Quantitive methods in derivatives pricing

Quantitative Methods in

derivatives pricing


Founded in 1807, John Wiley & Sons is the oldest independent publishing company in the United States. With offices in North America,
Europe, Australia, and Asia, Wiley is globally committed to developing and
marketing print and electronic products and services for our customers’
professional and personal knowledge and understanding.
The Wiley Finance series contains books written specifically for
finance and investment professionals as well as sophisticated individual
investors and their financial advisors. Book topics range from portfolio
management to e-commerce, risk management, financial engineering, valuation and financial instrument analysis, as well as much more.
For a list of available titles, visit our Web site at www.WileyFinance.com.


Quantitative Methods in

derivatives pricing
An Introduction
to Computational Finance

Domingo Tavella

John Wiley & Sons, Inc.


Copyright © 2002 by Domingo Tavella. All rights reserved.
Published by John Wiley & Sons, Inc., Hoboken, New Jersey
Published simultaneously in Canada

No part of this publication may be reproduced, stored in a retrieval system, or
transmitted in any form or by any means, electronic, mechanical, photocopying,
recording, scanning, or otherwise, except as permitted under Section 107 or 108 of
the 1976 United States Copyright Act, without either the prior written permission
of the Publisher, or authorization through payment of the appropriate per-copy fee
to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA
01923, 978-750-8400, fax 978-750-4470, or on the web at www.copyright.com.
Requests to the Publisher for permission should be addressed to the Permissions
Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030,
201-748-6011, fax 201-748-6008, e-mail:
Limit of Liability/Disclaimer of Warranty: While the publisher and author have
used their best efforts in preparing this book, they make no representations or
warranties with respect to the accuracy or completeness of the contents of this book
and specifically disclaim any implied warranties of merchantability or fitness for a
particular purpose. No warranty may be created or extended by sales
representatives or written sales materials. The advice and strategies contained
herein may not be suitable for your situation. You should consult with a
professional where appropriate. Neither the publisher nor author shall be liable for
any loss of profit or any other commercial damages, including but not limited to
special, incidental, consequential, or other damages.
For general information on our other products and services, or technical support,
please contact our Customer Care Department within the United States at
800-762-2974, outside the United States at 317-572-3993 or fax 317-572-4002.
Wiley also publishes its books in a variety of electronic formats. Some content that
appears in print may not be available in electronic books.
ISBN 0-471-39447-5
Printed in the United States of America
10 9 8 7 6 5 4 3 2 1

To
Rudolph and Natalie



preface

he emergence of computational finance as a discipline in its own right is
relatively recent. The first international conference on computational
finance took place in 1995 at Stanford University, where, as far as the
author is aware, the name for this new discipline was coined. The Journal
of Computational Finance was created shortly thereafter, and its success
and popularity soon demonstrated that there was a body of work of sufficient mass and extent to rightfully configure the emergence of a new discipline, complete with its views, paradigms, and methods.
The use of computational methods for solving engineering problems
allows us to analyze systems of such scale and complexity that their analysis would not be conceivable through empirical study through purely analytical means. Computational chemistry, computational fluid dynamics, the
numerical simulation of astronomical structures, structural analysis, and so
on, are examples where the use of sophisticated numerical techniques
allows us to gain a type of understanding of the nature of the problem that
could not be gained otherwise.
Just as physicists and engineers solve problems by solving so-called
“conservation equations,” financial engineers price financial instruments by
solving their corresponding pricing equations. The conservation equations
of physics establish relationships between the rates of convection, diffusion,
creation, and disappearance of mass, momentum, and energy. Typically,
these relationships are in the form of partial differential equations (PDEs).
The pricing equations of financial instruments state the way the price of the
instrument depends on time and the value of other instruments or processes.
Typically, these pricing equations are also PDEs.
While the conservation equations of physics are derived by considering
the detailed balance of mass, momentum, and energy flows, the pricing

equations of financial instruments are derived by considering arbitrage
(rather, the absence of arbitrage) and expectations. Are there significant differences in the computational challenges presented by physical problems
and financial problems? Although this question is hard to answer with generality, there are observations we can make about how financial engineers
perceive these challenges vis-à-vis their colleagues in other disciplines. In
engineering fields such as structural analysis or fluid dynamics, engineers
can deal with a relatively well-established set of PDEs with which he or she

T

vii


viii

PREFACE

can solve a very large number of problems by simply changing the boundary conditions. This relative consensus and stability of the mathematical
framework makes it possible to develop large and flexible software systems
to implement particular solution approaches applicable to particular areas
of engineering. These systems can be used to solve a large variety of problems by simply changing boundary values and the way boundaries are
treated. These systems will typically implement a particular numerical
approach, such as finite elements or finite differences, applicable to large
classes of problems. Structural engineers, for example, can deal with a large
array of problems using a single computational methodology, such as finite
elements. Aerodynamicists can work on projects ranging from small aircraft to reentry vehicles and still use the same methodology, such as finite
differences.
This situation is significantly different in financial engineering. The
pricing of financial contracts is not just a matter of repeatedly applying the
same numerical methodology with different boundary conditions. In many
cases, the pricing equation is very specific to the particular financial instrument being considered. In other cases, the pricing equation is not known.

Yet in other cases the pricing equation is extremely ill-suited for certain
types of numerical techniques. This means that the financial engineer must
be fluent in a number of computational techniques appropriate for dealing
with different instruments.
This book is designed as a graduate textbook in financial engineering.
It was motivated by the need to present the main techniques used in quantitative pricing in a single source adequate for Master level students. Students
are expected to have some background in algebra, elementary statistics, calculus, and elementary techniques of financial pricing, such as binomial
trees and simple Monte Carlo simulation. The book includes a brief introduction to the fundamentals of stochastic calculus.
The book is divided into seven chapters covering an introduction to
stochastic calculus, a summary of asset pricing theory, simulation applied
to pricing, and pricing using finite difference solutions. The topic of trees as
a tool for pricing is touched on at the end of the finite differences chapter.
Although trees are a popular pricing technique, finite differences, of which
trees are a particularly simple case, are a far more powerful and flexible
approach. Significant effort is dedicated to the fundamentals of early exercise simulation. This methodology is rapidly taking the lead as a preferred
way to price highly dimensional early exercise instruments.
Chapter 1 is a brief introduction to single-period pricing with the
objective of motivating the idea that the price of a financial instrument is
given by an expectation.


Preface

ix

Chapter 2 is a summary introduction to the basic elements of stochastic
calculus. The material is presented in a nonrigorous way and should be
easy to follow by anyone with a basic background in elementary calculus.
Chapter 3 is a brief description of pricing in continuous time, where
the main objectives are to more precisely determine the price as an expectation under a suitable measure and to derive the relevant pricing equation.

Chapter 4 focuses on the generation of scenarios for simulation. In
practical implementations of simulation, the generation of scenarios of
appropriate quality is essential. Issues of accuracy are discussed in detail.
Chapter 5 is dedicated to simulation applied to computing expectations
for European pricing. This chapter gives a summary with selected case
studies of the main approaches that have demonstrated practical value in
financial pricing.
Chapter 6 deals with simulation applied to early exercise pricing. At
the time of this writing, this is a rapidly evolving subject. For this reason,
this chapter must be viewed as an update of the most established aspects of
simulation for early exercise pricing. The chapter presents a brief historical
account of the various techniques, but the emphasis is on linear squares
Monte Carlo, the technique that has marked a breakthrough in this area.
Chapter 7 summarizes the use of finite differences in option pricing.
The material is presented in a concise manner, with an emphasis on the fundamentals.

DOMINGO TAVELLA
San Francisco, California
March 2002



acknowledgments

During the preparation of this book, I benefited from discussions with colleagues. I am especially indebted to Dr. Ervin Zhao for his valuable suggestions on the manuscript. My thanks are also due to Dr. Joshua Rosenberg
and Mr. Didier Vermeiren for their helpful comments.
D.T.

xi

contents

CHAPTER 1
Arbitrage and Pricing
The Pricing Problem
Arbitrage
State Prices
Present Value as an Expectation of Future Values

CHAPTER 2
Fundamentals of Stochastic Calculus
Basic Definitions
Probability Space
Sample Space
Filtration and the Revelation of Information
Probability Measure
Random Variables
Stochastic Process
Measurable Stochastic Process
Adapted Process
Conditional Expectation
Martingales
Wiener Process
First Variation of a Differentiable Function
First Variation of the Wiener Process
Second Variation of a Differentiable Function
Second Variation of the Wiener Process
Products of Infinitesimal Increments of Wiener Processes

Stochastic Integrals
Mean Square Limit
Ito Integral
Properties of the Ito Integral
Ito Processes

1
1
2
2
4

9
9
9
10
10
11
12
12
12
13
13
13
13
15
15
15
16
16

18
18
19
19
21

xiii


xiv

CONTENTS
Multidimensional Processes
Multidimensional Wiener Processes
Multidimensional Ito Processes
Ito’s Lemma
Multidimensional Ito’s Lemma
Stochastic Differential Equations
Moments of SDE Solutions
SDE Commonly Used in Finance
The Markov Property of Solutions of SDE
The Feynman-Kac Theorem
Measure Changes
Girsanov Theorem
Martingale Representation Theorem
Processes with Jumps
The Poisson Jump Model
Defining a Pure Jump Process
Defining a Jump-Diffusion Process
Ito’s Lemma in the Presence of Jumps

CHAPTER 3
Pricing in Continuous Time
One-Dimensional Risk Neutral Pricing
Multidimensional Market Model
Extension to Other Normalizing Assets
Deriving Risk-Neutralized Processes
The Pricing Equation
European Derivatives
Hedging Portfolio Approach
Feynman-Kac Approach
The Pricing Equation in the Presence of Jumps
An Application of Jump Processes: Credit Derivatives
Defaultable Bonds
Full Protection Credit Put
American Derivatives
Relationship between European and American Derivatives
American Options as Dynamic Optimization Problems
Conditions at Exercise Boundaries
Linear Complementarity Formulation of
American Option Pricing
Path Dependency
Discrete Sampling of Path Dependency

22
22
23
24
25
27

28
29
30
31
33
35
36
36
37
37
38
38

41
42
47
51
53
56
57
58
60
62
63
65
66
67
68
69
70

72
73
74


Contents

xv

CHAPTER 4
Scenario Generation

77

Scenario Nomenclature
Scenario Construction
Exact Solution Advancement
Sampling from the Joint Distribution of the Random Process
Generating Scenarios by Numerical Integration of the
Stochastic Differential Equations
Brownian Bridge
Brownian Bridge Construction
Generating Scenarios with Brownian Bridges
Joint Normals by the Choleski Decomposition Approach
Quasi-Random Sequences
The Concept of Discrepancy
Discrepancy and Convergence: The Koksma-Hlawka
Inequality
Proper Use of Quasi-Random Sequences
Interest Rate Scenarios

HJM for Instantaneous Forwards
LIBOR Rate Scenarios
Principal Component Analysis to Approximate Correlation
Matrices

CHAPTER 5
European Pricing with Simulation
Roles of Simulation in Finance
Monte Carlo in Pricing
Monte Carlo in Risk Management
The Workflow of Pricing with Monte Carlo
Estimators
Estimation of the Mean
Estimation of the Variance
Simulation Efficiency
Increasing Simulation Efficiency
Antithetic Variates
Efficiency of Antithetic Variates
Control Variates
Efficiency of Control Variates
Case Study: Application of Control Variates to Discretely
Sampled Step-Up Barrier Options

78
79
80
81
86
93
94

95
100
102
109
109
110
113
113
115
118

121
121
122
123
124
125
125
127
130
131
133
134
135
137
137


xvi

CONTENTS
Importance Sampling
Optimal Importance Density
Applying the Girsanov Theorem to Importance Sampling:
European Call Option
Importance Sampling by Direct Modeling of the Importance
Density: Credit Put
Moment Matching
Stratification
Stratified Standard Normals in One Dimension
Latin Hypercube Sampling
Case Study: Latin Hypercube Sampling Applied to Exotic
Basket Option
Effect of Discretization on Accuracy and the Emergence of
Computational Barriers
Discretization Error for the Log-Normal Process
Discretization Error and Computational Barriers
for a European Call

CHAPTER 6
Simulation for Early Exercise
The Basic Difficulty in Pricing Early Exercise with Simulation
Simulation Applied to Early Exercise
Dealing with Estimator Bias
Path-Bundling Algorithms
State Stratification Algorithms
Simulated Recombining Lattices
Simulated Bushy Trees
Least Squares Monte Carlo
Least Squares and Conditional Expectation

LSMC Algorithm
The Moneyness Criterion
Implementation Considerations
Case Study 1: Bermudan Call on Best-of-Three Assets
Specification
Basis Functions
The Benchmark
Numerical Results
Case Study 2: Bermudan Swaption
Specification
Scenario Generation
Basis Functions

140
142
143
149
152
155
159
161
162
164
167
171

177
177
179
180

183
185
186
187
188
189
192
196
197
198
198
199
200
201
202
203
203
204


Contents
The Benchmark
Numerical Results

CHAPTER 7
Pricing with Finite Differences
Fundamentals
Finite Difference Strategy
Constructing Finite Difference Space Discretizations
Implementation of Space Discretization

The Mechanics of Finite Differences
Stability and Accuracy Analysis
Analysis of Specific Algorithms
Time Advancement and Linear Solvers
Direct Solvers
Iterative Solvers
Finite Difference Approach for Early Exercise
The Linear Complementarity Problem
Boundary Conditions
Implementation of Boundary Conditions
Solving Alternative PDEs at Boundaries
Barriers
Coordinate Transformation Versus Process Transformation
Discrete Sampling of Barriers
Coordinate Transformations
Implementation of Coordinate Transformations
Discrete Events and Path Dependency
Displacement Shocks
Path Dependency and Discrete Sampling
Trees, Lattices, and Finite Differences
Connection Between the CRR Binomial Tree and Finite
Differences
Connection Between the Jarrow and Rudd Binomial Tree and
Finite Differences
Implications of the Correspondence Between Trees and Finite
Differences

xvii
205
206

207
207
210
212
213
215
217
224
228
229
231
233
233
237
238
240
242
243
247
252
254
259
260
262
267
268
270
272

BIBLIOGRAPHY

273

INDEX

277



CHAPTER

1

Arbitrage and Pricing

he purpose of this short chapter is to motivate the notion that the price
of a financial instrument is expressed in the form of an expectation of
suitably discounted future values or cash flows. To accomplish this, we will
work in a single period framework, where we will show that the price of a
security is an expectation where the probabilities used to compute the
expectation are determined by a normalizing asset, known as the numeraire
asset. We will not elaborate on discrete time pricing beyond this initial
chapter. The reader interested in additional details of discrete time pricing
can consult the excellent work by Dotham (1990). The reason we will not
dwell on discrete time modeling is that the power of the numerical pricing
methods we will consider originates in their application to continuous time
models.

T

THE PRICING PROBLEM
We will obtain intuitive derivation of pricing formulation by the following
line of reasoning.
■ Absence of arbitrage implies the existence of state prices. State
prices are the values of elementary securities known as ArrowDebreu securities.
■ State prices, when properly rescaled by the values of other instruments
or portfolios of instruments, can be interpreted as probabilities.
■ The derivative’s price is an expectation with respect to a probability
measure determined by the rescaling of state prices (a probability measure assigns probabilities to outcomes.)
■ When the underlying processes that determine the derivative’s price are
Ito processes, the expectation can be expressed as the solution to a parabolic partial differential equation. This is the pricing equation.

1


2

QUANTITATIVE METHODS IN DERIVATIVES PRICING

Arbitrage
We will consider a market that at payoff time T may achieve one of S states.
Assume there are N traded securities, whose values at t = 0 are denoted by
Vn(0), n = 1,…, N, and whose payoffs at time t = T are indicated by
Fs,n(T), s = 1,…, S, n = 1,…, N. The matrix F˜ (T), whose elements are
Fs,n(T), is called the payoff matrix. Each column of the payoff matrix represents the payoffs of a given security for the different market states. Each row
represents the payoffs of the different securities for a given market state.
We now define the concept of Arrow-Debreu securities. We will use
this concept in establishing the arbitrage conditions in the discrete time
model. An Arrow-Debreu security is a security that pays $1 at time T if a

particular state materializes and pays $0, otherwise.
If at time t = 0, we purchase the jth Arrow-Debreu security in the amount
Fj,n(T), we will get a payoff at time T equal to Fj,n(T) if the jth state materializes,
and zero, otherwise. This means that if we purchase the jth Arrow-Debreu security in the amount Fj,n(T), we will match the payoff of the nth asset in state j.
If we purchase a portfolio of Arrow-Debreu securities, such that F1,n(T)
is the amount of Arrow-Debreu security 1, F2,n(T) is the amount of ArrowDebreu security 2, and … FS,n(T) is the amount of Arrow-Debreu security S,
we will match the payoffs of the nth asset in all states at time T. The value of
s=S
this portfolio is equal to ⌺ s=1 Fs, n(T)␲s, where ␲s is the value of the sth
Arrow-Debreu security. The present value of the nth asset must be equal to the
value of this portfolio, because their payoffs are the same. (See Equation 1.1.)
s=S

Vn( 0 ) =

∑ Fs, n( T )␲s,

n = 1,…, N

(1.1)

s=1

If this relationship were not satisfied, it would be possible to make a
riskless profit. If the portfolio of Arrow-Debreu securities were more valuable than the asset, we would short-sell the portfolio of Arrow-Debreu
securities and buy the asset. If the asset were more valuable than the portfolio of Arrow-Debreu securities, we would do the opposite. In either case,
the difference would be a riskless profit.

State Prices
The values of the Arrow-Debreu securities are called state prices. If somehow we can determine these prices, we can use them to price other securities whose payoffs are known. If we limit our definition of arbitrage to

the situation we described in the last section and we assume that there are


3

Arbitrage and Pricing

as many independent assets as there are possible market states, we can
find the state prices from observed asset prices by solving the algebraic
system
s=S

∑ Fs, n( T ) ␲s

= V n( 0 ), n = 1,…, N

(1.2)

s=1

where N = S.
If there is a market for Arrow-Debreu securities, solving this system
will give us their prices. If we know their prices, we can price any other
security whose payoffs are known. We could then argue that the existence
of state prices implies the absence of arbitrage, and vice versa. Since the
state prices are the values of securities with positive payoffs, the state prices
are positive.
A more precise definition of absence of arbitrage is to say that any
investment with nonnegative payoff in every possible market outcome at
a future time must have a nonnegative initial cost. Loosely speaking,

this statement simply says that we cannot get something for nothing.
Mathematically, this means that if we hold amounts xn, n = 1,…, N of
assets whose initial values are Vn(0), n = 1,…, N, we must have the
conditional:
n=N

n=N

If

∑

F s, n( T ) x n ‡ 0, s = 1,…, S, then

n=1

∑ V n( 0 )xn ‡ 0

n=1

With this formulation of arbitrage, it is possible to show that also in
the case where the number of market states is greater than the number of
securities, S ‡ N, absence of arbitrage implies the existence of positive state
prices ␲s ‡ 0, s = 1,…, S, such that
s=S

V n( 0 ) =

∑ ␲sFs, n( T )

(1.3)

s=1

(The proof of this statement uses arguments from operations research: for
details, please refer to Varian (1987) or Duffie (1996).) It is clear that each
␲s can be interpreted as the values of a security that has a payoff of $1 at
time T if the state s materializes and $0, otherwise. We can see this by setting the payoff matrix F equal to the identity matrix.


4

QUANTITATIVE METHODS IN DERIVATIVES PRICING

In summary, the present value, V(0), of a security with payoff Vs(T) at
time T, if state s materializes, is given by
s=S

V(0) =

∑ ␲sVs( T )

(1.4)

s=1

If there are as many market states as there are independent securities,
the state prices ␲i are unique and the market is called complete. If there are
more market states than independent securities, the market is called incomplete. In this case, the state prices are not unique.
Equation 1.4 is the starting point for pricing a derivative as an expectation of future values.

Present Value as an Expectation of Future Values
Consider two instruments whose present values are denoted by A(0) and
B(0) and whose payoff vectors are A(T) and B(T). We write down the ratio
of their present values, using the last equation, as:
A 1( T )
␲ 1B 1 ( T )
A(0)
- -------------------------- = --------------------------------------------------------------␲ 1B1( T ) + … + ␲ SB S( T ) B1( T )
B(0)
␲ SB S( T )
A S( T )
- --------------+ … --------------------------------------------------------------␲ 1B1( T ) + … + ␲ SB S( T ) BS ( T )

(1.5)

This expression can be written as
A(0 )
------------ =
B( 0)

i=S

A (T)

i ∑ pi -------------Bi ( T )

(1.6)

i=1

where

␲iBi( T )
p i = --------------------------------------------------------------␲ 1 B 1( T ) + … + ␲ S B S ( T )

(1.7)

Notice that the pi s are all nonnegative, and they add up to 1. Hence,
since there are as many pis as there are possible market outcomes, the pis can
be interpreted as probability masses. The market outcomes have probabilities
of occurrence of their own, which we call objective or market probabilities.
The probabilities we have just derived are different from the objective probability of the market outcomes. In fact, the market objective probabilities do


5

Arbitrage and Pricing

not appear explicitly in the derivation. We refer to the probabilities in
Equation 1.7 as induced by the asset in the denominator of Equation 1.5. The
asset in the denominator is called the numeraire asset.
If the asset in the denominator of Equation 1.5 does not vanish for the market outcomes of interest, the induced probabilities will be different from zero
for the outcomes where the objective probabilities are different from zero. A
probability measure assigns probabilities to outcomes. Probability measures
that assign probabilities with this property are called equivalent probability
measures. We can infer that absence of arbitrage means that the price of a
traded asset, normalized with the price of another traded asset or portfolio of
traded assets, equals the expectation of the normalized value at time T with
respect to a probability measure induced by the normalizing asset.

This means that the present value of an asset can be written as
B A(T)
A ( 0 ) = B ( 0 )E ------------B( T)

(1.8)

where EB denotes expectation with respect to probabilities induced by B.
If asset B is an investment of $1 that pays a known compound return r
at time T, we get the more familiar formula
Q

A ( 0 ) = E exp ( – rT )A ( T )
Q

= exp ( – rT )E A ( T )

(1.9)

where Q indicates that the expectation is taken with respect to probabilities
induced by the continuously compounded $1 investment. This familiar formula says that the present value of an asset with uncertain payoffs is the
discounted expectation of the payoffs (assuming that the interest rate is
known), where the probabilities of market outcomes are said to be risk
neutral.
How are the objective probabilities of market outcomes related to the
probabilities induced by the numeraire asset? This relationship is captured
by the so-called “Radon-Nikodym derivative,” defined as
pi
Z i = ------pM
i

(1.10)

where p M
i is the objective, or market probability mass, for the ith market
outcome. The Radon-Nikodym derivative has the property
M

E Z = 1

(1.11)


6

QUANTITATIVE METHODS IN DERIVATIVES PRICING

where EM indicates expectation with respect to the market or objective
measure. For any random variable X,
B

M

E X = E ( ZX )

(1.12)

We can now summarize our observations:
■ If the number of possible market outcomes is equal to the number of
independent assets with payoffs associated with these market outcomes, the market is called complete. A unique set of state prices
determines a unique probability measure induced by a normalizing

asset, and there is no arbitrage.
■ If the number of possible market outcomes is greater than the number of
independent assets with payoffs associated with these outcomes, the market
is called incomplete. State prices rule out arbitrage but are not unique and
there is a nonunique probability measure induced by a normalizing asset.
■ If there are more independent assets than market states, there are no
state prices, there are no probabilities induced by a normalizing asset,
and there is arbitrage.
So far, we have motivated the formulation of the pricing problem as an
expectation of future payoffs. This expectation is taken with respect to probabilities associated with a given normalizing asset. The goal of quantitative
pricing is to compute this expectation. As we will see, this expectation can be
computed according to different methodologies. Each methodology for computing the expectation gives rise to a different specialization of quantitative
financial pricing. The main approaches are as follows.
■ Direct analytical evaluation of the expectation: This approach may
give closed-form solutions. We will see some simple examples in
Chapter 3.
■ Numerical computation of the expectation by simulation: A variety of
Monte Carlo techniques can be used with varying degrees of success.
We will discuss these techniques in Chapters 5 and 6.
■ Transformation of the expectation into a partial differential equation
(PDE) or an integro-partial differential equation (IPDE): This allows us
to resort to the vast field of numerical analysis applied to parabolic
PDEs. We will discuss this in Chapter 7.
Before tackling the pricing problem with any particular methodology,
we must enrich the framework for formulating the expectation we dis-