APFM Study Guide

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 243

Asset pricing and

financial markets
R. Payne
FN2190
2018

Undergraduate study in
Economics, Management,
Finance and the Social Sciences

This subject guide is for a 200 course offered as part of the University of London
undergraduate study in Economics, Management, Finance and the Social
Sciences. This is equivalent to Level 5 within the Framework for Higher Education
Qualifications in England, Wales and Northern Ireland (FHEQ).
For more information about the University of London, see: london.ac.uk
This guide was prepared for the University of London by:
Richard Payne, Professor of Finance, Cass Business School; City, University of London.

This is one of a series of subject guides published by the University. We regret that due to
pressure of work the author is unable to enter into any correspondence relating to, or arising
from, the guide. If you have any comments on this subject guide, favourable or unfavourable,
please use the form at the back of this guide.

University of London
Publications Office
Stewart House
32 Russell Square
London WC1B 5DN
United Kingdom
london.ac.uk

Published by: University of London


© University of London 2018

The University of London asserts copyright over all material in this subject guide except where
otherwise indicated. All rights reserved. No part of this work may be reproduced in any form,
or by any means, without permission in writing from the publisher. We make every effort to
respect copyright. If you think we have inadvertently used your copyright material, please let
us know.
Contents

1 Introduction 7

1.1 Route map to the guide . . . . . . . . . . . . . . . . . . . . . . 7

1.2 Course coverage and syllabus . . . . . . . . . . . . . . . . . . 8

1.3 Aims of the course . . . . . . . . . . . . . . . . . . . . . . . . 10

1.4 Learning outcomes for the course . . . . . . . . . . . . . . . . 10

1.5 Essential and further reading . . . . . . . . . . . . . . . . . . 11

1.6 Online study resources . . . . . . . . . . . . . . . . . . . . . . 12

1.7 Exam structure and revision strategy . . . . . . . . . . . . . . 14

1.8 Mathematical pre-requisites . . . . . . . . . . . . . . . . . . . 15

1.9 Statistical pre-requisites . . . . . . . . . . . . . . . . . . . . . 17

1.10 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

1.11 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

2 Present value computations 24

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

2.2 Foundation: interest rates and discount rates . . . . . . . . 25

2.3 Future values of current cashflows . . . . . . . . . . . . . . 28

1
2.4 Present values . . . . . . . . . . . . . . . . . . . . . . . . . . 30

2.5 Comparing single cashflows to be received at different dates


. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

2.6 Net Present Value . . . . . . . . . . . . . . . . . . . . . . . . 34

2.7 The optimality of the NPV rule . . . . . . . . . . . . . . . . 36

2.8 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

3 Valuing long-lived cashflow streams 42

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

3.2 Perpetuities . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

3.3 Annuities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

3.4 Perpetuities with growth . . . . . . . . . . . . . . . . . . . . 48

3.5 Annuities with growth . . . . . . . . . . . . . . . . . . . . . 50

3.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

4 Interest rates: quoted and effective, real and nominal 54

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

4.2 Quoted rates and APRs . . . . . . . . . . . . . . . . . . . . . . 55

4.3 Effective interest rates . . . . . . . . . . . . . . . . . . . . . . 56

4.4 Common terminology . . . . . . . . . . . . . . . . . . . . . . 58

4.5 Examples: APRs and discounting . . . . . . . . . . . . . . . 58

4.6 Continuous compounding . . . . . . . . . . . . . . . . . . . 60

4.7 Real and nominal interest rates . . . . . . . . . . . . . . . . 61

4.8 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

2
5 Government bonds: features and valuation 66

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

5.2 Valuing bonds when interest rates are constant . . . . . . . . 72

5.3 Valuing government bonds and the term structure of inter-


est rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

5.4 Bonds: yield to maturity . . . . . . . . . . . . . . . . . . . . . 77

5.5 Duration and dependence of bond prices on interest rates . 81

5.6 Spot and forward interest rates . . . . . . . . . . . . . . . . . 86

5.7 The term structure of interest rates . . . . . . . . . . . . . . . 89

5.8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

5.9 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

6 Stock markets and equity valuation 94

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

6.2 Equity valuation and discounting dividends . . . . . . . . . 96

6.3 Stock prices, earnings and dividends . . . . . . . . . . . . . 102

6.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

6.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109

7 Risk, return and stock markets 110

7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

7.2 Stock and bond market returns: the data . . . . . . . . . . . . 115

7.3 Portfolio risk and return . . . . . . . . . . . . . . . . . . . . 119

7.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

3
7.5 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129

8 Portfolio theory and the Capital Asset Pricing Model 130

8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

8.2 Investor preferences . . . . . . . . . . . . . . . . . . . . . . . 131

8.3 Basic portfolio theory . . . . . . . . . . . . . . . . . . . . . . 132

8.4 Optimal portfolios for investors: stocks only . . . . . . . . . 138

8.5 Optimal portfolio selection: stocks plus a risk-free asset . . . 143

8.6 Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148

8.7 The CAPM . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

8.8 Competing models . . . . . . . . . . . . . . . . . . . . . . . . 159

8.9 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161

8.10 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

9 Efficient markets 164

9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164

9.2 Expected returns versus profits . . . . . . . . . . . . . . . . . 165

9.3 Defining efficiency . . . . . . . . . . . . . . . . . . . . . . . . 165

9.4 Foundations of efficiency . . . . . . . . . . . . . . . . . . . . . 167

9.5 Efficiency: making it an operational concept . . . . . . . . . . 167

9.6 Efficiency and random returns . . . . . . . . . . . . . . . . . 171

9.7 Efficiency and empirical methods . . . . . . . . . . . . . . . . 173

9.8 Puzzles and anomalies from an efficient markets perspec-


tive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175

4
9.9 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176

9.10 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179

10 Forward and futures markets 180

10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180

10.2 Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181

10.3 Forwards and futures contracts . . . . . . . . . . . . . . . . . 183

10.4 Futures contracts . . . . . . . . . . . . . . . . . . . . . . . . . 186

10.5 Pricing derivatives . . . . . . . . . . . . . . . . . . . . . . . . 188

10.6 Forwards and futures prices . . . . . . . . . . . . . . . . . . . 191

10.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194

10.8 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

11 Options 197

11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197

11.2 Option payoffs . . . . . . . . . . . . . . . . . . . . . . . . . . . 200

11.3 Option combinations: payoffs and uses . . . . . . . . . . . . 206

11.4 Put-call parity . . . . . . . . . . . . . . . . . . . . . . . . . . . 211

11.5 Arbitrage and option price bounds . . . . . . . . . . . . . . . 212

11.6 Options pricing . . . . . . . . . . . . . . . . . . . . . . . . . . 214

11.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225

11.8 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227

12 Sample exam paper with worked solutions 228

5
12.1 Sample exam paper . . . . . . . . . . . . . . . . . . . . . . . . 229

12.2 Sketch solutions to sample exam . . . . . . . . . . . . . . . . 233

6
Chapter 1

Introduction

1.1 Route map to the guide

The aim of this subject guide is to help you to move through the Asset pric-
ing and financial markets syllabus and, in particular, to help you through
the required reading. As such, it is a supplement to, not a substitute for,
the recommended textbook. The guide contains (at least) one chapter for
each major topic in the syllabus, arranged in the same order as they are
found in the syllabus. Topics, and therefore chapters, are inter-dependent.
Therefore it is not advisable to begin studying, for example, Chapter 7,
without having read Chapters 1-6.

We would suggest that for each topic in the syllabus, you begin by read-
ing through the relevant chapter in the subject guide, before reading the
recommended parts of the textbook for that topic. Then engage in the ac-
tivities assigned to the topic. These activities will provide you with feed-
back indicating the parts of the topic that you need to put more work into.
You should then return to those parts in the guide and the textbook (and
perhaps the other suggested books) to cover them a second time.

7
1.2 Course coverage and syllabus

This guide is intended to accompany an undergraduate student’s first


course in investments. The goal of the guide is to introduce readers to
the techniques used to price financial assets. As a side product, readers
will be familiarised with the most common varieties of financial securities
and the ways in which they are traded.

A few key ideas underlie almost everything that follows. These are all
related to the assumptions we make regarding the basic motivations of
investors:

• Impatience and the time value of money. Investors value £1 re-


ceived today more highly than they value £1 to be received at any
point in the future (even if that future cash is certain to arrive).
• Greed. Investors are assumed to prefer assets with high expected re-
wards to those with low expected rewards, other things being equal.
• Risk aversion and risk premia. Investors, other things being equal,
prefer less risky situations to more risky situations. Thus, assets that
carry more risk must compensate investors by delivering greater ex-
pected rewards than those that carry less risk. The difference be-
tween the promised rewards from holding risky and risk-less assets
is called a risk premium.
• No-arbitrage. If two assets (or two combinations of assets) are cer-
tain to deliver exactly the same cashflows in every possible future
eventuality, then they must sell for the same price.1

Using these ideas we will price bonds, stocks and finally simple derivative
securities (e.g. futures and options).

We will begin by introducing the time value of money and showing how
it can be used to value and compare investments that deliver streams of
1 This result is a by-product of greed as, if two assets that always delivered the same
cashflows sold for different prices, a greedy investor would buy the cheaper asset and
sell the more expensive one, yielding an instant risk-free profit. This trade would place
buying pressure on the cheaper asset and selling pressure on the more expensive one
such that their prices would tend to equalise. The trading, and the associated profits,
would only cease once the prices were identical.

8
future cash (with the cashflows from different investments varying in both
their timing and size). We will then apply this valuation method to gov-
ernment bonds. We will show how they are priced and we will discuss
how to measure the risk an investor faces from holding bond positions.

We will then move on to the pricing of stocks and shares. Again this will
start with arguments based on the time value of money. We will develop
the famous Capital Asset Pricing Model for stocks, show how it fares when
confronted with real stock price data and then discuss modifications to it
that are required to square the model with the data.

At this point, we will digress slightly. We will review the academic litera-
ture that exists on the question of whether stock markets are informationally
efficient. This is essentially the same as asking whether investors can iden-
tify stocks that are wrongly priced and then trade those mis-priced stocks
so as to make a profit. We will discuss the theoretical underpinnings of
efficiency before reviewing the evidence on efficiency.

Finally, we will introduce the most common types of derivative assets.


These are forward contracts, futures contracts and option contracts. We
will discuss how they are structured, how they are used by investors and
finally how they are priced. In this case, pricing will use the no-arbitrage
method.

In list form, the syllabus for the course is as follows;

1. Present value methods

2. Valuing long-lived cashflow streams

3. Understanding interest rates

4. Understanding and pricing government bonds

5. Valuation of stocks

6. Risk, return and stock portfolios

7. Portfolio theory and the CAPM

8. Market efficiency

9. Derivative securities and derivative pricing

9
Inevitably, readers will require some previous exposure to mathematics
and statistics to follow the treatment here. The level is not high (maybe
advanced school level or first year undergraduate level), but the quantita-
tive tools are necessary to develop the required frameworks. In Sections
1.8 and 1.9 below, I give a brief overview of the most important pieces of
maths and stats for our purposes.

The up-to-date course syllabus for FN2190 Asset pricing and financial
markets can be found in the course information sheet, which is available
on the course VLE (virtual learning environment) page or on the LSE web-
site: www.lse.ac.uk/study-at-lse/uolip/course-information-sheets

1.3 Aims of the course

This course aims to:

• provide students with a thorough grounding in asset pricing.


• develop students’ skills in applying pricing methods to realistic sce-
narios.
• provide a critical overview of the research on financial market effi-
ciency.
• allow students to develop an understanding of how securities mar-
kets operate.

1.4 Learning outcomes for the course

Having completed the course and done the required reading and assess-
ments, a student should be able to:

• explain how to price assets using both present value and absence of
arbitrage methods.
• describe the important differences between stock, bond and deriva-
tive securities.

10
• apply present value techniques to price stocks and bonds.

• employ mathematical tools to compute risk and return for portfolios


of securities.

• evaluate portfolio choice problems.

• present, explain and apply the Capital Asset Pricing Model for com-
puting expected stock returns.

• critically evaluate the evidence for informational efficiency of stock


markets.

• price derivative securities using absence of arbitrage.

1.5 Essential and further reading

For each topic we cover, essential reading is given from the following text-
book:

• Brealey, Richard A., Stewart C. Myers, and Franklin Allen. 2017.


Principles of corporate finance. New York, NY: McGraw-Hill/Irwin.

There are many other excellent introductory finance textbooks available.


Another very good book, which you might use to supplement your read-
ing from the core textbook is:

• Berk, Jonathan and Peter DeMarzo. 2017. Corporate finance. Boston,


Mass.: Pearson.

Detailed reading references in this subject guide refer to the editions of the
set textbooks listed above. New editions of one or more of these textbooks
may have been published by the time you study this course. You can use
a more recent edition of any of the books; use the detailed chapter and
section headings and the index to identify relevant readings. Also check
the VLE regularly for updated guidance on readings.

11
1.6 Online study resources

In addition to the subject guide and the Essential reading, it is crucial that
you take advantage of the study resources that are available online for this
course, including the VLE and the Online Library.

You can access the VLE, the Online Library and your University of London
email account via the Student Portal at: https://my.london.ac.uk

You should have received your login details for the Student Portal with
your official offer, which was emailed to the address that you gave on your
application form. You have probably already logged in to the Student Por-
tal in order to register! As soon as you registered, you will automatically
have been granted access to the VLE, Online Library and your fully func-
tional University of London email account. If you have forgotten these
login details, please click on the ‘Forgotten your password’ link on the
login page.

1.6.1 The VLE

The VLE, which complements this subject guide, has been designed to en-
hance your learning experience, providing additional support and a sense
of community. It forms an important part of your study experience with
the University of London and you should access it regularly.

The VLE provides a range of resources for EMFSS courses:

• Course materials: Subject guides and other course materials avail-


able for download. In some courses, the content of the subject guide
is transferred into the VLE and additional resources and activities
are integrated with the text.
• Readings: Direct links, wherever possible, to essential readings in
the Online Library, including journal articles and e-books.
• Video content: Including introductions to courses and topics within
courses, interviews, lessons and debates.
• Screencasts: Videos of PowerPoint presentations, animated podcasts
and on-screen worked examples.

12
• External material: Links out to carefully selected 3rd-party resources

• Self-test activities: Multiple-choice, numerical and algebraic quizzes


to check your understanding.

• Collaborative activities: Work with fellow students to build a body


of knowledge.

• Discussion forums: A space where you can share your thoughts and
questions with fellow students. Many forums will be supported by
a ‘course moderator’, a subject expert employed by LSE to facilitate
the discussion and clarify difficult topics.

• Past examination papers: We provide up to three years’ of past exam-


inations alongside Examiners’ commentaries that provide guidance
on how to approach the questions.

• Study skills: Expert advice on getting started with your studies, prepar-
ing for examinations and developing your digital literacy skills.

Note: Students registered for Laws courses also receive access to the dedi-
cated Laws VLE. Some of these resources are available for certain courses
only, but we are expanding our provision all the time and you should
check the VLE regularly for updates.

1.6.2 Making use of the Online Library

The Online Library (http://onlinelibrary.london.ac.uk) contains a huge


array of journal articles and other resources to help you read widely and
extensively.

To access the majority of resources via the Online Library you will either
need to use your University of London Student Portal login details, or you
will be required to register and use an Athens login.

The easiest way to locate relevant content and journal articles in the Online
Library is to use the Summon search engine.

If you are having trouble finding an article listed in a reading list, try re-
moving any punctuation from the title, such as single quotation marks,
question marks and colons.

13
For further advice, please use the online help pages, (which are found
at http://onlinelibrary.london.ac.uk/resources/summon) or contact the
Online Library team: onlinelibrary@shl.london.ac.uk

1.7 Exam structure and revision strategy

Important: the information and advice given here are based on the ex-
amination structure used at the time this guide was written. Please note
that subject guides may be used for several years. Because of this we
strongly advise you to always check both the current Programme regula-
tions for relevant information about the examination, and the VLE where
you should be advised of any forthcoming changes. You should also care-
fully check the rubric/instructions on the paper you actually sit and follow
those instructions.

The final exam for this course is three hours long and requires you to an-
swer three questions from a set of four questions. A sample exam paper is
given at the end of this guide. This paper reflects the examination and as-
sessment arrangements for this course in the academic year 2016-17. The
format and structure of the examination may change in future years, and
any such changes will be publicised on the virtual learning environment
(VLE).

There are exercises for the reader to attempt at the end of each chapter. All
of these exercises should be regarded as possible (parts of) sample exam
questions. Solutions for these exercises will be provided on the VLE.

Many candidates are disappointed to find that their examination perfor-


mance is poorer than they expected. This may be due to a number of
reasons, but one particular failing is ‘question spotting’, that is, confin-
ing your examination preparation to a few questions and/or topics which
have come up in past papers for the course. This can have serious conse-
quences.

We recognise that candidates might not cover all topics in the syllabus
in the same depth, but you need to be aware that examiners are free to
set questions on any aspect of the syllabus. This means that you need to
study enough of the syllabus to enable you to answer the required number
of examination questions.

14
The syllabus can be found on the Course information sheet available on
the VLE. You should read the syllabus carefully and ensure that you cover
sufficient material in preparation for the examination. Examiners will vary
the topics and questions from year to year and may well set questions that
have not appeared in past papers. Examination papers may legitimately
include questions on any topic in the syllabus. So, although past papers
can be helpful during your revision, you cannot assume that topics or spe-
cific questions that have come up in past examinations will occur again.

If you rely on a question-spotting strategy, it is likely that you will find


yourself in difficulties when you sit the examination. We strongly ad-
vise you not to adopt this strategy.

Remember, it is important to check the VLE for:

• up-to-date information on examination and assessment arrangements


for this course

• where available, past examination papers and Examiners’ commen-


taries for the course which give advice on how each question might
best be answered.

1.8 Mathematical pre-requisites

We will use basic maths continuously. Amongst other things, we will


make a lot of use of the following:

• Sequences. Arithmetic and geometric progressions.

• Manipulation of basic algebraic expressions.

• Solving systems of equations.

• Graphing equations.

Finally, in some of the more advanced pieces of analysis we will use some
basic matrix algebra.

15
Summation and product notation

To avoid having to write out long sums of numbers or products of num-


bers, we will often make use of the following shorthand.

• Summation: if you want to write the sum of K numbers called Zi ,


where i runs from 1 to K, we will use the following notation:

ΣiK=1 Zi = Z1 + Z2 + . . . + ZK

• Product: similarly, if you want to write the product of Zi where i


runs from 1 to K, we define:

ΠiK=1 Zi = Z1 × Z2 × . . . × ZK

Basic calculus

The investors in our analysis will be looking to maximise well-being or


sometimes to minimise risk. Often they will want to know how sensitive
a particular outcome is to a particular input variable. In characterising all
of these situations we will require some basic differential calculus.

Take the generic function:

y = f (x)

dy
We will denote the derivative of y with respect to x either as dx or as f 0 ( x ).
The derivative should be thought of as giving the slope of the function
f ( x ), and this is why it is useful as a measure of the sensitivity of y to
changes in x.

For example, if:

y = cx k

16
Then:

dy
= ckx k−1
dx

If we had the natural log function:

f ( x ) = ln( x )

then the first derivative would be:

1
f 0 (x) =
x

d2 y
Extending this notation, the second derivative will be denoted by dx2
or
f 00 ( x ).

1.9 Statistical pre-requisites

Random variables and probabilities

Risks and gambles are central to theoretical and practical finance. We will
therefore rely heavily on tools that describe them. Here is a brief refresher.

First we must define what we mean when we describe something as a


random variable. Consider an experiment that has a (known) set of possi-
ble outcomes but that, before running the experiment, one does not know
which of the set of possible outcomes will actually come to pass. The ob-
vious example of this is a coin toss, where there are two possible outcomes
(heads and tails), but which of the two will actually occur when you toss
the coin is not known in advance.

A random variable is an experiment of this type, but where additionally


one has attached numerical labels to the outcomes. So in the coin toss
experiment, if the outcome ‘heads’ is labelled with a 1 and the outcome
‘tails’ is labelled with a zero, then we have a random variable. A more

17
finance-relevant example of a random variable is the change in the S&P-
500 stock index over the period starting today and ending one month from
now. This change already has a numerical value (e.g. if the S&P-500 falls
by 50 points over the month, then the value is -50) and there are lots of
possible outcomes for the evolution of the S&P-500 over the month, but
we don’t know for sure which of these outcomes will come to pass.

Next we need to define rules that govern how likely individual outcomes
of our random variable are i.e. we need to think about probabilities. To
begin, consider a random variable X and assume that:

• There are m possible outcomes for the variable, which we will label
as X1 , X2 , . . . , Xm−1 , Xm

• The probability of Xi occurring is Pi i.e. Pr( X = Xi ) = Pi .

These probabilities give us information about the relative likelihood of the


outcomes (i.e. if P4 is twice as big as P7 then X4 is twice as likely to be the
outcome than is X7 ). But the probabilities must obey certain rules. These
are:

• We must have Pi ≥ 0 for all i

• We must have Pi ≤ 1 for all i

• We must also have Σim=1 Pi = P1 + P2 + . . . + Pm = 1

So no outcome can have negative probability. An event can be impossible


(i.e. it has zero probability), but there’s no such thing as a negative likeli-
hood. No outcome can have probability larger than 1. A probability of 1
means that the outcome is certain to occur and thus the probability cannot
rise above this level for any possible outcome. Finally, if you aggregate
over all possible outcomes, the probabilities must sum to one. This final
requirement is essentially telling you that you have a complete description
of the set of outcomes (as if you summed probabilities and they came to a
number less than 1, then you must have missed a possible outcome from
your counting).

As an example of the above, consider a fair, six-sided dice. In this case m


is equal to 6 and the probability of any outcome is one-sixth, so that Pi = 16

18
for each i. So in this case the rules of probability are obeyed. No prob-
ability is below zero or bigger than one and when you sum probabilities
across all possible outcomes you get precisely 1.

Expected values

The expected value or mean of a random variable is its probability weighted


average. Thus, using the notation we set up previously, the expected value
of X is:

m
E( X ) = ∑ Pi Xi = P1 X1 + P2 X2 + . . . + Pm Xm
i =1

In a finance context, the mean outcome from an investment is the amount


that you should expect to win or lose on average. Another way of thinking
about the expected value is that it is equal to one’s best guess as to the
outcome of the investment.

If one has two random variables and forms a linear combination of them,
then the expected value of the linear combination is equal to the linear
combination of the expected values. So if X and Y are random variables
and a and b are constants we have:

E( aX + bY ) = aE( X ) + bE(Y ) (1.1)

This rule will be useful to us, as investors rarely place all of their funds
in one asset. They spread their money across investments and using the
rule above we can work out the expected payoff from investing in a set
of assets, given knowledge of the expected values of each member of the
set. Spreading one’s wealth across investments is usually called building
a portfolio of assets.

Variances and standard deviations

The variance of a random variable measures the (probability weighted)


dispersion of the outcomes around the mean.

19
h i
2
Var( X ) = E ( Xi − E( X ))
m
∑ Pi (Xi − E(X ))
2
=
i =1
= P1 ( X1 − E( X ))2 + P2 ( X2 − E( X ))2 + . . . + Pm ( Xm − E( X ))2

As squared deviations from the mean are always positive and probabilities
are always positive, the variance must always be positive. If two random
variables have the same mean but different variances, the variable with the
larger variance has outcomes that are, on average, more spread out around
the mean than the variable with smaller variance. We usually denote the
variance of a random variable with the Greek notation, σ2 .

The standard deviation is the square root of the variance and is usually
denoted by σ. We will often use the standard deviation, rather than the
variance, as the standard deviation is expressed in the same units as the
random variable itself (e.g. if your random variable has outcomes mea-
sured in US dollars, then your standard deviation is also measured in dol-
lars).

In our analysis, investors will use the variance or standard deviation of


their investment outcomes to measure the financial risk that they face and,
as we have mentioned above, we will assume that they prefer investments
with lower financial risk, other things being equal.

Covariance and correlation

The correlation, ρ, between two random variables tells us the sign and the
strength of the (linear) relationship between them. It always lies between
-1 and +1 and we have the following interpretation:

• A correlation of exactly zero means that the two variables are (lin-
early) unrelated
• A positive correlation means that two variables are positively related
• A negative correlation means that the two variables have a negative
relationship

20
• The magnitude of the correlation describes the strength of the rela-
tionship

Thus two variables with a correlation of -0.85 are very strongly negatively
related. This means that when one of them is above average we should
expect the other to be below average. It won’t always turn out to be true
that when one is above average the other is below average, but because of
the strength of the correlation this will be true most of the time.

The covariance between two variables, X and Y, is equal to:

m
Cov( X, Y ) = E [( X − E( X ))(Y − E(Y ))] = ∑ Pi (Xi − E(X )) (Yi − E(Y ))
i =1

so it is the expected value of the product of the deviations of X and Y from


their respective averages. Quick inspection of the equations defining the
covariance and the variance will tell you that Cov( X, X ) = Var( X ).

The correlation and covariance between X and Y are related. The correla-
tion is defined as:

Cov( X, Y )
ρ X,Y =
σX σY

where σX is the standard deviation of X and σY is defined similarly.

Because standard deviations are always positive, the sign of the covari-
ance is always the same as the sign of the correlation. Thus the signs of the
correlation and the covariance carry exactly the same information. The
magnitude of the covariance, however, depends on the scale of the vari-
ables X and Y and so it tells us nothing about the strength of the relation-
ship between the variables. Thus, the correlation is usually a more useful
number than the covariance.

Finally, now that we have defined correlation and covariance, we can talk
about the variance of a linear combination of random variables. We have:

Var( aX + bY ) = a2 σX2 + b2 σY2 + 2ab ρ X,Y σX σY (1.2)

21
So the variance of a linear combination of random variables depends on
the variances of the constituent random variables and the correlation be-
tween them. Holding everything else constant, the higher the correlation,
the higher the variance of the combination.

We will use this result repeatedly in order to work out the financial risk
associated with portfolios of financial assets.

1.10 Summary

In this opening chapter we have described the basic concepts that we will
use to price assets and given a brief overview of the path we will take in
the remaining chapters. Then we have spent some time reviewing mathe-
matical and statistical tools that will be essential for our analysis.

22
1.11 Exercises
1. Consider the function y = 6x2 − 3x + 14.
(a) What are the first and second derivatives of this function?
(b) At what value of x is there a turning point in this function? Is
this turning point a maximum or a minimum? What is the value
of y at the turning point? Justify all of your answers.
2. You simultaneously roll two six-sided dice and create a random vari-
able by summing the values shown on the upward faces.
(a) List the possible outcomes of this random variable.
(b) What is the probability of each possible event?
(c) What is the mean outcome?
(d) What is the variance of the outcomes?
3. A variable X has a mean of 12 and a variance of 10. Another variable
Y has a mean of 20 and a variance of 40. You form a new variable, Z,
which is equal to 12 ( X + Y ).
(a) What is the mean of Z?
(b) Assume that X and Y have a correlation of zero. What is the
variance of Z?
(c) What is the variance of Z if the correlation between X and Y is
+0.5 and what is the variance if, instead, the correlation is -0.5?
(d) What pattern do you see in your findings from parts (b) and (c)?
4. Give a TRUE/FALSE answer for each of the following statements
and in each case justify your answer.
(a) The first derivative of the function f ( x ), evaluated at the value
x = 5, gives the slope of the function at the value x = 5.
(b) The expected value of a random variable is equal to the value of
the most probable outcome.
(c) The expected value of a random variable must be positive.
(d) The variance of a linear combination of random variables can be
smaller than the variance of either of the constituent variables.
(e) If the correlation between two variables is +0.5, then when one
is above average the other will always be above average.

23
Chapter 2

Present value computations

Essential reading
Brealey, Myers and Allen (2017), Chapter 2 section 1.

Learning outcomes
By the end of this chapter, and having completed the Essential reading and
activities, you should be able to:

• use compounding and discounting techniques to value cashflow streams.

• use the Net Present Value rule to evaluate investment opportunities.

• explain why the NPV rule is optimal.

2.1 Introduction

This chapter builds a set of tools for valuing projects and assets based on
a single, simple foundation. That foundation is the result that, in a world
where interest rates are positive, £1 to be received today has a greater value
than £1 to be received at any future date. Or, to generalise slightly, £1 to
be received K months from now is preferred to £1 cash to be received M

24
months from now if M ≥ K. This result has come to be known as the time
value of money.

First we will show where this result comes from, using the notion of a bank
account paying a given rate of interest. We will then use the bank account
(or equivalently the interest rate) as a tool to derive the future value of a
current cashflow and the present value of a future cashflow. Finally, we will
focus on present values of future cashflows to derive the well-known Net
Present Value rule for investment valuation and appraisal.

2.2 Foundation: interest rates and discount rates

The most important building block in the analysis of this chapter is the
notion of a discount rate or interest rate.1

An interest rate is the reward that investors demand for accepting delayed
rather than immediate gratification. It is the compensation that a saver
demands for handing their money over to a bank for some period and
thus not being able to use that money immediately for the purposes of
consumption.

Interest rates may vary according to who is lending the money and who
is borrowing the money. If one lends money to a less trustworthy per-
son, then one should require a greater interest rate than if one lends to
an entirely trustworthy person. The increased interest rate here is com-
pensation for the possibility that the less trustworthy person refuses to (or
simply cannot) pay you back.

For now, we are going to assume that the interest rate that one receives
every period on one’s bank account is r and that this rate is fixed (over
time) and known to all borrowers and lenders.
1 We will use these terms interchangeably. We may also refer to the discount rate as

the required rate of return or opportunity cost of capital.

25
2.2.1 Compound interest versus simple interest

As we begin to think about bank accounts and securities that pay interest,
it is important to distinguish between two ways of paying interest. To
illustrate them, we assume a bank account that pays interest every period
at rate r and that a saver deposits an initial balance of X in the account.

Simple Interest

If the bank account pays simple interest, the interest earned by the depos-
itor (or paid by a borrower) each period is just the original balance of the
deposit times the interest rate.

So if I initially deposit X then after one period the balance is:

X + rX = X (1 + r )

After T periods, as in each period the interest is only paid on the initial
deposit, the balance in the account is:

X + rXT = X (1 + rT )

Thus, if one just deposits some cash and then one leaves the account alone,
the balance in the account will grow linearly over time.

But, normal bank accounts (and bank loans and credit cards) do not work
according to the rules of simple interest.

Compound interest

Real world bank accounts pay (and charge) interest using the idea of com-
pound interest. The crucial difference between simple and compound in-
terest is that in the compounding case, interest is paid on the entire current
balance of the bank account, not just the initial deposit.

26
So, again, if I initially deposit X in a bank account paying compound in-
terest at rate r then after one period the balance is:

X + rX = X (1 + r )

At the end of the second period, one earns interest at rate r on the total
balance that was in the account at the end of the first period. Thus at the
end of the second period one has:

X (1 + r ) × (1 + r ) = X (1 + r )2

By the same logic, after T periods, the total balance in your bank account
will be:

X (1 + r ) T

Clearly, a saver prefers a bank that pays compound interest at rate r to


one that pays simple interest at rate r. Why? Because in the compound
case, in every period after the first, one is earning interest on previously
paid interest as well as interest on the initial deposit. Figure 2.1 below
shows the evolution of the balances in two bank accounts that differ only
in whether they pay simple or compound interest.

Finally, for this section, a word of warning. Compound interest is clearly


good for savers but, conversely, it is bad for borrowers. Your credit card
company charges you compound interest, which means that your credit
card debt, if left untouched, will grow faster than it would if the company
charged you simple interest.

From this point onwards in this chapter, we will assume the existence of
a bank that takes deposits and offers loans. The interest rate paid on de-
posits is the same as that charged on loans, at r per period, and the bank
uses compound interest. The bank is risk-free, in the sense that it never
goes bankrupt.

27
Figure 2.1: The effects of compound versus simple interest on bank bal-
ance, assuming an initial deposit of £100 and an interest rate of 10%

700
Compound interest
Simple interest
600
Balance on account (£)

500

400

300

200

100
0 5 10 15 20
Years

2.3 Future values of current cashflows

Now that we have access to a bank account, we can answer the follow-
ing question. If I am given the choice between £100 today and £X to be
received one period from now, which do I prefer? Assume that both the
current and future payment are risk-less (i.e. the payments are guaran-
teed).

We will approach this question by using the bank account to transform the
£100 to be received today into a cashflow to be received in one period. If I
put my £100 in the bank for one period, at the end of that period the bank
balance is:

100(1 + r )

Thus we view £100 paid today as equivalent to £100(1 + r ) in one period’s


time.

28
Using this logic we can answer the original question. If the amount X I am
promised in one period is smaller than 100(1 + r ) then I prefer the £100
today while if X is larger than 100(1 + r ) then I prefer the payment of £X
in one period. Finally, if X = 100(1 + r ) then I am indifferent between the
alternatives.

What we have done here is to define and use the future value of a current
cashflow. The future value of £100 to be received today is £100(1 + r ) in one
period’s time. More generally, we can define the future value in T periods
of some cashflow Z to be received today as:

FV ( Z, r, T ) = Z (1 + r ) T

What we are essentially saying here is that, due to the existence of a risk-
free bank account paying compound interest at rate r, Z received today
is economically equivalent to Z (1 + r ) T to be received in T periods from
now.

Clearly, future values depend on three things: the initial cashflow (Z), the
interest rate (r) and the period of time over which you’re computing the
future value (T). The higher is any one of these three things, the greater
will be the future value.

29
Future values: example.

Scenario: you inherit £100,000 today and decide to place it in a bank


account for 10 years. At the end of that 10 year period you’ll withdraw
the funds and use the money to help buy your daughter a house.

• How much will you be able to contribute to her house purchase


assuming that the annual interest rate is 5% for the next 10 years?

• Sensitivity analysis: how would your answer change if interest


rates turn out to be 2% per year or 10% per year?

Solution: what the question is asking you for is the future value of your
inheritance under various different interest rate scenarios. The baseline
answer, assuming a 5% rate, and then the sensitivity analysis are given
below:

FV (100, 000, 5%, 10) = 100, 000 × (1 + 0.05)10 = 162, 889.46


FV (100, 000, 2%, 10) = 100, 000 × (1 + 0.02)10 = 121, 899.44
FV (100, 000, 10%, 10) = 100, 000 × (1 + 0.10)10 = 259, 374.25

2.4 Present values

When computing a future value of a current cashflow one is essentially us-


ing the bank account to transform the current cash into a (larger) amount
of cash at some specified point in the future. To compute the present value
of a future cashflow, one does the reverse.

Here is the intuition. Assume that you’re due to receive a payment of X in


T years. What current cashflow is equivalent to that future cashflow?

Using the same logic as in the previous section, the equivalent current
cashflow is the amount that you would have to place in a bank today such
that, with payment of interest, it would grow to be worth exactly X in T

30
years.2 This equivalent current cashflow is called the Present Value of X
and it is equal to:

X
PV ( X, r, T ) =
(1 + r ) T

Why is this quantity equivalent to the receipt of X in T years? Well if you


placed this amount in a bank account paying compound interest at rate r
for T years you would end up with:

X
PV ( X, r, T ) × (1 + r ) T = × (1 + r ) T = X
(1 + r ) T

Note that, in a world of positive interest rates, the present value is always
smaller than the actual future cashflow (X). So the present value of $100
to be received in 5 years is definitely less than $100.

Note also that, other things being equal, the present value will be smaller
when interest rates are higher and it will also be smaller when the cashflow
is to be received a long time from today (i.e. when T is large).

2.4.1 Discount factors

When computing present values we often make use of an object called a


discount factor. A discount factor is just the present value of £1. Thus, the
T period discount factor when interest rates are at rate r is:

1
DF (r, T ) =
(1 + r ) T

So, discount factors vary with the interest rate and with the investment
horizon. Higher interest rates lead to lower discount factors and longer
investment periods lead to lower discount factors
2 This is equivalent to asking how much a bank would lend you today if you promise

to give them a repayment of X in T years.

31
Discount factors are useful as to compute the present value of a particular
cashflow, X to be received in T periods one just has to multiply X by the
appropriate discount factor i.e.

PV ( X, r, T ) = X × DF (r, T )

Present values: example.

Scenario: you are a supplier of parts to Rolls Royce. In 2 years you are
due to receive £5,000,000 as payment for components you have supplied.

• How much is this worth in current terms assuming a current an-


nual interest rate of 1%?

• Sensitivity analysis: how would your answer change if interest


rates were to shift to 5% per year?

Solution: the baseline answer and then the sensitivity analysis are given
below:

PV (1%, 2) = 5, 000, 000 × (1 + 0.01)−2 = £ 4, 901, 480.25


PV (5%, 2) = 5, 000, 000 × (1 + 0.05)−2 = £ 4, 535, 147.39

Thus, clearly, even over short horizons, with reasonably high interest
rates, present values can be much smaller than actual future cashflows.

2.5 Comparing single cashflows to be received


at different dates

The basic use of either present value or future value calculations is to com-
pare cashflows received at different dates.

32
If one is trying to choose between the receipt of $100 today and the receipt
of $120 in two years from now, the answer is not immediately obvious.
The principle that must be followed is that if one is comparing cashflows,
those cashflows must always be expressed in terms of the same date. So,
in this case, what one should do is either:

• Compare the cashflow of $100 today with the present value of the
$120 to be received in two years.

• Compare the future value in two years’ time of the $100 received
today with the cashflow of $120 to be received in two years.

In either case, you are moving one of the cashflows through time until it
lines up, in date terms, with the other. Then you can directly compare their
values and choose the biggest positive cashflow. Note that whichever way
you choose to do the comparison (using future or present values), you’ll
always come to the same conclusion.

To complete our example, let’s assume that the one year interest rate is
5%. Then which of the two alternatives has greater value. Via the present
value approach we compute the present value of the $120 to be received
in two years. This is:

120
= 108.84
(1 + 0.05)2

Then, as $108.84 is great than $100, we conclude that the cashflow of $120
in two years is more valuable than the instant cashflow of $100.

If you use the future value approach, you need the value in two years of
$100 received today. This is:

100 × (1 + 0.05)2 = 110.25

Then, as $120 is bigger than $110.25, again we see that the cashflow of $120
in two years is worth more than the current cashflow of $100.

In both cases, your computations tell you that you should select the option
that pays you $120 in two years.

33
2.6 Net Present Value

Until now we’ve thought about single cashflows (current or future) only.
However, most investments involve multiple cashflows received and paid
at different points in time (where positive cashflows are receipts and neg-
ative cashflows are payments). Now we investigate how one computes
present values for these more complicated streams of money. The solution
is just a generalisation of the single cashflow examples above.

If one wants to compute the present value of an entire cashflow stream


then one:

• Takes each individual cashflow and computes its present value.

• Sums all of these individual present values.

This sum is called the Net Present Value or NPV. Note that receipts of cash
will contribute positively to a NPV and payments will contribute nega-
tively to the NPV.

Calculating NPV

Assume a constant interest rate of r per period and assume an investment


product which will deliver a cashflow of C1 at the end of period 1, C2 at the
end of two periods, continuing until it finishes by delivering a cashflow
of CK after K periods. We might also assume that there is an immediate
cashflow, C0 .

The NPV of this investment is:

K
C1 C2 CK Ci
NPV = C0 + +
(1 + r ) (1 + r ) 2
+ . . . +
(1 + r ) K
= ∑ i
i =0 (1 + r )

Again, note that some of the Ci might be negative (if they are payments
not receipts).

34
The NPV rule: evaluating investments

Consider a financial or physical investment project for which you have


calculated the NPV. If you obey the following rules then you will always
choose good investments:

• If the NPV is positive you should invest in the project.


• If the NPV is negative, you should turn down the investment oppor-
tunity.
• If you have a menu of investment projects, but you can only afford
to invest in one of them, then the right thing to do is to invest in the
project which has the largest NPV.

Often, when evaluating a project, the future cashflows are not known with
certainty. In that case, use the expected cashflows in the NPV calculation.
Also, as we will see later on in the course, if investments differ in their
level of risk then you should use a different discount rate when calculating
NPVs for those projects. When calculating the NPV of a more risky project
use a higher discount rate than if you’re calculating the NPV of a safe
project.

NPV: example.

Scenario: a building company is considering buying a plot of land cost-


ing £2,000,000 today. It plans to build five houses on the plot. Construc-
tion will take two years to complete, with each house’s building costs
comprising £75,000 in one year and £40,000 in two years. If the com-
pany expects to sell each house for £600,000 two years from now and
the interest rate is 8% per annum, should the project be taken on?

Solution: compute the NPV of the promised cashflow stream. This is:

5 × 75, 000 5 × (600, 000 − 40, 000)


NPV = −2, 000, 000 − + = 53, 326.47
(1 + 0.08) (1 + 0.08)2
As the NPV is positive, the company should go ahead with the project.

35
2.7 The optimality of the NPV rule

Why should we trust the NPV rule? Why is it optimal for individuals to
invest in projects with positive NPV and discard projects with negative
NPV?

It turns out that NPV is optimal (under some assumptions) in the sense
that use of the rule leads investors to maximise their expected wealth. This
is true regardless of how patient or impatient an investor is, and thus the
rule can be used for all investors (they will all agree on which investments
to choose and which to discard). Here is an example to give you the intu-
ition for this result

Consider two investors who live in a world with only two dates (’now’
and ’one year from now’):

• Nick (N) wants to consume now (he is impatient).

• Penny (P) wants to consume in one year (she is patient).

• Both have current income of £150,000 and expect zero income in a


year.

• Each can invest in an opportunity costing £150,000 now and return-


ing a guaranteed £170,000 at the end of the year.

• They also have access to a risk-free bank where the borrowing and
lending rate is 5% per annum.

We will show that, despite the fact that the two investors differ in when
they want to consume, they should both invest in the project. The reason
that they should both invest in the project is because its NPV is positive:

170, 000
NPV = − 150, 000 = 11, 904.76
(1 + 0.05)

36
Figure 2.2: Nick versus Penny and the optimality of NPV

$ ’000s Next Year


Penny invests in the project rather
than putting her cash in the bank
170 and consumes 170 in a year

157.5

Nick first invests and then


borrows against the proceeds
to consume 161.9 now

$ ’000s Now
150 161.9

37
2.7.1 Analysis of Penny’s situation:

Penny wants to consume at the end of the year. If she took her original
wealth, invested in the risk-free asset and consumed the proceeds, she
would have £157,500 to consume. Instead, if she invests in the project she
has £170,000 to consume at the end of the year. Thus, she is better off
investing in the project.

2.7.2 Analysis of Nick’s situation:

Nick wants to consume today. He could just consume his original wealth,
of £150,000.

Instead, he could do the following:

• He invests in the project, meaning that he has £170,000 in income at


the end of the year.

• He borrows the present value of £170,000 from the bank today. At


5% interest, this gives him £161,904.76 today.

• He consumes the £161,904.76 today and in one year uses the pro-
ceeds of the investment to repay the bank.

He is better off investing in the project and using the capital market (i.e.
borrowing and lending) to arrive at his desired consumption pattern.

2.7.3 Nick and Penny together

The outcome here is that both individuals end up investing in the project
as they are both better off if they do so. Why? Because, as we have shown
above, the project has positive NPV. They should both exploit the positive
NPV opportunity and then borrow/lend to move income through time
and arrive at an optimal consumption pattern.

The generalisation of this example is that everyone, regardless of their


preferences, should use the NPV rule to decide on which projects to invest

38
in. Also, the decision as to which projects to invest in can be separated
from the decision regarding when to consume. Use the NPV rule to de-
cide which projects to invest in and then use the bank account to move
money through time to suit preferences regarding when to consume.

Self-assessment exercise

Work through the preceding analysis once more, but under the assump-
tion that the investment has a negative NPV (e.g. assume that the project
pays off £155,000 not £170,000).

39
Reminder of learning outcomes
Having completed this chapter, and the Essential reading and activities,
you should be able to:

• use compounding and discounting techniques to value cashflow streams.

• use the Net Present Value rule to evaluate investment opportunities.

• explain why the NPV rule is optimal.

40
2.8 Exercises
1. An investment advisor tells you that a particular opportunity will
generate (certain) cashflows of $5,000 at the end of each of the next
four years. If the annual rate of interest is 7%, what is the maximum
amount that you should be prepared to pay in order to invest in this
opportunity? Should you invest if the advisor wants to charge you
$17,000 to participate.

2. Which of the following cashflows would you prefer (assuming that


all are certain and you can only choose one of them)?

• A: $26,000 now
• B: $30,000 to be received in 2 years
• C: $39,000 to be received in 5 years

Assume that the annual interest rate is 8%.

3. An oil pipeline will cost £19,000,000 to install. If installation starts to-


day, it will be complete in two years and half of the cost must be paid
at the end of each year of the installation process. Once installed, it
will allow the oil company installing it to save $3,000,000 per year at
the end of each of the first five years of operation, $2,000,000 per year
in the next five years of operation and zero thereafter. The annual in-
terest rate is 5%. Should the company invest in the pipeline? Justify
your answer.

4. Give a TRUE/FALSE answer for each of the following statements


and in each case justify your answer.

(a) If a standard bank account pays interest at annual rate of 5%,


then a deposit of $100 today will grow to be worth $110 in two
years.
(b) One should always choose to receive $110 in one year rather
than $100 today.
(c) If the annual interest rate is 6%, paying £53 in one year is prefer-
able to paying £50 today.

41
Chapter 3

Valuing long-lived cashflow


streams

Essential reading
Brealey, Myers and Allen (2017), Chapter 2 sections 2 and 3.

Learning outcomes
By the end of this chapter, and having completed the Essential reading and
activities, you should be able to:

• use compounding and discounting techniques to value perpetuities


and annuities.

• apply valuation techniques to practical situations.

3.1 Introduction

In the majority of cases, the investments we are interested in valuing gen-


erate multiple payments and/or receipts. For example:

• stocks pay dividends on a regular basis.

42
• bonds make regular interest payments.

• after making a loan to a friend you expect monthly repayments.

We can use PV techniques to arrive at a total value for these cashflow


streams. But in certain cases, we can use some shortcuts to make our com-
putations easier. The two key shortcuts are for:

• perpetuity cashflows (with and without growth).

• annuity cashflows (with and with growth).

In the remainder of this chapter, we will introduce these cashflow streams


and show how to value them in a world of constant, known discount rates.

3.2 Perpetuities

Consider an asset that promises a fixed nominal cashflow at the end of


every period from now until the end of time. This is a perpetuity i.e. a
perpetual cashflow stream.

Call the periodic cashflow C and assume that the one period interest rate
is fixed at r. Then using our standard rules of discounting, the PV of the
perpetuity is:

∞ ∞
C 1
PVP = ∑ i
=C∑ i
i =1 (1 + r ) i =1 (1 + r )

We will show that this infinite sum can be written as:

C
PVP =
r

Thus, if the discount rate is 5% and I promise to pay you $20 per year
forever, with the first payment one year from now, the present value of
this promise is:

43
C 20
PVP = = = $400
r 0.05

3.2.1 PV of a perpetuity: proof

Define x to be (1 + r )−1 . Then the present value of the perpetuity stream


is:

∞ ∞
!
PVP = C ∑ xi = Cx 1 + ∑ xi
i =1 i =1

Multiplying both sides of the original perpetuity value expression by x,


one can also see that:


xPVP = Cx ∑ xi
i =1

Now, combine these two equations to give:

∞ ∞
!
PVP − xPVP = Cx 1 + ∑ xi − Cx ∑ xi = Cx
i =1 i =1

and therefore:

x C
PVP = C =
1−x r

3.2.2 Perpetuities in the real world: undated UK govern-


ment debt

Perpetuities are simple to describe and, as we have shown, simple to value,


but what practical importance do they have? Have such cashflow streams
existed in the real world? The answer to this question turns out to be yes.

44
In late 2014 and early 2015 the UK government repaid undated (perpet-
ual) bonds that had been issued by previous UK governments to finance,
amongst other things, the First World War.

The First World War debt was, in total around £1.9 billion with a 3.5%
annual interest rate. The debt was perpetual (i.e. the holder was promised
interest every year forever) and was originally issued in 1932. Thus if you
held a nominal amount of £1000 of this debt, every year you would receive
an interest payment totalling £35.

The undated debt was replaced with standard fixed lifetime gilts issued at
relatively low interest rates (given the economic conditions in 2015). Plans
were made and executed to take all undated UK bonds and replace them
in the same way. Some of these other undated bonds were first issued in
the 18th century.

3.3 Annuities

These are more common cashflow streams, consisting of the payment of


a fixed nominal cashflow, once per period but only for a known, finite
number of periods. Thus they are like truncated perpetuities.

So, for example, if I borrow money from a mortgage company to buy a


house and in repayment of the loan I promise to pay the mortgage com-
pany £750 a month for the next 25 years, then I have promised the mort-
gage company an annuity.

3.3.1 Valuing annuities

To find the present value of an annuity, we use some simple intuition.


Assume a T year annuity, paying C per year and an annual interest rate of
r. This annuity is equivalent to a perpetuity, with first payment in one year,
and promising a cashflow of C per year, minus a perpetuity promising the
same annual cashflow, C, but with first payment T + 1 years from now.
Thus the value of the annuity must be equal to the difference in the values
of these two perpetuities.

45
The value of a perpetuity, paying C per year with the first payment in one
year is:

C
PV1 =
r

The present value of a perpetuity paying C per year and with a first pay-
ment in T + 1 years is:

1 C
PV2 = T
(1 + r ) r

The T year annuity has value equal to PV1 minus PV2 . So:

 
C 1 C C 1
PVA = PV1 − PV2 = − = 1−
r (1 + r ) T r r (1 + r ) T

3.3.2 Examples: annuity valuation

Scenario 1

A recent retiree wants to use some of her savings to buy an annuity prod-
uct. This product must deliver an annual income of $25,000 per year for
the next 15 years. [Assume that all payments are made at year end]. If the
annual interest rate is currently 3% and we expect this rate to stay constant,
how much should she expect to pay to the annuity provider today?

The annual payment (C) is $25,000, the constant interest rate (r) is 3% and
the term (T) is 15 years. So, using the formula derived above:

   
C 1 25, 000 1
PVA = 1− = 1− = $298, 448.38
r (1 + r ) T 0.03 1.0315

Thus the retiree will have to pay the annuity provider close to $300,000
today in exchange for annual payments of $25,000 for the next 15 years.

46
Scenario 2

Assume a world where interest rates are expected to be constant at 3% per


annum for the foreseeable future. You wish to buy a flat and a mortgage
company agrees to lend you £950,000 today. The mortgage has a term of
25 years. You promise the mortgage company a fixed repayment at the
end of each of the next 25 years. What is the fair repayment?

Well the repayment you promise is an annuity stream and for it to be fair
it must be worth £950,000 in PV terms. So we start with

 
C 1
PVA = 1−
r (1 + r ) T

and this yields:

 
C 1
950, 000 = 1−
0.03 (1.03)25

Solving this, we see that the fixed annual repayment, C is:

950, 000 × 0.03


C= h i = 54, 566.48
1
1 − (1.03)25

So, in return for lending you £950,000 today, you must pay the mortgage
provider roughly £54,566 at the end of each of the next 25 years.

Here is a follow up question. Just after you have made your tenth mort-
gage payment, what is the value of your outstanding debt to the mortgage
company?

After the tenth payment, you owe the mortgage provider fifteen more pay-
ments of £54,556.48 with the first payment one year away. The total PV of
these payments, and thus the size of your outstanding mortgage debt is:

47
 
54, 556.48 1
PVA = 1− = 651, 291.72
0.03 (1.03)15

Thus in the first 10 years of payments, you pay off roughly £300,000 of
your £950,000 debt.

3.4 Perpetuities with growth

The analysis contained in this chapter thus far looks at long-lived cashflow
streams where the periodic payment is fixed at C per period. We now want
to modify the cashflow stream to allow it to grow at a constant (expected)
rate every period. Why might we want to do this? Perhaps an investment
delivers cashflows that grow with the inflation rate and perhaps we expect
future inflation to be constant at, say, 2% per annum.

We start by allowing the cashflows of our perpetuity to grow at a constant


rate of g per period. We assume that at the end of the first period one is
promised a cashflow of C and at the end of each subsequent period, the
cashflow promised grows at rate g. Thus the second cashflow is C (1 + g),
the third is C (1 + g)2 and so on.

Again, if we assume that the constant one-period interest rate is r, then the
PV of the perpetuity with growth is:


C C (1 + g ) C (1 + g )2 C (1 + g ) i −1 C
PVP =
(1 + r )
+
(1 + r )2
+
(1 + r )3
+ ... = ∑ (1 + r ) i
=
r−g
i =1

We’ll prove this final equality now, using exactly the same trick that we
used to find the PV of the perpetuity stream without growth. Start with
our infinite sum again:

∞ ∞ ∞
" i # " #
C (1 + g ) i −1 1+g

C C
PVP = ∑ (1 + r ) i
=
1+r
1+ ∑
1+r
=
1+r
1 + ∑ xi
i =1 i =1 i =1

48
where x = (1 + g)/(1 + r ). This implies that:


C ∞ i
" #
C
x+∑x =
1 + r i∑
i
xPVP = x
1+r i =2 =1

Subtracting the result of this second equation from the first we get:


C ∞ i
" #
C C
PVP − xPVP = 1 + ∑ xi − ∑ x =
1+r i =1
1 + r i =1 (1 + r )

And finally:

1 C (1 + r ) C C
PVP = = =
1 − x (1 + r ) (r − g ) (1 + r ) (r − g )

Before we use this result, it’s worth thinking about it a little. First of all,
it is clear that, to make the answer sensible, we need r to be greater than
g. If this does not hold, then it seems that the answer is negative (and that
cannot be right)! So we can only use this equation when that restriction
(i.e. r > g) holds.1

3.4.1 Example

A financial advisor offers you a product that will pay you $1,000 in one
year, with this annual cashflow growing at a rate of 2% forever. If the
current annual interest rate is 4%, how much should you be willing to pay
to buy this product?

Using the formula above, we have:


1 When r < g the problem that one faces is that the cashflows grow so fast that the
discounted cashflows also grow every year. This leads to the present value of the entire
stream being infinite. So, for example, if your cashflows are growing at a rate of 5%
a year but the discount rate is only 3% per year, then the perpetuity will be worth an
infinite amount.

49
C 1, 000 1, 000
PVP = = = = $50, 000
(r − g ) 0.04 − 0.02 0.02

3.5 Annuities with growth

Finally, we can modify our annuity cashflow stream to allow growth in


the periodic payment.

Consider a cashflow stream such that at the end of one period one is
promised a cashflow of C and at the end of each subsequent period, the
cashflow promised grows at rate g. However, the cashflow stream termi-
nates after T payments. This is an annuity with growth.

To value this stream we use exactly the same trick as when valuing an
annuity without growth:

1. Value a perpetuity with growth where the first payment is made one
period from now.

2. Value a perpetuity with growth where the first payment is made T +


1 periods from now.

3. The cashflows from the annuity with growth equal the cashflows
from (1) minus the cashflows from (2).

4. Thus the PV of the annuity with growth is the PV of (1) minus the
PV of (2).

If the first cashflow is C, the cashflows grow at rate g and the one-period
discount rate is r then the PV of the growing annuity is:

T "  #
1+g 1+g T
 
C C C
PVA = − = 1−
(r − g ) 1+r (r − g ) (r − g ) 1+r

The proof of the final result follows exactly the same steps that we have
seen previously and hence it is omitted.

50
3.5.1 Example

The same financial advisor as before offers you a different product. The
product will pay you $1,000 in one year, with this annual cashflow grow-
ing at a rate of 2%, but the payments only last for 20 years. How much
should you pay for this product if the annual interest rate is 4%?

Using our result, we have:

"  # "  #
1+g T 1.02 20
 
C 1, 000
PVP = 1− = 1− = $16, 091.65
(r − g ) 1+r 0.04 − 0.02 1.04

51
Reminder of learning outcomes
Having completed this chapter, and the Essential reading and activities,
you should be able to:

• use compounding and discounting techniques to value perpetuities


and annuities.

• apply valuation techniques to practical situations.

52
3.6 Exercises
1. An investment advisor offers you a product that will deliver cash-
flows at the end of each of the next seven years. The first cashflow is
£5,000 and every year thereafter the cashflow will grow at a rate of
4%. If the annual interest rate is 2%, what is the present value of this
opportunity?

2. At the end of each of the next eight years, you plan to put £25,000 of
your annual salary in the bank. If the annual interest rate is 3%, what
is the present value of this planned savings stream? What will the
balance in your bank account be at the end of the eight year period?
[HINT: once you know the present value of the savings stream, it’s
easy to work out the future value.]

3. You need to borrow £800,000 to buy a house. You borrow the money
from a bank via a mortgage with a 25-year term. The mortgage re-
quires you to make monthly repayments, with the first payment one
month from now. If the monthly interest rate is 0.5%, work out the
fair monthly repayment that you will have to make.

4. Give a TRUE/FALSE answer for each of the following statements


and in each case justify your answer.

(a) A perpetuity paying $2,000 per year is currently valued at $80,000.


This implies that the one year interest rate is 4%.
(b) A perpetuity pays its initial cashflow of C today and then pays
a cashflow of C at the end of every year from now onwards. If
the annual interest rate is r, the present value of this perpetuity
is C/r.
(c) The annual interest rate is 5%. A perpetuity with growth has an
initial cashflow of £400, to be paid in one year. Its present value
is £10,000. This must mean that the growth rate of the cashflows
is 1%.

53
Chapter 4

Interest rates: quoted and


effective, real and nominal

Essential reading
Brealey, Myers and Allen (2017), Chapter 2 section 4.

Learning outcomes
By the end of this chapter, and having completed the Essential reading and
activities, you should be able to:

• explain the difference between quoted and effective interest rates.


• convert quoted to effective rates and use effective rates in discount-
ing problems.

4.1 Introduction

Before getting any deeper into asset pricing we need to talk more clearly
about interest rates. We need to describe how interest rates are quoted in
real financial markets and how these quoted interest rates translate into
actual interest payments or receipts. Finally, we will differentiate between
nominal and real interest rates.

54
4.2 Quoted rates and APRs

By convention, interest rates quoted in financial markets tend to be an-


nualised. This means that the rate you see quoted has been scaled up or
down so that it covers a period of exactly one year. Unfortunately, though,
the way in which this is done varies across markets and currencies. An-
other unfortunate side effect of this is that the conventions used mean that
the actual interest earned on a deposit over a year, for example, may not
equate to the quoted annual rate.

In this book we will adopt the convention below (which is also fairly com-
mon in real markets). Interest rates will always be given on an annualized
basis and the following rule will apply.

Quoted annual interest rates

If the K month interest rate is quoted at rate r

• The actual rate you will pay/receive on a K month loan/deposit is


K
12 r

The quoted rate, r, is also called the Annual Percentage Rate (APR). The
actual rate that you receive/pay on the K month deposit is also called a
K month effective rate.

So, for example, if someone quotes you an interest rate of 6% on a one-


month deposit, this does not mean that a deposit of $100 will be worth
$106 one month later. The actual rate that you will receive for the month
1
is 12 × 0.06 = 0.005 = 0.5%. Thus a deposit of $100 will grow to $100.50
after one month.

Here’s a second example. Your bank charges quarterly interest on loans


with an APR of 10%. Again, this is an annualized rate. The actual rate
3
that you would pay every quarter on a loan from the bank is 12 × 0.10 =
3
0.025 = 2.5%, with the 12 factor coming from the fact that this is a quarterly
rate. So if I was to borrow £500 from the bank for three months, at the end
of the quarter I would need to repay £512.50. If I was to borrow £100 for
six months, then at the end of the period I would have to repay £105.0625

55
(if we allow for fractional pennies).

4.3 Effective interest rates

The effective annual interest rate is the actual interest you would receive on
a deposit (or pay on a loan) on an annual basis, expressed as a percentage.

Example 1. Take the situation above where the quoted rate on a one-month
deposit is 6%. You make a deposit of $100 and at the end of every month
re-invest the balance at the one-month rate. At the end of 12 months you
have:

100 × (1 + r )12 = 1 × (1 + 0.005)12 = $106.17

So your effective annual interest rate is 6.17%.

Note that the effective annual rate (6.17%) is above the quoted rate (6%),
as late in the year you earn interest on interest paid earlier in the year.

The general lesson here is that the APR (or quoted annualized interest rate)
is not equal to the actual amount of interest you would earn on a deposit
over the course of a year (or pay on a loan over the course of a year). The
number that we, as economists, care about, is the effective rate. This is the
true rate that you earn on your bank account (over a period of a year). The
APR is an artificial, rescaled number that does not necessarily correspond
to the annual interest you earn on a deposit or pay on a loan.

This leads to a piece of advice. When given some interest rate data, the
first thing you should do is convert any APRs into effective annual rates.1
1 Of
course, if the interest rate you are given is for a 12 month deposit or loan then the
APR and effective annual rate coincide, so there’s nothing to worry about.

56
4.3.1 Non-annual effective rates

What about if you want to know the effective rate on a period not exactly
equal to 1 year. Say you want the effective M month interest rate. How do
you compute this?

If we denote the effective annual rate by EAR, then the calculation of the
effective K month rate is:

K
(1 + EAR) 12 − 1

Let’s illustrate using the data from Example 1 above. There we calculated
the EAR as 6.17%. What is the effective six-month rate? It is:

6
(1 + 0.0617) 12 − 1 = 0.0304 = 3.04%

Thus, the present value of £1,000 to be received in six months is:

1, 000
= 970.52
1 + 0.0304

Using the same data, what is the effective 18-month rate? It is:

18
(1 + 0.0617) 12 − 1 = 0.0939 = 9.39%

So the present value of £15,000 to be received in 18 months is:

15, 000
= 13, 712.04
1 + 0.0939

Finally, what’s the effective one-month rate. It is:

1
(1 + 0.0617) 12 − 1 = 0.005 = 0.5%

We already knew the answer to this. The effective one-month rate is 0.5%.

57
4.4 Common terminology

Later in this subject guide, when I quote interest rates to you, they will
always be annualised rates. Here’s how to understand some terms that I
will use.

• The quarterly compounded interest rate is 5%:


– This is an annualised rate (as always)
– The effective rate that would apply to a three-month deposit or
loan is 1.25%
– Thus the annual effective rate is 5.09%
• The stated interest rate with monthly compounding is 9%:
– This is an annualised rate (as always)
– The effective rate that would apply to a one-month deposit or
loan is 0.75%
– Thus the annual effective rate is 9.38%
• The semi-annually compounded interest rate is 7%:
– This is an annualised rate (as always)
– The effective rate that would apply to a six-month deposit or
loan is 3.5%
– Thus the annual effective rate is 7.12%

4.5 Examples: APRs and discounting

Here are a few examples which illustrate how to use quoted rates, or APRs,
properly when computing present values.

Example 1: a simple annuity

An investment opportunity offers you quarterly cashflows of £1000 for the


next five years. The three-month interest rate is quoted at 4%. What is the
maximum price you would pay to participate in this investment?

58
Solution: this is a non-growing annuity. We know that the PV of such a
cashflow stream is:

 
C 1
PVA = 1−
r (1 + r ) T

where C is the periodic cashflow, r is the one period interest rate and T is
the number of periods.

First of all, as the cashflows are quarterly, let’s define one period to be
one quarter. Then every period one receives £1000. What is the effective
one period interest rate? This is the effective quarterly rate and, using our
convention, the effective quarterly interest rate is equal to the quoted rate
3
multiplied by 12 , i.e. 1%. Finally, the number of periods the investment
covers is the number of quarters in five years, i.e. 20. So:

 
1 1
PVA = 1000 − = 18045.53
0.01 0.01(1.01)20

Example: a car loan

You need £25,000 to buy a car today. You take out a five-year loan to raise
the money. The loan requires you to make constant monthly payments
with the first payment one month from today. If the interest rate with
monthly compounding is 3% and we expect this rate not to move in the
near future, what is the monthly payment you must make?

Solution. This is a standard annuity. We know that the PV of such a


cashflow stream is:

 
C 1
PVA = 1−
r (1 + r ) T

In our case the PV of the repayments must be equal to £25,000 and the
effective one-month interest rate is 0.03/12 = 0.25%. There are 60 months
in the payment period so:

59
 
C 1
25, 000 = 1−
0.0025 (1.0025)60

Solving, we find that our monthly repayment is £449.22.

Example: computing PVs when payment frequencies and compounding


frequencies differ

You owe a supplier for materials. The payment schedule requires you to
deliver £1000 to the supplier at the end of each of the next nine months.
What is the PV of this cashflow stream if the quarterly compounded inter-
est rate is 7.5%?

Solution. First, define a period to be one month (as this is the frequency
of the cashflows). Next, we need the effective annual and then monthly
rates. The effective annual rate is:

 4
0.075
rA = 1+ − 1 = 0.0771 = 7.71%
4

The effective monthly rate is:

1 1
r M = (1 + r A ) 12 − 1 = 1.0771 12 − 1 = 0.0062 = 0.62%

Finally, compute the PV of the nine-period (i.e. nine-month) annuity:

 
1000 1
PVA = 1− = £8, 726.74
0.0062 (1.0062)9

4.6 Continuous compounding

In certain financial applications we will make the (rather unrealistic) as-


sumption that interest is not paid monthly or annually on a bank account,

60
but instead it is paid continuously. Usually this assumption is made be-
cause it makes solving a particular valuation problem simpler (although it
might introduce some approximation error).

To see how this works, fix a deposit size of X and an APR on a bank ac-
count of r. If there are m periods per year then, at the end of the year, the
balance on one’s account is:

 r m
X 1+
m

As the number of periods per year tends to infinity we get:

 r m
lim X 1 + = Xer
m→∞ m

We call this continuous compounding and in this situation r would be de-


scribed as the continuously compounded interest rate.

Thus, an investment for T years at continuously compounded rate r would


grow to XerT by the end of the holding period.

Obviously, if we can compound continuously, we can also discount con-


tinuously. Assume that you will receive a cashflow of Z in T years and you
want to work out the present value of this cashflow. If the continuously-
compounded interest rate is r then the present value is:

PVZ = Ze−rT

4.7 Real and nominal interest rates

Finally, for this discussion of interest rates, we need to discriminate be-


tween the interest rate that one earns in simple currency terms and the
interest rate that one earns in terms of purchasing power. To do this, de-
fine the following three rates:

61
• Inflation rate: the (usually annual) rate at which the level of prices
in the economy grows. Denote it by π.

• Nominal interest rate: the rate at which the balance of a deposit


grows in cash terms. Denote it by r.

• Real interest rate: the rate at which the balance of a deposit grows
in purchasing power terms. Denote it by i.

The inflation rate should be familiar to all. The nominal interest rate is the
interest rate that we have been studying throughout this chapter up to this
point. It is the real interest rate that is new. The real rate is the return that
your bank account (or other investment) delivers after accounting for the
erosion in the value of your deposit due to inflation.

Real interest rates

We define the real interest rate (i) as follows:

(1 + r )
(1 + i ) =
(1 + π )
Thus the real rate is increasing in the nominal interest rate and decreas-
ing in the inflation rate. If the nominal interest rate and the inflation rate
are identical, then the real rate is zero (i.e. if the interest rate your bank
account delivers just matches inflation, then in real terms you’re earning
exactly zero).

We can come up with a simple approximate real interest rate as follows.


By definition:

(1 + r ) = (1 + i ) × (1 + π ) ⇒ (1 + r ) = 1 + i + π + iπ

When rates and inflation are low, iπ will be very small. Thus we ignore it
and obtain:

(1 + r ) ≈ 1 + i + π ⇒ i ≈ r−π

62
Note that this approximation is good when the rates involved are small,
but when inflation or the nominal rate become large then the approxima-
tion becomes worse.2

So, a few comments to sum up:

• If nominal interest rates are above the inflation rate, then a deposit
will grow in real terms and debts will grow in real terms also.

• If the nominal interest rate is below the inflation rate, then a deposit
will shrink in purchasing power terms. On the flipside, a debt will
shrink in real terms if r < π.

• Inflation rates can be substantial, implying that the difference be-


tween real and nominal rates can be large.

– Assume nominal rates are 18% and inflation is running at 12%.


– The approximate real rate is 6%.
– The precise real rate is 5.36%.

2 Itis easy to see why this is true. In our approximation we have ignored the term iπ.
Thus the approximation will be poor when this term is large, which happens when at
least one of the components of the product becomes large.

63
Reminder of learning outcomes
Having completed this chapter, and the Essential reading and activities,
you should be able to:

• explain the difference between quoted and effective interest rates.

• convert quoted to effective rates and use effective rates in discount-


ing problems.

64
4.8 Exercises
1. The quarterly compounded interest rate is 8%. What is the present
value of $1,000 to be received in nine months? What is the future
value of $5,000 if it is deposited in a bank account today and remains
there for 15 months?

2. An investment product promises semi-annual cashflows of $3,500


for the next five years. If the stated interest rate with monthly com-
pounding is equal to 10%, what is the present value of the cashflows
that the investment promises?

3. You take out a loan today to pay for refurbishments to your house.
The loan requires you to make monthly repayments over the next
10 years. The cost of the refurbishments is £20,000. If the effective
annual interest rate is 8%, what will your monthly loan repayment
be?

4. Give a TRUE/FALSE answer for each of the following statements


and in each case justify your answer.

(a) Assuming that interest rates are positive, the effective annual
interest rate will always be larger than the quoted six-month
APR.
(b) You should always use effective rates to perform compounding
and discounting operations, and never use APRs.
(c) The real rate of interest is exactly equal to the difference be-
tween the nominal interest rate and the inflation rate.

65
Chapter 5

Government bonds: features and


valuation

Essential reading
Brealey, Myers and Allen (2017), Chapter 3.

Learning outcomes
By the end of this chapter, and having completed the Essential reading and
activities, you should be able to:

• describe the cashflow patterns of various types of bond.


• use NPV to price bonds.
• compute the yield to maturity on a bond.
• compute a bond’s duration and use it to assess that bond’s interest
rate risk.

5.1 Introduction

In this chapter we will familiarise ourselves with the structure and pric-
ing of government bonds. We restrict ourselves to government bonds as

66
this means we have to worry less about issues of default by the issuers
of bonds (we think that governments are less likely to go bankrupt than
companies), but all of the tools we meet here can be extended to the case
of corporate bonds.

We will cover how bonds are structured and how we can use present value
methods to value them. We start by assuming that interest rates are con-
stant across time but then relax this assumption to allow interest rates to
vary over time and over deposit/loan periods. We describe the variation
of interest rates with maturities of loans/deposits as the term structure of
interest rates and we will talk a little about how the term structure is deter-
mined. We will also talk a little about risk management of bond portfolios
using duration measures of interest rate risk.

5.1.1 Bonds: basics

Bonds are often called fixed-income securities. They are so named as the
cashflows they deliver to an investor, as well as the dates that these cash-
flows will arrive, tend to be known (fixed) in advance.

A bond is a security issued by a borrower (i.e. the issuer) and purchased


by an investor. Upon issue, the investor is obliged to pay the issuer the
bond price. Subsequently, the bond contract obligates the issuer to make
pre-specified payments to the bondholder at pre-specified future dates.

Essentially, bonds are borrowing/lending arrangements formalised in con-


tractual form. The bond issuer is borrowing money from the investor (in
the form of the bond price). In return, the issuer promises to pay the in-
vestor regular cashflows over some pre-specified period and this can be
thought of as gradual repayment of the loan plus interest.

5.1.2 Are bond markets that important?

In general, we hear much less about bond markets than we do equity (or
housing) markets in the financial media. Is this because bond markets are
unimportant or small? Clearly, as the table below shows, they are very
large.

67
Table 5.1: US bond markets: outstanding debt (SIFMA, Q1 2017)

Segment Issuer Debt ($ trillions)


Municipal Local Gov. 3.8
Treasury Treasury 14.0
Mortgage-related Private 9.0
Corporate Corporations 8.6
Federal agency Federal agencies 2.0
Money market Corporations, Financial inst. 0.9
Asset backed Financial institutions 1.4
Total 39.7

The different rows of the table above refer to different segments of the
bond market. For example, the second row (labelled ‘Treasury’) describes
the size of the US government bond market while the row labelled ’Corpo-
rate’ gives the scale of bond issues by US companies. Aggregating across
all segments we get a total size close to $40tn.

How about bond markets relative to equity markets? We use the Russell
3000 index of US stocks to measure US stock market capitalization, as the
Russell 3000 is over 98% of the US stock universe in market cap terms. In
February 2017 the Russell 3000 companies were worth $25.6 trillion

Thus, obviously, bond markets are around 50% larger in size than equity
markets.

5.1.3 The zero-coupon bond

The zero-coupon bond is the building block from which other bond struc-
tures are built. It has the following basic structure:

68
A K-year zero-coupon bond

An investor purchasing a K year zero-coupon bond (abbreviated as ’a


zero’) from the issuer, pays the issue price, P.

In return, the K-year zero promises the investor a single payment, called
the face value or par value, K years from the issue date. The date of the
payment is called the maturity date and K is the time to maturity.

Example: a five-year zero has a face value of £10,000 and is priced at £9750.
If one was to purchase this bond then:

• One would pay £9750 to buy the bond today

• One would be promised a cash inflow of £10,000, 5 years from today.

5.1.4 Coupon bonds

Coupon bonds deliver more complicated cashflow streams to investors

69
A K-year coupon bond

An investor purchasing a K year coupon bond from the issuer, pays the
issue price, P.

Subsequently, the investor in the K-year coupon bond receives the fol-
lowing cashflows:

• On the maturity date, the bondholder receives the face value.

• At regular intervals up to and including the maturity date the


bondholder receives a coupon payment

– These payments could be annual, semi-annual or quarterly.


– These coupon payments are usually the same at every pay-
ment date.
– The ratio of the total annual coupon payment to the face value
is called the coupon rate.

Thus a coupon bond promises the purchaser a stream of cashflows over


some specified period.

5.1.5 Some bond examples

Consider three bonds, All have face value £100 and two years to maturity:

• Bond A is a zero-coupon bond

• Bond B has annual coupon payments and coupon rate 6%

• Bond C is a bond with semi-annual coupon payments and coupon


rate 5%

The following pictures show the cashflows from each of these bonds.

70
A: Zero-coupon bond:

Cashflow 100
Year 0.5 1.0 1.5 2.0

B: Coupon bond with annual coupons:

Cashflow 6 106
Year 0.5 1.0 1.5 2.0

C: Coupon bond with semi-annual coupons:

Cashflow 2.5 2.5 2.5 102.5


Year 0.5 1.0 1.5 2.0

Thus, as described above, the zero-coupon bond delivers the purchaser a


single cashflow (the face value) on the maturity date. The coupon bonds,
in addition, deliver regular, fixed payments up to and including the matu-
rity date.

5.1.6 A real world example: UK gilts

Consider the 1.5% Treasury Gilt 2047 issued by the UK government in


July 2017. This is a bond issued by the government to fund its activities. It
has the following features:

71
• Every year until maturity, an individual receives an annual coupon
of £1.50 for every £100 face value of the bond that he/she holds.
• Coupon payments are semi-annual, so the £1.50 total annual coupon
is paid in 2 instalments:
– £0.75 on January 22nd every year
– £0.75 on July 22nd every year.
• The bond matures on July 22nd 2047, and on that date a holder of
£100 of face value of the bond issue will receive his/her £100 face
value plus the £0.75 coupon due on that date.

It’s worth asking at this juncture how issuers set coupon rates. They are
usually set such that, when the bond is issued the price investors are will-
ing to pay for the bond is roughly equal to its face value. As we will see
later, this is the same as saying that the coupon rate will be similar to the
current level of market interest rates.

5.2 Valuing bonds when interest rates are con-


stant

If you were an investor considering purchasing one (or all) of the three
bonds above, how much would you be willing to pay? Let’s start in a
simple world where the interest rate is constant over time and across ma-
turities. Call this interest rate r.

Then, take a generic coupon bond that promises you a coupon rate of c
with coupons paid every period for the next K periods and, additionally,
in period K also delivers its face value of F. What’s the right price for this
bond, P? Simple application of discounting tells us that:

cF cF cF cF + F
P= + + + . . .
(1 + r ) (1 + r )2 (1 + r )3 (1 + r ) K

We can use our prior knowledge to simplify this, however. Forget the
face value for now and focus only on the stream of coupon payments.
These are a constant payment of cF every period for the next K periods.

72
This cashflow stream is clearly an annuity. Thus we can use our annuity
valuation formula to rewrite the bond price as:

 
cF 1 F
P= 1− +
r (1 + r ) K (1 + r ) K

This formula tells us that bond prices have three key determinants:

• Face value: the higher this is, the higher is the price, other things
being equal. Higher face value increases all cashflows to the investor.
• Coupon rate: the higher this is, the higher is the price (other things
being equal) because it leads to higher coupon payments.
• Interest rates: the greater is the interest rate, the lower is the price
(holding cashflows constant). Higher interest rates mean heavier dis-
counting which means lower present values.

OK, so that was deceptively simple. Here is a quick example.

Example: assume that the (effective) annual interest rate is 4%. What is
the price of Bond B above?

• Well we know that the cashflows are £6 in one year and £106 in two
years. So the price must be:

6 106
PB = + = 103.77
(1 + 0.04) (1 + 0.04)2
• Thus you should be prepared to pay £103.77 for this bond.

Second example: again assuming that the (effective) annual interest rate
is 4%, what is the price of Bond C above?

• First, because the bond cashflows come every six months, we need
the effective six-month interest rate. This is:

r = (1 + 0.04)1/2 − 1 = 0.0198 = 1.98%

73
• Next, we know that the cashflows are £2.5 in 6, 12 and 18 months
respectively and £102.25 in 24 years. So the price must be:

2.5 2.5 2.5 102.5


PC = + + + = 101.98
(1 + 0.0198) (1 + 0.0198)2 (1 + 0.0198)3 (1 + 0.0198)4

• Thus you should be prepared to pay £101.98 for bond C.

This pricing analysis and these examples are simple, but they illustrate the
basic point. The value/price of a bond is nothing other than the present
value of its cashflows. Now we make this analysis a little more realistic by
allowing interest rates to vary with across maturities.

5.3 Valuing government bonds and the term struc-


ture of interest rates

The fundamental innovation in this section is that we allow cashflows to


be received at different future dates to be discounted using different dis-
count rates. We call the interest rate appropriate for discounting a cash-
flow to be received (or paid) K periods from now, the K period spot rate.

Example: call the two-year spot rate r2 and the three-year spot rate r3 .
Then:

1
• The PV of $1 to be received two years from now is (1+ r2 )2
.

1
• The PV of $1 to be received in three years is (1+r3 )3
.

So the interest rate we use to discount a cashflow depends on when that


cashflow arrives

• If you’re pricing a stream of cashflows, each cashflow has its own


relevant interest rate depending on when it is received.

74
• The PV of the entire stream is the sum of the present values of each
cashflow, where each one is discounted at the appropriate interest
rate.

In Figure 5.1 you can see a plot of the spot rates that were observed in
the UK government bond market on the 31st of May 2016 (the red curve)
and on the 30th of June 2016 (the blue curve). On the x-axis of the plot
you have period of time that the interest rate refers to and on the y-axis
the spot rate. Starting with the red curve, one can see that (in the main)
is it upward sloping. Thus, the interest rate appropriate for discounting
a cashflow to be received 15 years from 31st May 2016 is greater than the
appropriate rate for a cashflow to be received five years after May 31st.
To be more precise, the five-year spot rate is around 1% while the 15-year
spot rate is roughly 2%.

Additionally, we can see from Figure 5.1 that the entire term structure
moves over time. In the month between the end of May and the end of
June 2016, the entire UK term structure shifts downwards (from the red to
the blue curve). At the end of June, the 5 year spot rate has fallen to about
0.4% and the 15 year spot rate is around 1.5%.1

5.3.1 Term structure notation

Consider a basic time unit of 1 year. Cashflows received 1 year from today
have a different spot rate than cashflows received two or three years from
now. We use the following notation to distinguish between spot rates.

Years Rate Discount factor


1 r1 (1 + r 1 ) −1
2 r2 (1 + r 2 ) −2
3 r3 (1 + r 3 ) −3
....
N rN (1 + r N ) − N
1 The big UK political/economic event of June 2016 was the UK referendum on leaving

the EU. The reaction of the term structure, i.e. lower spot rates at all maturities, suggests
that markets believed that this was bad news for the UK economic outlook (as lower rates
are usually a response to weaker economic growth.

75
Figure 5.1: UK government term structure: 31st May 2016 and 30th June
2016

2.5
End May
End June

2
Yield, percentage points

1.5

0.5

0
0 5 10 15 20
Years

In a final piece of notation, the period that a particular spot rate is relevant
for is usually called the tenor of the rate.

5.3.2 Pricing bonds from the term structure

Now we can return to pricing bonds. Consider again the generic bond that
we priced earlier assuming constant interest rates. This bond promises a
coupon rate of c with coupons paid every year for the next K years and,
additionally, in year K also delivers its face value of F. Again we want
to compute the right price for this bond, but now we work under the as-
sumption that spot rates can vary with tenor.

Computing the present value of cashflows we get:

cF cF cF cF + F
P= + 2
+ 3
+...
(1 + r1 ) (1 + r2 ) (1 + r3 ) (1 + r K ) K

Again, bond prices depend positively on coupon rates and face values.
Now, the bond price depends on the entire term structure from 1 to K
years. A rise in any one of the relevant spot rates will reduce the price of

76
the bond (other things equal).

Also, note that we can no longer use the annuity formula to simplify the
pricing formula, as that formula relies on interest rates being constant
across tenors.

Thus the fact that spot rates vary across tenors leads to slightly more com-
plicated bond pricing. We will defer the question of why spot rates vary
across tenors to the last part of this chapter.

Example: pricing a three-year bond

Consider a bond with exactly three years to maturity. Face value is $100
and the coupon rate is 5% (coupons are paid annually). The one, two and
three year spot rates are 4%, 4.5% and 6% respectively. What is the bond
price?

To price the bond, we just perform the following PV computation:

5 5 105
P = + 2
+
(1 + r1 ) (1 + r2 ) (1 + r3 )3
5 5 105
= + 2
+
1.04 1.045 1.063
= 97.55

Thus the right price for the bond is $97.55.

5.4 Bonds: yield to maturity

It is much more common to talk about bonds using their yield to maturity
(YTM) or redemption yield rather than their price.

The YTM is the constant, hypothetical annual discount rate that, when
used to compute the PV of a bond’s cashflows, gives you the bond’s mar-
ket price as the answer. So the YTM is just a transform of the bond price.
The reason we talk about YTMs rather than prices is that prices can be

77
very different across bonds due to, for example, differing coupon rates,
but yields are annual discount rates and thus much more easily compared
across bonds.

Given the definition of the YTM, it should be clear that a higher bond price
must mean a lower YTM and vice versa.

5.4.1 YTM: mathematically

Consider our generic bond from earlier. This bond promises a coupon rate
of c with coupons paid every year for the next K years and, additionally,
in year K also delivers its face value of F. We know that the price of the
bond is:

cF cF cF cF + F
P= + + + . . . (5.1)
(1 + r1 ) (1 + r2 )2 (1 + r3 )3 (1 + r K ) K

Now, let’s take this price as given and ask what constant, maturity-independent
discount rate, y, equates the price with the discounted sum of the bond’s
cashflows. It must be the case that y solves:

cF cF cF cF + F
P= + 2
+ 3
+... (5.2)
(1 + y ) (1 + y ) (1 + y ) (1 + y ) K

So y, which is the bond’s yield to maturity, is the solution to this Kth order
polynomial. Solving this equation may seem daunting (and solving it by
hand is definitely daunting) but it is easily solved by a computer using,
for example, Excel. Equation (5.2) makes the relationship between prices
and YTMs clear – higher price means a lower YTM and lower price means
a higher YTM.

A final point to note from the definition above. Let’s assume that the term
structure of interest rates is completely flat, such that all spot rates are the
same. Call this spot rate for all tenors r. In this case, the yield to maturity
of the bond must simply be r. This can be seen by comparing equations
(5.1) and (5.2).

78
YTM example 1

Consider the three year coupon bond we priced above. Its face value is
$100 and the coupon rate is 5% (coupons are paid annually). The one, two
and three year spot rates are 4%, 4.5% and 6%. We showed that the bond
price is:

5 5 105
P= + 2
+ = 97.55
1.04 1.045 1.063

What is the YTM for this bond? It is the value of y that solves this equation:

5 5 105
97.55 = + 2
+
(1 + y ) (1 + y ) (1 + y )3

I have solved this equation for y using Excel’s Solver feature. The value
that one gets for the YTM is 5.915%. You can obviously check that this
solution is correct by substituting it into the right-hand side of the equation
above and verifying that it equates to the left-hand side.

YTM example 2

Assume that today is January 2018. What is the market price of the fol-
lowing bond? A Japanese Government bond pays a 1.50% annual coupon,
every year for six years. Its YTM is 2.8%. Assume a face value of 10,000
Yen.

Here we are going to go in the opposite direction to example 1. In example


1 we worked out the bond’s yield using its price and its cashflows. As
equation (5.2) makes clear, though, if you’re given a bond’s cashflows and
its yield, you can just as easily work out its price.

First, the cashflows delivered by the bond are:

Date Jan 19 Jan 20 Jan 21 Jan 22 Jan 23 Jan 24


Cashflow 150 150 150 150 150 10,150

79
We can thus compute the price as:

150 150 150 150 150 10, 150


P= + 2
+ 3
+ + 5
+
(1 + y ) (1 + y ) (1 + y ) (1 + y ) 4 (1 + y ) (1 + y )6

We are told that y = 0.028 and so we get P = 9291.07.

5.4.2 Yields, prices and coupon rates

Coupon rates are fixed at the issue date of a bond, while prices, and thus
YTMs, can vary through the bond’s lifetime. How do yields, prices and
coupon rates interact?

We’ve already talked about the relationship between yields and prices.
Let’s assume that just after I’ve bought a new bond, the term structure
of interest rates shifts upwards. This will cause the bond price to fall as
the price is just the PV of the bond’s cashflows and the rise in all spot
rates means that the PVs all fall. As the bond price falls, the YTM of the
bond will rise. In the converse case, i.e. if the term structure was to shift
downwards, we would instead see bond prices rising and yields falling.

How does the coupon rate affects prices and yields? It is relatively easy to
show that the following result holds:

• If the coupon rate on a bond is greater than the YTM, then the price
of the bond will always be above face value (or par).
• If a bond’s coupon rate is below the YTM, then the bond’s price will
always be below face value (below par).

So, for example, in a world of positive interest rates, zero-coupon bonds


must always be priced below par. Also, if you know three of the yield,
price, coupon rate and face value, you can use the results above to tell you
something about the fourth (e.g. if you tell me that a bond with face value
100 has a price of 102 and a yield of 4%, then I can tell you that the coupon
rate must be above 4%).

I will ask you to prove this result in the end of chapter problems.

80
5.4.3 Semi-annual coupons and yields

Finally, let us deal with the case of working out bond prices and YTMs
when bond coupons are paid semi-annually rather than annually.

Assume that you have a bond with face value F and annual coupon rate c.
You are also told that the YTM on the bond is y. Coupons are paid every
six months. In this case:

cF
• The actual coupon paid every six months is 2

• If you wanted to compute the bond price, the actual yield you would
use in your semi-annual discounting is y/2 (following the conven-
tion we described in Chapter 4).

So, assume for simplicity, that the bond has two years to maturity. Then
the price of the bond must be:

cF/2 cF/2 cF/2 cF/2 + F


P= + 2
+ 3
+
(1 + y/2) (1 + y/2) (1 + y/2) (1 + y/2)4

Using this formula, you should be able to show that if, for example, you
have a two-year bond with a 5% coupon rate, semi-annual coupon pay-
ments, face value $100 and a yield of 8%, then the price is $94.56.

5.5 Duration and dependence of bond prices on


interest rates

Assume that you are an investor who has purchased a portfolio of govern-
ment bonds. The key risk that you face is the risk that the term structure of
interest rates shifts. If this happens, the prices of your bonds will change
(as the present values of the cashflows change). If the term structure shifts
up, your bonds will drop in value and if it shifts down they will rise in
value.

In this section, we will compute a measure for the amount of interest rate
risk that a bond is exposed to. This measure is called Macaulay duration.

81
Bond Prices and Yields
Figure 5.2: Bond prices and yields

1600

1400

1200

1000
Price

800

600
5 Year 9% Bond
400

200 1 Year 9% Bond

0
0 2 4 6 8 10 12 14
Yield

FM212
Principles of Corporate Finance

It can be used to tell you (approximately) how much a bond’s price will
change as that bond’s yield changes. From it we will be able to see how
the maturity of a bond affects its interest rate sensitivity. We will also be
able to show how coupon rates affect interest rate risk.

5.5.1 Prices and yields

Consider a hypothetical one-year bond and a hypothetical five-year bond.


Both have face values of $1000 and both have yields of 9%. They are both
priced at par (which, using our earlier results tells us that the coupon rate
on each bond is 9%).

Let’s see hypothetically, how we would expect the prices of these two
bonds to change if the yield on each was to change. Figure 5.2 shows
this for yields between 0 and 14%.

The figure shows us very clearly that the longer dated bond has greater
sensitivity to yields. As yields rise, the price of the longer dated bond
declines more quickly than that of the short dated bond and as yields fall
the longer dated bond rises in value more quickly.

These plots thus seem to suggest that longer maturity bonds are more sen-

82
sitive to interest rate changes. Thinking about how present values are com-
puted supports this suggestion. Assume a flat term structure at rate r and
some zero-coupon bonds with face value $1. A one-period zero-coupon
1
bond has a price of (1+ r)
and a K period zero has price (1+1r)K . Clearly,
given the power of K in the denominator, small changes in r will affect the
price of the K period zero much more than they affect the price of the one
year zero.

Prices and yields for zero-coupon bonds: example

Consider three zero-coupon bonds, with 1, 10 and 20 years to maturity


and face value 100. Each starts with a yield of 6% and then the yield on
each rises to 7%. How do their prices change?

Bond Price at 6% Price at 7% Price Change Percentage change


1yr 94.34 93.46 -0.88 -0.93%
10yr 55.84 50.83 -5.00 -8.96 %
20yr 31.18 25.84 -5.34 -17.12%

Obviously, the size of the percentage fall in price is increasing in the


maturity of the bond.

5.5.2 Measuring the sensitivity of bond prices to interest


rates

Take a generic bond with annual cashflows (for simplicity). Let the cash-
flow on date i be Ci , the yield is y and there are k years to maturity. The
bond price is:

k
C
P= ∑ (1 +i y)i
i =1

We want to measure the sensitivity of prices to yields. Thus compute the


derivative of P with respect to (1 + y):

83
k k
dP Ci 1 Ci
=−∑i i +
= − ∑ i
d (1 + y ) i =1 (1 + y )
1 1 + y i =1 (1 + y ) i

If we wanted to work out the percentage change in price for a 1% change


in yields we would compute the elasticity of the bond price with respect to
yields. This is:

(1 + y) dP 1 k Ci
=− ∑i
P d (1 + y ) P i =1 (1 + y ) i

We define the Macaulay duration of the bond, D, to be the negative of this


elasticity. So:

1 k Ci
D= ∑ i
P i =1 (1 + y ) i
(5.3)

This measure, D, tells us precisely how a percentage movement in yields


translates into a percentage movement in price. High duration bonds ex-
perience greater percentage drops in price than low duration bonds when
yields rise. Thus an investor with a bond portfolio that has high duration
should be worried about the prospect of yield increases.

Looking at how the duration is computed, it is a weighted average of the


times to cashflows for the bond (i.e. in equation (5.3) it is a weighted av-
erage of the values of i). The weight associated with each measure of time
is equal to the present value of the cashflow to be received on that date
divided by the price of the bond. These weights must sum to exactly one,
as the bond price is the sum of the present values of the cashflows i.e.

1 k Ci k
Ci /(1 + y)i
P i∑ ∑
D= i i
= i × w i where w i =
=1 (1 + y ) i =1
P

Ci
and ∑ik=1 wi = 1 as P = ∑ik=1 (1+ y ) i
.

The implications of this are that:

84
• Bond durations should be positive.

• The bond duration will be somewhere between zero and the time to
maturity of the bond (i.e. a seven-year bond has a duration between
zero and seven).

• Given that for most bonds, most of the cash is paid at the end of
the bond’s lifetime, the bond’s duration is usually pretty close to the
time to the bond maturity.

• A zero-coupon bond with k periods to maturity always has a dura-


tion exactly equal to k.

Modified Duration (sometimes also confusingly referred to as ‘Volatility’)


is often reported for fixed income instruments. It is computed as:

D
DM =
1+y

This measure is useful as it can be employed to give approximate percent-


age changes in the price of a bond for a known change in yields. If yields
change by a small amount ∆y then the percentage change in price is:

∆P ∆y
≈ − D M ∆y = − D
P (1 + y )

Note that this relationship is an estimate and is not perfect. It relies on a


linear approximation to the relationship between prices and yields and we
know from Figure 5.2 that this relationship is not linear.

Duration: worked example

Consider a bond with five years to maturity paying annual coupons at


a rate of 7%. Its yield to maturity is 5%. The duration of this bond is
computed in the table below. Note that we have assumed a face value of
$100, but the duration figure does not depend on this (i.e. you would get
the same answer for duration if you had chosen a face value of $50 or $720
or $10,000).

85
Table 5.2: Duration: worked example

Year Cashflow PV(Cashflow) Weight Year×Weight


1 7.00 6.67 0.06 0.06
2 7.00 6.35 0.06 0.12
3 7.00 6.05 0.06 0.17
4 7.00 5.76 0.05 0.21
5 107.00 83.84 0.77 3.86
Total 108.66 1.00 4.41

Given the information above, the price of $100 worth of face value is
$108.66. The column labelled ‘Weight’ gives the ratio of the PV of the
cashflow at each particular date to the price of the bond. The final column
takes each weight and multiplies it by the time to receipt of that cashflow.
Finally the duration is the sum of values in the final column and thus, this
five-year bond has a duration of 4.41.

5.6 Spot and forward interest rates

Assume we’re looking at a set of spot interest rates for multiples of one
year (for simplicity):

• Today’s two year spot rate can be thought of as being built from:
– Today’s one year spot rate and ...
– the rate we would agree today for a one-year loan/deposit to
be made in exactly one year from today.
• The implied rate that we would agree today for a one-year deposit
to be made starting at some future date is called a forward rate.
• Today’s three-year spot rate can be thought of as being built from:
– Today’s one year spot rate and ...
– the rate we would agree today for a one-year loan/deposit to
be made in exactly one year from today and ...

86
– the rate we would agree today for a one-year loan/deposit to
be made exactly two years from today.

Notation:

• Let the current one year spot rate be r1

• Let the current N year spot rate be r N

• Forward rates: Call the implied one year spot rate covering the pe-
riod from one to two years from now f 2 , the implied spot rate from
the end of year two to the end of year three f 3 and so on.

Then we have:

(1 + r N ) N = (1 + r1 )(1 + f 2 )(1 + f 3 ) . . . (1 + f N )

We can also write:

(1 + r N ) N = (1 + r N −1 ) N −1 (1 + f N )

Note from the equation above:

• If I tell you r1 and r2 , you can work out f 2 .

• If, in addition, I now tell you r3 , you can work out f 3

• If I also tell you r4 you can infer f 4

• and so on . . .

Thus a forward rate curve showing implied future one-period rates can be
inferred from the set of spot rates of different maturities (and vice versa).
The relationship between spot rates and time to maturity is called the term
structure of interest rates.

87
Forward rates: worked example

Assume that the one period spot rate is 3%, the two period spot rate is 4%
and the three period spot rate is 4.5%. Then the forward rate covering the
period from one to two years from now is:

(1 + r2 )2 1.042
f2 = −1 = − 1 = 0.050 = 5%
(1 + r1 ) 1.03

The forward rate covering the period from two to three years from now is:

(1 + r3 )3 1.0453
f3 = − 1 = − 1 = 0.055 = 5.5%
(1 + r2 )2 1.042

5.6.1 Extracting forward rates from bond prices

Consider a two year bond with cashflows C1 and C2 . Its price is:

C1 C2 C1 C2
P= + = +
(1 + r1 ) (1 + r2 )2 (1 + r1 ) (1 + r1 )(1 + f 2 )

Clearly, if you have prices and cashflows for two bonds with different
cashflows streams one can infer the one and two year forward rates.

For simplicity define:

1 1
d1 = , d2 =
1 + r1 (1 + r1 )(1 + f 2 )

Data:

• Bond A has face value of 1000, an 8% coupon rate and two years to
maturity. Its price is 980.
• Bond B has face value of 1000, an 12% coupon rate and two years to
maturity. Its price is 1050.

88
Given the above we know that:

980 = 80d1 + 1080d2


1050 = 120d1 + 1120d2

Solving these equations gives d1 = 0.91 and d2 = 0.84. Therefore:

s
1 1
r1 = − 1 = 9.89% , r2 = − 1 = 9.11%
d1 d2

and then:

1 1
f2 = −1 = − 1 = 8.33%
(1 + r1 ) d2 1.0989 × 0.84

5.7 The term structure of interest rates

As we have already seen, a term structure is a mapping between tenors


and spot rates at those tenors. Graphically, it is a plot of spot rates on the
y-axis against tenors on the x-axis.

As we saw earlier, as of summer 2016, the UK term structure was up-


ward sloping (i.e. spot rates for longer maturities are greater than those
for smaller maturities). This is the normal state of affairs. But this is not al-
ways true. Sometimes term structures slope downwards, sometimes they
are flat, sometimes they are hump-shaped.

This leads to the following question. What determines the shape of the
term structure?

Below, we will briefly review three candidate theories that purport to ex-
plain the shape of the term structure.

89
5.7.1 Expectations theory

Basic idea:

• Agents construct forward rates as (unbiased, rational) forecasts of


future one-period interest rates.
• Spot rates are then just combinations of forward rates.
• An increasing term structure, for example, reflects expected future
one-period rates (and thus forward rates) that are higher than the
current one period spot rate.
• A flat term structure indicates that agents expect future one-period
spot rates to be close to today’s spot rate.

So:

• Today’s upward sloping term structure might be interpreted as sug-


gesting that markets think that interest rates are going to rise.

5.7.2 Liquidity premium theories

Extend the expectations idea:

• Start from the basic expectations theory (i.e. term structure slope
reflects expected future interest rate changes).
• Add the notion that savers have a preference for short-term securi-
ties over long-term securities.
• This means that long-term securities must offer a larger average re-
turn than short-term ones, leading to long-term rates naturally being
greater than short-term rates.
• Thus the natural shape of the term structure is upward sloping.
• Only if interest rates are expected to decline strongly should we see
a negatively sloped term structure.

90
5.7.3 Market segmentation theory

This is not much of a theory to be honest:

• Different clienteles of investor operate in the long- and short-term


deposit and loan markets.

• These investors don’t trade across short- and long-term securities.

• This implies that the short-rate set through trade amongst those in-
terested in short-term securities and long-rates set separately among
investors with a long-horizon.

• As there is no link between the two segments, this means there is no


link between equilibrium long- and short-rates.

• The outcome is that the term structure can look whichever way you
want it to.

5.8 Summary

In this chapter we have focussed on fixed income instruments. We have


covered:

• Basic bond structures and features

• The term structure and bond pricing

• Duration and sensitivity of bond prices to yield changes

• Links between spot and forward interest rates

• Term structure theories.

We have not had time to cover corporate bond pricing however. This is
made more complicated by the increased likelihood that a company will
default on its bond obligations.

91
Reminder of learning outcomes
Having completed this chapter, and the Essential reading and activities,
you should be able to:

• describe the cashflow patterns of various types of bond.

• use NPV to price bonds.

• compute the yield to maturity on a bond.

• compute a bond’s duration and use it to assess that bond’s interest


rate risk.

92
5.9 Exercises
1. A seven-year bond pays coupons annually, with the first due one
year from now. It has face value equal to £1000 and coupon rate 6%.
The term structure of interests is flat with all spot rates equal to 8%.
What is the price of the bond?

2. The effective annual interest rate is 5% and is constant across time


and maturities. A bond has four years to maturity, face value $100,
annual coupon rate equal to 7% and pays coupons semi-annually.
What is the fair price of this bond?

3. The current term structure of interest rates is such that the one-year
rate is 0.5%, the two-year rate is 1%, the three-year rate is 2% and
thereafter the term structure is flat at 2.25%. What is the price of a
five-year bond which pays coupons annually at rate 4%, assuming a
face value of £500?

4. Using the same term structure as in the previous question, price a


two-year bond with face value £1000 and coupon rate 2%. Compute
the yield to maturity on this bond.

5. Consider a treasury bond with eight years to maturity, coupon rate


4%, face value $100 and a quoted yield of 3.90%. If coupons are paid
semi-annually, show that the bond price is 100.68. [HINT: your for-
mula for valuing an annuity should come in handy here.]

6. A bond has exactly three years to maturity and pays coupons (annu-
ally) at a rate of 6%. Its yield to maturity is 5%. What is the price
of £100 of face value of this bond? What is the bond’s Macaulay du-
ration and, based on the duration, by how much you would expect
the bond’s price to change if its yield to maturity was to increase by
0.10%?

7. Using the data in question 3, compute the one-year forward rate cov-
ering the period between the end of year 1 and the end of year 2.
Compute the forward rate between the end of year 2 and the end of
year 3 as well. Comment on the shape of the forward rate curve with
reference to theories of the term structure of interest rates.

93
Chapter 6

Stock markets and equity


valuation

Essential reading
Brealey, Myers and Allen (2017), Chapter 4.

Learning outcomes
By the end of this chapter, and having completed the Essential reading and
activities, you should be able to:

• describe the key cashflow and control features of stocks and shares.

• value stocks using NPV techniques.

• compute the Present Value of Growth Opportunities implicit in a


stock’s price.

6.1 Introduction

We now move on to the valuation of stocks issued by corporations or, as


they are commonly called in the UK, shares. In aggregate, the outstand-

94
ing stock in a firm is referred to as its equity and those holding stock are
equityholders.

Holding a stock or share gives an investor two key rights:

• A right to vote on corporate affairs at company AGMs and EGMs. As


such, equityholders, in aggregate, are the owners of the firm. They
are involved in the key long-term decision-making process, although
they often delegate day-to-day running of the company to a set of
professional managers.

• A right to a share in the dividend paid by the firm.

Thus there is both a control and a cashflow benefit to being a shareholder.


It is worth noting that investors who have purchased bonds issued by a
company do not get a right to vote. Thus bondholders are creditors not
owners of the firm.

It is also worth noting that equityholders have a claim to the profits on the
firm (via their dividend payments), but only after all other claims have
been satisfied. So equityholders only get paid dividends after employees
have been paid wages, the government has been paid any tax owed to it,
bondholders have been paid any interest owing to them etc. Thus, equity-
holders are often called residual claimants to company cashflows which, in
plain English, means that a firm can only distribute dividends if all other
claims have been settled. This makes the cashflow to equityholders rather
risky.

Here is some terminology that we will come across when thinking about
equity markets:

• Common stock: a security representing a share in the ownership of


a corporation. Also called shares or equity.

• Market value of equity: if a firm issues N shares and each share is


priced at P, then the market value of that firm’s equity is N × P. It is
just the total value of all of the shares in issue from a particular firm.

• Initial public offering: the first sale of stock in a corporation to the


public.

95
• Secondary market: a market, often a stock exchange, in which pre-
viously issued shares are traded amongst investors.
• Dividends: payments made by companies to shareholders. These
are usually ex-ante uncertain (unlike bond coupons).
• Dividend yield: the ratio of annual dividend to share price. A mea-
sure of your cashflow as an equityholder relative to the value of your
investment.

The ultimate goal of this chapter is going to be to use data available to an


investor to value a particular equity share, for example to come up with
an estimate of the right price for a single share of Marks and Spencer PLC.
Thus, given information on the number of shares that the company has
issued, we will be able to estimate the market value of equity for the firm.

6.2 Equity valuation and discounting dividends

6.2.1 Defining expected returns

To begin to find the right price for a particular stock, let’s start with an
investor who buys a single share.

You buy a share in a corporation that has a current price of Pt . You expect
that, at the end of one year, the firm will pay you a dividend (Dt+1 ) and
after the payment of the dividend you will be left with a share worth Pt+1 .
You don’t know with certainty the values Dt+1 and Pt+1 today.

You wish to estimate the percentage one-year return you will obtain from
holding the stock. As the future dividend and the future price are un-
known, you must estimate them. Your expected return is:

Et [ Dt+1 + Pt+1 ] Et [ Dt+1 + Pt+1 − Pt ] Et ( Dt+1 ) Et ( Pt+1 − Pt )


r= −1 = = +
Pt Pt Pt Pt

This variable r is obviously uncertain and it measures the expected per-


centage increase in the value of this investment over the year that you

96
have held it. You started out buying a share worth Pt and ended the year
with a share that you expect to be worth Et ( Pt+1 ) and a cash dividend that
you believe will be Et ( Dt+1 ) and so the expected return measures the ex-
pected growth in the value of your cash invested. We call the first term on
the right-hand side of this equation the expected dividend yield and the
second term the expected capital gain.

Example

I buy a share of MKS for 299 pence today. I expect it to pay me a dividend
of 17 pence in one year and, after the dividend has been paid, I expect the
price of the share to be 320p. My expected return is:

Et [ Dt+1 + Pt+1 ] 17 + 320


r= −1 = − 1 = 0.127 = 12.7%
Pt 299

This expected return is comprised of an expected capital gain of approxi-


mately 7% and an expected dividend yield of 5.7%.

6.2.2 Constant expected returns and stock prices

Now we’re going to turn the expected return equation around.

Assume: for the particular stock we’re looking at, investors demand a
constant expected return of r̃. They set the value of r̃ on the basis of their
assessment of stock risk. As we will see in later chapters (and as common
sense suggests), investors will require larger expected returns from stocks
with larger risks.

Putting r̃ onto the left-hand side of our expected return equation and rear-
ranging we get:

Et [ Dt+1 + Pt+1 ] Et [ Dt+1 + Pt+1 ]


r̃ = −1 ⇒ Pt =
Pt (1 + r̃ )

This result has a very straightforward interpretation. The right price for
the stock today is equal to the discounted value of the price and the divi-

97
dend that one expects one year from now, where we have used investor’s
required rate of the return as the discount rate. So we’re back to the present
value computations that we’re completely comfortable with.

But we can go further. Consider the Et ( Pt+1 ) term on the right-hand side
of the preceding equation. By the same logic we just used:

Et [ Dt+2 + Pt+2 ]
Et ( Pt+1 ) =
(1 + r̃ )

i.e. one’s best guess of the right price for the stock at time t + 1 is equal to
the present value of the expected price at t + 2 and the expected dividend
at t + 2. So we can rewrite our pricing equation as:

E [D +P ]
h i
Et Dt+1 + t (t1++2 r̃) t+2 Et ( Dt+1 ) Et ( Dt+2 ) Et ( Pt+2 )
Pt = = + +
(1 + r̃ ) (1 + r̃ ) (1 + r̃ )2 (1 + r̃ )2

I can continue to use the same logic to substitute out Et ( Pt+2 ) and Et ( Pt+3 )
and so on. If I do this and if we make the assumption that Et ( Pt+k )/(1 +
r̃ )k tends to zero as k tends to infinity then we obtain the stock pricing
equation below:


E t ( Dt + i )
Pt = ∑ i
(6.1)
i =1 (1 + r̃ )

This method for estimating a stock’s price is called a dividend discount


model. It implies that:

• The stock price is the present value of an infinite stream of dividends


– The stream is infinite as, hypothetically, the company lives for-
ever.
– This infinite lifetime is clearly a false assumption, but it’s prob-
ably not too costly as the present value of the next 30 or 40 an-
nual dividends and the present value of the infinite stream are
not too different due to the large effect discounting has on the
value of distant cashflows.

98
• Prices are greater when expected dividends are greater

• Prices are lower when the expected return required by investors (r̃)
is greater.

– Thus, other things equal, riskier stocks will have lower prices as
their cashflows are discounted more heavily.

6.2.3 Special case 1: constant expected dividends

We can come up with simpler stock pricing equations if we’re willing to


make some assumptions about future dividends. Assume that we expect
all future dividends to be constant at a level of D. The price becomes:


D
Pt = ∑ (1 + r̃)i
i =1

Well this is just the PV of a perpetuity stream and we know how to com-
pute that. It is:

D
Pt = (6.2)

where D is the constant future dividend.

Example 1

You expect stock XYZ to pay an annual dividend of 8p per annum and the
first dividend is to be paid one year from now. Its required return is 12.5%.
Its price should therefore be:

D 8
Pt = = = 64p
r̃ 0.125

99
Example 2

Stock ABC is priced at 200p per share. It’s required return is 9%. If you
expect it to pay constant annual dividends in the future, with the first div-
idend one year from now, what is the implied annual dividend payment?

Well we have:

D D
Pt = ⇒ 200 = ⇒ D = 18p
r̃ 0.09

6.2.4 Special case 2: dividends expected to grow at a con-


stant rate

Assume now that we expect dividends to grow at a constant annual rate


every year in the future. Denote the next period’s expected dividend by D
and the growth rate by g, so that the stream of expected dividends will be
D, D (1 + g), D (1 + g)2 ......

Then the stock price is:


D (1 + g ) i −1
Pt = ∑ (1 + r̃)i
i =1

The right-hand side of this equation is just valuing a perpetuity with growth
and we know how to compute that. It gives us:

D
Pt = (6.3)
r̃ − g

This result is often called the Gordon Growth Formula for stock pricing. It
implies that stock prices are higher when dividends or their growth rates
are higher and stock prices fall when required returns rise. Of course,
though, it is only valid in a world where dividends are expected to grow
at a constant rate forever.

100
Example 3

You wish to value a stock. You estimate that it will pay a dividend of $8
in one year and that the dividends will grow at a rate of 5% per annum
thereafter. You require an annual return of 9% for holding a stock of this
risk level.

Using the formula above we have:

D 8 8
Pt = = = = $200
r̃ − g 0.09 − 0.05 0.04

How would your valuation change if you revised your view about divi-
dend growth and assumed it to be zero? In that case the price would be
8
equal to 0.09 = $88.89. This demonstrates the value of dividend growth
in that going from growth of 0 to 5% results in the stock price more than
doubling.

Example 4

A stock currently has a market price of $150. You believe that it will pay
a dividend of $6 at the end of the year and that dividends will grow at an
annual rate of 7% thereafter. What is the market’s required return on this
stock?

The Gordon Growth formula tells us that:

D
Pt =
r̃ − g

So:

6 6
150 = ⇒ r̃ = 0.07 + = 0.11 = 11%
r̃ − 0.07 150

Thus the stock has a required return of 11%.

101
6.2.5 Dividend discount models summary

Let us reflect on where we have reached so far. Under the assumption


of constant required returns from investors, we derived the dividend dis-
count mode, equation (6.1). If we believe in the constant expected returns
assumption (plus a couple of other more technical ones), then this equa-
tion can always be used to get the right price for a stock using a forecasted
stream of dividends and a required return.

In the special case where we expect future dividends to be constant we get


the pricing formula in equation (6.2). This formula is simpler, but it is only
valid under the constant expected dividends assumption. If you expect
dividends not to stay at a constant level it will give you the wrong price.

Finally, in the case where dividends are expected to grow at a constant


rate we get the Gordon Growth formula, equation (6.3). Again, it is worth
stressing that this formula will only give you the right price for the stock
if your assumption of constant growth for dividends is indeed correct. If
you think that dividends are going to change in the future, but not obey a
constant growth rule, you have to use equation (6.1).

6.3 Stock prices, earnings and dividends

Companies can use their earnings (i.e. their revenues net of all costs and
payments to creditors) to do two things. They can distribute earnings to
shareholders as dividends or, alternatively, they can retain the earnings
and invest them in building productive capacity that will be useful in gen-
erating future earnings and thus future dividends. Thus, the decision as
to how to use earnings is crucial in determining current and future cash-
flows to equityholders. A firm choosing to pay a lower dividend today
may be able to raise dividends in the future if today’s earnings have been
invested in projects that will generate future earnings. This might cause
stock prices to rise.

In order to think about the choice between retaining earnings and dis-
tributing them as dividends and the effect of this choice on prices, let’s
introduce some notation. We have:

102
• Payout ratio: is the proportion of earnings paid out as dividends.

• Plowback ratio: is the proportion of earnings retained by the firm


and used for investment.

Obviously the sum of the payout and plowback ratios is 1.

6.3.1 Book value of equity, return on equity and dividend


growth

To proceed, we will also need to define some further concepts:

• Return on equity: a measure of the amount of earnings that a dollar


of equity (book value) creates. Thus it is:

EPS
ROE =
Book value of Equity

ROE, retained earnings and dividend growth: the rate at which a firm’s
earnings can grow is governed by its ROE and by its plowback ratio. If we
denote the plowback ratio by ρ, earnings growth is:

g = ROE × ρ

Intuition:

• Assume constant ROE and plowback ratio.

• Plowback tells us how much of current earnings is retained for in-


vestment purposes and ...

• ROE tells us how much each dollar of equity contributes to earnings

• Put them together than you get growth in earnings (g)

103
Proof: start with the definition of growth in book value and then use the
definition of ROE:

BVt − BVt−1 = EPSt × ρ


= ROE × BVt−1 × ρ

So:

BVt − BVt−1
= ROE × ρ
BVt−1

Thus book value growth is determined by the product of ROE and the
Plowback ratio. Similarly, for dividends we have:

Dt = (1 − ρ) × EPSt = (1 − ρ) × ROE × BVt−1

Thus:

Dt−1 = (1 − ρ) × ROE × BVt−2

This means that growth in dividends must be the same as growth in book
value:

Dt − Dt − 1 BVt−1 − BVt−2
g= = = ROE × ρ
Dt − 1 BVt−2

So, if you plow back earnings into investment projects, this will enable
your dividends to grow faster. Growth will be greater when the ROE of
your investments is larger.

6.3.2 Prices, plowback and the ROE

Now that we understand how dividend growth is related to the ROE we


can plug that into our Gordon Growth pricing equation. The Gordon
Growth model is:

104
Dt + 1
Pt =
r−g

where Dt+1 is next period’s expected dividend and g is dividend growth.


We know from the above that dividend growth, g, is equal to the ROE
times plowback so:

Dt + 1
Pt =
r − (ROE × ρ)

Finally, we know that next period’s expected dividend is equal to next


period’s expected earnings multiplied by the payout ratio i.e. Dt+1 =
EPSt+1 × (1 − ρ). Thus, finally we have:

EPSt+1 × (1 − ρ)
Pt =
r − (ROE × ρ)

6.3.3 Example: the value of retaining earnings

To illustrate how the ROE and a firm’s payout decisions affect its value
let’s run a simple thought experiment with two firms that differ only in
their decisions regarding what to pay out and what to plow back.

• Firm 1: earns $20 per share every year and pays out all of this to
investors. Its required rate of return is 12.5%. What is its price?

Dt + 1 (1 − ρ) EPSt+1 EPSt+1 20
P= = = = = $160
r̃ r̃ r̃ 0.125
• Firm 2: has identical required return to firm 1 and an identical earn-
ings forecast for next year (i.e. $20). Instead of paying all of its earn-
ings as dividends, it commits to plowing back 25% of them. It’s ROE
is 18%. What is its price?
First, what is its dividend growth rate:

g = ROE × ρ = 0.18 × 0.25 = 0.045

105
Then what is its price:

Dt + 1 (1 − ρ) EPSt+1 0.75 × 20 15
P= = = = = $187.50
r̃ − g r̃ − g 0.125 − 0.045 0.08

What does this tell us? We see that:

• The firm that plows back earnings has a greater stock price.

• This is because its ROE is greater than the required return (r̃) so re-
taining earnings generates growth in earnings that is larger than the
discount rate.

A key factor here is the ROE. One can show that if the ROE is equal to
the required return, the two firms would be worth the same amount.1 So
a firm should only retain earnings if those retained earnings can be put
to work in a sufficiently strong manner to outweigh the benefit of paying
them as dividends.

The difference in value between the firm that plows back earnings and
the firm that does not is called the Present Value of Growth Opportunities
(PVGO). In our example the PVGO is $27.50 per share.

6.3.4 P/E ratios and required returns

Finally, we can use the definition of the PVGO, to write the following:

EPSt+1
P= + PVGO

1 If ROE is equal to r̃ then we have g = r̃ × ρ. This implies that:

(1 − ρ) EPSt+1 (1 − ρ) EPSt+1 EPSt+1


P= = =
r̃ − r̃ × ρ (1 − ρ)r̃ r̃
So the price of the firm that retains some earnings is equal to the price of the firm that
pays all earnings as dividends.

106
i.e. the stock price is equal to the present value of earnings under the
assumption that all earnings are paid as dividends (the first term) plus the
present value of growth opportunities (the second term).

Re-arranging, we get:

 
EPSt+1 PVGO
= r̃ 1 −
P P

This tells us that the ratio of a firm’s earnings to its price is driven by two
factors. The first is its required rate of return, r̃. Indeed, for a firm with
no PVGO, the earnings to price ratio is exactly equal to the return rate of
return. The second factor is the PVGO. Companies with large PVGOs will
have much smaller earnings to price ratios than similar firms with small
PVGOs. This is because a firm with a large PVGO is expecting consider-
able growth in earnings so that its price is way above its current earnings.

6.4 Summary

In this chapter we have explored the basics of stock pricing. We see that
stock prices depend on two fundamental factors:

• The future stream of dividends that a firm is expected to pay.


• The rate of return that investors require in order to hold stock.

Stock prices are equal to discounted values of future dividends, using in-
vestors’ required return as the discount rate.

We have explored how different assumptions about dividend paths affect


stock pricing equations and from that derived the Gordon Growth model.
Finally we examined how dividend growth rates can be linked to a firm’s
return on equity and its decision over what proportion of earnings to pay
out and what proportion to retain.

We have not, however, said much about how investor’s required rates of
return are determined. This topic is the subject of the next few chapters.

107
Reminder of learning outcomes
Having completed this chapter, and the Essential reading and activities,
you should be able to:

• describe the key cashflow and control features of stocks and shares.

• value stocks using NPV techniques.

• compute the Present Value of Growth Opportunities implicit in a


stock’s price.

108
6.5 Exercises
1. Stock X has a required return of 15%. Its earnings are expected to
grow at 5% per year forever and it pays out half of them as divi-
dends. If next year’s earnings are expected to be £20, what is its
current stock price?

2. Stock Y has a current price of £62.50. You believe that it will pay a
dividend of £5 per share in one year and that dividends will grow at
a nominal rate of 2% thereafter. What is the required return on the
stock?

3. Stock Z is expected to pay a dividend of £3 on one year, £8 in two


years and subsequently one expects dividends to grow at a rate of
4%. If the required rate of return on the stock is 10%, what is the
current stock price?

4. Firm A pays out all of its earnings as dividends and is expected to


earn $10 per share next year. Firm B has the same expected earnings
next year, but follows a policy of only paying 40% of its earnings as
dividends. It retains the remainder in the firm and invests them in
projects with a return on equity of 25%. If the required rate of return
for both firms is 20%, what are their prices and what is the PVGO
embedded in B’s price?

109
Chapter 7

Risk, return and stock markets

Essential reading
Brealey, Myers and Allen (2017), Chapter 7.

Learning outcomes
By the end of this chapter, and having completed the Essential reading and
activities, you should be able to:

• describe the history of risk and return in various asset classes.

• compute portfolio risk and return from mean returns and variances
of constituent stocks plus covariances between stock returns.

• show how and when diversification affects portfolio risk.

• define the beta of a stock with respect to a portfolio.

7.1 Introduction

We now keep our attention on equity markets and study the relationship
between the returns offered by equity portfolios and the risk carried by
those investments. First we look at the data on returns from stocks versus

110
bonds before moving on to a theoretical characterisation of how returns
on portfolios depend on returns on the underlying component assets.

We will cover:

• Portfolios

• The history of stock market risk and return

• How to measure portfolio risk

• Examples of portfolio risk computations

• Diversification

• Beta

7.1.1 Precursor: what is a portfolio?

In the introduction to this chapter I began talking about ‘portfolios’ of


stocks. But what is a portfolio?

In plain English, a portfolio is nothing more than a collection of asset hold-


ings. Unfortunately, we’ll need more than plain English in the remainder
of this chapter, though.

Portfolio weights

Assume that you wish to invest $V in a set of stocks. There are N stocks
and you split your money between them. If we denote by Vi the amount
of our original investment devoted to stock i then we must have:

N
∑ Vi = V1 + V2 + V3 + . . . + VN = V
i =1

i.e. when one adds up one’s investments across all stocks, the result must
be precisely the total amount invested.

111
If we divide the left- and right-hand side of the preceding expression by V
we get:

N N
Vi
∑ V ∑ wi = 1
=
i =1 i =1

where I have defined wi = VVi . We call wi the portfolio for stock i. It is the
proportion of each dollar invested that one has allocated to stock i. So, for
example, if w1 = 0.40, that means that one has invested 40% of one’s finds
in the first stock.

Equal weights and value weights

In an equally weighted portfolio, every stock is allocated the same amount of


funds. Thus every stock has the same portfolio weight and that portfolio
weight is:

1
wi = for all i
N

where N is the number of stocks in the portfolio.

In a value weighted portfolio each stock is assigned a weight equal to its mar-
ket value divided by the total value of all of the stocks in the portfolio.1 So
if we define Ei to be the market value of the ith company then the weights
in a value weighted portfolio are:

Ei
wi =
∑N
j =1 E j

Thus, in a value weighted portfolio one is devoting funds to an asset de-


pending on its relative size. Bigger stocks (in a market value sense) are
given greater weight than smaller stocks. Value weighted portfolios are
1 Asa reminder, the market value or market capitalisation of a stock is the price of
an individual share in the company multiplied by the number of shares issued. So it’s a
measure of the aggregate stock market value of a company.

112
popular choices for investors as it seems to make sense to invest more
money in ‘more important’ (i.e. larger) stocks.

Portfolio returns

We have previously defined the return on a stock. If we are currently at


the end of year t and are computing the return that a stock has delivered
over the previous year we have:

Pt + Dt Pt + Dt − Pt
Rt = −1 =
Pt−1 Pt−1

where Dt is the dividend paid (assumed at the end of the year), Pt is the
end of year stock price (after the dividend has been paid) and Pt−1 is the
stock price at the end of the preceding year. This return can be interpreting
as the percentage growth in one’s investment that one would have experi-
enced if one had invested in this stock alone.

But investors rarely dedicate all of their funds to just one stock. As indi-
cated above, they usually form portfolios of stocks. This prompts us to ask
how to measure the return on a portfolio of stock. The answer turns out to
be straightforward.

To illustrate, assume that one formed a portfolio of two stocks and to make
our computations easier, assume that these stocks paid no dividends. You
invested Vt−1 so that:

Vt−1 = V1,t−1 + V2,t−1

where Vi,t−1 is the amount of one’s investment devoted to the ith stock. If
at the end of the year, the value of one’s investment in stock i is Vi,t then
we also have:

Vt = V1,t + V2,t

What is the return on your portfolio? It is just the percentage increase in


the value of the portfolio i.e.

113
Vt Vt − Vt−1
R P,t = −1 =
Vt−1 Vt−1

This is simple, but we can go one step deeper. Assume that the price of
stock 1 at the beginning of the year was P1,t−1 . Given that we invested
V1,t−1 in stock 1 then we must have bought K1,t−1 shares where this is
defined by:

V1,t−1 = K1,t−1 × P1,t−1

We can define K2,t−1 similarly. This tells us that:

Vt−1 = K1,t−1 × P1,t−1 + K2,t−1 × P2,t−1

Given that we now know exactly how many shares we bought of both
stock 1 and stock 2 we can rewrite the portfolio value at the end of the
year. If Pi,t is the end of year price of stock i then:

Vt = K1,t−1 × P1,t + K2,t−1 × P2,t

So now let’s rewrite the portfolio return. We have:

Vt − Vt−1 (K × P1,t + K2,t−1 × P2,t ) − (K1,t−1 × P1,t−1 + K2,t−1 × P2,t−1 )


R P,t = = 1,t−1
Vt−1 Vt−1

Re-grouping the terms in the numerator we have:

K1,t−1 × ( P1,t − P1,t−1 ) + K2,t−1 × ( P2,t − P2,t−1 )


R P,t =
Vt−1

Then we can write:

( P1,t − P1,t−1 ) ( P2,t − P2,t−1 )


K1,t−1 × P1,t−1 × P1,t−1 + K2,t−1 × P2,t−1 × P2,t−1
R P,t =
Vt−1

114
( Pi,t − Pi,t−1 )
Now using the knowledge that Vi,t−1 = Ki,t−1 × Pi,t−1 and Ri,t = Pi,t−1
we have:

V1,t−1 × R1,t + V2,t−1 × R2,t


R P,t =
Vt−1

Vi,t−1
Last, using the definition of the portfolio weight, wi = Vt−1 , we have:

R P,t = w1 R1,t + w2 R2,t

Thus, the return on a portfolio of assets is just the sum of portfolio weights
multiplied by returns on the constituent assets. In an N asset portfolio we
get the same result and allowing for dividends doesn’t change anything.
In the N stock setting we have:

N
R P,t = ∑ wi Ri,t
i =1

This makes working out portfolio returns easy. Mathematically, the linear-
ity of this expression will also make our subsequent derivations nice and
clean.

Note that whether in the two stock case or the N stock case, it must be the
case that the sum of the portfolio weights is equal to one (i.e. ∑i wi = 1).

7.2 Stock and bond market returns: the data

7.2.1 Comparing the US stock market and T-bills

Now that we know what a value-weighted portfolio is, let’s look at the ev-
idence on how the value weighted portfolio of all (or almost all) US stocks
has performed relative to Treasury bills over the 20th century. Treasury
bills are short-maturity US government bonds and thus can be thought of
as risk-free assets.

115
Figure 7.1: Compounded returns on the US stock market and T-bills, plus
compounded inflation

10 5
Stock
T-bill
Inflation

10 4
Value of a $1 investment (Dollars)

10 3

10 2

10 1

10 0

10 -1
1900 1920 1940 1960 1980 2000 2020
Year

In Figure 7.1 we show the compounded return on the US stock market


versus that from T-bills. To understand this compounded return, assume
that you invested $1 in the US stock market in 1900. In every subsequent
year, your investment will grow or shrink according to the stock market
return in that year. The compound return at the end of year T is thus:

T
$1 × ∏ (1 + R t )
t=1901

P +D −P
where Rt = t Ptt−1 t−1 is the stock market return in year t. The com-
pounded stock market return in Figure 7.1 is the blue line. The red line
is another compounded return, but this time from T-bills. The yellow line
is the compounded value of annual inflation. This compounded inflation
line shows the price that a good costing $1 in 1900 would be selling for in
year t. On the x-axis of the Figure we have time and on the y-axis is the
compounded return. Note that the y-axis scale is logarithmic, not linear.

The first observation to make from Figure 7.1 is that an investment in US


stock markets has clearly out-performed an investment in T-bills over the
20th century. A $ investment in stocks in 1900 would have grown to over

116
$10,000 by the end of the century while a $1 investment in T-bills would
have reached only $100. In fact the T-bill investment only just outstrips
compounded inflation.2

However, it also appears to be the case that the stock market is a more
volatile investment than T-bills. The blue line is much more jagged than
the red line, reflecting the increased risk of stock market returns relative to
government bond returns.

If one looks at the annual returns from the stock market this volatility be-
comes clear. The one-year returns for every year across the 20th and 21st
centuries are shown in Figure 7.2. The mean stock market return from
these data is roughly 11% per year. The mean T-bill return is 3.5%, imply-
ing that on an average annual basis an investment in stocks outperformed
a bond investment by 6.5% per annum.

The risk of stock markets is very clear from Figure 7.2. There are individual
years when the US stock market delivered a return above 50% and single
years when an investment in the US stock market lost almost 50%. Thus
the spread of annual outcomes around the 11% average is considerable. If
one computes the standard deviation of stock market returns one gets a
number close to 20%. The comparable figure for T-bills is 3%. Thus bond
markets are an order of magnitude less risky than stock markets.3

Thus, to summarise. The US stock market has been, historically, a much


more profitable investment than an investment in T-bills. However it is
also a way more risky investment. While an investment in stocks clearly
outperforms bonds in the long-run, there are short periods of time when a
bond investment clears beats a stock investment (e.g. in the Great Depres-
sion or the recent Global Financial Crisis).
2 If the compounded return had been exactly equal to compounded inflation then the
total return across the century in purchasing power terms would be zero.
3 To get an idea of what a standard deviation of 20% means, note that if a variable is

Normally distributed there is a 10% chance that one gets an outcome outside of the range
[µ − 1.64σ , µ + 1.64σ] where µ and σ are the variable’s mean and standard deviation
respectively. Thus our US stock market return has a 10% chance of being outside the
range [-21.8% , +43.8%]. That’s a one in 10 chance of a very extreme number. One caveat
to this is that it relies on the assumption of a Normal distribution, which is debatable in
the stock market context.

117
Figure 7.2: Annual returns on the US stock market

60

40

20
Annual percentage return

-20

-40

-60
1900 1920 1940 1960 1980 2000 2020
Year

7.2.2 Risk and return for narrower US portfolios

When we break the broad universe of stocks into non-overlapping subsets


and also widen our bond universe to include corporate bonds, we can see
rather clear patterns in risk and returns. First, portfolios of small stocks
are, historically, much more risky and deliver a higher mean return than
portfolios of large stocks. Here we are measuring size (i.e. whether a
company is large or small) by its total market value of equity (i.e. the
price of each share multiplied by the number of shares in issue).

Table 7.1: Risk and return across asset classes

Asset class Mean return (%) Std Devn (%)


Large Company Stocks 9.6% 20.6%
Small Company Stocks 11.7% 33.0%
Long Term Corporate Bonds 5.9% 8.4%
Long Term Government Bonds 5.7% 9.4%
US Treasury Bills 3.7% 3.1%
Inflation 3.0% 4.2%

Looking at bonds, Treasury bills are the least risky and generate only

118
small mean returns, while corporate bonds are both much riskier and have
greater mean returns. Overall, there is a clear positive correlation between
risk and mean return. More risky portfolios tend to deliver greater ex-
pected rewards.

7.3 Portfolio risk and return

Now we have seen what the data have to tell us about portfolio risk and
return, let us ask (theoretically) how portfolio mean returns and portfolio
risk are generated. As before, we will measure portfolio risk using the
standard deviation of portfolio returns.4

The underlying question we ask in this section is as follows. Assume that


you’ve spread your invested wealth across N stocks. You know the mean
return and the return standard deviation on each stock and you also know
the correlation between all pairs of stock returns. How do you measure
portfolio mean return and variance?

Let’s start with some notation:

• Portfolio weights: xi is the proportion of your invested wealth that


you’ve allocated to asset i.

• Expected returns: µi is the mean/expected return on asset i.

• Variances: the variance of the return on asset i is σi2 . Thus, the stan-
dard deviation is σi .

• Correlations: define ρi,j to be the correlation between the returns on


stocks i and j.5

Using this notation, it is easy to show that the expected return on an arbi-
trary portfolio of assets is given by:
4 This is not the only way one could measure portfolio risk, but it is the most popular

choice and, analytically, a rather attractive one.


5 Note that we must have ρ = ρ and ρ = 1.
i,j j,i i,i

119
N
E( R P ) = x1 µ1 + x2 µ2 + . . . + x N µ N = ∑ xi µi (7.1)
i =1

So the expected portfolio return is just the sum of portfolio weights times
expected returns of individual assets.6 The more weight an asset has in a
portfolio, the more influence its mean return exerts on the portfolio mean
return.

The portfolio return variance is:

N N N
Var ( R P ) = ∑ ∑ xi x j ρi,j σi σj = ∑ xi2 σi2 + ∑ xi x j ρi,j σi σj (7.2)
i =1 j =1 i =1 i6= j

So the variance of portfolio returns depends on individual stock return


variances (there are N of these) and also on each of the possible corre-
lations between stock returns (and there are N ( N − 1) of these terms).7
Again, changes in portfolio weights change the importance of particular
variances and correlations.

7.3.1 Portfolio risk and return: two asset portfolios

To simplify, let’s look at these expressions for portfolios of only two assets.
The mean return is:

E( R P ) = x1 µ1 + x2 µ2

and if we use the requirement that the portfolio weights must sum to one
we can write this as:

E( R P ) = x1 µ1 + (1 − x1 ) µ2
6 This is a corollary of the statistical result that, if X and Y are random variables and a
and b are constants, then E( aX + bY ) = aE( X ) + bE(Y ).
7 If X and Y are random variables and a and b are constants, then Var( aX + bY ) =

a2 σx2 + b2 σy2 + 2 a bρ x,y σx σy .

120
The portfolio return variance is

Var( R P ) = x12 σ12 + x22 σ22 + 2x1 x2 ρ1,2 σ1 σ2


= x12 σ12 + (1 − x1 )2 σ22 + 2x1 (1 − x1 )ρ1,2 σ1 σ2

From this equation it is clear to see that, holding all else constant and as-
suming positive weights, increasing the correlation between stocks will
tend to make the portfolio return more volatile. Decreasing the correla-
tion will make the portfolio less volatile. Thus correlation is an important
ingredient of portfolio risk.

7.3.2 Portfolio characteristics: example

Consider the following scenario:

Suppose you invest 60% of your invested wealth in Vodafone (VOD) and
40% in Tesco (TSCO). You are told that:

• The expected return on your VOD stock is 20% and on TSCO it is


12%.
• The standard deviation of the annualized daily returns are 30% and
25.0%, respectively.
• The correlation coefficient between the two return series is 0.15.

What are the mean return and the risk of your chosen portfolio?

First the expected return:

E( R P ) = 0.60 × 0.20 + 0.40 × 0.12 = 0.168 = 16.8%

Now the variance:

σP2 = (0.62 × 0.302 ) + (0.42 × 0.252 ) + (2 × 0.6 × 0.4 × 0.15 × 0.30 × 0.25) = 0.0478

121
So the standard deviation is:


σP = 0.0478 = 0.2186 = 21.86%

Note that, in this example, the risk of the portfolio is lower than the risk of
either of the constituent assets. This will not always be true, but it is often
so. The fact that building portfolios of assets can often reduce risk will
be very important in what follows (because we will assume that investors
dislike risk).

7.3.3 Diversification

Let us build on the observation about portfolio risk often being lower than
the risk of individual securities from the end of the previous section.

Start with an empirical fact. That fact is that stock returns tend to be posi-
tively correlated, but their correlation is usually much less than +1.

The implication of this result is that building a portfolio containing many


stocks is a smart thing to do because it when you do so the resulting port-
folio has low risk. We call a portfolio consisting of many small positions
in many stocks a diversified portfolio.

What is the intuition for this result? It is as follows:

• If stocks aren’t perfectly correlated, then some stocks will be going


up at the same time that others are going down.
• Thus their individual fluctuations tend to cancel out when you put
them together.
• This leads to portfolio risk being low.
• The reduction in portfolio risk associated with holding many differ-
ent stocks works best when correlations between them are small.
• But: the reduction in variance can’t happen indefinitely. At some
point, even though you’ve washed away all individual stock-specific
fluctuations, you still have to bear the risk of the entire stock market
moving.

122
Economists call the reduction in risk brought about by the construction of
broad portfolios of stocks diversification.

A mathematical illustration

Assume for simplicity you have an equally weighted portfolio of N stocks


and they all have the same variance, σ2 . Also assume that all pairs of
stocks have the same return correlation, ρ. Then:

N
σP2 = ∑ xi2 σ2 + ∑ xi x j ρσσ
i =1 i6= j

1
Now if the portfolio is equally-weighted, this means that xi = x j = N. So:

N
1 1
σP2 = ∑ N 2 σ2 + ∑ N 2 ρσ2
i =1 i6= j

Now, simplifying the sums:

1 2 1 1 ( N − 1) 2
σP2 = N 2
σ + N ( N − 1) 2 ρσ2 = σ2 + ρσ
N N N N

And finally, as we let the number of stocks N tend to infinity:

σP2 → ρσ2

So, as you increase N the portfolio return variance decreases but hits a
limit given by the product of correlation and variance. Thus, in large port-
folios, correlations play a very important role. The smaller the (average)
correlation between pairs of stocks, the lower is the portfolio risk.

Figure 7.3 shows the effect on the number of stocks in a portfolio (N) on
that portfolio’s risk. To build this figure, I have assumed that the correla-
tion between all pairs of stock return series is 0.5 and each stock has a risk
of 20%.

123
Figure 7.3: Diversification and portfolio risk

0.05

0.045

0.04
Portfolio return variance

0.035
Diversifiable
or
0.03 idiosyncratic
risk
0.025

0.02

0.015 Undiversifiable
or
systematic
0.01 or market
risk
0.005

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

In Figure 7.3, the blue line shows how portfolio return variance changes
with N. The red line is the original level of risk when the portfolio only
contains one asset. Clearly portfolio risk drops quite sharply as one moves
from a portfolio containing one stock to one containing 10 stocks. But
thereafter as N increases to 100, further reductions in variance are small.
So diversification is powerful when initially spreading your wealth across
a few assets, but its power diminished when you go from investing in a
few to investing in a lot of assets. Finally, we often describe the level of
risk that you cannot escape, regardless of how big N is, as undiversifiable
or market risk.

7.3.4 Beta: contribution of a stock to portfolio risk

Now we know how to measure the risk of an entire portfolio, now I want
to approach a different question. Assume that you hold a diversified port-
folio of stocks (i.e. a portfolio containing many stocks each with a fairly
small weight). What is the contribution of one particular stock, say stock
i, to your portfolio’s overall risk?

The answer to this question is that the contribution of asset i to the overall
portfolio variance is:

124
σi Cov( Ri , R P )
β i = ρi,P =
σP Var( R P )

We call this quantity the beta for asset i. It measures the change that one
would observe in total portfolio variance if one was to slightly increase the
portfolio weight on stock i.8

As such, we think of an asset with a large beta as having high risk (as it
contributes a lot to the portfolio’s risk). Note that beta depends on the cor-
relation (or covariance) between the portfolio return and the asset return.
The higher the correlation that stock i has with the overall portfolio return,
the greater the stock’s contribution to risk.

Intuitively, a positive beta stock is risky as if you raise its weight it accentu-
ates portfolio fluctuations. A negative beta stock hedges portfolio risk. If
you increase the weight on a stock with negative beta then portfolio vari-
ance will fall (as the stock’s return tends to move the opposite way to the
market).

In Figure 7.4 we have a graphical view of a stock with a beta of 0.75 to a


particular portfolio. On the x-axis you have returns on the portfolio and on
the y-axis are concurrent returns on the stock. When the portfolio return
increases by 1%, the stock return tends to increase by 0.75%. The slope
of the red fitted line is the stock’s beta, i.e. 0.75. Data points do not lie
precisely on the line as the relationship between stock and portfolio is not
perfect.

Finally, it is worth emphasising that the value that stock i’s beta takes
depends on the portfolio that you are holding (P). If two investors are
holding different portfolios, call them P and Q, but both portfolios contain
stock i, then the investors are likely to disagree on the stock i’s beta. This is
because the correlation between P’s return and stock i’s return need not be
the same as the correlation between Q’s return and stock i’s return. Only
if two investors hold the same portfolio are they guaranteed to agree on
the risk of each asset.
8 Those who have done some econometrics will recognise beta as the slope coefficient
in a linear regression of stock i’s return on the portfolio return.

125
Figure 7.4: Beta from a scatter plot

0.6
Data points
Fitted relationship: slope 0.75
0.5

0.4

0.3

0.2
Stock return

0.1

-0.1

-0.2

-0.3

-0.4
-0.4 -0.2 0 0.2 0.4 0.6 0.8
Portfolio return

Beta: examples

Assume that the portfolio in question is the FTSE-100 and we are asking
about the risk of two particular stocks.

Example 1: a stock has a beta of 0.8 to the FTSE-100

• This stocks tends to move in the same direction as the FTSE-100.


• For every 1% rise (fall) in the FTSE-100, you expect the stock to rise
(fall) by 0.8%.
• For an investor who holds the FTSE-100, this stock is fairly risky.

Example 2: a stock has a beta of -0.25 to the FTSE-100

• This stocks tends to move in the opposite direction to the FTSE-100.


• For every 1% rise (fall) in the FTSE-100, you expect the stock to fall
(rise) by 0.25%.
• This stock is not risky. In fact adding that stock to a holding of the
FTSE-100 will tend to reduce overall risk.

126
7.3.5 Portfolio betas

Assume that you have two stocks, the first has a beta of β 1 and the second
has a beta of β 2 . You form a portfolio of these two stocks with weights x1
and x2 . The beta of the resulting portfolio is:

x1 β 1 + x2 β 2

In general, in a portfolio of N stocks where the weight on stock i is xi and


the beta of stock i is β i , the beta of the portfolio is:

N
∑ xi β i
i =1

7.4 Summary

In this chapter we have covered a lot of ground. We have discussed the


history of risk and return on various markets and asset classes. We then
proceeded to describe how to measure risk and return empirically and
then how to compute portfolio risk and return theoretically. Finally we
introduced the concept of beta.

127
Reminder of learning outcomes
Having completed this chapter, and the Essential reading and activities,
you should be able to:

• describe the history of risk and return in various asset classes.

• compute portfolio risk and return from mean returns and variances
of constituent stocks plus covariances between stock returns.

• show how and when diversification affects portfolio risk.

• define the beta of a stock with respect to a portfolio.

128
7.5 Exercises
1. Stocks X and Y have expected returns of 12% and 20% respectively.
Their return standard deviations are 25% and 40%.

(a) If the stocks’ returns are uncorrelated, work out the mean return
and the return standard deviation on a portfolio which places
weight 1/3 on X and 2/3 on Y.
(b) Using the same weights, work out the portfolio risk and return
under the assumption that:
i. The returns have perfect positive correlation.
ii. The returns have perfect negative correlation.
(c) Use your preceding answers to illustrate the effects of diversifi-
cation on portfolio risk.

2. Using the same data as in question (1), and assuming that the stocks’
returns are perfectly negatively correlated, compute the portfolio weights
for X and Y that must hold if the portfolio is to have zero risk.

3. Below are some data on the percentage return on stock Z and on the
market portfolio (M) in 5 consecutive years:

X M
1 12 6
2 -5 0
3 25 5
4 7 10
5 6 15

(a) Compute the beta on stock Z with respect to the market (to two
decimal places)
(b) What does this beta imply for the manner in which Z’s returns
vary as the market return varies?

129
Chapter 8

Portfolio theory and the Capital


Asset Pricing Model

Essential reading
Brealey, Myers and Allen (2017), Chapter 8.

Learning outcomes
By the end of this chapter, and having completed the Essential reading and
activities, you should be able to:

• derive the CAPM from first principles.

• explain the CAPM’s assumptions and empirical implications.

• use the CAPM to work out expected returns on stocks or portfolios


of stocks.

• discuss the empirical shortcomings of the CAPM.

130
8.1 Introduction

In this chapter we build on the characterisations of portfolio risk and re-


turn from the last chapter. We re-introduce investors to the picture and
ask, if they understand how to assess risk and return for portfolios, which
portfolios will they choose to hold?

Throughout we assume two things about investor preferences. First, other


things equal, they prefer portfolios with greater expected returns to those
with smaller expected returns. Second, they prefer low risk to high risk
portfolios, holding other things constant.

We start by characterising the set of opportunities available to investors in


risk-return terms. This will use the maths from the previous chapter. Then
we ask whether there are some portfolios that investors will definitely not
choose, and then rule those portfolios out of our thinking. Finally, we will
take the remaining portfolios and ask how investors select between them.

The outcome of all of this will be the Capital Asset Pricing Model (CAPM).
This model relates the expected return on each portfolio (or stock) to its
risk (measured by that stock’s beta to the overall stock market).

We will close by empirically evaluating the CAPM and discussing some


alternative models.

8.2 Investor preferences

It is worth re-iterating the assumptions that we will make regarding in-


vestors and their preferences as understanding them is crucial. Through-
out this analysis we assume that:

• When comparing two portfolios of equal expected return, investors


prefer the portfolio with the smallest return standard deviation (or
variance).

• When comparing two portfolios of equal return standard deviation,


investors prefer the one with the higher expected return.

131
Thus, loosely speaking, investors like rewards (mean returns) and dislike
risk (standard deviation or variance).

8.3 Basic portfolio theory

Let’s start with a simple scenario. An investor wishes to invest some cash
in the stock market and has access to two stocks. How should she build
the portfolio of those two stocks to suit her preferences best?

Before the analysis, remember that if an investor builds a diversified port-


folio, the investor eliminates idiosyncratic risk from the portfolio. Given
the investor’s preferences already discussed, reducing risk is a good thing.
Thus we might guess that building portfolios of stocks is going to be better
than simply investing all of one’s money in a single stock.

We give our investor the following data on the two stocks. These data
come from actual market returns for the 15 years ending just before the
2008 financial crisis. The two stocks are:

• Anglo American PLC (AAL) - a global mining company

• Barclays PCL (BARC) - a retail and investment bank

Means and standard deviations for the two annual return series are given
in Table 8.1. In addition to the data in the table we are told that the corre-
lation between the returns on AAL and BARC is 0.2.

Table 8.1: Stocks: mean returns and standard deviations

Stock E( R i ) σi
AAL 16.3% 36.2%
BARC 9.6% 26.2%

132
8.3.1 Charactering the feasible set of portfolios

Now we have these data we can use the portfolio risk and return equations
form the previous chapter to work out portfolio characteristics for every
possible choice of portfolio weights. In Table 8.2 I have done that for a
subset of possible portfolios, starting from a portfolio totally invested in
BARC and then shifting portfolio weight gradually (in steps of 0.1) until
the portfolio is invested wholly in AAL.

To illustrate the computations, take the portfolio where a weight of 0.3 is


placed on AAL and 0.7 is placed on BARC. Then we have:

E( R P ) = 0.3 × 0.163 + 0.7 × 0.096 = 0.116


p
σp = 0.32 × 0.3622 + 0.72 × 0.2622 + 2 × 0.2 × 0.3 × 0.7 × 0.362 × 0.262
= 0.231

So what does this table tell us? First focus on expected portfolio returns.
As we know, BARC has the lowest expected return and AAL the largest.
The table tells us that as the weight on AAL is increased from 0.0 to 1.0,
the expected return increases linearly from the expected return on BARC
to the expected return on AAL (as you move down by one row at a time,
the mean portfolio return increases by 0.67%).1

But, more interestingly, the portfolio return standard deviation does not
increase monotonically as we shift weight from the less risky stock (BARC)
to the more risky stock (AAL), initially it falls and subsequently it increases
again. This is diversification in action, driven by the fact that the correla-
tion between the two return series is only 0.2. Thus, forming portfolios of
stocks can substantially reduce risk.

On Figure 8.1, we plot the combinations of risk and return available from
these portfolios, with risk (i.e. standard deviation) on the x-axis and mean
returns on the y-axis. This plot summarises what we saw in the table.
Some portfolios have lower risk than either of the stocks and that makes
portfolios attractive investments for risk-averse individuals.
1 Thechange in mean return is 0.67% as we can write the mean return as wE( R A ) +
(1 − w)E( R B ) = E( R B ) + w (E( R A ) − E( R B )). So every time you increase w by 0.1 you
change the mean portfolio return by 0.1 × (E( R A ) − E( R B )) = 0.1 × (0.163 − 0.096) =
0.0067.

133
Table 8.2: Portfolios of two stocks: mean returns and standard deviations

ω AAL ω BARC E( R P ) σP
0.0 1.0 0.096 0.262
0.1 0.9 0.103 0.246
0.2 0.8 0.109 0.235
0.3 0.7 0.116 0.231
0.4 0.6 0.123 0.234
0.5 0.5 0.129 0.244
0.6 0.4 0.136 0.259
0.7 0.3 0.143 0.280
0.8 0.2 0.149 0.304
0.9 0.1 0.156 0.332
1.0 0.0 0.163 0.362

However, it is worth emphasising that this result that building portfolios


can lead to lower risk than displayed by either stock is not guaranteed. It is
driven by the level of correlation between the two stock return series. The
higher the correlation, the less benefit from diversification there is. To il-
lustrate, on Figure 8.2 I have added the set of risk-return combinations one
would get if the correlation between the two return series was +1.0. In this
case, all of the risk reduction benefits from building portfolios has disap-
peared. It is worth bearing this in mind, while at the same time keeping in
mind the fact that in the real world most pairs of stocks have correlations
substantially below 1.0.

Introducing short-sales

In real world financial markets, many investors are allowed to short-sell


financial assets. This possibility radically changes the set of risk-return
opportunities to investors as it allows them to take positions with negative
portfolio weights.

In what follows we will refer purchasing stock in order to take a positive


weight in that stock a long position.

First let’s describe how a short-sale works. Assume that you currently do
not own stock X, but you wish to take a position that will earn you money

134
Figure 8.1: The AAL-BARC portfolio frontier

0.17

AAL
0.16

0.15
Portfolio expected return

0.14

0.13

0.12

0.11

0.1

BARC
0.09
0.22 0.24 0.26 0.28 0.3 0.32 0.34 0.36 0.38
Portfolio return standard deviation

when X falls in price. This can be done via a short sale. The steps are as
follows:

• Borrow stock X from a friendly broker.


• Take the stock you’ve borrowed and sell it. Now you have gone
short X.

Now let’s think about the future. Assume that you close this position after
one month. You close the position by:

• Going to market and buying stock X.


• Then you take that stock you’ve just purchased and deliver it to the
broker you originally borrowed it from.

So now, everything is settled. You’ve sold and then purchased the stock
and thus initiated and then closed your short position. Also, the broker
has seen the stock he loaned you returned to him.

But have you made any money on the short position?

135
Figure 8.2: The AAL-BARC portfolio frontier: the effect of correlation

• If stock X fell in price over the month that you were short, then you
sold at a higher price than you purchased at, thus you made a profit
(or a positive return).

• If X rose in price over the month then you sold at a lower price than
you subsequently bought and you made a negative return i.e. a loss.

Finally, in a portfolio setting, how do we account for short sales? Intu-


itively, with a short sale you benefit if a stock price falls while with a
normal long position you benefit if the price rises so, if long positions
have positive portfolio weights then short positions should have negative
weights.

This can be shown a little more mechanically. Assume you have £100. You
intend to short sell £50 of MKS and then invest all of your available funds
in LLOY. First let’s consider the short sale. Because you are selling £50 of
MKS, you get a cash inflow of £50. Adding this to your original £100, you
now have £150 that can be invested in LLOY. So, relative to your original
£100, your portfolio weight in LLOY is:

136
150
wL = = 1.5
100

Your portfolio weight in MKS is:

−50
wM = = −0.5
100

Note that, even with short sales, portfolio weights must still always sum
to one (i.e. -0.5 + 1.5 = 1.0). But now we can have negative weights (for
short sales) and also weights greater than +1!

Figure 8.3 shows how the allowance of negative portfolio weights extends
the set of possible risk/return combinations. Note, that in building this
figure I’ve (arbitrarily) set the minimum weight on any stock to be -1.0.
Permitting short sales allows one to extend the risk-return frontier and to
expand the set of opportunities available to investors.

Figure 8.3: The AAL-BARC portfolio frontier with short-sales

0.25

0.2
Portfolio expected return

0.15

0.1

0.05

0
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Portfolio return standard deviation

137
Summary: the portfolio frontier and diversification

We have seen that:

• Portfolios of securities tend to have much smaller risk (as measured


by return standard deviations) than individual securities.

• This is due to the benefits of diversification.

• Thus risk-averse investors will tend to prefer ‘diversified’ portfolios


(i.e. portfolios containing many stocks).

What is the source of this result?

• The benefit of diversification is driven by the fact that security re-


turns are not perfectly positively correlated.

• If all stocks returns were perfectly correlated, there would be no risk-


reduction benefits from diversification.

8.4 Optimal portfolios for investors: stocks only

We are now in a position to describe the optimal investment policy of a


risk-averse investor. The two key components of this process are the re-
sults regarding portfolio risk and returns we derived in the previous sec-
tion and the assumptions we made about investor preferences earlier.

To find an investor’s optimal portfolio, we take the following steps:

• Derive the feasible set: the set of portfolio risk and return pairs that it
is possible to generate from a given set of stocks.

• Identify the portfolio frontier: this is the set of portfolios with smallest
risk for each level of expected return.

• Identify the optimal portfolio: it must lie on the portfolio frontier and
must maximise the investor’s utility.

138
Step 1: the feasible set

The first thing that an investor must do is to work out what investment
opportunities are available. Let’s assume that the investor has access to
three stocks (AAL, BARC and CBRY). The feasible set of portfolios available
to the investor consists of all possible portfolios of those three stocks.

Using historical data on the three stocks above (to compute mean returns,
standard deviations and correlations), in Figure 8.4 I display the feasible
set of portfolios available to an investor. The blue line is the portfolio fron-
tier, which consists of the set of portfolios which have smallest risk for
their particular level of expected return. When one can invest in more
than 2 stocks, the area to the right of the frontier consists of portfolios that
one can build, but which do not have the smallest possible risk for their
respective level of expected return. On the figure I have marked the po-
sition of the individual stocks and one can see that BARC, for example, is
far to the right of the frontier. This means that placing all of one’s wealth
into BARC is not a smart investment strategy as there are portfolios that
delivers the same expected return with much smaller risk.

Figure 8.4: The feasible set of portfolios: three stocks


0.25

0.2

AAL
Expected Portfolio return

0.15

0.1 BARC

CBRY

0.05

0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Portfolio return STDV

139
Step 2: identifying frontier portfolios

How does one mathematically identify the set of frontier portfolios? We


follow the procedure below.

Fix a level of expected return, E. We need to work out which portfolio that
has that level of expected return has the smallest risk. If there are N stocks,
the investor must solve the following problem:

min Var( R P )
x1 ,x2 ,...,x N

where xi is the portfolio weight on asset i and Var( R P ) is the portfolio


variance. This minimisation is subject to the constraints that:

N
∑ xi = 1
i =1
E( R P ) = E

where E( R P ) is expected portfolio return and E is our chosen expected


return target.

To remind you how to compute portfolio variance, we have:

N N N
Var( R P ) = ∑ ∑ xi x j ρi,j σi σj = ∑ xi2 σi2 + ∑ xi x j ρi,j σi σj
i =1 j =1 i =1 i6= j

Solving this problem gives us the frontier portfolio for the chosen expected
return, E. We now need to repeat the entire process for all possible differ-
ent values of E and from this we can work out the complete frontier. An
optimizing risk-averse investor must choose his portfolio from this set.

As we have already seen, in our three stock example, the frontier is the
blue line on Figure 8.4.

In general, in the N-asset case, the portfolio frontier is a hyperbola (once


we allow unlimited short sales). Points to the right of the portfolio frontier

140
are portfolios that can be formed, but which are not minimum variance.
Such portfolios have high risk for their respective levels of expected re-
turn. As we saw above, in most cases the original securities do not lie on
the portfolio frontier. This is due to the benefit of diversification.

Last, we define the efficient set as follows. It is the set of portfolios in which
all risk-averse investors will find their optimal portfolio. Graphically, it
is that part of the portfolio frontier that lies above and to the right of the
portfolio on the frontier that has minimum variance. Mean-variance in-
vestors will not choose portfolios that are on the frontier but not in the
efficient set, as for such portfolios one can find an alternative choice that is
in the efficient set and which has identical standard deviation but higher
expected return.

Step 3: optimal portfolio selection

Now all our investor has to do is use his indifference curves to select one
portfolio from the efficient set. An indifference curve connects a set of
points where the investor has exactly the same utility (i.e. he is indifferent,
in utility terms, between all the points on the line). In our setting, where
investor wellbeing is defined by expected returns and return volatility, in-
difference curves will be upward sloping. To see why, consider the fol-
lowing situation. An investor currently holds a portfolio with expected
return 10% and return standard deviation 20%. If you were to offer him
an alternative portfolio, with a return standard deviation of 30%, for him
to be indifferent (i.e. just as well off in utility terms) between this portfolio
and the original one you would have to ensure that this alternative port-
folio has an expected return above 10%. This is because higher risk makes
investors worse off and thus, to compensate, they need higher expected re-
turns. Thus, indifference curves slope upwards. Comparing indifference
curves on our graphs with risk on the x-axis and expected return on the
y-axis, indifference curves that are further North and further West repre-
sent higher utility levels (as the further North one is, the higher is expected
return and the further West one is the lower is risk).

The investor should choose the portfolio which is at the point where his
indifference curve is tangent to the efficient set (as at this point he’s getting
the highest possible utility).

On Figure 8.5 I’ve shown the optimal portfolios for two investors:

141
• Investor X: is very risk averse, thus has steep indifference curves and
chooses a portfolio with low return standard deviation

• Investor Y: is less risk averse, has fairly shallow indifference curves


and chooses a portfolio with higher return standard deviation

Obviously, investors with different levels of risk aversion choose different


portfolios of risky securities and more (less) risk-averse investors choose
less (more) risky portfolios.

Figure 8.5: Optimal portfolios

0.25

0.2
Expected Portfolio return

0.15

X
0.1

0.05

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Portfolio return STDV

So, to recap, we have solved the optimal investment problem for a risk-
averse investor choosing portfolios of N assets. We first derived and plot-
ted the portfolio frontier – the set of all portfolios that have minimal risk
for their level of expected return. Then we superimposed the investor’s
indifference map on the plot of the frontier. The optimal portfolio is that
point where an indifference curve is tangent to the efficient set.

Obviously this all relies on several assumptions. These assumptions in-


clude:

• mean-variance preferences

142
• unrestricted position taking (e.g. unlimited short sales)

• no trading frictions

• ... and others.

8.5 Optimal portfolio selection: stocks plus a risk-


free asset

Finally, we are going to complicate our M-V optimization by adding a risk-


free asset. The risk-free asset has the following properties:

• its return is known (in advance), to be R f

• the variance of the return on the risk-free asset is 0

• the covariance of the return on the risk-free asset with the return on
any other asset is 0.

Our investor can build any portfolio of the N risky assets plus the single
risk-free asset. To incorporate the risk-free asset into our mathematics, we
do the following:

• Concentrate on choosing portfolio weights for the risky assets only


(i.e. we choose for each stock i, the portfolio weight xi ).

• The portfolio weight on the risk-free asset will then be 1 − ∑iN=1 xi .

• Note that this automatically means that the weights sum to 1 across
risky and risk-free assets.

Reformulating the problem: portfolio return means and variances

With the introduction of the risk-free asset, nothing changes about the
portfolio return variance. This is because the risk-free asset has no vari-
ance and no covariance with any other asset. The expected portfolio return
does change, though. It becomes:

143
" #
N N
E( R P ) = ∑ x i E( R i ) + 1 − ∑ xi R f (8.1)
i =1 i =1

The implication of this is that, as the mean return has changed, we’re going
to have to completely redo our derivation of the frontier.

Portfolio frontier: with a risk-free asset

The minimization problem that the investor must solve is the same as be-
fore, although now the set of assets is different. If E is the investor’s target
level of expected return then (s)he solves:

min Var( R P )
x1 ,x2 ,...,x N

subject to the constraint that:

" #
N N
E( R P ) = ∑ x i E( R i ) + 1 − ∑ xi R f = E
i =1 i =1

and where:

N N N
Var( R P ) = ∑ ∑ xi x j ρi,j σi σj = ∑ xi2 σi2 + ∑ xi x j ρi,j σi σj
i =1 j =1 i =1 i6= j

Note that we’ve lost the constraint that the weights sum to 1, as we’ve
already imposed this condition through our definition of the weight on
the risk-free asset.

Again the investor solves this problem for all possible values for E to de-
rive the efficient set.

144
Solving a simpler problem

Consider a single risky asset (or portfolio) and a risk-free asset. What are
the characteristics of portfolios of these two securities?

If we call the expected return and return variance for our randomly se-
lected risky security E( Rk ) and σk2 and the portfolio weight on the risky
security ω then the expected return on the portfolio is:

E( R P ) = ωE( Rk ) + (1 − ω ) R f

where R f is the risk-free rate. The variance and standard deviation of the
portfolio return are:

Var( R P ) = ω 2 σk2 , σP = ωσk

If we combine these two expressions and eliminate ω we can describe the


relationship between portfolio risk and return (i.e. the relationship be-
tween E( Rk ) and σk ):

σP
E( R P ) = ω [E( R k ) − R F ] + R f = [E( R k ) − R F ] + R f
σK

Clearly this is a straight line between risk and return. The intercept of the
E( Rk )− R F 2
straight line is R f and the slope of the line is σK .

In Figure 8.6 we show this set of portfolios for the case of AAL and using
a risk-free return of 7.5%.

Points on this line to the left of AAL have positive weights on both AAL
and the risk-free asset. Points on the line to the right of AAK have weights
on the risk-free asset that are negative and the weight on AAL is thus
greater than one.
2 As an aside, this slope is often called the Sharpe ratio of security k. It is a measure
that scales the expected return of security k in excess of the risk-free rate by the risk of
the security. Investors often compare securities using their Sharpe ratios, with a larger
Sharpe ratio being preferable as it implies that the investor receives more return per unit
of risk taken on.

145
Figure 8.6: Combining a risky asset (AAL) and the risk-free security
0.18

0.16 AAL

0.14
Expected Portfolio return

0.12

0.1

0.08

0.06
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
Portfolio return STDV

This result works for any asset. If you take a risky portfolio and the risk-
free asset and you combine them, then as you change the weight on the
risk-free asset, the set of risk-return opportunities that you can achieve
trace out a straight line.

Solving the full problem: graphically

Given that we now know that the set of investments generated by com-
bining the risk-free asset and any portfolio is a straight line, what is the
optimal portfolio of risky assets to select?

It must be true that:

• The optimal portfolio of risky assets must be the point at which the
frontier of risky assets is tangent to a straight line through the risk-
free rate — call this portfolio of risky assets the tangency portfolio.

• Combinations of the tangency portfolio and the risk-free asset gen-


erate portfolios that have minimal risk for their level of expected re-
turn.

• Thus, when we include a risk-free asset, the straight line joining the
tangency portfolio to R f is the efficient set.

146
• We also call the straight line joining the tangency portfolio to R f the
capital market line.

Figure 8.7: The optimal risky portfolio in the presence of a risk-free asset

0.25

Efficient set

0.2
Y
Expected Portfolio return

0.15

0.1

0.05

Risky asset frontier

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Portfolio return STDV

Why is the tangency portfolio optimal? It allows you access to the set of
opportunities which have lowest risk for their levels of expected return. In
Figure 8.8, we contrast the tangency portfolio with a sub-optimal portfolio,
Q. Mixing the tangency portfolio with the risk-free asset puts the investor
on the straight black line. Mixing Q and the risk-free asset puts the in-
vestor on the red line. Every investor prefers to be on the black line rather
than the red line because at every level of risk it offers greater expected
return.

Features of the optimal portfolio choice

When a risk-free asset is available in addition to N risky assets, then at a


mean-variance optimum:

• All investors choose the same portfolio of risky assets – the tangency
portfolio (T).

147
Figure 8.8: The tangency portfolio and a sub-optimal risky portfolio

"#$%!
)*+!

,!

'!

!!!$(!

!!!!!&! ! !!

• They then satisfy their personal preferences for risk by combining T


with the risk-free security:
– Those who are very risk averse place a large amount of their
wealth in the risk-free asset and only a little in T – for example
point X on Figure 8.7. These people are lenders.
– Those who are only slightly risk averse may choose to borrow
money at the risk-free rate (i.e. short the risk-free asset) and in-
vest more than their wealth in portfolio T. See point Y on Figure
8.7.

Thus the optimal investment of every mean-variance investor is a combi-


nation of two portfolios (or funds) – T and the risk-free asset. This result
is called two-fund separation.

8.6 Equilibrium

We now know how to define the portfolio of risky assets, and the invest-
ment in a risk-free asset, that is optimal for a single investor.

We will now assume the following:

148
• As before, the existence of N risky assets and a risk-free asset.

• A population of investors, all of whom agree on the expected returns


and return covariances and variances of the risky assets. They also
face the same risk-free rate.

• Frictionless markets. As before, unlimited borrowing and lending at


r f and unlimited long and short positions on risky assets, no taxes
and no transaction costs.

These assumptions imply that all investors agree on the shape of the fron-
tier of risky portfolios and agree on the exact composition of the tangency
portfolio, T. So every investor has an identical optimal portfolio of risky
assets, T. They may differ in how much of their wealth they devote to T
versus the risk-free asset, but every dollar invested in risky assets is in-
vested with the weights dictated by T.

We add to this structure, by adding the notion of equilibrium in the markets


for risky assets. We impose the condition that demand for risky assets
must be equal to the supply of risky assets.

Demand side of market:

All investors hold risky assets in the proportions dictated by the tangency
portfolio – thus, in portfolio terms, the tangency portfolio summarises the
demand for risk assets.

Supply side of market:

The stocks issued by all of the N firms in our world make up the supply
of risky securities. The market portfolio summarises this supply of risky
securities where the market portfolio is the value weighted portfolio of
our N risky assets.

149
Equilibrium

In equilibrium the tangency portfolio must be exactly the market portfolio


of risky assets.

The implications of this are that:

• All investors choose the same optimal portfolio of risky assets and
this is the market portfolio.

• The market portfolio includes all risky securities (in principle it in-
cludes stocks, bonds, real estate and so on).

• The weight of an asset in the market portfolio is its market capital-


ization divided by the total market cap of all assets.

• All securities must be held in equilibrium (i.e. all securities have


positive weight in the portfolio T).

How does this equating of the market and the tangency portfolio work?

Assume, to illustrate, that the result does not hold. Assume that at its
current price, the tangency portfolio weight of BARC is below its market
portfolio weight:

• Then supply of BARC shares is greater than demand.

• The price of BARC would drop:

– weight of BARC in market portfolio will decrease


– weight of BARC in tangency portfolio will increase (investors
buy more as it’s cheaper).

• This process continues until market portfolio weight equals tangency


portfolio weight.

As the argument above makes clear, this model says that the correct price
for a stock is found once its tangency and market portfolio weights are
equalised.

150
8.7 The CAPM

We have seen that every investor holds risky assets according to the mar-
ket portfolio (M) weights and also holds some of the risk-free asset. Given
that we know this, how do we measure the risk of an individual stock?

Thinking back to our previous work on portfolio theory, we know that the
risk of a particular stock is measured by its contribution to the risk of the
portfolio that an investor holds. As all investors hold the same portfolio
(M), then they all agree on the risk of every individual stock. To be precise,
the risk of every stock is measured by its β to the market portfolio. Finally,
as investors require compensation for bearing risk, expected returns on
stocks are increasing in β.

In line with the intuition in the preceding paragraph, we can prove math-
ematically that under our assumptions the relationship between risk and
return is as follows:

 Cov( R j , R M )
E( R j ) = R f + β j E( R M ) − R f , βj = (8.2)
Var( R M )

First let’s focus on how to interpret the β that is on the right-hand side of
this equation. What is the beta of a risk-free asset? It is:

Cov( R f , R M )
βf = =0
Var( R M )

as the risk-free return has zero covariance with the return on any other
asset.

What is the beta of the market portfolio?

Cov( R M , R M ) Var( R M )
βM = = =1
Var( R M ) Var( R M )

Now consider some other asset. Let’s assume its β is between zero and
one. What does this imply?

151
• It means that the asset’s returns are positively correlated with the
market return, so when the market goes up (down), the asset’s price
tends to go up (down).

• But when the market goes up by 1%, we expect the asset price to go
up by less than 1% (in fact we expect it to rise by β%)

• As the β is positive, the asset is risky but because it is less than one it
is not very risky.

On the other hand, an asset with a β greater than one is such that:

• Its returns are positively correlated with the market return, so when
the market goes up (down), the asset’s price tends to go up (down).

• When the market goes up by 1%, we expect the asset price to go up


by more than 1% (in fact we expect it to rise by β%)

• As the β is positive, the asset is risky but because it is greater than


one it is very risky.

Now that’s look at the equation that determines expected returns. The
right way to interpret this equation is as telling us how expected returns
vary across stocks (i.e. across j) due to variations in β across stocks. We
see that:

• First, β j measures asset j’s risk and as equation (8.2) shows us that
assets with higher risk must deliver higher expected returns.

– A zero β asset carries no market risk and earns the risk-free rate.
– An asset with a β of one carries the same amount of risk as the
market and so delivers the same mean return as the market.
– An asset with a β between zero and one delivers an expected
return between the risk-free rate and the market return.
– A portfolio with β above one has a greater expected return than
the market.

• The extra expected return earned for a unit increase in risk is equal to
the difference between the expected return of the stock market and
the risk-free rate.

152
Example: CAPM computations

Assume that the risk-free rate is 3% and the expected return on the market
portfolio is 12%. If an asset has a β of 1.5, what is its expected return?

Via equation (8.2) we have:

E( Ri ) = 0.03 + 1.5 × (0.12 − 003) ⇒ E( Ri ) = 0.165 = 16.5%

What is the expected return on an asset with a β of 0.25? It is:

E( Ri ) = 0.03 + 0.25 × (0.12 − 003) ⇒ E( Ri ) = 0.0525 = 5.25%

The CAPM in pictures

One can illustrate the CAPM using two pictures. The first shows the Capi-
tal Market line (CML), which is our earlier depiction of the investor’s opti-
mal portfolio choice, but now in equilibrium:

Figure 8.9: The Capital Market Line

"#$%!
)'*!

'!

!!!$(!

!!!!!&! ! !!

153
All investors locate somewhere on the CML, but all risky portfolios other
than the market portfolio lie to its right. In equilibrium, every investor
holds the market portfolio and mixes it with borrowing or lending at the
risk-free rate.

The second picture associated with the CAPM plots the Security Market
line (SML). This is a plot of β versus expected return and is just a graphical
representation of equation (8.2).

Figure 8.10: The Security Market Line

"#$%!
)'*!

"#$'%!
'!

!!!$(!

!!!!!&! !!!+! ! !!

In equilibrium, all portfolios/securities lie on the SML. The only reason


why one stock’s expected return should differ from that of another, is if
their βs are different.

Deviations from the security market line

The CAPM tells us the return that a stock should earn based on its risk.
What if we find a stock that does not lie on the SML? Assume, for illus-
tration, that we have identified a stock that lies above the SML (see Figure
8.11):

• That stock has a higher expected return than the CAPM says it should

154
have.

• We can thus think of this stock as generating a positive abnormal re-


turn.

• Another way of thinking about this, is that the stock is currently


underpriced.

• Trading implication: give that stock higher weight in your portfolio


than it has in the market portfolio.

• Then you benefit, as you earn this extra expected return without
bearing the appropriate level of risk.

We call the difference between the return that you expect to earn from the
asset and the expected return implied by the CAPM the stock’s α. In this
case, the α is positive.

Figure 8.11: Security Market Line Deviations

#$%&!
*(+!
#$%"&! "!
!!!""!
!
-./(!#$%"&!

#$%(&!
(!

!!!%)!

!!!!!'! !!!,! ! !"! !!

Why do we call this an asset pricing model?

The CAPM equation determines expected returns, not prices, so why do


we call it an asset pricing model?

155
Expected returns and prices are inter-changeable:

• If, at the current market price of the stock, the expected return is
greater than the CAPM expected return, then the market price is too
low.
• If, at the current market price of the stock, the market expected return
is below the CAPM expected return, then the market price is too
high.

CAPM: further example calculations

A stock has an expected return in excess of the risk-free rate of 6% and a β


of 0.75.

• What is the expected excess return on the market over the risk-free
rate?
• If the risk-free rate is 2%, what is the total return on the market?
• What is the expected return on a stock with a β of 1.3?

Here are answers to the questions above:

• We know that:

  E( R j ) − R f
E( R j ) − R f = β j E( R M ) − R f ⇒ E( R M ) − R f =
βj

So in our case we have:

 0.06
E( R M ) − R f = = 0.08 = 8%
0.75
• If we know that the risk free rate is 2% then:


E( R M ) − R f = 0.08 ⇒ E( R M ) = 0.08 + 0.02 = 0.10 = 10%

156
• A stock with β of 1.3 has an expected return equal to:


E( R j ) = R f + β j E( R M ) − R f = 0.02 + 1.3 × 0.08 = 0.124 = 12.4%

8.7.1 How do we use the CAPM?

The CAPM is used in a number of ways:

• Valuation of stocks:

– Think of a stock as just a claim to a stream of future dividends.


– How should we value the stock? Compute the NPV of the fu-
ture estimated dividend stream.
– What discount rate should we use? Use the CAPM expected
return for a company.
– Why? Because the CAPM tells you what return you should ex-
pect from a company with that level of risk – it gives a risk-
adjusted discount rate.

• Valuation of projects:

– Similar intuition to the above.


– For an equity financed firm, the CAPM expected return is the
return that one should require any project that the firm takes
on to earn.
– Thus use it as the discount rate in a NPV calculation.

• Portfolio selection:

– If the CAPM is an entirely accurate description of the world,


then portfolio selection is straightforward:
∗ Hold the combination of the risk-free asset and the market
portfolio that maximises your personal utility.
– If you feel that the CAPM holds for the majority of stocks, but
is violated for a few:
∗ Start off with the market portfolio:

157
∗ Upweight those few stocks that have positive α.
∗ Downweight those few stocks with negative α.
∗ Combine the resulting risky portfolio with the risk-free as-
set so as to maximise your utility.

8.7.2 Empirical evaluation of the CAPM

What are the key implications of the CAPM for the behaviour of stock
return data? Consider the following version of the model:


E( R j ) − R f = α j + β j E( R M ) − R f

The key empirical implications of this model are as follows:

1. If the CAPM holds, estimated values for α should be zero for all
stocks.

2. Also, under the CAPM, expected returns should be linear in β.

3. The slope of the SML should be approximately the expected return


on the market portfolio minus the risk-free rate.

4. Finally, the only stock characteristic that affects expected returns is


the stock’s β.

Many of the results derived from data are broadly supportive of the CAPM,
but some very important rejections of these implications exist. Some of
these are as follows:

• The SML we see in the data seems to have a slope less than the mean
return on the market portfolio minus the risk-free rate.

• Factors other than β explain expected returns:

– Size: small stocks tend to have higher expected return than big
stocks (holding β constant)

158
– Value: mean returns on stocks with high book-to-market tend
to be larger than those on low book-to-market stocks (holding β
equal)

These latter results erode our confidence in the accuracy of the CAPM.

8.8 Competing models

The empirical failures of the CAPM have led researchers to study alterna-
tive models to the CAPM. Some of you will come across these models in if
you continue to study finance:

• Multi-factor models and the Arbitrage Pricing theory (APT)

• Inter-temporal CAPM (ICAPM)

• Consumption-based asset pricing models (CCAPM)

In current academic work (and in the finance industry), when most re-
searchers wish to estimate expected returns for stocks, they tend to use a
multi-factor model that explains stock returns with the CAPM β, but also
allows expected returns to vary with other firm-level characteristics.

Brief overview of the APT

The key innovation of the APT relative to the CAPM is the inclusion of
multiple sources of risk. The CAPM only has one source of risk (i.e. the
market). A 2 factor APT model looks something like this:

 
E( R j ) = R f + β 1,j E R F1 − R f + β 2,j E R F2 − R f

where E( R F1 − R f ) is the risk premium associated with the first factor,


E( R F2 − R f ) is the premium associated with the second factor, β 1,j mea-
sures stock j’s exposure to factor 1 and β 2,j measures stock j’s exposure to
factor 2.

159
In this model there are two sources of risk, represented by the two fac-
tors (F1 and F2 ) and a stock earns expected returns from its sensitivity to
both of these sources of risk. But what are these risk factors? The factors
might be any things that affect overall economic performance and com-
pany prospects. For example.

• Macroeconomic risks (e.g. GDP growth or unemployment or infla-


tion shocks)
• Market risk measurements (e.g. yield curve slopes or exchange rate
shocks).

The Fama-French three-factor model

A popular model in academic work that builds on the CAPM, is the 3


factor model. It looks like this:


E( R j ) − R f = β M,j E( R M ) − R f + β S,j (E( RSMB )) + β V,j (E( R HML ))

There are three sources of risk and thus three contributions to expected
returns:

• The market: the first term on the right-hand side, just like in the
CAPM.
• Size: β S,j is the stock’s beta to a size portfolio. The size portfolio
(called SMB) is a portfolio that is long small stocks and short large
stocks and has expected return equal to E( RSMB ).
• Value: β V,j is the stock’s beta to a value portfolio. The value portfolio
(called HML) is long stocks with low market value relative to book
value and short stocks with high market value relative to book value
and has expected return equal to E( R HML ).

We know that historically both the HML and SMB portfolios have positive
average returns and so a stock that has a positive beta to either of these
factors will get a positive contribution to its expected returns.

160
This model seems to perform better than the CAPM empirically (with the
value factor being especially important). Why size and value represent
sources of risk that investors care about is not entirely obvious, though.
Also, in very recent times, Fama and French have argued that their 3 factor
model should be replaced by a five-factor model, with similar structure to
that above, but where the factors are the market; size; value; profitability;
investment.

8.9 Summary

This chapter has covered a lot of ground. What have we learned?

• We began with simple portfolio theory, showing how portfolio risk


and return depends on the characteristics of constituent assets.

• We then characterised optimal portfolio for investors, showing that


all should invest in the tangency portfolio of risk assets and mix this
with borrowing and lending at the risk-free rate.

• The CAPM builds on modern portfolio theory and imposes equi-


librium in security markets to deliver a formula that allows one to
compute the return a stock should offer to cover its risk.

• Risk is measured by β, which is in turn related to the covariance


between a stock’s return and the market return.

• The relationship between β and expected returns is linear and posi-


tive.

• Abnormal returns are measured by α. Abnormal returns are differ-


ences between a stock’s return and the return that the CAPM says it
should earn.

• In equilibrium, the CAPM says that all αs should be zero.

• The CAPM shows some empirical failures and this has led researchers
to consider more complex models for expected returns.

161
Reminder of learning outcomes
Having completed this chapter, and the Essential reading and activities,
you should be able to:

• derive the CAPM from first principles.

• explain the CAPM’s assumptions and empirical implications.

• use the CAPM to work out expected returns on stocks or portfolios


of stocks.

• discuss the empirical shortcomings of the CAPM.

162
8.10 Exercises
1. If the beta of Amazon.com is 2.2, the risk-free rate is 5.5% and the
market risk premium is 8%, what is the expected rate of return for
Amazon.com stock?

2. If the required return on Exxon Mobil is 10.5%, the risk-free rate is


4% and the market rate of return is 14%, what value must Exxon’s
beta take?

3. True or False? If a stock is overpriced it plots above the security


market line.

4. Consider the following data for a stock: beta = 0.9; risk-free rate =
4%; market rate of return = 14%; and expected rate of return on the
stock = 15%. Is this stock underpriced, overpriced or fairly priced?

5. You are an investment manager who is running an aggressive UK


equity portfolio worth £12m. The portfolio has a beta of 1.4. You
are concerned that the market risk of this position is too large and
would prefer it if the beta of your portfolio was closer to 1.25. You
can also invest in a FTSE-100 tracker portfolio which has a beta of
exactly 1.0. Explain how you might redistribute the £12m between
the aggressive UK portfolio and the FTSE-100 tracker so as to reduce
risk to your desired level.

6. Assume that the CAPM holds. The risk-free rate is 3%. A portfolio Q
that is a combination of the risk-free asset and the market portfolio
has an expected return of 12% and a return standard deviation of
18%. The expected return on the market portfolio is 15%.

(a) What is the volatility of the market return?


(b) Consider a security X. Its returns have a 0.40 correlation with
the returns on the market portfolio and a standard deviation of
36%. What is its β?
(c) What rate of return would you expect for X?
(d) Comment on the market risk associated with security X.

163
Chapter 9

Efficient markets

Essential reading
Brealey, Myers and Allen (2017), Chapter 13.

Learning outcomes
By the end of this chapter, and having completed the Essential reading and
activities, you should be able to:

• define the notion of market efficiency.

• describe some simple tests for efficiency.

• discuss key pieces of evidence for and against efficiency.

9.1 Introduction

One can phrase the fundamental question posed in this chapter in various
ways:

• Does relevant information become reflected in financial securities in


both a timely and an accurate fashion?

164
• Can one make genuine profits through the trading of financial secu-
rities?

• Are financial securities correctly priced at all times?

• Can professional managers who promise to outperform simple in-


vestment strategies really deliver on such claims?

Given the large number of firms/funds that are devoted to making money
from trading, one might think that it is self-evident that there are profits
to be made from trading and that, therefore, information is not always
reflected quickly in prices of financial securities.

9.2 Expected returns versus profits

But before we approach the questions posed in the Introduction we must


define what we mean by ‘profit’? As our asset pricing studies have told
us, holding any security with a positive expected return (e.g. an asset with
a positive beta in a CAPM world) will deliver gains in the long-run, but is
this really making a profit?

The answer to that question is ‘no’. Securities deliver positive expected


returns due to the risk they carry. This is just like earning a wage for do-
ing a job one doesn’t like — you’re being paid to put up with something
unpleasant.

Thus, positive returns and ‘profits’ are not the same thing! Profit is only
earned if one obtains a return on a security that is greater than one would
expect due to that security’s risk.

9.3 Defining efficiency

The basic idea of market efficiency is that there are no ‘profits’ available
from the trading of financial securities. This is because in an efficient mar-
ket all relevant information is instantly and accurately reflected in prices.

165
Thus, as prices adjust instantly, no piece of information can be used by a
trader to make a profit. All securities are fairly valued at all times.1

The classic formal definition of efficiency is that given by Jensen (1978):

Jensen’s definition of efficiency

A market is said to be informationally efficient with respect to an informa-


tion set Ω, if an investor cannot make economic profits by trading on the
information contained in Ω.

• By economic profits, we mean risk-adjusted returns net of all costs.


• The object Ω comprises all the pieces of information that the investor
is using to try to earn abnormal returns.

As an example of how to think about Ω, if we’re trying to judge whether


the market for trade in HSBC is efficient, the information set might con-
tain current and past prices for HSBC and other banking stocks, and past
earnings too.

Malkiel’s efficiency definition

Malkiel (1992) gives a slightly different definition of efficiency:

• A capital market is said to be efficient if it fully and correctly reflects


all relevant information in determining security prices.
• Formally, the market is said to be efficient with respect to some in-
formation set, Ω, if security prices would be unaffected by revealing
that information to all participants.
• Moreover, efficiency with respect to an information set, Ω, implies
that it is impossible to make economic profits by trading on the basis
of Ω.
1 Strictlyspeaking, securities may not be 100% fairly priced at all times if it is not
possible to exploit the mis-pricing. For example, if a stock is overpriced by 5 pence, but it
costs 10p to complete a trade in that security then the stock is mis-priced but there is no
inefficiency as one cannot exploit the mis-pricing and make a profit.

166
Comparing these definitions

Both definitions require that in an efficient market, a specified set or piece


of information cannot be used to make a trading decision that delivers
positive economic profits. Statistically or econometrically speaking, no
piece of information can be used to forecast abnormal returns. Thus, in an
efficient market there is no way to tell, in advance, which securities will
deliver returns above their required return and which will deliver returns
below required returns.

9.4 Foundations of efficiency

Why should we think that markets might be efficient? What economic


conditions/assumptions might lead to efficiency? Shleifer (2000), argues
that any one of the conditions below should lead to efficiency:

1. Rationality: all investors are rational, receive information in a timely


fashion and all process it quickly. Thus all investors adjust stock val-
uations in the same way when news arrives and prices adjust in-
stantly.
2. Independent deviations from rationality: if not all people are ra-
tional, but optimists and pessimists have equal influence on mar-
kets, then their irrationality will balance out and stocks will be fairly
priced.
3. Dominance of rational, professional investors: if there is a large
group of investors (maybe working in banks and funds) who are ra-
tional and have a great amount of money to invest, then when they
see prices that are ‘wrong’, they trade and their trading (and size)
pushes prices to the ‘right’ level.

9.5 Efficiency: making it an operational concept

Thus far, our efficiency definitions leave certain things unspecified. These
are listed below:

167
1. The information set (Ω): Ω details which pieces of information we
are using to try to forecast returns on securities. Obviously there are
many possible definitions of this set.
2. Definition of risk and the returns to bearing risk: as we indicated
in the first section of this chapter, only if we know the returns that a
portfolio should deliver due to its risk can we discover if that port-
folio generates profits.
3. Trading costs: finally, to be able to say that a portfolio generates
profit we must account for the costs of forming and rebalancing that
portfolio in terms of trading costs (brokerage fees and commissions).

9.5.1 Information sets

Part of the definition of an efficient market was the set of information that
we are defining efficiency relative to. Different choices of information set
will result in one testing different kinds of efficiency. Roberts (1967) came
up with the following classification:

Variety Information set


Weak-form Past prices
Semi-strong form All public information
Strong form All public and private information

Clearly as we go down this list the efficiency concept becomes more strin-
gent. In fact we can see that:

Strong form ⇒ Semi-strong form ⇒ Weak-form

9.5.2 Returns, risks and profits

As indicated by our definitions, to assess whether a stock market is effi-


cient we need to understand whether genuine economic trading profits
can be made. This requires us to compute abnormal returns which, in turn,
requires an asset pricing model.

168
If the CAPM is the relevant asset pricing model then a security’s expected
abnormal return is the CAPM α that we discussed previously:

 
αi = R̄ P − R f − β i × R̄ M − R f

where R̄ P is the mean return on your portfolio and R̄ M is the mean return
on the market.

Thus we use the concept of α to measure the abnormal return from holding
a security and efficiency requires that we cannot identify which securities
are likely to have positive alphas and which have negative alphas (as if
we could do this, buying the securities with positive alphas and shorting
those with negative alphas would result in a portfolio with positive prof-
its).

Figure 9.1 shows stock X, which has delivered a positive α. This stock has
earned a mean return which is greater than that return the CAPM says it
should have earned — it has earned a positive abnormal return.

Figure 9.1: Graphical depiction of a security with positive CAPM α


#$%&!
*(+!
#$%"&! "!
!!!""!
!
-./(!#$%"&!

#$%(&!
(!

!!!%)!

!!!!!'! !!!,! ! !"! !!

Example: abnormal returns

A stock has a CAPM β of 0.6 and due to some research you have carried
out you believe its future return will be 10%. If the risk-free rate is 2% and
the expected return on the market minus the risk-free rate is 20%, do you
expect this stock to earn any abnormal return?

169
First work out the CAPM expected return on the stock. It is:


E( R j ) = R f + β j E( R M ) − R f ⇒ E( R j ) = 0.02 + 0.6 × 0.20 = 14%

As your research implies that the stock will return only 10%, you think the
stock will earn a negative abnormal return equal to -4%. If you are right,
then an investment in this stock does not deliver enough mean return to
cover its risk. You should consider short-selling it.

Problem: the joint hypothesis problem

Conducting a test of efficiency requires us to use an asset pricing model to


calculate abnormal returns. In the preceding example we used the CAPM,
for example. But the model we use could be the wrong one! For exam-
ple, perhaps the CAPM does not reflect the way that expected returns are
determined in the real world.

The immediate implication of this is that all tests of efficiency are a test of
a joint hypothesis containing 2 parts:

• The market under analysis is informationally efficient.

• The researcher knows and uses the correct asset pricing model to
generate abnormal returns.

If we reject the null hypothesis it can be because either (or both) of the parts
of the hypothesis are invalid. This creates uncertainty when interpreting
any test results. For example:

• A test indicates strong positive profits to buying stocks on announce-


ments of dividend increases and selling on announcements of divi-
dend cuts. Profits are computed using the CAPM to adjust for risk.

• One way to interpret this is that it leads us to reject the notion of


efficiency in this market.

170
• A second interpretation is that the profits we computed are not ac-
curate. They are based on the CAPM and the CAPM is not the right
model for expected returns.

Thus, the joint-hypothesis problem clouds our ability to interpret results from
efficiency tests and we need to keep this issue in mind at all times.

Practically, it is the researcher’s responsibility to make sure that if (s)he is


testing efficiency then (s)he has made sure that profits are computed using
state of the art asset pricing models.

9.6 Efficiency and random returns

Regardless of which variety of efficiency we are testing, an implication of


efficiency is as follows:

• If a market is efficient with respect to Ω then future abnormal re-


turns appear to be entirely random from the perspective of an indi-
vidual who has the information set given by Ω.

This interpretation often confuses people. If markets correctly and quickly


absorb all relevant information then it seems odd to assert that returns will
be random.

However, a bit more thought shows the interpretation to be valid: All


we’re really saying is that in an efficient market, one can’t forecast abnor-
mal returns: you can’t tell in advance whether the stock will deliver more
or less than its normal return.

The intuition for this result is that, given efficiency, all information has
already been absorbed into prices. All that can then move prices is the
arrival of new information. New information is by definition unpredictable
(otherwise it would not be information at all). Thus, any price changes
based on this new information will be unpredictable also.

Also, random-ness of future (risk-adjusted) returns does not imply that


there is no relationship between stock prices and fundamental variables

171
like dividends or earnings. Perhaps increased dividend payments lead to
increased stock prices, but in an efficient market stock prices adjust in-
stantly to news of increased dividends such that returns are still impossi-
ble to forecast.

9.6.1 Efficiency and statistics

In order to demonstrate inefficiency, we need to demonstrate that abnor-


mal returns are predictable/forecastable. Several statistical definitions will
help us characterise series that are unpredictable and thus may be useful
in describing abnormal returns in efficient markets. The most important
of these is the Fair game.

Consider a random variable Yt . The variable Yt is a fair game if:

E(Yt+1 |Ωt ) = 0 (9.1)

Given time t information, one’s best guess of the t + 1 value of a fair game
is zero. The statistical condition above will be satisfied by a gamble where
one’s expected profit is exactly equal to zero. For example, a gamble where
I give you £100 if a fair coin lands heads and I take £100 from you if the
coin lands tails is a fair game.

Previously, in a world where the CAPM is valid, we identified the CAPM


α as a stock’s abnormal return. True profit resides in the CAPM α. The
EMH implies that abnormal returns (i.e. α) are a fair game.

Thus, the EMH requires that:

E ( α t +1 | Ω t ) = 0 (9.2)

Equation (9.2) tells us that in an efficient market one’s best guess of abnor-
mal returns on the basis of Ωt is exactly zero — one can’t forecast whether
abnormal returns are likely to be positive or negative. Hence, pure trading
profits on the basis of Ωt are on average zero.

172
9.7 Efficiency and empirical methods

Below, to give some substance to some of the ideas presented above, we


discuss some statistical/econometric methods that some researchers have
used to study certain forms of efficiency.

9.7.1 Autocorrelations

Autocorrelation analysis can be used to test the proposition that future


(abnormal) returns are unrelated to current and past (abnormal) returns.
Consider a sequence returns on a stock or a portfolio, r̃t . These could, for
example, be the sequence of monthly returns on the FTSE-100 portfolio.
The kth autocorrelation of the returns series is defined as:

Cov(r̃t , r̃t−k )
ρ(k) =
Var(r̃t )

This is just the correlation of the series with its own past and we can in-
terpret it in exactly the same way that we would a correlation coefficient.
It is bounded between -1 and +1 and if there is no relationship between r̃t
and r̃t−k it is equal to zero. If the kth autocorrelation is above (below) zero
then it means that above average returns k periods ago lead, on average to
above (below) average returns in the current period.

Thus, if one finds that ρ(k) is reliably different from zero it means that
returns are predictable and this would cast doubt on the validity of the
efficient markets hypothesis. This observation would also suggest a prof-
itable trading rule. If, for example, ρ(1) is large and positive, then if the
portfolio return was positive (negative) in the period that just finished we
would expect it to be positive (negative) in the period just beginning and
thus we should buy (sell) the portfolio.

Note that the EMH requires that autocorrelations for all k are zero.

173
9.7.2 Event studies

Event studies directly look at price (or cumulative abnormal return) reac-
tions to specific types of corporate (or economic) event. For example, you
might wonder whether the market reacts to announcements of dividend
changes in a timely and accurate fashion and you want to test this. We
all agree that when companies announce dividends increases their stock
prices tend to rise (and vice versa), but is this rise completed so quickly
that it is essentially impossible for one to make money from trading on the
news?

To perform an event study, you would do something like the following:

• Collect a large set of announcements of dividend changes from a


large set of firms.

– Categorise them into dividend increases and dividend cuts. As-


sume that there are NI increases and NC cuts.

• For each announcement, collect abnormal returns for that firm in a


specified window around the announcement e.g. from 10 days be-
fore to 90 days afterwards (so 101 days, including the day of the an-
nouncement, denoted -10 to +90).2

• Take the NI dividend increase events. For each one you have abnor-
mal return data for the same sized window around the event. Av-
erage across announcements to give the mean abnormal return on
each day in the window (so that, in our example, you end up with
an abnormal return for each of days -10 to +90). Do the same for the
NC dividend cuts.

• Now study the averages.

– You would expect to see that all of the action in the abnormal
return is on the announcement day itself, so you would expect
the announcement day return to be big and positive for the div-
idend increases and big and negative for the cuts.
2 Usually you’ll have to compute abnormal returns from the actual returns on each
stock and an assumed asset pricing model, e.g. the CAPM, which you will have had to
estimate.

174
– But is there any evidence that in the days just after the announce-
ment, e.g. days +1 to +5, abnormal returns show some system-
atic pattern? If so, there’s evidence of inefficiency. For example,
if subsequent to dividend cuts prices continue to fall after the
event day itself, then the average abnormal returns will likely
be negative on days +1 and subsequently. This suggests ineffi-
ciency and a trading rule!3

Academic event studies generally follow the steps laid out above, with
the addition of some fancy statistical tests. But the logic is simple. Look
for patterns in abnormal returns after a specific type of event to see if the
market reacts efficiently to the release of that kind of information.

9.8 Puzzles and anomalies from an efficient mar-


kets perspective

In this final section we briefly describe some of the more important viola-
tions of efficiency that have been uncovered in stock (and bond and FX)
market data.

A word of warning though. The results described below are notewor-


thy precisely because consistent, identifiable violations of efficiency are so
rare. The vast majority of studies of efficiency conclude that markets are
efficient and thus it is very hard to consistently profit from security mis-
pricing.

Post-earnings announcement drift

Using event study techniques, researchers have found that markets don’t
react perfectly efficiently to the news in earnings announcements. Portfo-
lios of stocks that have released unexpectedly good earnings news deliver
much greater average abnormal returns than portfolios of stocks that have
released unexpectedly bad earnings. This outperformance persists for sev-
eral months after earnings release dates.
3 On the day of an announced dividend cut, sell the stock, and you’ll benefit (on aver-
age) if abnormal returns are negative in the few days after the cut.

175
This result is robust across markets and across time. It is also especially
perplexing as earnings announcements are incredibly important events in
stock markets and are scrutinized very heavily by analysts.

New issue puzzle

Companies that issue shares to the public for the first time (through an
IPO) or issue more, new shares to the public (through a seasoned equity
offering) underperform firms who do not issue shares for a period of up
to five years after the date of the issue.

Momentum

A strategy of buying stocks that have delivered positive returns in the 12


months up to the present and selling stocks that have delivered negative
returns in the year to the present day delivers, on average, significant posi-
tive returns. These returns seem to be present across time, across countries
and across asset classes (e.g. the strategy works in FX markets too).

Thus historical winners continue to significantly outperform historical losers!


Again, this result is perplexing as it is so widespread and so simple to ex-
ploit.

9.9 Summary

We have studied the underpinnings of the notion of market efficiency. We


have seen that:

• Efficiency implies that active trading is a fruitless pastime as no-one


can beat the market.

• This is because efficiency requires all information to be quickly and


accurately reflected in prices.

• We can define different varieties of efficiency (weak-form, semi-strong

176
form and strong form) based on the information being used to fore-
cast prices.

• Testing/understanding efficiency requires a researcher to understand


how expected returns are generated i.e. it requires researchers to
specify and have confidence in an asset pricing model.

• In a CAPM world, efficiency requires that CAPM αs are on average


zero and are not forecastable.

177
Reminder of learning outcomes
Having completed this chapter, and the Essential reading and activities,
you should be able to:

• define the notion of market efficiency.

• describe some simple tests for efficiency.

• discuss key pieces of evidence for and against efficiency.

178
9.10 Exercises
1. Is the following statement true or false? Justify your answer with a
short explanatory paragraph.

“In an efficient market changes in stock prices are unpredictable


and, therefore, stock prices cannot be related to fundamental in-
formation on company performance.”

2. Is the following statement true or false? Justify your answer with a


short explanatory paragraph.

“In an efficient stock market, no amount of fundamental or tech-


nical analysis will help an investor to achieve greater expected
future returns. The only way that an investor can expect to make
greater returns is by taking more risk.”

3. Empirical research suggests that company directors can made abnor-


mal trading profits by trading in their own companies’ stock. Does
this evidence suggest that markets are inefficient? If so, does the ev-
idence contravene the weak, semi-strong or strong form?

4. Recent work suggests that portfolios of companies with high operat-


ing profitability deliver significantly larger returns that portfolios of
companies with low profitability. Does this indicate inefficiency in
stock markets? Justify your answer.

5. An investment advisor approaches you with an opportunity to in-


vest in a product that has clearly outperformed the market portfolio
in recent years. Is it obvious that this product is delivering economic
profits and thus that you should invest in it?

179
Chapter 10

Forward and futures markets

Essential reading
Brealey, Myers and Allen (2017), Chapter 26 section 4 and Chapter 27 sec-
tions 1 and 2.

Learning outcomes
By the end of this chapter, and having completed the Essential reading and
activities, you should be able to:

• describe the payoff structures of forwards and futures.

• describe the differences between forward and futures contracts.

• compute no-arbitrage forward prices.

10.1 Introduction

In this chapter we will begin our analysis of derivative securities, their


using and how they are priced. We will begin with forward contracts.
Forwards are amongst the oldest financial contracts and are one of the
earliest known forms of derivative contract. In this chapter we will discuss
uses of and prices for forward and futures contracts.

180
10.2 Derivatives

Let’s start by defining some basic terminology and concepts. We start with
a definition:

Derivative assets

Definition: a derivative security is a security for which the payoff is gov-


erned entirely by the value of one or more underlying assets.

Given the above, it’s obvious that these assets are called derivatives as
their payoffs are derived from the prices of other assets. However, other
than specifying this feature, the definition gives little useful information.

To make things more concrete we need to specify the underlying asset and
to be able to deduce how the derivative payoff relates to the underlying
price.

Below is a list of some of the most common, also known as ‘vanilla’, deriva-
tives. These derivatives differ in terms of the payoff structures they promise.
In our later analysis we will focus on the members of this list. Beside each
item in the list is a brief description of its payoff profile:

• Forward and futures contracts: the obligation to buy/sell the under-


lying asset at a pre-specified price and on a pre-specified date.

• Call options: the right (but not obligation) to buy the underlying
asset for a pre-specified price and on (or before) a pre-specified date.

• Put options: the right (but not obligation) to sell the underlying asset
for a pre-specified price and on (or before) a pre-specified date.

• Swaps: the exchange of cashflows for a pre-specified period of time.


The cashflows are determined by some pre-defined rule and are based
on the value of some underlying assets or on observed market inter-
est rates.

More complicated, less common derivative securities are also available.


These are sometimes called exotic derivative contracts.

181
The underlying asset is the reference asset that, via its price, determines the
derivative’s payoff. Amongst other things, one can create derivatives on:

• Equities: single stocks (e.g. BP or MSFT) or equity indices (e.g. the


FTSE-100, S&P-500, Nikkei 225).

• Currencies: exchange rates such as USD/EUR or JPY/GBP.

• Bonds and Interest rates: for example, LIBOR, Bunds and Gilts.

• Commodities: such as Gold, Oil and Coffee.

Effectively, if one can observe the price of a security or the value of a par-
ticular indicator then one can write a derivative on it. These days we can
trade things as interesting as weather derivatives and energy derivatives.

10.2.1 Uses for derivatives

It is useful at this point to ask exactly what financial market participants


use derivatives for. Derivatives are used for a number of reasons with the
most common below:

• Hedging: many market participants, for example corporate treasur-


ers in oil firms or exporting firms, use derivatives to remove or re-
duce the risks associated with their economic activities or investment
portfolios.

• Making bets (or speculating): one can use derivatives to deliber-


ately gain exposure to certain risks e.g. if one has a view on the
future value of the FTSE-100, one could exploit this using FTSE fu-
tures.

• Create arbitrage portfolios: as derivatives are based on the prices of


underlying assets, one can sometimes construct portfolios of deriva-
tives and the underlying that yield risk-free arbitrage profits. This is
only possible if the derivative is mis-priced relative to the underlying
asset.

182
10.3 Forwards and futures contracts

Forwards are amongst the oldest derivative contracts, having been traded
(particularly in the agricultural sector) for hundreds of years. Futures are
really just slightly more standardized forward contracts.

Forward contract: definition

A forward contract is an agreement to buy/sell a certain quantity of an


asset for a pre-specified price (called the forward price or the delivery
price) at a particular future date.

Forwards tend to have the following characteristics.

• First, they are bilaterally agreed contracts and are not exchange traded.
Thus they are often described as ‘over the counter’ or OTC.

• Second, at inception the forward price (F), i.e. the price at which the
exchange is to occur in the future, is set such that the value of the
contract is zero and, thus, no money changes hands on the inception
date.

The individual contracted to purchase the asset is said to have a long posi-
tion in the forward while the individual selling the asset has a short position.

Long forward payoff

The individual holding the long-side has agreed to buy the specified asset
at price F at time T. Clearly, if the market price for the asset at T exceeds
the price he has promised to pay he has made a profit equal to the differ-
ence between the market price and the forward price. If at T, the market
price is lower than the forward price then the long side makes a loss equal
to the difference between forward price and market price.

In either case, the net gain, or payoff, to the long-side is given by:

183
ST − F

where ST is the market price of the underlying asset at the maturity date,
T.
3 Forwards and futures contracts 10

Long
Figure 10.1:Forward Payo↵
Long forward payoff

6 Profit

F -

ST

Short forward payoff


Derivatives: instruments and pricing Asset Pricing: lectures 9 and 10

The payoff to the short-side of the contract is obviously the negative of the
payoff of the long-side. If the long-side makes a profit of X, the individual
who is short must make a loss of X. Thus the payoff to the short-side of
the contract is given by:

F − ST

Example: forward contract

A corn farmer might arrange the following contract with a buyer.

184
3 Forwards and futures contracts 11

Short
Figure 10.2:Forward Payo↵
Short forward payoff

6 Profit
@
@
@
@
@
@
@
@
@
@ F ST
@ -
@
@
@
@
@
@
@
@
@
@
@@

She agrees to supply 5000 bushels of corn, six months from today (i.e. at
the date of harvest), at a price of $4.00 per bushel (i.e. for a total consider-
Derivatives: instruments and pricing Asset Pricing: lectures 9 and 10
ation of $20,000).

On the delivery date, if the market price of corn was to turn out to be:

• $3.00. The farmer makes a gain, relative to the market price, of $1.00
per bushel (as she is selling at $4.00 a bushel and the market price is
$3.00), or $5,000 in total.

• $4.75. The farmer makes a loss relative to the market price of $0.75
per bushel, or $3,750 in total.

But regardless of what happens in the market, the farmer’s revenue is


given. She will receive $20,000. She has eliminated any uncertainty about
the cashflow from her corn harvest and this may be useful in that it allows
her to plan her future investment better.

185
10.4 Futures contracts

Forwards and futures are very similar assets in terms of payoff. In both
cases, the total payoff to the long side is equal to the difference between
the underlying price at the delivery date and the agreed delivery price.
However they differ in a couple of key areas:

• Trading location:
– Forwards are bilateral agreements made between two parties.
– Futures are exchange traded, standardised contracts.
• Exchange of monies:
– Forwards: money changes hands only on the agreed delivery
date.
– Futures: money changes hands throughout the contract lifetime
via a process called marking to market.

Thus, while the total payoffs to long and short sides of a futures contract
are exactly as for the forward, money is exchanged over the lifetime of the
contract rather than in one lump on the delivery date.

Marking to market makes use of the following tools:

• Margin account: an account holding monies deposited by the long


party in the futures contract.
• Initial margin: the initial amount the buyer must deposit in the mar-
gin account when the contract is opened.
• Variation: as the market price of the futures contract changes the
balance in the margin account is altered accordingly. If the price of
the future rises, money is taken from the margin account of the short
party and deposited into the margin account of the long party (and
vice versa if the price of the future falls).
• Maintenance margin: a threshold value for the balance of the mar-
gin account. If the balance in the account falls below this value, the
buyer must top up the balance to the initial margin.

186
The margin account system and marking to market are overseen by the
exchange upon which the future is traded.

10.4.1 Marking to market: example

Consider a US investor who wishes to speculate. He believes that the price


of gold will rise over the coming weeks. Rather than buying gold itself
(and thus having to deal with storage and security), he takes a long posi-
tion in a gold futures contract with the following specifications:

• contract size: 100oz

• futures price: $1,300 per oz

• maturity: five weeks to maturity

• initial margin: $5,000

• maintenance margin: $2,500.

Below we give a hypothetical path for the futures price over the contract
lifetime and show how the margin account operates for our US speculator:

Week Futures Weekly Cumulative Margin Margin Balance


Price Change Change Balance Call after call
0 1, 300.00 5, 000 0 5, 000
1 1, 280.00 −2, 000 −2, 000 3, 000 0 3, 000
2 1, 270, 00 −1, 000 −3, 000 2, 000 3, 000 5, 000
3 1, 265.00 −500 −3, 500 4, 500 0 4, 500
4 1, 230.00 −3, 500 −7, 000 1, 000 4, 000 5, 000
5 1, 250.00 +2, 000 −5, 000 7, 000 0 7, 000

• At the opening of the position, the futures price is $1,300 per oz and
the speculator is required to deposit $5,000 in a margin account.

• In one week the price of the future has dropped to $1,280. Given the
100oz contract size, this makes a total loss of $2,000.

187
– The $2,000 loss is deducted from the speculator’s margin ac-
count and transferred to the seller’s margin account.

• The following week, the market price of the future drops to $1,270.

– The total loss associated with this price fall is another $1,000
which is also deducted from the balance in the margin account
and transferred to the account of the counter-party.
– The margin account balance is now $2,000.
– The balance in the margin account is below the maintenance
margin of $2,500 – thus the investor is required to top up the bal-
ance in the margin account to the initial margin level of $5,000.

• The rest of the example continues in similar fashion.

If we proceed to total up the gains and losses to the speculator over the
life of the contract. He gains $2,000 (= $7,000 - $5,000) on the balance in
the margin account but he’s lost $7,000 (= $3,000 + $4,000) in margin calls.
Finally, at maturity he pays the final futures price of $1,250 per oz. Thus in
total he pays (100 × $1, 250) + $7, 000 − $2, 000 = $130, 000. Note that this
is exactly the same as the cashflow that would have been paid in a forward
contract (as with the forward one would just pay the initial forward price,
of $1,300 per oz, multiplied by 100oz). This is no accident, it will always
be so regardless of how the futures prices evolves.

The key advantage of marking to market is that, by spreading the $130,000


payment over time, the problems associated with one of the counter-parties
defaulting on their obligations is reduced. In our example, if our spec-
ulator was to go bankrupt on the day before the delivery date then the
counter-party would still have received most of the profits that (s)he has
earned through the margin account. In the case of a simple forward con-
tract, bankruptcy might result in the seller not receiving a penny.

10.5 Pricing derivatives

Now that we know how some derivative assets are structured, we can
start to think about how we might price them. Throughout this section we
will use the same underlying pricing principle - called absence of arbitrage

188
pricing - for all of our examples. We briefly review the mechanics of this
method of pricing in this section.

First we must define what we mean by arbitrage.

Arbitrage portfolios

An arbitrage portfolio is said to exist when one can construct one of the
following:

• A portfolio with zero set-up cost (i.e. a zero price) but a chance of
positive subsequent payoffs (and no chance of a negative payoff).

• A portfolio with a negative set-up cost and zero payoffs thereafter.

Obviously both of these portfolios seem too good to be true. The first
promises positive future cashflow for no upfront cost and the second de-
livers cash today and no risk in the future. In both cases, there is no down-
side risk and any wealth-loving investor should invest heavily in them.

A simple example of an arbitrage is a stock trading on two difference ex-


changes at two different prices. Assume, for example, that stock X is trad-
ing at £5 per share on exchange A and £4.50 per share on exchange B.
Assume, for simplicity, that the cost of trading stock on either exchange is
zero. Then an arbitrage exists. By buying X on exchange B and immedi-
ately selling it on A you make £0.50 and you bear no risk. You should try
to exploit this opportunity with as much cash as you can muster.

Given that these arbitrage portfolios are too good to be true, the way we’re
going to price derivative securities is as follows.

• We assume that investors are smart (i.e. greedy) enough to see any
arbitrage opportunities and take them.

• These smart investors, or arbitrageurs as they are sometimes known,


will be buying cheap assets and selling expensive assets in their ex-
ploitation of the arbitrage opportunity.1
1 Inthe example above, we buy stock X on the exchange where it is cheap and sell it
on the exchange where it is expensive.

189
• This will tend to raise the prices of the cheaper assets and reduce
those of the expensive securities and thus reduce the scope for arbi-
trage.

• Only when the arbitrage opportunity has completely disappeared


will the asset prices have no tendency to move upwards or down-
wards.2

• Thus, we focus on these no-arbitrage situations to price our deriva-


tive assets.

So we price assets by assuming that no arbitrage opportunities can exist. This


pricing rule means that no risk-less profits are available and leads to two
important implications.

• The law of one price: if two portfolios have identical payoffs in all
states of nature, they must have the same price.

• The law of payoff dominance: if portfolio A guarantees a payoff at


least as great as portfolio B in all states of nature, then portfolio A
must command a greater price than portfolio B.

If, for example, the law of one price was violated, that would mean that
two portfolios existed that always promised the same payoff but which
had different prices. If one was then to buy the cheaper portfolio and
sell the more expensive one, then one would earn money today and be
guaranteed a zero future payoff. This is a risk-less profit and thus violates
absence of arbitrage.

Similarly, if the law of payoff dominance was violated than one could find
two portfolios with the same price but where one always delivered weakly
greater payoffs than the other. Shorting the low payoff portfolio and buy-
ing the high payoff portfolio would cost nothing today (as their prices are
the same) but would deliver positive cashflows in the future. Again, this
is a risk-less profit and thus violates absence of arbitrage.
2 Inthe example above, the arbitrage disappears when the prices of the stock on the
two exchanges are the same.

190
10.6 Forwards and futures prices

We will ignore the differences in the timing of cashflows between futures


and forwards and treat forwards and futures as being identical.

10.6.1 Pricing forwards on assets that pay no cashflows

Consider a T-period forward contract on an asset that pays no cashflows


to an investor between the current date and T. The delivery price agreed
in the forward contract is F. and the current price of the underlying is S.
We denote by r T the T-period spot rate.

Finally, denote the underlying price at maturity with ST . This price is


obviously unknown when the contract is written.

What is the no-arbitrage delivery price of the forward contract? Well, we


know that for a forward contract no cashflows are exchanged at the date
on which the contract is struck. To rule out arbitrage we manipulate the
delivery price of the contract (F).

First we construct a portfolio that replicates the payoff from being long the
forward, assuming that T-year zero-coupon bonds exist with face value
of £1. We buy one unit of stock and sell F T-year zero-coupon bonds.
The costs and payoffs of the forward and the replicating portfolio are as
follows:

Forward Replicating p/f


Cost at t0 0 S − (1+Fr )T
T
Payoff at t0 + T ST − F ST − F

By design, the replicating portfolio’s payoff is always the same as the for-
ward contract’s. However, the price of entering into the forward is zero.
Thus, by the law of one price, the cost of the replicating portfolio must be
zero. Hence:

F
S− =0
(1 + r T ) T

191
This nails down the no-arbitrage value for the delivery price. The no-
arbitrage forward price on an asset that pays no dividends or coupons is:

F = S (1 + r T ) T (10.1)

If the forward price was any different from this value an arbitrage oppor-
tunity would be available.

10.6.2 Forward pricing example

Assume that the current price of MSFT is $90. This stock never pays div-
idends. The current two-year spot rate is 2.5%. What is the no-arbitrage
price of a two-year forward on MSFT stock?

F = 90 (1 + 0.025)2 = $94.556

Here is a follow up question. If the current market price of a two-year


forward on MSFT stock is $98. How might one exploit this?

• Well, the market price of the forward is too high.

• So enter a contract to sell MSFT forward two years from now for $98.

• Simultaneously, borrow $90 for two years. Use this to buy MSFT
stock today.

• In two years, deliver the stock you bought today to the forward
counter-party in return for $98 and repay the loan plus interest i.e.
$90 (1 + 0.025)2 = $94.556.

• You’ve made a (guaranteed) profit of $98 − $94.556 = $3.444.

10.6.3 Forward prices for dividend paying assets

We can extend our analysis to deal with forwards on assets that will pay a
known dividend during the forward’s lifetime.

192
Retain the same notation as in the previous section, but additionally, de-
note the present value of the dividend that the asset pays by I. Then the no-
arbitrage T-period forward price on the dividend paying asset is:

F = ( S − I ) (1 + r T ) T (10.2)

This can be easily generalised. When there are multiple dividend pay-
ments, I is then the present value of the entire set of dividend payments
When dividend payments are uncertain, I should be interpreted as the
present value of the expected dividend payments.

10.6.4 Commodity forwards

Pricing forwards on commodities differs from pricing forwards on finan-


cial assets as most commodities have a value in production or consump-
tion that financial securities do not have. For example, coffee beans can be
processed and consumed, oil can be refined and burned and copper can be
used as a component in the construction of many goods. Financial assets
have no alternative use. They cannot be used in production processes or
directly consumed.

This alternative value of commodities leads to what financial professionals


call a convenience yield. This is the value that someone obtains from holding
a physical inventory of a commodity due to its value in either production
or consumption (e.g. an airline might prefer to keep a store of jet fuel to
guard against disruptions to its service caused by future jet fuel shortages).
The convenience yield can vary across time and across commodities (e.g.
yields might move around harvest time or might be different in summer
versus winter).

A second difference between financial assets and commodities it that it


is expensive to store commodities (e.g. oil or wheat). It costs virtually
nothing to hold an inventory of a financial asset like a stock, on the other
hand. Commodities have significant storage costs.

Taking these two factors into account, denoting the single period storage
cost by c and the single period convenience yield by CY, we can write the
no-arbitrage T period forward price for a commodity as:

193
F = S (1 + r T − (CY − c)) T (10.3)

10.7 Summary

We have introduced forward and futures contracts here. We have de-


scribed their uses and payoff structures and introduced the technique of
pricing via absence of arbitrage. We started by pricing forwards on assets
with no income, then moved on to pricing forwards on assets with income
and finally we priced forwards on commodities. We also covered the basic
operation of the process called marking to market for futures contracts.

194
Reminder of learning outcomes
Having completed this chapter, and the Essential reading and activities,
you should be able to:

• describe the payoff structures of forwards and futures.

• describe the differences between forward and futures contracts.

• compute no-arbitrage forward prices.

195
10.8 Exercises
1. An economy contains two risky assets. Both of them cost $1. Asset A
will pay the holder $3 if it rains tomorrow and $0.50 if it does not rain
tomorrow. Asset B will pay the holder $3 if it rains tomorrow and
$0.75 if it does not rain. Is there an arbitrage portfolio available in
this economy (assuming that you can go long or short both assets)?
If so, how is it constructed? If you think that there is an arbitrage
available what would you expect to happen to the prices of the two
securities to eliminate the arbitrage?

2. Another economy also contains two risky assets. Asset X will pay
off $1 in one year if between now and then UK GDP grows and zero
otherwise. Asset Y will pay nothing in one year if UK GDP grows
over the next 12 months and $1 otherwise. The price of asset X is
$0.75 and the price of asset Y is $0.20.

(a) What is the no-arbitrage price of a new risky asset that promises
to pay off $5 one year from now if the UK grows over the next
12 months and $3 otherwise?
(b) What is the no-arbitrage price today of a risk-free bond that
promises to pay $1 in one year’s time (regardless of what UK
GDP does)?
(c) Thus, what is the implied annual risk-free rate in this economy?

3. A non-dividend paying stock has a current market price of £45. What


is the no-arbitrage delivery price of a two-year forward contract on
this stock if the effective annual interest rate is 5%?

4. A stock has current price S. One period from now the stock will pay
a known dividend of D. The one-period risk-free rate is r. Derive an
expression for the no-arbitrage delivery price of a 1 period forward
contract on this stock. Assume that the forward contract matures
just after the payment of the dividend. Also, note that the long party
in the forward contract will not receive the dividend as he does not
hold the stock at the time the dividend is paid.

5. The current spot exchange rate allows one to exchange £1 for JPY
140. The 10-year UK spot interest rate is 1.5% and the 10-year spot
rate in Japan is 0.1%. What is the no arbitrage value of the 10-year
forward exchange rate of Sterling for Yen?

196
Chapter 11

Options

Essential reading
Brealey, Myers and Allen (2017), Chapters 20 and 21.

Learning outcomes
By the end of this chapter, and having completed the Essential reading and
activities, you should be able to:

• describe the key payoff features of put and call options.


• compute payoffs for option portfolios.
• provide some no-arbitrage bounds on option prices.
• price options via no-arbitrage in a binomial setting.
• discuss the key features of the Black-Scholes call option pricing model
and use the model to price options.

11.1 Introduction

Options are among the most liquid financial instruments. In order to un-
derstand their usefulness and features let’s start with the definition of a

197
standard option contract.

Option contract: definition

An option gives its buyer the right, but not the obligation, to buy or sell
a given quantity of a specified asset at (or before) some particular future
date for a pre-determined price, known as the strike price or exercise price.

It is worth contrasting options with forwards at this point. The option


gives the holder a choice. It does not force the holder to buy, for example.
Forwards, on the other hand, require the counter-parties to buy (or sell)
once the contract has been entered into, otherwise the contract has been
broken.

Fundamentally, choice is a valuable thing. As such, we should expect (and


we will see that) options cost money to enter into. The cost of entering an
option position is called the premium

An individual who has purchased an option contract is said to be long the


option and an individual who has sold an option contract is said to have
written the option. The long party pays the premium to the writer of the
option at the date the contract is entered into.

From the definition above, we have five fundamental things to specify in


the option contract terms. These are:

• The underlying asset


• The quantity of the asset to be bought or sold
• The maturity date
• The strike price
• Whether the option holder can buy or sell.

Calls and puts

We can fundamentally distinguish two types of option according to whether


they give the right to buy or sell the underlying asset:

198
• Call option: gives the owner the right to buy the underlying asset.

• Put option: gives the owner the right to sell the underlying asset.

If one pays a premium and purchases a call option on stock X, one has
bought the right to buy stock X at the maturity date for the strike price.
The writer of this option receives the premium when the contract is en-
tered into and, if the long side wishes to buy the stock on the maturity
date, must supply to stock to him (or her) in return for the strike price.

Similarly, if one purchases a put option on stock Y today, one has bought
the right to sell the stock to the writer on the maturity date for the strike
price. The writer receives the premium today and, if the long side wants
to exercise on the maturity date, must purchase the stock Y from her for
the strike price.

European versus American

Another fundamental distinction can be made on the basis of times at


which the holder can exercise his right to buy/sell:

• European option: the owner can buy/sell on the maturity date only.

• American option: the holder can choose to buy/sell at any date up


to and including the maturity date.

As choice is a valuable thing and American options are more flexible than
European ones, then one would expect the premium on an American op-
tion to be at least as large as the premium on a European option (assuming
that the two options are identical in all other respects).

Option contract example

On 29/03/2018 the following contract was being traded on Eurex.

A June 18 call option on Vodafone, with strike price 200 pence, and giv-
ing rights to 1000 shares, was trading for 4.50 pence. At the same time,

199
the underlying price of Vodafone shares was 194.22 pence. The call was
American in nature.

Thus if one was to spend £45.00 (= 0.045 × 1000):

• You’d have the right to buy 1000 shares of Vodafone.

• You could exercise on any business day up to the maturity date (which
was the third Friday in June 2018).

• If exercised you’d pay 2.00 × 1000 = 2, 000 for 1000 Vodafone shares.

11.2 Option payoffs

Consider a call option with strike price K and where at the current time
the price of the underlying is S. We have the following terminology:

• The call is in the money if the underlying price exceeds the strike
price (i.e. S − K > 0). Then the option could be immediately exer-
cised at a profit.

• The call is at the money if the underlying price is the same as the
spot price (i.e. S − K = 0).

• The call is out of the money if the strike price is above the underlying
price (i.e. S − K < 0) as then the holder can make no money through
immediate exercise and would thus choose not to exercise.

Based on the above we can characterise the payoff, at the maturity date, to
the investor with a long European call option position.

• When the underlying price (S) exceeds the exercise price (K), this
investor will choose to exercise his right to buy the asset and will
make a gain on exercise of S − K.

• However, when the current underlying price is below the exercise


price the individual will choose not to exercise (as it would be cheaper
to buy the asset in the open market) such that his gain is zero.

200
Putting these two arguments together, the payoff to the long position is:

Long call payoff = max[S − K, 0]

The key feature of the preceding payoff is its non-linearity. The fact that
the holder of the option is not obliged to purchase the asset means that
he only does so when it is profitable and this creates a kink in the payoff
profile.

However, as yet, we have accounted for the option premium. Unlike when
entering a forward contract, the long side of the option contract is required
to pay a premium to the individual writing the contract when the agree-
ment is made. Why is this? Well, as we have seen above, an individual
who is long an option never receives a negative payoff at expiry and some-
times gets a positive payoff. Thus the individual who is long the option
has a position that never costs money at maturity and for such a position
one must pay a positive price up front. Conversely, the individual who
has written the option has a payoff that is at best zero and sometimes neg-
ative at the expiry date. To induce someone to write an option, therefore,
one must compensate them up front with a premium.

Combining the payoff and premium information, we define the total profit
to the long side as the payoff in the equation above less the option pre-
mium.1 If the call premium is c then the profit is:

Long call profit = max[S − K, 0] − c

The writer of the option has a profit that is equal to the mirror image of the
long-side. He receives the option premium when the contract is struck but,
when the underlying price exceeds the strike price, his payoff is eroded
by the difference between the underlying price and the strike price as his
counter-party exercises the option. Thus the profit to the writer is c −
max[S − K, 0]
1 It
is worth noting that the option premium is paid at the entry of the option position,
while the payoff is only received later when the option is exercised. This in our profit
calculation, the two component cashflows actually occur at different dates.

201
11.2.1 Long put payoff

Similar arguments to those above tell us that the payoff from holding a
long position in a put option is given by:

Long put payoff = max[K − S, 0]

as when the spot price exceeds the strike price the investor chooses not to
exercise his right to sell and when the spot price is below the strike price
the option is exercised. This means that the definition of in the money
(and the other money-ness definitions) for a put option are the other way
round relative to their definitions for a call option. For example, a put
option is in the money if the current underlying price is below the strike
price (as then one could exercise and sell the underlying for K when it is
only worth S).

Again, one must deduct the put option premium from this payoff to get
the total profit to the long-side. So the profit to the long put position is:

Long put profit = max[K − S, 0] − p

where p is the put premium.

The payoff to the writer is once again the negative of profit to the long
side.

11.2.2 Payoff properties

We see that:

• In either put or call case, the long-side of the option has limited
downside risk as she can choose not to exercise.

– She loses at most the premium she paid for the option.

• A long position in a call has unlimited upside potential, however.

202
• On the flip-side, an individual who has written a call has the possi-
bility of unbounded losses.
5 Options 26

Long
Figure 11.1:Call
LongProfit
Call Profit

6 Profit

K -

ST

5 Options 27

Short
Figure 11.2: call
ShortProfit
Call Profit
Derivatives: instruments and pricing Asset Pricing: lectures 9 and 10

6 Profit

@
ST
-
@
@
K @
@
@
@
@
@
@
@@

203
Derivatives: instruments and pricing Asset Pricing: lectures 9 and 10
5 Options 28

Long
Figure 11.3:Put
LongProfit
Put Profit

6 Profit
@
@
@
@
@
@
@
@
@ ST
@ K -
@
@

5 Options 29

Short
Figure 11.4:Put
ShortProfit
Put Profit
Derivatives: instruments and pricing Asset Pricing: lectures 9 and 10

6 Profit

K ST

Derivatives: instruments and pricing Asset Pricing: lectures 9 and 10

204
11.2.3 Options, leverage and risk

Consider the case of an individual who is long a call option.

• Given that his payoff is bounded below at the premium he has paid
for the option it would seem that this position is not very risky.
• If he was long the underlying instead, the value of this investment
could go to zero.

Based on the observation that the most you can lose from buying an op-
tion is the premium, some argue that they are not very risky. Does this
argument that the option position is not very risky stand up? The answer
is no as the example below should demonstrate.

The stock of Corporation A currently trades at £48 per share. A call option
with strike price £40 is also trading and currently each option costs £8. The
following two portfolios both cost £48:

• Portfolio A: a long position in 1 unit of stock


• Portfolio B: the purchase of 6 calls with strike price £40.

Figure 11.5 below shows returns from the two portfolios:

• The y-axis measures the return on each portfolio


• The x-axis is the price of the underlying at exercise.
• This figure clearly demonstrates the greatly increased risk of the op-
tion portfolio over the stock portfolio.
– For any underlying price at maturity below £30, the option port-
folio loses its entire value while the stock portfolio always re-
tains some value (aside from in the unlikely case that the stock
price falls to exactly zero).
– On the upside, the option portfolio returns reach great levels
for relatively small up moves of the stock price (e.g. at £40 the
option portfolio makes a 100% return). For similarly sized stock
moves, the return on the stock portfolio is much smaller.

205
The increased risk of the option portfolio over the stock portfolio is due to
leverage, a concept we will return to when pricing options.

Figure 11.5: Option versus stock portfolio risk

700
Long Stock
Long Options
600

500

400
Portfolio Return

300

200

100

-100

0 10 20 30 40 50 60 70 80 90 100
Underlying price at maturity

11.3 Option combinations: payoffs and uses

Because of the non-linearity in their payoff profiles (at maturity), options


can be combined to yield some very interesting payoff structures and some
useful portfolios. These portfolios can be built to exploit rising or falling
markets, limit upside and downside exposure and create exposure to volatil-
ity (i.e. big moves in the price of the underlying in either direction).

We’ll briefly look at the following combined option positions:

• Bull and bear spreads


• Butterfly spreads
• Straddles and strangles

206
11.3.1 Bull and bear spreads

Consider an investor with a mildly bullish view (i.e. he thinks the price of
the underlying is going to rise) but who wants a portfolio that’s protected
against extreme price moves. He can choose one of the following option
combinations.

• Bull call spread: combine a long position in a call option with a low
strike (K1 ) with a short position in a call option with a higher strike
(K2 ).
• Bull put spread: combine a long position in a put option with a low
strike (K1 ) with a short position in a put option with a higher strike
(K2 ).

Both of these positions yield the desired portfolio payoff profile, i.e. a port-
folio value that rises when the underlying price rises but without extreme
gains or losses. An explanation for the payoff profile of the bull call spread
is given below.

Define the underlying price at maturity to be ST . The bull call spread


consists of a long call position with strike K1 and a short call position with
strike K2 where K2 > K1 :

Position Pay-Off
S T ≤ K1 K1 ≤ S T ≤ K2 K2 ≤ S T

Long 1 Call @ K1 0 S T − K1 S T − K1
Short 1 Call @ K2 0 0 K2 − S T

Total 0 S T − K1 K2 − K1

The profit on the bull spread equals this payoff minus the premium on the
call that has been purchased plus the premium on the call that has been
written.

Self-assessment exercise: demonstrate that the bull put spread defined


previously gives a similar portfolio payoff profile to the call bull spread
using a payoff table such as that above.

207
6 Option combinations: payo↵s and uses 36

Figure 11.6:
Bull Call
spread BullCalls
using Spread

6 Profit

@ -
@ ST
K1 K2 @
@
@
@
@
@
@
@
@
@

6 Option combinations: payo↵s and uses 37

Figure
Bull11.7: Put
spread BullPuts
using Spread
Derivatives: instruments and pricing Asset Pricing: lectures 9 and 10

6 Profit
@
@
@
@
@
@
@
@ K1 K2
@
@ - ST
@

11.3.2 Bear spreads


Derivatives: instruments and pricing Asset Pricing: lectures 9 and 10

An investor who holds a mildly bearish view about the price of the under-
lying (i.e. she thinks that the price of the underlying is going to fall) but
wants a position that’s not too sensitive to extreme market movements
might construct a bear spread.

• Bear call spread: combine a short position in a call option with a low
strike (K1 ) with a long position in a call option with a higher strike
(K2 ).

208
• Bear put spread: combine a short position in a put option with a low
strike (K1 ) with a long position in a put option with a higher strike
(K2 ).

Self-assessment exercise: demonstrate that both bear spread positions


have the desired payoff profiles using a payoff table such as that above.
6 Option combinations: payo↵s and uses 39

Figure 11.8:
Bear Callusing
spread BearCalls
Spread

6 Profit

@
@
K1@@@@ K2
@ @ - ST
@
@
@ @
@ @
@ @
@
@@
@
@ @
@ @
@
@
@
@

6 Option combinations: payo↵s and uses 40

Figure 11.9:
Bear Putusing
spread BearPuts
Spread
Derivatives: instruments and pricing Asset Pricing: lectures 9 and 10

6 Profit @
@
@
@ @
@ @
@ @
@ @
@ @
@ @
@ @
@ @
@ @ - ST
@ @
K 1 @ @ K2
@@
@

11.3.3 The Butterfly spread


Derivatives: instruments and pricing Asset Pricing: lectures 9 and 10

Consider an investor who believes that volatility in the underlying will be


low until maturity (i.e. the price of the underlying will change little from

209
it’s current value, either up or down). That investor wants a portfolio that
generates money for small moves in the underlying. The butterfly spread
is such as portfolio.

• Butterfly spread using calls: go long one call with a low exercise
price (K1 ), short two calls with a medium strike (K2 ) and long one
call with a high exercise price (K3 ).
• Butterfly spread using puts: go long one put with a low exercise
price (K1 ), short two puts with a medium strike (K2 ) and long one
put with a high exercise price (K3 ).

We’ll assume that the underlying price is currently equal to K2 and that K2
is equal to the average of K1 and K3 . The worked example of the payoffs
for the butterfly spread is as follows.

Position Pay-Off
S T ≤ K1 K1 ≤ S T ≤ K2 K2 ≤ S T ≤ K3 K3 ≤ S T

Long 1 Call @ K1 0 S T − K1 S T − K1 S T − K1
Short 2 Calls @ K2 0 0 2 ( K2 − S T ) 2 ( K2 − S T )
Long 1 Call @ K3 0 0 0 S T − K3

Total 0 S T − K1 K3 − S T ∗ 0∗

(∗ : Since 2K2 = K1 + K3 )
6 Option combinations: payo↵s and uses 43

Figure 11.10: Butterfly


Butterfly spread Spread using calls
using Calls

6 Profit

A
A
A
@ A
@A
K1 @A K
@ 3
A -
A@ ST
K A@
2 A
A
A
A
A
A
A
A
A
A

210
Derivatives: instruments and pricing Asset Pricing: lectures 9 and 10
11.3.4 The Straddle

Consider an investor who wants to bet on volatility in the underlying asset


i.e. he wants a portfolio that is valuable both when there are extreme up
and down movements in the underlying price.

• Straddle: go long a put and long a call with the same strike (K) and
where K is equal to the current underlying price.

• Strangle: go long a put with a low strike price (K1 ) and long a call
with a high strike price (K2 ). We assume that K1 is below the current
underlying price and K2 is above

Both will pay off when there are extreme up or down moves in the under-
lying price. The strangle, however, increases the range of the underlying
price where the buyer of the straddle makes no profits.
6 Option combinations: payo↵s and uses 46

Figure 11.11:
LongStraddle
Straddle portfolio

6 Profit
@
@ @@
@ @
@ @
@ @
@ @
@ @
@ @
@ @
@ @ K -
@
@
@
@
ST
@ @
@ @
@ @
@
@
@

Derivatives: instruments and pricing Asset Pricing: lectures 9 and 10

11.4 Put-call parity

We can use no-arbitrage to relate the prices of puts and calls on the same
stock.

211
The following condition must hold for a put and a call on the same under-
lying and with the same strike price (X) and time to maturity (T):

S + p = c + Xe−rT (11.1)

In the above I have used continuous discounting but the condition is un-
affected by the use of discrete compounding/discounting.

To make clear why this condition must hold, consider two portfolios.

• Portfolio A consists of the underlying and a put option with T peri-


ods to maturity and struck at X.

• Portfolio B consists of a call struck at X and with time to maturity T


plus cash to the value of Xe−rT .

The payoff profiles of these two portfolios are given below:

Portfolio ST ≤ X ST > X
A ST + max[0, X − ST ] = X ST + max[0, X − ST ] = ST
B max[0, ST − X ] + X = X max[0, ST − X ] + X = ST

Note that, regardless of the final value of ST , both portfolios always give
the same payoff. Thus, no-arbitrage tells us that their prices must be iden-
tical. The price of portfolio A is S + p and that of B is c + Xe−rT . Equating
these gives us the put-call parity condition.

11.5 Arbitrage and option price bounds

Before trying to derive precise values for the premia on options, in this sec-
tion we provide information on some relationships that place restrictions
on options prices with respect to the underlying. These are all based on

212
no-arbitrage and we present them for call options only (although similar
arguments exist for put options).

We look at:

• lower bounds on call prices

• upper bounds on call prices.

Lower bounds

An obvious bound on the price of a call is that it must always be positive.


As the payoff of a call option is never negative, the price of a call can never
be negative. An asset with weakly positive payoffs must always have a
weakly positive price:

c≥0

But we can deliver a sharper lower bound. Consider the following two
portfolios:

• Portfolio A: consists of a call option with time to maturity T and


struck at X, and cash to the value of Xe−rT .

• Portfolio B: a unit of the stock.

What are the payoffs of these two portfolios at the maturity date?

Portfolio ST ≤ X ST > X
A max[0, ST − X ] + X = X max[0, ST − X ] + X = ST
B ST ST

Note that the payoff of portfolio A is always at least as large as that of


portfolio B. Thus, the price of portfolio A must be at least as big as that of
B. This implies that:

213
c + Xe−rT ≥ S ⇒ c ≥ S − Xe−rT

Thus, the price of the call is always weakly greater than the current price
of the stock less the present value of the exercise price.

Upper bound

The payoff of the call option is always smaller than the terminal stock
price. This implies that holding the stock is always at least as good as hold-
ing a call and, accordingly, the price of the stock must be weakly greater
than the price of the call. Thus:

c≤S

Combining these bounds

Putting the analysis above together we can see that:

max[0, S − Xe−rT ] ≤ c ≤ S

The call premium must always lie in this interval, otherwise there is arbi-
trage available.

11.6 Options pricing

Now we will aim to get precise answers to the question ‘what is the right
premium for this option’? We will use a couple of models, which differ
in the way that they describe the evolution of the price of the underlying.
For both approaches, though, the option will be priced using absence of
arbitrage.

214
11.6.1 Binomial option pricing

To start, we will price options (by which we mean work out the right pre-
mium) using binomial models.

The binomial assumption

In a single period, the underlying price can move from its current level
(S) to one of only two new levels. Either the price moves up by a factor
u, thus reaching the level uS, or it moves down by a factor d to the level
dS (where u > d). We call the first state the ‘up state’ and the second the
’down state’. Both u and d are known today.

This seems pretty restrictive – prices can only move to one of two different
levels in the next period – but it’s easy to generalise to many periods and
to think of the period length shrinking to cover a matter of seconds so that
the process above starts to resemble something more reasonable.

An example of a one-step binomial process is shown in Figure 11.12.

The option to be priced

Based on this process for the underlying, we will use absence of arbitrage
to price

• A one-period call option.

• The exercise price for the call option is X.

• It is European such that it can only be exercised at the end of the


period.

Figure 11.13 shows the payoff structure of this call in the binomial world.
We denote by cu the option payoff in the up state and the option payoff in
the down state is cd .

215
The risk-free asset:

We assume that there also exists a one-period zero-coupon bond with face
value £1. The one-period spot rate is r and so the price dynamics of the
bond are as shown in Figure 11.14 below. Note that we require that d <
(1 + r ) < u.
10 Binomial option pricing 70

Figure 11.12: Binomial process for underlying: S = 100, u = 1.25, d = 0.80


Figure 2: Binomial process for underlying: S = 100, u = 1.25, d = 0.80

t0 = 0 t1 = 1

Stock Price = uS = 125

‘Up’

v
Stock Price = S = 100 HH
H
HH
H
HH ‘Down’
H
HH
H
HH
Stock Price = dS = 80

10 Binomial option pricing 71

Derivatives: instruments and pricing Asset Pricing: lectures 9 and 10


Figure 11.13: Binomial setting: payoff to call option struck at X
Figure 3: Binomial setting: payo↵ to call option struck at X

t0 = 0 t1 = 1

Payo↵ = cu = max[0, uS X]

‘Up’

w
Premium = c =? HH
H
HH
H
HH ‘Down’
H
HH
H
HH
H
H Payo↵ = cd = max[0, dS X]

Derivatives: instruments and pricing Asset Pricing: lectures 9 and 10

216
10 Binomial option pricing 72

Figure 11.14: Binomial setting: dynamics for one-period zero-coupon


bond
Figure 4: Binomial setting: dynamics for one-period zero-coupon bond
t0 = 0 t1 = 1

Payo↵ = 1
‘Up’

Price = 1 u
1+r HH
H
HH
H
HH‘Down’
H
HH
HH
Payo↵ = 1

Pricing: building a replicating portfolio for the option


Derivatives: instruments and pricing Asset Pricing: lectures 9 and 10

We want to build a portfolio of the underlying stock and some risk free
bonds which has exactly the same payoff as the option. If, to build this
portfolio, we purchase ∆ units of stock and we sell N one-period zero-
coupon bonds, it has payoffs as shown in Figure 11.15. So, in one period
from now, if the stock has risen in value to uS then the portfolio is worth
∆uS − N. The first term here is the value of your stock holding in one
period and the second term is due to the fact that, having sold N bonds
now, you are required to pay the holders $N in aggregate in one period.
Similarly, the value of the replicating portfolio in the state when the stock
has moved in price to dS is equal to ∆dS − N.

By comparison of Figures 11.15 and 11.13, for this portfolio to exactly repli-
cate the payoff of the call option, the following conditions must hold:

cu = ∆uS − N
cd = ∆dS − N

Given that cu , cd , u, d and S are known, we can solve this pair of equations
for the quantities of stock (∆) and bonds (N) in the replicating portfolio.

The solutions are:

217
10 Binomial option pricing 73

Figure 11.15: Binomial setting: dynamics for replicating portfolio


Figure 5: Binomial setting: dynamics for replicating portfolio
t0 = 0 t1 = 1

Payo↵ = uS N
‘Up’

Price = S N u
1+r HH
H
HH
H
HH‘Down’
H
HH
HH
Payo↵ = dS N

Derivatives: instruments and pricing cu − cd dcu − ucd Asset Pricing: lectures 9 and 10
∆= , N=
(u − d)S u−d

Pricing the option

Now we know the precise composition of the replicating portfolio, we im-


pose no-arbitrage and argue that the price of the call option must be ex-
actly equal to the cost of setting up the replicating portfolio (because the
option and the portfolio always have the same payoff).

Thus the no-arbitrage call price is:

N cu − cd dcu − ucd
c = ∆S − , where ∆ = , N= (11.2)
1+r (u − d)S u−d

This is easy to compute using the values of the parameters of the model
that we know today (i.e. S, u, d, r) and the payoffs of the option (i.e. cu and
cd ). We call this method pricing by replication, as one explicitly works out
the composition of the replicating portfolio (i.e. ∆ and N) and then solves
for the price of the call.

Note the following features of this derivative price:

218
1. The final price for the call is entirely unrelated to the probabilities of
the up or down move in the binomial process for the underlying.

• Indeed, we have not even mentioned probabilities until this


point.
• Intuitively, this is due to the fact that our replicating portfolio
is designed to match the option payoff exactly in all possible
outcomes. It doesn’t matter how likely any outcome is or how
relatively likely any pair of outcomes are.

2. The above method does not rely on the derivative we’re pricing be-
ing a call option at all. It works for any derivative.

• As long as you can work out the payoff of the derivative in the
up-state (cu ) and the payoff in the down-state (cd ) then you can
directly apply the final formula (i.e. equation (11.3)) to what-
ever derivative you want.

An alternative interpretation

(1+r )−d
If we define q = u−d then the preceding equation can be made to look
a bit more straightforward. We can write the no-arbitrage call price as:

1
c= [qcu + (1 − q)cd ] (11.3)
(1 + r )

Note that this is just equation (11.2) re-arranged so that equation and equa-
tion (11.3) will always deliver exactly the same option price.

We call this way of working out the option price risk neutral pricing. Why
do we give it this name?

Well, first, note that as we required d < (1 + r ) < u, it must be the case
that q is between zero and one. Thus we can interpret q as if it were a
probability. Indeed, q is referred to as the risk-neutral probability of an up-
move (and 1 − q is the risk-neutral probability of a down-move). This is
because, in a world where all investors were risk-neutral, q would have
to be the probability of an up-move and 1 − q the probability of a down-
move.

219
Given this interpretation for q, then equation (11.3) can be thought of as
saying that the option premium is equal to the expected payoff of the call
under the risk-neutral probability structure (the numerator), discounted
back to the current time at the risk-free rate (the denominator).

Again, it’s worth noting that any derivative can be priced by calculating its
discounted expected payoff under the risk-neutral probability structure.
Nothing we have done relies on the derivative we’re looking at being a
call option.

BUT: note again that the true probabilities of up and down-moves in the
underlying price are entirely irrelevant. The only probabilities that enter
are the risk-neutral probabilities and these are not ‘real’ probabilities.

Example: binomial call pricing

The risk-free rate is 3%. A stock has current price $60. In one period its
price will either rise to $75 or fall to $48. Price a one-period call option on
this stock with exercise price $65.

Well, first we must work out the size of the up and down moves. We have:

75 48
u= = 1.25 , d= = 0.8
60 60

Next we need to know the call option payoffs in the two states:

cu = max[0, uS − X ] = max[0, 10] = 10


cd = max[0, dS − X ] = max[0, −17] = 0

Now, to price the option by replication we compute values for ∆ and N:

cu − cd 10 dcu − ucd 0.8 × 10 − 1.25 × 0


∆= = = 0.37 , N = = = 17.78
(u − d)S 0.45 × 60 u−d 0.45

220
Finally, the price of the call is:

N 17.78
c = ∆S − = 0.37 × 60 − = 4.96
1+r 1.03

We can also work out the option price via the risk-neutral method. The
risk-neutral probability of an up-move is given by:

(1 + r ) − d 1.03 − 0.8
q= = = 0.51
u−d 1.25 − 0.8

Then the call price is:

1 1
c= [qcu + (1 − q)cd ] = [10 × 0.51 + 0 × 0.49] = 4.96
(1 + r ) 1.03

11.6.2 Black-Scholes option pricing

The previous section relied on a discrete time description of the underly-


ing price for a derivative, in the sense that time was assumed to pass in
steps from one period to the next and during that period the price of the
underlying could move either a fixed amount upwards or a fixed amount
downwards.

The analysis of the current section makes the alternative assumption that
time passes continuously and the price of the underlying evolves on a
continuous basis. We can view this continuous time setting as the limit of
a binomial world where the number of steps in the binomial tree tends to
∞ and the amount of time that elapses during each step approaches zero.

Using this framework and absence of arbitrage, one can derive the famous
Black-Scholes price for a call option. However, the mathematics behind this
derivation are beyond the scope of the current course. Thus we will just
present the pricing formula, analyze its properties and look at some exam-
ples.

221
Continuous compounding and discounting

Earlier we saw that in a continuous time setting, the continuously com-


pounded value of £A from time t to T is given by:

Aer(T −t)

where r is the continuously compounded interest rate. Similarly, continu-


ously discounting an amount B back to the current time (t) from a given
point in the future (T) gives:

Be−r(T −t)

We will use both of these formulae in what follows.

The Black-Scholes formula

It we are currently at time t, then the Black-Scholes price for a call option
with maturity date T and strike price X is given by:

c = SN (d1 ) − Xe−r(T −t) N (d2 ) (11.4)

where S is the current underlying price, σ is the volatility of the underlying


price process, r is the interest rate, ( T − t) is the time to maturity and:

ln(S/X ) + (r + σ2 /2)( T − t)
q
d1 = √ , d2 = d1 − σ ( T − t ).
σ T−t

Finally, the function N (·) is the cumulative Normal distribution function.

We can (loosely) interpret that Black-Scholes equation as follows.

• It turns out that N (d2 ) can be interpreted as the risk-neutral probabil-


ity that the option will be exercised such that the term Xe−r(T −t) N (d2 )

222
can be thought of as the expected present value of the cost of exercis-
ing the option (as it’s a probability × a discount factor × the exercise
price).
• Similarly, the first term SN (d1 ) can be interpreted as the expected
value of a variable that pays off max[S − X, 0] is a risk-neutral world.
So it’s the expected benefit from exercising the option.
• Thus the option price is just the expected gain from exercising less
the expected cost of exercise.

Note that we can use put-call parity to derive the Black-Scholes price for
a put option. If we plug the Black-Scholes price for a call, equation (11.4),
into equation (11.1) and re-arrange we get:

p = Xe−r(T −t) N (−d2 ) − SN (−d1 ) (11.5)

where d1 and d2 are as previously defined.

There are five unknowns that must be input to the B-S equation in order
to give the call price. The following table tells us how the call price (and
also the price of a B-S put) vary with each of these parameters:

Table 11.1: Properties of the B-S pricing equations

Increase in Call Put


S Increase Decrease
X Decrease Increase
σ Increase Increase
T Increase Increase
r Increase Decrease

Some of these effects are obvious and others more subtle. Explanations are
given below:

• Call premia increase as the underlying price rises as, holding every-
thing else fixed, a rise in the underlying price means that the call is a
claim on a more valuable asset. (The converse holds for puts.)

223
• Call prices fall as the exercise price is increased as, other things equal,
it means that the cost of exercising the option is increased. (Again,
the converse holds for puts.)

• Rises in both σ and T tend to increase call and put prices as both
increase the chances that the option ends up in the money.

• Increased interest rates increase the value of calls as this reduces the
present value of the exercise price that the holder must pay. An in-
terest rate rise reduces put values as it reduces the present value of
the exercise price that the holder receives if exercise occurs.

Measuring volatility

Of the five parameters in the preceding table all are easy to obtain aside
from the volatility parameter σ. There are two main ways to obtain values
for this parameter.

• Estimate volatility from the history of the underlying price. Given


a time-series of daily prices on the underlying, first compute daily
returns and then compute their time-series standard deviation. An
annualised volatility measure for that underlying can then
√ be calcu-
lated as the estimated standard deviation multiplied by 252.

• Retrieve the volatility parameter from the price of an already traded


option on the same underlying. If one knows that option’s price, the
option’s strike, its time to maturity, the current interest rate and the
underlying price then the Black-Scholes formula can be inverted to
give the volatility parameter that prices the option. This is called
an implied volatility. Then use the implied volatility to price another
option on the underlying.

Example: B-S call pricing

Assume that the risk-free rate is 5%. A stock has current price $20 and its
annualized historical volatility is 22.5%. Price a call with exercise price $22
and one-year to maturity.

224
First we should work out the values for d1 and d2 using the previous equa-
tions. We get the following:

ln(20/22) + (0.05 + 0.2252 /2)(1)


d1 = √ = −0.0889
0.225 × 1

d2 = d1 − 0.225 × 1 = −0.31388

Then using the Excel function NORMSDIST, that returns the value of the
standard Normal CDF we get:

N (d1 ) = 0.4646 , N (d2 ) = 0.3768

Finally, plugging into equation (11.4) we get:

c = SN (d1 ) − Xe−r(T −t) N (d2 ) = 20 × 0.4646 − 22 × e−0.05×1 × 0.3768 = 1.4063

11.7 Summary

We have finally discussed option contracts and option pricing. We started


by introducing basic option contract structure and then introduced their
payoffs. We went on to discuss the payoff structures of some important
option portfolios. Finally we turned to pricing. We started with some sim-
ple bounds on option premia using absence of arbitrage. Then we moved
on to binomial and finally Black-Scholes pricing

225
Reminder of learning outcomes
Having completed this chapter, and the Essential reading and activities,
you should be able to:

• describe the key payoff features of put and call options.

• compute payoffs for option portfolios.

• provide some no-arbitrage bounds on option prices.

• price options via no-arbitrage in a binomial setting.

• discuss the key features of the Black-Scholes call option pricing model
and use the model to price options.

226
11.8 Exercises
1. Why would we typically expect the prices of American options to be
greater than those for European options?

2. The current price of XPZ stock is 200p. It never pays dividends and,
over the next six months, will either rise in price by 15% or fall in
price by 10%. The risk-free rate is 2% per annum. What are the
prices of European call and put options each with strike price 220p
and each with six months to maturity?

3. Show that put-call parity holds for the option prices you computed
in the question above.

4. Using the data from the previous question, compute and plot the
payoff function for an at the money straddle on XPZ.

5. How would you expect the price of a European call option to depend
on the payment of dividends? Should dividends paid prior to the
maturity date cause call prices to rise or fall? Explain your reasoning.

6. The price of a stock is $40. The price of a one-year put with strike
price $30 is $0.70 and a call with the same time to maturity and a
strike of $50 costs $0.50. Both options are European.

(a) An investor buys one share, shorts one call and buys one put.
Draw and comment upon the payoff of this portfolio at maturity
as a function of the underlying price.
(b) How would your answer to (a) change if the investor buys one
share, shorts two calls and buys two puts instead.

7. How do European option premia vary as time to maturity varies?


Explain why this relationship holds. Does the sign of the relationship
differ for puts and calls?

227
Chapter 12

Sample exam paper with worked


solutions

In the two sections that follow in this chapter, there is a sample exam paper
and then a set of sketch solutions to that exam.

In order to obtain the most value from this exam, I would encourage all
students to attempt the questions before looking at the solutions.

228
12.1 Sample exam paper

Exam instructions: answer any three of the four questions below. All
questions carry equal marks. You have three hours to complete the exam.

Question 1

(a) Is the following statement true or false? Justify your answer


with a short explanatory paragraph.
In an efficient stock market, no amount of fundamental or
technical analysis will help an investor to achieve greater ex-
pected future returns. The only way that an investor can
expect to make greater returns is by taking more risk.
(6 marks)

(b) The quarterly compounded interest rate is 8%. What is the effec-
tive six-month interest rate and what is the present value of an
annuity cashflow stream of $500 to be received every six months
for the next five years? Give answers to two decimal places and
show all your workings. (6 marks)

(c) Two firms have identical expected earnings next year, equal to
$10 per share, and identical required returns of 10%. Firm 1
chooses to pay out all future earnings as dividends and firm 2
decides to pay out only 75% of its earnings, investing the rest
into projects with an ROE of 16%. What is the difference in the
share prices of the two firms? (7 marks)

(d) An economy contains two risky assets. Both of them cost $1.
Asset A will pay the holder $3 if it rains tomorrow and $0.50
if it does not rain tomorrow. Asset B will pay the holder $3 if it
rains tomorrow and $0.75 if it does not rain. Is there an arbitrage
portfolio available in this economy (assuming that you can go
long and short both assets)? If so, how is it constructed? If
you think that there is an arbitrage available what would you
expect to happen to the prices of the two securities to eliminate
the arbitrage? (6 marks)

229
Question 2
In all parts of the question below, assume that the underlying asset
pays no dividends.

(a) A straddle is an option portfolio that is constructed by buying


one at the money call option and one at the money put option.
Draw the profit profile for a straddle position in a stock cur-
rently priced at S. If an investor buys a straddle portfolio, what
can you infer about her beliefs regarding the future evolution of
the underlying stock’s price? (5 marks)

(b) Draw the profit profile from buying two at the money call op-
tions and one at the money put. Contrast the picture you get
here with that you obtained in (a) and indicate how the beliefs
of an investor who builds this portfolio might differ from the
investor who buys a straddle. (5 marks)

(c) A non-dividend paying stock has a current price of $50. In one


period, the price will either increase to $60 or fall to $45. If
the one period interest rate is 5%, what is the price of a strad-
dle built from call and put options with one period to matu-
rity? (10 marks)

(d) Explain how volatility in the underlying stock price affects the
prices of both call and put options written on the stock. What is
the intuition behind this relationship? (5 marks)

Question 3

(a) At present, Lloyds Banking Group is estimated to have a beta of


0.62, while Vodafone has a beta of 1.26. Explain what the differ-
ences in these betas imply for the market risk that the stocks
carry and the stocks’ expected returns. [For the purposes of
this analysis assume that the ‘market portfolio’ is the FT All-
Share.] (6 marks)

(b) If one wished to build a portfolio of Lloyds and Vodafone that


had a beta of exactly one, what portfolio weights would one
have to place on each stock? (3 marks)

(c) The standard deviation of Lloyds Banking Group’s stock return


is 20% and the standard deviation of Vodafone’s return is 35%.

230
If the correlation between their returns is 0.377, what is the stan-
dard deviation of the return on the portfolio that you calculated
in (b)? (3 marks)

(d) If the expected return on Vodafone is 13.6% and the expected re-
turn of Lloyds is 7.2%, what is the expected return on the port-
folio that you calculated in (b)? (2 marks)

(e) The risk-free rate is 1%. An investor ranks investments using


the Sharpe ratio and can only choose to invest in Lloyds or
Vodafone or the portfolio in (b). Which of the three will the
investor choose? (3 marks)

(f) Guided by the data and your calculations above, draw an ap-
proximation to the portfolio frontier that is available from in-
vesting in Vodafone and Lloyds only. Show on the graph how
an investor with access to a risk-free asset returning 1% per
year would build her portfolio of Vodafone and Lloyds. [Note:
there is no need to perform any more calculations here. Draw
a smoothed approximate frontier that fits the data that you al-
ready have and work from that]. (8 marks)

Question 4

(a) i. Two bonds, both with 10 years to maturity, have the same
yield to maturity (equal to 5%) but different coupon rates.
Coupon payments are annual. Bond A has a Macaulay du-
ration of 8.66 and bond B has a Macaulay duration of 7.05.
Based on this information, explain which bond you think
has the higher coupon rate. Be sure to present your reason-
ing. (3 marks)

ii. Bond A in part (i) has a face value of $100 and a coupon rate
of 3%. What is its price? (4 marks)

iii. If the yield to maturity of bond B was to decrease by 0.5%,


what would be your estimate of the percentage change in
its price? (3 marks)

iv. Explain whether the answer you arrived at in part (iii) is


likely to be exact or an approximation. Use a diagram show-

231
ing the relationship between prices and yields to support
your arguments. (4 marks)

(b) i. Two bonds both have two years to maturity, face value of
$100 and pay coupons annually. The coupon rate on bond
A is 3% and that on bond B is 8%. Bond A sells for $94.55
and bond B’s price is $103.81. What are the one and two
year spot interest rates that can be inferred from this infor-
mation? (6 marks)

ii. Using your estimated one- and two-year spot rates, com-
pute the one-year forward interest rate between the end of
year 1 and the end of year 2. Give an economic interpre-
tation of the information that is contained in this forward
rate. (5 marks)

232
12.2 Sketch solutions to sample exam
Question 1

(a) This statement is essentially true. In an efficient market the in-


vestor only gets rewarded for bearing systematic risk. There
is no scope for earning abnormal returns that are positive (on
average) as all information is correctly reflected in prices. Tech-
nical analysis doesn’t help in a weak form efficient market and,
in addition, in a semi-strong form efficient market fundamental
analysis is useless. Support the answer with either Jensen’s or
Malkiel’s definition of efficiency.
(b) If the quarterly compounded rate is 8% then the actual three-
month rate is 2%. This means that the six-month rate is:

r = (1 + 0.02)2 − 1 = 0.0404 = 4.04%


Now value the annuity. The semi-annual cashflow is $500 and
this arrives at the end of every (six month) period for 10 periods.
So:

   
C 1 500 1
PV = 1− = 1− = 4047.38
r (1 + r ) T 0.0404 (1.0404)10

(c) Firm 1 pays out all earnings as dividends and thus has zero
dividend growth. Its price is:

D1 10
P1 = = = 100
r 0.10
Firm 2 retains 25% of earnings and they are invested at a ROE
of 16%. Thus dividend growth is:

g = 0.25 × 0.16 = 0.04


The stock price of firm 2 is thus:

0.75 × 10 7.5
P2 = = = 125
0.10 − 0.04 0.06
So the price difference is 25.

233
(d) Here are the payoffs on the two securities tomorrow, both of
which cost $1;
Rainy Sunny
A 3.00 0.50
B 3.00 0.75

Is there an arbitrage available? Yes, via the following argument;


• The payoff on B is always at least as big as that on A and,
when it is sunny, the payoff on B is larger than that on A.
• So we would expect the price of B to exceed the price of A.
• But the prices are identical (they’re both $1)
• Thus there is an arbitrage available.
Here you can make some money by buying one unit of B and
short selling one unit of A. As they both cost $1, this costs noth-
ing (you pay $1 for the unit of B and receive $1 when you short
sell A). The payoffs from this portfolio are;

Rainy Sunny
B-A 3.00 - 3.00 = 0 0.75 - 0.50 = 0.25

So you’ve paid nothing to get a portfolio which never costs you


money and sometimes pays you money. This is an arbitrage.
To correct the arbitrage, we would expect the price of B to rise
relative to that of A.

234
Question 2

(a) If you buy both a put and a call, then one immediately pays
out c + p (where these are the option premia). Call the current
underlying price S and the price at maturity ST . Then if ST > S,
the call option payoff is ST − S and the put payoff is zero so total
profit is ST − S − c − p. If ST < S then the put payoff is S − ST ,
the call payoff is zero and the total profit is S − ST − c − p. This
means that the profit, as a function of ST , is V-shaped with a
minimum at ST = S. It looks something like this.
Profit

ST
S

This profile would be of value to someone who thought that the


stock price was either going to go a reasonably long way up or a
reasonably long way down (as then the portfolio makes a profit)
but also thinks that the chances of the stock price not moving at
all are fairly small (as then the portfolio makes a loss). Thus it is
a portfolio that allows one to take a bet on volatility being high.
(b) The cost of buying this portfolio is 2c + p. In contrast to part
(a), the profit profile is not symmetric, though. When ST < S,
the profit is S − ST − 2c − p but when ST > S the profit is
2(ST − S) − 2c − p. So the profits when the stock price rises
are larger than when the stock price falls (in fact the gradient of
the upward sloping part of the profit function is +2 while that
of the downward sloping part of the profile is -1). This portfo-
lio would be useful to an investor who things that volatility is
going to be high (as before) but in addition thinks that a large

235
price rise is more likely than a large price fall. See below for a
picture:

Profit

S ST

(c) Let’s use the risk neutral method. First calculate the risk neutral
probability of an up move. It is:

1 + rf − d
q=
u−d
The question says that r f = 0.05, while u = 60/50 = 1.2 and
d = 45/50 = 0.9. So:

1.05 − 0.90
q= = 0.5
1.20 − 0.90
Now, the call price is:

1 1
c= [qcu + (1 − q)cd ] = [0.5 × 10 + 0.5 × 0] = 4.76
1 + rf 1.05

Note that in the above cu is the call payoff in the up state i.e.
max (60 − 50, 0) = 10 and cd is the call payoff in the down state
i.e. max (45 − 50, 0) = 0.
We can price the put via put-call parity. It is:

X 50
p = c+ − S = 4.76 + − 50 = 2.38
(1 + r f ) 1.05
Thus the cost of the straddle is 2.38 + 4.76 = 7.14.

236
(d) Higher volatility in the underlying increases the premia on both
puts and calls. What is the intuition for this? Higher volatil-
ity means that it is more likely that one will see more extreme
movements up and down in the underlying. Consider the holder
of a call option. When the underlying becomes more volatile,
the call holder enjoys the fact that the likelihood of an extreme
positive payoff is now greater (when the underlying rises a long
way) and, importantly, is not worried about increased likeli-
hood of the underlying falling a long way as on the downside
his payoff never goes below zero. Thus, he is better off and
the call in more valuable under high volatility and this means
that the call premium must rise. The identical argument can be
made for put options.

237
Question 3

(a) VOD clearly carries about twice the market risk than does LLOY.
For each +1% return on the market, one expects VOD to expe-
rience a return of +1.26% on average. For LLOY, though, an
extra +1% return on the market only delivers a +0.62% rise in
the stock price on average (and obviously symmetric arguments
hold for drops in the market). Under the CAPM, the greater
beta for VOD will lead to a greater expected return.
(b) We want a portfolio with beta equal to 1. Call the weight on
VOD, x. Then we must have:

β VOD x + β LLOY (1 − x ) = 1.26x + 0.62(1 − x ) = 1


This implies that:

0.38
x= = 0.59
0.64
and so the weight on VOD is 0.59 and the weight on LLOY is
0.41.
(c) This is just a portfolio risk computation. We have:

2
σP2 = σVOD x2 + σLLOY
2
(1 − x )2 + 2x (1 − x ) ρ σVOD σLLOY

Plugging in the numbers we have:

σP2 = 352 0.592 + 202 0.412 + 2(0.59)(0.41) (0.377) (35)(20) = 625

Thus:

σP = 625 = 25
(d) Now just mean portfolio return is required:

E( R P ) = E( RVOD ) × x + E( R LLOY ) × (1 − x )
= 13.6 × 0.59 + 7.2 × 0.41 = 11

(where I’ve rounded to the nearest percentage point in the an-


swer).

238
(e) The Sharpe ratio for a portfolio Q is defined as:

E( R Q ) − R F
SQ =
σQ
Doing this computation for VOD one gets 0.36, while for LLOY
it is 0.31 and for portfolio P it is 0.40. As a higher Sharpe ra-
tio means a ‘better’ portfolio (as one gets higher mean return
for each unit of risk taken on), the ‘best’ of the three choices is
portfolio P.
(f) In the plot below I’ve drawn a smooth frontier through the points
representing VOD, LLOY and P. Then I’ve drawn the CML, cut-
ting the y-axis at the risk-free rate (of 1%) and having a slope
which makes it tangent to the frontier. This gives an optimal
portfolio for VOD and LLOY marked at point T on the graph.
Any approximation that looked something like this was marked
as correct (as long as there were no errors in labelling or descrip-
tion).

20

18

16 CML

14
VOD

12
Mean return

T P
10

LLOY
6

0
0 5 10 15 20 25 30 35 40
Standard deviation

239
Question 4
(a) i. The bonds have the same maturities and yields. Their coupon
rates differ. Which one has the lowest duration? The bond
with the highest coupon rate will have the lower duration,
as a bigger proportion of its cashflows come early on in its
lifetime. Thus, bond B is the high coupon rate bond.
ii. Discounting its cashflows we get:
10
3 100
PA = ∑ 1.05i + 1.0510 = 84.56
i =1
The easy way to do this is to compute the PV of coupons
using the annuity valuation formula (which equals 23.17)
and then add on the PV of the face value (61.39).
iii. Here we can approximate the percentage change in price
using the duration:
∆P D 7.05
≈− ∆y = − × (−0.005) = 0.034
P 1+y 1.05
iv. This answer is approximate as the relationship between prices
and yields is non-linear, but the duration-based estimate
of price change assumes a linear relationship (as it’s based
only on the first derivative of the relationship between prices
and yields). Draw the standard diagram here, where one
is approximating the change along a curve with a straight
line. Bonus marks for pointing out that one could make the
approximation better by using second derivatives.
(b) i. So for Bond A we have:
3 103
+ = 94.55
1 + r1 (1 + r2 )2
For Bond B we have:
8 108
+ = 103.81
1 + r1 (1 + r2 )2
Now, let’s define x = 1/(1 + r1 ) and y = 1/(1 + r2 )2 . Then
we can write:

3x + 103y = 94.55
8x + 108y = 103.81

240
This is just a pair of simultaneous equations and we can
solve them for:

x = 0.962 , y = 0.890
Then inverting the definitions of x and y we find that r1 =
4% and r2 = 6%.
ii. The forward rate between the end of years 1 and 2, f 2 is
defined through the following equation:

(1 + r2 )2 1.062
(1 + f 2 ) = = = 1.08
(1 + r1 ) 1.04
Therefore the forward rate is 8%. If you believe the pure
expectations theory of the term structure, this is interpreted
as implying that the one year spot rate is expected to rise
to 8% next year. Alternatively, the difference between the
4% spot rate for this year and the 8% forward rate could
contain, at least partially, a positive term premium as well
as an expected rise in the one year spot rate.

241

You might also like