Portfolio Theory and Risk Management

With its emphasis on examples, exercises and calculations, this book suits advanced
undergraduates as well as postgraduates and practitioners. It provides a clear treatment
of the scope and limitations of mean-variance portfolio theory and introduces popular
modern risk measures. Proofs are given in detail, assuming only modest mathematical
background, but with attention to clarity and rigour. The discussion of VaR and its
more robust generalizations, such as AVaR, brings recent developments in risk measures
within range of some undergraduate courses and includes a novel discussion of reducing
VaR and AVaR by means of hedging techniques.
A moderate pace, careful motivation and more than 70 exercises give students confi-
dence in handling risk assessments in modern finance. Solutions and additional materi-
als for instructors are available at www.cambridge.org/9781107003675.

maciej j. capi ński is an Associate Professor in the Faculty of Applied Mathematics

at AGH University of Science and Technology in Kraków, Poland. His interests include
mathematical finance, financial modelling, computer-assisted proofs in dynamical sys-
tems and celestial mechanics. He has authored 10 research publications, one book, and
supervised over 30 MSc dissertations, mostly in mathematical finance.

ekkehard kopp is Emeritus Professor of Mathematics at the University of Hull,

where he taught courses at all levels in analysis, measure and probability, stochastic
processes and mathematical finance between 1970 and 2007. His editorial experience
includes service as founding member of the Springer Finance series (1998–2008) and
the Cambridge University Press AIMS Library Series. He has taught in the UK, Canada
and South Africa and he has authored more than 50 research publications and five
books.
Mastering Mathematical Finance

Mastering Mathematical Finance is a series of short books that cover all core topics
and the most common electives offered in Master’s programmes in mathematical or
quantitative finance. The books are closely coordinated and largely self-contained, and
can be used efficiently in combination but also individually.

The MMF books start financially from scratch and mathematically assume only under-
graduate calculus, linear algebra and elementary probability theory. The necessary
mathematics is developed rigorously, with emphasis on a natural development of math-
ematical ideas and financial intuition, and the readers quickly see real-life financial
applications, both for motivation and as the ultimate end for the theory. All books are
written for both teaching and self-study, with worked examples, exercises and solutions.

[DMFM] Discrete Models of Financial Markets,

Marek Capiński, Ekkehard Kopp
[PF] Probability for Finance,
Ekkehard Kopp, Jan Malczak, Tomasz Zastawniak
[SCF] Stochastic Calculus for Finance,
Marek Capiński, Ekkehard Kopp, Janusz Traple
[BSM] The Black–Scholes Model,
Marek Capiński, Ekkehard Kopp
[PTRM] Portfolio Theory and Risk Management,
Maciej J. Capiński, Ekkehard Kopp
[NMFC] Numerical Methods in Finance with C++,
Maciej J. Capiński, Tomasz Zastawniak
[SIR] Stochastic Interest Rates,
Daragh McInerney, Tomasz Zastawniak
[CR] Credit Risk,
Marek Capiński, Tomasz Zastawniak
[FE] Financial Econometrics,
Marek Capiński
[SCAF] Stochastic Control Applied to Finance,
Szymon Peszat, Tomasz Zastawniak

Series editors Marek Capiński, AGH University of Science and Technology, Kraków;
Ekkehard Kopp, University of Hull; Tomasz Zastawniak, University of York
Portfolio Theory and Risk Management

MACIEJ J. CAPI ŃSKI

AGH University of Science and Technology, Kraków, Poland

EKKEHARD KOPP
University of Hull, Hull, UK
 University Printing House, Cambridge CB2 8BS, United Kingdom

Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning and research at the highest international levels of excellence.

www.cambridge.org
Information on this title: www.cambridge.org/9781107003675
© Maciej J. Capiński and Ekkehard Kopp 2014
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2014
Printed in the United Kingdom by TJ International Ltd, Padstow Cornwall
A catalogue record for this publication is available from the British Library
Library of Congress Cataloguing in Publication data
Capiński, Maciej J.
Portfolio theory and risk management / Maciej J. Capiński, AGH University of Science and
Technology, Kraków, Poland, Ekkehard Kopp, University of Hull, Hull, UK.
pages cm – (Mastering mathematical finance)
Includes bibliographical references and index.
ISBN 978-1-107-00367-5 (Hardback) – ISBN 978-0-521-17714-6 (Paperback)
1. Portfolio management. 2. Risk management. 3. Investment analysis.
I. Kopp, P. E., 1944– II. Title.
HG4529.5.C366 2014
332.6–dc23 2014006178
ISBN 978-1-107-00367-5 Hardback
ISBN 978-0-521-17714-6 Paperback
Additional resources for this publication at www.cambridge.org/9781107003675
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication,
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
To Anna, Emily, Staś, Weronika and Helenka
Contents
Preface page ix
1 Risk and return 1
1.1 Expected return 2
1.2 Variance as a risk measure 5
1.3 Semi-variance 9
2 Portfolios consisting of two assets 11
2.1 Return 12
2.2 Attainable set 15
2.3 Special cases 20
2.4 Minimum variance portfolio 23
2.5 Adding a risk-free security 25
2.6 Indifference curves 28
2.7 Proofs 31
3 Lagrange multipliers 35
3.1 Motivating examples 35
3.2 Constrained extrema 40
3.3 Proofs 44
4 Portfolios of multiple assets 48
4.1 Risk and return 48
4.2 Three risky securities 52
4.3 Minimum variance portfolio 54
4.4 Minimum variance line 57
4.5 Market portfolio 62
5 The Capital Asset Pricing Model 67
5.1 Derivation of CAPM 68
5.2 Security market line 71
5.3 Characteristic line 73
6 Utility functions 76
6.1 Basic notions and axioms 76
6.2 Utility maximisation 80
6.3 Utilities and CAPM 92
6.4 Risk aversion 95

vii
viii Contents
7 Value at Risk 98
7.1 Quantiles 99
7.2 Measuring downside risk 102
7.3 Computing VaR: examples 104
7.4 VaR in the Black–Scholes model 109
7.5 Proofs 120
8 Coherent measures of risk 124
8.1 Average Value at Risk 125
8.2 Quantiles and representations of AVaR 127
8.3 AVaR in the Black–Scholes model 136
8.4 Coherence 146
8.5 Proofs 154
Index 159
Preface
In this fifth volume of the series ‘Mastering Mathematical Finance’ we
present a self-contained rigorous account of mean-variance portfolio the-
ory, as well as a simple introduction to utility functions and modern risk
measures.
Portfolio theory, exploring the optimal allocation of wealth among dif-
ferent assets in an investment portfolio, based on the twin objectives of
maximising return while minimising risk, owes its mathematical formula-
tion to the work of Harry Markowitz1 in 1952; for which he was awarded
the Nobel Prize in Economics in 1990. Mean-variance analysis has held
sway for more than half a century, and forms part of the core curriculum
in financial economics and business studies. In these settings mathematical
rigour may suffer at times, and our aim is to provide a carefully motivated
treatment of the mathematical background and content of the theory, as-
suming only basic calculus and linear algebra as prerequisites.
Chapter 1 provides a brief review of the key concepts of return and risk,
while noting some defects of variance as a risk measure. Considering a
portfolio with only two risky assets, we show in Chapter 2 how the mini-
mum variance portfolio, minimum variance line, market portfolio and cap-
ital market line may be found by elementary calculus methods. Chapter 3
contains a careful account of the method of Lagrange multipliers, includ-
ing a discussion of sufficient conditions for extrema in the special case of
quadratic forms. These techniques are applied in Chapter 4 to generalise
the formulae obtained for two-asset portfolios to the general case.
The derivation of the Capital Asset Pricing Model (CAPM) follows in
Chapter 5, including two proofs of the CAPM formula, based, respectively,
on the underlying geometry (to elucidate the role of beta) and linear alge-
bra (leading to the security market line), and introducing performance mea-
sures such as the Jensen index and Sharpe ratio. The security characteristic
line is shown to aid the least-squares estimation of beta using historical
portfolio returns and the market portfolio.
Chapter 6 contains a brief introduction to utility theory. To keep matters
simple we restrict to finite sample spaces to discuss preference relations.

1
H. Markowitz, Portfolio selection, Journal of Finance 7 (1), (1952), 77–91.

ix
x Preface
We consider examples of von Neumann–Morgenstern utility functions, link
utility maximisation with the No Arbitrage Principle and explain the key
role of state price vectors. Finally, we explore the link between utility max-
imisation and the CAPM and illustrate the role of the certainty equivalent
for the risk averse investor.
In the final two chapters the emphasis shifts from variance to measures
of downside risk. Chapter 7 contains an account of Value at Risk (VaR),
which remains popular in practice despite its well-documented shortcom-
ings. Following a careful look at quantiles and the algebraic properties of
VaR, our emphasis is on computing VaR, especially for assets within the
Black–Scholes framework. A novel feature is an account of VaR-optimal
hedging with put options, which is shown to reduce to a linear program-
ming problem if the parameters are chosen with care.
In Chapter 8 we examine how the defects of VaR can be addressed using
coherent risk measures. The principal example discussed is Average Value
at Risk (AVaR), which is described in detail, including a careful proof of
sub-additivity. AVaR is placed in the context of coherent risk measures, and
generalised to yield spectral risk measures. The analysis of hedging with
put options in the Black–Scholes setting is revisited, with AVaR in place of
VaR, and the outcomes are compared in examples.
Throughout this volume the emphasis is on examples, applications and
computations. The underlying theory is presented rigorously, but as simply
as possible. Proofs are given in detail, with the more demanding ones left to
the end of each chapter to avoid disrupting the flow of ideas. Applications
presented in the final chapters make use of background material from the
earlier volumes [PF] and [BSM] in the current series. The exercises form
an integral part of the volume, and range from simple verification to more
challenging problems. Solutions and additional material can be found at
www.cambridge.org/9781107003675, which will be updated regularly.
1
Risk and return

1.1 Expected return

1.2 Variance as a risk measure
1.3 Semi-variance

Financial investors base their activity on the expectation that their invest-
ment will increase over time, leading to an increase in wealth. Over a fixed
time period, the investor seeks to maximise the return on the investment,
that is, the increase in asset value as a proportion of the initial investment.
The final values of most assets (other than loans at a fixed rate of interest)
are uncertain, so that the returns on these investments need to be expressed
in terms of random variables. To estimate the return on such an asset by a
single number it is natural to use the expected value of the return, which
averages the returns over all possible outcomes.
Our uncertainty about future market behaviour finds expression in the
second key concept in finance: risk. Assets such as stocks, forward con-
tracts and options are risky because we cannot predict their future values
with certainty. Assets whose possible final values are more ‘widely spread’
are naturally seen as entailing greater risk. Thus our initial attempt to mea-
sure the riskiness of a random variable will measure the spread of the re-
turn, which rational investors will seek to minimise while maximising their
return.
In brief, return reflects the efficiency of an investment, risk is concerned
with uncertainty. The balance between these two is at the heart of portfo-
lio theory, which seeks to find optimal allocations of the investor’s initial
wealth among the available assets: maximising return at a given level of
risk and minimising risk at a given level of expected return.

1
2 Risk and return

1.1 Expected return

We are concerned with just two time instants: the present time, denoted
by 0, and the future time 1, where 1 may stand for any unit of time. Sup-
pose we make a single-period investment in some stock with the current
price S (0) known, and the future price S (1) unknown, hence assumed to
be represented by a random variable
S (1) : Ω → [0, +∞),
where Ω is the sample space of some probability space (Ω, F , P) . The
members of Ω are often called states or scenarios. (See [PF] for basic
definitions.)
When Ω is finite, Ω = {ω1 , . . . , ωN }, we shall adopt the notation
S (1, ωi ) = S (1)(ωi ) for i = 1, . . . , N,
for the possible values of S (1). In this setting it is natural to equip Ω with
the σ-field F = 2Ω of all its subsets. To define a probability measure P :
F → [0, 1] it is sufficient to give its values on single element sets, P({ωi }) =
pi , by choosing pi ∈ (0, 1] such that i=1 pi = 1. We can then compute the
PN
expected price at the end of the period
N
X
E(S (1)) = S (1, ωi )pi ,
i=1

and the variance of the price

N
X
Var(S (1)) = (S (1, ωi ) − E(S (1)))2 pi .
i=1

Example 1.1
Assume that S (0) = 100 and
with probability 12 ,
(
120
S (1) =
90 with probability 12 .
Then E(S (1)) = 12 120 + 12 90 = 105 and Var(S (1)) = (120 − 105)2 12 +
(90 − 105)2 12 = 152 . Observe also that the
√ standard deviation, which is the
square root of the variance, is equal to Var(S (1)) = 15.
 1.1 Expected return 3

Exercise 1.1 Assume that U, D ∈ R are such that −1 < D < U.

Assume also that S has a binomial distribution, that is
!
  N
P S (1) = S (0) (1 + U) (1 + D)
k N−k
= pk (1 − p)N−k ,
k
for k ∈ {0, 1, . . . , N}. Compute E(S (1)) and Var(S (1)).

When S (1) is continuously distributed, with density function f : R → R,

then
Z ∞
E(S (1)) = x f (x)dx,
−∞

and
Z ∞
Var(S (1)) = (x − E(S (1)))2 f (x)dx.
−∞

Example 1.2
Assume that S (1) = S (0) exp (m + sZ) , where Z is a random variable with
standard normal distribution N(0, 1). This means that S (1) has lognormal
distribution. The density function of S (1) is equal to
x −m 2
1 (ln S (0) )
f (x) = √ e− 2s2 for x > 0,
xs 2π
and 0 for x ≤ 0. We can compute the expected price as
Z ∞
E(S (1)) = x f (x)dx
0
2
∞
1 − (ln S (0)2−m)
Z x

= √ e 2s dx
0 s 2π
Z ∞ !
sy+m 1 1 x
2
− y2
= S (0)e √ e dy (taking y = ln −m )
−∞ 2π s S (0)
Z ∞
s2 1 (y−s)2
= S (0)em+ 2 √ e− 2 dy
−∞ 2π
s2
= S (0)em+ 2 .
4 Risk and return

Exercise 1.2 Consider S (1) from Example 1.2. Show that

 2  2
Var(S (1)) = S (0)2 e s − 1 e2m+s .

While we may allow any probability space, we must make sure that
negative values of the random variable S (1) are excluded since negative
prices make no sense from the point of view of economics. This means
that the distribution of S (1) has to be supported on [0, +∞) (meaning that
P(S (1) ≥ 0) = 1).
The return (also called the rate of return) on the investment S is a ran-
dom variable K : Ω → R, defined as
S (1) − S (0)
K= .
S (0)
By the linearity of mathematical expectation, the expected (or mean) re-
turn is given by
E(S (1)) − S (0)
E(K) = .
S (0)
We introduce the convention of using the Greek letter µ for expectations of
various random returns
µ = E(K),

with various subscripts indicating the context, if necessary.

The relationships between the prices and returns can be written as

S (1) = S (0)(1 + K),

E(S (1)) = S (0)(1 + µ),

which illustrates the possibility of reversing the approach: given the returns
we can find the prices.
The requirement that S (1) is nonnegative implies that we must have
K ≥ −1. This in particular excludes the possibility of considering K with
Gaussian (normal) distribution.
At time 1 a dividend may be paid. In practice, after the dividend is paid,
the stock price drops by this amount, which is logical. Thus we have to
determine the price that includes the dividend; more precisely, we must
distinguish between the right to receive that price (the cum dividend price)
and the price after the dividend is paid (the ex dividend price). We assume
 1.2 Variance as a risk measure 5
that S (1) denotes the latter, hence the definition of the return has to be
modified to account for dividends:
S (1) + Div(1) − S (0)
K= .
S (0)
A bond is a special security that pays a certain sum of money, known
in advance, at maturity; this sum is the same in each state. The return on a
bond is not random (recall that we are dealing with a single time period).
Consider a bond paying a unit of home currency at time 1, that is B(1) = 1,
which is purchased for B(0) < 1. Then
1 − B(0)
R=
B(0)
defines the risk-free return. The bond price can be expressed as
1
B(0) = ,
1+R
giving the present value of a unit at time 1.

Exercise 1.3 Compute the expected returns for the stocks described
in Exercise 1.1 and Example 1.2.

Exercise 1.4 Assume that S (0) = 80 and that the ex dividend price
is
with probability 16 ,


 60
S (1) =  with probability 36 ,

80

with probability 26 .


 90

The company will pay out a constant dividend (independent of the fu-
ture stock price). Compute the dividend for which the expected return
on stock would be 20%.

1.2 Variance as a risk measure

The concept of risk in finance is captured in many ways. The basic and
most widely used one is concerned with risk as uncertainty of the unknown
6 Risk and return
future value of some quantity in question (here we are concerned with re-
turn). This uncertainty is understood as the scatter around some reference
point. A natural candidate for the reference value is the mathematical ex-
pectation (though other benchmarks are sometimes considered). The extent
of scatter is conveniently measured by the variance. This notion takes care
of two aspects of risk:
(i) The distances between possible values and the expectation.
(ii) The probabilities of attaining the various possible values.
Definition 1.3
By (the measure of) risk we mean the variance of the return
Var(K) = E(K − µ)2 = E(K 2 ) − µ2 ,
√
or the standard deviation Var(K).
The variance of the return can be computed from the variance of S (1),
!
S (1) − S (0)
Var(K) = Var
S (0)
1
= Var (S (1) − S (0))
S (0)2
1
= Var (S (1)) .
S (0)2
We use the Greek letter σ for standard deviations of various random
returns
σ = Var(K),
p

qualified by subscripts, as required.

Exercise 1.5 In a market with risk-free return R > 0, we buy a

‘leveraged’ stock S at time 0 with a mixture of cash and a loan at
rate R. To buy the stock for S (0) we use wS (0) of our own cash and
borrow (1 − w)S (0), for some w ∈ (0, 1). Denote the returns at time 1
on the stock and leveraged position by KS and Klev respectively.