Q = 1, ∆TC ≈ MC(15) · ∆Q = 4 · 1 = 4,

132
 • ∆Q = 1, ∆TC ≈ MC(20) · ∆Q = 4 · 1 = 4,

• ∆Q = 5, ∆TC ≈ MC(15) · ∆Q = 4 · 5 = 20,

• ∆Q = −1, ∆TC ≈ MC(15) · ∆Q = 4 · (−1) = −4.

All approximations are exact. For straight lines this is always the case. We also see
that MC does not depend on Q. If we ask the same question, but instead of 15 and
16 we use 20 and 21 we again get ∆TC = 4.

b) We compute the exact results first:

• ∆Q = 1, TC(15) = 1 125 and TC(16) = 1 237.3̄, therefore ∆TC = 112.3̄,

• ∆Q = 1, TC(20) = 1 866.6̄ and TC(21) = 2 079, therefore ∆TC = 212.3̄,

• ∆Q = 5, TC(15) = 1 125 and TC(20) = 1 866.6̄, therefore ∆TC = 741.6̄,

• ∆Q = −1, TC(15) = 1 125 and TC(14) = 1 026.6̄, therefore ∆TC = −98.3̄.

Next we use the marginal costs MC(Q) = TC′ (Q) = 120 − 16Q + Q2 , so we get
MC(15) = 105, MC(20) = 200 and:

• ∆Q = 1, ∆TC ≈ MC(15) · ∆Q = 105 · 1 = 105,

• ∆Q = 1, ∆TC ≈ MC(20) · ∆Q = 200 · 1 = 200,

• ∆Q = 5, ∆TC ≈ MC(15) · ∆Q = 105 · 5 = 525,

• ∆Q = −1, ∆TC ≈ MC(15) · ∆Q = 105 · (−1) = −105.

If we compare with the exact results we find out that the approximation is not exact
anymore. This is because TC is no straight line. Also we see that MC depends on Q,
therefore it makes a difference if we consider the situation from 15 and 16 or 20 and
21. Additionally we compute the relative errors:

105−112.3̄
• 112.3̄
= 0.06528 . . . ≈ 6.53%,

200−212.3̄
• 212.3̄
= −0.05808 . . . ≈ 5.81%,

525−741.6̄
• 741.6̄
= 0.29181 . . . ≈ 29.18%,

133
 −105−(−98.3̄)
• −98.3̄
= 0.06779 . . . ≈ 6.78%.

We see, that the relative errors are getting larger if ∆Q is getting larger. Of course
having approximations instead of exact results is a disadvantage. But using marginal
costs allows us to e.g. plot the graph easily and the approximation is in most cases
good enough. In Figure 69 we can see the behaviour of TC. In the graph we imme-
diately see what happens if Q is increased by 1 for many Q.

Figure 69: Total costs and marginal costs

Remark. In the following example we use the demand function P (Q), so the price P
depends on the quantity Q. Sometimes this relation is reversed, so Q(P ) is used. For
example if P (Q) = 6 − 0.5Q we can rewrite this by using Q(P ) = 12 − 2P . ■

Example 69. Given the demand function P (Q) = 6 − 0.5Q.

a) Compute a function which describes the total revenue.

b) Find the value of MR for Q0 = 4 and explain the result.

c) Find an approximation of Q0 such that an increase of Q0 by ∆Q = 1 leads to an

increase of TR by ∆TR = 4.

Solution:

134
a) TR(Q) = P (Q) · Q = 6Q − 0.5Q2

b) MR(Q) = TR′ (Q) = 6 − Q and therefore MR(4) = 2. ∆TR ≈ MR(4) · ∆Q = 2 · ∆Q.

Explanation: Starting at a value of Q0 = 4: if Q0 is increased by ∆Q = 1, TR
increases by 2. If Q0 is increasing by ∆Q = 2, TR increases by 2 · 2 = 4. And so on.

c) We know that ∆TR ≈ MR(Q0 ) · ∆Q, ∆TR = 4, ∆Q = 1 and MR(Q0 ) = 6 − Q0 , so

we get 4 ≈ (6 − Q0 ) · 1. Therefore we conclude that Q0 ≈ 2 satisfies the mentioned
claim. This is not 100% exact, as the following calculation shows: TR(3) − TR(2) =
13.5 − 10 = 3.5 ̸= 4.

The interested reader can solve this example exactly by using the equation

4 = TR(Q0 + 1) − TR(Q0 ) = . . . = 5.5 − Q0 .

Hence the exact solution would be Q0 = 1.5. We can finally check the result:

TR(2.5) − TR(1.5) = 11.875 − 7.875 = 4.

In Section 4.5 a similar approach leads to the concept of elasticity. There we will ask
the question:

Let’s assume that x settled at a fixed value x0 with corresponding function

value f (x0 ). What relative increase/decrease of f can be expected if x0
increases by 1%?

As usual we conclude the section by further exercises:

Exercise 96. Given the profit function

π(Q) = −100 000 + 1 100Q − Q2 .

First compute the change in π exactly and second approximate the computation if Q0 is
increasing from 200 to 201 and from 200 to 250.

Solution: ∆π = 699, ∆π ≈ 700, ∆π = 32 500 and ∆π ≈ 35 000 ■

Exercise 97. Given the profit function

π(Q) = −100 000 + 1 100Q − Q2 .

For which Q0 the following is true:

135
a) An increase of Q0 by 20 approximately leads to a decrease of π by 6 000?

b) A decrease of Q0 by 20 approximately leads to an increase of π by 8 000?

Solution: Q ≈ 700, Q ≈ 750 ■

4.3 Economic Applications of Maxima and Minima

Let’s start with two examples to get a first idea about optimisation - the topic of this
section.

Example 70. The demand function of a product is given by Q(P ) = 100 − P where Q
is the demanded quantity and P the price.

a) What is the meaningful domain of definition?

b) Which price maximises the total revenue TR?

c) Find P such that TR is increasing.

d) Find P such that TR is decreasing.

Solution: We will solve this example by using the graph. Later we will learn how to solve
this by computation. We know that the total revenue is the product of the quantity and
the price, therefore there holds

TR(P ) = Q(P ) · P = (100 − P ) · P = 100P − P 2 .

In Figure 70 we can see the graph and we immediately conclude:

a) The domain of definition is given by the interval [0, 100]. For P not in this domain
the corresponding TR would be negative.

b) For P = 50 we get the maximal total revenue TR(50) = 2 500.

c) For P ∈ [0, 50) the total revenue is increasing.

d) For P ∈ (50, 100] the total revenue is decreasing.

136
 Figure 70: Total revenue

Example 71. Let’s assume that the total cost function TC is given by

TC(Q) = 40 000 + 500Q − 3Q2 + 0.01Q3 ,

see Figure 71. Find the quantity Q such that the average costs AC are minimal.

Solution: The average costs are given by

TC 40 000 + 500Q − 3Q2 + 0.01Q3

AC(Q) = =
Q Q
and plotted in Figure 72. Again we will solve this example by using the graph. We can

Figure 71: Total costs Figure 72: Average costs

see that for Q = 200 we get the desired minimum of the average costs.

Till now we have spoken of maxima and minima. From now on we are a little bit more
precise and for a function f (x) we define the following:

137
a) A point (x0 , f (x0 )) is called local/relative maximum if and only if for all x “around
x0 ” the corresponding function value f (x) is less or equal than f (x0 ).16

b) A point (x0 , f (x0 )) is called global/absolute maximum if and only if for all x in
the whole domain of definition the corresponding function value f (x) is less or equal
than f (x0 ).

c) A point (x0 , f (x0 )) is called local/relative minimum if and only if for all x “around
x0 ” the corresponding function value f (x) is greater or equal than f (x0 ).17

d) A point (x0 , f (x0 )) is called global/absolute minimum if and only if for all x in
the whole domain of definition the corresponding function value f (x) is greater or
equal than f (x0 ).

Example 72. Find the highest mountain (global maximum) of

a) Austria,

b) the whole world.

Solution: In Lower Austria the highest mountain is the Schneeberg (2.076 meters). This
is a local maximum because around the Schneeberg there is no higher mountain. But
the Schneeberg is no global maximum in Austria, because the Großglockner (3 798 me-
ters) is higher. In Austria there is no higher mountain than the Großglockner, so the
Großglockner is a local but also a global maximum.

But if the domain of interest is the whole world, the Großglockner is not a global maxi-
mum because there are higher mountains in Europe: The highest mountain is the Elbrus
(5 642 meters) or the Mont Blanc (4 810 meters), depending on the border definition.
Both mountains are local maxima. But the Mount Everest (8 848 meters) is the highest
mountain of the world. So the Mount Everest is not only a local maximum but also a
global maximum.

In Figure 73 we can see some pictures of the mountains18 .

Example 73. Let’s analyse the graph of Figure 74 to find local and global maxima and
local and global minima.
16
“around” is not very precise. In calculus we normally use the more precise definition: there exists a
δ > 0 such that f (x) ≤ f (x0 ) in the interval (x0 − δ, x0 + δ).
17
The comment concerning local maxima can be applied here too. Instead of ≤ we use ≥ of course.
18
Pictures were taken from www.wikipedia.com.

138
Figure 73: Top from left to right: Schneeberg, Großglockner, Mont Blanc
Bottom from left to right: Elbrus and Mount Everest

Solution: Points P2 , P4 and P6 are local maxima, but only point P6 is a global maximum.
Points P1 , P3 and P5 are local minima, but only point P3 is a global minimum. ■

Figure 74: Local/global maxima and local/global minima

Having Example 73 in mind, we are now looking for methods to get the local maxi-
mum/minimum by calculation. If we know all local maxima/minima, it is easy to get
the global maximum/minimum. The sparkling idea is to use the tangent. So we draw
the tangents in the local maxima and local minima of Figure 74, which leads to Figure
75. We observe that all tangents of the local maxima/minima - except the ones on the
boundary of the domain - have slope zero. This is somehow clear because the tangent g
of a function f in (x0 , y0 ) approximates the function for x close to x0 . So if the tangent

139
 Figure 75: Tangents

has a slope greater than zero, the function is increasing in x0 and in a small area around
x0 , see the tangents of P1 and P6 in Figure 75. If the tangent has a slope less than zero,
the function is decreasing in x0 and in a small area around x0 .
Maybe the function left to P1 would be smaller than in P1 but this area is not anymore
part of the domain of definition. Maybe the function right to P6 would be greater than
in P6 but this area is again not part of the domain of definition. So the two points on
the boundary of the domain of definition are local maxima/minima.

Summing up: If for a given function f the point (x0 , y0 ) is a local maximum/minimum
then one of the two statements must be true:

a) The tangent of f in (x0 , y0 ) has slope zero. This is the case if

f ′ (x0 ) = 0. (42)

b) (x0 , y0 ) is located on the boundary of the domain of definition.

Our strategy therefore is to find all points (x0 , y0 ) such that f ′ (x0 ) = 0 (first derivative
test) or (x0 , y0 ) is located on the boundary of the domain of definition.

Not all candidates (x0 , y0 ) with f ′ (x0 ) = 0 have to be local maxima/minima, see Figure
76. In this situation there is a horizontal tangent, but the corresponding point is not a
local maximum neither a local minimum. Such points are called saddle points.

140
 Figure 76: Saddle point

So finally we have to find out if a candidate with f ′ (x0 ) = 0 is a local maximum/minimum

or a saddle point. The following second derivative test helps to figure this out:

a) If f ′ (x0 ) = 0 and f ′′ (x0 ) < 0 the point (x0 , f (x0 )) is a local maximum.

b) If f ′ (x0 ) = 0 and f ′′ (x0 ) > 0 the point (x0 , f (x0 )) is a local minimum.

c) If f ′ (x0 ) = 0 and f ′′ (x0 ) = 0 and f ′′′ (x0 ) ̸= 0 the point (x0 , f (x0 )) is a saddle point.

d) If none of the first three cases comes up, we would have to consider higher derivatives.

Remark. In most cases we can argue by using the graph or by thinking to find out, if
the point is a local maximum/minimum or saddle point. ■

Before we apply this theory to some applications we summarise: To find the local and
global maxima/minima of a function f we consider:

a) the points (x0 , f (x0 )) on the boundary of the domain of interest. E.g. if the domain
of definition is the interval [0, 20] we have to check (0, f (0)) and (20, f (20)). If the
domain of definition is the interval [0, ∞) we have to check (0, f (0)) and “(∞, f (∞)”
and if the domain of definition is R we have to check “(−∞, f (−∞)” and “(∞, f (∞)”.

b) the points (x0 , f (x0 )) in the domain of definition such that f ′ (x0 ) = 0. To find out if
(x0 , f (x0 )) is a local maximum/minimum or saddle point we use the second derivative
test.

c) To find out the global maximum/minimum we simply take the candidates of a) and

141
 b). The one with the highest function value is the global maximum. The one with
the lowest function value is the global minimum.

Let’s consider the following computation-expensive example before we look at some

applications.

Example 74. Find all relative minima/maxima and saddle points of

x2 · (x − 1)3 .

a) What is the global minimum/maximum, if the domain of interest is R?

b) What is the global minimum/maximum, if the domain of interest is the interval

[0, 1.5]?

The first derivative can be computed using the product rule:

f ′ (x) = 2x · (x − 1)3 + x2 · 3 · (x − 1)2

= x · (x − 1)2 · [2 · (x − 1) + 3x]
= x · (x − 1)2 · (5x − 2)

Now please do not expand this, because we want to find the roots, i.e. we want to find
x such that f ′ (x) = 0. But a product is zero if and only if one of the factors is zero.
Therefore we get: x1 = 0, x2 = 1 and x3 = 52 = 0.4. Next we can compute the second
and the third derivative. To do that it is easier to expand the first derivative, so we have

f ′ (x) = (5x2 − 2x) · (x2 − 2x + 1) = −2x + 9x2 − 12x3 + 5x4 ,

f ′′ (x) = −2 + 18x − 36x2 + 20x3 ,
f ′′′ (x) = 18 − 72x + 60x2 .

We see that for x1 = 0 we get f ′′ (0) = −2 < 0, therefore (0, 0) is a local maximum.
For x2 = 1 we get f ′′ (1) = 0 and f ′′′ (1) = 6 ̸= 0, therefore (1, 0) is a saddle point.
For x3 = 0.4 we get f ′′ (0.4) = 0.72 > 0, therefore (0.2, −0.03456) is a local minimum.

a) We know that for f (x) = x2 · (x − 1)3 the function value tends to +∞, if x gets
bigger and bigger and −∞, if x gets smaller and smaller.Therefore the global mini-
mum/maximum are at x = −∞ and x = +∞.

b) The function values on the boundaries are f (0) = 0 and f (1.5) = 0.28125. If we
consider all candidates, we get the global minimum at (0.4, −0.03456) and the global
maximum at (1.5, 0.28125).

In Figure 77 we can verify our calculation. ■

142
 Figure 77: Optimisation problem

Example 70 and Example 71 are solved again, but now we use our theoretical tools:

Example 70 (Again). Let the demand function of a product is given by Q(P ) = 100−P
where Q is the demanded quantity and P the price.

a) What is the meaningful domain of definition?

b) Which price maximises the total revenue TR?

c) Find P such that TR is increasing.

d) Find P such that TR is decreasing.

Solution:

a) Of course P ≥ 0 has to be satisfied. As Q(P ) = 100 − P the function is only

meaningful for P ≤ 100. Therefore the domain of definition is given by [0, 100].

b) Let’s consider TR and its derivatives:

TR(P ) = 100P − P 2 ,
TR′ (P ) = 100 − 2P,
TR′′ (P ) = −2.

Boundary: TR(0) = 0 and TR(100) = 0.

First derivative test: 0 = TR′ (P ) = 100 − 2P and therefore P = 50.

143
 Corresponding second derivative test: TR′′ (50) = −2 < 0.
Therefore (50, 2 500) is a local maximum, but also a global one, because the other two
candidates have a smaller function value.

c) TR is increasing for x if TR′ (x) > 0:

TR′ (P ) > 0,
100 − 2P > 0,
100 > 2P,
50 > P.
TR is increasing in the interval [0, 50).

d) TR is decreasing for x if TR′ (x) < 0:

TR′ (P ) < 0,
100 − 2P < 0,
100 < 2P,
50 < P.
TR is decreasing in the interval (50, 100].

Example 71 (Again). Let’s assume that the total cost function TC is given by
TC(Q) = 40 000 + 500Q − 3Q2 + 0.01Q3 ,
see Figure 71. Find the quantity Q such that the average costs AC are minimal.

Solution: Let’s consider AC and its derivatives:

40 000
AC(Q) = + 500 − 3Q + 0.01Q2
Q
−40 000
AC′ (Q) = − 3 + 0.02Q
Q2
80 000Q 80 000
AC′′ (Q) = + 0.02 = + 0.02.
Q4 Q3
In general Q ≥ 0 has to be satisfied, therefore the domain of interest is the interval
[0, ∞).

Boundary: AC(0) = ∞
−40 000−3Q2 +0.02Q3
First derivative test19 : AC′ (Q) = Q2
, so we have to solve −40 000 − 3Q2 +
19
A fraction is zero if and only if the numerator is zero.

144
0.02Q3 = 0. As this is hard to solve we can use Microsoft Excel’s solver: Q = 200
Second derivative test: AC ′′ (200) = 0.03 > 0
Therefore (200, 500) is a local minimum
If Q tends to ∞ we see that the average costs tend to ∞ too.
Therefore (200, 500) is a global minimum.

Interpretation:

a) If Q = 200 (optimal case) and P = 500 the total costs are covered by the total revenue:
TC(200) = 100 000 and TR(200) = 500 · 200 = 100 000.

b) If P < 500 the total costs are not covered. Not even in the optimal case Q = 200.

c) If P = 500 but Q ̸= 200 we are not in the optimal case and therefore the total costs
are not covered!

We have seen: Asking a price lower than the minimum AC is in general not recommended
because not even in the optimal case all costs are covered.

However, sometimes lower prices are asked: Let’s assume that the total costs are covered
by already existing orders. We were asked to perform an additional price campaign. But
at least we want to cover the variable costs. Hence, we now want to know the minimal
price, such that the variable costs are covered, at least in the optimal case.

Example 75. As in Example 71 we assume that the total cost function is given by

TC(Q) = 40 000 + 500Q − 3Q2 + 0.01Q3 .

Find the quantity Q such that the average variable costs are minimal.

Solution: The variable costs, the average variable costs and its derivatives are given by

VC(Q) = 500Q − 3Q2 + 0.01Q3 ,

VC 500Q − 3Q2 + 0.01Q3
AVC(Q) = = = 500 − 3Q + 0.01Q2 ,
Q Q
AVC′ (Q) = −3 + 0.02Q,
AVC′′ (Q) = 0.02.

The domain of interest is given by the interval [0, ∞)

Boundary: AVC(0) = 500
First derivative test: 0 = AVC′ (Q) = −3 + 0.02Q and therefore Q = 150.

145
Corresponding second derivative test: AVC′′ (150) = 0.02 > 0.
Therefore (150, 275) is a local minimum.
If Q tends to ∞ we see that the average costs tend to ∞ too.
Therefore (150, 275) is a global minimum.

Interpretation:

a) If Q = 150 (optimal case) and P = 275 the variable costs are covered by the total
revenue: VC(150) = 41 250 and TR(150) = 275 · 150 = 41 250.

b) The fixed costs are not covered!

c) If P < 275 the variable costs are not covered. Not even in the optimal case Q = 150.

d) If P = 275 but Q ̸= 150 we are not in the optimal case and therefore not even the
variable costs are covered!

In the next example, a company’s profit will be optimised.

Example 76. We consider a hotel’s demand and total cost function:

80 P
Q(P ) =− ,
3 3
TC(Q) = 120 + 8Q.

a) What is the domain of interest for P and Q?

b) Find the function for the total revenue TR(Q) and the profit π(Q).

c) Find the break-even points (Q such that π(Q) = 0).

d) Find Q such that the profit is maximal.

Solution:

a) It is clear that Q ≥ 0 and P ≥ 0. Looking at Q(P ) we can see that

 80P > 80 lead to
negative quantities. Hence we get P ∈ [0, 80]. This leads to Q ∈ 0, 3 = [0, 26.6̄].

b) We know TR(Q) = Q · P (Q) but instead of P (Q) we only know Q(P ). Therefore we
first reshape the demand function: P (Q) = 80 − 3Q. So we get the total revenue
TR(Q) = (80 − 3Q) · Q = 80Q − 3Q2 .

146
 Next we define the profit function:

π(Q) = TR(Q) − TC(Q) = 80Q − 3Q2 − (120 + 8Q)

= −120 + 72Q − 3Q2 .

c) We solve 0 = π(Q) = −120 + 72Q − 3Q2 , therefore we get

p
−72 ± 722 − 4 · (−3) · (−120) −72 ± 61.188 . . .
Q1,2 = =
2 · (−3) 2
and hence the break-even points are Q1 = 1.801 . . . and Q2 = 22.198 . . .

d) Boundaries: π(0) = −120 and π(26.6̄) = −333.3̄

The first two derivatives of π are given by

π ′ (Q) = 72 − 6Q,
π ′′ (Q) = −6.

First derivative test: 0 = π ′ (Q) = 72 − 6Q, therefore Q = 12.

Second derivative test: π ′′ (12) = −6 < 0.
Therefore (12, 312) is a local maximum, but also a global one, because the other can-
didates lead to less profit.

All information are plotted in Figure 78.

Figure 78: Maximise profit

147
Exercise 98. The demand function for a good is given as P (Q) = 50 − 2Q.

a) Write down the domain of definition.

b) Write down an expression for TR (total revenue).

c) Calculate the output at which TR is a global maximum.

d) Sketch the TR function.

Solution: [0, 25], TR(Q) = 50Q − 2Q2 , TR′ (Q) = 50 − 4Q, TR′′ (Q) = −4, candidates
for maximum: (0, 0) (global min), (25, 0) (global min), (12.5, 312.5) (global max), graph
see Figure 79

Figure 79: Total revenue

Exercise 99. Let the total cost function be given by

1
TC(Q) = 100 + · Q3 .
100
Find Q such that the average costs are minimal. Compute the average costs for this
quantity.

Solution: Q = 17.099 . . . and AC(17.099 . . .) = 8.772 . . . ■

Exercise 100. Let the total cost function be given by

TC(Q) = 98 + 60Q − 12Q2 + Q3 .

148
Find Q such that the average variable costs are minimal. Compute the average variable
costs for this quantity.

Solution: Q = 6 and AVC(6) = 24 ■

Exercise 101. Let’s consider a constant demand function P (Q) = 121 (i.e. P does not
depend on Q) and the variable cost function
1
VC(Q) = 175Q − 15Q2 + Q3 .
2
The fixed costs are given by FC = 500.

a) What is the domain of interest?

b) Write down the equations for TC (total costs), TR (total revenue) and π (profit).

c) Compute the break-even points. You can use the solver of Microsoft Excel.

d) Find the output Q, at which profit is minimised and maximised.

e) Sketch TC, TR and π on one diagram. Include all candidates for maxima/minima,
the break-even points and the profit zone.

Solution: [0, ∞), TC(Q) = 500 + 175Q − 15Q2 + 21 Q3 , TR(Q) = 121Q, π(Q) = −500 −
54Q + 15Q2 − 12 Q3 , 10.417 . . ., 23.643 . . ., π ′ (Q) = −54 + 30Q − 23 Q2 , π ′′ (Q) = 30 −
3Q, candidates for maximum/minimum: (0, −500) (local max), (2, −552) (local min),
(18, 472) (global max), for Q → ∞ we get profit −∞. Graph see Figure 80. ■

4.4 Point of Inflection, Progressive and Degressive Behaviour

First we are interested in points on a graph of a function f where the so-called curva-
ture20 changes its sign. In these points the behaviour of the graph changes from the
brown to the orange behaviour of Figure 81 or vice versa from the orange to the brown
behaviour. These points are called points of inflection. In case of the brown function
the curvature is positive and the behaviour is called convex. In case of the orange
function the curvature is negative and the behaviour is called concave.
20
For the sake of completeness the formula for the curvature κ of a function f (x) is stated here:

f ′′ (x)
κ(x) = n o3/2
2
1 + [f ′ (x)]

149
 Figure 80: Maximise profit

Figure 81: Behaviour changes in points of inflection

150
To get a better idea we are looking at the following example:

Example 77. Let’s assume that we are sitting in a helicopter. On the ground a car
is driving on a road, see Figure 82. 5 positions are highlighted on the graph. In which

Figure 82: Car on the road and point of inflection

positions the steering wheel is turned to the right (concave behaviour) and in which to
the left (convex behaviour)? Is there a position, such that the steering wheel is in the
neutral position? This point is an inflection point.

Solution: In P1 and P2 the steering wheel is turned to the left (convex behaviour). In P3
the steering wheel is in the neutral position. In P4 and P5 the steering wheel is turned
to the right (concave behaviour). So P3 is a point of inflection. See Figure 83 to analyse
the behaviour of the curvature. ■

To be able to compute points of inflection we use the following criteria:

If a function f is given and there holds f ′′ (x0 ) = 0 and f ′′′ (x0 ) ̸= 0, the point
(x0 , f (x0 )) is a point of inflection.21

If f ′′ (x0 ) = 0 and f ′′′ (x0 ) = 0 we would have to consider higher derivatives. But we
don’t care about these cases.

21
We have defined the curvature κ in the above footnote. We can see immediately, that the denominator
is always positive, so the sign of κ depends on f ′′ only. And therefore a necessary condition for a
point of inflection is f ′′ (x0 ) = 0. The condition f ′′′ (x0 ) ̸= 0 guarantees that the sign really changes.

151
 Figure 83: Car on the road and point of inflection

Example 78. Let’s consider the function

f (x) = −0.27 + 0.47x − 0.09x2 + 0.01x3 .

Find inflection points if there are any.

Solution: Consider the derivatives:

f ′ (x) = 0.47 − 0.18x + 0.03x2 ,

f ′′ (x) = −0.18 + 0.06x,
f ′′′ (x) = −0.06.

The root of f ′′ is given by x = 3. Due to the fact that f ′′′ (x) = −0.06 ̸= 0, the point
(3, 0.6) is a point of inflection. In Figure 84 our calculation is verified. ■

Example 79. Let’s consider the functions

f (x) = x2 and g(x) = x4 .

Find inflection points if there are any.

Solution: Consider the derivatives:

f ′ (x) = 2x, g ′ (x) = 4x3 ,

f ′′ (x) = 2, g ′′ (x) = 12x2 ,
f ′′′ (x) = 0, g ′′′ (x) = 24x.

152
 Figure 84: Point of inflection

Figure 85: No point of inflection Figure 86: No point of inflection

153
There are no roots of f ′′ , therefore there are no points of inflection for f , see Figure 85.
The only root of g ′′ is for x = 0. But g ′′′ (0) = 0, therefore we do not know, if (0, 0) is a
point of inflection. Looking at the graph in Figure 86, we see, that (0, 0) is no point of
inflection but a minimum. ■

Next we want to analyse increasing functions in more details. So let us assume that an
increasing function f is given, i.e. f (x1 ) ≤ f (x2 ) if x1 ≤ x2 . We already know that if
f ′ (x) ≥ 0 the function is increasing, because all tangents have a positive slope.

Of course one could be interested in decreasing functions as well, i.e functions such that
f (x1 ) ≥ f (x2 ) if x1 ≤ x2 . This would be the case, e.g. if f ′ (x) ≤ 0.

Example 80. Let’s consider the function

f (x) = −960 + 368x − 36x2 + x3 .

Find all x such that f (x) is increasing. Finally find all x such that f (x) is decreasing.

Solution: The first derivative is given by

f ′ (x) = 368 − 72x + 3x2 .

Next we compute the roots:

√
72 ± 722 − 4 · 3 · 368 72 ± 27.712 . . .
x1,2 = = ,
2·3 6
x1 = 7.281 . . . and x2 = 16.618 . . . As the first derivative is only zero in this two points,
we know, that f ′ does not change sign in the intervals (−∞, x1 ), (x1 , x2 ) and (x2 , ∞).
To find out the sign of f ′ in the three intervals we only have to consider one arbitrary x
in the corresponding interval:

(−∞, x1 ) : choose e.g. x = 0 : f ′ (0) = 368 > 0,

(x1 , x2 ) : choose e.g. x = 10 : f ′ (10) = −52 < 0,
(x2 , ∞) : choose e.g. x = 20 : f ′ (20) = 128 > 0.

Therefore we know that f ′ is positive in (−∞, x1 ), negative in (x1 , x2 ) and again positive
in (x2 , ∞). Therefore f is increasing in (−∞, x1 ), decreasing in (x1 , x2 ) and again
increasing in (x2 , ∞). In Figure 87 we can verify the calculation. ■

For an increasing function f we now we want to distinguish between:

154
Figure 87: Intervals for increasing and decreasing behaviour

155
a) The function f and the value of the growth is increasing in an interval (a, b) (pro-
gressive behaviour). f is increasing in (a, b) if f ′ is positive in (a, b). The value of
the growth (f ′ ) is increasing in (a, b), if its derivative f ′′ is positive in (a, b). Summing
up: if f ′ and f ′′ are positive in (a, b) the behaviour is progressive. See Figure 88 for
an example.

b) The function f is increasing but the value of the growth is decreasing in an interval
(a, b) (degressive behaviour). f is increasing in (a, b) if f ′ is positive in (a, b). The
value of the growth (f ′ ) is decreasing in (a, b) if its derivative f ′′ is negative in (a, b).
Summing up: if f ′ is positive in (a, b) and f ′′ is negative in (a, b) the behaviour is
degressive. See Figure 89 for an example.

Figure 88: Progressive behaviour Figure 89: Degressive behaviour

Example 81. Let’s consider the total costs: TC(Q) = 100 + 3Q − 0.3Q2 + 0.01Q3 . Find
the intervals such that TC is progressive or degressive.

Solution: First we mention that TC is only defined in the interval [0, ∞) (negative Q do
not make sense). The derivatives of TC are given by:
TC′ (Q) = 0.03Q2 − 0.6Q + 3,
TC′′ (Q) = 0.06Q − 0.6,
TC′′′ (Q) = 0.06.
To find out if TC is increasing we have a look at the roots of TC′ :
√
0.6 ± 0.62 − 4 · 3 · 0.03 0.6
Q1,2 = = = 10.
2 · 0.03 0.06
Therefore TC does not change sign in the intervals (0, 10) and (10, ∞). To find out the
sign, we only have to evaluate for one Q in both intervals:
(0, 10) : choose e.g. Q = 1 : TC′ (1) = 2.43 > 0,
(10, ∞) : choose e.g. Q = 100 : TC′ (100) = 243 > 0.

156
Due to the positive sign in both intervals, TC′ is positive in both intervals and therefore
TC is increasing in both intervals.

To find out the intervals for progressive and degressive behaviour we have a look at the
roots of TC′′ :

Q = 10.

So the sign of TC′′ does not change in the interval (0, 10) and (10, ∞). To find out the
signs, we only have to evaluate for one Q in both intervals:

(0, 10) : choose e.g. Q = 1 : TC′′ (1) = −0.54 < 0,

(10, ∞) : choose e.g. Q = 100 : TC′′ (100) = 5.4 > 0.

Therefore TC is degressive in (0, 10) and progressive in (10, ∞). Finally we check that
the point (10, TC(10)) = (10, 110) is a point of inflection: As TC′′′ (10) = 0.06 ̸= 0, this
is a point of inflection. See Figure 90 to verify the computational results. ■

Figure 90: Progressive and degressive behaviour

Remark. In the previously example and especially in Figure 90 we can see the difference
between progressive and degressive behaviour.

A progressive behaviour often leads to difficulties in real life. Think about the rapid
spread of viruses during the Corona time. In Section 1 we have analyzed another very
famous example for progressive behaviour: compound interest. In case of high debts this
can be existence-threatening too.

157
Examples for degressive behaviours are given by e.g. learning curves or biological growth:
When learning a new skill or subject, you often learn a lot in the beginning (fast increase).
However, over time, the rate at which you learn new things tends to decrease (degressive
increase). Many biological processes, such as the growth of plants or animals, display
degressive growth. At first, the organism grows rapidly. But as it gets larger, the rate of
growth slows down, until it eventually stops growing altogether. ■

Exercise 102. Find x such that the function

f (x) = 10 + 10x2 − 10x3 + 2x4
is increasing/decreasing.

Solution: f ′ (x) = 20x − 30x2 + 8x3 = (20 − 30x + 8x2 ) · x, roots: x1 = 0, x2 =

0.867 . . ., x3 = 2.882 . . ., f ′ (−1) = −58 < 0, therefore decreasing in (−∞, 0), f (0.5) =
3.5 > 0, therefore increasing in (0, 0.867 . . .), f ′ (1) = −1 <=, therefore decreasing in
(0.867 . . . , 2.882 . . .) and f ′ (4) = 112 > 0, therefore increasing in (2.882 . . . , ∞).

See Figure 91 to verify the solution.

Figure 91: Intervals for increasing and decreasing behaviour

Exercise 103. Let’s consider the total costs: TC(Q) = 1 000 + 20Q − 0.3Q2 + 0.004Q3 .
Find the intervals such that TC is progressive or degressive.

Solution: The domain of definition is (0, ∞), TC′ (Q) = 20 − 0.6Q + 0.012Q2 , TC′′ (Q) =
−0.6 + 0.024Q, TC′′ (Q) = 0.024.

158
There are no roots of TC′ , TC′ (10) = 25.2 > 0, therefore TC′ is positive in (0, ∞),
therefore TC is increasing in (0, ∞).
The only root of TC′′ is for Q = 25, TC′′ (10) = −0.36 < 0, therefore TC′′ is negative
in (0, 25), therefore TC′ is decreasing and TC is increasing in (0, 25), therefore TC
is degressive in (0, 25), TC′′ (1 000) = 23.4 > 0, therefore TC′′ is positive in (25, ∞),
therefore TC′ and TC are increasing in (25, ∞), therefore TC is progressive in (25, ∞),
(25, 1 375) is a point of inflection due to TC′′′ (25) = 0.024 ̸= 0.

See Figure 92 to verify the solution. ■

Figure 92: Progressive and degressive behaviour

4.5 Elasticity

Let’s assume that a function f (x) be given. For example f describes the demand de-
pending on the price. In this case we would prefer the notation Q(P ) instead of f (x),
but first we will stay with the notation f (x). In Section 4.2 we considered and analysed
the question:

Let’s assume that x settled at a fixed value x0 with corresponding function

value f (x0 ). What increase/decrease of f can be expected if x0 increases by
1?

We were able to solve this question exactly but it turned out that the solution can be
approximated quite good by using the concept of tangents. Finally we came up with

159
the formula:
∆f = f (x0 + ∆x) − f (x0 ) ≈ f ′ (x0 ) · ∆x.
Set ∆x = 1 to answer the above question. We are now focusing on the following
question:

Let’s assume that x settled at a fixed value x0 with corresponding function

value f (x0 ). What relative increase/decrease can be expected if x0
increases by 1%?

We use the notation ∆x to describe the absolute change for x and ∆f to describe
the corresponding absolute change in the function value. The corresponding relative
changes are therefore given by ∆x
x0
and f∆f
(x0 )
. Before we analyse the mentioned question,
we discuss one easy example first to get used to the notation:

Example 82. Let’s assume that a function f is defined by

f (x) = 100 + x2 .
Starting point is x0 = 4. We further consider ∆x = 2, so we are interested in x0 + ∆x =
4 + 2 = 6. What is the corresponding relative increase? Compute the absolute and
relative increase of the function value.

Solution: The relative increase is given by ∆x

x0
= 24 = 50%. Next we compute ∆f = f (6)−
f (4) = 136 − 116 = 20 and the corresponding relative increase f∆f (x0 )
20
= 116 ≈ 17.24%.
Therefore a relative increase of x0 by 50% corresponds with a relative increase of the
function value by only 17.24%. ■

Lets consider this topic in more detail. We start with the well known formula:
∆f = f (x + ∆x) − f (x0 ) ≈ f ′ (x0 ) · ∆x,
i.e.
∆f ≈ f ′ (x0 ) · ∆x.
As we are interested in the relative change we divide by f (x0 ) and get
∆f f ′ (x0 )
≈ · ∆x.
f (x0 ) f (x0 )
One last easy trick yields
∆f f ′ (x0 ) x0 f ′ (x0 ) · x0 ∆x
≈ · ·∆x = · .
f (x0 ) f (x0 ) x0 f (x0 ) x0
|{z}
=1

160
 ∆x
So we have answered the mentioned question: If x0 increases relatively by x0
then the
function value increase relatively by
∆f f ′ (x0 ) · x0 ∆x
≈ · .
f (x0 ) f (x0 ) x0
We use the notation
f ′ (x0 ) · x0
ε(x0 ) =
f (x0 )
and conclude that ε(x0 ) is the proportionality constant which is called (point) elastic-
ity:
∆f ∆x
≈ ε(x0 ) · .
f (x0 ) x0

∆x ∆f
Let’s have a look at the special case x0
= 1%, then we see that the relative change f (x0 )
is given by ε(x0 ).

We now want to have a look at special elastic behaviours:

a) If ε(x0 ) is positive, then a positive relative change of x0 causes a positive relative

change of the function value. Vice versa: If ε(x0 ) is positive, then a negative relative
change of x0 causes a negative relative change of the function value.

b) If ε(x0 ) is negative, then a positive relative change of x0 causes a negative relative

change of the function value. Vice versa: If ε(x0 ) is negative, then a negative relative
change of x0 causes a positive relative change of the function value.

Remark. The relation between demand and price of the same good leads to negative
point elasticity in most cases: A relative increase of price for coffee in general leads to
a relative decrease of demand for coffee. And of course the other way round: a relative
decrease of price for coffee lead to a relative increase of demand for coffee. Sometimes this
is not the case, e.g. if I want to show my guests that I have lots of money. Let’s assume
that the price for coffee has increased dramatically. To demonstrate my wealthiness I
especially buy and offer large amounts of coffee.

But we can also think about the demand of a good, e.g. bicycles and the price of another
good, e.g. fuel. In this case a relative increase of price for fuel can lead to a relative
increase of demand for bikes and the other way round because cars are substituted by
bicycles. ■

Another classification is the following:

161
a) If |ε(x0 )| > 1 which means ε(x0 ) > 1 or ε(x0 ) < −1 then a relative change of x0 will
be amplified. This situation is called elastic behaviour.

b) If |ε(x0 )| < 1 which means −1 < ε(x0 ) < 1 then a relative change of x0 will be
damped. This situation is called inelastic behaviour.

c) Finally, if |ε(x0 )| = 1 which means ε(x0 ) = 1 or ε(x0 ) = −1 then a relative change

of x0 will not be amplified and not damped. This situation is called unit elastic
behaviour.

Remark. Let’s assume a person is addicted to heroin and that the price for heroin is
increasing by 10%. Due to the addiction the consumed amount of heroin stays almost
the same. The behaviour is inelastic.

But if we consider the demand for loans depending on the interest rate and assuming that
there are lots of banks with comparable conditions: If one bank increases their interest
rate by a small value, many people will switch to a competitor. The behaviour is elastic.
■

Example 83. Let’s assume that f (x) is given by f (x) = 10 − x + x2 . Compute the
elasticity for x0 = 10. Give an interpretation of this result.

Solution: f (10) = 100, the derivative of f is given by f ′ (x) = −1 + 2x and f ′ (10) = 19.
Therefore the elasticity is given by
19 · 10
ε(10) = = 1.9.
100
We see that ε(10) = 1.9 > 1 and therefore the situation is elastic. A 1% increase of x0
corresponds approximately with an increase of f by 1.9%. Vice versa, a 1% decrease of
x0 corresponds approximately with an decrease of f by 1.9%. ■

If we consider the demand function depending on the price we use the notation Q(P )
and ε(P ) instead of f (x) and ε(x) and call ε(P ) the price elasticity of demand.
Considering the price elasticity of demand we get an interesting view, as the following
example shows:

Example 84. Let’s assume that the demand of cigarettes is given by

500
Q(P ) = + 1 000,
P +2
see Figure 93. Further we assume that today’s price is 4.00 € Euro. Due to taxes the

162
 Figure 93: Demand of cigarettes depending on price

price will increase tomorrow by 15%. Compute the relative change of the demand by
using elasticity. Do we have an elastic, inelastic or unit elastic behaviour? Before doing
the computation guess if the relative change of the demand is

• larger than 15%,

• equal to 15%,

• less than −15%,

• equal to −15%,

• between 0% and 15%,

• between −15% and 0%.

Solution: My personal guess is that the demand function will decrease. Further I guess
that many people are addicted to cigarettes, therefore they won’t be able to stop smoking
although they do not like the higher price. Therefore my guess is that the demand will
change relatively between −15% and 0%.

Next, let’s compute the derivative of Q(P )

−500
Q′ (P ) =
(P + 2)2
and compute Q(4) = 1 083.3̄, Q′ (4) = −13.8̄ and so we get
Q′ (4) · 4 −13.8̄ · 4
ε(4) = = = −0.05128 . . .
Q(4) 1 083.3̄

163
We see that ε(4) is negative and −1 < ε(4) < 1, therefore inelastic. Finally we approxi-
mate the relative change of demand:
∆Q ∆P
≈ ε(P0 ) · = −0.05128 . . . · 15% = −0.769 . . . % ≈ −0.77%,
Q0 P0
where P0 = 4 and Q0 = 1 083.3̄. So if the price increases by 15% the demand only
decreases by ≈ 0.77%, this is really inelastic. ■

Of course we are not always interested in only one special x0 or P0 . So it would be a

good idea to compute the elasticity generally, i.e. ε(x) or ε(P ) instead of ε(x0 ) or ε(P0 ).
Of course computation can then be more exhausting but we are then able to compute
elasticity for arbitrary x or P and additionally we can plot the elasticity function. So
let’s have a look at Example 83 and Example 84 again.

Example 85. Use the instructions of Example 83 and compute the elasticity function
for all x.

Solution:

f (x) = 10 − x + x2 ,
f ′ (x) = −1 + 2x.

So elasticity is given by
f ′ (x) · x (−1 + 2x) · x
ε(x) = = .
f (x) 10 − x + x2
If we insert x = 10 we again get
190
ε(10) = = 1.9.
100
In Figure 94 we can see the graph of elasticity. We immediately see the elastic and
inelastic area ■

Figure 94: elasticity Figure 95: elasticity (cigarettes)

164
Example 86. Use the instructions of Example 84 and compute the elasticity function
for all P .

Solution:
500
Q(P ) = + 1 000,
P +2
−500
Q′ (P ) = .
(P + 2)2
Doing some simplifications, elasticity is given by
−500 −1 −1
Q′ (P ) · P (P +2)2
·P 500 · (P +2)2
·P (P +2)2
·P
ε(P ) = = 500 = 1

= 1
Q(P ) P +2
+ 1 000 500 · P +2
+2 P +2
+2
−1
(P +2)2
·P (P + 2)2 −P
= 1 · =
P +2
+2 (P + 2)2 (P + 2) + 2 · (P + 2)2
| {z }
=1
−P −P
= 2
= .
P + 2 + 2 · (P + 4P + 4) 10 + 9P + 2P 2
Therefore we get
−4
ε(4) = = −0.05128 . . . ,
10 + 9 · 4 + 2 · 42
which is the same result as in the original example of course. But now we are able
to compute the elasticity for arbitrary P and of course one can plot the graph too, see
Figure 95. ■

One big advantage of this approach is, that we can solve the following type of question:
Both relative changes are now given i.e. ∆xx0
and f∆f
(x0 )
are given. For which x0 does a
relative change of x0 by ∆x
x0
cause a relative change of f (x0 ) by f∆f
(x0 )
?

Example 87. Let’s think of a demand function

P
Q(P ) = 12 − .
2

a) Compute ε(P ), i.e. the price elasticity of demand.

b) Compute the price elasticity of demand for the price 5 €. Do we have an elastic,
inelastic or unit elastic behaviour?

c) Which relative change of demand can be expected, if the price increases from 5 Euro
by 20%?

165
d) For which price the following is true: Considering a 3% increase in price we can
expect a decrease in demand by 12%?

Solution:

a) Q′ (P ) = − 12 , therefore we get
− 12 · P − 12 · P 2 −P
ε(P ) = P
= P
· = .
12 − 2 12 − 2 |{z}2 24 − P
=1

b) ε(5) = −0.263 . . ., therefore inelastic.

c) We know that ∆Q
Q0
≈ ε(P0 ) · ∆P
P0
, so we get ∆Q
Q0
≈ (−0.263 . . .) · 20% = −5.263 . . . %

d) We do not know P0 , therefore we have to solve

−P0
−12% ≈ · 3% : 3%
24 − P0
−P0
−4 ≈ ·(24 − P0 )
24 − P0
−4 · (24 − P0 ) = −96 + 4P0 ≈ −P0 +P0 + 96
5P0 ≈ 96 :5
96
≈ P0 ≈ = 19.2.
5
For P0 ≈ 19.2 € a 3% increase in price causes a 12% decrease in demand.

Exercise 104. Let’s assume that demand depends on the price in the following way:
Q(P ) = 560 − 40P.
Compute the elasticity and evaluate the result for the prices P = 1.50 €, P = 7.00 € and
P = 12.00 € and decide if this is related to an elastic, inelastic or unit elastic behaviour.

Solution: ε(P ) = −14+P

P
, ε(1.50) = −0.12 (inelastic), ε(7.00) = −1 (unit elastic),
ε(12.00) = −6 (elastic) ■

Exercise 105. Let’s assume that the demand of a chest of avocados is given by
5 000
Q(P ) = − 50,
P +1
for P ∈ [0, 99).

166
a) Compute the elasticity function ε(P ).

b) Evaluate ε(P ) for P = 60 €, do we have an elastic, inelastic or unit elastic behaviour?

c) If the price decreases by 5%, which relative increase/decrease in demand can be ex-
pected?
−5 000
·P
Solution: Q (P ) =
′ −5 000
(P +1)2
, ε(P ) = (P +1)2
5 000
−50
= ... = −100P
99+98P −P 2
, ε(60) = −2.522 . . .
P +1
(elastic), 12.610 . . . % ■

Exercise 106. Let’s think of an demand function

Q(P ) = 1 000 − 2P − P 2 .

a) Compute ε(P ), i.e. the price elasticity of demand.

b) Compute the price elasticity of demand for the price 14 € Euro. Do we have an
elastic, inelastic or unit elastic behaviour?

c) Which relative change of demand can be expected, if the price increases from 14 € by
10%?

d) For which price the following is true: Considering a 2% increase in price we can
expect a decrease in demand by 3%?

−2P −2P 2
Solution: Q′ (P ) = −2 − 2P , ε(P ) = 1 000−2P −P 2
, ε(14) = −0.541 . . . (inelastic), ≈
−5.41%, ≈ 20.85 € ■

167
References
Bradley, Teresa (2013). Essential Mathematics for Economics and Business. New York:
Wiley.
Croft, Anthony and Robert Davison (2019). Mathematics for Engineers. Harlow, Eng-
land [etc.]: Pearson.
Die Fantastischen Vier (2007). Einfach sein. Track 2 on Fornika. Columbia Records.
Euler, Leonhard (1748). Introductio in analysin infinitorum. Vol. 1. Marcum-Michaelem
Bousquet & socios.
Handmann, Jakob Emanuel (1753). Bildnis des Leonhard Euler. Kunstmuseum Basel,
Collection Online. https://www.kunstmuseumbasel.ch/en/collection/online-
collection. Accessed: 2016-27-11.
Lapham, Lewis H. (1988). Money and Class in America. Weidenfeld & Nicolson.
Price, Richard (1771). “An Appeal to the Public, on the Subject of the National Debt”.
In: The McMaster Collection. McMaster University.
Reni, Guido (1640). Joseph and the Christ Child. https://en.wikipedia.org/wiki/
File:Guido_Reni_003.jpg. Oil on canvas. Accessed: 2023-07-10.
Swigert, Jack (1970). Apollo 13 mission. NASA Mission Transcripts, available at NASA
archives. Quote: "Houston, we’ve had a problem here".
Sydsæter, Knut et al. (2016). Essential mathematics for economic analysis. Harlow [etc.]:
Pearson Education.
Turner, Tina (1971). Proud Mary. Track 2 on Workin’ Together. Capitol Records.

168