- Số trang:
**514**| - Loại file:
**PDF**| - Lượt xem:
**20**| - Lượt tải:
**0**

tranphuong

Đã đăng **59088** tài liệu

Mô tả:

Introduction to the Economics and Mathematics of Financial Markets
Jakša Cvitanić and Fernando Zapatero
The MIT Press
Cambridge, Massachusetts
London, England
c 2004 Massachusetts Institute of Technology
All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means
(including photocopying, recording, or information storage and retrieval) without permission in writing from the
publisher.
This book was set in 10/13 Times Roman by ICC and was printed and bound in the United States of America.
Library of Congress Cataloging-in-Publication Data
Cvitanić, Jakša
Introduction to the economics and mathematics of ﬁnancial markets / Jakša Cvitanić and
Fernando Zapatero.
p.
cm.
Includes bibliographical references and index.
ISBN 0-262-03320-8
ISBN 0-262-53265-4 (International Student Edition)
1. Finance—Mathematical models—Textbooks. I. Zapatero, Fernando. II. Title.
HG106.C86 2004
332.632 01 515—dc22
2003064872
To Vesela, Lucia, Toni
and
Maitica, Nicolás, Sebastián
This page intentionally left blank
Contents
Preface
I
1
1.1
1.2
1.3
1.4
1.5
1.6
2
2.1
THE SETTING: MARKETS, MODELS, INTEREST RATES,
UTILITY MAXIMIZATION, RISK
xvii
1
Financial Markets
Bonds
1.1.1 Types of Bonds
1.1.2 Reasons for Trading Bonds
1.1.3 Risk of Trading Bonds
Stocks
1.2.1 How Are Stocks Different from Bonds?
1.2.2 Going Long or Short
Derivatives
1.3.1 Futures and Forwards
1.3.2 Marking to Market
1.3.3 Reasons for Trading Futures
1.3.4 Options
1.3.5 Calls and Puts
1.3.6 Option Prices
1.3.7 Reasons for Trading Options
1.3.8 Swaps
1.3.9 Mortgage-Backed Securities; Callable Bonds
Organization of Financial Markets
1.4.1 Exchanges
1.4.2 Market Indexes
Margins
1.5.1 Trades That Involve Margin Requirements
Transaction Costs
Summary
Problems
Further Readings
3
3
5
5
6
7
8
9
9
10
11
12
13
13
15
16
17
19
20
20
21
22
23
24
25
26
29
Interest Rates
Computation of Interest Rates
2.1.1 Simple versus Compound Interest; Annualized Rates
2.1.2 Continuous Interest
31
31
32
34
viii
2.2
2.3
3
3.1
3.2
3.3
3.4
3.5
Contents
Present Value
2.2.1 Present and Future Values of Cash Flows
2.2.2 Bond Yield
2.2.3 Price-Yield Curves
Term Structure of Interest Rates and Forward Rates
2.3.1 Yield Curve
2.3.2 Calculating Spot Rates; Rates Arbitrage
2.3.3 Forward Rates
2.3.4 Term-Structure Theories
Summary
Problems
Further Readings
35
36
39
39
41
41
43
45
47
48
49
51
Models of Securities Prices in Financial Markets
Single-Period Models
3.1.1 Asset Dynamics
3.1.2 Portfolio and Wealth Processes
3.1.3 Arrow-Debreu Securities
Multiperiod Models
3.2.1 General Model Speciﬁcations
3.2.2 Cox-Ross-Rubinstein Binomial Model
Continuous-Time Models
3.3.1 Simple Facts about the Merton-Black-Scholes Model
3.3.2 Brownian Motion Process
3.3.3 Diffusion Processes, Stochastic Integrals
3.3.4 Technical Properties of Stochastic Integrals∗
3.3.5 Itô’s Rule
3.3.6 Merton-Black-Scholes Model
3.3.7 Wealth Process and Portfolio Process
Modeling Interest Rates
3.4.1 Discrete-Time Models
3.4.2 Continuous-Time Models
Nominal Rates and Real Rates
3.5.1 Discrete-Time Models
3.5.2 Continuous-Time Models
53
54
54
55
57
58
58
60
62
62
63
66
67
69
74
78
79
79
80
81
81
83
Contents
3.6
3.7
4
4.1
4.2
4.3
4.4
4.5
4.6
ix
Arbitrage and Market Completeness
3.6.1 Notion of Arbitrage
3.6.2 Arbitrage in Discrete-Time Models
3.6.3 Arbitrage in Continuous-Time Models
3.6.4 Notion of Complete Markets
3.6.5 Complete Markets in Discrete-Time Models
3.6.6 Complete Markets in Continuous-Time Models∗
Appendix
3.7.1 More Details for the Proof of Itô’s Rule
3.7.2 Multidimensional Itô’s Rule
Summary
Problems
Further Readings
83
84
85
86
87
88
92
94
94
97
97
98
101
Optimal Consumption / Portfolio Strategies
Preference Relations and Utility Functions
4.1.1 Consumption
4.1.2 Preferences
4.1.3 Concept of Utility Functions
4.1.4 Marginal Utility, Risk Aversion, and Certainty Equivalent
4.1.5 Utility Functions in Multiperiod Discrete-Time Models
4.1.6 Utility Functions in Continuous-Time Models
Discrete-Time Utility Maximization
4.2.1 Single Period
4.2.2 Multiperiod Utility Maximization: Dynamic Programming
4.2.3 Optimal Portfolios in the Merton-Black-Scholes Model
4.2.4 Utility from Consumption
Utility Maximization in Continuous Time
4.3.1 Hamilton-Jacobi-Bellman PDE
Duality/Martingale Approach to Utility Maximization
4.4.1 Martingale Approach in Single-Period Binomial Model
4.4.2 Martingale Approach in Multiperiod Binomial Model
4.4.3 Duality/Martingale Approach in Continuous Time∗
Transaction Costs
Incomplete and Asymmetric Information
4.6.1 Single Period
103
103
104
105
107
108
112
112
113
114
116
121
122
122
122
128
128
130
133
138
139
139
x
Contents
4.6.2 Incomplete Information in Continuous Time∗
4.6.3 Power Utility and Normally Distributed Drift∗
Appendix: Proof of Dynamic Programming Principle
Summary
Problems
Further Readings
140
142
145
146
147
150
Risk
Risk versus Return: Mean-Variance Analysis
5.1.1 Mean and Variance of a Portfolio
5.1.2 Mean-Variance Efﬁcient Frontier
5.1.3 Computing the Optimal Mean-Variance Portfolio
5.1.4 Computing the Optimal Mutual Fund
5.1.5 Mean-Variance Optimization in Continuous Time∗
VaR: Value at Risk
5.2.1 Deﬁnition of VaR
5.2.2 Computing VaR
5.2.3 VaR of a Portfolio of Assets
5.2.4 Alternatives to VaR
5.2.5 The Story of Long-Term Capital Management
Summary
Problems
Further Readings
153
153
154
157
160
163
164
167
167
168
170
171
171
172
172
175
II
PRICING AND HEDGING OF DERIVATIVE SECURITIES
177
6
6.1
6.2
Arbitrage and Risk-Neutral Pricing
Arbitrage Relationships for Call and Put Options; Put-Call Parity
Arbitrage Pricing of Forwards and Futures
6.2.1 Forward Prices
6.2.2 Futures Prices
6.2.3 Futures on Commodities
Risk-Neutral Pricing
6.3.1 Martingale Measures; Cox-Ross-Rubinstein (CRR) Model
6.3.2 State Prices in Single-Period Models
6.3.3 No Arbitrage and Risk-Neutral Probabilities
179
179
184
184
186
187
188
188
192
193
4.7
5
5.1
5.2
6.3
Contents
6.4
7
7.1
7.2
7.3
7.4
7.5
xi
6.3.4 Pricing by No Arbitrage
6.3.5 Pricing by Risk-Neutral Expected Values
6.3.6 Martingale Measure for the Merton-Black-Scholes Model
6.3.7 Computing Expectations by the Feynman-Kac PDE
6.3.8 Risk-Neutral Pricing in Continuous Time
6.3.9 Futures and Forwards Revisited∗
Appendix
6.4.1 No Arbitrage Implies Existence of a Risk-Neutral Probability∗
6.4.2 Completeness and Unique EMM∗
6.4.3 Another Proof of Theorem 6.4∗
6.4.4 Proof of Bayes’ Rule∗∗
Summary
Problems
Further Readings
194
196
197
201
202
203
206
206
207
210
211
211
213
215
Option Pricing
Option Pricing in the Binomial Model
7.1.1 Backward Induction and Expectation Formula
7.1.2 Black-Scholes Formula as a Limit of the Binomial
Model Formula
Option Pricing in the Merton-Black-Scholes Model
7.2.1 Black-Scholes Formula as Expected Value
7.2.2 Black-Scholes Equation
7.2.3 Black-Scholes Formula for the Call Option
7.2.4 Implied Volatility
Pricing American Options
7.3.1 Stopping Times and American Options
7.3.2 Binomial Trees and American Options
7.3.3 PDEs and American Options∗
Options on Dividend-Paying Securities
7.4.1 Binomial Model
7.4.2 Merton-Black-Scholes Model
Other Types of Options
7.5.1 Currency Options
7.5.2 Futures Options
7.5.3 Exotic Options
217
217
217
220
222
222
222
225
227
228
229
231
233
235
236
238
240
240
242
243
xii
7.6
7.7
7.8
7.9
8
8.1
8.2
8.3
8.4
9
9.1
Contents
Pricing in the Presence of Several Random Variables
7.6.1 Options on Two Risky Assets
7.6.2 Quantos
7.6.3 Stochastic Volatility with Complete Markets
7.6.4 Stochastic Volatility with Incomplete Markets; Market Price
of Risk∗
7.6.5 Utility Pricing in Incomplete Markets∗
Merton’s Jump-Diffusion Model∗
Estimation of Variance and ARCH/GARCH Models
Appendix: Derivation of the Black-Scholes Formula
Summary
Problems
Further Readings
247
248
252
255
Fixed-Income Market Models and Derivatives
Discrete-Time Interest-Rate Modeling
8.1.1 Binomial Tree for the Interest Rate
8.1.2 Black-Derman-Toy Model
8.1.3 Ho-Lee Model
Interest-Rate Models in Continuous Time
8.2.1 One-Factor Short-Rate Models
8.2.2 Bond Pricing in Afﬁne Models
8.2.3 HJM Forward-Rate Models
8.2.4 Change of Numeraire∗
8.2.5 Option Pricing with Random Interest Rate∗
8.2.6 BGM Market Model∗
Swaps, Caps, and Floors
8.3.1 Interest-Rate Swaps and Swaptions
8.3.2 Caplets, Caps, and Floors
Credit/Default Risk
Summary
Problems
Further Readings
275
275
276
279
281
286
287
289
291
295
296
299
301
301
305
306
308
309
312
Hedging
Hedging with Futures
9.1.1 Perfect Hedge
313
313
313
256
257
260
262
265
267
268
273
Contents
9.2
9.3
9.4
9.5
9.1.2 Cross-Hedging; Basis Risk
9.1.3 Rolling the Hedge Forward
9.1.4 Quantity Uncertainty
Portfolios of Options as Trading Strategies
9.2.1 Covered Calls and Protective Puts
9.2.2 Bull Spreads and Bear Spreads
9.2.3 Butterﬂy Spreads
9.2.4 Straddles and Strangles
Hedging Options Positions; Delta Hedging
9.3.1 Delta Hedging in Discrete-Time Models
9.3.2 Delta-Neutral Strategies
9.3.3 Deltas of Calls and Puts
9.3.4 Example: Hedging a Call Option
9.3.5 Other Greeks
9.3.6 Stochastic Volatility and Interest Rate
9.3.7 Formulas for Greeks
9.3.8 Portfolio Insurance
Perfect Hedging in a Multivariable Continuous-Time Model
Hedging in Incomplete Markets
Summary
Problems
Further Readings
10 Bond Hedging
10.1 Duration
10.1.1 Deﬁnition and Interpretation
10.1.2 Duration and Change in Yield
10.1.3 Duration of a Portfolio of Bonds
10.2 Immunization
10.2.1 Matching Durations
10.2.2 Duration and Immunization in Continuous Time
10.3 Convexity
Summary
Problems
Further Readings
xiii
314
316
317
317
318
318
319
321
322
323
325
327
327
330
332
333
333
334
335
336
337
340
341
341
341
345
346
347
347
350
351
352
352
353
xiv
Contents
11 Numerical Methods
11.1 Binomial Tree Methods
11.1.1 Computations in the Cox-Ross-Rubinstein Model
11.1.2 Computing Option Sensitivities
11.1.3 Extensions of the Tree Method
11.2 Monte Carlo Simulation
11.2.1 Monte Carlo Basics
11.2.2 Generating Random Numbers
11.2.3 Variance Reduction Techniques
11.2.4 Simulation in a Continuous-Time Multivariable Model
11.2.5 Computation of Hedging Portfolios by Finite Differences
11.2.6 Retrieval of Volatility Method for Hedging and
Utility Maximization∗
11.3 Numerical Solutions of PDEs; Finite-Difference Methods
11.3.1 Implicit Finite-Difference Method
11.3.2 Explicit Finite-Difference Method
Summary
Problems
Further Readings
355
355
355
358
359
361
362
363
364
367
370
III
381
EQUILIBRIUM MODELS
12 Equilibrium Fundamentals
12.1 Concept of Equilibrium
12.1.1 Deﬁnition and Single-Period Case
12.1.2 A Two-Period Example
12.1.3 Continuous-Time Equilibrium
12.2 Single-Agent and Multiagent Equilibrium
12.2.1 Representative Agent
12.2.2 Single-Period Aggregation
12.3 Pure Exchange Equilibrium
12.3.1 Basic Idea and Single-Period Case
12.3.2 Multiperiod Discrete-Time Model
12.3.3 Continuous-Time Pure Exchange Equilibrium
12.4 Existence of Equilibrium
12.4.1 Equilibrium Existence in Discrete Time
371
373
374
376
377
378
380
383
383
383
387
389
389
389
389
391
392
394
395
398
399
Contents
xv
12.4.2 Equilibrium Existence in Continuous Time
12.4.3 Determining Market Parameters in Equilibrium
Summary
Problems
Further Readings
400
403
406
406
407
13 CAPM
13.1 Basic CAPM
13.1.1 CAPM Equilibrium Argument
13.1.2 Capital Market Line
13.1.3 CAPM formula
13.2 Economic Interpretations
13.2.1 Securities Market Line
13.2.2 Systematic and Nonsystematic Risk
13.2.3 Asset Pricing Implications: Performance Evaluation
13.2.4 Pricing Formulas
13.2.5 Empirical Tests
13.3 Alternative Derivation of the CAPM∗
13.4 Continuous-Time, Intertemporal CAPM∗
13.5 Consumption CAPM∗
Summary
Problems
Further Readings
409
409
409
411
412
413
413
414
416
418
419
420
423
427
430
430
432
14
14.1
14.2
14.3
433
433
436
438
438
439
441
442
445
445
445
Multifactor Models
Discrete-Time Multifactor Models
Arbitrage Pricing Theory (APT)
Multifactor Models in Continuous Time∗
14.3.1 Model Parameters and Variables
14.3.2 Value Function and Optimal Portfolio
14.3.3 Separation Theorem
14.3.4 Intertemporal Multifactor CAPM
Summary
Problems
Further Readings
xvi
Contents
15 Other Pure Exchange Equilibria
15.1 Term-Structure Equilibria
15.1.1 Equilibrium Term Structure in Discrete Time
15.1.2 Equilibrium Term Structure in Continuous Time; CIR Model
15.2 Informational Equilibria
15.2.1 Discrete-Time Models with Incomplete Information
15.2.2 Continuous-Time Models with Incomplete Information
15.3 Equilibrium with Heterogeneous Agents
15.3.1 Discrete-Time Equilibrium with Heterogeneous Agents
15.3.2 Continuous-Time Equilibrium with Heterogeneous Agents
15.4 International Equilibrium; Equilibrium with Two Prices
15.4.1 Discrete-Time International Equilibrium
15.4.2 Continuous-Time International Equilibrium
Summary
Problems
Further Readings
447
447
447
449
451
451
454
457
458
459
461
462
463
466
466
467
16 Appendix: Probability Theory Essentials
16.1 Discrete Random Variables
16.1.1 Expectation and Variance
16.2 Continuous Random Variables
16.2.1 Expectation and Variance
16.3 Several Random Variables
16.3.1 Independence
16.3.2 Correlation and Covariance
16.4 Normal Random Variables
16.5 Properties of Conditional Expectations
16.6 Martingale Deﬁnition
16.7 Random Walk and Brownian Motion
469
469
469
470
470
471
471
472
472
474
476
476
References
Index
479
487
Preface
Why We Wrote the Book
The subject of ﬁnancial markets is fascinating to many people: to those who care about
money and investments, to those who care about the well-being of modern society, to those
who like gambling, to those who like applications of mathematics, and so on. We, the
authors of this book, care about many of these things (no, not the gambling), but what
we care about most is teaching. The main reason for writing this book has been our belief
that we can successfully teach the fundamentals of the economic and mathematical aspects
of ﬁnancial markets to almost everyone (again, we are not sure about gamblers). Why are
we in this teaching business instead of following the path of many of our former students,
the path of making money by pursuing a career in the ﬁnancial industry? Well, they don’t
have the pleasure of writing a book for the enthusiastic reader like yourself!
Prerequisites
This text is written in such a way that it can be used at different levels and for different groups
of undergraduate and graduate students. After the ﬁrst, introductory chapter, each chapter
starts with sections on the single-period model, goes to multiperiod models, and ﬁnishes
with continuous-time models. The single-period and multiperiod models require only basic
calculus and an elementary introductory probability/statistics course. Those sections can
be taught to third- and fourth-year undergraduate students in economics, business, and
similar ﬁelds. They could be taught to mathematics and engineering students at an even
earlier stage. In order to be able to read continuous-time sections, it is helpful to have been
exposed to an advanced undergraduate course in probability. Some material needed from
such a probability course is brieﬂy reviewed in chapter 16.
Who Is It For?
The book can also serve as an introductory text for graduate students in ﬁnance, ﬁnancial economics, ﬁnancial engineering, and mathematical ﬁnance. Some material from continuoustime sections is, indeed, usually considered to be graduate material. We try to explain much
of that material in an intuitive way, while providing some of the proofs in appendixes to
the chapters. The book is not meant to compete with numerous excellent graduate-level
books in ﬁnancial mathematics and ﬁnancial economics, which are typically written in a
mathematically more formal way, using a theorem-proof type of structure. Some of those
more advanced books are mentioned in the references, and they present a natural next step
in getting to know the subject on a more theoretical and advanced level.
xviii
Preface
Structure of the Book
We have divided the book into three parts. Part I goes over the basic securities, organization
of ﬁnancial markets, the concept of interest rates, the main mathematical models, and
ways to measure in a quantitative way the risk and the reward of trading in the market.
Part II deals with option pricing and hedging, and similar material is present in virtually
every recent book on ﬁnancial markets. We choose to emphasize the so-called martingale,
probabilistic approach consistently throughout the book, as opposed to the differentialequations approach or other existing approaches. For example, the one proof of the BlackScholes formula that we provide is done calculating the corresponding expected value.
Part III is devoted to one of the favorite subjects of ﬁnancial economics, the equilibrium
approach to asset pricing. This part is often omitted from books in the ﬁeld of ﬁnancial
mathematics, having fewer direct applications to option pricing and hedging. However, it is
this theory that gives a qualitative insight into the behavior of market participants and how
the prices are formed in the market.
What Can a Course Cover?
We have used parts of the material from the book for teaching various courses at the University of Southern California: undergraduate courses in economics and business, a masterslevel course in mathematical ﬁnance, and option and investment courses for MBA students.
For example, an undergraduate course for economics/business students that emphasizes
option pricing could cover the following (in this order):
The ﬁrst three chapters without continuous-time sections; chapter 10 on bond hedging
could also be done immediately after chapter 2 on interest rates
•
The ﬁrst two chapters of part II on no-arbitrage pricing and option pricing, without most
of the continuous-time sections, but including basic Black-Scholes theory
•
•
Chapters on hedging in part II, with or without continuous-time sections
The mean-variance section in chapter 5 on risk; chapter 13 on CAPM could also be done
immediately after that section
•
If time remains, or if this is an undergraduate economics course that emphasizes
equilibrium/asset pricing as opposed to option pricing, or if this is a two-semester course,
one could also cover
•
discrete-time sections in chapter 4 on utility.
•
discrete-time sections in part III on equilibrium models.
Preface
xix
Courses aimed at more mathematically oriented students could go very quickly through
the discrete-time sections, and instead spend more time on continuous-time sections. A
one-semester course would likely have to make a choice: to focus on no-arbitrage option
pricing methods in part II or to focus on equilibrium models in part III.
Web Page for This Book, Excel Files
The web page http://math.usc.edu/∼cvitanic/book.html will be regularly updated with
material related to the book, such as corrections of typos. It also contains Microsoft Excel
ﬁles, with names like ch1.xls. That particular ﬁle has all the ﬁgures from chapter 1, along
with all the computations needed to produce them. We use Excel because we want the reader
to be able to reproduce and modify all the ﬁgures in the book. A slight disadvantage of this
choice is that our ﬁgures sometimes look less professional than if they had been done by a
specialized drawing software. We use only basic features of Excel, except for Monte Carlo
simulation for which we use the Visual Basic programming language, incorporated in Excel.
The readers are expected to learn the basic features of Excel on their own, if they are not
already familiar with them. At a few places in the book we give “Excel Tips” that point out
the trickier commands that have been used for creating a ﬁgure. Other, more mathematically
oriented software may be more efﬁcient for longer computations such as Monte Carlo, and
we leave the choice of the software to be used with some of the homework problems to the
instructor or the reader. In particular, we do not use any optimization software or differential
equations software, even though the instructor could think of projects using those.
Notation
Asterisk Sections and problems designated by an asterisk are more sophisticated in mathematical terms, require extensive use of computer software, or are otherwise somewhat
unusual and outside of the main thread of the book. These sections and problems could
be skipped, although we suggest that students do most of the problems that require use of
computers.
Dagger End-of-chapter problems that are solved in the student’s manual are preceded by
a dagger.
Greek Letters We use many letters from the Greek alphabet, sometimes both lowercase
and uppercase, and we list them here with appropriate pronunciation: α (alpha), β (beta),
γ , (gamma), δ, (delta), ε (epsilon), ζ (zeta), η (eta), θ (theta), λ (lambda), μ (mu),
ξ (xi), π, (pi), ω, (omega), ρ (rho), σ, (sigma), τ (tau), ϕ, (phi).
xx
Preface
Acknowledgments
First and foremost, we are immensely grateful to our families for the support they provided
us while working on the book. We have received great help and support from the staff of our
publisher, MIT Press, and, in particular, we have enjoyed working with Elizabeth Murry,
who helped us go through the writing and production process in a smooth and efﬁcient
manner. J. C.’s research and the writing of this book have been partially supported by
National Science Foundation grant DMS-00-99549. Some of the continuous-time sections
in parts I and II originated from the lecture notes prepared in summer 2000 while J. C. was
visiting the University of the Witwatersrand in Johannesburg, and he is very thankful to
his host, David Rod Taylor, the director of the Mathematical Finance Programme at Wits.
Numerous colleagues have made useful comments and suggestions, including Krzysztof
Burdzy, Paul Dufresne, Neil Gretzky, Assad Jalali, Dmitry Kramkov, Ali Lazrak, Lionel
Martellini, Adam Ostaszewski, Kaushik Ronnie Sircar, Costis Skiadas, Halil Mete Soner,
Adam Speight, David Rod Taylor, and Mihail Zervos. In particular, D. Kramkov provided
us with proofs in the appendix of chapter 6. Some material on continuous-time utility
maximization with incomplete information is taken from a joint work with A. Lazrak and
L. Martellini, and on continuous-time mean-variance optimization from a joint work with
A. Lazrak. Moreover, the following students provided their comments and pointed out
errors in the working manuscript: Paula Guedes, Frank Denis Hiebsch, and Chulhee Lee.
Of course, we are solely responsible for any remaining errors.
A Prevailing Theme: Pricing by Expected Values
Before we start with the book’s material, we would like to give a quick illustration here in
the preface of a connection between a price of a security and the optimal trading strategy of
an investor investing in that security. We present it in a simple model, but this connection is
present in most market models, and, in fact, the resulting pricing formula is of the form that
will follow us through all three parts of this book. We will repeat this type of argument later
in more detail, and we present it early here only to give the reader a general taste of what
the book is about. The reader may want to skip the following derivation, and go directly to
equation (0.3).
Consider a security S with today’s price S(0), and at a future time 1 its price S(1) either
has value s u with probability p, or value s d with probability 1 − p. There is also a risk-free
security that returns 1 + r dollars at time 1 for every dollar invested today. We assume that
s d < (1 + r )S(0) < s u . Suppose an investor has initial capital x, and has to decide how
many shares δ of security S to hold, while depositing the rest of his wealth in the bank
Preface
xxi
account with interest rate r . In other words, his wealth X (1) at time 1 is
X (1) = δS(1) + [x − δS(0)](1 + r )
The investor wants to maximize his expected utility
E[U (X (1))] = pU (X u ) + (1 − p)U (X d )
where U is a so-called utility function, while X u , X d is his ﬁnal wealth in the case S(1) = s u ,
S(1) = s d , respectively. Substituting for these values, taking the derivative with respect to
δ and setting it equal to zero, we get
pU (X u )[s u − S(0)(1 + r )] + (1 − p)U (X d )[s d − S(0)(1 + r )] = 0
The left-hand side can be written as E[U (X (1)){S(1) − S(0)(1 + r )}], which, when made
equal to zero, implies, with arbitrary wealth X replaced by optimal wealth X̂ ,
U ( X̂ (1)) S(1)
(0.1)
S(0) = E
E(U [X̂ (1)]) 1 + r
If we denote
Z (1) :=
U (X̂ (1))
E{U (X̂ (1))}
we see that the today’s price of our security S is given by
S(1)
S(0) = E Z (1)
1+r
(0.2)
(0.3)
We will see that prices of most securities (with some exceptions, like American options)
in the models of this book are of this form: the today’s price S(0) is an expected value of
the future price S(1), multiplied (“discounted”) by a certain random factor. Effectively, we
get the today’s price as a weighted average of the discounted future price, but with weights
that depend on the outcomes of the random variable Z (1). Moreover, in standard optionpricing models (having a so-called completeness property) we will not need to use utility
functions, since Z (1) will be independent of the investor’s utility. The random variable Z (1)
is sometimes called change of measure, while the ratio Z (1)/(1 + r ) is called state-price
density, stochastic discount factor, pricing kernel, or marginal rate of substitution,
depending on the context and interpretation. There is another interpretation of this formula,
using a new probability; hence the name “change of (probability) measure.” For example,
if, as in our preceding example, Z (1) takes two possible values Z u (1) and Z d (1) with
xxii
Preface
probabilities p, 1 − p, respectively, we can deﬁne
p ∗ := p Z u (1),
1 − p ∗ = (1 − p)Z d (1)
The values of Z (1) are such that p ∗ is a probability, and we interpret p ∗ and 1 − p ∗ as
modiﬁed probabilities of the movements of asset S. Then, we can write equation (0.3) as
S(1)
(0.4)
S(0) = E ∗
1+r
where E ∗ denotes the expectation under the new probabilities, p ∗ , 1 − p ∗ . Thus the price
today is the expected value of the discounted future value, where the expected value is
computed under a special, so-called risk-neutral probability, usually different from the
real-world probability.
Final Word
We hope that we have aroused your interest about the subject of this book. If you turn out to
be a very careful reader, we would be thankful if you could inform us of any remaining
typos and errors that you ﬁnd by sending an e-mail to our current e-mail addresses. Enjoy
the book!
Jakša Cvitanić and Fernando Zapatero
E-mail addresses: cvitanic@math.usc.edu, zapatero@usc.edu

- Xem thêm -