An Introduction to Stochastic Calculus with Applications

to Finance

Ovidiu Calin
Department of Mathematics
Eastern Michigan University
Ypsilanti, MI 48197 USA
[email protected]
Preface

The goal of this work is to introduce elementary Stochastic Calculus to senior undergraduate
as well as to master students with Mathematics, Economics and Business majors. The author’s
goal was to capture as much as possible of the spirit of elementary Calculus, at which the stu-
dents have been already exposed in the beginning of their majors. This assumes a presentation
that mimics similar properties of deterministic Calculus, which facilitates the understanding of
more complicated concepts of Stochastic Calculus. Since deterministic Calculus books usually
start with a brief presentation of elementary functions, and then continue with limits, and other
properties of functions, we employed here a similar approach, starting with elementary stochas-
tic processes, different types of limits and pursuing with properties of stochastic processes. The
chapters regarding differentiation and integration follow the same pattern. For instance, there
is a product rule, a chain-type rule and an integration by parts in Stochastic Calculus, which
are modifications of the well-known rules from the elementary Calculus.
Since deterministic Calculus can be used for modeling regular business problems, in the
second part of the book we deal with stochastic modeling of business applications, such as
Financial Derivatives, whose modeling are solely based on Stochastic Calculus.
In order to make the book available to a wider audience, the rigor was sacrificed for clarity.
Most of the time we assumed maximal regularity conditions for which the computations hold
and the statements are valid. This will be found attractive by both Business and Economics
students, who might get lost otherwise in a very profound mathematical textbook where the
forest’s scenary is obscured by the sight of the trees.
An important feature of this textbook is the large number of solved problems and examples
from which will benefit both the beginner as well as the advanced student.
This book grew from a series of lectures and courses given by the author at the Eastern
Michigan University (USA), Kuwait University (Kuwait) and Fu-Jen University (Taiwan). Sev-
eral students read the first draft of these notes and provided valuable feedback, supplying a
list of corrections, which is by far exhaustive. Any typos or comments regarding the present
material are welcome.

The Author,
Ann Arbor, October 2012

i
ii O. Calin
Contents

I Stochastic Calculus 3

1 Basic Notions 5
1.1 Probability Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2 Sample Space . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3 Events and Probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.4 Random Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.5 Distribution Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.6 Basic Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.7 Independent Random Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
1.8 Integration in Probability Measure . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.9 Expectation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.10 Radon-Nikodym’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.11 Conditional Expectation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.12 Inequalities of Random Variables . . . . . . . . . . . . . . . . . . . . . . . . . . 16
1.13 Limits of Sequences of Random Variables . . . . . . . . . . . . . . . . . . . . . 22
1.14 Properties of Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
1.15 Stochastic Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

2 Useful Stochastic Processes 29

2.1 The Brownian Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.2 Geometric Brownian Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.3 Integrated Brownian Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
2.4 Exponential Integrated Brownian Motion . . . . . . . . . . . . . . . . . . . . . 37
2.5 Brownian Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.6 Brownian Motion with Drift . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.7 Bessel Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.8 The Poisson Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.8.1 Definition and Properties . . . . . . . . . . . . . . . . . . . . . . . . . . 40
2.8.2 Interarrival times . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.8.3 Waiting times . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.8.4 The Integrated Poisson Process . . . . . . . . . . . . . . . . . . . . . . . 45
2.9 Submartingales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

iii
iv O. Calin

3 Properties of Stochastic Processes 51

3.1 Stopping Times . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.2 Stopping Theorem for Martingales . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.3 The First Passage of Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
3.4 The Arc-sine Laws . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.5 More on Hitting Times . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.6 The Inverse Laplace Transform Method . . . . . . . . . . . . . . . . . . . . . . 65
3.7 Limits of Stochastic Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.8 Convergence Theorems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.8.1 The Martingale Convergence Theorem . . . . . . . . . . . . . . . . . . . 75
3.8.2 The Squeeze Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.9 Quadratic Variation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.9.1 The Quadratic Variation of Wt . . . . . . . . . . . . . . . . . . . . . . . 77
3.9.2 The Quadratic Variation of Nt − λt . . . . . . . . . . . . . . . . . . . . 79

4 Stochastic Integration 83
4.0.3 Nonanticipating Processes . . . . . . . . . . . . . . . . . . . . . . . . . . 83
4.0.4 Increments of Brownian Motions . . . . . . . . . . . . . . . . . . . . . . 83
4.1 The Ito Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.2 Examples of Ito integrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.2.1 The case Ft = c, constant . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.2.2 The case Ft = Wt . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.3 Properties of the Ito Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.4 The Wiener Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
4.5 Poisson Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
4.5.1 A Work Out Example: the case Ft = Mt . . . . . . . . . . . . . . . . . . 93
RT
4.6 The distribution function of XT = 0 g(t) dNt . . . . . . . . . . . . . . . . . . 97

5 Stochastic Differentiation 99
5.1 Differentiation Rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
5.2 Basic Rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
5.3 Ito’s Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
5.3.1 Ito’s formula for diffusions . . . . . . . . . . . . . . . . . . . . . . . . . . 102
5.3.2 Ito’s formula for Poisson processes . . . . . . . . . . . . . . . . . . . . . 104
5.3.3 Ito’s multidimensional formula . . . . . . . . . . . . . . . . . . . . . . . 105

6 Stochastic Integration Techniques 107

6.1 Fundamental Theorem of Stochastic Calculus . . . . . . . . . . . . . . . . . . . 107
6.2 Stochastic Integration by Parts . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
6.3 The Heat Equation Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
6.4 Table of Usual Stochastic Integrals . . . . . . . . . . . . . . . . . . . . . . . . . 118

7 Stochastic Differential Equations 121

7.1 Definitions and Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
7.2 Finding Mean and Variance from the Equation . . . . . . . . . . . . . . . . . . 122
7.3 The Integration Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
7.4 Exact Stochastic Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
7.5 Integration by Inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
An Introd. to Stoch. Calc. with Appl. to Finance 1

7.6 Linear Stochastic Differential Equations . . . . . . . . . . . . . . . . . . . . . . 136

7.7 The Method of Variation of Parameters . . . . . . . . . . . . . . . . . . . . . . 141
7.8 Integrating Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
7.9 Existence and Uniqueness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145

8 Applications of Brownian Motion 147

8.1 The Generator of an Ito Diffusion . . . . . . . . . . . . . . . . . . . . . . . . . . 147
8.2 Dynkin’s Formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
8.3 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
8.3.1 Kolmogorov’s Backward Equation . . . . . . . . . . . . . . . . . . . . . 151
8.3.2 Exit Time from an Interval . . . . . . . . . . . . . . . . . . . . . . . . . 151
8.3.3 Transience and Recurrence of Brownian Motion . . . . . . . . . . . . . . 152
8.4 Application to Solving Parabolic Equations . . . . . . . . . . . . . . . . . . . . 156

9 Martingales 159
9.1 Examples of Martingales . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
9.2 Girsanov’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
2 O. Calin
 Part I

Stochastic Calculus

3
Chapter 1

Basic Notions

1.1 Probability Space

The modern theory of probability stems from the work of A. N. Kolmogorov published in
1933. Kolmogorov associates a random experiment with a probability space, which is a triplet,
(Ω, F , P ), consisting of the set of outcomes, Ω, a σ-field, F , with Boolean algebra properties,
and a probability measure, P . In the following sections, each of these elements will be discussed
in more detail.

1.2 Sample Space

A random experiment in the theory of probability is an experiment whose outcomes cannot be
determined in advance. These experiments are done mentally most of the time.
When an experiment is performed, the set of all possible outcomes is called the sample
space, and we shall denote it by Ω. In financial markets one can regard this also as the states
of the world, understanding by this all possible states the world might have. The number of
states the world that affect the stock market is huge. These would contain all possible values
for the vector parameters that describe the world, and is practically infinite.
For some simple experiments the sample space is much smaller. For instance, flipping a coin
will produce the sample space with two states {H, T }, while rolling a die yields a sample space
with six states. Choosing randomly a number between 0 and 1 corresponds to a sample space
which is the entire segment (0, 1).
All subsets of the sample space Ω form a set denoted by 2Ω . The reason for this notation
is that the set of parts of Ω can be put into a bijective correspondence with the set of binary
functions f : Ω → {0, 1}. The number of elements of this set is 2|Ω| , where |Ω| denotes the
cardinal of Ω. If the set is finite, |Ω| = n, then 2Ω has 2n elements. If Ω is infinitely countable
(i.e. can be put into a bijective correspondence with the set of natural numbers), then 2|Ω| is
infinite and its cardinal is the same as that of the real number set R. As a matter of fact, if Ω
represents all possible states of the financial world, then 2Ω describes all possible events, which
might happen in the market; this is a supposed to be a fully description of the total information
of the financial world.
The following couple of examples provide instances of sets 2Ω in the finite and infinite cases.
Example 1.2.1 Flip a coin and measure the occurrence of outcomes by 0 and 1: associate a

5
6

0 if the outcome does not occur and a 1 if the outcome occurs. We obtain the following four
possible assignments:

{H → 0, T → 0}, {H → 0, T → 1}, {H → 1, T → 0}, {H → 1, T → 1},

so the set of subsets of {H, T } can be represented as 4 sequences of length 2 formed with 0
and 1: {0, 0}, {0, 1}, {1, 0}, {1, 1}. These correspond in order to the sets Ø, {T }, {H}, {H, T },
which is the set 2{H,T } .

Example 1.2.2 Pick a natural number at random. Any subset of the sample space corresponds
to a sequence formed with 0 and 1. For instance, the subset {1, 3, 5, 6} corresponds to the
sequence 10101100000 . . . having 1 on the 1st, 3rd, 5th and 6th places and 0 in rest. It is known
that the number of these sequences is infinite and can be put into a bijective correspondence with
the real number set R. This can be also written as |2N | = |R|, and stated by saying that the set
of all subsets of natural numbers N has the same cardinal as the real numbers set R.

1.3 Events and Probability

The set of parts 2Ω satisfies the following properties:

1. It contains the empty set Ø;

2. If it contains a set A, then it also contains its complement Ā = Ω\A;
3. It is closed with regard to unions, i.e., if A1 , A2 , . . . is a sequence of sets, then their union
A1 ∪ A2 ∪ · · · also belongs to 2Ω .

Any subset F of 2Ω that satisfies the previous three properties is called a σ-field. The sets
belonging to F are called events. This way, the complement of an event, or the union of events
is also an event. We say that an event occurs if the outcome of the experiment is an element
of that subset.
The chance of occurrence of an event is measured by a probability function P : F → [0, 1]
which satisfies the following two properties:
1. P (Ω) = 1;
2. For any mutually disjoint events A1 , A2 , · · · ∈ F ,

P (A1 ∪ A2 ∪ · · · ) = P (A1 ) + P (A2 ) + · · · .

The triplet (Ω, F , P ) is called a probability space. This is the main setup in which the
probability theory works.

Example 1.3.1 In the case of flipping a coin, the probability space has the following elements:
Ω = {H, T }, F = {Ø, {H}, {T }, {H, T }} and P defined by P (Ø) = 0, P ({H}) = 12 , P ({T }) =
1
2 , P ({H, T }) = 1.

Example 1.3.2 Consider a finite sample space Ω = {s1 , . . . , sn }, with the σ-field F = 2Ω , and
probability given by P (A) = |A|/n, ∀A ∈ F . Then (Ω, 2Ω , P ) is called the classical probability
space.
 7

Figure 1.1: If any pullback X −1 (a, b) is known, then the random variable X : Ω → R is
2Ω -measurable.

1.4 Random Variables

Since the σ-field F provides the knowledge about which events are possible on the considered
probability space, then F can be regarded as the information component of the probability
space (Ω, F , P ). A random variable X is a function that assigns a numerical value to each
state of the world, X : Ω → R, such that the values taken by X are known to someone who
has access to the information F . More precisely, given any two numbers a, b ∈ R, then all the
states of the world for which X takes values between a and b forms a set that is an event (an
element of F ), i.e.
{ω ∈ Ω; a < X(ω) < b} ∈ F .
Another way of saying this is that X is an F -measurable function. It is worth noting that
in the case of the classical field of probability the knowledge is maximal since F = 2Ω , and
hence the measurability of random variables is automatically satisfied. From now on instead
of measurable terminology we shall use the more suggestive word predictable. This will make
more sense in a future section when we shall introduce conditional expectations.

Example 1.4.1 Let X(ω) be the number of people who want to buy houses, given the state of
the market ω. Is X predictable? This would mean that given two numbers, say a = 10, 000 and
b = 50, 000, we know all the market situations ω for which there are at least 10, 000 and at most
50, 000 people willing to purchase houses. Many times, in theory, it makes sense to assume that
we have enough knowledge to assume X predictable.

Example 1.4.2 Consider the experiment of flipping three coins. In this case Ω is the set of
all possible triplets. Consider the random variable X which gives the number of tails obtained.
For instance X(HHH) = 0, X(HHT ) = 1, etc. The sets

{ω; X(ω) = 0} = {HHH}, {ω; X(ω) = 1} = {HHT, HT H, T HH},

{ω; X(ω) = 3} = {T T T }, {ω; X(ω) = 2} = {HT T, T HT, T T H}

belong to 2Ω , and hence X is a random variable.

8

Example 1.4.3 A graph is a set of elements, called nodes, and a set of unordered pairs of
nodes, called edges. Consider the set of nodes N = {n1 , n2 , . . . , nk } and the set of edges
E = {(ni , nj ), 1 ≤ i, j ≤ n, i 6= j}. Define the probability space (Ω, F , P ), where
◦ the sample space is the the complete graph, Ω = N ∪ E;
◦ the σ-field F is the set of all subgraphs of Ω;
◦ the probability is given by P (G) = n(G)/k, where n(G) is the number of nodes of the
graph G.
As an example of a random variable we consider Y : F → R, Y (G) = the total number of edges
of the graph G. Since given F , one can count the total number of edges of each subgraph, it
follows that Y is F -measurable, and hence it is a random variable.

1.5 Distribution Functions

Let X be a random variable on the probability space (Ω, F , P ). The distribution function of X
is the function FX : R → [0, 1] defined by

FX (x) = P (ω; X(ω) ≤ x).

It is worth observing that since X is a random variable, then the set {ω; X(ω) ≤ x} belongs to
the information set F .
The distribution function is non-decreasing and satisfies the limits

lim FX (x) = 0, lim FX (x) = 1.

x→−∞ x→+∞

If we have
d
F (x) = p(x),
dx X
then we say that p(x) is the probability density function of X. A useful property which follows
from the Fundamental Theorem of Calculus is
Z b
P (a < X < b) = P (ω; a < X(ω) < b) = p(x) dx.
a

In the case of discrete random variables the aforementioned integral is replaced by the following
sum X
P (a < X < b) = P (X = x).
a<x<b

For more details the reader is referred to a traditional probability book, such as Wackerly et.
al. [?].

1.6 Basic Distributions

We shall recall a few basic distributions, which are most often seen in applications.
Normal distribution A random variable X is said to have a normal distribution if its prob-
ability density function is given by
1 2 2
p(x) = √ e−(x−µ) /(2σ ) ,
σ 2π
 9

0.4
0.5

0.3 0.4

0.3
0.2

0.2

0.1
0.1

-4 -2 0 2 4 0 2 4 6 8

a b
3.5

Α = 3, Β = 9 Α = 8, Β = 3
3.0
0.20

Α = 3, Β = 2 2.5

0.15

2.0

0.10 1.5
Α = 4, Β = 3
1.0

0.05

0.5

0 5 10 15 20 0.0 0.2 0.4 0.6 0.8 1.0

c d

Figure 1.2: a Normal distribution; b Log-normal distribution; c Gamma distributions; d Beta

distributions.

with µ and σ > 0 constant parameters, see Fig.1.2a. The mean and variance are given by

E[X] = µ, V ar[X] = σ 2 .

If X has a normal distribution with mean µ and variance σ 2 , we shall write

X ∼ N (µ, σ 2 ).

Exercise 1.6.1 Let α, β ∈ R. Show that if X is normal distributed, with X ∼ N (µ, σ 2 ), then
Y = αX + β is also normal distributed, with Y ∼ N (αµ + β, α2 σ 2 ).

Log-normal distribution Let X be normally distributed with mean µ and variance σ 2 . Then
the random variable Y = eX is said to be log-normal distributed. The mean and variance of Y
are given by
σ2
E[Y ] = eµ+ 2

2 2
V ar[Y ] = e2µ+σ (eσ − 1).

The density function of the log-normal distributed random variable Y is given by

1 (ln x−µ)2
p(x) = √ e− 2σ2 , x > 0,
xσ 2π
see Fig.1.2b.

Exercise 1.6.2 Given that the moment generating function of a normally distributed random
2 2
variable X ∼ N (µ, σ 2 ) is m(t) = E[etX ] = eµt+t σ /2 , show that
10

2 2
(a) E[Y n ] = enµ+n σ /2 , where Y = eX .
(b) Show that the mean and variance of the log-normal random variable Y = eX are
2 2 2
E[Y ] = eµ+σ /2
, V ar[Y ] = e2µ+σ (eσ − 1).
Gamma distribution A random variable X is said to have a gamma distribution with pa-
rameters α > 0, β > 0 if its density function is given by
xα−1 e−x/β
p(x) = , x ≥ 0,
β α Γ(α)
where Γ(α) denotes the gamma function,1 see Fig.1.2c. The mean and variance are
E[X] = αβ, V ar[X] = αβ 2 .
The case α = 1 is known as the exponential distribution, see Fig.1.3a. In this case
1 −x/β
p(x) = e , x ≥ 0.
β
The particular case when α = n/2 and β = 2 becomes the χ2 −distribution with n degrees of
freedom. This characterizes also a sum of n independent standard normal distributions.
Beta distribution A random variable X is said to have a beta distribution with parameters
α > 0, β > 0 if its probability density function is of the form
xα−1 (1 − x)β−1
p(x) = , 0 ≤ x ≤ 1,
B(α, β)
where B(α, β) denotes the beta function.2 See see Fig.1.2d for two particular density functions.
In this case
α αβ
E[X] = , V ar[X] = 2
.
α+β (α + β) (α + β + 1)
Poisson distribution A discrete random variable X is said to have a Poisson probability
distribution if
λk −λ
P (X = k) = e , k = 0, 1, 2, . . . ,
k!
with λ > 0 parameter, see Fig.1.3b. In this case E[X] = λ and V ar[X] = λ.
Pearson 5 distribution Let α, β > 0. A random variable X with the density function
1 e−β/x
p(x) = , x≥0
βΓ(α) (x/β)α+1
is said to have a Pearson 5 distribution3 with positive parameters α and β. It can be shown
that
 
 β  β2
, if α > 1 , if α > 2
E[X] = α−1 V ar(X) = (α − 1)2 (α − 2)
 ∞, otherwise, 
∞, otherwise.
R
the definition of the gamma function Γ(α) = 0∞ y α−1 e−y dy; if α = n, integer, then Γ(n) = (n − 1)!
1 Recall
R
2 Two definition formulas for the beta functions are B(α, β) = Γ(α)Γ(β) = 1 y α−1 (1 − y)β−1 dy.
Γ(α+β) 0
3 The Pearson family of distributions was designed by Pearson between 1890 and 1895. There are several

Pearson distributions, this one being distinguished by the number 5.

 11

0.35 0.10

0.30 Λ = 15, 0 < k < 30

0.08

0.25

0.20
Β=3 0.06

0.15
0.04

0.10

0.02
0.05

0 2 4 6 8 10 5 10 15 20 25 30

a b

Figure 1.3: a Exponential distribution; b Poisson distribution.

β
The mode of this distribution is equal to .
α+1
The Inverse Gaussian distribution Let µ, λ > 0. A random variable X has an inverse
Gaussian distribution with parameters µ and λ if its density function is given by
2
λ − λ(x−µ)
p(x) = e 2µ2 x , x > 0. (1.6.1)
2πx3
We shall write X ∼ IG(µ, λ). Its mean, variance and mode are given by
r
µ3  9µ2 3µ 
E[X] = µ, V ar(X) = , M ode(X) = µ 1+ 2 − .
λ 4λ 2λ
This distribution will be used to model the time instance when a Brownian motion with drift
exceeds a certain barrier for the first time.

1.7 Independent Random Variables

Roughly speaking, two random variables X and Y are independent if the occurrence of one
of them does not change the probability density of the other. More precisely, if for any sets
A, B ⊂ R, the events
{ω; X(ω) ∈ A}, {ω; Y (ω) ∈ B}
are independent,4 then X and Y are called independent random variables.

Proposition 1.7.1 Let X and Y be independent random variables with probability density
functions pX (x) and pY (y). Then the joint probability density function of (X, Y ) is given by
pX,Y (x, y) = pX (x) pY (y).

Proof: Using the independence of sets, we have5

pX,Y (x, y) dxdy = P (x < X < x + dx, y < Y < y + dy)

= P (x < X < x + dx)P (y < Y < y + dy)
= pX (x) dx pY (y) dy
= pX (x)pY (y) dxdy.
4 In Probability Theory two events A1 and A2 are called independent if P (A1 ∩ A2 ) = P (A1 )P (A2 ).
R
5 We are using the useful approximation P (x < X < x + dx) = xx+dx p(u) du = p(x)dx.
12

Dropping the factor dxdy yields the desired result. We note that the converse holds true.

1.8 Integration in Probability Measure

The notion of expectation is based on integration on measure spaces. In this section we recall
briefly the definition of an integral with respect to the probability measure P .
Let X : Ω → R be a random variable on the probability space (Ω, F , P ). A partition
(Ωi )1≤i≤n of Ω is a family of subsets Ωi ⊂ Ω satisfying
1. Ωi ∩ Ωj = Ø, for i 6= j;
[n
2. Ωi = Ω.
i

Each Ωi is an event with the Pnassociated probability P (Ωi ). A simple function is a sum of
characteristic functions f = i ci χΩi . This means f (ω) = ck for ω ∈ Ωk . The integral of the
simple function f is defined by
Z n
X
f dP = ci P (Ωi ).
Ω i

If X : Ω → R is a random variable such that there is a sequence of simple functions (fn )n≥1
satisfying:
1. fn is fundamental in probability: ∀ǫ > 0 limn,m→∞ P (ω; |fn (ω) − fm (ω)| ≥ ǫ) → 0,
2. fn converges to X in probability: ∀ǫ > 0 limn→∞ P (ω; |fn (ω) − X(ω)| ≥ ǫ) → 0
then the integral of X is defined as the following limit of integrals
Z Z
X dP = lim fn dP.
Ω n→∞ Ω
R R
From now on, the integral notations Ω X dP or Ω X(ω) dP (ω) will be used interchangeably.
In the rest of the chapter the integral notation will be used formally, without requiring a direct
use of the previous definition.

1.9 Expectation
A random variable X : Ω → R is called integrable if
Z Z
|X(ω)| dP (ω) = |x|p(x) dx < ∞,
Ω R

where p(x) denotes the probability density function of X. The previous identity is based on
changing the domain of integration from Ω to R.
The expectation of an integrable random variable X is defined by
Z Z
E[X] = X(ω) dP (ω) = x p(x) dx.
Ω R
 13

Customarily, the expectation of X is denoted by µ and it is also called the mean. In general,
for any continuous6 function h : R → R, we have
Z Z


E[h(X)] = h X(ω) dP (ω) = h(x)p(x) dx.
Ω R

Proposition 1.9.1 The expectation operator E is linear, i.e. for any integrable random vari-
ables X and Y
1. E[cX] = cE[X], ∀c ∈ R;
2. E[X + Y ] = E[X] + E[Y ].

Proof: It follows from the fact that the integral is a linear operator.

Proposition 1.9.2 Let X and Y be two independent integrable random variables. Then

E[XY ] = E[X]E[Y ].

Proof: This is a variant of Fubini’s theorem, which in this case states that a double integral
is a product of two simple integrals. Let pX , pY , pX,Y denote the probability densities of X, Y
and (X, Y ), respectively. Since X and Y are independent, by Proposition 1.7.1 we have
ZZ Z Z
E[XY ] = xypX,Y (x, y) dxdy = xpX (x) dx ypY (y) dy = E[X]E[Y ].

1.10 Radon-Nikodym’s Theorem

This section is concerned with existence and uniqueness results that will be useful later in
defining conditional expectations. Since this section is rather theoretical, it can be skipped at
a first reading.

Proposition 1.10.1 Consider the probability space (Ω, F , P ), and let G be a σ-field included
in F . If X is a G-predictable random variable such that
Z
X dP = 0 ∀A ∈ G,
A

then X = 0 a.s.


Proof: In order to show that X = 0 almost surely, it suffices to prove that P ω; X(ω) = 0 = 1.
We shall show first that X takes values as small as possible with probability one, i.e. ∀ǫ > 0
we have P (|X| < ǫ) = 1. To do this, let A = {ω; X(ω) ≥ ǫ}. Then
Z Z Z
1 1
0 ≤ P (X ≥ ǫ) = dP = ǫ dP ≤ X dP = 0,
A ǫ A ǫ A
and hence P (X ≥ ǫ) = 0. Similarly P (X ≤ −ǫ) = 0. Therefore

P (|X| < ǫ) = 1 − P (X ≥ ǫ) − P (X ≤ −ǫ) = 1 − 0 − 0 = 1.

6 in general, measurable
14

1
Taking ǫ → 0 leads to P (|X| = 0) = 1. This can be formalized as follows. Let ǫ = n and
consider Bn = {ω; |X(ω)| ≤ ǫ}, with P (Bn ) = 1. Then
∞
\
P (X = 0) = P (|X| = 0) = P ( Bn ) = lim P (Bn ) = 1.
n→∞
n=1

Corollary 1.10.2 If X and Y are G-predictable random variables such that

Z Z
X dP = Y dP ∀A ∈ G,
A A

then X = Y a.s.
R
Proof: Since A (X − Y ) dP = 0, ∀A ∈ G, by Proposition 1.10.1 we have X − Y = 0 a.s.

Theorem 1.10.3 (Radon-Nikodym) Let (Ω, F , P ) be a probability space and G be a σ-field

included in F . Then for any random variable X there is a G-predictable random variable Y
such that Z Z
X dP = Y dP, ∀A ∈ G. (1.10.2)
A A

We shall omit the proof but discuss a few aspects.

1. All σ-fields G ⊂ F contain impossible and certain events Ø, Ω ∈ G. Making A = Ω yields
Z Z
X dP = Y dP,
Ω Ω

which is E[X] = E[Y ].

2. Radon-Nikodym’s theorem states the existence of Y . In fact this is unique almost surely.
In order to show that, assume there are two G-predictable random variables Y1 and Y2 with the
aforementioned property. Then from (1.10.2) yields
Z Z
Y1 dP = Y2 dP, ∀A ∈ G.
A A

Applying Corollary (1.10.2) yields Y1 = Y2 a.s.

R
3. Since E[X] = Ω X dP is the expectation of the random variable X, given the full
knowledge F , then the random variable Y plays the role of the expectation of X given the
partial information G. The next section will deal with this concept in detail.

1.11 Conditional Expectation

Let X be a random variable on the probability space (Ω, F , P ). Let G be a σ-field contained in
F . Since X is F -predictable, the expectation of X, given the information F must be X itself.
This shall be written as E[X|F ] = X (for details see Example 1.11.3). It is natural to ask what
is the expectation of X, given the information G. This is a random variable denoted by E[X|G]
satisfying the following properties:
 15

1. E[X|G] is G-predictable;
R R
2. A E[X|G] dP = A X dP, ∀A ∈ G.
E[X|G] is called the conditional expectation of X given G.
We owe a few explanations regarding the correctness of the aforementioned definition. The
existence of the G-predictable random variable E[X|G] is assured by the Radon-Nikodym theo-
rem. The almost surely uniqueness is an application of Proposition (1.10.1) (see the discussion
point 2 of section 1.10).
It is worth noting that the expectation of X, denoted by E[X] is a number, while the
conditional expectation E[X|G] is a random variable. When are they equal and what is their
relationship? The answer is inferred by the following solved exercises.

Example 1.11.1 Show that if G = {Ø, Ω}, then E[X|G] = E[X].

Proof: We need to show that E[X] satisfies conditions 1 and 2. The first one is obviously
satisfied since any constant is G-predictable. The latter condition is checked on each set of G.
We have
Z Z Z
X dP = E[X] = E[X] dP = E[X]dP
ZΩ Z Ω Ω

X dP = E[X]dP.
Ø Ø

Example 1.11.2 Show that E[E[X|G]] = E[X], i.e. all conditional expectations have the same
mean, which is the mean of X.

Proof: Using the definition of expectation and taking A = Ω in the second relation of the
aforementioned definition, yields
Z Z
E[E[X|G]] = E[X|G] dP = XdP = E[X],
Ω Ω

which ends the proof.

Example 1.11.3 The conditional expectation of X given the total information F is the random
variable X itself, i.e.
E[X|F ] = X.

Proof: The random variables X and E[X|F ] are both F -predictable (from the definition of
the random variable). From the definition of the conditional expectation we have
Z Z
E[X|F ] dP = X dP, ∀A ∈ F .
A A

Corollary (1.10.2) implies that E[X|F ] = X almost surely.

General properties of the conditional expectation are stated below without proof. The proof
involves more or less simple manipulations of integrals and can be taken as an exercise for the
reader.
16

Proposition 1.11.4 Let X and Y be two random variables on the probability space (Ω, F , P ).
We have
1. Linearity:
E[aX + bY |G] = aE[X|G] + bE[Y |G], ∀a, b ∈ R;
2. Factoring out the predictable part:

E[XY |G] = XE[Y |G]

if X is G-predictable. In particular, E[X|G] = X.

3. Tower property:
E[E[X|G]|H] = E[X|H], if H ⊂ G;
4. Positivity:
E[X|G] ≥ 0, if X ≥ 0;
5. Expectation of a constant is a constant:

E[c|G] = c.

6. An independent condition drops out:

E[X|G] = E[X],

if X is independent of G.

Exercise 1.11.5 Prove the property 3 (tower property) given in the previous proposition.

Exercise 1.11.6 Let X be a random variable on the probability space (Ω, F , P ), which is in-
dependentof the σ-field G ⊂ F . Consider the characteristic function of a set A ⊂ Ω defined by
1, if ω ∈ A
χA (ω) = Show the following:
0, if ω ∈ /A
(a) χA is G-predictable for any A ∈ G;
(b) P (A) = E[χA ];
(c) X and χA are independent random variables;
(d) E[χA X] = E[X]P (A) for any A ∈ G;
(e) E[X|G] = E[X].

1.12 Inequalities of Random Variables

This section prepares the reader for the limits of sequences of random variables and limits of
stochastic processes. We shall start with a classical inequality result regarding expectations:

Theorem 1.12.1 (Jensen’s inequality) Let ϕ : R → R be a convex function and let X be

an integrable random variable on the probability space (Ω, F , P ). If ϕ(X) is integrable, then

ϕ(E[X]) ≤ E[ϕ(X)]

almost surely (i.e. the inequality might fail on a set of probability zero).
 17

Figure 1.4: Jensen’s inequality ϕ(E[X]) < E[ϕ(X)] for a convex function ϕ.

Proof: We shall assume ϕ twice differentiable with ϕ′′ continuous. Let µ = E[X]. Expand ϕ
in a Taylor series about µ and get
1
ϕ(x) = ϕ(µ) + ϕ′ (µ)(x − µ) + ϕ′′ (ξ)(ξ − µ)2 ,
2
with ξ in between x and µ. Since ϕ is convex, ϕ′′ ≥ 0, and hence

ϕ(x) ≥ ϕ(µ) + ϕ′ (µ)(x − µ),

which means the graph of ϕ(x) is above the tangent line at x, ϕ(x) . Replacing x by the
random variable X, and taking the expectation yields

E[ϕ(X)] ≥ E[ϕ(µ) + ϕ′ (µ)(X − µ)] = ϕ(µ) + ϕ′ (µ)(E[X] − µ)

= ϕ(µ) = ϕ(E[X]),

which proves the result.

Fig.1.4 provides a graphical interpretation of Jensen’s inequality. If the distribution of X is
symmetric, then the distribution of ϕ(X) is skewed, with ϕ(E[X]) < E[ϕ(X)].
It is worth noting that the inequality is reversed for ϕ concave. We shall next present a
couple of applications.
A random variable X : Ω → R is called square integrable if
Z Z
2 2
E[X ] = |X(ω)| dP (ω) = x2 p(x) dx < ∞.
Ω R

Application 1.12.2 If X is a square integrable random variable, then it is integrable.

Proof: Jensen’s inequality with ϕ(x) = x2 becomes

E[X]2 ≤ E[X 2 ].

Since the right side is finite, it follows that E[X] < ∞, so X is integrable.
18

Application 1.12.3 If mX (t) denotes the moment generating function of the random variable
X with mean µ, then
mX (t) ≥ etµ .

Proof: Applying Jensen inequality with the convex function ϕ(x) = ex yields

eE[X] ≤ E[eX ].

Substituting tX for X yields

eE[tX] ≤ E[etX ]. (1.12.3)
Using the definition of the moment generating function mX (t) = E[etX ] and that E[tX] =
tE[X] = tµ, then (1.12.3) leads to the desired inequality.
The variance of a square integrable random variable X is defined by

V ar(X) = E[X 2 ] − E[X]2 .

By Application 1.12.2 we have V ar(X) ≥ 0, so there is a constant σX > 0, called standard

deviation, such that
2
σX = V ar(X).

Exercise 1.12.4 Prove the following identity:

V ar[X] = E[(X − E[X])2 ].

Exercise 1.12.5 Prove that a non-constant random variable has a non-zero standard devia-
tion.

Exercise 1.12.6 Prove the following extension of Jensen’s inequality: If ϕ is a convex function,
then for any σ-field G ⊂ F we have

ϕ(E[X|G]) ≤ E[ϕ(X)|G].

Exercise 1.12.7 Show the following:

(a) |E[X]| ≤ E[|X|];
(b) |E[X|G]| ≤ E[|X| |G], for any σ-field G ⊂ F ;
(c) |E[X]|r ≤ E[|X|r ], for r ≥ 1;
(d) |E[X|G]|r ≤ E[|X|r |G], for any σ-field G ⊂ F and r ≥ 1.

Theorem 1.12.8 (Markov’s inequality) For any λ, p > 0, we have the following inequality:
1
P (ω; |X(ω)| ≥ λ) ≤ E[|X|p ].
λp
Proof: Let A = {ω; |X(ω)| ≥ λ}. Then
Z Z Z
E[|X|p ] = |X(ω)|p dP (ω) ≥ |X(ω)|p dP (ω) ≥ λp dP (ω)
Ω A A
Z
p
= λ dP (ω) = λp P (A) = λp P (|X| ≥ λ).
A
p
Dividing by λ leads to the desired result.
 19

Theorem 1.12.9 (Tchebychev’s inequality) If X is a random variable with mean µ and

variance σ 2 , then
σ2
P (ω; |X(ω) − µ| ≥ λ) ≤ 2 .
λ
Proof: Let A = {ω; |X(ω) − µ| ≥ λ}. Then
Z Z
σ2 = V ar(X) = E[(X − µ)2 ] = (X − µ)2 dP ≥ (X − µ)2 dP
Ω A
Z
≥ λ2 dP = λ2 P (A) = λ2 P (ω; |X(ω) − µ| ≥ λ).
A

Dividing by λ2 leads to the desired inequality.

The next result deals with exponentially decreasing bounds on tail distributions.

Theorem 1.12.10 (Chernoff bounds) Let X be a random variable. Then for any λ > 0 we
have
E[etX ]
1. P (X ≥ λ) ≤ , ∀t > 0;
eλt
E[etX ]
2. P (X ≤ λ) ≤ , ∀t < 0.
eλt
Proof: 1. Let t > 0 and denote Y = etX . By Markov’s inequality

E[Y ]
P (Y ≥ eλt ) ≤ .
eλt
Then we have

P (X ≥ λ) = P (tX ≥ λt) = P (etX ≥ eλt )

E[Y ] E[etX ]
= P (Y ≥ eλt ) ≤ λt = .
e eλt
2. The case t < 0 is similar.
In the following we shall present an application of the Chernoff bounds for the normal
distributed random variables.
Let X be a random variable normally distributed with mean µ and variance σ 2 . It is known
that its moment generating function is given by
1 2
σ2
m(t) = E[etX ] = eµt+ 2 t .

Using the first Chernoff bound we obtain

m(t) 1 2 2
P (X ≥ λ) ≤ = e(µ−λ)t+ 2 t σ , ∀t > 0,
eλt
which implies
1
min[(µ − λ)t + t2 σ 2 ]
P (X ≥ λ) ≤ e t>0 2 .
20

It is easy to see that the quadratic function f (t) = (µ − λ)t + 21 t2 σ 2 has the minimum value
λ−µ
reached for t = . Since t > 0, λ needs to satisfy λ > µ. Then
σ2
λ − µ (λ − µ)2
min f (t) = f 2
=− .
t>0 σ 2σ 2
Substituting into the previous formula, we obtain the following result:

Proposition 1.12.11 If X is a normally distributed variable, with X ∼ N (µ, σ 2 ), then for

any λ > µ
(λ − µ)2
−
P (X ≥ λ) ≤ e 2σ 2 .

Exercise 1.12.12 Let X be a Poisson random variable with mean λ > 0.

t
(a) Show that the moment generating function of X is m(t) = eλ(e −1) ;
(b) Use a Chernoff bound to show that
t
P (X ≥ k) ≤ eλ(e −1)−tk
, t > 0.

Markov’s, Tchebychev’s and Chernoff’s inequalities will be useful later when computing
limits of random variables.
The next inequality is called Tchebychev’s inequality for monotone sequences of numbers.

Lemma 1.12.13 Let (ai ) and (bi ) be two sequences of real numbers such that either

a1 ≤ a2 ≤ · · · ≤ an , b1 ≤ b2 ≤ · · · ≤ bn

or
a1 ≥ a2 ≥ · · · ≥ an ,b1 ≥ b2 ≥ · · · ≥ bn
Xn
If (λi ) is a sequence of non-negative numbers such that λi = 1, then
i=1

n
X n
 X  Xn
λi ai λi bi ≤ λi ai bi .
i=1 i=1 i=1

Proof: Since the sequences (ai ) and (bi ) are either both increasing or both decreasing

(ai − aj )(bi − bj ) ≥ 0.

Multiplying by the positive quantity λi λj and summing over i and j we get

X
λi λj (ai − aj )(bi − bj ) ≥ 0.
i,j

Expanding yields
X  X  X  X  X  X 
λj λi ai bi − λi ai λj bj − λj aj λi bi
j i i j j i
X  X 
+ λi λj aj bj ≥ 0.
i j
 21
X
Using λj = 1 the expression becomes
j

X X  X 
λi ai bi ≥ λi ai λj bj ,
i i j

which ends the proof.

Next we present a meaningful application of the previous inequality.

Proposition 1.12.14 Let X be a random variable and f and g be two functions, both increas-
ing or both decreasing. Then

E[f (X)g(X)] ≥ E[f (X)]E[g(X)]. (1.12.4)

Proof: If X is a discrete random variable, with outcomes {x1 , · · · , xn }, inequality (1.12.4)

becomes X X X
f (xj )g(xj )p(xj ) ≥ f (xj )p(xj ) g(xj )p(xj ),
j j j

where p(xj ) = P (X = xj ). Denoting aj = f (xj ), bj = g(xj ), and λj = p(xj ), the inequality

transforms into X X X
aj bj λj ≥ aj λj bj λj ,
j j j

which holds true by Lemma 1.12.13.

If X is a continuous random variable with the density function p : I → R, the inequality
(1.12.4) can be written in the integral form
Z Z Z
f (x)g(x)p(x) dx ≥ f (x)p(x) dx g(x)p(x) dx. (1.12.5)
I I I

Let x0 < x1 < · · · < xn be a partition of the interval I, with ∆x = xk+1 − xk . Using Lemma
1.12.13 we obtain the following inequality between Riemann sums
X X  X 
f (xj )g(xj )p(xj )∆x ≥ f (xj )p(xj )∆x g(xj )p(xj )∆x ,
j j j

where aj = f (xj ), bj = g(xj ), and λj = p(xj )∆x. Taking the limit k∆xk → 0 we obtain
(1.12.5), which leads to the desired result.

Exercise 1.12.15 Show the following inequalities:

(a) E[X 2 ] ≥ E[X]2 ;
(b) E[X sinh(X)] ≥ E[X]E[sinh(X)];
(c) E[X 6 ] ≥ E[X]E[X 5 ];
(d) E[X 6 ] ≥ E[X 3 ]2 .

Exercise 1.12.16 For any n, k ≥ 1, show that

E[X 2(n+k+1) ] ≥ E[X 2k+1 ]E[X 2n+1 ].

22

1.13 Limits of Sequences of Random Variables

Consider a sequence (Xn )n≥1 of random variables defined on the probability space (Ω, F , P ).
There are several ways of making sense of the limit expression X = lim Xn , and they will be
n→∞
discussed in the following sections.

Almost Certain Limit

The sequence Xn converges almost certainly to X, if for all states of the world ω, except a set
of probability zero, we have
lim Xn (ω) = X(ω).
n→∞

More precisely, this means

 
P ω; lim Xn (ω) = X(ω) = 1,
n→∞

and we shall write ac-lim Xn = X. An important example where this type of limit occurs is
n→∞
the Strong Law of Large Numbers:
If Xn is a sequence of independent and identically distributed random variables with the
X1 + · · · + Xn
same mean µ, then ac-lim = µ.
n→∞ n
It is worth noting that this type of convergence is also known under the name of strong
convergence. This is the reason that the aforementioned theorem bares its name.

Example 1.13.1 Let Ω = {H, T } be the sample space obtained when a coin is flipped. Consider
the random variables Xn : Ω → {0, 1}, where Xn denotes the number of heads obtained at the n-
th flip. Obviously, Xn are i.i.d., with the distribution given by P (Xn = 0) = P (Xn = 1) = 1/2,
and the mean E[Xn ] = 0 · 12 + 1 · 21 = 12 . Then X1 + · · · + Xn is the number of heads obtained
after n flips of the coin. By the law of large numbers, n1 (X1 + · · · + Xn ) tends to 1/2 strongly,
as n → ∞.

Mean Square Limit

Another possibility of convergence is to look at the mean square deviation of Xn from X. We
say that Xn converges to X in the mean square if

lim E[(Xn − X)2 ] = 0.

n→∞

More precisely, this should be interpreted as

Z

2
lim Xn (ω) − X(ω) dP (ω) = 0.
n→∞ Ω

This limit will be abbreviated by ms-lim Xn = X. The mean square convergence is useful
n→∞
when defining the Ito integral.

Example 1.13.1 Consider a sequence Xn of random variables such that there is a constant k
with E[Xn ] → k and V ar(Xn ) → 0 as n → ∞. Show that ms-lim Xn = k.
n→∞
 23

Proof: Since we have

E[|Xn − k|2 ] = E[Xn2 − 2kXn + k 2 ] = E[Xn2 ] − 2kE[Xn ] + k 2

= E[Xn2 ] − E[Xn ]2 + E[Xn ]2 − 2kE[Xn ] + k 2

2
= V ar(Xn ) + E[Xn ] − k ,

the right side tends to 0 when taking the limit n → ∞.

Exercise 1.13.2 Show the following relation

2
E[(X − Y )2 ] = V ar[X] + V ar[Y ] + E[X] − E[Y ] − 2Cov(X, Y ).

Exercise 1.13.3 If Xn tends to X in mean square, with E[X 2 ] < ∞, show that:
(a) E[Xn ] → E[X] as n → ∞;
(b) E[Xn2 ] → E[X 2 ] as n → ∞;
(c) V ar[Xn ] → V ar[X] as n → ∞;
(d) Cov(Xn , X) → V ar[X] as n → ∞.

Exercise 1.13.4 If Xn tends to X in mean square, show that E[Xn |H] tends to E[X|H] in
mean square.

Limit in Probability or Stochastic Limit

The random variable X is the stochastic limit of Xn if for n large enough the probability of
deviation from X can be made smaller than any arbitrary ǫ. More precisely, for any ǫ > 0


lim P ω; |Xn (ω) − X(ω)| ≤ ǫ = 1.
n→∞

This can be written also as

lim P ω; |Xn (ω) − X(ω)| > ǫ = 0.
n→∞

This limit is denoted by st-lim Xn = X.

n→∞
It is worth noting that both almost certain convergence and convergence in mean square
imply the stochastic convergence. Hence, the stochastic convergence is weaker than the afore-
mentioned two convergence cases. This is the reason that it is also called the weak convergence.
One application is the Weak Law of Large Numbers:
If X1 , X2 , . . . are identically distributed with expected value µ and if any finite number of
X1 + · · · + Xn
them are independent, then st-lim = µ.
n→∞ n
Proposition 1.13.5 The convergence in the mean square implies the stochastic convergence.

Proof: Let ms-lim Yn = Y . Let ǫ > 0 be arbitrarily fixed. Applying Markov’s inequality with
n→∞
X = Yn − Y , p = 2 and λ = ǫ, yields
1
0 ≤ P (|Yn − Y | ≥ ǫ) ≤ E[|Yn − Y |2 ].
ǫ2
24

The right side tends to 0 as n → ∞. Applying the Squeeze Theorem we obtain

lim P (|Yn − Y | ≥ ǫ) = 0,
n→∞

which means that Yn converges stochastically to Y .

Example 1.13.6 Let Xn be a sequence of random variables such that E[|Xn |] → 0 as n → ∞.

Prove that st-lim Xn = 0.
n→∞

Proof: Let ǫ > 0 be arbitrarily fixed. We need to show

lim P ω; |Xn (ω)| ≥ ǫ = 0. (1.13.6)
n→∞

From Markov’s inequality (see Exercise 1.12.8) we have

E[|Xn |]
0 ≤ P ω; |Xn (ω)| ≥ ǫ ≤ .
ǫ

Using Squeeze Theorem we obtain (1.13.6).

Remark 1.13.7 The conclusion still holds true even in the case when there is a p > 0 such
that E[|Xn |p ] → 0 as n → ∞.

Limit in Distribution

We say the sequence Xn converges in distribution to X if for any continuous bounded function
ϕ(x) we have
lim ϕ(Xn ) = ϕ(X).
n→∞

This type of limit is even weaker than the stochastic convergence, i.e. it is implied by it.
An application of the limit in distribution is obtained if we consider ϕ(x) = eitx . In this
case, if Xn converges in distribution to X, then the characteristic function of Xn converges to
the characteristic function of X. In particular, the probability density of Xn approaches the
probability density of X.
It can be shown that the convergence in distribution is equivalent with

lim Fn (x) = F (x),

n→∞

whenever F is continuous at x, where Fn and F denote the distribution functions of Xn and

X, respectively. This is the reason that this convergence bares its name.

Remark 1.13.8 The almost certain convergence implies the stochastic convergence, and the
stochastic convergence implies the limit in distribution. The proof of these statements is beyound
the goal of this book. The interested reader can consult a graduate text in probability theory.
 25

1.14 Properties of Limits

Lemma 1.14.1 If ms-lim Xn = 0 and ms-lim Yn = 0, then
n→∞ n→∞

1. ms-lim (Xn + Yn ) = 0
n→∞

2. ms-lim (Xn Yn ) = 0.
n→∞

Proof: Since ms-lim Xn = 0, then lim E[Xn2 ] = 0. Applying the Squeeze Theorem to the
n→∞ n→∞
inequality7
0 ≤ E[Xn ]2 ≤ E[Xn2 ]
yields lim E[Xn ] = 0. Then
n→∞
 
lim V ar[Xn ] = lim E[Xn2 ] − lim E[Xn ]2
n→∞ n→∞ n→∞
= lim E[Xn2 ] − lim E[Xn ]2
n→∞ n→∞
= 0.

Similarly, we have lim E[Yn2 ] = 0, lim E[Yn ] = 0 and lim V ar[Yn ] = 0. Then lim σXn =
n→∞ n→∞ n→∞ n→∞
lim σYn = 0. Using the correlation definition formula of two random variables Xn and Yn
n→∞

Cov(Xn , Yn )
Corr(Xn , Yn ) = ,
σXn σYn

and the fact that |Corr(Xn , Yn )| ≤ 1, yields

0 ≤ |Cov(Xn , Yn )| ≤ σXn σYn .

Since lim σXn σXn = 0, from the Squeeze Theorem it follows that
n→∞

lim Cov(Xn , Yn ) = 0.
n→∞

Taking n → ∞ in the relation

Cov(Xn , Yn ) = E[Xn Yn ] − E[Xn ]E[Yn ]

yields lim E[Xn Yn ] = 0. Using the previous relations, we have

n→∞

lim E[(Xn + Yn )2 ] = lim E[Xn2 + 2Xn Yn + Yn2 ]

n→∞ n→∞
= lim E[Xn2 ] + 2 lim E[Xn Yn ] + lim E[Yn2 ]
n→∞ n→∞ n→∞
= 0,

which means ms-lim (Xn + Yn ) = 0.

n→∞

7 This follows from the fact that V ar[Xn ] ≥ 0.

26

Proposition 1.14.2 If the sequences of random variables Xn and Yn converge in the mean
square, then

1. ms-lim (Xn + Yn ) = ms-lim Xn + ms-lim Yn

n→∞ n→∞ n→∞
2. ms-lim (cXn ) = c · ms-lim Xn , ∀c ∈ R.
n→∞ n→∞

Proof: 1. Let ms-lim Xn = L and ms-lim Yn = M . Consider the sequences Xn′ = Xn − L

n→∞ n→∞
and Yn′ = Yn − M . Then ms-lim Xn′ = 0 and ms-lim Yn′ = 0. Applying Lemma 1.14.1 yields
n→∞ n→∞
ms-lim (Xn′ + Yn′ ) = 0.
n→∞
This is equivalent with
ms-lim (Xn − L + Yn − M ) = 0,
n→∞
which becomes
ms-lim (Xn + Yn ) = L + M .
n→∞

1.15 Stochastic Processes

A stochastic process on the probability space (Ω, F , P ) is a family of random variables Xt
parameterized by t ∈ T, where T ⊂ R. If T is an interval we say that Xt is a stochastic
process in continuous time. If T = {1, 2, 3, . . . } we shall say that Xt is a stochastic process in
discrete time. The latter case describes a sequence of random variables. The aforementioned
types of convergence can be easily extended to continuous time. For instance, Xt converges in
the strong sense to X as t → ∞ if
 
P ω; lim Xt (ω) = X(ω) = 1.
t→∞

The evolution in time of a given state of the world ω ∈ Ω given by the function t 7−→ Xt (ω) is
called a path or realization of Xt . The study of stochastic processes using computer simulations
is based on retrieving information about the process Xt given a large number of it realizations.
Consider that all the information accumulated until time t is contained by the σ-field Ft .
This means that Ft contains the information of which events have already occurred until time
t, and which did not. Since the information is growing in time, we have

Fs ⊂ Ft ⊂ F

for any s, t ∈ T with s ≤ t. The family Ft is called a filtration.

A stochastic process Xt is called adapted to the filtration Ft if Xt is Ft - predictable, for any
t ∈ T.

Example 1.15.1 Here there are a few examples of filtrations:

1. Ft represents the information about the evolution of a stock until time t, with t > 0.
2. Ft represents the information about the evolution of a Black-Jack game until time t, with
t > 0.

Example 1.15.2 If X is a random variable, consider the conditional expectation

Xt = E[X|Ft ].
 27

From the definition of conditional expectation, the random variable Xt is Ft -predictable, and
can be regarded as the measurement of X at time t using the information Ft . If the accumulated
knowledge Ft increases and eventually equals the σ-field F , then X = E[X|F ], i.e. we obtain
the entire random variable. The process Xt is adapted to Ft .

Example 1.15.3 Don Joe is asking a doctor how long he still has to live. The age at which he
will pass away is a random variable, denoted by X. Given his medical condition today, which
is contained in Ft , the doctor infers that Mr. Joe will die at the age of Xt = E[X|Ft ]. The
stochastic process Xt is adapted to the medical knowledge Ft .

We shall define next an important type of stochastic process.8

Definition 1.15.4 A process Xt , t ∈ T, is called a martingale with respect to the filtration Ft

if
1. Xt is integrable for each t ∈ T;
2. Xt is adapted to the filtration Ft ;
3. Xs = E[Xt |Fs ], ∀s < t.

Remark
Z 1.15.5 The first condition states that the unconditional forecast is finite E[|Xt ]] =
|Xt | dP < ∞. Condition 2 says that the value Xt is known, given the information set Ft .
Ω
This can be also stated by saying that Xt is Ft -predictable. The third relation asserts that the
best forecast of unobserved future values is the last observation on Xt .

Remark 1.15.6 If the third condition is replaced by

3′ . Xs ≤ E[Xt |Fs ], ∀s ≤ t
then Xt is called a submartingale; and if it is replaced by
3′′ . Xs ≥ E[Xt |Fs ], ∀s ≤ t
then Xt is called a supermartingale.
It is worth noting that Xt is a submartingale if and only if −Xt is a supermartingale.

Example 1.15.1 Let Xt denote Mr. Li Zhu’s salary after t years of work at the same company.
Since Xt is known at time t and it is bounded above, as all salaries are, then the first two
conditions hold. Being honest, Mr. Zhu expects today that his future salary will be the same as
today’s, i.e. Xs = E[Xt |Fs ], for s < t. This means that Xt is a martingale.
If Mr. Zhu is optimistic and believes as of today that his future salary will increase, then
Xt is a submartingale.

Exercise 1.15.7 If X is an integrable random variable on (Ω, F , P ), and Ft is a filtration.

Prove that Xt = E[X|Ft ] is a martingale.

Exercise 1.15.8 Let Xt and Yt be martingales with respect to the filtration Ft . Show that for
any a, b, c ∈ R the process Zt = aXt + bYt + c is a Ft -martingale.

Exercise 1.15.9 Let Xt and Yt be martingales with respect to the filtration Ft .

(a) Is the process Xt Yt always a martingale with respect to Ft ?
(b) What about the processes Xt2 and Yt2 ?
8 The concept of martingale was introduced by Lévy in 1934.
28

Exercise 1.15.10 Two processes Xt and Yt are called conditionally uncorrelated, given Ft , if

E[(Xt − Xs )(Yt − Ys )|Fs ] = 0, ∀ 0 ≤ s < t < ∞.

Let Xt and Yt be martingale processes. Show that the process Zt = Xt Yt is a martingale if and
only if Xt and Yt are conditionally uncorrelated. Assume that Xt , Yt and Zt are integrable.

In the following, if Xt is a stochastic process, the minimum amount of information resulted

from knowing the process Xt until time t is denoted by Ft = σ(Xs ; s ≤ t). In the case of a
discrete process, we have Fn = σ(Xk ; k ≤ n).

Exercise 1.15.11 Let Xn , n ≥ 0 be a sequence of integrable independent random variables,

with E[Xn ] < ∞, for all n ≥ 0. Let S0 = X0 , Sn = X0 + · · · + Xn . Show the following:
(a) Sn − E[Sn ] is an Fn -martingale.
(b) If E[Xn ] = 0 and E[Xn2 ] < ∞, ∀n ≥ 0, then Sn2 − V ar(Sn ) is an Fn -martingale.
(c) If E[Xn ] ≥ 0, then Sn is an Fn -submartingale.

Exercise 1.15.12 Let Xn , n ≥ 0 be a sequence of independent, integrable random variables

such that E[Xn ] = 1 for n ≥ 0. Prove that Pn = X0 · X1 · · · · Xn is an Fn -martingale.

Exercise 1.15.13 (a) Let X be a normally distributed random variable with mean µ 6= 0 and
variance σ 2 . Prove that there is a unique θ 6= 0 such that E[eθX ] = 1.
(b) Let (Xi )i≥0 be a sequence ofPidentically normally distributed random variables with mean
n
µ 6= 0. Consider the sum Sn = j=0 Xj . Show that Zn = eθSn is a martingale, with θ defined
in part (a).

In section 9.1 we shall encounter several processes which are martingales.

Chapter 2

Useful Stochastic Processes

This chapter deals with the most common used stochastic processes and their basic properties.
The two main basic processes are the Brownian motion and the Poisson process. The other
processes described in this chapter are derived from the previous two.

2.1 The Brownian Motion

The observation made first by the botanist Robert Brown in 1827, that small pollen grains
suspended in water have a very irregular and unpredictable state of motion, led to the definition
of the Brownian motion, which is formalized in the following:

Definition 2.1.1 A Brownian motion process is a stochastic process Bt , t ≥ 0, which satisfies

1. The process starts at the origin, B0 = 0;
2. Bt has stationary, independent increments;
3. The process Bt is continuous in t;
4. The increments Bt − Bs are normally distributed with mean zero and variance |t − s|,

Bt − Bs ∼ N (0, |t − s|).

The process Xt = x + Bt has all the properties of a Brownian motion that starts at x. Since
Bt − Bs is stationary, its distribution function depends only on the time interval t − s, i.e.

P (Bt+s − Bs ≤ a) = P (Bt − B0 ≤ a) = P (Bt ≤ a).

It is worth noting that even if Bt is continuous, it is nowhere differentiable. From condition 4

we get that Bt is normally distributed with mean E[Bt ] = 0 and V ar[Bt ] = t

Bt ∼ N (0, t).

This implies also that the second moment is E[Bt2 ] = t. Let 0 < s < t. Since the increments
are independent, we can write

E[Bs Bt ] = E[(Bs − B0 )(Bt − Bs ) + Bs2 ] = E[Bs − B0 ]E[Bt − Bs ] + E[Bs2 ] = s.

Consequently, Bs and Bt are not independent.

29
30

Condition 4 has also a physical explanation. A pollen grain suspended in water is kicked
by a very large numbers of water molecules. The influence of each molecule on the grain is
independent of the other molecules. These effects are average out into a resultant increment
of the grain coordinate. According to the Central Limit Theorem, this increment has to be
normal distributed.

Proposition 2.1.2 A Brownian motion process Bt is a martingale with respect to the infor-
mation set Ft = σ(Bs ; s ≤ t).

Proof: The integrability of Bt follows from Jensen’s inequality

E[|Bt |]2 ≤ E[Bt2 ] = V ar(Bt ) = |t| < ∞.

Bt is obviously Ft -predictable. Let s < t and write Bt = Bs + (Bt − Bs ). Then

E[Bt |Fs ] = E[Bs + (Bt − Bs )|Fs ]

= E[Bs |Fs ] + E[Bt − Bs |Fs ]
= Bs + E[Bt − Bs ] = Bs + E[Bt−s − B0 ] = Bs ,

where we used that Bs is Fs -predictable (from where E[Bs |Fs ] = Bs ) and that the increment
Bt − Bs is independent of previous values of Bt contained in the information set Ft = σ(Bs ; s ≤
t).

A process with similar properties as the Brownian motion was introduced by Wiener.

Definition 2.1.3 A Wiener process Wt is a process adapted to a filtration Ft such that

1. The process starts at the origin, W0 = 0;
2. Wt is an Ft -martingale with E[Wt2 ] < ∞ for all t ≥ 0 and

E[(Wt − Ws )2 ] = t − s, s ≤ t;

3. The process Wt is continuous in t.

Since Wt is a martingale, its increments are unpredictable1 and hence E[Wt − Ws ] = 0; in

particular E[Wt ] = 0. It is easy to show that

V ar[Wt − Ws ] = |t − s|, V ar[Wt ] = t.

Exercise 2.1.4 Show that a Brownian process Bt is a Winer process.

The only property Bt has and Wt seems not to have is that the increments are normally
distributed. However, there is no distinction between these two processes, as the following
result states.

Theorem 2.1.5 (Lévy) A Wiener process is a Brownian motion process.

In stochastic calculus we often need to use infinitesimal notation and its properties. If dWt
denotes the infinitesimal increment of a Wiener process in the time interval dt, the aforemen-
tioned properties become dWt ∼ N (0, dt), E[dWt ] = 0, and E[(dWt )2 ] = dt.
1 This follows from E[Wt − Ws ] = E[Wt − Ws |Fs ] = E[Wt |Fs ] − Ws = Ws − Ws = 0.
 31

Proposition 2.1.6 If Wt is a Wiener process with respect to the information set Ft , then
Yt = Wt2 − t is a martingale.

Proof: Yt is integrable since

E[|Yt |] ≤ E[Wt2 + t] = 2t < ∞, t > 0.

Let s < t. Using that the increments Wt −Ws and (Wt −Ws )2 are independent of the information
set Fs and applying Proposition 1.11.4 yields

E[Wt2 |Fs ] = E[(Ws + Wt − Ws )2 |Fs ]

= E[Ws2 + 2Ws (Wt − Ws ) + (Wt − Ws )2 |Fs ]
= E[Ws2 |Fs ] + E[2Ws (Wt − Ws )|Fs ] + E[(Wt − Ws )2 |Fs ]
= Ws2 + 2Ws E[Wt − Ws |Fs ] + E[(Wt − Ws )2 |Fs ]
= Ws2 + 2Ws E[Wt − Ws ] + E[(Wt − Ws )2 ]
= Ws2 + t − s,

and hence E[Wt2 − t|Fs ] = Ws2 − s, for s < t.

The following result states the memoryless property of Brownian motion2 Wt .
Proposition 2.1.7 The conditional distribution of Wt+s , given the present Wt and the past
Wu , 0 ≤ u < t, depends only on the present.
Proof: Using the independent increment assumption, we have

P (Wt+s ≤ c|Wt = x, Wu , 0 ≤ u < t)

= P (Wt+s − Wt ≤ c − x|Wt = x, Wu , 0 ≤ u < t)
= P (Wt+s − Wt ≤ c − x)
= P (Wt+s ≤ c|Wt = x).

Since Wt is normally distributed with mean 0 and variance t, its density function is
1 − x2
φt (x) = √ e 2t .
2πt
Then its distribution function is
Z x
1 u2
Ft (x) = P (Wt ≤ x) = √ e− 2t du
2πt −∞

The probability that Wt is between the values a and b is given by

Z b
1 u2
P (a ≤ Wt ≤ b) = √ e− 2t du, a < b.
2πt a

Even if the increments of a Brownian motion are independent, their values are still corre-
lated.
2 These type of processes are called Marcov processes.
32

Proposition 2.1.8 Let 0 ≤ s ≤ t. Then

1. Cov(Ws , Wt ) = s;r
s
2. Corr(Ws , Wt ) = .
t
Proof: 1. Using the properties of covariance

Cov(Ws , Wt ) = Cov(Ws , Ws + Wt − Ws )
= Cov(Ws , Ws ) + Cov(Ws , Wt − Ws )
= V ar(Ws ) + E[Ws (Wt − Ws )] − E[Ws ]E[Wt − Ws ]
= s + E[Ws ]E[Wt − Ws ]
= s,

since E[Ws ] = 0.
We can also arrive at the same result starting from the formula

Cov(Ws , Wt ) = E[Ws Wt ] − E[Ws ]E[Wt ] = E[Ws Wt ].

Using that conditional expectations have the same expectation, factoring the predictable part
out, and using that Wt is a martingale, we have

E[Ws Wt ] = E[E[Ws Wt |Fs ]] = E[Ws E[Wt |Fs ]]

= E[Ws Ws ] = E[Ws2 ] = s,

so Cov(Ws , Wt ) = s.
2. The correlation formula yields

r
Cov(Ws , Wt ) s s
Corr(Ws , Wt ) = = √ √ = .
σ(Wt )σ(Ws ) s t t

Remark 2.1.9 Removing the order relation between s and t, the previous relations can also be
stated as

Cov(Ws , Wt ) = min{s, t};

s
min{s, t}
Corr(Ws , Wt ) = .
max{s, t}

The following exercises state the translation and the scaling invariance of the Brownian
motion.
Exercise 2.1.10 For any t0 ≥ 0, show that the process Xt = Wt+t0 − Wt0 is a Brownian
motion. This can be also stated as saying that the Brownian motion is translation invariant.

Exercise 2.1.11 For any λ > 0, show that the process Xt = √1 Wλt is a Brownian motion.
λ
This says that the Brownian motion is invariant by scaling.
 33

50 100 150 200 250 300 350

4

3
-0.5

-1.0

50 100 150 200 250 300 350

a b

Figure 2.1: a Three simulations of the Brownian motion process Wt ; b Two simulations of the
geometric Brownian motion process eWt .

Exercise 2.1.12 Let 0 < s < t < u. Show the following multiplicative property

Corr(Ws , Wt )Corr(Wt , Wu ) = Corr(Ws , Wu ).

Exercise 2.1.13 Find the expectations E[Wt3 ] and E[Wt4 ].

Exercise 2.1.14 (a) Use the martingale property of Wt2 − t to find E[(Wt2 − t)(Ws2 − s)];
(b) Evaluate E[Wt2 Ws2 ];
(c) Compute Cov(Wt2 , Ws2 );
(d) Find Corr(Wt2 , Ws2 ).

Exercise 2.1.15 Consider the process Yt = tW 1t , t > 0, and define Y0 = 0.

(a) Find the distribution of Yt ;
(b) Find the probability density of Yt ;
(c) Find Cov(Ys , Yt );
(d) Find E[Yt − Ys ] and V ar(Yt − Ys ) for s < t.

Exercise 2.1.16 The process Xt = |Wt | is called Brownian motion reflected at the origin.
Show that p
(a) E[|Wt |] = 2t/π;
(b) V ar(|Wt |) = (1 − π2 )t.

Exercise 2.1.17 Let 0 < s < t. Find E[Wt2 |Fs ].

Exercise 2.1.18 Let 0 < s < t. Show that

(a) E[Wt3 |Fs ] = 3(t − s)Ws + Ws3 ;
(b) E[Wt4 |Fs ] = 3(t − s)2 + 6(t − s)Ws2 + Ws4 .

Exercise 2.1.19 Show that the following processes are Brownian motions
(a) Xt = WT − WT −t , 0 ≤ t ≤ T ;
(b) Yt = −Wt , t ≥ 0.
34

2.2 Geometric Brownian Motion

The process Xt = eWt , t ≥ 0 is called geometric Brownian motion. A few simulations of this
process are contained in Fig.2.1 b. The following result will be useful in the following.
2
Lemma 2.2.1 E[eαWt ] = eα t/2
, for α ≥ 0.

Proof: Using the definition of expectation

Z Z
1 x2
E[eαWt ] = eαx φt (x) dx = √ e− 2t +αx dx
2πt
2
= eα t/2
,

where we have used the integral formula

Z r
−ax2 +bx π b2
e dx = e 4a , a>0
a
1
with a = 2t and b = α.

Proposition 2.2.2 The geometric Brownian motion Xt = eWt is log-normally distributed with
mean et/2 and variance e2t − et .

Proof: Since Wt is normally distributed, then Xt = eWt will have a log-normal distribution.
Using Lemma 2.2.1 we have

E[Xt ] = E[eWt ] = et/2

E[Xt2 ] = E[e2Wt ] = e2t ,

and hence the variance is

V ar[Xt ] = E[Xt2 ] − E[Xt ]2 = e2t − (et/2 )2 = e2t − et .

The distribution function of Xt = eWt can be obtained by reducing it to the distribution

function of a Brownian motion.

FXt (x) = P (Xt ≤ x) = P (eWt ≤ x)

= P (Wt ≤ ln x) = FWt (ln x)
Z ln x
1 u2
= √ e− 2t du.
2πt −∞

The density function of the geometric Brownian motion Xt = eWt is given by

 1 2

 √ e−(ln x) /(2t) , if x > 0,
d x 2πt
p(x) = F (x) =
dx Xt 

0, elsewhere.
 35

Exercise 2.2.3 Show that

t−s
E[eWt −Ws ] = e 2 , s < t.

Exercise 2.2.4 Let Xt = eWt .

(a) Show that Xt is not a martingale.
t
(b) Show that e− 2 Xt is a martingale.
1 2
(c) Show that for any constant c ∈ R, the process Yt = ecWt − 2 c t is a martingale.

Exercise 2.2.5 If Xt = eWt , find Cov(Xs , Xt )

(a) by direct computation;
(b) by using Exercise 2.2.4 (b).

Exercise 2.2.6 Show that


2Wt2 (1 − 4t)−1/2 , 0 ≤ t < 1/4
E[e ]=
∞, otherwise.

2.3 Integrated Brownian Motion

The stochastic process
Z t
Zt = Ws ds, t≥0
0

is called integrated Brownian motion. Obviously, Z0 = 0.

Let 0 = s0 < s1 < · · · < sk < · · · sn = t, with sk = kt
n . Then Zt can be written as a limit of
Riemann sums
n
X Ws1 + · · · + Wsn
Zt = lim Wsk ∆s = t lim ,
n→∞ n→∞ n
k=1
t
where ∆s = sk+1 − sk = n.
We are tempted to apply the Central Limit Theorem at this point,
but Wsk are not independent, so we first need to transform the sum into a sum of independent
normally distributed random variables. A straightforward computation shows that

Ws1 + · · · + Wsn
= n(Ws1 − W0 ) + (n − 1)(Ws2 − Ws1 ) + · · · + (Wsn − Wsn−1 )
= X1 + X2 + · · · + Xn . (2.3.1)

Since the increments of a Brownian motion are independent and normally distributed, we have


X1 ∼ N 0, n2 ∆s


X2 ∼ N 0, (n − 1)2 ∆s


X3 ∼ N 0, (n − 2)2 ∆s
..
.


Xn ∼ N 0, ∆s .

Recall now the following variant of the Central Limit Theorem:

36

Theorem 2.3.1 If Xj are independent random variables normally distributed with mean µj
and variance σj2 , then the sum X1 +· · ·+Xn is also normally distributed with mean µ1 +· · ·+µn
and variance σ12 + · · · + σn2 .
Then
   n(n + 1)(2n + 1) 
X1 + · · · + Xn ∼ N 0, (1 + 22 + 32 + · · · + n2 )∆s = N 0, ∆s ,
6
t
with ∆s = . Using (2.3.1) yields
n
Ws + · · · + Wsn  (n + 1)(2n + 1) 
t 1 ∼ N 0, t3 .
n 6n2
“Taking the limit” we get
 t3 
Zt ∼ N 0, .
3
Proposition 2.3.2 The integrated Brownian motion Zt has a normal distribution with mean
0 and variance t3 /3.
Remark 2.3.3 The aforementioned limit was taken heuristically, without specifying the type
of the convergence. In order to make this to work, the following result is usually used:
If Xn is a sequence of normal random variables that converges in mean square to X, then
the limit X is normal distributed, with E[Xn ] → E[X] and V ar(Xn ) → V ar(X), as n → ∞.
The mean and the variance can also be computed in a direct way as follows. By Fubini’s
theorem we have
Z t Z Z t
E[Zt ] = E[ Ws ds] = Ws ds dP
0 R 0
Z tZ Z t
= Ws dP ds = E[Ws ] ds = 0,
0 R 0

since E[Ws ] = 0. Then the variance is given by

V ar[Zt ] = E[Zt2 ] − E[Zt ]2 = E[Zt2 ]
Z t Z t Z tZ t
= E[ Wu du · Wv dv] = E[ Wu Wv dudv]
0 0 0 0
Z tZ t ZZ
= E[Wu Wv ] dudv = min{u, v} dudv
0 0 [0,t]×[0,t]
ZZ ZZ
= min{u, v} dudv + min{u, v} dudv, (2.3.2)
D1 D2

where
D1 = {(u, v); u > v, 0 ≤ u ≤ t}, D2 = {(u, v); u < v, 0 ≤ u ≤ t}
The first integral can be evaluated using Fubini’s theorem
ZZ ZZ
min{u, v} dudv = v dudv
D1 D1
Z tZ u Z t
 u2 t3
= v dv du = du = .
0 0 0 2 6
 37

Similarly, the latter integral is equal to

ZZ
t3
min{u, v} dudv = .
D2 6

Substituting in (2.3.2) yields

t3 t3 t3
V ar[Zt ] = + = .
6 6 3

Exercise 2.3.4 (a) Prove that the moment generating function of Zt is

2 3
m(u) = eu t /6
.

(b) Use the first part to find the mean and variance of Zt .

Exercise 2.3.5 Let s < t. Show that the covariance of the integrated Brownian motion is given
by
  t s
Cov Zs , Zt = s2 − , s < t.
2 6
Exercise 2.3.6 Show that
(a) Cov(Zt , Zt −Zt−h ) = 12 t2 h+o(h), where o(h) denotes a quantity such that limh→0 o(h)/h =
0;

t2
(b) Cov(Zt , Wt ) = .
2
Exercise 2.3.7 Show that
u+s
E[eWs +Wu ] = e 2 emin{s,u} .
Z t
Exercise 2.3.8 Consider the process Xt = eWs ds.
0
(a) Find the mean of Xt ;
(b) Find the variance of Xt .
Z t
Exercise 2.3.9 Consider the process Zt = Wu du, t > 0.
0
(a) Show that E[ZT |Ft ] = Zt + Wt (T − t), for any t < T ;
(b) Prove that the process Mt = Zt − tWt is an Ft -martingale.

2.4 Exponential Integrated Brownian Motion

Rt
If Zt = 0 Ws ds denotes the integrated Brownian motion, the process

Vt = eZt

is called exponential integrated Brownian motion. The process starts at V0 = e0 = 1. Since

Zt is normally distributed, then Vt is log-normally distributed. We compute the mean and the
38

variance in a direct way. Using Exercises 2.2.5 and 2.3.4 we have

t3
E[Vt ] = E[eZt ] = m(1) = e 6
4t3 2t3
E[Vt2 ] = E[e2Zt ] = m(2) = e 6 =e 3

2t3 t3
V ar(Vt ) = E[Vt2 ] − E[Vt ]2 = e 3 −e 3
t+3s
Cov(Vs , Vt ) = e 2 .
(T −t)3
Exercise 2.4.1 Show that E[VT |Ft ] = Vt e(T −t)Wt + 3 for t < T .

2.5 Brownian Bridge

The process Xt = Wt − tW1 is called the Brownian bridge fixed at both 0 and 1. Since we can
also write

Xt = Wt − tWt − tW1 + tWt

= (1 − t)(Wt − W0 ) − t(W1 − Wt ),

using that the increments Wt − W0 and W1 − Wt are independent and normally distributed,
with
Wt − W0 ∼ N (0, t), W1 − Wt ∼ N (0, 1 − t),
it follows that Xt is normally distributed with

E[Xt ] = (1 − t)E[(Wt − W0 )] − tE[(W1 − Wt )] = 0

V ar[Xt ] = (1 − t)2 V ar[(Wt − W0 )] + t2 V ar[(W1 − Wt )]
= (1 − t)2 (t − 0) + t2 (1 − t)
= t(1 − t).

This can be also stated by saying that the Brownian bridge

tied at 0 and 1 is a Gaussian process
with mean 0 and variance t(1 − t), so Xt ∼ N 0, t(1 − t) .

Exercise 2.5.1 Let Xt = Wt − tW1 , 0 ≤ t ≤ 1 be a Brownian bridge fixed at 0 and 1. Let

Yt = Xt2 . Show that Y0 = Y1 = 0 and find E[Yt ] and V ar(Yt ).

2.6 Brownian Motion with Drift

The process Yt = µt + Wt , t ≥ 0, is called Brownian motion with drift. The process Yt tends to
drift off at a rate µ. It starts at Y0 = 0 and it is a Gaussian process with mean

E[Yt ] = µt + E[Wt ] = µt

and variance
V ar[Yt ] = V ar[µt + Wt ] = V ar[Wt ] = t.

Exercise 2.6.1 Find the distribution and the density functions of the process Yt .
 39

2.7 Bessel Process

This section deals with the process satisfied by the Euclidean distance from the origin to a
particle following a Brownian motion in Rn . More precisely, if W1 (t), · · · , Wn (t) are independent
Brownian motions, let W (t) = (W1 (t), · · · , Wn (t)) be a Brownian motion in Rn , n ≥ 2. The
process p
Rt = dist(O, W (t)) = W1 (t)2 + · · · + Wn (t)2
is called n-dimensional Bessel process.
The probability density of this process is given by the following result.
Proposition 2.7.1 The probability density function of Rt , t > 0 is given by

ρ2
 (2t)n/22Γ(n/2) ρn−1 e− 2t , ρ ≥ 0;

pt (ρ) =


0, ρ<0
with  n
n  ( 2 − 1)! for n even;
Γ = √
2
( n2 − 1)( n2 − 2) · · · 32 21 π, for n odd.


Proof: Since the Brownian motions W1 (t), . . . , Wn (t) are independent, their joint density
function is
fW1 ···Wn (t) = fW1 (t) · · · fWn (t)
1 2 2
= n/2
e−(x1 +···+xn )/(2t) , t > 0.
(2πt)
In the next computation we shall use the following formula of integration that follows from
the use of polar coordinates
Z Z ρ
n−1
f (x) dx = σ(S ) rn−1 g(r) dr, (2.7.3)
{|x|≤ρ} 0
n
where f (x) = g(|x|) is a function on R with spherical symmetry, and where
2π n/2
σ(Sn−1 ) =
Γ(n/2)
is the area of the (n − 1)-dimensional sphere in Rn .
Let ρ ≥ 0. The distribution function of Rt is
Z
FR (ρ) = P (Rt ≤ ρ) = fW1 ···Wn (t) dx1 · · · dxn
{Rt ≤ρ}
Z
1 2 2
= n/2
e−(x1 +···+xn )/(2t) dx1 · · · dxn
2 2
x1 +···+xn ≤ρ 2 (2πt)
Z ρ Z !
1 2 2
= rn−1 n/2
e−(x1 +···+xn )/(2t) dσ dr
0 S(0,1) (2πt)
Z
σ(Sn−1 ) ρ n−1 −r2 /(2t)
= r e dr.
(2πt)n/2 0
40

Differentiating yields

d σ(Sn−1 ) n−1 − ρ2
pt (ρ) = FR (ρ) = ρ e 2t
dρ (2πt)n/2
2 ρ2
= n/2
ρn−1 e− 2t , ρ > 0, t > 0.
(2t) Γ(n/2)

It is worth noting that in the 2-dimensional case the aforementioned density becomes a
particular case of a Weibull distribution with parameters m = 2 and α = 2t, called Wald’s
distribution
1 x2
pt (x) = xe− 2t , x > 0, t > 0.
t

Exercise 2.7.2 Let P (Rt ≤ t) be the probability of a 2-dimensional Brownian motion to be

inside of the disk D(0, ρ) at time t > 0. Show that

ρ2  ρ2  ρ2
1− < P (Rt ≤ t) < .
2t 4t 2t

Exercise 2.7.3 Let Rt be a 2-dimensional Bessel process. Show that

√
(a) E[Rt ] = 2πt/2;


(b) V ar(Rt ) = 2t 1 − π4 .

Rt
Exercise 2.7.4 Let Xt = , t > 0, where Rt is a 2-dimensional Bessel process. Show that
t
Xt → 0 as t → ∞ in mean square.

2.8 The Poisson Process

A Poisson process describes the number of occurrences of a certain event before time t, such
as: the number of electrons arriving at an anode until time t; the number of cars arriving at a
gas station until time t; the number of phone calls received on a certain day until time t; the
number of visitors entering a museum on a certain day until time t; the number of earthquakes
that occurred in Chile during the time interval [0, t]; the number of shocks in the stock market
from the beginning of the year until time t; the number of twisters that might hit Alabama
from the beginning of the century until time t.

2.8.1 Definition and Properties

The definition of a Poisson process is stated more precisely in the following.
 41

Definition 2.8.1 A Poisson process is a stochastic process Nt , t ≥ 0, which satisfies

1. The process starts at the origin, N0 = 0;
2. Nt has stationary, independent increments;
3. The process Nt is right continuous in t, with left hand limits;
4. The increments Nt − Ns , with 0 < s < t, have a Poisson distribution with parameter
λ(t − s), i.e.
λk (t − s)k −λ(t−s)
P (Nt − Ns = k) = e .
k!
It can be shown that condition 4 in the previous definition can be replaced by the following
two conditions:
P (Nt − Ns = 1) = λ(t − s) + o(t − s) (2.8.4)
P (Nt − Ns ≥ 2) = o(t − s), (2.8.5)
where o(h) denotes a quantity such that limh→0 o(h)/h = 0. Then the probability that a jump
of size 1 occurs in the infinitesimal interval dt is equal to λdt, and the probability that at least
2 events occur in the same small interval is zero. This implies that the random variable dNt
may take only two values, 0 and 1, and hence satisfies
P (dNt = 1) = λ dt (2.8.6)
P (dNt = 0) = 1 − λ dt. (2.8.7)
Exercise 2.8.2 Show that if condition 4 is satisfied, then conditions (2.8.4 − 2.8.5) hold.
Exercise 2.8.3 Which of the following expressions are o(h)?
(a) f (h) = 3h2 + h;
√
(b) f (h) = h + 5;
(c) f (h) = h ln |h|;
(d) f (h) = heh .
The fact that Nt − Ns is stationary can be stated as
n
X (λt)k
P (Nt+s − Ns ≤ n) = P (Nt − N0 ≤ n) = P (Nt ≤ n) = e−λt .
k!
k=0

From condition 4 we get the mean and variance of increments

E[Nt − Ns ] = λ(t − s), V ar[Nt − Ns ] = λ(t − s).
In particular, the random variable Nt is Poisson distributed with E[Nt ] = λt and V ar[Nt ] = λt.
The parameter λ is called the rate of the process. This means that the events occur at the
constant rate λ.
Since the increments are independent, we have for 0 < s < t
E[Ns Nt ] = E[(Ns − N0 )(Nt − Ns ) + Ns2 ]
= E[Ns − N0 ]E[Nt − Ns ] + E[Ns2 ]
= λs · λ(t − s) + (V ar[Ns ] + E[Ns ]2 )
= λ2 st + λs. (2.8.8)
As a consequence we have the following result:
42

Proposition 2.8.4 Let 0 ≤ s ≤ t. Then

1. Cov(Ns , Nt ) = λs;
r
s
2. Corr(Ns , Nt ) = .
t
Proof: 1. Using (2.8.8) we have

Cov(Ns , Nt ) = E[Ns Nt ] − E[Ns ]E[Nt ]

= λ2 st + λs − λsλt
= λs.

2. Using the formula for the correlation yields

r
Cov(Ns , Nt ) λs s
Corr(Ns , Nt ) = = = .
(V ar[Ns ]V ar[Nt ])1/2 (λsλt)1/2 t

It worth noting the similarity with Proposition 2.1.8.

Proposition 2.8.5 Let Nt be Ft -adapted. Then the process Mt = Nt − λt is an Ft -martingale.

Proof: Let s < t and write Nt = Ns + (Nt − Ns ). Then

E[Nt |Fs ] = E[Ns + (Nt − Ns )|Fs ]

= E[Ns |Fs ] + E[Nt − Ns |Fs ]
= Ns + E[Nt − Ns ]
= Ns + λ(t − s),

where we used that Ns is Fs -predictable (and hence E[Ns |Fs ] = Ns ) and that the increment
Nt − Ns is independent of previous values of Ns and the information set Fs . Subtracting λt
yields
E[Nt − λt|Fs ] = Ns − λs,
or E[Mt |Fs ] = Ms . Since it is obvious that Mt is integrable and Ft -adapted, it follows that Mt
is a martingale.
It is worth noting that the Poisson process Nt is not a martingale. The martingale process
Mt = Nt − λt is called the compensated Poisson process.

Exercise 2.8.6 Compute E[Nt2 |Fs ] for s < t. Is the process Nt2 an Fs -martingale?

Exercise 2.8.7 (a) Show that the moment generating function of the random variable Nt is
x
mNt (x) = eλt(e −1)
.

(b) Deduct the expressions for the first few moments

E[Nt ] = λt
E[Nt2 ] = λ2 t2 + λt
E[Nt3 ] = λ3 t3 + 3λ2 t2 + λt
E[Nt4 ] = λ4 t4 + 6λ3 t3 + 7λ2 t2 + λt.
 43

(c) Show that the first few central moments are given by

E[Nt − λt] = 0
E[(Nt − λt)2 ] = λt
E[(Nt − λt)3 ] = λt
E[(Nt − λt)4 ] = 3λ2 t2 + λt.

Exercise 2.8.8 Find the mean and variance of the process Xt = eNt .

Exercise 2.8.9 (a) Show that the moment generating function of the random variable Mt is
x
mMt (x) = eλt(e −x−1)
.

(b) Let s < t. Verify that

E[Mt − Ms ] = 0,
E[(Mt − Ms )2 ] = λ(t − s),
E[(Mt − Ms )3 ] = λ(t − s),
E[(Mt − Ms )4 ] = λ(t − s) + 3λ2 (t − s)2 .

Exercise 2.8.10 Let s < t. Show that

V ar[(Mt − Ms )2 ] = λ(t − s) + 2λ2 (t − s)2 .

2.8.2 Interarrival times

For each state of the world, ω, the path t → Nt (ω) is a step function that exhibits unit jumps.
Each jump in the path corresponds to an occurrence of a new event. Let T1 be the random
variable which describes the time of the 1st jump. Let T2 be the time between the 1st jump
and the second one. In general, denote by Tn the time elapsed between the (n − 1)th and nth
jumps. The random variables Tn are called interarrival times.

Proposition 2.8.11 The random variables Tn are independent and exponentially distributed
with mean E[Tn ] = 1/λ.

Proof: We start by noticing that the events {T1 > t} and {Nt = 0} are the same, since both
describe the situation that no events occurred until after time t. Then

P (T1 > t) = P (Nt = 0) = P (Nt − N0 = 0) = e−λt ,

and hence the distribution function of T1 is

FT1 (t) = P (T1 ≤ t) = 1 − P (T1 > t) = 1 − e−λt .

Differentiating yields the density function

d
fT1 (t) = FT (t) = λe−λt .
dt 1
It follows that T1 is has an exponential distribution, with E[T1 ] = 1/λ.
44

In order to show that the random variables T1 and T2 are independent, if suffices to show that

P (T2 ≤ t) = P (T2 ≤ t|T1 = s),

i.e. the distribution function of T2 is independent of the values of T1 . We note first that from
the independent increments property


P 0 jumps in (s, s + t], 1 jump in (0, s] = P (Ns+t − Ns = 0, Ns − N0 = 1)



= P (Ns+t − Ns = 0)P (Ns − N0 = 1) = P 0 jumps in (s, s + t] P 1 jump in (0, s] .

Then the conditional distribution of T2 is

F (t|s) = P (T2 ≤ t|T1 = s) = 1 − P (T2 > t|T1 = s)

P (T2 > t, T1 = s)
= 1−
P (T1 = s)


P 0 jumps in (s, s + t], 1 jump in (0, s]
= 1−
P (T1 = s)



P 0 jumps in (s, s + t] P 1 jump in (0, s]

= 1−
= 1 − P 0 jumps in (s, s + t]
P 1 jump in (0, s]
= 1 − P (Ns+t − Ns = 0) = 1 − e−λt ,

which is independent of s. Then T2 is independent of T1 and exponentially distributed. A

similar argument for any Tn leads to the desired result.

2.8.3 Waiting times

The random variable Sn = T1 + T2 + · · · + Tn is called the waiting time until the nth jump.
The event {Sn ≤ t} means that there are n jumps that occurred before or at time t, i.e. there
are at least n events that happened up to time t; the event is equal to {Nt ≥ n}. Hence the
distribution function of Sn is given by
∞
X (λt)k
FSn (t) = P (Sn ≤ t) = P (Nt ≥ n) = e−λt ·
k!
k=n
Differentiating we obtain the density function of the waiting time Sn
d λe−λt (λt)n−1
fSn (t) = FSn (t) = ·
dt (n − 1)!
Writing
tn−1 e−λt
fSn (t) = ,
(1/λ)n Γ(n)
it turns out that Sn has a gamma distribution with parameters α = n and β = 1/λ. It follows
that
n n
E[Sn ] = , V ar[Sn ] = 2 ·
λ λ
The relation lim E[Sn ] = ∞ states that the expectation of the waiting time is unbounded as
n→∞
n → ∞.
 45

Figure 2.2: The Poisson process Nt and the waiting times S1 , S2 , · · · Sn . The shaded rectangle
has area n(Sn+1 − t).

d λe−λt (λt)n−1
Exercise 2.8.12 Prove that FSn (t) = ·
dt (n − 1)!
Exercise 2.8.13 Using that the interarrival times T1 , T2 , · · · are independent and exponentially
distributed, compute directly the mean E[Sn ] and variance V ar(Sn ).

2.8.4 The Integrated Poisson Process

The function u → Nu is continuous with the exception of a set of countable jumps of size 1. It
is known that such functions are Riemann integrable, so it makes sense to define the process
Z t
Ut = Nu du,
0

called the integrated Poisson process. The next result provides a relation between the process
Ut and the partial sum of the waiting times Sk .

Proposition 2.8.14 The integrated Poisson process can be expressed as

Nt
X
Ut = tNt − Sk .
k=1

Let Nt = n. Since Nt is equal to k between the waiting times Sk and Sk+1 , the process Ut ,
which is equal to the area of the subgraph of Nu between 0 and t, can be expressed as
Z t
Ut = Nu du = 1 · (S2 − S1 ) + 2 · (S3 − S2 ) + · · · + n(Sn+1 − Sn ) − n(Sn+1 − t).
0

Since Sn < t < Sn+1 , the difference of the last two terms represents the area of last the
rectangle, which has the length t − Sn and the height n. Using associativity, a computation
46

yields

1 · (S2 − S1 ) + 2 · (S3 − S2 ) + · · · + n(Sn+1 − Sn ) = nSn+1 − (S1 + S2 + · · · + Sn ).

Substituting in the aforementioned relation yields

Ut = nSn+1 − (S1 + S2 + · · · + Sn ) − n(Sn+1 − t)

= nt − (S1 + S2 + · · · + Sn )
Nt
X
= tNt − Sk ,
k=1

where we replaced n by Nt .
The conditional distribution of the waiting times is provided by the following useful result.

Theorem 2.8.15 Given that Nt = n, the waiting times S1 , S2 , · · · , Sn have the the joint den-
sity function given by
n!
f (s1 , s2 , · · · , sn ) = , 0 < s1 ≤ s2 ≤ · · · ≤ sn < t.
tn
This is the same as the density of an ordered sample of size n from a uniform distribution on
the interval (0, t). A naive explanation of this result is as follows. If we know that there will
be exactly n events during the time interval (0, t), since the events can occur at any time, each
of them can be considered uniformly distributed, with the density f (sk ) = 1/t. Since it makes
sense to consider the events independent, taking into consideration all possible n! permutaions,
n!
the joint density function becomes f (s1 , · · · , sn ) = n!f (s1 ) · · · f (sn ) = n .
t
Exercise 2.8.16 Find the following means
(a) E[Ut ].
Nt
X 
(b) E Sk .
k=1

λt3
Exercise 2.8.17 Show that V ar(Ut ) = .
3
Exercise 2.8.18 Can you apply a similar proof as in Proposition 2.3.2 to show that the inte-
grated Poisson process Ut is also a Poisson process?

Exercise 2.8.19 Let Y : Ω → N be a discrete random variable. Then for any random variable
X we have X
E[X] = E[X|Y = y]P (Y = y).
y≥0

Exercise 2.8.20 Use Exercise 2.8.19 to solve Exercise 2.8.16 (b).

Exercise 2.8.21 (a) Let Tk be the kth interarrival time. Show that
λ
E[e−σTk ] = , σ > 0.
λ+σ
 47

(b) Let n = Nt . Show that

Ut = nt − [nT1 + (n − 1)T2 + · · · + 2Tn−1 + Tn ].

(c) Find the conditional expectation

h i
E e−σUt Nt = n .

(Hint: If know that there are exactly n jumps in the interval [0, T ], it makes sense to consider
the arrival time of the jumps Ti independent and uniformly distributed on [0, T ]).
(d) Find the expectation h i
E e−σUt .

2.9 Submartingales
A stochastic process Xt on the probability space (Ω, F , P ) is called submartingale with respect
to the filtration Ft if:
R
(a) Ω |Xt | dP < ∞ (Xt integrable);

(b) Xt is known if Ft is given (Xt is adaptable to Ft );

(c) E[Xt+s |Ft ] ≥ Xt , ∀t, s ≥ 0 (future predictions exceed the present value).

Example 2.9.1 We shall prove that the process Xt = µt+ σWt , with µ > 0 is a submartingale.

The integrability follows from the inequality |Xt (ω)| ≤ µt + |Wt (ω)| and integrability of Wt .
The adaptability of Xt is obvious, and the last property follows from the computation:

E[Xt+s |Ft ] = E[µt + σWt+s |Ft ] + µs > E[µt + σWt+s |Ft ]

= µt + σE[Wt+s |Ft ] = µt + σWt = Xt ,

where we used that Wt is a martingale.

Example 2.9.2 We shall show that the square of the Brownian motion, Wt2 , is a submartin-
gale.

Using that Wt2 − t is a martingale, we have

2 2
E[Wt+s |Ft ] = E[Wt+s − (t + s)|Ft ] + t + s = Wt2 − t + t + s
= Wt2 + s ≥ Wt2 .

The following result supplies examples of submatingales starting from martingales or sub-
martingales.

Proposition 2.9.3 (a) If Xt is a martingale and φ a convex function such that φ(Xt ) is
integrable, then the process Yt = φ(Xt ) is a submartingale.
(b) If Xt is a submartingale and φ an increasing convex function such that φ(Xt ) is integrable,
then the process Yt = φ(Xt ) is a submartingale.
48

Proof: (a) Using Jensen’s inequality for conditional probabilities, Exercise 1.12.6, we have
 
E[Yt+s |Ft ] = E[φ(Xt+s )|Ft ] ≥ φ E[Xt+s |Ft ] = φ(Xt ) = Yt .

(b) From the submatingale property and monotonicity of φ we have

 
φ E[Xt+s |Ft ] ≥ φ(Xt ).

Then apply a similar computation as in part (a).

Corollary 2.9.4 (a) Let Xt be a martingale. Then Xt2 , |Xt |, eXt are submartingales.
(b) Let µ > 0. Then eµt+σWt is a submartingale.

Proof: (a) Results from part (a) of Proposition 2.9.3.

(b) It follows from Example 2.9 and part (b) of Proposition 2.9.3.

The following result provides important inequalities involving submartingales.

Proposition 2.9.5 (Doob’s Submartingale Inequality) (a) Let Xt be a non-negative sub-

martingale. Then
E[Xt ]
P (sup Xs ≥ x) ≤ , ∀x > 0.
s≤t x
(b) If Xt is a right continuous submartingale, then for any x > 0

E[Xt+ ]
P (sup Xt ≥ x) ≤ ,
s≤t x

where Xt+ = max{Xt , 0}.

Exercise 2.9.6 Let x > 0. Show the inequalities:

t
(a) P (sup Ws2 ≥ x) ≤ .
s≤t x
p
2t/π
(b) P (sup |Ws | ≥ x) ≤ .
s≤t x

sups≤t |Ws |
Exercise 2.9.7 Show that st-lim = 0.
t→∞ t
Exercise 2.9.8 Show that for any martingale Xt we have the inequality
E[Xt2 ]
P (sup Xt2 > x) ≤ , ∀x > 0.
s≤t x

It is worth noting that Doob’s inequality implies Markov’s inequality. Since sup Xs ≥ Xt ,
s≤t
then P (Xt ≥ x) ≤ P (sup Xs ≥ x). Then Doob’s inequality
s≤t

E[Xt ]
P (sup Xs ≥ x) ≤
s≤t x
 49

implies Markov’s inequality (see Theorem 1.12.8)

E[Xt ]
P (Xt ≥ x) ≤ .
x
Exercise 2.9.9 Let Nt denote the Poisson process and consider the information set Ft =
σ{Ns ; s ≤ t}.
(a) Show that Nt is a submartingale;
(b) Is Nt2 a submartingale?

Exercise 2.9.10 It can be shown that for any 0 < σ < τ we have the inequality
h X N 2 i 4τ λ
t
E −λ ≤ 2 ·
t σ
σ≤t≤τ

Nt
Using this inequality prove that ms–lim = λ.
t→∞ t
50
Chapter 3

Properties of Stochastic
Processes

3.1 Stopping Times

Consider the probability space (Ω, F , P ) and the filtration (Ft )t≥0 , i.e.

Ft ⊂ Fs ⊂ F , ∀t < s.

Assume that the decision to stop playing a game before or at time t is determined by the
information Ft available at time t. Then this decision can be modeled by a random variable
τ : Ω → [0, ∞] which satisfies
{ω; τ (ω) ≤ t} ∈ Ft .
This means that given the information set Ft , we know whether the event {ω; τ (ω) ≤ t} had
occurred or not. We note that the possibility τ = ∞ is included, since the decision to continue
the game for ever is also a possible event. However, we ask the condition P (τ < ∞) = 1. A
random variable τ with the previous properties is called a stopping time.
The next example illustrates a few cases when a decision is or is not a stopping time. In order
to accomplish this, think of the situation that τ is the time when some random event related
to a given stochastic process occurs first.

Example 3.1.1 Let Ft be the information available until time t regarding the evolution of a
stock. Assume the price of the stock at time t = 0 is $50 per share. The following decisions are
stopping times:
(a) Sell the stock when it reaches for the first time the price of $100 per share;
(b) Buy the stock when it reaches for the first time the price of $10 per share;
(c) Sell the stock at the end of the year;
(d) Sell the stock either when it reaches for the first time $80 or at the end of the year.
(e) Keep the stock either until the initial investment doubles or until the end of the year;
The following decisions are not stopping times:
(f ) Sell the stock when it reaches the maximum level it will ever be;
(g) Keep the stock until the initial investment at least doubles.

51
52

Part (f ) is not a stopping time because it requires information about the future that is not
contained in Ft . For part (g), since the initial stock price is S0 = $50, the general theory of
stock prices state
P (St ≥ 2S0 ) = P (St ≥ 100) < 1,
i.e. there is a positive probability that the stock never doubles its value. This contradicts the
condition P (τ = ∞) = 0. In part (e) there are two conditions; the latter one has the occurring
probability equal to 1.

Exercise 3.1.2 Show that any positive constant, τ = c, is a stopping time with respect to any
filtration.

Exercise 3.1.3 Let τ (ω) = inf{t > 0; |Wt (ω)| > K}, with K > 0 constant. Show that τ is a
stopping time with respect to the filtration Ft = σ(Ws ; s ≤ t).

The random variable τ is called the first exit time of the Brownian motion Wt from the interval
(−K, K). In a similar way one can define the first exit time of the process Xt from the interval
(a, b):
τ (ω) = inf{t > 0; Xt (ω) ∈
/ (a, b)} = inf{t > 0; Xt (ω) > b or Xt (ω) < a)}.
Let X0 < a. The first entry time of Xt in the interval (a, b) is defined as

τ (ω) = inf{t > 0; Xt (ω) ∈ (a, b)} = inf{t > 0; Xt (ω) > a or Xt (ω) < b)}.

If let b = ∞, we obtain the first hitting time of the level a

τ (ω) = inf{t > 0; Xt (ω) > a)}.

We shall deal with hitting times in more detail in section 3.3.

Exercise 3.1.4 Let Xt be a continuous stochastic process. Prove that the first exit time of Xt
from the interval (a, b) is a stopping time.

We shall present in the following some properties regarding operations with stopping times.
Consider the notations τ1 ∨ τ2 = max{τ1 , τ2 }, τ1 ∧ τ2 = min{τ1 , τ2 }, τ̄n = supn≥1 τn and
τ n = inf n≥1 τn .
Proposition 3.1.5 Let τ1 and τ2 be two stopping times with respect to the filtration Ft . Then
1. τ1 ∨ τ2
2. τ1 ∧ τ2
3. τ1 + τ2
are stopping times.
Proof: 1. We have

{ω; τ1 ∨ τ2 ≤ t} = {ω; τ1 ≤ t} ∩ {ω; τ2 ≤ t} ∈ Ft ,

since {ω; τ1 ≤ t} ∈ Ft and {ω; τ2 ≤ t} ∈ Ft . From the subadditivity of P we get

   
P {ω; τ1 ∨ τ2 = ∞} = P {ω; τ1 = ∞} ∪ {ω; τ2 = ∞}
   
≤ P {ω; τ1 = ∞} + P {ω; τ2 = ∞} .
 53

Since    
P {ω; τi = ∞} = 1 − P {ω; τi < ∞} = 0, i = 1, 2,
   
it follows that P {ω; τ1 ∨ τ2 = ∞} = 0 and hence P {ω; τ1 ∨ τ2 < ∞} = 1. Then τ1 ∨ τ2 is
a stopping time.
2. The event {ω; τ1 ∧ τ2 ≤ t} ∈ Ft if and only if {ω; τ1 ∧ τ2 > t} ∈ Ft .

{ω; τ1 ∧ τ2 > t} = {ω; τ1 > t} ∩ {ω; τ2 > t} ∈ Ft ,

since {ω; τ1 > t} ∈ Ft and {ω; τ2 > t} ∈ Ft (the σ-algebra Ft is closed to complements). The
fact that τ1 ∧ τ2 < ∞ almost surely has a similar proof.

3. We note that τ1 + τ2 ≤ t if there is a c ∈ (0, t) such that

τ1 ≤ c, τ2 ≤ t − c.

Using that the rational numbers are dense in R, we can write

[  
{ω; τ1 + τ2 ≤ t} = {ω; τ1 ≤ c} ∩ {ω; τ2 ≤ t − c} ∈ Ft ,
0<c<t,c∈Q

since
{ω; τ1 ≤ c} ∈ Fc ⊂ Ft , {ω; τ2 ≤ t − c} ∈ Ft−c ⊂ Ft .
Writing
{ω; τ1 + τ2 = ∞} = {ω; τ1 = ∞} ∪ {ω; τ2 = ∞}
yields
     
P {ω; τ1 + τ2 = ∞} ≤ P {ω; τ1 = ∞} + P {ω; τ2 = ∞} = 0,
 
Hence P {ω; τ1 + τ2 < ∞} = 1. It follows that τ1 + τ2 is a stopping time.
∞
\
A filtration Ft is called right-continuous if Ft = Ft+ n1 , for t > 0. This means that the
n=1
information available at time t is a good approximation for any future infinitesimal information
Ft+ǫ ; or, equivalently, nothing more can be learned by peeking infinitesimally far into the
future.

Exercise 3.1.6 (a) Let Ft = σ{Ws ; s ≤ t}, where Wt is a Brownian motion. Show that Ft is
right-continuous.
(b) Let Nt = σ{Ns ; s ≤ t}, where Nt is a Poisson motion. Is Nt right-continuous?

Proposition 3.1.7 Let Ft be right-continuous and (τn )n≥1 be a sequence of bounded stopping
times.
(a) Then supn τ and inf τn are stopping times.
(b) If the sequence (τn )n≥1 converges to τ , τ 6= 0, then τ is a stopping time.
54

Proof: (a) The fact that τ̄n is a stopping time follows from
\
{ω; τ̄n ≤ t} ⊂ {ω; τn ≤ t} ∈ Ft ,
n≥1

and from the boundedness, which implies P (τn < ∞) = 1.

In order to show that τ n is a stopping time we shall proceed as in the following. Using that
Ft is right-continuous and closed to complements, it suffices to show that {ω; τ n ≥ t} ∈ Ft .
This follows from \
{ω; τ n ≥ t} = {ω; τn > t} ∈ Ft .
n≥1

(b) Let lim τn = τ . Then there is an increasing (or decreasing) subsequence τnk of stopping
n→∞
times that tends to τ , so sup τnk = τ (or inf τnk = τ ). Since τnk are stopping times, by part
(a), it follows that τ is a stopping time.
The condition that τn is bounded is significant, since if take τn = n as stopping times, then
supn τn = ∞ with probability 1, which does not satisfy the stopping time definition.
Exercise 3.1.8 Let τ be a stopping time.
(a) Let c ≥ 1 be a constant. Show that cτ is a stopping time.
(b) Let f : [0, ∞) → R be a continuous, increasing function satisfying f (t) ≥ t. Prove that
f (τ ) is a stopping time.
(c) Show that eτ is a stopping time.
Exercise 3.1.9 Let τ be a stopping time and c > 0 a constant. Prove that τ + c is a stopping
time.
Exercise 3.1.10 Let a be a constant and define τ = inf{t ≥ 0; Wt = a}. Is τ a stopping time?
Exercise 3.1.11 Let τ be a stopping time. Consider the following sequence τn = (m + 1)2−n
if m2−n ≤ τ < (m + 1)2−n (stop at the first time of the form k2−n after τ ). Prove that τn is
a stopping time.

3.2 Stopping Theorem for Martingales

The next result states that in a fair game, the expected final fortune of a gambler, who is
using a stopping time to quit the game, is the same as the expected initial fortune. From
the financially point of view, the theorem says that if you buy an asset at some initial time
and adopt a strategy of deciding when to sell it, then the expected price at the selling time is
the initial price; so one cannot make money by buying and selling an asset whose price is a
martingale. Fortunately, the price of a stock is not a martingale, and people can still expect to
make money buying and selling stocks.
If (Mt )t≥0 is an Ft -martingale, then taking the expectation in
E[Mt |Fs ] = Ms , ∀s < t
and using Example 1.11.2 yields
E[Mt ] = E[Ms ], ∀s < t.
In particular, E[Mt ] = E[M0 ], for any t > 0. The next result states necessary conditions under
which this identity holds if t is replaced by any stopping time τ .
 55

Theorem 3.2.1 (Optional Stopping Theorem) Let (Mt )t≥0 be a right continuous Ft -martingale
and τ be a stopping time with respect to Ft . If either one of the following conditions holds:
1. τ is bounded, i.e. ∃N < ∞ such that τ ≤ N ;
2. ∃c > 0 such that E[|Mt |] ≤ c, ∀t > 0,
then E[Mτ ] = E[M0 ].
Proof: We shall sketch the proof for the case 1 only. Taking the expectation in relation
Mτ = Mτ ∧t + (Mτ − Mt )1{τ >t} ,
see Exercise 3.2.3, yields
E[Mτ ] = E[Mτ ∧t ] + E[Mτ 1{τ >t} ] − E[Mt 1{τ >t} ].
Since Mτ ∧t is a martingale, see Exercise 3.2.4 (b), then E[Mτ ∧t ] = E[M0 ]. The previous relation
becomes
E[Mτ ] = E[M0 ] + E[Mτ 1{τ >t} ] − E[Mt 1{τ >t} ], ∀t > 0.
Taking the limit yields
E[Mτ ] = E[M0 ] + lim E[Mτ 1{τ >t} ] − lim E[Mt 1{τ >t} ]. (3.2.1)
t→∞ t→∞

We shall show that both limits are equal to zero.

Since |Mτ 1{τ >t} | ≤ |Mτ |, ∀t > 0, and Mτ is integrable, see Exercise 3.2.4 (a), by the
dominated convergence theorem we have
Z Z
lim E[Mτ 1{τ >t} ] = lim Mτ 1{τ >t} dP = lim Mτ 1{τ >t} dP = 0.
t→∞ t→∞ Ω Ω t→∞
For the second limit
Z
lim E[Mt 1{τ >t} ] = lim Mt 1{τ >t} dP = 0,
t→∞ t→∞ Ω

since for t > N the integrand vanishes. Hence relation (3.2.1) yields E[Mτ ] = E[M0 ].
It is worth noting that the previous theorem is a special case of the more general Optional
Stopping Theorem of Doob:
Theorem 3.2.2 Let Mt be a right continuous martingale and σ, τ be two bounded stopping
times, with σ ≤ τ . Then Mσ , Mτ are integrable and
E[Mτ |Fσ ] = Mσ a.s.
Exercise 3.2.3 Show that
Mτ = Mτ ∧t + (Mτ − Mt )1{τ >t} ,
where 
1, τ (ω) > t;
1{τ >t} (ω) =
0, τ (ω) ≤ t
is the indicator function of the set {τ > t}.
Exercise 3.2.4 Let Mt be a right continuous martingale and τ be a stopping time. Show that
(a) Mτ is integrable;
(b) Mτ ∧t is a martingale.
Exercise 3.2.5 Show that if let σ = 0 in Theorem 3.2.2 yields Theorem 3.2.1.
56

2.5

Wt
2.0
a

1.5

50 100 150 200 250 300 350

Ta

Figure 3.1: The first hitting time Ta given by WTa = a.

3.3 The First Passage of Time

The first passage of time is a particular type of hitting time, which is useful in finance when
studying barrier options and lookback options. For instance, knock-in options enter into exis-
tence when the stock price hits for the first time a certain barrier before option maturity. A
lookback option is priced using the maximum value of the stock until the present time. The
stock price is not a Brownian motion, but it depends on one. Hence the need for studying the
hitting time for the Brownian motion.
The first result deals with the first hitting time for a Brownian motion to reach the barrier
a ∈ R, see Fig.3.1.

Lemma 3.3.1 Let Ta be the first time the Brownian motion Wt hits a. Then the distribution
function of Ta is given by
Z ∞
2 2
P (Ta ≤ t) = √ √ e−y /2
dy.
2π |a|/ t

Proof: If A and B are two events, then

P (A) = P (A ∩ B) + P (A ∩ B)
= P (A|B)P (B) + P (A|B)P (B). (3.3.2)

Let a > 0. Using formula (3.3.2) for A = {ω; Wt (ω) ≥ a} and B = {ω; Ta(ω) ≤ t} yields

P (Wt ≥ a) = P (Wt ≥ a|Ta ≤ t)P (Ta ≤ t)

+P (Wt ≥ a|Ta > t)P (Ta > t) (3.3.3)

If Ta > t, the Brownian motion did not reach the barrier a yet, so we must have Wt < a.
Therefore
P (Wt ≥ a|Ta > t) = 0.
If Ta ≤ t, then WTa = a. Since the Brownian motion is a Markov process, it starts fresh at Ta .
Due to symmetry of the density function of a normal variable, Wt has equal chances to go up
or go down after the time interval t − Ta . It follows that
1
P (Wt ≥ a|Ta ≤ t) = .
2
 57

Substituting into (3.3.3) yields

P (Ta ≤ t) = 2P (Wt ≥ a)
Z ∞ Z ∞
2 2 2 2
= √ e−x /(2t) dx = √ e−y /2 dy.
2πt a 2π a/√t

If a < 0, symmetry implies that the distribution of Ta is the same as that of T−a , so we get
Z ∞
2 −y 2 /2
P (Ta ≤ t) = P (T−a ≤ t) = √ √ e dy.
2π −a/ t

Remark 3.3.2 The previous proof is based on a more general principle called the Reflection
Principle: If τ is a stopping time for the Brownian motion Wt , then the Brownian motion
reflected at τ is also a Brownian motion.

Theorem 3.3.3 Let a ∈ R be fixed. Then the Brownian motion hits a (in a finite amount of
time) with probability 1.

Proof: The probability that Wt hits a (in a finite amount of time) is

Z ∞
2 2
P (Ta < ∞) = lim P (Ta ≤ t) = lim √ e−y /2 dy
t→∞ t→∞ 2π |a|/√t
Z ∞
2 2
= √ e−y /2 dy = 1,
2π 0
where we used the well known integral
Z ∞ Z
2 1 ∞ −y2 /2 1√
e−y /2 dy = e dy = 2π.
0 2 −∞ 2

The previous result stated that the Brownian motion hits the barrier a almost surely. The
next result shows that the expected time to hit the barrier is infinite.

Proposition 3.3.4 The random variable Ta has a Pearson 5 distribution given by

|a| a2 3
p(t) = √ e− 2t t− 2 , t > 0.
2π

a2
It has the mean E[Ta ] = ∞ and the mode .
3
Proof: Differentiating in the formula of distribution function1
Z ∞
2 2
FTa (t) = P (Ta ≤ t) = √ e−y /2 dy
2π a/√t
Z ψ(t)
1 One d
may use Leibnitz’s formula f (u) du = f (ψ(t))ψ′ (t) − f (ϕ(t))ϕ′ (t).
dt ϕ(t)
58

0.4

0.3

0.2

0.1

2
a3
1 2 3 4

Figure 3.2: The distribution of the first hitting time Ta .

yields the following probability density function

dFTa (t) a a2 3
p(t) = = √ e− 2t t− 2 , t > 0.
dt 2π
This is a Pearson 5 distribution with parameters α = 1/2 and β = a2 /2. The expectation is
Z ∞ Z ∞
a 1 a2
E[Ta ] = tp(t) dt = √ √ e− 2t dt.
0 2π 0 t
a2 a2
Using the inequality e− 2t > 1 − , t > 0, we have the estimation
2t
Z ∞ Z ∞
a 1 a3 1
E[Ta ] > √ √ dt − √ dt = ∞,
2π 0 t 2 2π 0 t3/2
R∞ 1
R∞ 1
since 0
√
t
dt is divergent and 0 t3/2
dt is convergent.
The mode of Ta is given by

β a2 a2
= 1 = .
α+1 2( 2 + 1) 3

Remark 3.3.5 The distribution has a peak at a2 /3. Then if we need to pick a small time
interval [t − dt, t + dt] in which the probability that the Brownian motion hits the barrier a is
maximum, we need to choose t = a2 /3.

Remark 3.3.6 The expected waiting time for Wt to reach the barrier a is infinite. However,
the expected waiting time for the Brownian motion Wt to hit either a or −a is finite, see Exercise
3.3.9.

Corollary 3.3.7 A Brownian motion process returns to the origin in a finite amount time with
probability 1.

Proof: Choose a = 0 and apply Theorem 3.3.3.

 59

Exercise 3.3.8 Try to apply the proof of Lemma 3.3.1 for the following stochastic processes
(a) Xt = µt + σWt , with µ, σ > 0 constants;
Z t
(b) Xt = Ws ds.
0
Where is the difficulty?
Exercise 3.3.9 Let a > 0 and consider the hitting time
τa = inf{t > 0; Wt = a or Wt = −a} = inf{t > 0; |Wt | = a}.
Prove that E[τa ] = a2 .
Exercise 3.3.10 (a) Show that the distribution function of the process
Xt = max Ws
s∈[0,t]

is given by √
Z a/ t
2 2
P (Xt ≤ a) = √ e−y /2
dy.
2π 0
√
(b) Show that E[Xt ] = 0 and V ar(Xt ) = 2 t.
Exercise 3.3.11 (a) Find the distribution of Yt = |Wt |, t ≥ 0;
r
  πT
(b) Show that E max |Wt | = .
0≤t≤T 2
The fact that a Brownian motion returns to the origin or hits a barrier almost surely is
a property characteristic to the first dimension only. The next result states that in larger
dimensions this is no longer possible.


Theorem 3.3.12 Let (a, b) ∈ R2 . The 2-dimensional Brownian motion W (t) = W1 (t), W2 (t)
(with W1 (t) and W2 (t) independent) hits the point (a, b) with probability zero. The same result
is valid for any n-dimensional Brownian motion, with n ≥ 2.
However, if the point (a, b) is replaced by the disk Dǫ (x0 ) = {x ∈ R2 ; |x − x0 | ≤ ǫ}, there is
a difference in the behavior of the Brownian motion from n = 2 to n > 2.


Theorem 3.3.13 The 2-dimensional Brownian motion W (t) = W1 (t), W2 (t) hits the disk
Dǫ (x0 ) with probability one.


Theorem 3.3.14 Let n > 2. The n-dimensional Brownian motion W (t) = W1 (t), · · · Wn (t)
hits the ball Dǫ (x0 ) with probability
 |x | 2−n
0
P = < 1.
ǫ
The previous results can be stated by saying that that Brownian motion is transient in Rn ,
for n > 2. If n = 2 the previous probability equals 1. We shall come back with proofs to the
aforementioned results in a later chapter.
Remark 3.3.15 If life spreads according to a Brownian motion, the aforementioned results
explain why life is more extensive on earth rather than in space. The probability for a form of
life to reach a planet of radius R situated at distance d is R
d . Since d is large the probability is
very small, unlike in the plane, where the probability is always 1.
Exercise 3.3.16 Is the one-dimensional Brownian motion transient or recurrent in R?
60

0.4

0.2

t1 t2
50 100 150 200 250 300 350

-0.2
a
-0.4

Wt
-0.6

Figure 3.3: The event A(a; t1 , t2 ) in the Law of Arc-sine.

3.4 The Arc-sine Laws

In this section we present a few results which provide certain probabilities related with the
behavior of a Brownian motion in terms the arc-sine of a quotient of two time instances. These
results are generally known under the name of law of arc-sines.
The following result will be used in the proof of the first law of Arc-sine.

Proposition 3.4.1 (a) If X : Ω → N is a discrete random variable, then for any subset A ⊂ Ω,
we have X
P (A) = P (A|X = x)P (X = x).
x∈N

(b) If X : Ω → R is a continuous random variable, then

Z Z
P (A) = P (A|X = x)dP = P (A|X = x)fX (x) dx.

Proof: (a) The sets X −1 (x) = {X = x} = {ω; X(ω) = x} form a partition of the sample space
Ω, i.e.:
S
(i) Ω = x X −1 (x);
(ii) X −1 (x) ∩ X −1 (y) = Ø for x 6= y.
[  [ 
Then A = A ∩ X −1 (x) = A ∩ {X = x} , and hence
x x

X  
P (A) = P A ∩ {X = x}
x
X P (A ∩ {X = x})
= P ({X = x})
x
P ({X = x})
X
= P (A|X = x)P (X = x).
x

(b) In the case when X is continuous, the sum is replaced by an integral and the probability
P ({X = x}) by fX (x)dx, where fX is the density function of X.
 61

The zero set of a Brownian motion Wt is defined by {t ≥ 0; Wt = 0}. Since Wt is continuous,

the zero set is closed with no isolated points almost surely. The next result deals with the
probability that the zero set does not intersect the interval (t1 , t2 ).
Theorem 3.4.2 (The law of Arc-sine) The probability that a Brownian motion Wt does not
have any zeros in the interval (t1 , t2 ) is equal to
r
2 t1
P (Wt 6= 0, t1 ≤ t ≤ t2 ) = arcsin .
π t2

Proof: Let A(a; t1 , t2 ) denote the event that the Brownian motion Wt takes on the value a
between t1 and t2 . In particular, A(0; t1 , t2 ) denotes the event that Wt has (at least) a zero
between t1 and t2 . Substituting A = A(0; t1 , t2 ) and X = Wt1 into the formula provided by
Proposition 3.4.1
Z
P (A) = P (A|X = x)fX (x) dx

yields
Z



P A(0; t1 , t2 ) = P A(0; t1 , t2 )|Wt1 = x fWt (x) dx (3.4.4)
1
Z ∞
1
− x2
= √ P A(0; t1 , t2 )|Wt1 = x e 2t1 dx
2πt1 −∞
Using the properties of Wt with respect to time translation and symmetry we have



P A(0; t1 , t2 )|Wt1 = x = P A(0; 0, t2 − t1 )|W0 = x


= P A(−x; 0, t2 − t1 )|W0 = 0


= P A(|x|; 0, t2 − t1 )|W0 = 0


= P A(|x|; 0, t2 − t1 )


= P T|x| ≤ t2 − t1 ,
the last identity stating that Wt hits |x| before t2 − t1 . Using Lemma 3.3.1 yields
Z ∞

2 − y2
P A(0; t1 , t2 )|Wt1 = x = p e 2(t2 −t1 ) dy.
2π(t2 − t1 ) |x|
Substituting into (3.4.4) we obtain
Z ∞  Z ∞  x2

1 2 − y2
P A(0; t1 , t2 ) = √ p e 2(t2 −t1 ) dy e− 2t1 dx
2πt1 −∞ 2π(t2 − t1 ) |x|
Z ∞Z ∞
1 − y2
−x
2
= p e 2(t2 −t1 ) 2t1 dydx.
π t1 (t2 − t1 ) 0 |x|

The above integral can be evaluated to get (see Exercise 3.4.3 )

r

2 t1
P A(0; t1 , t2 ) = 1 − arcsin .
π t2


Using P (Wt 6= 0, t1 ≤ t ≤ t2 ) = 1 − P A(0; t1 , t2 ) we obtain the desired result.
62

Exercise 3.4.3 Use polar coordinates to show

Z ∞Z ∞ r
1 2
x2
− 2(t y−t ) − 2t 2 t1
p e 2 1 1 dydx = 1 − arcsin .
π t1 (t2 − t1 ) 0 |x| π t2


Exercise 3.4.4 Find the probability that a 2-dimensional Brownian motion W (t) = W1 (t), W2 (t)
stays in the same quadrant for the time interval t ∈ (t1 , t2 ).

Exercise 3.4.5 Find the probability that a Brownian motion Wt does not take the value a in
the interval (t1 , t2 ).

Exercise 3.4.6 Let a 6= b. Find the probability that a Brownian motion Wt does not take any
of the values {a, b} in the interval (t1 , t2 ). Formulate and prove a generalization.

We provide below without proof a few similar results dealing with arc-sine probabilities. The
first result deals with the amount of time spent by a Brownian motion on the positive half-axis.
Rt
Theorem 3.4.7 (Arc-sine Law of Lévy) Let L+ +
t = 0 sgn Ws ds be the amount of time a
Brownian motion Wt is positive during the time interval [0, t]. Then
r
2 τ
P (L+t ≤ τ ) = arcsin .
π t
The next result deals with the Arc-sine law for the last exit time of a Brownian motion from
0.
Theorem 3.4.8 (Arc-sine Law of exit from 0) Let γt = sup{0 ≤ s ≤ t; Ws = 0}. Then
r
2 τ
P (γt ≤ τ ) = arcsin , 0 ≤ τ ≤ t.
π t
The Arc-sine law for the time the Brownian motion attains its maximum on the interval
[0, t] is given by the next result.
Theorem 3.4.9 (Arc-sine Law of maximum) Let Mt = max Ws and define
0≤s≤t

θt = sup{0 ≤ s ≤ t; Ws = Mt }.

Then r
2 s
P (θt ≤ s) = arcsin , 0 ≤ s ≤ t, t > 0.
π t

3.5 More on Hitting Times

In this section we shall deal with results regarding hitting times of Brownian motion with drift.
They will be useful in the sequel when pricing barrier options.

Theorem 3.5.1 Let Xt = µt + Wt denote a Brownian motion with nonzero drift rate µ, and
consider α, β > 0. Then

e2µβ − 1
P (Xt goes up to α before down to − β) = .
e2µβ − e−2µα
 63

Proof: Let T = inf{t > 0; Xt = α or Xt = −β} be the first exit time of Xt from the interval
(−β, α), which is a stopping time, see Exercise 3.1.4. The exponential process
c2
Mt = ecWt − 2 t
, t≥0

is a martingale, see Exercise 2.2.4(c). Then E[Mt ] = E[M0 ] = 1. By the Optional Stopping
Theorem (see Theorem 3.2.1), we get E[MT ] = 1. This can be written as
1 2 1 2
1 = E[ecWT − 2 c T
] = E[ecXT −(cµ+ 2 c )T
]. (3.5.5)

Choosing c = −2µ yields E[e−2µXT ] = 1. Since the random variable XT takes only the values
α and −β, if let pα = P (XT = α), the previous relation becomes

e−2µα pα + e2µβ (1 − pα ) = 1.

Solving for pα yields

e2µβ − 1
pα = . (3.5.6)
e2µβ − e−2µα
Noting that
pα = P (Xt goes up to α before down to − β)
leads to the desired answer.
It is worth noting how the previous formula changes in the case when the drift rate is zero,
i.e. when µ = 0, and Xt = Wt . The previous probability is computed by taking the limit µ → 0
and using L’Hospital’s rule

e2µβ − 1 2βe2µβ β
lim = lim = .
µ→0 e2µβ − e−2µα µ→0 2βe2µβ + 2αe−2µα α+β

Hence
β
P (Wt goes up to α before down to − β) = .
α+β

Taking the limit β → ∞ we recover the following result

P (Wt hits α) = 1.

If α = β we obtain
1
P (Wt goes up to α before down to − α) = ,
2
which shows that the Brownian motion is equally likely to go up or down an amount α in a
given time interval.
If Tα and Tβ denote the times when the process Xt reaches α and β, respectively, then the
aforementioned probabilities can be written using inequalities. For instance the first identity
becomes
e2µβ − 1
P (Tα ≤ T−β ) = 2µβ .
e − e−2µα
64

Exercise 3.5.2 Let Xt = µt + Wt denote a Brownian motion with nonzero drift rate µ, and
consider α > 0.
(a) If µ > 0 show that
P (Xt goes up to α) = 1.
(b) If µ < 0 show that
P (Xt goes up to α) = e2µα < 1.

Formula (a) can be written equivalently as

P (sup(Wt + µt) ≥ α) = 1, µ ≥ 0,
t≥0

while formula (b) becomes

P (sup(Wt + µt) ≥ α) = e2µα , µ < 0,

t≥0

or
P (sup(Wt − γt) ≥ α) = e−2γα , γ > 0,
t≥0

which is known as one of the Doob’s inequalities. This can be also described in terms of stopping
times as follows. Define the stopping time τα = inf{t > 0; Wt − γt ≥ α}. Using


P (τα < ∞) = P sup(Wt − γt) ≥ α
t≥0

yields the identities

P (τα < ∞) = e−2αγ , γ > 0,

P (τα < ∞) = 1, γ ≤ 0.

Exercise 3.5.3 Let Xt = µt + Wt denote a Brownian motion with nonzero drift rate µ, and
consider β > 0. Show that the probability that Xt never hits −β is given by

1 − e−2µβ , if µ > 0
0, if µ < 0.
Recall that T is the first time when the process Xt hits α or −β.

Exercise 3.5.4 (a) Show that

αe2µβ + βe−2µα − α − β
E[XT ] = .
e2µβ − e−2µα
(b) Find E[XT2 ];
(c) Compute V ar(XT ).

The next result deals with the time one has to wait (in expectation) for the process Xt =
µt + Wt to reach either α or −β.
Proposition 3.5.5 The expected value of T is
αe2µβ + βe−2µα − α − β
E[T ] = .
µ(e2µβ − e−2µα )
 65

Proof: Using that Wt is a martingale, with E[Wt ] = E[W0 ] = 0, applying the Optional
Stopping Theorem, Theorem 3.2.1, yields
0 = E[WT ] = E[XT − µT ] = E[XT ] − µE[T ].
Then by Exercise 3.5.4(a) we get
E[XT ] αe2µβ + be−2µα − α − β
E[T ] = = .
µ µ(e2µβ − e−2µα )

Exercise 3.5.6 Take the limit µ → 0 in the formula provided by Proposition 3.5.5 to find the
expected time for a Brownian motion to hit either α or −β.
Exercise 3.5.7 Find E[T 2 ] and V ar(T ).
Exercise 3.5.8 (Wald’s identities) Let T be a finite stopping time for the Brownian motion
Wt . Show that
(a) E[WT ] = 0;
(b) E[WT2 ] = E[T ].
The previous techniques can be also applied to right continuous martingales. Let a > 0 and
consider the hitting time of the Poisson process of the barrier a
τ = inf{t > 0; Nt ≥ a}.
Proposition 3.5.9 The expected waiting time for Nt to reach the barrier a is E[τ ] = λa .
Proof: Since Mt = Nt −λt is a right continuous martingale, by the Optional Stopping Theorem
E[Mτ ] = E[M0 ] = 0. Then E[Nτ − λτ ] = 0 and hence E[τ ] = λ1 E[Nτ ] = λa .

3.6 The Inverse Laplace Transform Method

In this section we shall use the Optional Stopping Theorem in conjunction with the inverse
Laplace transform to obtain the probability density for hitting times.
A. The case of standard Brownian motion Let x > 0. The first hitting time τ = Tx =
1 2
inf{t > 0; Wt = x} is a stopping time. Since Mt = ecWt − 2 c t , t ≥ 0, is a martingale, with
E[Mt ] = E[M0 ] = 1, by the Optional Stopping Theorem, see Theorem 3.2.1, we have
E[Mτ ] = 1.
1 2
This can be written equivalently as E[ecWτ e− 2 c τ ] = 1. Using Wτ = x, we get
1 2
E[e− 2 c τ ] = e−cx .
1 2
It is worth noting that c > 0. This is implied from the fact that e− 2 c τ
< 1 and τ, x > 0.
Substituting s = 12 c2 , the previous relation becomes
√
E[e−sτ ] = e− 2sx
. (3.6.7)
This relation has a couple of useful applications.
66

Proposition 3.6.1 The moments of the first hitting time are all infinite E[τ n ] = ∞, n ≥ 1.

Proof: The nth moment of τ can be obtained by differentiating and taking s = 0

dn
E[e−sτ ] = E[(−τ )n e−sτ ] = (−1)n E[τ n ].
dsn s=0 s=0

Using (3.6.7) yields

dn −√2sx
E[τ n ] = (−1)n e .
dsn
s=0

Since by induction we have

n−1
dn −√2sx √ X Mk xn−k
n
e = (−1)n e− 2sx ,
ds
k=0
2rk /2 s(n+k)/2

with Mk , rk positive integers, it easily follows that E[τ n ] = ∞.

For instance, in the case n = 1 we have
d −√2sx √ x
E[τ ] = − e = lim+ e− 2sx
√ = +∞.
ds s=0
s→0 2 2sx

Another application involves the inverse Laplace transform to get the probability density.
This way we can retrieve the result of Proposition 3.3.4.
Proposition 3.6.2 The probability density of the hitting time τ is given by
|x| − x2
p(t) = √ e 2t , t > 0. (3.6.8)
2πt3
Proof: Let x > 0. The expectation
Z ∞
E[e−sτ ] = e−sτ p(τ ) dτ = L{p(τ )}(s)
0

is the Laplace transform of p(τ ). Applying the inverse Laplace transform yields
√
p(τ ) = L−1 {E[e−sτ ]}(τ ) = L−1 {e− 2sx
}(τ )
x x2
= √ e− 2τ , τ > 0.
2πτ 3

In the case x < 0 we obtain

−x − x2
p(τ ) = √ e 2τ , τ > 0,
2πτ 3
which leads to (3.6.8).
Another application of formula (3.6.7) is the following inequality.
Exercise 3.6.3 Find the following inverse Laplace transforms:
√
(a) L−1 {e−c s }(τ ), c > 0 constant;
√
(b) L−1 {e− 2sx
}(τ ), x > 0.
 67

Proposition 3.6.4 (Chernoff bound) Let τ denote the first hitting time when the Brownian
motion Wt hits the barrier x, x > 0. Then
x2
P (τ ≤ λ) ≤ e− 2λ , ∀λ > 0.
Proof: Let s = −t in the part 2 of Theorem 1.12.10 and use (3.6.7) to get
E[etX ] E[e−sX ] √
P (τ ≤ λ) ≤ λt
= −λs
= eλs−x 2s , ∀s > 0.
e e
√
Then P (τ ≤ λ) ≤ emins>0 f (s) , where f (s) = λs − x 2s. Since f ′ (s) = λ − √x , then f (s)
2s
2
x
reaches its minimum at the critical point s0 = 2λ2 . The minimum value is

x2
min f (s) = f (s0 ) = − .
s>0 2λ
Substituting in the previous inequality leads to the required result.

B. The case of Brownian motion with drift

Consider the Brownian motion with drift Xt = µt + σWt , with µ, σ > 0. Let
τ = inf{t > 0; Xt = x}
denote the first hitting time of the barrier x, with x > 0. We shall compute the distribution of
the random variable τ and its first two moments.
1 2
Applying the Optional Stopping Theorem (Theorem 3.2.1) to the martingale Mt = ecWt − 2 c t
yields
E[Mτ ] = E[M0 ] = 1.
1
Using that Wτ = σ (Xτ − µτ ) and Xτ = x, the previous relation becomes
cµ 1 2 c
E[e−( σ + 2 c )τ
] = e− σ x . (3.6.9)
cµ 1 2
Substituting s = + c and completing to a square yields
σ 2
µ2  µ 2
2s + 2 = c + .
σ σ
Solving for c we get the solutions
r r
µ µ2 µ µ2
c=− + 2s + 2 , c=− − 2s + .
σ σ σ σ2
Assume c < 0. Then substituting the second solution into (3.6.9) yields
1
√ 2 2
E[e−sτ ] = e σ2 (µ+ 2sσ +µ )x .
1
√ 2
This relation is contradictory since e−sτ < 1 while e σ2 (µ+ 2sσ+µ )x > 1, where we used that
s, x, τ > 0. Hence it follows that c > 0. Substituting the first solution into (3.6.9) leads to
1
√ 2 2
E[e−sτ ] = e σ2 (µ− 2sσ +µ )x .
We arrived at the following result:
68

Proposition 3.6.5 Let τ be the time the process Xt = µt + σWt hits x for the first time. Then
for s > 0, x > 0 we have √ 2 2
1
E[e−sτ ] = e σ2 (µ− 2sσ +µ )x . (3.6.10)
Proposition 3.6.6 Let τ be the time the process Xt = µt + σWt hits x, with x > 0 and µ > 0.
(a) Then the density function of τ is given by
x (x−µτ )2
p(τ ) = √ e− 2τ σ2 , τ > 0. (3.6.11)
σ 2πτ 3/2
(b) The mean and variance of τ are
x xσ 2
E[τ ] = , V ar(τ ) = .
µ µ3
Proof: (a) Let p(τ ) be the density function of τ . Since
Z ∞
E[e−sτ ] = e−sτ p(τ ) dτ = L{p(τ )}(s)
0

is the Laplace transform of p(τ ), applying the inverse Laplace transform yields
1
√ 2 2
p(τ ) = L−1 {E[e−sτ ]} = L−1 {e σ2 (µ− 2sσ +µ )x }
x (x−µτ )2
= √ e− 2τ σ2 , τ > 0.
σ 2πτ 3/2
(b) The moments are obtained by differentiating the moment generating function and taking
the value at s = 0

d d 1 √ 2 2
E[τ ] = − E[e−sτ ] = − e σ2 (µ− 2sσ +µ )x
ds s=0 ds s=0
x 1
√ 2 2
= p e σ2 (µ− 2sσ +µ )x
µ2 + 2sµ s=0

x
= .
µ

d2 d2 12 (µ−√2sσ2 +µ2 )x
E[τ 2 ] = (−1)2 E[e −sτ
] = eσ
ds2 s=0 ds2 s=0

x(σ 2 + x µ2 + 2sσ 2 ) 12 (µ−√2sσ2 +µ2 )x

p
= eσ
(µ2 + 2sσ 2 )3/2 s=0

xσ 2 x2
= + 2.
µ3 µ
Hence
xσ 2
V ar(τ ) = E[τ 2 ] − E[τ ]2 =.
µ3
It is worth noting that we can arrive at the formula E[τ ] = µx in the following heuristic way.
Taking the expectation in the equation µτ + σWτ = x yields µE[τ ] = x, where we used that
E[Wτ ] = 0 for any finite stopping time τ (see Exercise 3.5.8 (a)). Solving for E[τ ] yields the
aforementioned formula.
 69

Exercise 3.6.7 Find the mode of distribution (3.6.11).

Exercise 3.6.8 Let Xt = 2t + 3Wt and Yt = 6t + Wt .

(a) Show that the expected times for Xt and Yt to reach any barrier x > 0 are the same.
(b) If Xt and Yt model the prices of two stocks, which one would you like to own?

Exercise 3.6.9 Does 4t + 2Wt hit 9 faster (in expectation) than 5t + 3Wt hits 14?

Exercise 3.6.10 Let τ be the first time the Brownian motion with drift Xt = µt + Wt hits x,
where µ, x > 0. Prove the inequality
x2 +λ2 µ2
P (τ ≤ λ) ≤ e− 2λ +µx
, ∀λ > 0.

C. The double barrier case

In the following we shall consider the case of double barrier. Consider the Brownian motion
with drift Xt = µt + Wt , µ > 0. Let α, β > 0 and define the stopping time

T = inf{t > 0; Xt = α or Xt = −β}.

Relation (3.5.5) states

1 2
E[ecXT e−(cµ+ 2 c )T
] = 1.

Since the random variables T and XT are independent (why?), we have

1 2
E[ecXT ]E[e−(cµ+ 2 c )T
] = 1.

Using E[ecXT ] = ecα pα + e−cβ (1 − pα ), with pα given by (3.5.6), then

1 2 1
E[e−(cµ+ 2 c )T
]= ·
ecα pα + e−cβ (1 − pα )

If substitute s = cµ + 12 c2 , then

1
E[e−sT ] = √ √ · (3.6.12)
(−µ+ 2s+µ2 )α −(−µ+ 2s+µ2 )β
e pα + e (1 − pα )

The probability density of the stopping time T is obtained by taking the inverse Laplace trans-
form of the right side expression
n 1 o
p(T ) = L−1 √ √ (τ ),
2s+µ2 )α 2s+µ2 )β
e(−µ+ pα + e−(−µ+ (1 − pα )

an expression which is not feasible for having closed form solution. However, expression (3.6.12)
would be useful for computing the price for double barrier derivatives.

Exercise 3.6.11 Use formula (3.6.12) to find the expectation E[T ].

70

Exercise 3.6.12 Denote by Mt = Nt − λt the compensated Poisson process and let c > 0 be a
constant.
(a) Show that
c
Xt = ecMt −λt(e −c−1)

is an Ft -martingale, with Ft = σ(Nu ; u ≤ t).

(b) Let a > 0 and T = inf{t > 0; Mt > a} be the first hitting time of the level a. Use the
Optional Stopping Theorem to show that

E[e−λsT ] = e−ϕ(s)a , s > 0,

where ϕ : [0, ∞) → [0, ∞) is the inverse function of f (x) = ex − x − 1.

(c) Show that E[T ] = ∞.
(d) Can you use the inverse Laplace transform to find the probability density function of T ?

3.7 Limits of Stochastic Processes

Let (Xt )t≥0 be a stochastic process. We can make sense of the limit expression X = lim Xt ,
t→∞
in a similar way as we did in section 1.13 for sequences of random variables. We shall rewrite
the definitions for the continuous case.

Almost Certain Limit

The process Xt converges almost certainly to X, if for all states of the world ω, except a set of
probability zero, we have
lim Xt (ω) = X(ω).
t→∞

We shall write ac-lim Xt = X. It is also sometimes called strong convergence.

t→∞

Mean Square Limit

We say that the process Xt converges to X in the mean square if

lim E[(Xt − X)2 ] = 0.

t→∞

In this case we write ms-lim Xt = X.

t→∞

Limit in Probability or Stochastic Limit

The stochastic process Xt converges in stochastic limit to X if


lim P ω; |Xt (ω) − X(ω)| > ǫ = 0.
t→∞

This limit is abbreviated by st-lim Xt = X.

t→∞
It is worth noting that, like in the case of sequences of random variables, both almost certain
convergence and convergence in mean square imply the stochastic convergence, which implies
the limit in distribution.
 71

Limit in Distribution
We say that Xt converges in distribution to X if for any continuous bounded function ϕ(x) we
have
lim ϕ(Xt ) = ϕ(X).
t→∞
It is worth noting that the stochastic convergence implies the convergence in distribution.

3.8 Convergence Theorems

The following property is a reformulation of Exercise 1.13.1 in the continuous setup.
Proposition 3.8.1 Consider a stochastic process Xt such that E[Xt ] → k, a constant, and
V ar(Xt ) → 0 as t → ∞. Then ms-lim Xt = k.
t→∞

Next we shall provide a few applications.

Application 3.8.2 If α > 1/2, then
Wt
ms-lim = 0.
t→∞ tα
Wt E[Wt ] 1 t 1
Proof: Let Xt = . Then E[Xt ] = = 0, and V ar[Xt ] = 2α V ar[Wt ] = 2α = 2α−1 ,
tα tα t t t
1 Wt
for any t > 0. Since 2α−1 → 0 as t → ∞, applying Proposition 3.8.1 yields ms-lim α = 0.
t t→∞ t

Wt
Corollary 3.8.3 ms-lim = 0.
t→∞t
Rt
Application 3.8.4 Let Zt = 0 Ws ds. If β > 3/2, then
Zt
ms-lim = 0.
t→∞ tβ
Zt E[Zt ] 1 t3 1
Proof: Let Xt = β
. Then E[X t ] = β
= 0, and V ar[X t ] = 2β
V ar[Z t ] = 2β
= 2β−3 ,
t t t 3t 3t
1
for any t > 0. Since 2β−3 → 0 as t → ∞, applying Proposition 3.8.1 leads to the desired
3t
result.

Application 3.8.5 For any p > 0, c ≥ 1 we have

eWt −ct
ms-lim = 0.
t→∞ tp
eWt −ct eWt
Proof: Consider the process Xt = p
= p ct . Since
t t e
E[eWt ] et/2 1 1
E[Xt ] = p ct
= p ct
= (c− 1 )t p → 0, as t → ∞
t e t e e 2 t
V ar[eWt ] e2t − et 1  1 1 
V ar[Xt ] = = = − → 0,
t2p e2ct t2p e2ct t2p e2t(c−1) et(2c−1)
as t → ∞, Proposition 3.8.1 leads to the desired result.
72

Application 3.8.6 Show that

max Ws
0≤s≤t
ms-lim = 0.
t→∞ t
max Ws
0≤s≤t
Proof: Let Xt = . Since by Exercise 3.3.10
t
E[ max Ws ] = 0
0≤s≤t
√
V ar( max Ws ) = 2 t,
0≤s≤t

then

E[Xt ] = 0
√
2 t
V ar[Xt ] = → 0, t → ∞.
t2
Apply Proposition 3.8.1 to get the desired result.

Remark 3.8.7 One of the strongest result regarding limits of Brownian motions is called the
law of iterated logarithms and was first proved by Lamperti:

Wt
lim sup p = 1,
t→∞ 2t ln(ln t)

almost certainly.

Exercise 3.8.8 Use the law of iterated logarithms to show

Wt
lim p = 1.
t→0+ 2t ln ln(1/t)

Wt
Application 3.8.9 We shall show that ac-lim = 0.
t→∞ t
Wt
From the law of iterated logarithms p < 1 for t large. Then
2t ln(ln t)
p p
Wt Wt 2 ln(ln t) 2 ln(ln t)
= p √ < √ .
t 2t ln(ln t) t t
p
2 ln(ln t) Wt
Let ǫt = √ . Then < ǫt for t large. As an application of the l’Hospital rule, it is
t t
not hard to see that ǫt satisfies the following limits

ǫ → 0, t→∞
√ t
ǫt t → ∞, t → ∞.
 73

Wt
In order to show that ac-lim = 0, it suffices to prove
t→∞ t
 Wt (ω) 
P ω; < ǫt → 1, t → ∞. (3.8.13)
t
We have
 Wt (ω)    Z tǫt
1 − u2
P ω; < ǫt = P ω; −tǫt < Wt (ω) < tǫt = √ e 2t du
t −tǫt 2πt
Z √ Z
ǫt t ∞
1 v2 1 v2
= √
√ e− 2 dv → √ e− 2 dv = 1, t → ∞,
−ǫt t 2π −∞ 2π
√
where we used ǫt t → ∞, as t → ∞, which proves (3.8.13).
Proposition 3.8.10 Let Xt be a stochastic process. Then

ms-lim Xt = 0 ⇐⇒ ms-lim Xt2 = 0.

t→∞ t→∞

Proof: Left as an exercise.

Exercise 3.8.11 Let Xt be a stochastic process. Show that

ms-lim Xt = 0 ⇐⇒ ms-lim |Xt | = 0.

t→∞ t→∞

Another convergence result can be obtained if we consider the continuous analog of Example
1.13.6:
Proposition 3.8.12 Let Xt be a stochastic process such that there is a p > 0 such that
E[|Xt |p ] → 0 as t → ∞. Then st-lim Xt = 0.
t→∞

Application 3.8.13 We shall show that for any α > 1/2

Wt
st-lim = 0.
t→∞ tα
Wt
Proof: Consider the process Xt = . By Proposition 3.8.10 it suffices to show ms-lim Xt2 =
tα t→∞
0. Since the mean square convergence implies the stochastic convergence, we get st-lim Xt2 = 0.
t→∞
Since h W 2 i E[W 2 ] t 1
E[|Xt |2 ] = E[Xt2 ] = E 2αt = 2α
t
= 2α = 2α−1 → 0, t → ∞,
t t t t
then Proposition 3.8.12 yields st-lim Xt = 0.
t→∞

The following result can be regarded as the L’Hospital’s rule for sequences:
Lemma 3.8.14 (Cesaró-Stoltz) Let xn and yn be two sequences of real numbers, n ≥ 1. If
xn+1 − xn
the limit lim exists and is equal to L, then the following limit exists
n→∞ yn+1 − yn

xn
lim = L.
n→∞ yn
74

Proof: (Sketch) Assume there are differentiable functions f and g such that f (n) = xn and
g(n) = yn . (How do you construct these functions?) From Cauchy’s theorem2 there is a
cn ∈ (n, n + 1) such that

xn+1 − xn f (n + 1) − f (n) f ′ (cn )

L = lim = lim = lim ′ .
n→∞ yn+1 − yn n→∞ g(n + 1) − g(n) n→∞ g (cn )

Since cn → ∞ as n → ∞, we can also write the aforementioned limit as

f ′ (t)
lim = L.
t→∞ g ′ (t)

(Here one may argue against this, but we recall the freedom of choice for the functions f and
g such that cn can be any number between n and n + 1). By l’Hospital’s rule we get

f (t)
lim = L.
t→∞ g(t)
xn
Making t = n yields lim = L.
yn
t→∞

The next application states that if a sequence is convergent, then the arithmetic average of
its terms is also convergent, and the sequences have the same limit.

Example 3.8.1 Let an be a convergent sequence with lim an = L. Let

n→∞

a1 + a2 + · · · + an
An =
n
be the arithmetic average of the first n terms. Then An is convergent and

lim An = L.
n→∞

Proof: This is an application of Cesaró-Stoltz lemma. Consider the sequences xn = a1 + a2 +

· · · + an and yn = n. Since

xn+1 − xn (a1 + · · · + an+1 ) − (a1 + · · · + an ) an+1

= = ,
yn+1 − yn (n + 1) − n 1

then
xn+1 − xn
lim = lim an+1 = L.
n→∞ yn+1 − yn n→∞

Applying the Cesaró-Stoltz lemma yields

xn
lim An = lim = L.
n→∞ n→∞ yn

2 This
says that if f and g are differentiable on (a, b) and continuous on [a, b], then there is a c ∈ (a, b) such
f (a) − f (b) f ′ (c)
that = ′ .
g(a) − g(b) g (c)
 75

Exercise 3.8.15 Let bn be a convergent sequence with lim bn = L. Let

n→∞

Gn = (b1 · b2 · · · · · bn )1/n
be the geometric average of the first n terms. Show that Gn is convergent and
lim Gn = L.
n→∞

The following result extends the Cesaró-Stoltz lemma to sequences of random variables.
Proposition 3.8.16 Let Xn be a sequence of random variables on the probability space (Ω, F , P ),
such that
Xn+1 − Xn
ac- lim = L.
n→∞ Yn+1 − Yn
Then
Xn
ac- lim = L.
n→∞ Yn

Proof: Consider the sets

Xn+1 (ω) − Xn (ω)
A = {ω ∈ Ω; lim = L}
Yn+1 (ω) − Yn (ω)
n→∞

Xn (ω)
B = {ω ∈ Ω; lim = L}.
n→∞ Yn (ω)

Since for any given state of the world ω, the sequences xn = Xn (ω) and yn = Yn (ω) are
numerical sequences, Lemma 3.8.14 yields the inclusion A ⊂ B. This implies P (A) ≤ P (B).
Since P (A) = 1, it follows that P (B) = 1, which leads to the desired conclusion.

Example 3.8.17 Let Sn denote the price of a stock on day n, and assume that
ac-lim Sn = L.
n→∞
Then
S1 + · · · + Sn
ac-lim = L and ac-lim (S1 · · · · · Sn )1/n = L.
n→∞ n n→∞

This says that if almost all future simulations of the stock price approach the steady state
limit L, the arithmetic and geometric averages converge to the same limit. The statement is a
consequence of Proposition 3.8.16 and follows a similar proof as Example 3.8.1. Asian options
have payoffs depending on these type of averages, as we shall see in Part II.

3.8.1 The Martingale Convergence Theorem

We state now, without proof, a result which is a powerful way of proving the almost certain
convergence. We start with the discrete version:
Theorem 3.8.18 Let Xn be a martingale with bounded means
∃M > 0 such that E[|Xn |] ≤ M, ∀n ≥ 1. (3.8.14)
Then there is L < ∞ such that ac- lim Xn = L, i.e.
n→∞


P ω; lim Xn (ω) = L = 1.
n→∞
76

Since E[|Xn |]2 ≤ E[Xn2 ], the condition (3.8.14) can be replaced by its stronger version

∃M > 0 such that E[Xn2 ] ≤ M, ∀n ≥ 1.

The following result deals with the continuous

 version
 of the Martingale Convergence The-
orem. Denote the infinite knowledge by F∞ = σ ∪t Ft .

Theorem 3.8.19 Let Xt be an Ft -martingale such that

∃M > 0 such that E[|Xt |] < M, ∀t > 0.

Then there is an F∞ -measurable random variable X∞ such that Xt → X∞ a.c. as t → ∞.

The next exrcise deals with a process that is a.c-convergent but is not ms-convergent.

Exercise 3.8.20 It is known that Xt = eWt −t/2 is a martingale. Since

E[|Xt |] = E[eWt −t/2 ] = e−t/2 E[eWt ] = e−t/2 et/2 = 1,

by the Martingale Convergence Theorem there is a number L such that Xt → L a.c. as t → ∞.

(a) What is the limit L? How did you make your guess?
(b) Show that
 2
E[|Xt − 1|2 ] = V ar(Xt ) + E(Xt ) − 1 .

(c) Show that Xt does not converge in the mean square to 1.

(d) Prove that the sequence Xt is a.c-convergent but it is not ms-convergent.

3.8.2 The Squeeze Theorem

The following result is the analog of the Squeeze Theorem from usual Calculus.
Theorem 3.8.21 Let Xn , Yn , Zn be sequences of random variables on the probability space
(Ω, F , P ) such that
Xn ≤ Yn ≤ Zn a.s. ∀n ≥ 1.
If Xn and Zn converge to L as n → ∞ almost certainly (or in mean square, or stochastic or
in distribution), then Yn converges to L in a similar mode.
Proof: For any state of the world ω ∈ Ω consider the sequences xn = Xn (ω), yn = Yn (ω) and
zn = Zn (ω) and apply the usual Squeeze Theorem to them.

Remark 3.8.22 The previous theorem remains valid if n is replaced by a continuous positive
parameter t.
Wt sin(Wt )
Example 3.8.2 Show that ac- lim = 0.
t→∞ t
Wt sin(Wt ) Wt
Proof: Consider the sequences Xt = 0, Yt = and Zt = . From Application 3.8.9
t t
we have ac-lim Zt = 0. Applying the Squeeze Theorem we obtain the desired result.
t→∞
 77

Exercise 3.8.23 Use the Squeeze Theorem to find the following limits:
sin(Wt )
(a) ac-lim ;
t→∞ t
(b) ac-lim t cos Wt ;
t→0
(c) ac- lim et (sin Wt )2 .
t→−∞

3.9 Quadratic Variation

For some stochastic processes the sum of squares of consecutive increments tends in mean square
to a finite number, as the norm of the partition decreases to zero. We shall encounter in the
following a few important examples that will be useful when dealing with stochastic integrals.

3.9.1 The Quadratic Variation of Wt

The next result states that the quadratic variation of the Brownian motion Wt on the interval
[0, T ] is T . More precisely, we have:
Proposition 3.9.1 Let T > 0 and consider the equidistant partition 0 = t0 < t1 < · · · tn−1 <
tn = T . Then
n−1
X
ms- lim (Wti+1 − Wti )2 = T. (3.9.15)
n→∞
i=0

Proof: Consider the random variable

n−1
X
Xn = (Wti+1 − Wti )2 .
i=0

Since the increments of a Brownian motion are independent, Proposition 4.0.7 yields
n−1
X n−1
X
E[Xn ] = E[(Wti+1 − Wti )2 ] = (ti+1 − ti )
i=0 i=0
= tn − t0 = T ;

n−1
X n−1
X
V ar(Xn ) = V ar[(Wti+1 − Wti )2 ] = 2(ti+1 − ti )2
i=0 i=0
2T 2
 T 2
= n·2 = ,
n n
where we used that the partition is equidistant. Since Xn satisfies the conditions
E[Xn ] = T, ∀n ≥ 1;
V ar[Xn ] → 0, n → ∞,
by Proposition 3.8.1 we obtain ms-lim Xn = T , or
n→∞

n−1
X
ms- lim (Wti+1 − Wti )2 = T. (3.9.16)
n→∞
i=0
78

Exercise 3.9.2 Prove that the quadratic variation of the Brownian motion Wt on [a, b] is equal
to b − a.

The Fundamental Relation dWt2 = dt

The relation discussed in this section can be regarded as the fundamental relation of Stochastic
Calculus. We shall start by recalling relation (3.9.16)
n−1
X
ms- lim (Wti+1 − Wti )2 = T. (3.9.17)
n→∞
i=0
The right side can be regarded as a regular Riemann integral
Z T
T = dt,
0

while the left side can be regarded as a stochastic integral with respect to dWt2
Z T n−1
X
(dWt )2 = ms- lim (Wti+1 − Wti )2 .
0 n→∞
i=0

Substituting into (3.9.17) yields

Z T Z T
(dWt )2 = dt, ∀T > 0.
0 0

The differential form of this integral equation is

dWt2 = dt.

In fact, this expression also holds in the mean square sense, as it can be inferred from the next
exercise.
Exercise 3.9.3 Show that
(a) E[dWt2 − dt] = 0;
(b) V ar(dWt2 − dt) = o(dt);
(c) ms- lim (dWt2 − dt) = 0.
dt→0

Roughly speaking, the process dWt2 , which is the square of infinitesimal increments of a
Brownian motion, is totally predictable. This relation plays a central role in Stochastic Calculus
and will be useful when dealing with Ito’s lemma.
The following exercise states that dtdWt = 0, which is another important stochastic relation
useful in Ito’s lemma.
Exercise 3.9.4 Consider the equidistant partition 0 = t0 < t1 < · · · tn−1 < tn = T . Then
n−1
X
ms- lim (Wti+1 − Wti )(ti+1 − ti ) = 0. (3.9.18)
n→∞
i=0
 79

3.9.2 The Quadratic Variation of Nt − λt

The following result deals with the quadratic variation of the compensated Poisson process
Mt = Nt − λt.

Proposition 3.9.5 Let a < b and consider the partition a = t0 < t1 < · · · < tn−1 < tn = b.
Then
n−1
X
ms– lim (Mtk+1 − Mtk )2 = Nb − Na , (3.9.19)
k∆n k→0
k=0

where k∆n k := sup (tk+1 − tk ).

0≤k≤n−1

Proof: For the sake of simplicity we shall use the following notations:

∆tk = tk+1 − tk , ∆Mk = Mtk+1 − Mtk , ∆Nk = Ntk+1 − Ntk .

The relation we need to prove can also be written as

n−1
X  
ms-lim (∆Mk )2 − ∆Nk = 0.
n→∞
k=0

Let
Yk = (∆Mk )2 − ∆Nk = (∆Mk )2 − ∆Mk − λ∆tk .
It suffices to show that
h n−1
X i
E Yk = 0, (3.9.20)
k=0
h n−1
X i
lim V ar Yk = 0. (3.9.21)
n→∞
k=0

The first identity follows from the properties of Poisson processes (see Exercise 2.8.9)

h n−1
X i n−1
X n−1
X
E Yk = E[Yk ] = E[(∆Mk )2 ] − E[∆Nk ]
k=0 k=0 k=0
n−1
X
= (λ∆tk − λ∆tk ) = 0.
k=0

For the proof of the identity (3.9.21) we need to find first the variance of Yk .

V ar[Yk ] = V ar[(∆Mk )2 − (∆Mk + λ∆tk )] = V ar[(∆Mk )2 − ∆Mk ]

= V ar[(∆Mk )2 ] + V ar[∆Mk ] − 2Cov[∆Mk2 , ∆Mk ]
= λ∆tk + 2λ2 ∆t2k + λ∆tk
h i
−2 E[(∆Mk )3 ] − E[(∆Mk )2 ]E[∆Mk ]
= 2λ2 (∆tk )2 ,
80

where we used Exercise 2.8.9 and the fact that E[∆Mk ] = 0. Since Mt is a process with
independent increments, then Cov[Yk , Yj ] = 0 for i 6= j. Then

h n−1
X i n−1
X X n−1
X
V ar Yk = V ar[Yk ] + 2 Cov[Yk , Yj ] = V ar[Yk ]
k=0 k=0 k6=j k=0
n−1
X n−1
X
= 2λ2 (∆tk )2 ≤ 2λ2 k∆n k ∆tk = 2λ2 (b − a)k∆n k,
k=0 k=0

h n−1
X i
and hence V ar Yn → 0 as k∆n k → 0. According to the Example 1.13.1, we obtain the
k=0
desired limit in mean square.
The previous result states that the quadratic variation of the martingale Mt between a and
b is equal to the jump of the Poisson process between a and b.

The Fundamental Relation dMt2 = dNt

Recall relation (3.9.19)

n−1
X
ms- lim (Mtk+1 − Mtk )2 = Nb − Na . (3.9.22)
n→∞
k=0

The right side can be regarded as a Riemann-Stieltjes integral

Z b
Nb − Na = dNt ,
a

while the left side can be regarded as a stochastic integral with respect to (dMt )2
Z b n−1
X
(dMt )2 := ms- lim (Mtk+1 − Mtk )2 .
a n→∞
k=0

Substituting in (3.9.22) yields

Z b Z b
2
(dMt ) = dNt ,
a a

for any a < b. The equivalent differential form is

(dMt )2 = dNt . (3.9.23)

The Relations dt dMt = 0, dWt dMt = 0

In order to show that dt dMt = 0 in the mean square sense, we need to prove the limit
n−1
X
ms- lim (tk+1 − tk )(Mtk+1 − Mtk ) = 0. (3.9.24)
n→∞
k=0
 81

This can be thought as a vanishing integral of the increment process dMt with respect to dt
Z b
dMt dt = 0, ∀a, b ∈ R.
a

Denote
n−1
X n−1
X
Xn = (tk+1 − tk )(Mtk+1 − Mtk ) = ∆tk ∆Mk .
k=0 k=0

In order to show (3.9.24) it suffices to prove that

1. E[Xn ] = 0;

2. lim V ar[Xn ] = 0.
n→∞

Using the additivity of the expectation and Exercise 2.8.9, (ii)

h n−1
X i n−1
X
E[Xn ] = E ∆tk ∆Mk = ∆tk E[∆Mk ] = 0.
k=0 k=0

Since the Poisson process Nt has independent increments, the same property holds for the
compensated Poisson process Mt . Then ∆tk ∆Mk and ∆tj ∆Mj are independent for k 6= j, and
using the properties of variance we have

h n−1
X i n−1
X n−1
X
V ar[Xn ] = V ar ∆tk ∆Mk = (∆tk )2 V ar[∆Mk ] = λ (∆tk )3 ,
k=0 k=0 k=0

where we used
V ar[∆Mk ] = E[(∆Mk )2 ] − (E[∆Mk ])2 = λ∆tk ,
see Exercise 2.8.9 (ii). If we let k∆n k = max ∆tk , then
k

n−1
X n−1
X
3 2
V ar[Xn ] = λ (∆tk ) ≤ λk∆n k ∆tk = λ(b − a)k∆n k2 → 0
k=0 k=0

as n → ∞. Hence we proved the stochastic differential relation

dt dMt = 0. (3.9.25)

For showing the relation dWt dMt = 0, we need to prove

ms- lim Yn = 0, (3.9.26)

n→∞

where we have denoted

n−1
X n−1
X
Yn = (Wk+1 − Wk )(Mtk+1 − Mtk ) = ∆Wk ∆Mk .
k=0 k=0
82

Since the Brownian motion Wt and the process Mt have independent increments and ∆Wk is
independent of ∆Mk , we have
n−1
X n−1
X
E[Yn ] = E[∆Wk ∆Mk ] = E[∆Wk ]E[∆Mk ] = 0,
k=0 k=0

where we used E[∆Wk ] = E[∆Mk ] = 0. Using also E[(∆Wk )2 ] = ∆tk , E[(∆Mk )2 ] = λ∆tk ,
and invoking the independence of ∆Wk and ∆Mk , we get

V ar[∆Wk ∆Mk ] = E[(∆Wk )2 (∆Mk )2 ] − (E[∆Wk ∆Mk ])2

= E[(∆Wk )2 ]E[(∆Mk )2 ] − E[∆Wk ]2 E[∆Mk ]2
= λ(∆tk )2 .

Then using the independence of the terms in the sum, we get

n−1
X n−1
X
V ar[Yn ] = V ar[∆Wk ∆Mk ] = λ (∆tk )2
k=0 k=0
n−1
X
≤ λk∆n k ∆tk = λ(b − a)k∆n k → 0,
k=0

as n → ∞. Since Yn is a random variable with mean zero and variance decreasing to zero, it
follows that Yn → 0 in the mean square sense. Hence we proved that

dWt dMt = 0. (3.9.27)

Exercise 3.9.6 Show the following stochastic differential relations:

(a) dt dNt = 0; (b) dWt dNt = 0; (c) dt dWt = 0;
(d) (dNt )2 = dNt ; (e) (dMt )2 = dNt ; (f ) (dMt )4 = dNt .

The relations proved in this section will be useful in the Part II when developing the stochas-
tic model of a stock price that exhibits jumps modeled by a Poisson process.
Chapter 4

Stochastic Integration

This chapter deals with one of the most useful stochastic integrals, called the Ito integral.
This type of integral was introduced in 1944 by the Japanese mathematician K. Ito, and was
originally motivated by a construction of diffusion processes.

4.0.3 Nonanticipating Processes

Consider the Brownian motion Wt . A process Ft is called a nonanticipating process if Ft is
independent of the increment Wt′ − Wt for any t and t′ with t < t′ . Consequently, the process
Ft is independent of the behavior of the Brownian motion in the future, i.e. it cannot anticipate
the future. For instance, Wt , eWt , Wt2 − Wt + t are examples of nonanticipating processes, while
Wt+1 or 12 (Wt+1 − Wt )2 are not.
Nonanticipating processes are important because the Ito integral concept applies only to
them.
If Ft denotes the information known until time t, where this information is generated by
the Brownian motion {Ws ; s ≤ t}, then any Ft -adapted process Ft is nonanticipating.

4.0.4 Increments of Brownian Motions

In this section we shall discuss a few basic properties of the increments of a Brownian motion,
which will be useful when computing stochastic integrals.

Proposition 4.0.7 Let Wt be a Brownian motion. If s < t, we have

1. E[(Wt − Ws )2 ] = t − s.
2. V ar[(Wt − Ws )2 ] = 2(t − s)2 .

Proof: 1. Using that Wt − Ws ∼ N (0, t − s), we have

E[(Wt − Ws )2 ] = E[(Wt − Ws )2 ] − (E[Wt − Ws ])2 = V ar(Wt − Ws ) = t − s.

Wt − Ws
2. Dividing by the standard deviation yields the standard normal random variable √ ∼
t−s
2
(Wt − Ws )
N (0, 1). Its square, is χ2 -distributed with 1 degree of freedom.1 Its mean is 1 and
t−s
1A χ2 -distributed random variable with n degrees of freedom has mean n and variance 2n.

83
84

its variance is 2. This implies

h (W − W )2 i
t s
E = 1 =⇒ E[(Wt − Ws )2 ] = t − s;
t−s
h (W − W )2 i
t s
V ar = 2 =⇒ V ar[(Wt − Ws )2 ] = 2(t − s)2 .
t−s

Remark 4.0.8 The infinitesimal version of the previous result is obtained by replacing t − s
with dt
1. E[dWt2 ] = dt;
2. V ar[dWt2 ] = 2dt2 .
We shall see in an upcoming section that in fact dWt2 and dt are equal in a mean square sense.

Exercise 4.0.9 Show that

(a) E[(Wt − Ws )4 ] = 3(t − s)2 ;
(b) E[(Wt − Ws )6 ] = 15(t − s)3 .

4.1 The Ito Integral

The Ito integral is defined in a way that is similar to the Riemann integral. The Ito integral is
taken with respect to infinitesimal increments of a Brownian motion, dWt , which are random
variables, while the Riemann integral considers integration with respect to the predictable
infinitesimal changes dt. It is worth noting that the Ito integral is a random variable, while the
Riemann integral is just a real number. Despite this fact, there are several common properties
and relations between these two types of integrals.
Consider 0 ≤ a < b and let Ft = f (Wt , t) be a nonanticipating process with
hZ b i
E Ft2 dt < ∞. (4.1.1)
a

Divide the interval [a, b] into n subintervals using the partition points

a = t0 < t1 < · · · < tn−1 < tn = b,

and consider the partial sums

n−1
X
Sn = Fti (Wti+1 − Wti ).
i=0

We emphasize that the intermediate points are the left endpoints of each interval, and this is
the way they should be always chosen. Since the process Ft is nonanticipative, the random
variables Fti and Wti+1 − Wti are independent; this is an important feature in the definition of
the Ito integral.
The Ito integral is the limit of the partial sums Sn
 85

Z b
ms-lim Sn = Ft dWt ,
n→∞ a

provided the limit exists. It can be shown that the choice of partition does not influence the
value of the Ito integral. This is the reason why, for practical purposes, it suffices to assume
the intervals equidistant, i.e.

(b − a)
ti+1 − ti = a + , i = 0, 1, · · · , n − 1.
n
The previous convergence is in the mean square sense, i.e.
h Z b 2 i
lim E Sn − Ft dWt = 0.
n→∞ a

Existence of the Ito integral

Z b
It is known that the Ito stochastic integral Ft dWt exists if the process Ft = f (Wt , t) satisfies
a
the following properties:
1. The paths t → Ft (ω) are continuous on [a, b] for any state of the world ω ∈ Ω;
2. The process Ft is nonanticipating for t ∈ [a, b];
hZ b i
3. E Ft2 dt < ∞.
a

For instance, the following stochastic integrals exist:

Z T Z T Z b
cos(Wt )
Wt2 dWt , sin(Wt ) dWt , dWt .
0 0 a t

4.2 Examples of Ito integrals

As in the case of the Riemann integral, using the definition is not an efficient way of computing
integrals. The same philosophy applies to Ito integrals. We shall compute in the following two
simple Ito integrals. In later sections we shall introduce more efficient methods for computing
Ito integrals.

4.2.1 The case Ft = c, constant

In this case the partial sums can be computed explicitly
n−1
X n−1
X
Sn = Fti (Wti+1 − Wti ) = c(Wti+1 − Wti )
i=0 i=0
= c(Wb − Wa ),

and since the answer does not depend on n, we have

Z b
c dWt = c(Wb − Wa ).
a
86

In particular, taking c = 1, a = 0, and b = T , since the Brownian motion starts at 0, we have

the following formula:
Z T
dWt = WT .
0

4.2.2 The case Ft = Wt

We shall integrate the process Wt between 0 and T . Considering an equidistant partition, we
kT
take tk = , k = 0, 1, · · · , n − 1. The partial sums are given by
n
n−1
X
Sn = Wti (Wti+1 − Wti ).
i=0

Since
1
xy = [(x + y)2 − x2 − y 2 ],
2
letting x = Wti and y = Wti+1 − Wti yields
1 2 1 1
Wti (Wti+1 − Wti ) = W − W 2 − (Wti+1 − Wti )2 .
2 ti+1 2 ti 2
Then after pair cancelations the sum becomes
n−1 n−1 n−1
1X 2 1X 2 1X
Sn = W − W − (Wti+1 − Wti )2
2 i=0 ti+1 2 i=0 ti 2 i=0
n−1
1 2 1X
= Wtn − (Wti+1 − Wti )2 .
2 2 i=0

Using tn = T , we get
n−1
1 2 1X
Sn = WT − (Wti+1 − Wti )2 .
2 2 i=0
Since the first term is independent of n, using Proposition 3.9.1, we have
n−1
1 2 1X
ms- lim Sn = WT − ms- lim (Wti+1 − Wti )2 (4.2.2)
n→∞ 2 n→∞ 2
i=0
1 2 1
= W − T. (4.2.3)
2 T 2
We have now obtained the following explicit formula of a stochastic integral:
Z T
1 2 1
Wt dWt = W − T.
0 2 T 2

In a similar way one can obtain

Z b
1 1
Wt dWt = (W 2 − Wa2 ) − (b − a).
a 2 b 2
 87

It is worth noting that the right side contains random variables depending on the limits of
integration a and b.

Exercise 4.2.1 Show the following identities:

RT
(a) E[ 0
dWt ] = 0;
RT
(b) E[ 0 Wt dWt ] = 0;
Z T
T2
(c) V ar[ Wt dWt ] = .
0 2

4.3 Properties of the Ito Integral

We shall start with some properties which are similar with those of the Riemannian integral.

Proposition 4.3.1 Let f (Wt , t), g(Wt , t) be nonanticipating processes and c ∈ R. Then we
have

1. Additivity:

Z T Z T Z T
[f (Wt , t) + g(Wt , t)] dWt = f (Wt , t) dWt + g(Wt , t) dWt .
0 0 0

2. Homogeneity:
Z T Z T
cf (Wt , t) dWt = c f (Wt , t) dWt .
0 0

3. Partition property:

Z T Z u Z T
f (Wt , t) dWt = f (Wt , t) dWt + f (Wt , t) dWt , ∀0 < u < T.
0 0 u

Proof: 1. Consider the partial sum sequences

n−1
X
Xn = f (Wti , ti )(Wti+1 − Wti )
i=0
n−1
X
Yn = g(Wti , ti )(Wti+1 − Wti ).
i=0
88

Z T Z T
Since ms-lim Xn = f (Wt , t) dWt and ms-lim Yn = g(Wt , t) dWt , using Proposition
n→∞ 0 n→∞ 0
1.14.2 yields
Z T  
f (Wt , t) + g(Wt , t) dWt
0
n−1
X 
= ms-lim f (Wti , ti ) + g(Wti , ti ) (Wti+1 − Wti )
n→∞
i=0
h n−1
X n−1
X i
= ms-lim f (Wti , ti )(Wti+1 − Wti ) + g(Wti , ti )(Wti+1 − Wti )
n→∞
i=0 i=0
= ms-lim (Xn + Yn ) = ms-lim Xn + ms-lim Yn
n→∞ n→∞ n→∞
Z T Z T
= f (Wt , t) dWt + g(Wt , t) dWt .
0 0

The proofs of parts 2 and 3 are left as an exercise for the reader.
Some other properties, such as monotonicity, do not hold in general. It is possible to have a
RT
nonnegative random variable Ft for which the random variable 0 Ft dWt has negative values.
Some of the random variable properties of the Ito integral are given by the following result:
Proposition 4.3.2 We have
1. Zero mean:
hZ b i
E f (Wt , t) dWt = 0.
a
2. Isometry:
h Z b 2 i hZ b i
E f (Wt , t) dWt =E f (Wt , t)2 dt .
a a
3. Covariance:
h Z b  Z b i hZ b i
E f (Wt , t) dWt g(Wt , t) dWt =E f (Wt , t)g(Wt , t) dt .
a a a

We shall discuss the previous properties giving rough reasons why they hold true. The
detailed proofs are beyond the goal of this book.
Pn−1
1. The Ito integral is the mean square limit of the partial sums Sn = i=0 fti (Wti+1 − Wti ),
where we denoted fti = f (Wti , ti ). Since f (Wt , t) is a nonanticipative process, then fti is
independent of the increments Wti+1 − Wti , and hence we have

h n−1
X i n−1
X
E[Sn ] = E fti (Wti+1 − Wti ) = E[fti (Wti+1 − Wti )]
i=0 i=0
n−1
X
= E[fti ]E[(Wti+1 − Wti )] = 0,
i=0

because the increments have mean zero. Since each partial sum has zero mean, their limit,
which is the Ito Integral, will also have zero mean.
 89

2. Since the square of the sum of partial sums can be written as

 n−1
X 2
Sn2 = fti (Wti+1 − Wti )
i=0
n−1
X X
= ft2i (Wti+1 − Wti )2 + 2 fti (Wti+1 − Wti )ftj (Wtj+1 − Wtj ),
i=0 i6=j

using the independence yields

n−1
X
E[Sn2 ] = E[ft2i ]E[(Wti+1 − Wti )2 ]
i=0
X
+2 E[fti ]E[(Wti+1 − Wti )]E[ftj ]E[(Wtj+1 − Wtj )]
i6=j
n−1
X
= E[ft2i ](ti+1 − ti ),
i=0
Z b hZ b i
which are the Riemann sums of the integral E[ft2 ] dt = E ft2 dt , where the last identity
a a
follows from Fubini’s theorem. Hence E[Sn2 ] converges to the aforementioned integral. It is yet
to be shown that the convergence holds also in mean square.
3. Consider the partial sums
n−1
X n−1
X
Sn = fti (Wti+1 − Wti ), Vn = gtj (Wtj+1 − Wtj ).
i=0 j=0

Their product is
 n−1
X  n−1
X 
Sn Vn = fti (Wti+1 − Wti ) gtj (Wtj+1 − Wtj )
i=0 j=0
n−1
X n−1
X
= fti gti (Wti+1 − Wti )2 + fti gtj (Wti+1 − Wti )(Wtj+1 − Wtj )
i=0 i6=j

Using that ft and gt are nonanticipative and that

E[(Wti+1 − Wti )(Wtj+1 − Wtj )] = E[Wti+1 − Wti ]E[Wtj+1 − Wtj ] = 0, i 6= j

2
E[(Wti+1 − Wti ) ] = ti+1 − ti ,

it follows that
n−1
X
E[Sn Vn ] = E[fti gti ]E[(Wti+1 − Wti )2 ]
i=0
n−1
X
= E[fti gti ](ti+1 − ti ),
i=0
90

Rb
which is the Riemann sum for the integral E[ft gt ] dt. a
Rb
From 1 and 2 it follows that the random variable a f (Wt , t) dWt has mean zero and variance

hZ b i hZ b i
V ar f (Wt , t) dWt = E f (Wt , t)2 dt .
a a

From 1 and 3 it follows that

hZ b Z b i Z b
Cov f (Wt , t) dWt , g(Wt , t) dWt = E[f (Wt , t)g(Wt , t)] dt.
a a a

Corollary 4.3.3 (Cauchy’s integral inequality) Let f (t) = f (Wt , t) and g(t) = g(Wt , t).
Then
Z b 2  Z b  Z b 
E[ft gt ] dt ≤ E[ft2 ] dt E[gt2 ] dt .
a a a

Proof: It follows from the previous theorem and from the correlation formula |Corr(X, Y )| =
|Cov(X, Y )|
≤ 1.
[V ar(X)V ar(Y )]1/2
Let Ft be the information set at time t. This implies that fti and Wti+1 − Wti are known
n−1
X
at time t, for any ti+1 ≤ t. It follows that the partial sum Sn = fti (Wti+1 − Wti ) is
i=0
Ft -predictable. The following result states that this is also valid in mean square:
Rt
Proposition 4.3.4 The Ito integral 0 fs dWs is Ft -predictable.

The following two results state that if the upper limit of an Ito integral is replaced by the
parameter t we obtain a continuous martingale.

Proposition 4.3.5 For any s < t we have

hZ t i Z s
E f (Wu , u) dWu |Fs = f (Wu , u) dWu .
0 0

Proof: Using part 3 of Proposition 4.3.2 we get

hZ t i
E f (Wu , u) dWu |Fs
0
hZ s Z t i
= E f (Wu , u) dWu + f (Wu , u) dWu |Fs
0 s
hZ s i hZ t i
= E f (Wu , u) dWu |Fs + E f (Wu , u) dWu |Fs . (4.3.4)
0 s
Rs
Since 0
f (Wu , u) dWu is Fs -predictable (see Proposition 4.3.4), by part 2 of Proposition 1.11.4
hZ s i Z s
E f (Wu , u) dWu |Fs = f (Wu , u) dWu .
0 0
 91
Rt
Since s f (Wu , u) dWu contains only information between s and t, it is unpredictable given the
information set Fs , so
hZ t i
E f (Wu , u) dWu |Fs = 0.
s
Substituting into (4.3.4) yields the desired result.
Rt
Proposition 4.3.6 Consider the process Xt = 0 f (Ws , s) dWs . Then Xt is continuous, i.e.
for almost any state of the world ω ∈ Ω, the path t → Xt (ω) is continuous.
Proof: A rigorous proof is beyond the purpose of this book. We shall provide a rough sketch.
Assume the process f (Wt , t) satisfies E[f (Wt , t)2 ] < M , for some M > 0. Let t0 be fixed and
consider h > 0. Consider the increment Yh = Xt0 +h −Xt0 . Using the aforementioned properties
of the Ito integral we have
h Z t0 +h i
E[Yh ] = E[Xt0 +h − Xt0 ] = E f (Wt , t) dWt = 0
t0
h Z t0 +h 2 i Z t0 +h
E[Yh2 ] = E f (Wt , t) dWt = E[f (Wt , t)2 ] dt
t0 t0
Z t0 +h
< M dt = M h.
t0

The process Yh has zero mean for any h > 0 and its variance tends to 0 as h → 0. Using a
convergence theorem yields that Yh tends to 0 in mean square, as h → 0. This is equivalent
with the continuity of Xt at t0 .

4.4 The Wiener Integral

The Wiener integral is a particular case of the Ito stochastic integral. It is obtained by replacing
the nonanticipating stochastic process f (Wt , t) by the deterministic function f (t). The Wiener
Rb
integral a f (t) dWt is the mean square limit of the partial sums
n−1
X
Sn = f (ti )(Wti+1 − Wti ).
i=0

All properties of Ito integrals also hold for Wiener integrals. The Wiener integral is a random
variable with zero mean
hZ b i
E f (t) dWt = 0
a
and variance
h Z b 2 i Z b
E f (t) dWt = f (t)2 dt.
a a
However, in the case of Wiener integrals we can say something about their distribution.
Rb
Proposition 4.4.1 The Wiener integral I(f ) = a f (t) dWt is a normal random variable with
mean 0 and variance Z b
V ar[I(f )] = f (t)2 dt := kf k2L2 .
a
92

Proof: Since increments Wti+1 − Wti are normally distributed with mean 0 and variance
ti+1 − ti , then
f (ti )(Wti+1 − Wti ) ∼ N (0, f (ti )2 (ti+1 − ti )).
Since these random variables are independent, by the Central Limit Theorem (see Theorem
2.3.1), their sum is also normally distributed, with
n−1
X  n−1
X 
Sn = f (ti )(Wti+1 − Wti ) ∼ N 0, f (ti )2 (ti+1 − ti ) .
i=0 i=0

Taking n → ∞ and max kti+1 − ti k → 0, the normal distribution tends to

i

 Z b 
N 0, f (t)2 dt .
a

The previous convergence holds in distribution, and it still needs to be shown in the mean
square. However, we shall omit this essential proof detail.
RT
Exercise 4.4.2 Show that the random variable X = 1 √1t dWt is normally distributed with
mean 0 and variance ln T .
RT √
Exercise 4.4.3 Let Y = 1 t dWt . Show that Y is normally distributed with mean 0 and
variance (T 2 − 1)/2.
Rt
Exercise 4.4.4 Find the distribution of the integral 0 et−s dWs .
Rt Rt
Exercise 4.4.5 Show that Xt = 0 (2t − u) dWu and Yt = 0 (3t − 4u) dWu are Gaussian
processes with mean 0 and variance 37 t3 .
Z
1 t
Exercise 4.4.6 Show that ms- lim u dWu = 0.
t→0 t 0

Rt 
Exercise 4.4.7 Find all constants a, b such that Xt = 0 a + bu
t dWu is normally distributed
with variance t.

Exercise 4.4.8 Let n be a positive integer. Prove that

Z t
  tn+1
Cov Wt , un dWu = .
0 n+1
Formulate and prove a more general result.

4.5 Poisson Integration

In this section we deal with the integration with respect to the compensated Poisson process
Mt = Nt − λt, which is a martingale. Consider 0 ≤ a < b and let Ft = F (t, Mt ) be a
nonanticipating process with
hZ b i
E Ft2 dt < ∞.
a
 93

Consider the partition

a = t0 < t1 < · · · < tn−1 < tn = b
of the interval [a, b], and associate the partial sums
n−1
X
Sn = Fti− (Mti+1 − Mti ),
i=0

where Fti− is the left-hand limit at ti− . For predictability reasons, the intermediate points are
the left-handed limit to the endpoints of each interval. Since the process Ft is nonanticipative,
the random variables Fti− and Mti+1 − Mti are independent.
The integral of Ft− with respect to Mt is the mean square limit of the partial sum Sn
Z T
ms-lim Sn = Ft− dMt ,
n→∞ 0
provided the limit exists. More precisely, this convergence means that
h Z b 2 i
lim E Sn − Ft− dMt = 0.
n→∞ a

4.5.1 A Work Out Example: the case Ft = Mt

We shall integrate the process Mt− between 0 and T with respect to Mt . Considering the
kT
partition points tk = , k = 0, 1, · · · , n − 1. The partial sums are given by
n
n−1
X
Sn = Mti− (Mti+1 − Mti ).
i=0

1
Using xy = [(x + y)2 − x2 − y 2 ], by letting x = Mti− and y = Mti+1 − Mti , we get (Where
2
does a minus go?)
1 2 1 1
Mti− (Mti+1 − Mti ) = Mti+1 − Mt2i − (Mti+1 − Mti )2 .
2 2 2
After pair cancelations we have
n−1 n−1 n−1
1X 2 1X 2 1X
Sn = M − M − (Mti+1 − Mti )2
2 i=0 ti+1 2 i=0 ti 2 i=0
n−1
1 2 1X
= Mtn − (Mti+1 − Mti )2
2 2 i=0

Since tn = T , we get
n−1
1 2 1X
Sn = MT − (Mti+1 − Mti )2 .
2 2 i=0
The second term on the right is the quadratic variation of Mt . Using formula (3.9.19) yields
that Sn converges in mean square towards 21 MT2 − 12 NT , since N0 = 0.
94

Hence we have arrived at the following formula

Z T
1 2 1
Mt− dMt = M − NT .
0 2 T 2

Similarly, one can obtain

Z b
1 1
Mt− dMt = (Mb2 − Ma2 ) − (Nb − Na ).
a 2 2

hZ b i
Exercise 4.5.1 (a) Show that E Mt− dMt = 0,
a
hZ b i
(b) Find V ar Mt dMt .
a

Exercise 4.5.2 Let ω be a fixed state of the world and assume the sample path t → Nt (ω) has
a jump in the interval (a, b). Show that the Riemann-Stieltjes integral
Z b
Nt (ω) dNt
a

does not exist.

Exercise 4.5.3 Let Nt− denote the left-hand limit of Nt . Show that Nt− is predictable, while
Nt is not.

The previous exercises provide the reason why in the following we shall work with Mt− instead
Z b Z b
of Mt : the integral Mt dNt might not exist, while Mt− dNt does exist.
a a

Exercise 4.5.4 Show that

Z T Z t
1
Nt− dMt = (Nt2 − Nt ) − λ Nt dt.
0 2 0

Exercise 4.5.5 Find the variance of

Z T
Nt− dMt .
0

The following integrals with respect to a Poisson process Nt are considered in the Riemann-
Stieltjes sense.
Proposition 4.5.6 For any continuous function f we have
hZ t i Z t
(a) E f (s) dNs = λ f (s) ds;
0 0
h Z t 2 i Z t Z t 2
(b) E f (s) dNs =λ f (s)2 ds + λ2 f (s) ds ;
h R t0 i R t f (s)
0 0
f (s) dN λ (e −1) ds
(c) E e 0 s
=e 0 .
 95

Proof: (a) Consider the equidistant partition 0 = s0 < s1 < · · · < sn = t, with sk+1 − sk = ∆s.
Then

hZ t i h n−1
X i n−1
X h i
E f (s) dNs = lim E f (si )(Nsi+1 − Nsi ) = lim f (si )E Nsi+1 − Nsi
0 n→∞ n→∞
i=0 i=0
n−1
X Z t
= λ lim f (si )(si+1 − si ) = λ f (s) ds.
n→∞ 0
i=0

(b) Using that Nt is stationary and has independent increments, we have respectively
E[(Nsi+1 − Nsi )2 ] = E[Ns2i+1 −si ] = λ(si+1 − si ) + λ2 (si+1 − si )2
= λ ∆s + λ2 (∆s)2 ,
E[(Nsi+1 − Nsi )(Nsj+1 − Nsj )] = E[(Nsi+1 − Nsi )]E[(Nsj+1 − Nsj )]
= λ(si+1 − si )λ(sj+1 − sj ) = λ2 (∆s)2 .
Applying the expectation to the formula
 n−1
X 2 n−1
X
f (si )(Nsi+1 − Nsi ) = f (si )2 (Nsi+1 − Nsi )2
i=0 i=0
X
+2 f (si )f (sj )(Nsi+1 − Nsi )(Nsj+1 − Nsj )
i6=j

yields
h n−1
X 2 i n−1
X X
E f (si )(Nsi+1 − Nsi ) = f (si )2 (λ∆s + λ2 (∆s)2 ) + 2 f (si )f (sj )λ2 (∆s)2
i=0 i=0 i6=j
n−1
X h n−1
X X i
= λ f (si )2 ∆s + λ2 f (si )2 (∆s)2 + 2 f (si )f (sj )(∆s)2
i=0 i=0 i6=j
n−1
X  n−1
X 2
= λ f (si )2 ∆s + λ2 f (si ) ∆s
i=0 i=0
Z t Z t 2
→ λ f (s)2 ds + λ2 f (s) ds , as n → ∞.
0 0

(c) Using that Nt is stationary with independent increments and has the moment generating
k
function E[ekNt ] = eλ(e −1)t , we have
h Rt i h Pn−1 i h n−1
Y i
E e 0 f (s) dNs = lim E e i=0 f (si )(Nsi+1 −Nsi ) = lim E ef (si )(Nsi+1 −Nsi )
n→∞ n→∞
i=0
n−1
Y h i n−1
Y h i
= lim E ef (si )(Nsi+1 −Nsi ) = lim E ef (si )(Nsi+1 −si )
n→∞ n→∞
i=0 i=0
n−1
Y Pn−1
f (si )
(ef (si ) −1)(si+1 −si )
= lim eλ(e −1)(si+1 −si )
= lim eλ i=0
n→∞ n→∞
i=0
Rt
λ (ef (s) −1) ds
= e 0 .
96

Z t
Since f is continuous, the Poisson integral f (s) dNs can be computed in terms of the
0
waiting times Sk
Z t Nt
X
f (s) dNs = f (Sk ).
0 k=1

This formula can be used to give a proof for the previous result. For instance, taking the
expectation and using conditions over Nt = n, yields

hZ t i Nt
hX n
i X hX i
E f (s) dNs = E f (Sk ) = E f (Sk )|Nt = n P (Nt = n)
0 k=1 n≥0 k=1
XnZ t (λt)n −λt
Z t
1 X (λt)n
−λt
= f (x) dx e =e f (x) dx
t 0 n! 0 t (n − 1)!
n≥0 n≥0
Z t Z t
= e−λt f (x) dx λeλt = λ f (x) dx.
0 0

Exercise 4.5.7 Solve parts (b) and (c) of Proposition 4.5.6 using a similar idea with the one
presented above.

Exercise 4.5.8 Show that

h Z t 2 i Z t
E f (s) dMs =λ f (s)2 ds,
0 0

where Mt = Nt − λt is the compensated Poisson process.

Exercise 4.5.9 Prove that

Z t  Z t
V ar f (s) dNs = λ f (s)2 dNs .
0 0

Exercise 4.5.10 Find h Rt i

E e 0 f (s) dMs .

Proposition 4.5.11 Let Ft = σ(Ns ; 0 ≤ s ≤ t). Then for any constant c, the process
c
Mt = ecNt +λ(1−e )t
, t≥0

is an Ft -martingale.

Proof: Let s < t. Since Nt − Ns is independent of Fs and Nt is stationary, we have

E[ec(Nt −Ns ) |Fs ] = E[ec(Nt −Ns ) ] = E[ecNt−s ]

c
= eλ(e −1)(t−s) .

On the other side, taking out the predictable part yields

E[ec(Nt −Ns ) |Fs ] = e−cNs E[ecNt |Fs ].

 97

Equating the last two relations we arrive at

c c
E[ecNt +(1−e )t
|Fs ] = ecNs +λ(1−e )s
,

which is equivalent with the martingale condition E[Mt |Fs ] = Ms .

We shall present an application of the previous result. Consider the waiting time until the
nth jump, Sn = inf{t > 0; Nt = n}, which is a stopping time, and the filtration Ft = σ(Ns ; 0 ≤
s ≤ t). Since
c
Mt = ecNt +λ(1−e )t
is an Ft -martingale, by the Optional Stopping Theorem (Theorem 3.2.1) we have E[MSn ] =
c
E[M0 ] = 1, which is equivalent with E[eλ(1−e )Sn ] = e−cn . Substituting s = −λ(1 − ec ), then
c = ln(1 + λs ). Since s, λ > 0, then c > 0. The previous expression becomes

s
 λ x
E[e−sSn ] = e−n ln(1+ λ ) = .
λ+s
Since the expectation on the left side is the Laplace transform of the probability density of Sn ,
then
n λ x o
p(Sn ) = L−1 {E[e−sSn ]} = L−1
λ+s
e−tλ tn−1 λn
= ,
Γ(n)

which shows that Sn has a gamma distribution.

RT
4.6 The distribution function of XT = 0 g(t) dNt
In this section we consider the function g(t) continuous. Let S1 < S2 < · · · < SNt denote the
waiting times until time t. Since the increments dNt are equal to 1 at Sk and 0 otherwise, the
integral can be written as
Z T
XT = g(t) dNt = g(S1 ) + · · · + g(SNt ).
0
RT
The distribution function of the random variable XT = 0 g(t) dNt can be obtained condition-
ing over the Nt
X
P (XT ≤ u) = P (XT ≤ u|NT = k) P (NT = k)
k≥0
X
= P (g(S1 ) + · · · + g(SNt ) ≤ u|NT = k) P (NT = k)
k≥0
X
= P (g(S1 ) + · · · + g(Sk ) ≤ u) P (NT = k). (4.6.5)
k≥0

Considering S1 , S2 , · · · , Sk independent and uniformly distributed over the interval [0, T ], we

have
98

Z

1 vol(Dk )
P g(S1 ) + · · · + g(Sk ) ≤ u = k
dx1 · · · dxk = ,
Dk T Tk
where
Dk = {g(x1 ) + g(x2 ) + · · · + g(xk ) ≤ u} ∩ {0 ≤ x1 , · · · , xk ≤ T }.
Substituting back in (4.6.5) yields
X
P (XT ≤ u) = P (g(S1 ) + · · · + g(Sk ) ≤ u) P (NT = k)
k≥0
X vol(Dk ) λk T k X λk vol(Dk )
−λT −λT
= e = e . (4.6.6)
Tk k! k!
k≥0 k≥0

In general, the volume of the k-dimensional solid Dk is not obvious easy to obtain. However,
there are simple cases when this can be computed explicitly.
A Particular Case. We shall do an explicit computation of the partition function of XT =
RT 2
s dNs . In this case the solid Dk is the intersection between the k-dimensional ball of radius
√0
u centered at the origin and the k-dimensional
√ cube [0, T ]k . There are three possible shapes
for Dk , which depend on the size of u:
√
(a) if 0 ≤ u < T , then Dk is a 21k -part of a k-dimensional sphere;
√ √
(b) if T ≤ u < T k, then Dk has a complicated shape;
√ √
(c) if T k ≤ u, then Dk is the entire k-dimensional cube, and then vol(Dk ) = T k .
π k/2 Rk
Since the volume of the k-dimensional ball of radius R is given by , then the volume
Γ( k2 + 1)
of Dk in case (a) becomes
π k/2 uk/2
vol(Dk ) = .
2k Γ( k2 + 1)
Substituting in (4.6.6) yields
X (λ2 πu)k/2 √
P (XT ≤ u) = e−λT , 0≤ u < T.
k≥0
k!Γ( k2 + 1)
√ √
It is worth noting that for u → ∞, the inequality T k ≤ u is satisfied for all k ≥ 0; hence
relation (4.6.6) yields
X λk T k
lim P (XT ≤ u) = e−λT = e−kT ekT = 1.
u→∞ k!
k≥0

The computation in case (b) is more complicated and will be omitted.

hR i R 
T T
Exercise 4.6.1 Calculate the expectation E 0 eks dNs and the variance V ar 0 eks dNs .
RT
Exercise 4.6.2 Compute the distribution function of Xt = 0 s dNs .
Chapter 5

Stochastic Differentiation

5.1 Differentiation Rules

Most stochastic processes are not differentiable. For instance, the Brownian motion process
Wt is a continuous process which is nowhere differentiable. Hence, derivatives like dW
dt do not
t

make sense in stochastic calculus. The only quantities allowed to be used are the infinitesimal
changes of the process, in our case dWt .
The infinitesimal change of a process
The change in the process Xt between instances t and t + ∆t is given by ∆Xt = Xt+∆t − Xt .
When ∆t is infinitesimally small, we obtain the infinitesimal change of a process Xt

dXt = Xt+dt − Xt .

Sometimes it is useful to use the equivalent formula Xt+dt = Xt + dXt .

5.2 Basic Rules

The following rules are the analog of some familiar differentiation rules from elementary Cal-
culus.
The constant multiple rule
If Xt is a stochastic process and c is a constant, then

d(c Xt ) = c dXt .

The verification follows from a straightforward application of the infinitesimal change formula

d(c Xt ) = c Xt+dt − c Xt = c(Xt+dt − Xt ) = c dXt .

The sum rule

If Xt and Yt are two stochastic processes, then

d(Xt + Yt ) = dXt + dYt .

99
100

The verification is as in the following:

d(Xt + Yt ) = (Xt+dt + Yt+dt ) − (Xt + Yt )
= (Xt+dt − Xt ) + (Yt+dt − Yt )
= dXt + dYt .
The difference rule
If Xt and Yt are two stochastic processes, then

d(Xt − Yt ) = dXt − dYt .

The proof is similar to the one for the sum rule.

The product rule
If Xt and Yt are two stochastic processes, then

d(Xt Yt ) = Xt dYt + Yt dXt + dXt dYt .

The proof is as follows:

d(Xt Yt ) = Xt+dt Yt+dt − Xt Yt
= Xt (Yt+dt − Yt ) + Yt (Xt+dt − Xt ) + (Xt+dt − Xt )(Yt+dt − Yt )
= Xt dYt + Yt dXt + dXt dYt ,
where the second identity is verified by direct computation.
If the process Xt is replaced by the deterministic function f (t), then the aforementioned
formula becomes
d(f (t)Yt ) = f (t) dYt + Yt df (t) + df (t) dYt .
Since in most practical cases the process Yt is an Ito diffusion
dYt = a(t, Wt )dt + b(t, Wt )dWt ,
using the relations dt dWt = dt2 = 0, the last term vanishes
df (t) dYt = f ′ (t)dtdYt = 0,
and hence
d(f (t)Yt ) = f (t) dYt + Yt df (t).
This relation looks like the usual product rule.
The quotient rule
If Xt and Yt are two stochastic processes, then
 X  Y dX − X dY − dX dY Xt
t t t t t t t
d = 2 + 3 (dYt )2 .
Yt Yt Yt

The proof follows from Ito’s formula and will be addressed in section 5.3.3.
When the process Yt is replaced by the deterministic function f (t), and Xt is an Ito diffusion,
then the previous formula becomes
 X  f (t)dX − X df (t)
t t t
d = .
f (t) f (t)2
 101

Example 5.2.1 We shall show that

d(Wt2 ) = 2Wt dWt + dt.

Applying the product rule and the fundamental relation (dWt )2 = dt, yields
d(Wt2 ) = Wt dWt + Wt dWt + dWt dWt = 2Wt dWt + dt.
Example 5.2.2 Show that
d(Wt3 ) = 3Wt2 dWt + 3Wt dt.
Applying the product rule and the previous exercise yields
d(Wt3 ) = d(Wt · Wt2 ) = Wt d(Wt2 ) + Wt2 dWt + d(Wt2 ) dWt
= Wt (2Wt dWt + dt) + Wt2 dWt + dWt (2Wt dWt + dt)
= 2Wt2 dWt + Wt dt + Wt2 dWt + 2Wt (dWt )2 + dt dWt
= 3Wt2 dWt + 3Wt dt,
where we used (dWt )2 = dt and dt dWt = 0.
Example 5.2.3 Show that d(tWt ) = Wt dt + t dWt .
Using the product rule and dt dWt = 0, we get
d(tWt ) = Wt dt + t dWt + dt dWt
= Wt dt + t dWt .
Rt
Example 5.2.4 Let Zt = 0 Wu du be the integrated Brownian motion. Show that

dZt = Wt dt.
The infinitesimal change of Zt is
Z t+dt
dZt = Zt+dt − Zt = Ws ds = Wt dt,
t

since Ws is a continuous function in s.

Rt
Example 5.2.5 Let At = 1t Zt = 1t 0 Wu du be the average of the Brownian motion on the
time interval [0, t]. Show that
1 1 
dAt = Wt − Zt dt.
t t
We have
1 1 1
dAt = d Zt + dZt + d dZt
t t t
−1 1 −1
= 2
dt2
Zt dt + Wt dt + 2 Wt |{z}
t t t
=0
1 1 
= Wt − Zt dt.
t t
1
Rt
Exercise 5.2.1 Let Gt = t 0 eWu du be the average of the geometric Brownian motion on
[0, t]. Find dGt .
102

5.3 Ito’s Formula

Ito’s formula is the analog of the chain rule from elementary Calculus. We shall start by
reviewing a few concepts regarding function approximations.
Let f be a differentiable function of a real variable x. Let x0 be fixed and consider the
changes ∆x = x − x0 and ∆f (x) = f (x) − f (x0 ). It is known from Calculus that the following
second order Taylor approximation holds
1
∆f (x) = f ′ (x)∆x + f ′′ (x)(∆x)2 + O(∆x)3 .
2
When x is infinitesimally close to x0 , we replace ∆x by the differential dx and obtain
1
df (x) = f ′ (x)dx + f ′′ (x)(dx)2 + O(dx)3 . (5.3.1)
2
In the elementary Calculus, all terms involving terms of equal or higher order to dx2 are
neglected; then the aforementioned formula becomes

df (x) = f ′ (x)dx.

Now, if we consider x = x(t) to be a differentiable function of t, substituting into the previous

formula we obtain the differential form of the well known chain rule




df x(t) = f ′ x(t) dx(t) = f ′ x(t) x′ (t) dt.

We shall present a similar formula for the stochastic environment. In this case the deter-
ministic function x(t) is replaced by a stochastic process Xt . The composition between the
differentiable function f and the process Xt is denoted by Ft = f (Xt ). Since the increments
involving powers of dt2 or higher are neglected, we may assume that the same holds true for
the increment dXt . Then the expression (5.3.1) becomes

1

2
dFt = f ′ Xt dXt + f ′′ Xt dXt . (5.3.2)
2

In the computation of dXt we may take into the account stochastic relations such as dWt2 = dt,
or dt dWt = 0.

5.3.1 Ito’s formula for diffusions

The previous formula is a general case of Ito’s formula. However, in most cases the increments
dXt are given by some particular relations. An important case is when the increment is given
by
dXt = a(Wt , t)dt + b(Wt , t)dWt .
A process Xt satisfying this relation is called an Ito diffusion.

Theorem 5.3.1 (Ito’s formula for diffusions) If Xt is an Ito diffusion, and Ft = f (Xt ),
then
h b(Wt , t)2 ′′ i
dFt = a(Wt , t)f ′ (Xt ) + f (Xt ) dt + b(Wt , t)f ′ (Xt ) dWt . (5.3.3)
2
 103

Proof: We shall provide a formal proof. Using relations dWt2 = dt and dt2 = dWt dt = 0, we
have
 2
(dXt )2 = a(Wt , t)dt + b(Wt , t)dWt
= a(Wt , t)2 dt2 + 2a(Wt , t)b(Wt , t)dWt dt + b(Wt , t)2 dWt2
= b(Wt , t)2 dt.

Substituting into (5.3.2) yields

1

2
dFt = f ′ Xt dXt + f ′′ Xt dXt
2

  1

= ′
f Xt a(Wt , t)dt + b(Wt , t)dWt + f ′′ Xt b(Wt , t)2 dt
2
h
′ b(Wt , t)2 ′′ i
= a(Wt , t)f (Xt ) + f (Xt ) dt + b(Wt , t)f ′ (Xt ) dWt .
2

In the case Xt = Wt we obtain the following consequence:

Corollary 5.3.2 Let Ft = f (Wt ). Then

1 ′′
dFt = f (Wt )dt + f ′ (Wt ) dWt . (5.3.4)
2

Particular cases

1. If f (x) = xα , with α constant, then f ′ (x) = αxα−1 and f ′′ (x) = α(α − 1)xα−2 . Then (5.3.4)
becomes the following useful formula

1
d(Wtα ) = α(α − 1)Wtα−2 dt + αWtα−1 dWt .
2

A couple of useful cases easily follow:

d(Wt2 ) = 2Wt dWt + dt

d(Wt3 ) = 3Wt2 dWt + 3Wt dt.

2. If f (x) = ekx , with k constant, f ′ (x) = kekx , f ′′ (x) = k 2 ekx . Therefore

1
d(ekWt ) = kekWt dWt + k 2 ekWt dt.
2

In particular, for k = 1, we obtain the increments of a geometric Brownian motion

1
d(eWt ) = eWt dWt + eWt dt.
2
3. If f (x) = sin x, then
1
d(sin Wt ) = cos Wt dWt − sin Wt dt.
2
104

Exercise 5.3.3 Use the previous rules to find the following increments
(a) d(Wt eWt )
(b) d(3Wt2 + 2e5Wt )
2
(c) d(et+Wt )


(d) d (t + Wt )n .
1 Z t 
(e) d Wu du
t 0
1 Z t 
(f ) d α eWu du , where α is a constant.
t 0

In the case when the function f = f (t, x) is also time dependent, the analog of (5.3.1) is
given by
1
df (t, x) = ∂t f (t, x)dt + ∂x f (t, x)dx + ∂x2 f (t, x)(dx)2 + O(dx)3 + O(dt)2 . (5.3.5)
2
Substituting x = Xt yields
1
df (t, Xt ) = ∂t f (t, Xt )dt + ∂x f (t, Xt )dXt + ∂x2 f (t, Xt )(dXt )2 . (5.3.6)
2
If Xt is an Ito diffusion we obtain an extra-term in formula (5.3.3)
h b(Wt , t)2 2 i
dFt = ∂t f (t, Xt ) + a(Wt , t)∂x f (t, Xt ) + ∂x f (t, Xt ) dt
2
+b(Wt , t)∂x f (t, Xt ) dWt . (5.3.7)

Exercise 5.3.4 Show that

d(tWt2 ) = (t + Wt2 )dt + 2tWt dWt .

Exercise 5.3.5 Find the following increments

(a) d(tWt ) (c) d(t2 cos Wt )

(b) d(et Wt ) (d) d(sin t Wt2 ).

5.3.2 Ito’s formula for Poisson processes

Consider the process Ft = F (Mt ), where Mt = Nt − λt is the compensated Poisson process.
Ito’s formula for the process Ft takes the following integral form. For a proof the reader can
consult Kuo [?].

Proposition 5.3.6 Let F be a twice differentiable function. Then for any a < t we have
Z t X  
Ft = Fa + F ′ (Ms− ) dMs + ∆F (Ms ) − F ′ (Ms− )∆Ms ,
a a<s≤t

where ∆Ms = Ms − Ms− and ∆F (Ms ) = F (Ms ) − F (Ms− ).

 105

We shall apply the aforementioned result for the case Ft = F (Mt ) = Mt2 . We have
Z t X  
Mt2 = Ma2 + 2 Ms− dMs + Ms2 − Ms−2
− 2Ms− (Ms − Ms− ) . (5.3.8)
a a<s≤t

Since the jumps in Ns are of size 1, we have (∆Ns )2 = ∆Ns . Since the difference of the
processes Ms and Ns is continuous, then ∆Ms = ∆Ns . Using these formulas we have
   
2
Ms2 − Ms− − 2Ms− (Ms − Ms− ) = (Ms − Ms− ) Ms + Ms− − 2Ms−
= (Ms − Ms− )2 = (∆Ms )2 = (∆Ns )2
= ∆Ns = Ns − Ns− .
P
Since the sum of the jumps between s and t is a<s≤t ∆Ns = Nt −Na , formula (5.3.8) becomes
Z t
Mt2 = Ma2 + 2 Ms− dMs + Nt − Na . (5.3.9)
a
The differential form is
d(Mt2 ) = 2Mt− dMt + dNt ,
which is equivalent with
d(Mt2 ) = (1 + 2Mt− ) dMt + λdt,
since dNt = dMt + λdt.
Exercise 5.3.7 Show that Z T
1
Mt− dMt = (M 2 − NT ).
0 2 T
Exercise 5.3.8 Use Ito’s formula for the Poison process to find the conditional expectation
E[Mt2 |Fs ] for s < t.

5.3.3 Ito’s multidimensional formula

If the process Ft depends on several Ito diffusions, say Ft = f (t, Xt , Yt ), then a similar formula
to (5.3.7) leads to
∂f ∂f ∂f
dFt = (t, Xt , Yt )dt + (t, Xt , Yt )dXt + (t, Xt , Yt )dYt
∂t ∂x ∂y
1 ∂2f 1 ∂2f
+ 2
(t, Xt , Yt )(dXt )2 + (t, Xt , Yt )(dYt )2
2 ∂x 2 ∂y 2
∂2f
+ (t, Xt , Yt )dXt dYt .
∂x∂y
Particular cases
In the case when Ft = f (Xt , Yt ), with Xt = Wt1 , Yt = Wt2 independent Brownian motions, we
have
∂f ∂f 1 ∂2f 1 ∂2f
dFt = dWt1 + dWt2 + (dW 1 2
t ) + (dWt2 )2
∂x ∂y 2 ∂x2 2 ∂y 2
1 ∂2f
+ dWt1 dWt2
2 ∂x∂y
∂f ∂f 1  ∂ 2f ∂2f 
= dWt1 + dWt2 + + dt
∂x ∂y 2 ∂x2 ∂y 2
106

The expression
1  ∂2f ∂2f 
∆f = 2
+ 2
2 ∂x ∂y
is called the Laplacian of f . We can rewrite the previous formula as
∂f ∂f
dFt = dWt1 + dWt2 + ∆f dt
∂x ∂y
A function f with ∆f = 0 is called harmonic. The aforementioned formula in the case of
harmonic functions takes the simple form
∂f ∂f
dFt = dWt1 + dWt2 . (5.3.10)
∂x ∂y
Exercise 5.3.9 Let Wt1 , Wt2 be two independent Brownian motions. If the function f is har-
monic, show that Ft = f (Wt1 , Wt2 ) is a martingale. Is the converse true?
Exercise 5.3.10 Use the previous formulas to find dFt in the following cases
(a) Ft = (Wt1 )2 + (Wt2 )2
(b) Ft = ln[(Wt1 )2 + (Wt2 )2 ].
p
Exercise 5.3.11 Consider the Bessel process Rt = (Wt1 )2 + (Wt2 )2 , where Wt1 and Wt2 are
two independent Brownian motions. Prove that
1 W1 W2
dRt = dt + t dWt1 + t dWt2 .
2Rt Rt Rt
Example 5.3.1 (The product rule) Let Xt and Yt be two processes. Show that
d(Xt Yt ) = Yt dXt + Xt dYt + dXt dYt .
Consider the function f (x, y) = xy. Since ∂x f = y, ∂y f = x, ∂x2 f = ∂y2 f = 0, ∂x ∂y = 1, then
Ito’s multidimensional formula yields


d(Xt Yt ) = d f (X, Yt ) = ∂x f dXt + ∂y f dYt
1 1
+ ∂x2 f (dXt )2 + ∂y2 f (dYt )2 + ∂x ∂y f dXt dYt
2 2
= Yt dXt + Xt dYt + dXt dYt .
Example 5.3.2 (The quotient rule) Let Xt and Yt be two processes. Show that
 X  Y dX − X dY − dX dY Xt
t t t t t t t
d = + 2 (dYt )2 .
Yt Yt2 Yt
Consider the function f (x, y) = xy . Since ∂x f = y1 , ∂y f = − yx2 , ∂x2 f = 0, ∂y2 f = − yx2 , ∂x ∂y = 1
y2 ,
then applying Ito’s multidimensional formula yields
X 

t
d = d f (X, Yt ) = ∂x f dXt + ∂y f dYt
Yt
1 1
+ ∂x2 f (dXt )2 + ∂y2 f (dYt )2 + ∂x ∂y f dXt dYt
2 2
1 Xt 1
= dXt − 2 dYt − 2 dXt dYt
Yt Yt Yt
Yt dXt − Xt dYt − dXt dYt Xt
= 2 + 2 (dYt )2 .
Yt Yt
Chapter 6

Stochastic Integration Techniques

Computing a stochastic integral starting from the definition of the Ito integral is a quite ineffi-
cient method. Like in the elementary Calculus, several methods can be developed to compute
stochastic integrals. In order to keep the analogy with the elementary Calculus, we have called
them Fundamental Theorem of Stochastic Calculus and Integration by Parts. The integration
by substitution is more complicated in the stochastic environment and we have considered only
a particular case of it, which we called The method of heat equation.

6.1 Fundamental Theorem of Stochastic Calculus

Consider a process Xt whose increments satisfy the equation dXt = f (t, Wt )dWt . Integrating
formally between a and t yields
Z t Z t
dXs = f (s, Ws )dWs . (6.1.1)
a a
The integral on the left side can be computed as in the following. If we consider the partition
0 = t0 < t1 < · · · < tn−1 < tn = t, then
Z t n−1
X
dXs = ms-lim (Xtj+1 − Xtj ) = Xt − Xa ,
a n→∞
j=0

since we canceled the terms in pairs. Substituting into formula (6.1.1) yields Xt = Xa +
Z t Z t 
f (s, Ws )dWs , and hence dXt = d f (s, Ws )dWs , since Xa is a constant.
a a
Theorem 6.1.1 (The Fundamental Theorem of Stochastic Calculus)
(i) For any a < t, we have
Z t 
d f (s, Ws )dWs = f (t, Wt )dWt .
a

(ii) If Yt is a stochastic process, such that Yt dWt = dFt , then

Z b
Yt dWt = Fb − Fa .
a

107
108

We shall provide a few applications of the aforementioned theorem.

Example 6.1.1 Verify the stochastic formula

Z t
W2 t
Ws dWs = t − .
0 2 2
Rt Wt2 t
Let Xt = 0 Ws dWs and Yt = − . From Ito’s formula
2 2
W2  t 1 1
t
dYt = d −d = (2Wt dWt + dt) − dt = Wt dWt ,
2 2 2 2
and from the Fundamental Theorem of Stochastic Calculus
Z t 
dXt = d Ws dWs = Wt dWt .
0

Hence dXt = dYt , or d(Xt − Yt ) = 0. Since the process Xt − Yt has zero increments, then
Xt − Yt = c, constant. Taking t = 0, yields
Z 0 W2
0 0
c = X0 − Y0 = Ws dWs − − = 0,
0 2 2
and hence c = 0. It follows that Xt = Yt , which verifies the desired relation.

Example 6.1.2 Verify the formula

Z t Z
t 2 t 1 t 2
sWs dWs = Wt − − Ws ds.
0 2 2 2 0
Rt t 2 
1
Rt
Consider the stochastic processes Xt = 0 sWs dWs , Yt = Wt − 1 , and Zt = 2 0
Ws2 ds.
2
The Fundamental Theorem yields

dXt = tWt dWt

1 2
dZt = W dt.
2 t
Applying Ito’s formula, we get
t t  1  t2 
dYt = d Wt2 − = d(tWt2 ) − d
2 2 2 4
1h 2
i 1
= (t + Wt )dt + 2tWt dWt − tdt
2 2
1 2
= W dt + tWt dWt .
2 t
We can easily see that
dXt = dYt − dZt .
This implies d(Xt − Yt + Zt ) = 0, i.e. Xt − Yt + Zt = c, constant. Since X0 = Y0 = Z0 = 0, it
follows that c = 0. This proves the desired relation.
 109

Example 6.1.3 Show that

Z t
1 3
(Ws2 − s) dWs = W − tWt .
0 3 t

Consider the function f (t, x) = 31 x3 − tx, and let Ft = f (t, Wt ). Since ∂t f = −x, ∂x f = x2 − t,
and ∂x2 f = 2x, then Ito’s formula provides

1
dFt = ∂t f dt + ∂x f dWt + ∂x2 f (dWt )2
2
2 1
= −Wt dt + (Wt − t) dWt + 2Wt dt
2
= (Wt2 − t)dWt .

From the Fundamental Theorem we get

Z t Z t
1 3
(Ws2 − s) dWs = dFs = Ft − F0 = Ft = W − tWt .
0 0 3 t

6.2 Stochastic Integration by Parts

Consider the process Ft = f (t)g(Wt ), with f and g differentiable. Using the product rule yields

dFt = df (t) g(Wt ) + f (t) dg(Wt )

1

= f ′ (t)g(Wt )dt + f (t) g ′ (Wt )dWt + g ′′ (Wt )dt
2
1
= ′
f (t)g(Wt )dt + f (t)g (Wt )dt + f (t)g ′ (Wt )dWt .
′′
2
Writing the relation in the integral form, we obtain the first integration by parts formula:
Z b Z b Z b
′
b
′ 1
f (t)g (Wt ) dWt = f (t)g(Wt ) − f (t)g(Wt ) dt − f (t)g ′′ (Wt ) dt.
a a a 2 a

This formula is to be used when integrating a product between a function of t and a function
of the Brownian motion Wt , for which an antiderivative is known. The following two particular
cases are important and useful in applications.
1. If g(Wt ) = Wt , the aforementioned formula takes the simple form
Z b t=b
Z b
f (t) dWt = f (t)Wt − f ′ (t)Wt dt. (6.2.2)
a t=a a

It is worth noting that the left side is a Wiener integral.

2. If f (t) = 1, then the formula becomes
Z b t=b
Z b
′ 1
g (Wt ) dWt = g(Wt ) − g ′′ (Wt ) dt. (6.2.3)
a t=a 2 a
110

Z T
Application 1 Consider the Wiener integral IT = t dWt . From the general theory, see
0
Proposition 4.4.1, it is known that I is a random variable normally distributed with mean 0
and variance Z T
T3
V ar[IT ] = t2 dt = .
0 3
Recall the definition of integrated Brownian motion
Z t
Zt = Wu du.
0

Formula (6.2.2) yields a relationship between I and the integrated Brownian motion
Z T Z T
IT = t dWt = T WT − Wt dt = T WT − ZT ,
0 0

and hence IT + ZT = T WT . This relation can be used to compute the covariance between IT
and ZT .

Cov(IT + ZT , IT + ZT ) = V ar[T WT ] ⇐⇒
V ar[IT ] + V ar[ZT ] + 2Cov(IT , ZT ) = T 2 V ar[WT ] ⇐⇒
T 3 /3 + T 3 /3 + 2Cov(IT , ZT ) = T 3 ⇐⇒
Cov(IT , ZT ) = T 3 /6,

where we used that V ar[ZT ] = T 3 /3. The processes It and Zt are not independent. Their
correlation coefficient is 0.5 as the following calculation shows

Cov(IT , ZT ) T 3 /6
Corr(IT , ZT ) =  1/2 =
T 3 /3
V ar[IT ]V ar[ZT ]
= 1/2.

x2
Application 2 If we let g(x) = 2 in formula (6.2.3), we get

Z b
Wb2 − Wa2 1
Wt dWt = − (b − a).
a 2 2

It is worth noting that letting a = 0 and b = T , we retrieve a formula that was proved by direct
methods in chapter 2
Z T
W2 T
Wt dWt = T − .
0 2 2

x3
Similarly, if we let g(x) = 3 in (6.2.3) yields

Z b Z b
W3 b
Wt2 dWt = t − Wt dt.
a 3 a a
 111

Application 3
RT
Choosing f (t) = eαt and g(x) = cos x, we shall compute the stochastic integral 0 eαt cos Wt dWt
using the formula of integration by parts
Z T Z T
eαt cos Wt dWt = eαt (sin Wt )′ dWt
0 0
Z T Z
T 1 T αt
= eαt sin Wt − (eαt )′ sin Wt dt −
e (cos Wt )′′ dt
0 0 2 0
Z T Z
1 T αt
= eαT sin WT − α αt
e sin Wt dt + e sin Wt dt
0 2 0
Z
 1  T αt
= eαT sin WT − α − e sin Wt dt.
2 0
1
The particular case α = 2 leads to the following exact formula of a stochastic integral
Z T
t T
e 2 cos Wt dWt = e 2 sin WT . (6.2.4)
0

RT
In a similar way, we can obtain an exact formula for the stochastic integral 0 eβt sin Wt dWt
as follows
Z T Z T
eβt sin Wt dWt = − eβt (cos Wt )′ dWt
0 0
T
Z T Z T
βt βt 1
= −e cos Wt +β e cos Wt dt − eβt cos Wt dt.
0 0 2 0
1
Taking β = 2 yields the closed form formula
Z T
t T
e 2 sin Wt dWt = 1 − e 2 cos WT . (6.2.5)
0

A consequence of the last two formulas and of Euler’s formula

eiWt = cos Wt + i sin Wt ,
is
Z T
t T
e 2 +iWt dWt = i(1 − e 2 +iWT ).
0

The proof details are left to the reader.

A general form of the integration by parts formula
In general, if Xt and Yt are two Ito diffusions, from the product formula
d(Xt Yt ) = Xt dYt + Yt dXt + dXt dYt .
Integrating between the limits a and b
Z b Z b Z b Z b
d(Xt Yt ) = Xt dYt + Yt dXt + dXt dYt .
a a a a
112

From the Fundamental Theorem

Z b
d(Xt Yt ) = Xb Yb − Xa Ya ,
a

so the previous formula takes the following form of integration by parts

Z b Z b Z b
Xt dYt = Xb Yb − Xa Ya − Yt dXt − dXt dYt .
a a a

This formula is of theoretical value. In practice, the term dXt dYt needs to be computed using
the rules Wt2 = dt, and dt dWt = 0.
Exercise 6.2.1 (a) Use integration by parts to get
Z T Z T
1 −1 Wt
2 dWt = tan (WT ) + 2 2 dt, T > 0.
0 1 + Wt 0 (1 + Wt )

(b) Show that

Z T h i
−1 Wt
E[tan (WT )] = − E dt.
0 (1 + Wt2 )2
(c) Prove the double inequality
√ √
3 3 x 3 3
− ≤ ≤ , ∀x ∈ R.
16 (1 + x2 )2 16
(d) Use part (c) to obtain
√ Z T √
3 3 Wt 3 3
− T ≤ 2 2 dt ≤ T.
16 0 (1 + Wt ) 16
(e) Use part (d) to get √ √
3 3 −1 3 3
− T ≤ E[tan (WT )] ≤ T.
16 16
(f ) Does part (e) contradict the inequality
π π
− < tan−1 (WT ) < ?
2 2
Exercise 6.2.2 (a) Show the relation
Z T Z
1 T Wt
eWt dWt = eWT − 1 − e dt.
0 2 0
(b) Use part (a) to find E[eWt ].
Exercise 6.2.3 (a) Use integration by parts to show
Z T Z
1 T Wt
Wt eWt dWt = 1 + WT eWT − eWT − e (1 + Wt ) dt;
0 2 0
(b) Use part (a) to find E[Wt eWt ];
(c) Show that Cov(Wt , eWt ) = tet/2 ;
r
t
(d) Prove that Corr(Wt , eWt ) = , and compute the limits as t → 0 and t → ∞.
et − 1
 113

Exercise 6.2.4 (a) Let T > 0. Show the following relation using integration by parts
Z T Z T
2Wt 1 − Wt2
dWt = ln(1 + WT2 ) − dt.
0 1 + Wt2 0 (1 + Wt2 )2

(b) Show that for any real number x the following double inequality holds

1 1 − x2
− ≤ ≤ 1.
8 (1 + x2 )2

(c) Use part (b) to show that

Z T
1 1 − Wt2
− T ≤ dt ≤ T.
8 0 (1 + Wt2 )2

(d) Use parts (a) and (c) to get

T
− ≤ E[ln(1 + WT2 )] ≤ T.
8
(e) Use Jensen’s inequality to get

E[ln(1 + WT2 )] ≤ ln(1 + T ).

Does this contradict the upper bound provided in (d)?

6.3 The Heat Equation Method

In elementary Calculus, integration by substitution is the inverse application of the chain rule.
In the stochastic environment, this will be the inverse application of Ito’s formula. This is
difficult to apply in general, but there is a particular case of great importance.
Let ϕ(t, x) be a solution of the equation

1
∂t ϕ + ∂x2 ϕ = 0. (6.3.6)
2
This is called the heat equation without sources. The non-homogeneous equation
1
∂t ϕ + ∂x2 ϕ = G(t, x) (6.3.7)
2
is called heat equation with sources. The function G(t, x) represents the density of heat sources,
while the function ϕ(t, x) is the temperature at point x at time t in a one-dimensional wire. If
the heat source is time independent, then G = G(x), i.e. G is a function of x only.

Example 6.3.1 Find all solutions of the equation (6.3.6) of type ϕ(t, x) = a(t) + b(x).

Substituting into equation (6.3.6) yields

1 ′′
b (x) = −a′ (t).
2
114

Since the left side is a function of x only, while the right side is a function of variable t, the
only where the previous equation is satisfied is when both sides are equal to the same constant
C. This is called a separation constant. Therefore a(t) and b(x) satisfy the equations
1 ′′
a′ (t) = −C, b (x) = C.
2
Integrating yields a(t) = −Ct + C0 and b(x) = Cx2 + C1 x + C2 . It follows that

ϕ(t, x) = C(x2 − t) + C1 x + C3 ,

with C0 , C1 , C2 , C3 arbitrary constants.

Example 6.3.2 Find all solutions of the equation (6.3.6) of the type ϕ(t, x) = a(t)b(x).

Substituting into the equation and dividing by a(t)b(x) yields

a′ (t) 1 b′′ (x)

+ = 0.
a(t) 2 b(x)

a′ (t) b′′ (x)

There is a separation constant C such that = −C and = 2C. There are three
a(t) b(x)
distinct cases to discuss:
1. C = 0. In this case a(t) = a0 and b(x) = b1 x + b0 , with a0 , a1 , b0 , b1 real constants. Then

ϕ(t, x) = a(t)b(x) = c1 x + c0 , c0 , c1 ∈ R

is just a linear function in x.

2
2. C > 0. Let λ > 0 such that 2C = λ2 . Then a′ (t) = − λ2 a(t) and b′′ (x) = λ2 b(x), with
solutions
2
a(t) = a0 e−λ t/2

b(x) = c1 eλx + c2 e−λx .

The general solution of (6.3.6) is

2
ϕ(t, x) = e−λ t/2
(c1 eλx + c2 e−λx ), c1 , c2 ∈ R.
λ2
3. C < 0. Let λ > 0 such that 2C = −λ2 . Then a′ (t) = 2 a(t) and b′′ (x) = −λ2 b(x).
Solving yields
2
a(t) = a0 eλ t/2
b(x) = c1 sin(λx) + c2 cos(λx).

The general solution of (6.3.6) in this case is

2

ϕ(t, x) = eλ t/2
c1 sin(λx) + c2 cos(λx) , c1 , c2 ∈ R.

In particular, the functions x, x2 − t, ex−t/2 , e−x−t/2 , et/2 sin x and et/2 cos x, or any linear
combination of them are solutions of the heat equation (6.3.6). However, there are other
solutions which are not of the previous type.
 115

Exercise 6.3.1 Prove that ϕ(t, x) = 31 x3 − tx is a solution of the heat equation (6.3.6).
2
Exercise 6.3.2 Show that ϕ(t, x) = t−1/2 e−x /(2t)
is a solution of the heat equation (6.3.6)
for t > 0.
x
Exercise 6.3.3 Let ϕ = u(λ), with λ = √ , t > 0. Show that ϕ satisfies the heat equation
2 t
(6.3.6) if and only if u′′ + 2λu′ = 0.
Z ∞ √
2 2
Exercise 6.3.4 Let erf c(x) = √ e−r dr. Show that ϕ = erf c(x/(2 t)) is a solution of
π x
the equation (6.3.6).
x2
Exercise 6.3.5 (the fundamental solution) Show that ϕ(t, x) = √ 1 e− 4t , t > 0 satisfies
4πt
the equation (6.3.6).

Sometimes it is useful to generate new solutions for the heat equation from other solutions.
Below we present a few ways to accomplish this:
(i) by linear combination: if ϕ1 and ϕ2 are solutions, then a1 ϕ1 + a1 ϕ2 is a solution, where
a1 , a2 constants.
(ii) by translation: if ϕ(t, x) is a solution, then ϕ(t − τ, x − ξ) is a solution, where (τ, ξ) is
a translation vector.
(iii) by affine transforms: if ϕ(t, x) is a solution, then ϕ(λτ, λ2 x) is a solution, for any
constant λ.
∂ n+m
(iv) by differentiation: if ϕ(t, x) is a solution, then n m ϕ(t, x) is a solution.
∂ x∂ t
(v) by convolution: if ϕ(t, x) is a solution, then so are
Z b
ϕ(t, x − ξ)f (ξ) dξ
a
Z b
ϕ(t − τ, x)g(τ ) dt.
a

For more detail on the subject the reader can consult Widder [?] and Cannon [?].

Theorem 6.3.6 Let ϕ(t, x) be a solution of the heat equation (6.3.6) and denote f (t, x) =
∂x ϕ(t, x). Then
Z b
f (t, Wt ) dWt = ϕ(b, Wb ) − ϕ(a, Wa ).
a

Proof: Let Ft = ϕ(t, Wt ). Applying Ito’s formula we get

 1 
dFt = ∂x ϕ(t, Wt ) dWt + ∂t ϕ + ∂x2 ϕ dt.
2
Since ∂t ϕ + 12 ∂x2 ϕ = 0 and ∂x ϕ(t, Wt ) = f (t, Wt ), we have

dFt = f (t, Wt ) dWt .

116

Applying the Fundamental Theorem yields

Z b Z b
f (t, Wt ) dWt = dFt = Fb − Fa = ϕ(b, Wb ) − ϕ(a, Wa ).
a a

Application 6.3.7 Show that

Z T
1 2 1
Wt dWt = W − T.
0 2 T 2

Choose the solution of the heat equation (6.3.6) given by ϕ(t, x) = x2 − t. Then f (t, x) =
∂x ϕ(t, x) = 2x. Theorem 6.3.6 yields
Z T Z T T
2Wt dWt = f (t, Wt ) dWt = ϕ(t, x) = WT2 − T.
0 0 0

Dividing by 2 leads to the desired result.

Application 6.3.8 Show that

Z T
1 3
(Wt2 − t) dWt = W − T WT .
0 3 T

Consider the function ϕ(t, x) = 13 x3 − tx, which is a solution of the heat equation (6.3.6), see
Exercise 6.3.1. Then f (t, x) = ∂x ϕ(t, x) = x2 − t. Applying Theorem 6.3.6 yields
Z T Z T T 1 3
(Wt2 − t) dWt = f (t, Wt ) dWt = ϕ(t, Wt ) = W − T WT .
0 0 0 3 T

Application 6.3.9 Let λ > 0. Prove the identities

Z T
λ2 t 1  − λ2 T ±λWT 
e− 2 ±λWt dWt = e 2 −1 .
0 ±λ

λ2 t
Consider the function ϕ(t, x) = e− 2 ±λx , which is a solution of the homogeneous heat equation
λ2 t
(6.3.6), see Example 6.3.2. Then f (t, x) = ∂x ϕ(t, x) = ±λe− 2 ±λx . Apply Theorem 6.3.6 to
get
Z T Z T T
λ2 t λ2 T
±λe− 2 ±λx dWt = f (t, Wt ) dWt = ϕ(t, Wt ) = e− 2 ±λWT − 1.
0 0 0

Dividing by the constant ±λ ends the proof.

In particular, for λ = 1 the aforementioned formula becomes
Z T
t T
e− 2 +Wt dWt = e− 2 +WT − 1. (6.3.8)
0
 117

Application 6.3.10 Let λ > 0. Prove the identity

Z T
λ2 t 1 λ2 T
e 2 cos(λWt ) dWt = e 2 sin(λWT ).
0 λ

λ2 t
From the Example 6.3.2 we know that ϕ(t, x) = e 2 sin(λx) is a solution of the heat equation.
λ2 t
Applying Theorem 6.3.6 to the function f (t, x) = ∂x ϕ(t, x) = λe 2 cos(λx), yields
Z T Z T T
λ2 t
λe 2 cos(λWt ) dWt = f (t, Wt ) dWt = ϕ(t, Wt )
0 0 0
λ2 t T λ2 T
= e 2 sin(λWt ) =e 2 sin(λWT ).
0

Divide by λ to end the proof.

If we choose λ = 1 we recover a result already familiar to the reader from section 6.2
Z T
t T
e 2 cos(Wt ) dWt = e 2 sin WT . (6.3.9)
0

Application 6.3.11 Let λ > 0. Show that

Z T
λ2 t 1 λ2 T

e 2 sin(λWt ) dWt = 1 − e 2 cos(λWT ) .
0 λ

λ2 t
Choose ϕ(t, x) = e 2 cos(λx) to be a solution of the heat equation. Apply Theorem 6.3.6 for
λ2 t
the function f (t, x) = ∂x ϕ(t, x) = −λe 2 sin(λx) to get
Z T T
λ2 t
(−λ)e 2 sin(λWt ) dWt = ϕ(t, Wt )
0 0
λ2 T T λ2 T
= e 2 cos(λWt ) =e 2 cos(λWT ) − 1,
0

and then divide by −λ.

Application 6.3.12 Let 0 < a < b. Show that

Z b Wb2
3 Wt2 1
2
Wa 1
t− 2 Wt e− 2t dWt = a− 2 e− 2a − b− 2 e− 2b . (6.3.10)
a

2
From Exercise 6.3.2 we have that ϕ(t, x) = t−1/2 e−x /(2t) is a solution of the homogeneous heat
2
equation. Since f (t, x) = ∂x ϕ(t, x) = −t−3/2 xe−x /(2t) , applying Theorem 6.3.6 yields to the
desired result. The reader can easily fill in the details.
Integration techniques will be used when solving stochastic differential equations in the next
chapter.
118

Exercise 6.3.13 Find the value of the following stochastic integrals

Z 1 √
(a) et cos( 2Wt ) dWt
0
Z 3
(b) e2t cos(2Wt ) dWt
0
Z 4 √
(c) e−t+ 2Wt dWt .
0

Exercise 6.3.14 Let ϕ(t, x) be a solution of the following non-homogeneous heat equation with
time-dependent and uniform heat source G(t)
1
∂t ϕ + ∂x2 ϕ = G(t).
2
Denote f (t, x) = ∂x ϕ(t, x). Show that
Z b Z b
f (t, Wt ) dWt = ϕ(b, Wb ) − ϕ(a, Wa ) − G(t) dt.
a a

How does the formula change if the heat source G is constant?

6.4 Table of Usual Stochastic Integrals

Since there are so many formulas in this chapter that we need to use in next chapters, we shall
present them in the following in a single user-friendly table. Let a < b and 0 < T . Then we
have:
Z b
1. dWt = Wb − Wa ;
a
Z T
WT2 T
2. Wt dWt = − ;
0 2 2
Z T
WT2
3. (Wt2 − t) dWt = − T WT ;
0 3
Z T Z T
4. t dWt = T WT − Wt dt, 0 < T ;
0 0
Z T Z T
W3
5. Wt2 dWt = T − Wt dt;
0 3 0
Z T
t T
6. e 2 cos Wt dWt = e 2 sin WT ;
0
Z T
t T
7. e 2 sin Wt dWt = 1 − e 2 cos WT ;
0
Z T
t T
8. e− 2 +Wt = e− 2 +WT − 1;
0
 119

Z T
λ2 t 1 λ2 T
9. e 2 cos(λWt ) dWt = e 2 sin(λWT );
0 λ
Z T
λ2 t 1 λ2 T

10. e 2 sin(λWt ) dWt = 1 − e 2 cos(λWT ) ;
0 λ
Z T
λ2 t 1  − λ2 T ±λWT 
11. e− 2 ±λWt dWt = e 2 −1 ;
0 ±λ
Z b Wb2
3 Wt2 1
2
Wa 1
12. t− 2 Wt e− 2t dWt = a− 2 e− 2a − b− 2 e− 2b ;
a
Z t 
13. d f (s, Ws ) dWs = f (t, Wt ) dWt ;
a
Z b
14. Yt dWt = Fb − Fa , if Yt dWt = dFt ;
a
Z b Z b
15. f (t) dWt = f (t)Wt |ba − f ′ (t)Wt dt;
a a
Z b b
Z b
′ 1
16. g (Wt ) dWt = g(Wt ) − g ′′ (Wt ) dt.
a a 2 a
120
Chapter 7

Stochastic Differential Equations

7.1 Definitions and Examples

Let Xt be a continuous stochastic process. If small changes in the process Xt can be written
as a linear combination of small changes in t and small increments of the Brownian motion Wt ,
we may write
dXt = a(t, Wt , Xt )dt + b(t, Wt , Xt ) dWt (7.1.1)
and call it a stochastic differential equation. In fact, this differential relation has the following
integral meaning:
Z t Z t
Xt = X0 + a(s, Ws , Xs ) ds + b(s, Ws , Xs ) dWs , (7.1.2)
0 0

where the last integral is taken in the Ito sense. Relation (7.1.2) is taken as the definition for
the stochastic differential equation (7.1.1). However, since it is convenient to use stochastic
differentials informally, we shall approach stochastic differential equations by analogy with the
ordinary differential equations, and try to present the same methods of solving equations in the
new stochastic environment.
The functions a(t, Wt , Xt ) and b(t, Wt , Xt ) are called drift rate and volatility, respectively.
A process Xt is called a (strong) solution for the stochastic equation (7.1.1) if it satisfies the
equation. We shall start with an example.

Example 7.1.1 (The Brownian bridge) Let a, b ∈ R. Show that the process
Z t
1
Xt = a(1 − t) + bt + (1 − t) dWs , 0 ≤ t < 1
0 1 − s
is a solution of the stochastic differential equation
b − Xt
dXt = dt + dWt , 0 ≤ t < 1, X0 = a.
1−t
We shall perform a routine verification to show that Xt is a solution. First we compute the
b − Xt
quotient :
1−t

121
122

Z t
1
b − Xt = b − a(1 − t) − bt − (1 − t) dWs
0 1 − s
Z t
1
= (b − a)(1 − t) − (1 − t) dWs ,
0 1−s

and dividing by 1 − t yields

Z t
b − Xt 1
=b−a− dWs . (7.1.3)
1−t 0 1−s

Using
Z t
1  1
d dWs = dWt ,
0 1−s 1−t
the product rule yields
Z t Z t 1 
1
dXt = a d(1 − t) + bdt + d(1 − t) dWs + (1 − t)d dWs
0 1−s 0 1−s
 Z t 
1
= b−a− dWs dt + dWt
0 1−s
b − Xt
= dt + dWt ,
1−t
where the last identity comes from (7.1.3). We just verified that the process Xt is a solution of
the given stochastic equation. The question of how this solution was obtained in the first place,
is the subject of study for the next few sections.

7.2 Finding Mean and Variance from the Equation

For most practical purposes, the most important information one needs to know about a process
is its mean and variance. These can be found directly from the stochastic equation in some
particular cases without solving explicitly the equation. We shall deal in the present section
with this problem.
Taking the expectation in (7.1.2) and using the property of the Ito integral as a zero mean
random variable yields Z t
E[Xt ] = X0 + E[a(s, Ws , Xs )] ds. (7.2.4)
0

Applying the Fundamental Theorem of Calculus we obtain

d
E[Xt ] = E[a(t, Wt , Xt )].
dt
We note that Xt is not differentiable, but its expectation E[Xt ] is. This equation can be solved
exactly in a few particular cases.
d
1. If a(t, Wt , Xt ) = a(t), then dt E[Xt ] = a(t) with the exact solution E[Xt ] = X0 +
Rt
0
a(s) ds.
 123

2. If a(t, Wt , Xt ) = α(t)Xt + β(t), with α(t) and β(t) continuous deterministic function,
then
d
E[Xt ] = α(t)E[Xt ] + β(t),
dt
which is a linear differential equation in E[Xt ]. Its solution is given by
 Z t 
E[Xt ] = eA(t) X0 + e−A(s) β(s) ds , (7.2.5)
0
Rt
where A(t) = 0 α(s) ds. It is worth noting that the expectation E[Xt ] does not depend on the
volatility term b(t, Wt , Xt ).
Exercise 7.2.1 If dXt = (2Xt + e2t )dt + b(t, Wt , Xt )dWt , then
E[Xt ] = e2t (X0 + t).
Proposition 7.2.2 Let Xt be a process satisfying the stochastic equation
dXt = α(t)Xt dt + b(t)dWt .
Then the mean and variance of Xt are given by
E[Xt ] = eA(t) X0
Z t
2A(t)
V ar[Xt ] = e e−A(s) b2 (s) ds,
0
Rt
where A(t) = 0
α(s) ds.
Proof: The expression of E[Xt ] follows directly from formula (7.2.5) with β = 0. In order to
compute the second moment we first compute
(dXt )2 = b2 (t) dt;
d(Xt2 ) = 2Xt dXt + (dXt )2


= 2Xt α(t)Xt dt + b(t)dWt + b2 (t)dt


= 2α(t)Xt2 + b2 (t) dt + 2b(t)Xt dWt ,
where we used Ito’s formula. If we let Yt = Xt2 , the previous equation becomes

p
dYt = 2α(t)Yt + b2 (t) dt + 2b(t) Yt dWt .
Applying formula (7.2.5) with α(t) replaced by 2α(t) and β(t) by b2 (t), yields
 Z t 
2A(t)
E[Yt ] = e Y0 + e−2A(s) b2 (s) ds ,
0

which is equivalent with

 Z t 
E[Xt2 ] =e 2A(t)
X02 + e−2A(s) b2 (s) ds .
0

It follows that the variance is

Z t
V ar[Xt ] = E[Xt2 ] − (E[Xt ]) = e 2 2A(t)
e−2A(s) b2 (s) ds.
0
124

Remark 7.2.3 We note that the previous equation is of linear type. This shall be solved
explicitly in a future section.

The mean and variance for a given stochastic process can be computed by working out the
associated stochastic equation. We shall provide next a few examples.

Example 7.2.1 Find the mean and variance of ekWt , with k constant.

From Ito’s formula

1
d(ekWt ) = kekWt dWt + k 2 ekWt dt,
2
and integrating yields
Z t Z t
1
ekWt = 1 + k ekWs dWs + k 2 ekWs ds.
0 2 0

Taking the expectations we have

Z t
1
E[ekWt ] = 1 + k 2 E[ekWs ] ds.
2 0

If we let f (t) = E[ekWt ], then differentiating the previous relations yields the differential equa-
tion
1
f ′ (t) = k 2 f (t)
2
2
with the initial condition f (0) = E[ekW0 ] = 1. The solution is f (t) = ek t/2
, and hence

2
E[ekWt ] = ek t/2
.

The variance is
2 2
V ar(ekWt ) = E[e2kWt ] − (E[ekWt ])2 = e4k t/2
− ek t
2 2
= ek t (ek t − 1).

Example 7.2.2 Find the mean of the process Wt eWt .

We shall set up a stochastic differential equation for Wt eWt . Using the product formula and
Ito’s formula yields

d(Wt eWt ) = eWt dWt + Wt d(eWt ) + dWt d(eWt )

1
= eWt dWt + (Wt + dWt )(eWt dWt + eWt dt)
2
1
= ( Wt eWt + eWt )dt + (eWt + Wt eWt )dWt .
2
Integrating and using that W0 eW0 = 0 yields
Z t Z t
Wt 1
Wt e = ( Ws eWs + eWs ) ds + (eWs + Ws eWs ) dWs .
0 2 0
 125

Since the expectation of an Ito integral is zero, we have

Z t
1

E[Wt eWt ] = E[Ws eWs ] + E[eWs ] ds.
0 2

Let f (t) = E[Wt eWt ]. Using E[eWs ] = es/2 , the previous integral equation becomes
Z t
1
f (t) = ( f (s) + es/2 ) ds,
0 2

Differentiating yields the following linear differential equation

1
f ′ (t) = f (t) + et/2
2
with the initial condition f (0) = 0. Multiplying by e−t/2 yields the following exact equation

(e−t/2 f (t))′ = 1.

The solution is f (t) = tet/2 . Hence we obtained that

E[Wt eWt ] = tet/2 .

Exercise 7.2.4 Find a E[Wt2 eWt ]; (b) E[Wt ekWt ].

Example 7.2.3 Show that for any integer k ≥ 0 we have

(2k)! k
E[Wt2k ] = t , E[Wt2k+1 ] = 0.
2k k!

In particular, E[Wt4 ] = 3t2 , E[Wt6 ] = 15t3 .

From Ito’s formula we have

n(n − 1) n−2
d(Wtn ) = nWtn−1 dWt + Wt dt.
2
Integrate and get Z Z
t t
n(n − 1)
Wtn = n Wsn−1 dWs + Wsn−2 ds.
0 2 0
Since the expectation of the first integral on the right side is zero, taking the expectation yields
the following recursive relation
Z
n n(n − 1) t
E[Wt ] = E[Wsn−2 ] ds.
2 0

Using the initial values E[Wt ] = 0 and E[Wt2 ] = t, the method of mathematical induction
implies that E[Wt2k+1 ] = 0 and E[Wt2k ] = (2k)! k
2k k! t .

Exercise 7.2.5 (a) Is Wt4 − 3t2 an Ft -martingale?

(b) What about Wt3 ?
126

Example 7.2.4 Find E[sin Wt ].

From Ito’s formula

1
d(sin Wt ) = cos Wt dWt − sin Wt dt,
2
then integrating yields
Z t Z t
1
sin Wt = cos Ws dWs − sin Ws ds.
0 2 0

Taking expectations we arrive at the integral equation

Z
1 t
E[sin Wt ] = − E[sin Ws ] ds.
2 0

Let f (t) = E[sin Wt ]. Differentiating yields the equation f ′ (t) = − 12 f (t) with f (0) = E[sin W0 ] =
0. The unique solution is f (t) = 0. Hence

E[sin Wt ] = 0.

Exercise 7.2.6 Let σ be a constant. Show that

(a) E[sin(σWt )] = 0;
2
(b) E[cos(σWt )] = e−σ t/2
;
−σ2 t/2
(c) E[sin(t + σWt )] = e sin t;
2
(d) E[cos(t + σWt )] = e−σ t/2
cos t;

Exercise 7.2.7 Use the previous exercise and the definition of expectation to show that
Z ∞
2 π 1/2
(a) e−x cos x dx = 1/4 ;
−∞ e
Z ∞ r
−x2 /2 2π
(b) e cos x dx = .
−∞ e

Exercise 7.2.8 Using expectations show that

Z ∞ r  
2 π b b2 /(4a)
(a) xe−ax +bx dx = e ;
−∞ a 2a
Z ∞ r
2 −ax2 +bx π 1 b2  b2 /(4a)
(b) x e dx = 1+ e ;
−∞ a 2a 2a
(c) Can you apply a similar method to find a close form expression for the integral
Z ∞
2
xn e−ax +bx dx?
−∞

Exercise 7.2.9 Using the result given by Example 7.2.3 show that
3
(a) E[cos(tWt )] = e−t /2
;
(b) E[sin(tWt )] = 0;
(c) E[etWt ] = 0.
 127

For general drift rates we cannot find the mean, but in the case of concave drift rates we
can find an upper bound for the expectation E[Xt ]. The following result will be useful.

Lemma 7.2.10 (Gronwall’s inequality) Let f (t) be a non-negative function satisfying the
inequality
Z t
f (t) ≤ C + M f (s) ds
0

for 0 ≤ t ≤ T , with C, M constants. Then

f (t) ≤ CeMt , 0 ≤ t ≤ T.

Proposition 7.2.11 Let Xt be a continuous stochastic process such that

dXt = a(Xt )dt + b(t, Wt , Xt ) dWt ,

with the function a(·) satisfying the following conditions

1. a(x) ≥ 0, for 0 ≤ x ≤ T ;
2. a′′ (x) < 0, for 0 ≤ x ≤ T ;
3. a′ (0) = M .
Then E[Xt ] ≤ X0 eMt , for 0 ≤ Xt ≤ T .

Proof: From the mean value theorem there is ξ ∈ (0, x) such that

a(x) = a(x) − a(0) = (x − 0)a′ (ξ) ≤ xa′ (0) = M x, (7.2.6)

where we used that a′ (x) is a decreasing function. Applying Jensen’s inequality for concave
functions yields
E[a(Xt )] ≤ a(E[Xt ]).
Combining with (7.2.6) we obtain E[a(Xt )] ≤ M E[Xt ]. Substituting in the identity (7.2.4)
implies
Z t
E[Xt ] ≤ X0 + M E[Xs ] ds.
0

Applying Gronwall’s inequality we obtain E[Xt ] ≤ X0 eMt .

Exercise 7.2.12 State the previous result in the particular case when a(x) = sin x, with 0 ≤
x ≤ π.

Not in all cases can the mean and the variance be obtained directly from the stochastic
equation. In these cases we need more powerful methods that produce closed form solutions.
In the next sections we shall discuss several methods of solving stochastic differential equation.

7.3 The Integration Technique

We shall start with the simple case when both the drift and the volatility are just functions of
time t.
128

Proposition 7.3.1 The solution Xt of the stochastic differential equation

dXt = a(t)dt + b(t)dWt

Rt Rt
is Gaussian distributed with mean X0 + 0 a(s) ds and variance 0 b2 (s) ds.

Proof: Integrating in the equation yields

Z t Z t Z t
Xt − X0 = dXs = a(s) ds + b(s) dWs .
0 0 0
Rt
Using the property of Wiener integrals, 0 b(s) dWs is Gaussian distributed with mean 0 and
Rt
variance 0 b2 (s) ds. Then Xt is Gaussian (as a sum between a predictable function and a
Gaussian), with
Z t Z t
E[Xt ] = E[X0 + a(s) ds + b(s) dWs ]
0 0
Z t Z t
= X0 + a(s) ds + E[ b(s) dWs ]
0 0
Z t
= X0 + a(s) ds,
0
Z t Z t
V ar[Xt ] = V ar[X0 + a(s) ds + b(s) dWs ]
0 0
hZ t i
= V ar b(s) dWs
0
Z t
= b2 (s) ds,
0

which ends the proof.

Exercise 7.3.2 Solve the following stochastic differential equations for t ≥ 0 and determine
the mean and the variance of the solution
(a) dXt = cos t dt − sin t dWt , X0 = 1.
√
(b) dXt = et dt + t dWt , X0 = 0.
t
(c) dXt = 1+t2 dt + t3/2 dWt , X0 = 1.

If the drift and the volatility depend on both variables t and Wt , the stochastic differential
equation
dXt = a(t, Wt )dt + b(t, Wt )dWt , t≥0
defines an Ito diffusion. Integrating yields the solution
Z t Z t
Xt = X0 + a(s, Ws ) ds + b(s, Ws ) dWs .
0 0

There are several cases when both integrals can be computed explicitly.
 129

Example 7.3.1 Find the solution of the stochastic differential equation

dXt = dt + Wt dWt , X0 = 1.

Integrate between 0 and t and get

Z t Z t
Wt2 t
Xt = 1+ ds + Ws dWs = t + −
0 0 2 2
1
= (W 2 + t).
2 t
Example 7.3.2 Solve the stochastic differential equation

dXt = (Wt − 1)dt + Wt2 dWt , X0 = 0.

Rt
Let Zt = 0 Ws ds denote the integrated Brownian motion process. Integrating the equation
between 0 and t yields
Z s Z t Z t
Xt = dXs = (Ws − 1)ds + Ws2 dWs
0 0 0
1 1 t
= Zt − t + Wt3 − Wt2 −
3 2 2
1 1 t
= Zt + Wt3 − Wt2 − .
3 2 2
Example 7.3.3 Solve the stochastic differential equation

dXt = t2 dt + et/2 cos Wt dWt , X0 = 0,

and find E[Xt ] and V ar(Xt ).

Integrating yields
Z t Z t
Xt = s2 ds + es/2 cos Ws dWs
0 0
3
t
= + et/2 sin Wt , (7.3.7)
3
where we used (6.3.9). Even if the process Xt is not Gaussian, we can still compute its mean
and variance. By Ito’s formula we have
1
d(sin Wt ) = cos Wt dWt − sin Wt dt
2
Integrating between 0 and t yields
Z t Z t
1
sin Wt = cos Ws dWs − sin Ws ds,
0 2 0

where we used that sin W0 = sin 0 = 0. Taking the expectation in the previous relation yields
hZ t i 1Z t
E[sin Wt ] = E cos Ws dWs − E[sin Ws ] ds.
0 2 0
130

From the properties of the Ito integral, the first expectation on the right side is zero. Denoting
µ(t) = E[sin Wt ], we obtain the integral equation
Z
1 t
µ(t) = − µ(s) ds.
2 0
Differentiating yields the differential equation
1
µ′ (t) = − µ(t)
2
with the solution µ(t) = ke−t/2 . Since k = µ(0) = E[sin W0 ] = 0, it follows that µ(t) = 0.
Hence
E[sin Wt ] = 0.
Taking expectation in (7.3.7) leads to
h t3 i t3
E[Xt ] = E + et/2 E[sin Wt ] = .
3 3
Since the variance of predictable functions is zero,
h t3 i
V ar[Xt ] = V ar + et/2 sin Wt = (et/2 )2 V ar[sin Wt ]
3
et
= et E[sin2 Wt ] = (1 − E[cos 2Wt ]). (7.3.8)
2
In order to compute the last expectation we use Ito’s formula
d(cos 2Wt ) = −2 sin 2Wt dWt − 2 cos 2Wt dt
and integrate to get
Z t Z t
cos 2Wt = cos 2W0 − 2 sin 2Ws dWs − 2 cos 2Ws ds
0 0

Taking the expectation and using that Ito integrals have zero expectation, yields
Z t
E[cos 2Wt ] = 1 − 2 E[cos 2Ws ] ds.
0

If we denote m(t) = E[cos 2Wt ], the previous relation becomes an integral equation
Z t
m(t) = 1 − 2 m(s) ds.
0

Differentiate and get

m′ (t) = −2m(t),
with the solution m(t) = ke−2t . Since k = m(0) = E[cos 2W0 ] = 1, we have m(t) = e−2t .
Substituting into (7.3.8) yields
et et − e−t
V ar[Xt ] = (1 − e−2t ) = = sinh t.
2 2
In conclusion, the solution Xt has the mean and the variance given by
t3
E[Xt ] = , V ar[Xt ] = sinh t.
3
 131

Example 7.3.4 Solve the following stochastic differential equation

et/2 dXt = dt + eWt dWt , X0 = 0,

and find the distribution of the solution Xt and its mean and variance.

Dividing by et/2 , integrating between 0 and t, and using formula (6.3.8) yields
Z t Z t
Xt = e−s/2 ds + e−s/2+Ws dWs
0 0
= 2(1 − e−t/2 ) + e−t/2 eWt − 1
= 1 + e−t/2 (eWt − 2).

Since eWt is a geometric Brownian motion, using Proposition 2.2.2 yields

E[Xt ] = E[1 + e−t/2 (eWt − 2)] = 1 − 2e−t/2 + e−t/2 E[eW

t ]
= 2 − 2e−t/2 .
V ar(Xt ) = V ar[1 + e−t/2 (eWt − 2)] = V ar[e−t/2 eWt ] = e−t V ar[eWt ]
= e−t (e2t − et ) = et − 1.

The process Xt has the following distribution:

F (y) = P (Xt ≤ y) = P 1 + e−t/2 (eWt − 2) ≤ y

 W 1

t
= P Wt ≤ ln 2 + et/2 (y − 1) = P √ ≤ √ ln 2 + et/2 (y − 1)
t t
 1

t/2
= N √ ln 2 + e (y − 1) ,
t
Z u
1 2
where N (u) = √ e−s /2
ds is the distribution function of a standard normal distributed
2π −∞
random variable.

Example 7.3.5 Solve the stochastic differential equation

2
dXt = dt + t−3/2 Wt e−Wt /(2t) dWt , X1 = 1.

Integrating between 1 and t and applying formula (6.3.10) yields

Z t Z t
2
Xt = X1 + ds + s−3/2 Ws e−Ws /(2s) dWs
1 1
2 1 2
= 1 + t − 1 − e−W1 /2 − e−Wt /(2t)
t1/2
2 1 2
= t − e−W1 /2 − e−Wt /(2t) , ∀t ≥ 1.
t1/2
132

7.4 Exact Stochastic Equations

The stochastic differential equation

dXt = a(t, Wt )dt + b(t, Wt )dWt (7.4.9)

is called exact if there is a differentiable function f (t, x) such that

1
a(t, x) = ∂t f (t, x) + ∂x2 f (t, x) (7.4.10)
2
b(t, x) = ∂x f (t, x). (7.4.11)

Assume the equation is exact. Then substituting in (7.4.9) yields

 1 
dXt = ∂t f (t, Wt ) + ∂x2 f (t, Wt ) dt + ∂x f (t, Wt )dWt .
2
Applying Ito’s formula, the previous equations becomes


dXt = d f (t, Wt ) ,

which implies Xt = f (t, Wt ) + c, with c constant.

Solving the partial differential equations system (7.4.10)–(7.4.10) requires the following
steps:
1. Integrating partially with respect to x in the second equation to obtain f (t, x) up to an
additive function T (t);
2. Substitute into the first equation and determine the function T (t);
3. The solution is Xt = f (t, Wt ) + c, with c determined from the initial condition on Xt .

Example 7.4.1 Solve the stochastic differential equation

dXt = et (1 + Wt2 )dt + (1 + 2et Wt )dWt , X0 = 0.

In this case a(t, x) = et (1 + x2 ) and b(t, x) = 1 + 2et x. The associated system is

1
et (1 + x2 ) = ∂t f (t, x) + ∂x2 f (t, x)
2
1 + 2et x = ∂x f (t, x).

Integrate partially in x in the second equation yields

Z
f (t, x) = (1 + 2et x) dx = x + et x2 + T (t).

Then ∂t f = et x2 + T ′ (t) and ∂x2 f = 2et . Substituting in the first equation yields

et (1 + x2 ) = et x2 + T ′ (t) + et .

This implies T ′ (t) = 0, or T = c constant. Hence f (t, x) = x + et x2 + c, and Xt = f (t, Wt ) =

Wt + et Wt2 + c. Since X0 = 0, it follows that c = 0. The solution is Xt = Wt + et Wt2 .
 133

Example 7.4.2 Find the solution of

dXt = 2tWt3 + 3t2 (1 + Wt ) dt + (3t2 Wt2 + 1)dWt , X0 = 0.

The coefficient functions are a(t, x) = 2tx3 + 3t2 (1 + x) and b(t, x) = 3t2 x2 + 1. The associated
system is given by
1
2tx3 + 3t2 (1 + x) = ∂t f (t, x) + ∂x2 f (t, x)
2
3t2 x2 + 1 = ∂x f (t, x).

Integrating partially in the second equation yields

Z
f (t, x) = (3t2 x2 + 1) dx = t2 x3 + x + T (t).

Then ∂t f = 2tx3 + T ′ (t) and ∂x2 f = 6t2 x, and substituting into the first equation we get
1
2tx3 + 3t2 (1 + x) = 2tx3 + T ′ (t) + 6t2 x.
2
After cancellations we get T ′ (t) = 3t2 , so T (t) = t3 + c. Then

f (t, x) = t2 x3 + x + 3t2 = t2 (x3 + 1) + x + c.

The solution process is given by Xt = f (t, Wt ) = t2 (Wt3 + 1) + Wt + c. Using X0 = 0 we get

c = 0. Hence the solution is Xt = t2 (Wt3 + 1) + Wt .
The next result deals with a condition regarding the closeness of the stochastic differential
equation.

Theorem 7.4.1 If the stochastic differential equation (7.4.9) is exact, then the coefficient func-
tions a(t, x) and b(t, x) satisfy the condition

1
∂x a = ∂t b + ∂x2 b. (7.4.12)
2

Proof: If the stochastic equation is exact, there is a function f (t, x) satisfying the system
(7.4.10)–(7.4.10). Differentiating the first equation of the system with respect to x yields
1
∂x a = ∂t ∂x f + ∂x2 ∂x f.
2
Substituting b = ∂x f yields the desired relation.

Remark 7.4.2 The equation (7.4.12) has the meaning of a heat equation. The function b(t, x)
represents the temperature measured at x at the instance t, while ∂x a is the density of heat
sources. The function a(t, x) can be regarded as the potential from which the density of heat
sources is derived by taking the gradient in x.

It is worth noting that equation (7.4.12) is a just necessary condition for exactness. This
means that if this condition is not satisfied, then the equation is not exact. In that case we
need to try a different method to solve the equation.
134

Example 7.4.3 Is the stochastic differential equation

dXt = (1 + Wt2 )dt + (t4 + Wt2 )dWt

exact?
Collecting the coefficients, we have a(t, x) = 1 + x2 , b(t, x) = t4 + x2 . Since ∂x a = 2x, ∂t b = 4t3 ,
and ∂x2 b = 2, the condition (7.4.12) is not satisfied, and hence the equation is not exact.

7.5 Integration by Inspection

When solving a stochastic differential equation by inspection we look for opportunities to apply
the product or the quotient formulas:

d(f (t)Yt ) = f (t) dYt + Yt df (t).

 X  f (t)dX − X df (t)
t t t
d = .
f (t) f (t)2
For instance, if a stochastic differential equation can be written as

dXt = f ′ (t)Wt dt + f (t)dWt ,

the product rule brings the equation into the exact form
 
dXt = d f (t)Wt ,

which after integration leads to the solution

Xt = X0 + f (t)Wt .

Example 7.5.1 Solve

dXt = (t + Wt2 )dt + 2tWt dWt , X0 = a.

We can write the equation as

dXt = Wt2 dt + t(2Wt dWt + dt),

which can be contracted to

dXt = Wt2 dt + td(Wt2 ).
Using the product rule we can bring it to the exact form

dXt = d(tWt2 ),

with the solution Xt = tWt2 + a.

Example 7.5.2 Solve the stochastic differential equation

dXt = (Wt + 3t2 )dt + tdWt .

 135

If we rewrite the equation as

dXt = 3t2 dt + (Wt dt + tdWt ),
we note the exact expression formed by the last two terms Wt dt + tdWt = d(tWt ). Then
dXt = d(t3 ) + d(tWt ),
which is equivalent with d(Xt ) = d(t3 + tWt ). Hence Xt = t3 + tWt + c, c ∈ R.
Example 7.5.3 Solve the stochastic differential equation
e−2t dXt = (1 + 2Wt2 )dt + 2Wt dWt .
Multiply by e2t to get
dXt = e2t (1 + 2Wt2 )dt + 2e2t Wt dWt .
After regrouping, this becomes
dXt = (2e2t dt)Wt2 + e2t (2Wt dWt + dt).
Since d(e2t ) = 2e2t dt and d(Wt2 ) = 2Wt dWt + dt, the previous relation becomes
dXt = d(e2t )Wt2 + e2t d(Wt2 ).
By the product rule, the right side becomes exact
dXt = d(e2t Wt2 ),
and hence the solution is Xt = e2t Wt2 + c, c ∈ R.
Example 7.5.4 Solve the equation
t3 dXt = (3t2 Xt + t)dt + t6 dWt , X1 = 0.
The equation can be written as
t3 dXt − 3Xt t2 dt = tdt + t6 dWt .
Divide by t6 :
t3 dXt − Xt d(t3 )
= t−5 dt + dWt .
(t3 )2
Applying the quotient rule yields
X   t−4 
t
d 3 = −d + dWt .
t 4
Integrating between 1 and t, yields
Xt t−4
= − + Wt − W1 + c
t3 4
so
1
Xt = ct3 − + t3 (Wt − W1 ), c ∈ R.
4t
Using X1 = 0 yields c = 1/4 and hence the solution is
1 3 1
Xt = t − + t3 (Wt − W1 ), c ∈ R.
4 t
136

Exercise 7.5.1 Solve the following stochastic differential equations by the inspection method
(a) dXt = (1 + Wt )dt + (t + 2Wt )dWt , X0 = 0;
(b) t2 dXt = (2t3 − Wt )dt + tdWt , X1 = 0;
(c) e−t/2 dXt = 12 Wt dt + dWt , X0 = 0;
(d) dXt = 2tWt dWt + Wt2 dt, X0 = 0;

1
 √
(e) dXt = 1 + 2√ t
Wt dt + t dWt , X1 = 0.

7.6 Linear Stochastic Differential Equations

Consider the stochastic differential equation with drift term linear in Xt


dXt = α(t)Xt + β(t) dt + b(t, Wt )dWt , t ≥ 0.
This also can be written as
dXt − α(t)Xt dt = β(t)dt + b(t, Wt )dWt .
Rt
Let A(t) = 0 α(s) ds. Multiplying by the integrating factor e−A(t) , the left side of the previous
equation becomes an exact expression
 
e−A(t) dXt − α(t)Xt dt = e−A(t) β(t)dt + e−A(t) b(t, Wt )dWt
 
d e−A(t) Xt = e−A(t) β(t)dt + e−A(t) b(t, Wt )dWt .

Integrating yields
Z t Z t
−A(t) −A(s)
e Xt = X0 + e β(s) ds + e−A(s) b(s, Ws ) dWs
0 0
Z t Z t 
A(t) A(t) −A(s)
Xt = X0 e +e e β(s) ds + e−A(s) b(s, Ws ) dWs .
0 0

The first integral within the previous parentheses is a Riemann integral, and the latter one is
an Ito stochastic integral. Sometimes, in practical applications these integrals can be computed
explicitly.
When b(t, Wt ) = b(t), the latter integral becomes a Wiener integral. In this case the solution
Xt is Gaussian with mean and variance given by
Z t
E[Xt ] = X0 eA(t) + eA(t) e−A(s) β(s) ds
0
Z t
V ar[Xt ] = e2A(t) e−2A(s) b(s)2 ds.
0

Another important particular case is when α(t) = α 6= 0, β(t) = β are constants and
b(t, Wt ) = b(t). The equation in this case is


dXt = αXt + β dt + b(t)dWt , t ≥ 0,
and the solution takes the form
Z t
β αt
Xt = X0 eαt + (e − 1) + eα(t−s) b(s) dWs .
α 0
 137

Example 7.6.1 Solve the linear stochastic differential equation

dXt = (2Xt + 1)dt + e2t dWt .

Write the equation as

dXt − 2Xt dt = dt + e2t dWt
and multiply by the integrating factor e−2t to get
 
d e−2t Xt = e−2t dt + dWt .

Integrate between 0 and t and multiply by e2t , and obtain

Z t Z t
2t 2t −2s 2t
Xt = X0 e + e e ds + e dWs
0 0
1 2t
= X0 e + (e2t − 1) + e2t Wt .
2
Example 7.6.2 Solve the linear stochastic differential equation

dXt = (2 − Xt )dt + e−t Wt dWt .

Multiplying by the integrating factor et yields

et (dXt + Xt dt) = 2et dt + Wt dWt .

Since et (dXt + Xt dt) = d(et Xt ), integrating between 0 and t we get

Z t Z t
e t Xt = X0 + 2et dt + Ws dWs .
0 0
t
Dividing by e and performing the integration yields
1
Xt = X0 e−t + 2(1 − e−t ) + e−t (Wt2 − t).
2
Example 7.6.3 Solve the linear stochastic differential equation
1
dXt = ( Xt + 1)dt + et cos Wt dWt .
2
Write the equation as
1
dXt − Xt dt = dt + et cos Wt dWt
2
and multiply by the integrating factor e−t/2 to get

d(e−t/2 Xt ) = e−t/2 dt + et/2 cos Wt dWt .

Integrating yields Z Z
t t
−t/2 −s/2
e Xt = X0 + e ds + es/2 cos Ws dWs
0 0

Multiply by et/2 and use formula (6.3.9) to obtain the solution

Xt = X0 et/2 + 2(et/2 − 1) + et sin Wt .

138

Exercise 7.6.1 Solve the following linear stochastic differential equations

(a) dXt = (4Xt − 1)dt + 2dWt ;
(b) dXt = (3Xt − 2)dt + e3t dWt ;
(c) dXt = (1 + Xt )dt + et Wt dWt ;
(d) dXt = (4Xt + t)dt + e4t dWt ;
 
(e) dXt = t + 12 Xt dt + et sin Wt dWt ;
(f ) dXt = −Xt dt + e−t dWt .
In the following we present an important example of stochastic differential equations, which
can be solved by the method presented in this section.
Proposition 7.6.2 (The mean-reverting Ornstein-Uhlenbeck process) Let m and α be
two constants. Then the solution Xt of the stochastic equation

dXt = (m − Xt )dt + αdWt (7.6.13)

is given by
Z t
−t
Xt = m + (X0 − m)e +α es−t dWs . (7.6.14)
0

Xt is Gaussian with mean and variance given by

E[Xt ] = m + (X0 − m)e−t
α2
V ar(Xt ) = (1 − e−2t ).
2
Proof: Adding Xt dt to both sides and multiplying by the integrating factor et we get
d(et Xt ) = met dt + αet dWt ,
which after integration yields
Z t
t t
e Xt = X0 + m(e − 1) + α es dWs .
0

Hence
Z t
Xt = X0 e−t + m − e−t + αe−t es dWs
0
Z t
= m + (X0 − m)e−t + α es−t dWs .
0

Since Xt is the sum between a predictable function and a Wiener integral, then we can use
Proposition 4.4.1 and it follows that Xt is Gaussian, with
h Z t i
E[Xt ] = m + (X0 − m)e−t + E α es−t dWs = m + (X0 − m)e−t
0
h Z t i Z t
V ar(Xt ) = V ar α es−t dWs = α2 e−2t e2s ds
0 0
e2t − 1 1
= α2 e−2t = α2 (1 − e−2t ).
2 2
 139

The name mean-reverting comes from the fact that

lim E[Xt ] = m.
t→∞

The variance also tends to zero exponentially, lim V ar[Xt ] = 0. According to Proposition
t→∞
3.8.1, the process Xt tends to m in the mean square sense.

Proposition 7.6.3 (The Brownian bridge) For a, b ∈ R fixed, the stochastic differential
equation
b − Xt
dXt = dt + dWt , 0 ≤ t < 1, X0 = a
1−t
has the solution
Z t
1
Xt = a(1 − t) + bt + (1 − t) dWs , 0 ≤ t < 1. (7.6.15)
0 1−s

The solution has the property limt→1 Xt = b, almost certainly.

Proof: If we let Yt = b − Xt the equation becomes linear in Yt

1
dYt + Yt dt = −dWt .
1−t
1
Multiplying by the integrating factor ρ(t) = 1−t yields
 Y  1
t
d =− dWt ,
1−t 1−t
which leads by integration to Z t
Yt 1
=c− dWs .
1−t 0 1−s
Making t = 0 yields c = a − b, so
Z t
b − Xt 1
=a−b− dWs .
1−t 0 1−s
Solving for Xt yields
Z t
1
Xt = a(1 − t) + bt + (1 − t) dWs , 0 ≤ t < 1.
0 1−s
Rt 1
Let Ut = (1 − t) 0 1−s dWs . First we notice that
Z t
 1 
E[Ut ] = (1 − t)E dWs = 0,
0 1−s
Z Z t
 t 1  1
V ar(Ut ) = (1 − t)2 V ar dWs = (1 − t)2 ds
0 1 − s 0 (1 − s)2
 1 
= (1 − t)2 − 1 = t(1 − t).
1−t
140

In order to show ac-limt→1 Xt = b, we need to prove

P ω; lim Xt (ω) = b = 1.
t→1

Since Xt = a(1 − t) + bt + Ut , it suffices to show that

P ω; lim Ut (ω) = 0 = 1. (7.6.16)
t→1

We evaluate the probability of the complementary event

P ω; lim Ut (ω) 6= 0 = P ω; |Ut (ω)| > ǫ, ∀t ,
t→1

for some ǫ > 0. Since by Markov’s inequality

V ar(Ut ) t(1 − t)
P ω; |Ut (ω)| > ǫ) < 2
=
ǫ ǫ2
holds for any 0 ≤ t < 1, choosing t → 1 implies that

P ω; |Ut (ω)| > ǫ, ∀t) = 0,

which implies (7.6.16).

The process (7.6.15) is called the Brownian bridge because it joins X0 = a with X1 = b.
Since Xt is the sum between a deterministic linear function in t and a Wiener integral, it follows
that it is a Gaussian process, with mean and variance

E[Xt ] = a(1 − t) + bt
V ar(Xt ) = V ar(Ut ) = t(1 − t).

It is worth noting that the variance is maximum at the midpoint t = (b − a)/2 and zero at the
end points a and b.

Example 7.6.4 Find Cov(Xs , Xt ), 0 < s < t for the following cases:
(a) Xt is a mean reverting Orstein-Uhlenbeck process;
(b) Xt is a Brownian bridge process.

Stochastic equations with respect to a Poisson process

Similar techniques can be applied in the case when the Brownian motion process Wt is
replaced by a Poisson process Nt with constant rate λ. For instance, the stochastic differential
equation

dXt = 3Xt dt + e3t dNt

X0 = 1

can be solved multiplying by the integrating factor e−3t to obtain

d(e−3t Xt ) = dNt .

Integrating yields e−3t Xt = Nt + C. Making t = 0 yields C = 1, so the solution is given by

Xt = e3t (1 + Nt ).
 141

The following equation

dXt = (m − Xt )dt + αdNt
is similar with the equation defining the mean-reverting Ornstein-Uhlenbeck process. As we
shall see, in this case, the process is no more mean-reverting. It reverts though to a certain
constant. A similar method yields the solution
Z t
−t −t
Xt = m + (X0 − m)e + αe es dNs .
0

Since from Proposition 4.5.6 and Exercise 4.5.9 we have

Z Z t
 t s 
E e dNs = λ es ds = λ(et − 1)
0 0
Z t Z t
s

λ
V ar e dNs = λ e2s ds = (e2t − 1),
0 0 2
it follows that
E[Xt ] = m + (X0 − m)e−t + αλ(1 − e−t )
λα2
V ar(Xt ) = (1 − e−2t ).
2
It is worth noting that in this case the process Xt is no more Gaussian.

7.7 The Method of Variation of Parameters

Let’s start by considering the following stochastic equation
dXt = αXt dWt , (7.7.17)
with α constant. This is the equation which, in physics is known to model the linear noise.
Dividing by Xt yields
dXt
= α dWt .
Xt
Switch to the integral form Z Z
dXt
= α dWt ,
Xt
and integrate “blindly” to get ln Xt = αWt + c, with c an integration constant. This leads to
the “pseudo-solution”
Xt = eαWt +c .
The nomination “pseudo” stands for the fact that Xt does not satisfy the initial equation. We
shall find a correct solution by letting the parameter c be a function of t. In other words, we
are looking for a solution of the following type:
Xt = eαWt +c(t) , (7.7.18)
where the function c(t) is subject to be determined. Using Ito’s formula we get
dXt = d(eWt +c(t) ) = eαWt +c(t) (c′ (t) + α2 /2)dt + αeαWt +c(t) dWt
= Xt (c′ (t) + α2 /2)dt + αXt dWt .
142

Substituting the last term from the initial equation (7.7.17) yields

dXt = Xt (c′ (t) + α2 /2)dt + dXt ,

which leads to the equation

c′ (t) + α2 /2 = 0
2
with the solution c(t) = − α2 t + k. Substituting into (7.7.18) yields
α2
Xt = eαWt − 2 t+k
.

The value of the constant k is determined by taking t = 0. This leads to X0 = ek . Hence we

have obtained the solution of the equation (7.7.17)
α2
Xt = X0 eαWt − 2 t
.

Example 7.7.1 Use the method of variation of parameters to solve the equation

dXt = Xt Wt dWt .

Dividing by Xt and converting the differential equation into the equivalent integral form, we
get Z Z
1
dXt = Wt dWt .
Xt
The right side is a well-known stochastic integral given by
Z
W2 t
Wt dWt = t − + C.
2 2
The left side will be integrated “blindly” according to the rules of elementary Calculus
Z
1
dXt = ln Xt + C.
Xt
Equating the last two relations and solving for Xt we obtain the “pseudo-solution”
Wt2
− 2t +c
Xt = e 2 ,

with c constant. In order to get a correct solution, we let c to depend on t and Wt . We shall
assume that c(t, Wt ) = a(t) + b(Wt ), so we are looking for a solution of the form
Wt2
− 2t +a(t)+b(Wt )
Xt = e 2 .

Applying Ito’s formula, we have

 1 1 

dXt = Xt − + a′ (t) + (1 + b′′ (Wt )) dt + Xt Wt + b′ (Wt ) dWt .
2 2
Subtracting the initial equation dXt = Xt Wt dWt yields
1

0 = Xt a′ (t) + b′′ (Wt ) dt + Xt b′ (Wt )dWt .
2
 143

This equation is satisfied if we are able to choose the functions a(t) and b(Wt ) such that the
coefficients of dt and dWt vanish

1
b′ (Wt ) = 0, a′ (t) + b′′ (Wt ) = 0.
2
From the first equation b must be a constant. Substituting into the second equation it follows
that a is also a constant. It turns out that the aforementioned “pseudo-solution” is in fact a
solution. The constant c = a + b is obtained letting t = 0. Hence the solution is given by
Wt2
− 2t
Xt = X0 e 2 .

Example 7.7.2 Use the method of variation of parameters to solve the stochastic differential
equation
dXt = µXt dt + σXt dWt ,

with µ and σ constants.

After dividing by Xt we bring the equation into the equivalent integral form
Z Z Z
dXt
= µ dt + σ dWt .
Xt

Integrate on the left “blindly” and get

ln Xt = µt + σWt + c,

where c is an integration constant. We arrive at the following “pseudo-solution”

Xt = eµt+σWt +c .

Assume the constant c is replaced by a function c(t), so we are looking for a solution of the
form
Xt = eµt+σWt +c(t) . (7.7.19)

Apply Ito’s formula and get

σ2

dXt = Xt µ + c′ (t) + dt + σXt dWt .
2
Subtracting the initial equation yields

σ2

c′ (t) + dt = 0,
2
2 2
which is satisfied for c′ (t) = − σ2 , with the solution c(t) = − σ2 t + k, k ∈ R. Substituting into
(7.7.19) yields the solution

σ2 σ2 σ2
Xt = eµt+σWt − 2 t+k
= e(µ− 2 )t+σWt +k
= X0 e(µ− 2 )t+σWt
.
144

7.8 Integrating Factors

The method of integrating factors can be applied to a class of stochastic differential equations
of the type
dXt = f (t, Xt )dt + g(t)Xt dWt , (7.8.20)
where f and g are continuous deterministic functions. The integrating factor is given by
Rt Rt
g(s) dWs + 21 g2 (s) ds
ρt = e − 0 0 .
The equation can be brought into the following exact form
d(ρt Xt ) = ρt f (t, Xt )dt.
Substituting Yt = ρt Xt , we obtain that Yt satisfies the deterministic differential equation
dYt = ρt f (t, Yt /ρt )dt,
which can be solved by either integration or as an exact equation. We shall exemplify the
method of integrating factors with a few examples.
Example 7.8.1 Solve the stochastic differential equation
dXt = rdt + αXt dWt , (7.8.21)
with r and α constants.
1 2
The integrating factor is given by ρt = e 2 α t−αWt . Using Ito’s formula, we can easily check
that
dρt = ρt (α2 dt − αdWt ).
Using dt2 = dt dWt = 0, (dWt )2 = dt we obtain
dXt dρt = −α2 ρt Xt dt.
Multiplying by ρt , the initial equation becomes
ρt dXt − αρt Xt dWt = rρt dt,
and adding and subtracting α2 ρt Xt dt from the left side yields
ρt dXt − αρt Xt dWt + α2 ρt Xt dt − α2 ρt Xt dt = rρt dt.
This can be written as
ρt dXt + Xt dρt + dρt dXt = rρt dt,
which, with the virtue of the product rule, becomes
d(ρt Xt ) = rρt dt.
Integrating yields Z t
ρt X t = ρ0 X 0 + r ρs ds
0
and hence the solution is
Z t
1 r
Xt = X0 + ρs ds
ρt ρt 0
Z t
1 2 1 2
= X0 eαWt − 2 α t + r e− 2 α (t−s)+α(Wt −Ws ) ds.
0
 145

Exercise 7.8.1 Let α be a constant. Solve the following stochastic differential equations by the
method of integrating factors
(a) dXt = αXt dWt ;
(b) dXt = Xt dt + αXt dWt ;
1
(c) dXt = dt + αXt dWt , X0 > 0.
Xt

Exercise 7.8.2 Let X R tt be the solution of the stochastic equation dXt = σXt dWt , with σ
constant. Let At = 1t 0 Xs dWs be the stochastic average of Xt . Find the stochastic equation
satisfied by At , the mean and variance of At .

7.9 Existence and Uniqueness

An exploding solution
Consider the non-linear stochastic differential equation

dXt = Xt3 dt + Xt2 dWt , X0 = 1/a. (7.9.22)

We shall look for a solution of the type Xt = f (Wt ). Ito’s formula yields

1
dXt = f ′ (Wt )dWt + f ′′ (Wt )dt.
2
Equating the coefficients of dt and dWt in the last two equations yields

f ′ (Wt ) = Xt2 =⇒ f ′ (Wt ) = f (Wt )2 (7.9.23)

1 ′′
f (Wt ) = Xt3 =⇒ f ′′ (Wt ) = 2f (Wt )3 (7.9.24)
2
We note that equation (7.9.23) implies (7.9.24) by differentiation. So it suffices to solve only
the ordinary differential equation

f ′ (x) = f (x)2 , f (0) = 1/a.

Separating and integrating we have

Z Z
df 1
2
= ds =⇒ f (x) = .
f (x) a−x

Hence a solution of equation (7.9.22) is

1
Xt = .
a − Wt

Let Ta be the first time the Brownian motion Wt hits a. Then the process Xt is defined only
for 0 ≤ t < Ta . Ta is a random variable with P (Ta < ∞) = 1 and E[Ta ] = ∞, see section 3.3.
The following theorem is the analog of Picard’s uniqueness result from ordinary differential
equations:
146

Theorem 7.9.1 (Existence and Uniqueness) Consider the stochastic differential equation

dXt = b(t, Xt )dt + σ(t, Xt )dWt , X0 = c

where c is a constant and b and σ are continuous functions on [0, T ] × R satisfying

1. |b(t, x)| + |σ(t, x)| ≤ C(1 + |x|); x ∈ R, t ∈ [0, T ]
2. |b(t, x) − b(t, y)| + |σ(t, x) − σ(t, y)| ≤ K|x − y|, x, y ∈ R, t ∈ [0, T ]
with C, K positive constants. There there is a unique solution process Xt that is continuous
and satisfies
hZ T i
E Xt2 dt < ∞.
0

The first condition says that the drift and volatility increase no faster than a linear function
in x. The second conditions states that the functions are Lipschitz in the second argument.

Example 7.9.1 Consider the stochastic differential equation

q 1 
q
dXt = 1 + Xt2 + Xt dt + 1 + Xt2 dWt , X0 = x0 .
2
(a) Solve the equation;
(b) Show that there a unique solution.
Chapter 8

Applications of Brownian Motion

8.1 The Generator of an Ito Diffusion

Let (Xt )t≥0 be a stochastic process with X0 = x0 . In this section we shall deal with the operator
associated with Xt . This is an operator which describes infinitesimally the rate of change of a
function which depends smoothly on Xt .
More precisely, the generator of the stochastic process Xt is the second order partial differ-
ential operator A defined by
E[f (Xt )] − f (x)
Af (x) = lim ,
tց0 t
for any smooth function (at least of class C 2 ) with compact support f : Rn → R. Here E
stands for the expectation operator taken at t = 0, i.e.,
Z
E[f (Xt )] = f (y)pt (x, y) dy,
Rn

where pt (x, y) is the transition density of Xt .

In the following we shall find the generator associated with the Ito diffusion

dXt = b(Xt )dt + σ(Xt )dW (t), t ≥ 0, X0 = x, (8.1.1)

where W (t) = W1 (t), . . . , Wm (t) is an m-dimensional Brownian motion, b : Rn → Rn and
σ : Rn → Rn×m are measurable functions.
The main tool used in deriving the formula for the generator A is Ito’s formula in several
variables. If let Ft = f (Xt ), then using Ito’s formula we have
X ∂f 1 X ∂2f
dFt = (Xt ) dXti + (Xt ) dXti dXtj , (8.1.2)
i
∂xi 2 i,j
∂xi ∂xj

where Xt = (Xt1 , · · · , Xtn ) satisfies the Ito diffusion (8.1.1) on components, i.e.,

dXti = bi (Xt )dt + [σ(Xt ) dW (t)]i

X
= bi (Xt )dt + σik dWk (t). (8.1.3)
k

147
148

Using the stochastic relations dt2 = dt dWk (t) = 0 and dWk (t) dWr (t) = δkr dt, a computation
provides

 X  X 
dXti dXtj = bi dt + σik dWk (t) bj dt + σjk dWk (t)
k k
X  X 
= σik dWk (t) σjr dWr (t)
k r
X X
= σik σjr dWk (t)dWr (t) = σik σjk dt
k,r k

= (σσ T )ij dt.

Therefore
dXti dXtj = (σσ T )ij dt. (8.1.4)
Substitute (8.1.3) and (8.1.4) into (8.1.2) yields

" #
1 X ∂2f T
X ∂f
dFt = (Xt )(σσ )ij + bi (Xt ) (Xt ) dt
2 i,j ∂xi ∂xj i
∂xi
X ∂f
+ (Xt )σik (Xt ) dWk (t).
∂xi
i,k

Integrate and obtain

Z t" X #
1 ∂2f T
X ∂f
Ft = F0 + (σσ )ij + bi (Xs ) ds
0 2 i,j ∂xi ∂xj i
∂xi
XZ t X ∂f
+ σik (Xs ) dWk (s).
0 i
∂xi
k

Since F0 = f (X0 ) = f (x) and E(f (x)) = f (x), applying the expectation operator in the
previous relation we obtain
"Z #
t X
1 T ∂2f X ∂f 
E(Ft ) = f (x) + E (σσ )ij + bi (Xs ) ds . (8.1.5)
0 2 i,j ∂xi ∂xj i
∂xi
Rt
Using the commutation between the operator E and the integral 0 yields

E(Ft ) − f (x) 1X ∂ 2 f (x) X ∂f (x)

lim = (σσ T )ij + bk .
tց0 t 2 i,j ∂xi ∂xj ∂xk
k

We conclude the previous computations with the following result.

Theorem 8.1.1 The generator of the Ito diffusion (8.1.1) is given by
1X ∂2 X ∂
A= (σσ T )ij + bk . (8.1.6)
2 i,j ∂xi ∂xj ∂xk
k
 149

The matrix σ is called dispersion and the product σσ T is called diffusion matrix. These
names are related with their physical significance Substituting (8.1.6) in (8.1.5) we obtain the
following formula
hZ t i
E[f (Xt )] = f (x) + E Af (Xs ) ds , (8.1.7)
0

for any f ∈ C02 (Rn ).

Exercise 8.1.2 Find the generator operator associated with the n-dimensional Brownian mo-
tion.

Exercise 8.1.3 Find the Ito diffusion corresponding to the generator Af (x) = f ′′ (x) + f ′ (x).

Exercise 8.1.4 Let ∆G = 12 (∂x21 + x21 ∂x22 ) be the Grushin’s operator.

(a) Find the diffusion process associated with the generator ∆G .
(b) Find the diffusion and dispersion matrices and show that they are degenerate.

8.2 Dynkin’s Formula

Formula (8.1.7) holds under more general conditions, when t is a stopping time. First we need
the following result, which deals with a continuity-type property in the upper limit of a Ito
integral.

Lemma 8.2.1 Let g be a bounded measurable function and τ be a stopping time for Xt with
E[τ ] < ∞. Then
h Z τ ∧k i hZ τ i
lim E g(Xs ) dWs = E g(Xs ) dWs ; (8.2.8)
k→∞ 0 0

hZ τ ∧k i hZ τ i
lim E g(Xs ) ds = E g(Xs ) ds . (8.2.9)
k→∞ 0 0

Proof: Let |g| < K. Using the properties of Ito integrals, we have
h Z τ Z τ ∧k 2 i h Z τ 2 i
E g(Xs ) dWs − g(Xt ) dWs =E g(Xs ) dWs
0 0 τ ∧k
hZ τ i
= E g 2 (Xs ) ds ≤ K 2 E[τ − τ ∧ k] → 0, k → ∞.
τ ∧k

Since E[X 2 ] ≤ E[X]2 , it follows that

hZ τ Z τ ∧k i
E g(Xs ) dWs − g(Xt ) dWs → 0, k → ∞,
0 0

which is equivalent with relation (8.2.8).

The second relation can be proved similar and is left as an exercise for the reader.
150

Exercise 8.2.2 Assume the hypothesis of the previous lemma. Let 1{s<τ } be the characteristic
function of the interval (−∞, τ )

1, if u < τ
1{s<τ } (u) =
0, otherwise.

Show that
Z τ ∧k Z k
(a) g(Xs ) dWs = 1{s<τ } g(Xs ) dWs ,
0 0
Z τ ∧k Z k
(b) g(Xs ) ds = 1{s<τ } g(Xs ) ds.
0 0

Theorem 8.2.3 (Dynkin’s formula) Let f ∈ C02 (Rn ), and Xt be an Ito diffusion starting at
x. If τ is a stopping time with E[τ ] < ∞, then
hZ τ i
E[f (Xτ )] = f (x) + E Af (Xs ) ds , (8.2.10)
0

where A is the infinitesimal generator of Xt .

Proof: Replace t by k and f by 1{s<τ } f in (8.1.7) and obtain

hZ k i
E[1s<τ f (Xk )] = 1{s<τ } f (x) + E A(1{s<τ } f )(Xs ) ds ,
0

which can be written as

hZ k i
E[f (Xk∧τ )] = 1{s<τ } f (x) + E 1{s<τ } (s)A(f )(Xs ) ds
0
hZ k∧τ i
= 1{s<τ } f (x) + E A(f )(Xs ) ds . (8.2.11)
0

Since by Lemma 8.2.1 (b)

E[f (Xk∧τ )] → E[f (Xτ )], k→∞

hZ k∧τ i hZ τ i
E A(f )(Xs ) ds → E A(f )(Xs ) ds , k → ∞,
0 0

using Exercise (8.2.2) relation (8.2.11) yields (8.2.10).

8.3 Applications
In this section we shall present a few important results of stochastic calculus which can be
obtained as a consequence of Dynkin’s formula.
 151

8.3.1 Kolmogorov’s Backward Equation

For any function f ∈ C02 (Rn ) let v(t, x) = E[f (Xt )], given that X0 = x. As usual, E denotes
the expectation at time t = 0. Then v(0, x) = f (x), and differentiating in Dynkin’s formula
(8.1.7)
Z t
v(t, x) = f (x) + E[Af (Xs )] ds
0
yields
∂v
= E[Af (Xt )] = AE[f (Xt )] = Av(t, x).
∂t
We arrived at the following result.

Theorem 8.3.1 (Kolmogorov’s backward equation) For any f ∈ C02 (Rn ) the function
v(t, x) = E[f (Xt )] satisfies the following Cauchy’s problem
∂v
= Av, t>0
∂t
v(0, x) = f (x),

where A denotes the generator of the Ito’s diffusion (8.1.1).

8.3.2 Exit Time from an Interval

Let Xt = x0 + Wt e a one-dimensional Brownian motion starting at x0 . Consider the exit time
of the process Xt from the strip (a, b)

τ = inf{t > 0; Xt 6∈ (a, b)}.

Assuming E[τ ] < 0, applying Dynkin’s formula yields

h i h Z τ 1 d2 i
E f (Xτ ) = f (x0 ) + E 2
f (X s ) ds . (8.3.12)
0 2 dx

Choosing f (x) = x in (8.3.12) we obtain

E[Xτ ] = x0 (8.3.13)

Exercise 8.3.2 Prove relation (8.3.13) using the Optional Stopping Theorem for the martin-
gale Xt .
Let pa = P (Xτ = a) and pb = P (Xτ = b) be the exit probabilities from the interval (a, b).
Obviously, pa + pb = 1, since the probability that the Brownian motion will stay for ever inside
the bounded interval is zero. Using the expectation definition, relation (8.3.13) yields

apa + b(1 − pa ) = x0 .

Solving for pa and pb we get the following exit probabilities

b − x0
pa = (8.3.14)
b−a
x0 − a
pb = 1 − pa = . (8.3.15)
b−a
152

Figure 8.1: The Brownian motion Xt in the ball B(0, R).

It is worth noting that if b → ∞ then pa → 1 and if a → −∞ then pb → 1. This can be stated

by saying that a Brownian motion starting at x0 reaches any level (below or above x0 ) with
probability 1.
Next we shall compute the mean of the exit time, E[τ ]. Choosing f (x) = x2 in (8.3.12)
yields
E[(Xτ )2 ] = x20 + E[τ ].
From the definition of the mean and formulas (8.3.14)-(8.3.15) we obtain

b − x0 x0 − a
E[τ ] = a2 pa + b2 pb − x20 = a2 + b2 − x0
b−a b−a
1 h i
= ba2 − ab2 + x0 (b − a)(b + a) − x20
b−a
= −ab + x0 (b + a) − x20
= (b − x0 )(x0 − a). (8.3.16)

Exercise 8.3.3 (a) Show that the equation x2 − (b − a)x + E[τ ] = 0 cannot have complex roots;

(b − a)2
(b) Prove that E[τ ] ≤ ;
4
(c) Find the point x0 ∈ (a, b) such that the expectation of the exit time, E[τ ], is maximum.

8.3.3 Transience and Recurrence of Brownian Motion

We shall consider first the expectation of the exit time from a ball. Then we shall extend it to
an annulus and compute the transience probabilities.


1. Consider the process Xt = a + W (t), where W (t) = W1 (t), . . . , Wn (t) is an n-dimensional
Brownian motion, and a = (a1 , . . . , an ) ∈ Rn is a fixed vector, see Fig. 8.1. Let R > 0 such
 153

that R > |a|. Consider the exit time of the process Xt from the ball B(0, R)

τ = inf{t > 0; |Xt | > R}. (8.3.17)

Assuming E[τ ] < ∞ and letting f (x) = |x|2 = x21 + · · · + x2n in Dynkin’s formula
  hZ τ 1 i
E f (Xτ ) = f (x) + E ∆f (Xs ) ds
0 2

yields
hZ τ i
R2 = |a|2 + E n ds ,
0

and hence
R2 − |a|2
E[τ ] = . (8.3.18)
n
In particular, if the Brownian motion starts from the center, i.e. a = 0, the expectation of the
exit time is
R2
E[τ ] = .
n
(i) Since R2 /2 > R2 /3, the previous relation implies that it takes longer for a Brownian motion
to exit a disk of radius R rather than a ball of the same radius.
(ii) The probability that a Brownian motion leaves the interval (−R, R) is twice the probability
that a 2-dimensional Brownian motion exits the disk B(0, R).

Exercise 8.3.4 Prove that E[τ ] < ∞, where τ is given by (8.3.17).

Exercise 8.3.5 Apply the Optional Stopping Theorem for the martingale Wt = Wt2 − t to show
that E[τ ] = R2 , where
τ = inf{t > 0; |Wt | > R}
is the first exit time of the Brownian motion from (−R, R).

2. Let b ∈ Rn such that b ∈

/ B(0, R), i.e. |b| > R, and consider the annulus

Ak = {x; R < |x| < kR}

where k > 0 such that b ∈ Ak . Consider the process Xt = b + W (t) and let

τk = inf{t > 0; Xt ∈
/ Ak }

be the first exit time of Xt from the annulus Ak . Let f : Ak → R be defined by

(
− ln |x|, if n = 2
f (x) = 1
, if n > 2.
|x|n−2

A straightforward computation shows that ∆f = 0. Substituting into Dynkin’s formula

  h Z τk  1  i
E f (Xτk ) = f (b) + E ∆f (Xs ) ds
0 2
154

yields  
E f (Xτk ) = f (b). (8.3.19)
This can be stated by saying that the value of f at a point b in the annulus is equal to the
expected value of f at the first exit time of a Brownian motion starting at b.
Since |Xτk | is a random variable with two outcomes, we have
 
E f (Xτk ) = pk f (R) + qk f (kR),

where pk = P (|Xτk | = R), qk = P (|XXτk |) = kR and pk + qk = 1. Substituting in (8.3.19)

yields
pk f (R) + qk f (kR) = f (b). (8.3.20)
There are the following two distinguished cases:
(i) If n = 2 we obtain
−pk ln R − qk (ln k + ln R) = − ln b.
Using pk = 1 − qk , solving for pk yields

ln( Rb )
pk = 1 − .
ln k
Hence
P (τ < ∞) = lim pk = 1,
k→∞

where τ = inf{t > 0; |Xt | < R} is the first time Xt hits the ball B(0, R). Hence in R2 a
Brownian motion hits with probability 1 any ball. This is stated equivalently by saying that
the Brownian motion is recurrent in R2 .
(ii) If n > 2 the equation (8.3.20) becomes

pk qk 1
+ n−2 n−2 = n−2 .
Rn−2 k R b
Taking the limit k → ∞ yields
 R n−2
lim pk = < 1.
k→∞ b
Then in Rn , n > 2, a Brownian motion starting outside of a ball hits it with a probability less
than 1. This is usually stated by saying that the Brownian motion is transient.
3. We shall recover the previous results using the n-dimensional Bessel process

p
Rt = dist 0, W (t) = W1 (t)2 + · · · + Wn (t)2 .

Consider the process Yt = α + Rt , with 0 ≤ α < R. The generator of Yt is the Bessel operator
of order n
1 d2 n−1 d
A= + ,
2 dx2 2x dx
see section 2.7. Consider the exit time

τ = {t > 0; Yt > R}.

 155

Figure 8.2: The Brownian motion Xt in the annulus Ar,R .

Applying Dynkin’s formula

h Z τ i
 
E f (Yτ ) = f (Y0 ) + E (Af )(Ys ) ds
0
Rτ 
for f (x) = x2 yields R2 = α2 + E 0 nds . This leads to

R2 − α2
E[τ ] = .
n
which recovers (8.3.18) with α = |a|.
In the following assume n ≥ 3 and consider the annulus

Ar,R = {x ∈ Rn ; r < |x| < R}.

Consider the stopping time τ = inf{t > 0; Xt ∈ / Ar,R } = inf{t > 0;Yt ∈ /  (r, R)}, where
|Y0 | = α ∈ (r, R). Applying Dynkin’s formula for f (x) = x2−n yields E f (Yτ = f (α). This
can be written as
pr r2−n + pR R2−n = α2−n ,
where
pr = P (|Xt | = r), pR = P (|Xt | = R), pr + pR = 1.
Solving for pr and pR yields

R n−2


r n−2
α −1 α −1
pr =
, pR =
.
R n−2 r n−2
r −1 R −1
The transience probability is obtained by taking the limit to infinity
α2−n Rn−2 − 1  r n−2
pr = lim pr,R = lim = ,
R→∞ R→∞ r2−n Rn−2 − 1 α
156

where pr is the probability that a Brownian motion starting outside the ball of radius r will hit
the ball, see Fig. 8.2.

Exercise 8.3.6 Solve the equation 12 f ′′ (x)+ n−1 ′

2x f (x) = 0 by looking for a solution of monomial
k
type f (x) = x .

8.4 Application to Solving Parabolic Equations

This section deals with solving first and second order parabolic equations using the integral of
the cost function along a certain characteristic solution. The first order equations are related
to predictable characteristic curves, while the second order equations depend on stochastic
characteristic curves.
1. Predictable characteristics Let ϕ(s) be the solution of the following one-dimensional ODE

dX(s)

= a s, X(s) , t≤s≤T
ds
X(t) = x,

and define the cumulative cost between t and T along the solution ϕ
Z T


u(t, x) = c s, ϕ(s) ds, (8.4.21)
t

where c denotes a continuous cost function. Differentiate both sides with respect to t
Z T
∂
∂

u t, ϕ(t) = c s, ϕ(s) ds
∂t ∂t t


∂t u + ∂x u ϕ′ (t) = −c t, ϕ(t) .

Hence (8.4.21) is a solution of the following final value problem

∂t u(t, x) + a(t, x)∂x u(t, x) = −c(t, x)

u(T, x) = 0.

It is worth mentioning that this is a variant of the method of characteristics.1 The curve given
by the solution ϕ(s) is called a characteristic curve.

Exercise 8.4.1 Using the previous method solve the following final boundary problems:
(a)

∂t u + x∂x u = −x
u(T, x) = 0.

(b)

∂t u + tx∂x u = ln x, x>0
u(T, x) = 0.
1 This is a well known method of solving linear partial differential equations.
 157

2. Stochastic characteristics Consider the diffusion

dXs = a(s, Xs )ds + b(s, Xs )dWs , t≤s≤T

Xt = x,

and define the stochastic cumulative cost function

Z T
u(t, Xt ) = c(s, Xs ) ds, (8.4.22)
t

with the conditional expectation

h i
u(t, x) = E u(t, Xt )|Xt = x
hZ T i
= E c(s, Xs ) ds|Xt = x .
t

Taking increments in both sides of (8.4.22) yields

Z T
du(t, Xt ) = d c(s, Xs ) ds.
t

Applying Ito’s formula on one side and the Fundamental Theorem of Calculus on the other, we
obtain
1
∂t u(t, x)dt + ∂x u(t, Xt )dXt + ∂x2 u(t, t, Xt )dXt2 = −c(t, Xt )dt.
2
Taking the expectation E[ · |Xt = x] on both sides yields
1
∂t u(t, x)dt + ∂x u(t, x)a(t, x)dt + ∂x2 u(t, x)b2 (t, x)dt = −c(t, x)dt.
2
Hence, the expected cost
hZ T i
u(t, x) = E c(s, Xs ) ds|Xt = x
t

is a solution of the following second order parabolic equation

1
∂t u + a(t, x)∂x u + b2 (t, x)∂x2 u(t, x) = −c(t, x)
2
u(T, x) = 0.

This represents the probabilistic interpretation of the solution of a parabolic equation.

Exercise 8.4.2 Solve the following final boundary problems:

(a)

1
∂t u + ∂x u + ∂x2 u = −x
2
u(T, x) = 0.
158

(b)
1
∂t u + ∂x u + ∂x2 u = ex ,
2
u(T, x) = 0.

(c)
1
∂t u + µx∂x u + σ 2 x2 ∂x2 u = −x,
2
u(T, x) = 0.
Chapter 9

Martingales

9.1 Examples of Martingales

In this section we shall use the knowledge acquired in previous chapters to present a few
important examples of martingales. These will be useful in the proof of Girsanov’s theorem in
the next section.
We start by recalling that a process Xt , 0 ≤ t ≤ T , is an Ft -martingale if
1. E[|Xt |] < ∞ (Xt integrable for each t);
2. Xt is Ft -predictable;
3. the forecast of future values is the last observation: E[Xt |Fs ] = Xs , ∀s < t.
We shall present three important examples of martingales and some of their particular cases.

Example 9.1.1 If v(s) is a continuous function on [0, T ], then

Z t
Xt = v(s) dWs
0

is an Ft -martingale.

Taking out the predictable part

hZ s Z t i
E[Xt |Fs ] = E v(τ ) dWτ + v(τ ) dWτ Fs
0 s
hZ t i
= Xs + E v(τ ) dWτ Fs = Xs ,
s

Rt
where we used that s v(τ ) dWτ is independent of Fs and the conditional expectation equals
the usual expectation
hZ t i hZ t i
E v(τ ) dWτ Fs = E v(τ ) dWτ = 0.
s s

159
160
Z t
Example 9.1.2 Let Xt = v(s) dWs be a process as in Example 9.1.1. Then
0

Z t
Mt = Xt2 − v 2 (s) ds
0

is an Ft -martingale.

The process Xt satisfies the stochastic equation dXt = v(t)dWt . By Ito’s formula

d(Xt2 ) = 2Xt dXt + (dXt )2 = 2v(t)Xt dWt + v 2 (t)dt. (9.1.1)

Integrating between s and t yields

Z t Z t
Xt2 − Xs2 = 2 Xτ v(τ ) dWτ + v 2 (τ ) dτ.
s s

Then separating the predictable from the unpredictable we have

h Z t i
E[Mt |Fs ] = E Xt2 − v 2 (τ ) dτ Fs
0
h Z t Z s i
2 2 2 2
= E Xt − Xs − v (τ ) dτ + Xs − v 2 (τ ) dτ Fs
s 0
Z s h Z t i
= Xs2 − v 2 (τ ) dτ + E Xt2 − Xs2 − v 2 (τ ) dτ Fs
0 s
hZ t i
= Ms + 2E Xτ v(τ ) dWτ Fs = Ms ,
s
Z t
where we used relation (9.1.1) and that Xτ v(τ ) dWτ is totally unpredictable given the
s
information set Fs . In the following we shall mention a few particular cases.
1. If v(s) = 1, then Xt = Wt . In this case Mt = Wt2 − t is an Ft -martingale.
Rt
2. If v(s) = s, then Xt = 0 s dWs , and hence
Z t 2 t3
Mt = s dWs −
0 3

is an Ft -martingale.

Example 9.1.3 Let u : [0, T ] → R be a continuous function. Then

Rt Rt
u(s) dWs − 21 u2 (s) ds
Mt = e 0 0

is an Ft -martingale for 0 ≤ t ≤ T .
Z t Z t
1
Consider the process Ut = u(s) dWs − u2 (s) ds. Then
0 2 0
 161

1
dUt = u(t)dWt − u2 (t)dt
2
(dUt )2 = u(t)dt.

Then Ito’s formula yields

1
dMt = d(eUt ) = eUt dUt + eUt (dUt )2
2
 1 2 1 
Ut
= e u(t)dWt − u (t)dt + u2 (t)dt
2 2
= u(t)Mt dWt .

Integrating between s and t yields

Z t
Mt = Ms + u(τ )Mτ dWτ
s
Z t
Since u(τ )Mτ dWτ is independent of Fs , then
s

hZ t i 
Z t 
E u(τ )Mτ dWτ |Fs = E u(τ )Mτ dWτ = 0,
s s

and hence Z t
E[Mt |Fs ] = E[Ms + u(τ )Mτ dWτ |Fs ] = Ms .
s

Remark 9.1.4 The condition that u(s) is continuous on [0, T ] can be relaxed by asking only
Z t
2
u ∈ L [0, T ] = {u : [0, T ] → R; measurable and |u(s)|2 ds < ∞}.
0

It is worth noting that the conclusion still holds if the function u(s) is replaced by a stochastic
process u(t, ω) satisfying Novikov’s condition
 1 RT 2 
E e 2 0 u (s,ω) ds < ∞.

The previous process has a distinguished importance in the theory of martingales and it will
be useful when proving the Girsanov theorem.

Definition 9.1.5 Let u ∈ L2 [0, T ] be a deterministic function. Then the stochastic process
Rt Rt
u(s) dWs − 21 u2 (s) ds
Mt = e 0 0

is called the exponential process induced by u.

Particular cases of exponential processes.

σ2
1. Let u(s) = σ, constant, then Mt = eσWt − 2 t is an Ft -martingale.
162

2. Let u(s) = s. Integrating in d(tWt ) = tdWt − Wt dt yields

Z t Z t
s dWs = tWt − Ws ds.
0 0
Rt
Let Zt = 0
Ws ds be the integrated Brownian motion. Then
Rt Rt
d dWs − 12 s2 ds
Mt = e 0 0

t3
= etWt − 6 −Zt

is an Ft -martingale.

Example 9.1.6 Let Xt be a solution of dXt = u(t)dt + dWt , with u(s) a bounded function.
Consider the exponential process
Rt Rt
u(s) dWs − 12 u2 (s) ds
Mt = e− 0 0 . (9.1.2)

Then Yt = Mt Xt is an Ft -martingale.

In Example 9.1.3 we obtained dMt = −u(t)Mt dWt . Then

dMt dXt = −u(t)Mt dt.

The product rule yields

dYt = Mt dXt + Xt dMt + dMt dXt

= Mt u(t)dt + dWt − Xt u(t)Mt dWt − u(t)Mt dt


= Mt 1 − u(t)Xt dWt .

Integrating between s and t yields

Z t

Yt = Ys + Mτ 1 − u(τ )Xτ dWτ .
s
Rt

Since s
Mτ 1 − u(τ )Xτ dWτ is independent of Fs ,
hZ t
i h Z t
i
E Mτ 1 − u(τ )Xτ dWτ |Fs = E Mτ 1 − u(τ )Xτ dWτ = 0,
s s

and hence
E[Yt |Fs ] = Ys .
1
Exercise 9.1.7 Prove that (Wt + t)e−Wt − 2 t is an Ft -martingale.

Exercise 9.1.8 Let h be a continuous function. Using the properties of the Wiener integral
and log-normal random variables, show that
h Rt i 1
Rt 2
E e 0 h(s) dWs = e 2 0 h(s) ds .
 163

Exercise 9.1.9 Let Mt be the exponential process (9.1.2). Use the previous exercise to show
that for any t > 0 Rt 2
(a) E[Mt ] = 1 (b) E[Mt2 ] = e 0 u(s) ds .
Exercise 9.1.10 Let Ft = σ{Wu ; u ≤ t}. Show that the following processes are Ft -martingales:
(a) et/2 cos Wt ;
(b) et/2 sin Wt .
Pn
Recall that the Laplacian of a twice differentiable function f is defined by ∆f (x) = j=1 ∂x2j f .
Example 9.1.11 Consider the smooth function f : Rn → R, such that
(i) ∆f = 0;
 
(ii) E |f (Wt )| < ∞, ∀t > 0 and x ∈ R.
Then the process Xt = f (Wt ) is an Ft -martingale.
Proof: It follows from the more general Example 9.1.13.
Exercise 9.1.12 Let W1 (t) and W2 (t) be two independent Brownian motions. Show that Xt =
eW1 (t) cos W2 (t) is a martingale.
Example 9.1.13 Let f : Rn → R be a smooth function such that
 
(i) E |f (Wt )| < ∞;
hR i
t
(ii) E 0 |∆f (Ws )| ds < ∞.
Rt
Then the process Xt = f (Wt ) − 21 0 ∆f (Ws ) ds is a martingale.
Proof: For 0 ≤ s < t we have
    h1 Z t i
E Xt |Fs = E f (Wt )|Fs − E ∆f (Wu ) du|Fs
2 0
Z Z t h
  1 s 1 i
= E f (Wt )|Fs − ∆f (Wu ) du − E ∆f (Wu )|Fs du (9.1.3)
2 0 s 2
Let p(t, y, x) be the probability density function of Wt . Integrating by parts and using that p
satisfies the Kolmogorov’s backward equation, we have
h1 i Z Z
1 1
E ∆f (Wu )|Fs = p(u − s, Ws , x) ∆f (x) dx = ∆x p(u − s, Ws , x)f (x) dx
2 2 2
Z
∂
= p(u − s, Ws , x)f (x) dx.
∂u
Then, using the Fundamental Theorem of Calculus, we obtain
Z t h i Z t Z 
1 ∂
E ∆f (Wu )|Fs du = p(u − s, Ws , x)f (x) dx du
s 2 ∂u
Zs Z
= p(t − s, Ws , x)f (x) dx − lim p(ǫ, Ws , x)f (x) dx
ǫց0
Z
 
= E f (Wt )|Fs − δ(x = Ws )f (x) dx
 
= E f (Wt )|Fs − f (Ws ).
164

Substituting in (9.1.3) yields

Z
    1 s  
E Xt |Fs = E f (Wt )|Fs − ∆f (Wu ) du − E f (Wt )|Fs + f (Ws ) (9.1.4)
2 0
Z
1 s
= f (Ws ) − ∆f (Wu ) du (9.1.5)
2 0
= Xs . (9.1.6)

Hence Xt is an Ft -martingale.

Exercise 9.1.14 Use the Example 9.1.13 to show that the following processes are martingales:
(a) Xt = Wt2 − t;
Rt
(b) Xt = Wt3 − 3 0 Ws ds;
1
Rt
(c) Xt = n(n−1) Wtn − 12 0 Wsn−2 ds;
Rt
(d) Xt = ecWt − 12 c2 0 eWs ds, with c constant;
Rt
(e) Xt = sin(cWt ) + 12 c2 0 sin(cWs ) ds, with c constant.

Exercise 9.1.15 Let f : Rn → R be a function such that

 
(i) E |f (Wt )| < ∞;
(ii) ∆f = λf , λ constant.
λ
Rt
Show that the process Xt = f (Wt ) − 2 0 f (Ws ) ds is a martingale.

9.2 Girsanov’s Theorem

In this section we shall present and prove a particular version of Girsanov’s theorem, which will
suffice for the purpose of later applications. The application of Girsanov’s theorem to finance is
to show that in the study of security markets the differences between the mean rates of return
can be removed.
We shall recall first a few basic notions. Let (Ω, F , P ) be a probability space. When dealing
with an Ft -martingale on the aforementioned probability space, the filtration Ft is considered to
be the sigma-algebra generated by the given Brownian motion Wt , i.e. Ft = σ{Wu ; 0 ≤ u ≤ s}.
By default, a martingale is considered with respect to the probability measure P , in the sense
that the expectations involve an integration with respect to P
Z
E P [X] = X(ω)dP (ω).
Ω

We have not used the upper script until now since there was no doubt which probability measure
was used. In this section we shall use also another probability measure given by

dQ = MT dP,

where MT is an exponential process. This means that Q : F → R is given by

Z Z
Q(A) = dQ = MT dP, ∀A ∈ F .
A A
 165

Since MT > 0, M0 = 1, using the martingale property of Mt yields

Q(A) > 0, A 6= Ø;
Z
Q(Ω) = MT dP = E P [MT ] = E P [MT |F0 ] = M0 = 1,
Ω

which shows that Q is a probability on F , and hence (Ω, F , Q) becomes a probability space.
The following transformation of expectation from the probability measure Q to P will be useful
in part II. If X is a random variable
Z Z
E Q [X] = X(ω) dQ(ω) = X(ω)MT (ω) dP (ω)
Ω Ω
= E P [XMT ].
The following result will play a central role in proving Girsanov’s theorem:
Lemma 9.2.1 Let Xt be the Ito process
dXt = u(t)dt + dWt , X0 = 0, 0 ≤ t ≤ T,
with u(s) a bounded function. Consider the exponential process
Rt Rt
u(s) dWs − 12 u2 (s) ds
Mt = e− 0 0 .
Then Xt is an Ft -martingale with respect to the measure
dQ(ω) = MT (ω)dP (ω).
Proof: We need to prove that Xt is an Ft -martingale with respect to Q, so it suffices to show
the following three properties:
1. Integrability of Xt . This part usually follows from standard manipulations of norms esti-
mations. We shall do it here in detail, but we shall omit it in other proofs. Integrating in the
equation of Xt between 0 and t yields
Z t
Xt = u(s) ds + Wt . (9.2.7)
0

We start with an estimation of the expectation with respect to P

h Z t 2 Z t i
E P [Xt2 ] = E P u(s) ds + 2 u(s) ds Wt + Wt2
0 0
Z t 2 Z t
= u(s) ds + 2 u(s) ds E P [Wt ] + E P [Wt2 ]
0 0
Z t 2
= u(s) ds + t < ∞, ∀0 ≤ t ≤ T
0

where the last inequality follows from the norm estimation

Z t Z t h Z t i1/2
u(s) ds ≤ |u(s)| ds ≤ t |u(s)|2 ds
0 0 0
h Z T i1/2
2
≤ t |u(s)| ds = T 1/2 kukL2 [0,T ] .
0
166

Next we obtain an estimation with respect to Q

Z 2 Z Z
E Q [|Xt |]2 = |Xt |MT dP ≤ |Xt |2 dP MT2 dP
Ω Ω Ω
= E P [Xt2 ] E P [MT2 ] < ∞,
RT
u(s)2 ds kuk2L2 [0,T ]
since E P [Xt2 ] < ∞ and E P [MT2 ] = e 0 =e , see Exercise 9.1.9.
2. Ft -predictability of Xt . This follows from equation (9.2.7) and the fact that Wt is Ft -
predictable.
3. Conditional expectation of Xt . From Examples 9.1.3 and 9.1.6 recall that for any 0 ≤ t ≤ T

Mt is an Ft -martingale with respect to probability measure P ;

Xt Mt is an Ft -martingale with respect to probability measure P.

We need to verify that

E Q [Xt |Fs ] = Xs , ∀s ≤ t.
which can be written as Z Z
Xt dQ = Xs dQ, ∀A ∈ Fs .
A A
Since dQ = MT dP , the previous relation becomes
Z Z
Xt MT dP = Xs MT dP, ∀A ∈ Fs .
A A

This can be written in terms of conditional expectation as

E P [Xt MT |Fs ] = E P [Xs MT |Fs ]. (9.2.8)

We shall prove this identity by showing that both terms are equal to Xs Ms . Since Xs is
Fs -predictable and Mt is a martingale, the right side term becomes

E P [Xs MT |Fs ] = Xs E P [MT |Fs ] = Xs Ms , ∀s ≤ T.

Let s < t. Using the tower property (see Proposition 1.11.4, part 3), the left side term becomes
   
E P [Xt MT |Fs ] = E P E P [Xt MT |Ft ]|Fs = E P Xt E P [MT |Ft ]|Fs
 
= E P Xt Mt |Fs = Xs Ms ,

where we used that Mt and Xt Mt are martingales and Xt is Ft -predictable. Hence (9.2.8) holds
and Xt is an Ft -martingale with respect to the probability measure Q.

Lemma 9.2.2 Consider the process

Z t
Xt = u(s) ds + Wt , 0 ≤ t ≤ T,
0

with u ∈ L2 [0, T ] a deterministic function, and let dQ = MT dP . Then

E Q [Xt2 ] = t.
 167
Rt
Proof: Denote U (t) = 0
u(s) ds. Then

E Q [Xt2 ] = E P [Xt2 MT ] = E P [U 2 (t)MT + 2U (t)Wt MT + Wt2 MT ]

= U 2 (t)E P [MT ] + 2U (t)E P [Wt MT ] + E P [Wt2 MT ]. (9.2.9)
P P
From Exercise 9.1.9 (a) we have E [MT ] = 1. In order to compute E [Wt MT ] we use the
tower property and the martingale property of Mt

E P [Wt MT ] = E[E P [Wt MT |Ft ]] = E[Wt E P [MT |Ft ]]

= E[Wt Mt ]. (9.2.10)
Using the product rule
d(Wt Mt ) = Mt dWt + Wt dMt + dWt dMt


= Mt − u(t)Mt Wt dWt − u(t)Mt dt,
where we used dMt = −u(t)Mt dWt . Integrating between 0 and t yields
Z t Z t


Wt Mt = Ms − u(s)Ms Ws dWs − u(s)Ms ds.
0 0

Taking the expectation and using the property of Ito integrals we have
Z t Z t
E[Wt Mt ] = − u(s)E[Ms ] ds = − u(s) ds = −U (t). (9.2.11)
0 0

Substituting into (9.2.10) yields

E P [Wt MT ] = −U (t). (9.2.12)
P
For computing E [Wt2 MT ] we proceed in a similar way
P
E [Wt2 MT ] = E[E P [Wt2 MT |Ft ]] = E[Wt2 E P [MT |Ft ]]
= E[Wt2 Mt ]. (9.2.13)
Using the product rule yields
d(Wt2 Mt ) = Mt d(Wt2 ) + Wt2 dMt + d(Wt2 )dMt


= Mt (2Wt dWt + dt) − Wt2 u(t)Mt dWt


−(2Wt dWt + dt) u(t)Mt dWt



= Mt Wt 2 − u(t)Wt dWt + Mt − 2u(t)Wt Mt dt.
Integrate between 0 and t
Z t Z t



Wt2 Mt = [Ms Ws 2 − u(s)Ws ] dWs + Ms − 2u(s)Ws Ms ds,
0 0

and take the expected value to get

Z t

E[Wt2 Mt ] = E[Ms ] − 2u(s)E[Ws Ms ] ds
0
Z t

= 1 + 2u(s)U (s) ds
0
= t + U 2 (t),
168

where we used (9.2.11). Substituting into (9.2.13) yields

E P [Wt2 MT ] = t + U 2 (t). (9.2.14)
Substituting (9.2.12) and (9.2.14) into relation (9.2.9) yields
E Q [Xt2 ] = U 2 (t) − 2U (t)2 + t + U 2 (t) = t, (9.2.15)
which ends the proof of the lemma.
Now we are prepared to prove one of the most important results of stochastic calculus.
Theorem 9.2.3 (Girsanov Theorem) Let u ∈ L2 [0, T ] be a deterministic function. Then
the process Z t
Xt = u(s) ds + Wt , 0≤t≤T
0
is a Brownian motion with respect to the probability measure Q given by
RT RT
u(s) dWs − 21 u(s)2 ds
dQ = e− 0 0 dP.
Proof: In order to prove that Xt is a Brownian motion on the probability space (Ω, F , Q)
we shall apply Lévy’s characterization theorem, see Theorem 2.1.5. Hence it suffices to show
that Xt is a Wiener process. Lemma 9.2.1 implies that the process Xt satisfies the following
properties:
1. X0 = 0;
2. Xt is continuous in t;
3. Xt is a square integrable Ft -martingale on the space (Ω, F , Q).
The only property we still need to show is
4. E Q [(Xt − Xs )2 ] = t − s, s < t.
Using Lemma 9.2.2, the martingale property of Wt , the additivity and the tower property of
expectations, we have
E Q [(Xt − Xs )2 ] = E Q [Xt2 ] − 2E Q [Xt Xs ] + E Q [Xs2 ]
= t − 2E Q [Xt Xs ] + s
= t − 2E Q [E Q [Xt Xs |Fs ]] + s
= t − 2E Q [Xs E Q [Xt |Fs ]] + s
= t − 2E Q [Xs2 ] + s
= t − 2s + s = t − s,
which proves relation 4.
Choosing u(s) = λ, constant, we obtain the following consequence that will be useful later
in finance applications.
Corollary 9.2.4 Let Wt be a Brownian motion on the probability space (Ω, F , P ). Then the
process
Xt = λt + Wt , 0≤t≤T
is a Brownian motion on the probability space (Ω, F , Q), where
1 2
dQ = e− 2 λ T −λWT
dP.
 169

Chapter 1
1.6.1 Let X ∼ N (µ, σ 2 ). Then the distribution function of Y is
 y −β
FY (y) = P (Y < y) = P (αX + β < y) = P X <
α
Z y−β Z y
2
1 α
− (x−µ)
2
1 −
z−(αµ+β)

= √ e 2σ2 dx = √ e 2α2 σ2 dz
2πσ −∞ 2πασ −∞
Z y ′ 2
1 − (z−µ )
= √ e 2(σ′ )2 dz,
2πσ ′ −∞
with µ′ = αµ + β, σ ′ = ασ.
2 2
1.6.2 (a) Making t = n yields E[Y n ] = E[enX ] = eµn+n σ /2 .
(b) Let n = 1 and n = 2 in (a) to get the first two moments and then use the formula of
variance.
1.11.5 The tower property
 
E E[X|G]|H = E[X|H], H⊂G

is equivalent with Z Z
E[X|G] dP = X dP, ∀A ∈ H.
A

Since A ∈ G, the previous relation holds by the definition on E[X|G].

1.11.6 (a) Direct application of the definition.
R R
(b) P (A) = A dP = Ω χA (ω) dP (ω) = E[χA ].
(d) E[χA X] = E[χA ]E[X] = P (A)E[X].
(e) We have the sequence of equivalencies
Z Z
E[X|G] = E[X] ⇔ E[X] dP = X dP, ∀A ∈ G ⇔
A A
Z
E[X]P (A) = X dP ⇔ E[X]P (A) = E[χA ],
A

which follows from (d).

1.12.4 If µ = E[X] then

E[(X − E[X])2 ] = E[X 2 − 2µX + µ2 ] = E[X 2 ] − 2µ2 + µ2

= E[X 2 ] − E[X]2 = V ar[X].

1.12.5 From 1.12.4 we have V ar(X) = 0 ⇔ X = E[X], i.e. X is a constant.

1.12.6 The same proof as the one in Jensen’s inequality.
1.12.7 It follows from Jensen’s inequality or using properties of integrals.
P k −λ t
1.12.12 (a) m(t) = E[etX ] = k etk λ k! e
= eλ(e −1) ;
(b) It follows from the first Chernoff bound.
170

1.12.16 Choose f (x) = x2k+1 and g(x) = x2n+1 .

1.13.2 By direct computation we have

E[(X − Y )2 ] = E[X 2 ] + E[Y 2 ] − 2E[XY ]

= V ar(X) + E[X]2 + V ar[Y ] + E[Y ]2 − 2E[X]E[Y ]
+2E[X]E[Y ] − 2E[XY ]
= V ar(X) + V ar[Y ] + (E[X] − E[Y ])2 − 2Cov(X, Y ).

1.13.3 (a) Since

E[(X − Xn )2 ] ≥ E[X − Xn ]2 = (E[Xn ] − E[X])2 ≥ 0,

the Squeeze theorem yields limn→∞ (E[Xn ] − E[X]) = 0.

(b) Writing
Xn2 − X 2 = (Xn − X)2 − 2X(X − Xn ),
and taking the expectation we get

E[Xn2 ] − E[X 2 ] = E[(Xn − X)2 ] − 2E[X(X − Xn )].

The right side tends to zero since

E[(Xn − X)2 ] → 0
Z
|E[X(X − Xn )]| ≤ |X(X − Xn )| dP
Ω
 Z 1/2  Z 1/2
2
≤ X dP (X − Xn )2 dP
pΩ Ω
= E[X 2 ]E[(X − Xn )2 ] → 0.

(c) It follows from part (b).

(d) Apply Exercise 1.13.2.
1.13.4 Using Jensen’s inequality we have

2  
2 
E E[Xn |H] − E[X|H] = E E[Xn − X|H]
 
≤ E E[(Xn − X)2 |H]
= E[(Xn − X)2 ] → 0,

as n → 0.
1.15.7 The integrability of Xt follows from
   
E[|Xt |] = E |E[X|Ft ]| ≤ E E[|X| |Ft ] = E[|X|] < ∞.

Xt is Ft -predictable by the definition of the conditional expectation. Using the tower property
yields  
E[Xt |Fs ] = E E[X|Ft ]|Fs = E[X|Fs ] = Xs , s < t.
 171

1.15.8 Since

E[|Zt |] = E[|aXt + bYt + c|] ≤ |a|E[|Xt |] + |b|E[|Yt |] + |c| < ∞

then Zt is integrable. For s < t, using the martingale property of Xt and Yt we have
E[Zt |Fs ] = aE[Xt |Fs ] + bE[Yt |Fs ] + c = aXs + bYs + c = Zs .

1.15.9 In general the answer is no. For instance, if Xt = Yt the process Xt2 is not a martingale,
since the Jensen’s inequality
 2
E[Xt2 |Fs ] ≥ E[Xt |Fs ] = Xs2

is not necessarily an identity. For instance Bt2 is not a martingale, with Bt the Brownian motion
process.
1.15.10 It follows from the identity

E[(Xt − Xs )(Yt − Ys )|Fs ] = E[Xt Yt − Xs Ys |Fs ].

1.15.11 (a) Let Yn = Sn − E[Sn ]. We have

Yn+k = Sn+k − E[Sn+k ]
k
X k
X
= Yn + Xn+j − E[Xn+j ].
j=1 j=1

Using the properties of expectation we have

k
X k
X
E[Yn+k |Fn ] = Yn + E[Xn+j |Fn ] − E[E[Xn+j ]|Fn ]
j=1 j=1
k
X k
X
= Yn + E[Xn+j ] − E[Xn+j ]
j=1 j=1
= Yn .
(b) Let Zn = Sn2 − V ar(Sn ). The process Zn is an Fn -martingale iff
E[Zn+k − Zn |Fn ] = 0.

Let U = Sn+k − Sn . Using the independence we have

2


Zn+k − Zn = (Sn+k − Sn2 ) − V ar(Sn+k − V ar(Sn )


= (Sn + U )2 − Sn2 − V ar(Sn+k − V ar(Sn )
= U 2 + 2U Sn − V ar(U ),
so
E[Zn+k − Zn |Fn ] = E[U 2 ] + 2Sn E[U ] − V ar(U )
= E[U 2 ] − (E[U 2 ] − E[U ]2 )
= 0,
172

since E[U ] = 0.
1.15.12 Let Fn = σ(Xk ; k ≤ n). Using the independence

E[|Pn |] = E[|X0 |] · · · E[|Xn |] < ∞,

so |Pn | integrable. Taking out the predictable part we have

E[Pn+k |Fn ] = E[Pn Xn+1 · · · Xn+k |Fn ] = Pn E[Xn+1 · · · Xn+k |Fn ]

= Pn E[Xn+1 ] · · · E[Xn+k ] = Pn .

1.15.13 (a) Since the random variable Y = θX is normally distributed with mean θµ and
variance θ2 σ 2 , then
1 2 2
E[eθX ] = eθµ+ 2 θ σ .
Hence E[eθX ] = 1 iff θµ + 12 θ2 σ 2 = 0 which has the nonzero solution θ = −2µ/σ 2 .
(b) Since eθXi are independent, integrable and satisfy E[eθXi ] = 1, by Exercise 1.15.12 we
get that the product Zn = eθSn = eθX1 · · · eθXn is a martingale.
Chapter 2
2.1.4 Bt starts at 0 and is continuous in t. By Proposition 2.1.2 Bt is a martingale with
E[Bt2 ] = t < ∞. Since Bt − Bs ∼ N (0, |t − s|), then E[(Bt − Bs )2 ] = |t − s|.
2.1.10 It is obvious that Xt = Wt+t0 − Wt0 satisfies X0 = 0 and that Xt is continuous in t. The
increments are normal distributed Xt −Xs = Wt+t0 −Ws+t0 ∼ N (0, |t−s|). If 0 < t1 < · · · < tn ,
then 0 < t0 < t1 + t0 < · · · < tn + t0 . The increments Xtk+1 − Xtk = Wtk+1 +t0 − Wtk +t0 are
obviously independent and stationary.
2.1.11 For any λ > 0, show that the process Xt = √1λ Wλt is a Brownian motion. This says that
the Brownian motion is invariant by scaling. Let s < t. Then Xt − Xs = √1λ (Wλt − Wλs ) ∼


√1 N 0, λ(t − s) = N (0, t − s). The other properties are obvious.
λ

2.1.12 Apply Property 2.1.8.

2.1.13 Using the moment generating function, we get E[Wt3 ] = 0, E[Wt4 ] = 3t2 .
2.1.14 (a) Let s < t. Then
h i h i
E[(Wt2 − t)(Ws2 − s)] = E E[(Wt2 − t)(Ws2 − s)]|Fs = E (Ws2 − s)E[(Wt2 − t)]|Fs
h i h i
= E (Ws2 − s)2 = E Ws4 − 2sWs2 + s2
= E[Ws4 ] − 2sE[Ws2 ] + s2 = 3s2 − 2s2 + s2 = 2s2 .

(b) Using part (a) we have

2s2 = E[(Wt2 − t)(Ws2 − s)] = E[Ws2 Wt2 ] − sE[Wt2 ] − tE[Ws2 ] + ts

= E[Ws2 Wt2 ] − st.

Therefore E[Ws2 Wt2 ] = ts + 2s2 .

(c) Cov(Wt2 , Ws2 ) = E[Ws2 Wt2 ] − E[Ws2 ]E[Wt2 ] = ts + 2s2 − ts = 2s2 .
 173

2s2
(d) Corr(Wt2 , Ws2 ) = 2ts = st , where we used

V ar(Wt2 ) = E[Wt4 ] − E[Wt2 ]2 = 3t2 − t2 = 2t2 .

2.1.15 (a) The distribution function of Yt is given by

F (x) = P (Yt ≤ x) = P (tW1/t ≤ x) = P (W1/t ≤ x/t)

Z x/t Z x/t p
2
= φ1/t (y) dy = t/(2π)ety /2 dy
0 0
Z √
x/ t
1 2
= √ e−u /2 du.
0 2π
(b) The probability density of Yt is obtained by differentiating F (x)
Z √
x/ t
′d 1 2 1 −x2 /(2t)
p(x) = F (x) = √ e−u /2 du = √ e ,
dx 0 2π 2πt
and hence Yt ∼ N (0, t).
(c) Using that Yt has independent increments we have

Cov(Ys , Yt ) = E[Ys Yt ] − E[Ys ]E[Yt ] = E[Ys Yt ]

h i
= E Ys (Yt − Ys ) + Ys2 = E[Ys ]E[Yt − Ys ] + E[Ys2 ]
= 0 + s = s.

(d) Since
Yt − Ys = (t − s)(W1/t − W0 ) − s(W1/s − W1/t )
E[Yt − Ys ] = (t − s)E[W1/t ] − sE[W1/s − W1/t ] = 0,
and
1 1 1
V ar(Yt − Ys ) = E[(Yt − Ys )2 ] = (t − s)2 + s2 ( − )
t s t
(t − s)2 + s(t − s)
= = t − s.
t

2.1.16 (a) Applying the definition of expectation we have

Z ∞ Z ∞
1 − x2 1 − x2
E[|Wt |] = |x| √ e 2t dx = 2x √ e 2t dx
−∞ 2πt 0 2πt
Z ∞
1 y p
= √ e− 2t dy = 2t/π.
2πt 0
(b) Since E[|Wt |2 ] = E[Wt2 ] = t, we have
2t 2
V ar(|Wt |) = E[|Wt |2 ] − E[|Wt |]2 = t − = t(1 − ).
π π

2.1.17 By the martingale property of Wt2 − t we have

E[Wt2 |Fs ] = E[Wt2 − t|Fs ] + t = Ws2 + t − s.

174

2.1.18 (a) Expanding

(Wt − Ws )3 = Wt3 − 3Wt2 Ws + 3Wt Ws2 − Ws3

and taking the expectation

E[(Wt − Ws )3 |Fs ] = E[Wt3 |Fs ] − 3Ws E[Wt2 ] + 3Ws2 E[Wt |Fs ] − Ws3
= E[Wt3 |Fs ] − 3(t − s)Ws − Ws3 ,

so
E[Wt3 |Fs ] = 3(t − s)Ws + Ws3 ,
since
E[(Wt − Ws )3 |Fs ] = E[(Wt − Ws )3 ] = E[Wt−s
3
] = 0.

(b) Hint: Start from the expansion of (Wt − Ws )4 .

2.2.3 Using that eWt −Ws is stationary, we have
1
E[eWt −Ws ] = E[eWt−s ] = e 2 (t−s) .

2.2.4 (a)

E[Xt |Fs ] = E[eWt |Fs ] = E[eWt −Ws eWs |Fs ]

= eWs E[eWt −Ws |Fs ] = eWs E[eWt −Ws ]
= eWs et/2 e−s/2 .

(b) This can be written also as

E[e−t/2 eWt |Fs ] = e−s/2 eWs ,

which shows that e−t/2 eWt is a martingale.

(c) From the stationarity we have
1 2
E[ecWt −cWs ] = E[ec(Wt −Ws ) ] = E[ecWt−s ] = e 2 c (t−s)
.

Then for any s < t we have

E[ecWt |Fs ] = E[ec(Wt −Ws ) ecWs |Fs ] = ecWs E[ec(Wt −Ws ) |Fs ]
1 2 1 2
= ecWs E[ec(Wt −Ws ) ] = ecWs e 2 c (t−s)
= Ys e 2 c t .
1 2
Multiplying by e− 2 c t
yields the desired result.
2.2.5 (a) Using Exercise 2.2.3 we have

Cov(Xs , Xt ) = E[Xs Xt ] − E[Xs ]E[Xt ] = E[Xs Xt ] − et/2 es/2

= E[eWs +Wt ] − et/2 es/2 = E[eWt −Ws e2(Ws −W0 ) ] − et/2 es/2
t−s
= E[eWt −Ws ]E[e2(Ws −W0 ) ] − et/2 es/2 = e 2 e2s − et/2 es/2
t+3s
= e 2 − e(t+s)/2 .
 175

(b) Using Exercise 2.2.4 (b), we have

h i h i
E[Xs Xt ] = E E[Xs Xt |Fs ] = E Xs E[Xt |Fs ]
h i h i
= et/2 E Xs E[e−t/2 Xt |Fs ] = et/2 E Xs e−s/2 Xs ]
= e(t−s)/2 E[Xs2 ] = e(t−s)/2 E[e2Ws ]
t+3s
= e(t−s)/2 e2s = e 2 ,

and continue like in part (a).

2.2.6 Using the definition of the expectation we have

Z Z
2Wt2 2x2 1 2 x2
E[e ] = e φt (x) dx = √ e2x e− 2t dx
2πt
Z
1 2x2 − 1−4t x 2 1
= √ e e 2t dx = √ ,
2πt 1 − 4t
Z p
2
if 1−4t > 0. Otherwise, the integral is infinite. We used the standard integral e−ax = π/a,
a > 0.
2.3.4 It follows from the fact that Zt is normally distributed.
2.3.5 Using the definition of covariance we have
 
Cov Zs , Zt = E[Zs Zt ] − E[Zs ]E[Zt ] = E[Zs Zt ]
hZ s Z t i hZ s Z t i
= E Wu du · Wv dv = E Wu Wv dudv
0 0 0 0
Z sZ t Z sZ t
= E[Wu Wv ] dudv = min{u, v} dudv
0 0 0 0
t s
= s2 − , s < t.
2 6

2.3.6 (a) Using Exercise 2.3.5

Cov(Zt , Zt − Zt−h ) = Cov(Zt , Zt ) − Cov(Zt , Zt−h )

t3 t t−h
= − (t − h)2 ( − )
3 2 6
1 2
= t h + o(h).
2
Rt
(b) Using Zt − Zt−h = t−h
Wu du = hWt + o(h),

1
Cov(Zt , Wt ) = Cov(Zt , Zt − Zt−h )
h
1 1 2
1
= t h + o(h) = t2 .
h 2 2
176

2.3.7 Let s < u. Since Wt has independent increments, taking the expectation in
eWs +Wt = eWt −Ws e2(Ws −W0 )
we obtain
u−s
E[eWs +Wt ] = E[eWt −Ws ]E[e2(Ws −W0 ) ] = e 2 e2s
u+s u+s
= e 2 es = e 2 emin{s,t} .
Rt Rt
2.3.8 (a) E[Xt ] = 0 E[eWs ] ds = 0 E[es/2 ] ds = 2(et/2 − 1)
(b) Since V ar(Xt ) = E[Xt2 ] − E[Xt ]2 , it suffices to compute E[Xt2 ]. Using Exercise 2.3.7 we
have
hZ t Z t i hZ tZ t i
2 Wt Wu
E[Xt ] = E e ds · e du = E eWs eWu dsdu
0 0 0 0
Z tZ t Z tZ t
u+s
= E[eWs +Wu ] dsdu = e 2 emin{s,t} dsdu
0 0 0 0
ZZ ZZ
u+s u+s
= e 2 es duds + e 2 eu duds
D1 D2
ZZ
u+s
u 4  1 2t 3
= 2 e 2 e duds = e − 2et/2 + ,
D2 3 2 2
where D1 {0 ≤ s < u ≤ t} and D2 {0 ≤ u < s ≤ t}. In the last identity we applied Fubini’s
theorem. For the variance use the formula V ar(Xt ) = E[Xt2 ] − E[Xt ]2 .
2.3.9 (a) Splitting the integral at t and taking out the predictable part, we have

Z T Z t Z T
E[ZT |Ft ] = E[ Wu du|Ft ] = E[ Wu du|Ft ] + E[ Wu du|Ft ]
0 0 t
Z T
= Zt + E[ Wu du|Ft ]
t
Z T
= Zt + E[ (Wu − Wt + Wt ) du|Ft ]
t
Z T
= Zt + E[ (Wu − Wt ) du|Ft ] + Wt (T − t)
t
Z T
= Zt + E[ (Wu − Wt ) du] + Wt (T − t)
t
Z T
= Zt + E[Wu − Wt ] du + Wt (T − t)
t
= Zt + Wt (T − t),

since E[Wu − Wt ] = 0.
(b) Let 0 < t < T . Using (a) we have
E[ZT − T WT |Ft ] = E[ZT |Ft ] − T E[WT |Ft ]
= Zt + Wt (T − t) − T Wt
= Zt − tWt .
 177

2.4.1
h Rt RT i
E[VT |Ft ] = E e 0 Wu du+ t Wu du |Ft
Rt h RT i
= e 0 Wu du E e t Wu du |Ft
Rt h RT i
= e 0 Wu du E e t (Wu −Wt ) du+(T −t)Wt |Ft
h RT i
= Vt e(T −t)Wt E e t (Wu −Wt ) du |Ft
h RT i
= Vt e(T −t)Wt E e t (Wu −Wt ) du
h R T −t i
= Vt e(T −t)Wt E e 0 Wτ dτ
3
1 (T −t)
= Vt e(T −t)Wt e 2 3 .

2.6.1
F (x) = P (Yt ≤ x) = P (µt + Wt ≤ x) = P (Wt ≤ x − µt)
Z x−µt
1 − u2
= √ e 2t du;
0 2πt
1 − (x−µt)2
f (x) = F ′ (x) = √ e 2t .
2πt

2.7.2 Since Z ρ
1 − x2
P (Rt ≤ ρ) = xe 2t dx,
0 t
use the inequality
x2 x2
1− < e− 2t < 1
2t
to get the desired result.
2.7.3
Z ∞ Z ∞
1 2 − x2
E[Rt ] = xpt (x) dx = x e 2t dx
0 0 t
Z √ Z ∞ 3 −1 −z
1 ∞ 1/2 − y
= y e 2t dy = 2t z 2 e dz
2t 0 0
√
√ 3
2πt
= 2t Γ = .
2 2
Since E[Rt2 ] = E[W1 (t)2 + W2 (t)2 ] = 2t, then
2πt π

V ar(Rt ) = 2t − = 2t 1 − .
4 4
2.7.4
√ r
E[Xt ] 2πt π
E[Xt ] = = = → 0, t → ∞;
t 2t 2t
1 2 π
V ar(Xt ) = 2
V ar(Rt ) = (1 − ) → 0, t → ∞.
t t 4
178

By Example 1.13.1 we get Xt → 0, t → ∞ in mean square.

2.8.2

P (Nt − Ns = 1) = λ(t − s)e−λ(t−s)

= λ(t − s) 1 − λ(t − s) + o(t − s)
= λ(t − s) + o(t − s).

P (Nt − Ns ≥ 1) = 1 − P (Nt − Ns = 0) + P (Nt − Ns = 1)

= 1 − e−λ(t−s) − λ(t − s)e−λ(t−s)
 
= 1 − 1 − λ(t − s) + o(t − s)
 
−λ(t − s) 1 − λ(t − s) + o(t − s)
= λ2 (t − s)2 = o(t − s).

2.8.6 Write first as

Nt2 = Nt (Nt − Ns ) + Nt Ns
= (Nt − Ns )2 + Ns (Nt − Ns ) + (Nt − λt)Ns + λtNs ,

then

E[Nt2 |Fs ] = E[(Nt − Ns )2 |Fs ] + Ns E[Nt − Ns |Fs ] + E[Nt − λt|Fs ]Ns + λtNs
= E[(Nt − Ns )2 ] + Ns E[Nt − Ns ] + E[Nt − λt]Ns + λtNs
= λ(t − s) + λ2 (t − s)2 + λtNs + Ns2 − λsNs + λtNs
= λ(t − s) + λ2 (t − s)2 + 2λ(t − s)Ns + Ns2
= λ(t − s) + [Ns + λ(t − s)]2 .

Hence E[Nt2 |Fs ] 6= Ns2 and hence the process Nt2 is not an Fs -martingale.
2.8.7 (a)
X
mNt (x) = E[exNt ] = exh P (Nt = k)
k≥0
X λk tk
= exk e−λt
k!
k≥0
x x
= e−λt eλte = eλt(e −1)
.

(b) E[Nt2 ] = m′′Nt (0) = λ2 t2 + λt. Similarly for the other relations.
2.8.8 E[Xt ] = E[eNt ] = mNt (1) = eλt(e−1) .
2.8.9 (a) Since exMt = ex(Nt −λt) = e−λtx exNt , the moment generating function is

mMt (x) = E[exMt ] = e−λtx E[exNt ]

x x
= e−λtx eλt(e −1) = eλt(e −x−1) .
 179

(b) For instance

E[Mt3 ] = m′′′
Mt (0) = λt.
Since Mt is a stationary process, E[(Mt − Ms )3 ] = λ(t − s).
2.8.10
V ar[(Mt − Ms )2 ] = E[(Mt − Ms )4 ] − E[(Mt − Ms )2 ]2
= λ(t − s) + 3λ2 (t − s)2 − λ2 (t − s)2
= λ(t − s) + 2λ2 (t − s)2 .
hR i R Rt
t t λt2
2.8.16 (a) E[Ut ] = E 0
Nu du = 0 E[Nu ] du = 0 λu du = 2 .
Nt
X  λt2 λt2
(b) E Sk = E[tNt − Ut ] = tλt − = .
2 2
k=1
2.8.18 The proof cannot go through because a product between a constant and a Poisson
process is not a Poisson process.
2.8.19 Let pX (x) be the P
probability density of X. If p(x, y) is the joint probability density of
X and Y , then pX (x) = y p(x, y). We have
X XZ
E[X|Y = y]P (Y = y) = xpX|Y =y (x|y)P (Y = y) dx
y≥0 y≥0
XZ xp(x, y)
Z X
= P (Y = y) dx = x p(x, y) dx
P (Y = y)
y≥0 y≥0
Z
= xpX (x) dx = E[X].

2.8.21 (a) Since Tk has an exponential distribution with parameter λ

Z ∞
λ
E[e−σTk ] = e−σx λe−λx dx = .
0 λ+σ
(b) We have
Ut = T2 + 2T3 + 3T4 + · · · + (n − 2)Tn−1 + (n − 1)Tn + (t − Sn )n
= T2 + 2T3 + 3T4 + · · · + (n − 2)Tn−1 + (n − 1)Tn + nt − n(T1 + T2 + · · · + Tn )
= nt − [nT1 + (n − 1)T2 + · · · + Tn ].
(c) Using that the arrival times Sk , k = 1, 2, . . . n, have the same distribution as the order
statistics U(k) corresponding to n independent random variables uniformly distributed on the
interval (0, t), we get
h i PN t
E e−σUt Nt = n = E[e−σ(tNt − k=1 Sk ) |Nt = n]
Pn Pn
= e−nσt E[eσ i=1 U(i)
] = e−nσt E[eσ i=1 Ui
]
−nσt σU1 σUn
= e E[e ] · · · E[e ]
Z t Z t
1 1
= e−nσt eσx1 dx1 · · · eσxn dxn
t 0 t 0
(1 − e−σt )n
= ·
σ n tn
180

(d) Using Exercise 2.8.19 we have

h X h i
E e−σUt ] = P (Nt = n)E e−σUt Nt = n
n≥0
X e−λ tn λn (1 − e−σt )n
=
n! σ n tn
n≥0
−σt
= eλ(1−e )/σ−λ
·

3.9.6 (a) dt dNt = dt(dMt + λdt) = dt dMt + λdt2 = 0

(b) dWt dNt = dWt (dMt + λdt) = dWt dMt + λdWt dt = 0
2.9.6 Use Doob’s
p inequality for the submartingales Wt2 and |Wt |, and use that E[Wt2 ] = t and
E[|Wt |] = 2t/π, see Exercise 2.1.16 (a).
2.9.7 Divide by t in the inequality from Exercise 2.9.6 part (b).
2.9.10 Let σ = n and τ = n + 1. Then
h N 2 i 4λ(n + 1)
t
E sup −λ ≤ ·
n≤t≤n=1 t n2

The result follows by taking n → ∞ in the sequence of inequalities

h N 2 i h N 2 i 4λ(n + 1)
t t
0≤E −λ ≤E sup −λ ≤ ·
t n≤t≤n=1 t n2

Chapter 3
3.1.2 We have 
Ω, if c ≤ t
{ω; τ (ω) ≤ t} =
Ø, if c > t
and use that Ø, Ω ∈ Ft .
3.1.3 First we note that
[
{ω; τ (ω) < t} = {ω; |Ws (ω)| > K}. (9.2.16)
0<s<t

This can be shown by double inclusion. Let As = {ω; |Ws (ω)| > K}.
“ ⊂ ” Let ω ∈ {ω; τ (ω) < t}, so inf{s > 0; |Ws (ω)| > K} < t. Then exists τ < u < t such that
|Wu (ω)| > K, and
S hence ω ∈ Au .
“ ⊃ ” Let ω ∈ 0<s<t {ω; |Ws (ω)| > K}. Then there is 0 < s < t such that |Ws (ω)| > K. This
implies τ (ω) < s and since s < t it follows that τ (ω) < t.
Since Wt is continuous, then (9.2.16) can be also written as
[
{ω; τ (ω) < t} = {ω; |Wr (ω)| > K},
0<r<t,r∈Q
 181

which implies {ω; τ (ω) < t} ∈ Ft since {ω; |Wr (ω)| > K} ∈ Ft , for 0 < r < t. Next we shall
show that P (τ < ∞) = 1.
[ 
P ({ω; τ (ω) < ∞}) = P {ω; |Ws (ω)| > K} > P ({ω; |Ws (ω)| > K})
0<s
Z
1 y2 2K
= 1− √ e− 2s dy > 1 − √ → 1, s → ∞.
|x|<K 2πs 2πs

Hence τ is a stopping time.

1 1
3.1.4 Let Km = [a + m, b − m ]. We can write
\ [
{ω; τ ≤ t} = {ω; Xr ∈
/ K m } ∈ Ft ,
m≥1 r<t,r∈Q

/ Km } = {ω; Xr ∈ K m } ∈ Fr ⊂ Ft .
since {ω; Xr ∈
3.1.8 (a) We have {ω; cτ ≤ t} = {ω; τ ≤ t/c} ∈ Ft/c ⊂ Ft . And P (cτ < ∞) = P (τ < ∞) = 1.
(b) {ω; f (τ ) ≤ t} = {ω; τ ≤ f −1 (t)} = Ff −1 (t) ⊂ Ft , since f −1 (t) ≤ t. If f is bounded, then it is


obvious that P (f (τ ) < ∞) = 1. If limt→∞ f (t) = ∞, then P (f (τ ) < ∞) = P τ < f −1 (∞) =
P (τ < ∞) = 1.
(c) Apply (b) with f (x) = ex .
T
3.1.10 If let G(n) = {x; |x − a| < n1 }, then {a} = n≥1 G(n). Then τn = inf{t ≥ 0; Wt ∈ G(n)}
are stopping times. Since supn τn = τ , then τ is a stopping time.
3.2.3 The relation is proved by verifying two cases:
(i) If ω ∈ {ω; τ > t} then (τ ∧ t)(ω) = t and the relation becomes

Mτ (ω) = Mt (ω) + Mτ (ω) − Mt (ω).

(ii) If ω ∈ {ω; τ ≤ t} then (τ ∧ t)(ω) = τ (ω) and the relation is equivalent with the obvious
relation
Mτ = Mτ .

3.2.5 Taking the expectation in E[Mτ |Fσ ] = Mσ yields E[Mτ ] = E[Mσ ], and then make σ = 0.
3.3.9 Since Mt = WT2 − t is a martingale with E[Mt ] = 0, by the Optional Stopping Theorem
we get E[Mτa ] = E[M0 ] = 0, so E[Wτ2a − τa ] = 0, from where E[τa ] = E[Wτ2a ] = a2 , since
Wτa = a.
3.3.10 (a)

F (a) = P (Xt ≤ a) = 1 − P (Xt > a) = 1 − P ( max Ws > a)

0≤s≤t
Z ∞
2 −y 2 /2
= 1 − P (Ta ≤ t) = 1 − √ √ e dy
2π |a|/ t
Z ∞ Z ∞
2 2 2 2
= √ e−y /2 dy − √ e−y /2 dy
2π 0 2π |a|/√t
Z |a|/√t
2 2
= √ e−y /2 dy.
2π 0
182

2
(b) The density function is p(a) = F ′ (a) = √ 2 e−a /(2t) , a > 0. Then
2πt
Z ∞ Z ∞
1 −y2 /(2t) 1
E[Xt ] = xp(x) dx = √ e dy = .
0 0 2πt 2

3.3.16 It is recurrent since P (∃t > 0 : a < Wt < b) = 1.

3.4.4 Since
r
1 1 t1
P (Wt > 0; t1 ≤ t ≤ t2 ) = P (Wt 6= 0; t1 ≤ t ≤ t2 ) = arcsin ,
2 π t2
using the independence
r
1  t1 2
P (Wt1 > 0, Wt2 ) = P (Wt1 > 0)P (Wt2 > 0) = 2 arcsin .
π t2
r
4 t1 2
The probability for Wt = (Wt1 , Wt2 ) to be in one of the quadrants is arcsin .
π2 t2
3.5.2 (a) We have

e2µβ − 1
P (Xt goes up to α) = P (Xt goes up to α before down to − ∞) = lim = 1.
β→∞ e2µβ − e−2µα

3.5.3
e2µβ − 1
P (Xt never hits − β) = P (Xt goes up to ∞ before down to − β) = lim .
α→∞ e2µβ − e−2µα

3.5.4 (a) Use that E[XT ] = αpα − β(1 − pα ); (b) E[XT2 ] = α2 pα + β 2 (1 − pα ), with pα =
e2µβ −1
e2µβ −e−2µα
; (c) Use V ar(T ) = E[T 2 ] − E[T ]2 .
3.5.7 Since Mt = Wt2 − t is a martingale, with E[Mt ] = 0, by the Optional Stopping Theorem
we get E[WT2 − T ] = 0. Using WT = XT − µT yields

E[XT2 − 2µT XT + µ2 T 2 ] = E[T ].

Then
E[T ](1 + 2µE[XT ]) − E[XT2 ]
E[T 2 ] = .
µ2
Substitute E[XT ] and E[XT2 ] from Exercise 3.5.4 and E[T ] from Proposition 3.5.5.
3.6.10 See the proof of Proposition 3.6.4.
3.6.12 (b) Applying the Optional Stopping Theorem
c
E[ecMT −λT (e −c−1)
] = E[X0 ] = 1
ca−λT f (c)
E[e ] = 1
E[e−λT f (c) ] = e−ac .

Let s = f (c), so c = ϕ(s). Then E[e−λsT ] = e−aϕ(s) .

 183

(c) Differentiating and taking s = 0 yields

−λE[T ] = −ae−aϕ(0)ϕ′ (0)

1
= −a ′ = −∞,
f (0)

so E[T ] = ∞.
 
(d) The inverse Laplace transform L−1 e−aϕ(s) cannot be represented by elementary functions.

3.8.8 Use that if Wt is a Brownian motion then also tW1/t is a Brownian motion.
3.8.11 Use E[(|Xt | − 0)2 ] = E[|Xt |2 ] = E[(Xt − 0)2 ].
a1 +···+an
3.8.15 Let an = ln bn . Then Gn = eln Gn = e n → eln L = L.
3.8.20 (a) L = 1. (b) A computation shows

E[(Xt − 1)2 ] = E[Xt2 − 2Xt + 1] = E[Xt2 ] − 2E[Xt ] + 1

= V ar(Xt ) + (E[Xt ] − 1)2 .

(c) Since E[Xt ] = 1, we have E[(Xt − 1)2 ] = V ar(Xt ). Since

V ar(Xt ) = e−t E[eWt ] = e−t (e2t − et ) = et − 1,

then E[(Xt − 1)2 ] does not tend to 0 as t → ∞.

Chapter 4
4.0.9 (a) Use either the definition or the moment generation function to show that E[Wt4 ] = 3t2 .
Using stationarity, E[(Wt − Ws )4 ] = E[Wt−s 4
] = 3(t − s)2 .
Z T
4.2.1 (a) E[ dWt ] = E[WT ] = 0.
Z T 0
1 1
(b) E[ Wt dWt ] = E[ WT2 − T ] = 0.
0Z 2 2
T Z T
1 1 1 T2
(c) V ar( Wt dWt ) = E[( Wt dWt )2 ] = E[ WT2 + T 2 − T WT2 ] = .
0 0 4 4 2 2
3.9.3 (a) E[(dWt )2 − dt2 ] = E[(dWt )2 ] − dt2 = 0.

(b) V ar((dWt )2 − dt) = E[(dWt2 − dt)2 ] = E[(dWt )4 − 2dtdWt + dt2 ]

= 3st2 − 2dt · 0 + dt2 = 4dt2 .

RT 1
4.4.2 X ∼ N (0, 1 t dt) = N (0, ln T ).

RT  
4.4.3 Y ∼ N (0, 1
tdt) = N 0, 12 (T 2 − 1)
Z t
1 2t
4.4.4 Normally distributed with zero mean and variance e2(t−s) ds = (e − 1).
0 2
184

7t3
4.4.5 Using the property of Wiener integrals, both integrals have zero mean and variance .
3
4.4.6 The mean is zero and the variance is t/3 → 0 as t → 0.
4.4.7 Since it is a Wiener integral, Xt is normally distributed with zero mean and variance
Z t
bu 2 b2
a+ du = (a2 + + ab)t.
0 t 3
b2
Hence a2 + 3 + ab = 1.
Rt
4.4.8 Since both Wt and 0 f (s) dWs have the mean equal to zero,
 Z t  Z t Z t Z t
Cov Wt , f (s) dWs = E[Wt , f (s) dWs ] = E[ dWs f (s) dWs ]
0 0 0 0
Z t Z t
= E[ f (u) ds] = f (u) ds.
0 0

The general result is Z t

  Z t
Cov Wt , f (s) dWs = f (s) ds.
0 0

Choosing f (u) = un yields the desired identity.

4.5.7 Apply the expectation to
Nt
X Nt
2 X Nt
X
f (Sk ) = f 2 (Sk ) + 2 f (Sk )f (Sj ).
k=1 k=1 k6=j

4.6.1
hZ T i λ kT
E eks dNs = (e − 1)
0 k
Z T  λ 2kT
V ar eks dNs = (e − 1).
0 2k

Chapter 5
Rt Wu
5.2.1 Let Xt = 0
e du. Then

Xt tdXt − Xt dt teWt dt − Xt dt 1  Wt 
dGt = d( )= = = e − Gt dt.
t t2 t2 t

5.3.3
(a) eWt (1 + 12 Wt )dt + eWt (1 + Wt )dWt ;
(b) (6Wt + 10e5Wt )dWt + (3 + 25e5Wt )dt;
2 2
(c) 2et+Wt (1 + Wt2 )dt + 2et+Wt Wt dWt ;
 
(d) n(t + Wt )n−2 (t + Wt + n−12 )dt + (t + Wt )dWt ;
 185

Z !
t
1 1
(e) Wt − Wu du dt;
t t 0
Z !
t
1 α
(f ) α eWt − e Wu
du dt.
t t 0

5.3.4

d(tWt2 ) = td(Wt2 ) + Wt2 dt = t(2Wt dWt + dt) + Wt2 dt = (t + Wt2 )dt + 2tWt dWt .

5.3.5 (a) tdWt + Wt dt;

(b) et (Wt dt + dWt );
(c) (2 − t/2)t cos Wt dt − t2 sin Wt dWt ;
(d) (sin t + Wt2 cos t)dt + 2 sin t Wt dWt ;
5.3.7 It follows from (5.3.9).
5.3.8 Take the conditional expectation in
Z t
Mt2 = Ms2 + 2 Mu− dMu + Nt − Ns
s

and obtain

Z t
E[Mt2 |Fs ] = Ms2 + 2E[ Mu− dMu |Fs ] + E[Nt |Fs ] − Ns
s
= Ms2 + E[Mt + λt|Fs ] − Ns
= Ms2 + Ms + λt − Ns
= Ms2 + λ(t − s).

5.3.9 Integrating in (5.3.10) yields

Z t Z t
∂f ∂f
Ft = Fs + dWt1 + dWt2 .
s ∂x s ∂y

One can check that E[Ft |Fs ] = Fs .

2Wt1 dWt1 + 2Wt2 dWt2
5.3.10 (a) dFt = 2Wt1 dWt1 + 2Wt2 dWt2 + 2dt; (b) dFt = .
(Wt1 )2 + (Wt2 )2
p ∂f x ∂f y
5.3.11 Consider the function f (x, y) = x2 + y 2 . Since = p , = p ,
∂x x + y ∂y
2 2 x + y2
2
1
∆f = p , we get
2 x2 + y 2

∂f ∂f W1 W2 1
dRt = dWt1 + dWt2 + ∆f dt = t dWt1 + t dWt2 + dt.
∂x ∂y Rt Rt 2Rt

Chapter 6
186

6.2.1 (a) Use integration formula with g(x) = tan−1 (x).

Z T Z T Z T
1 1 2Wt
dWt = (tan−1 )′ (Wt ) dWt = tan−1 WT + dt.
0 1 + Wt2 0 2 0 (1 + Wt )2
hZ T
1 i
(b) Use E 2 dWt = 0.
0 1 + Wt
x
(c) Use Calculus to find minima and maxima of the function ϕ(x) = .
(1 + x2 )2

6.2.2 (a) Use integration by parts with g(x) = ex and get

Z T Z T
1
eWt dWt = eWT − 1 − eWt dt.
0 2 0

(b) Applying the expectation we obtain

Z T
1
E[eWT ] = 1 + E[eWt ] dt.
2 0

If let φ(T ) = E[eWT ], then φ satisfies the integral equation

Z T
1
φ(T ) = 1 + φ(t) dt.
2 0

Differentiating yields the ODE φ′ (T ) = 12 φ(T ), with φ(0) = 1. Solving yields φ(T ) = eT /2 .
6.2.3 (a) Apply integration by parts with g(x) = (x − 1)ex to get
Z T Z T Z T
1
Wt eWt dWt = g ′ (Wt ) dWt = g(WT ) − g(0) − g ′′ (Wt ) dt.
0 0 2 0

(b) Applying the expectation yields

Z
1 Th i
E[WT eWT ] = E[eWT ] − 1 + E[eWt ] + E[Wt eWt ] dt
2 0
Z T 
1
= eT /2 − 1 + et/2 + E[Wt eWt ] dt.
2 0

Then φ(T ) = E[WT eWT ] satisfies the ODE φ′ (T ) − φ(T ) = eT /2 with φ(0) = 0.
6.2.4 (a) Use integration by parts with g(x) = ln(1 + x2 ).
(e) Since ln(1 + T ) ≤ T , the upper bound obtained in (e) is better than the one in (d), without
contradicting it.
6.3.1 By straightforward computation.
6.3.2 By computation.
√ √
6.3.13 (a) √e2 sin( 2W1 ); (b) 21 e2T sin(2W3 ); (c) √1 (e 2W4 −4
2
− 1).
 187

6.3.14 Apply Ito’s formula to get

1
dϕ(t, Wt ) = (∂t ϕ(t, Wt ) + ∂x2 ϕ(t, Wt ))dt + ∂x ϕ(t, Wt )dWt
2
= G(t)dt + f (t, Wt ) dWt .

Integrating between a and b yields

Z b Z b
ϕ(t, Wt )|ba = G(t) dt + f (t, Wt ) dWt .
a a

Chapter 7
Rt Rt
7.2.1 Integrating yields Xt = X0 + (2Xs + e2s ) ds + 0 b dWs . Taking the expectation we get
0
Z t
E[Xt ] = X0 + (2E[Xs ] + e2s ) ds.
0
′ 2t
Differentiating we obtain f (t) = 2f (t) + e , where f (t) = E[Xt ], with f (0) = X0 . Multiplying
by the integrating factor e−2t yields (e−2t f (t))′ = 1. Integrating yields f (t) = e2t (t + X0 ).
7.2.4 (a) Using product rule and Ito’s formula, we get
1
d(Wt2 eWt ) = eWt (1 + 2Wt + Wt2 )dt + eWt (2Wt + Wt2 )dWt .
2
Integrating and taking expectations yields
Z t 
2 Wt 1
E[Wt e ] = E[eWs ] + 2E[Ws eWs ] + E[Ws2 eWs ] ds.
0 2

Since E[eWs ] = et/2 , E[Ws eWs ] = tet/2 , if let f (t) = E[Wt2 eWt ], we get by differentiation
1
f ′ (t) = et/2 + 2tet/2 + f (t), f (0) = 0.
2
Multiplying by the integrating factor e−t/2 yields (f (t)e−t/2 )′ = 1 + 2t. Integrating yields the
solution f (t) = t(1 + t)et/2 . (b) Similar method.
7.2.5 (a) Using Exercise 2.1.18

E[Wt4 − 3t2 |Ft ] = E[Wt4 |Ft ] − 3t2

= 3(t − s)2 + 6(t − s)Ws2 + Ws4 − 3t2
= (Ws4 − 3s2 ) + 6s2 − 6ts + 6(t − s)Ws2 6= Ws4 − 3s2

Hence Wt4 − 3t2 is not a martingale. (b)

E[Wt3 |Fs ] = Ws3 + 3(t − s)Ws 6= Ws3 ,

and hence Wt3 is not a martingale.

7.2.6 (a) Similar method as in Example 7.2.4.
(b) Applying Ito’s formula
1
d(cos(σWt )) = −σ sin(σWt )dWt − σ 2 cos(σWt )dt.
2
188

2 σ2
Let f (t) = E[cos(σWt )]. Then f ′ (t) = − σ2 f (t), f (0) = 1. The solution is f (t) = e− 2 t .
(c) Since sin(t + σWt ) = sin t cos(σWt ) + cos t sin(σWt ), taking the expectation and using (a)
and (b) yields
σ2
E[sin(t + σWt )] = sin tE[cos(σWt )] = e− 2 t
sin t.
(d) Similarly starting from cos(t + σWt ) = cos t cos(σWt ) − sin t sin(σWt ).
7.2.7 From Exercise 7.2.6 (b) we have E[cos(Wt )] = e−t/2 . From the definition of expectation
Z ∞
1 − x2
E[cos(Wt )] = cos x √ e 2t dx.
−∞ 2πt

Then choose t = 1/2 and t = 1 to get (a) and (b), respectively.

2
7.2.8 (a) Using a standard method involving Ito’s formula we can get E(Wt ebWt ) = bteb t/2
.
Let a = 1/(2t). We can write
Z √ Z
−ax2 +bx 1 − x2
xe dx = 2πt xebx √ e 2t dx
2πt
r  
√ bWt
√ b2 t/2 π b b2 /(4a)
= 2πtE(Wt e ) = 2πtbte = e .
a 2a

The same method for (b) and (c).

7.2.9 (a) We have

X Wt2n t2n  X E[Wt2n ]t2n

E[cos(tWt )] = E (−1)n = (−1)n
(2n)! (2n)!
n≥0 n≥0
2n n 3n
X t (2n)!t X
n t
= (−1)n = (−1)
(2n)! 2n n! 2n n!
n≥0 n≥0
−t3 /2
= e .

(b) Similar computation using E[Wt2n+1 ] = 0.

Rt Rt t
7.3.2 (a) Xt = 1 + sin t− 0 sin s dWs , E[Xt ] = 1 + sin t, V ar[Xt ] = 0 (sin s)2 ds = 2 − 14 sin(2t);
Rt√ 2
(b) Xt = et − 1 + 0 s dWs , E[Xt ] = et − 1, V ar[Xt ] = t2 ;
R t 4
(c) Xt = 1 + 12 ln(1 + t2 ) + 0 s3/2 dWs , E[Xt ] = 1 + 21 ln(1 + t2 ), V ar(Xt ) = t4 .
7.3.4
Z t
s
Xt = 2(1 − e−t/2 ) + e− 2 +Ws dWs = 1 − 2e−t/2 + e−t/2+Wt
0
= 1 + et/2 (eWt − 2);

Its distribution function is given by

F (y) = P (Xt ≤ y) = P (1 + et/2 (eWt − 2) ≤ y) = P Wt ≤ ln(2 + (y − 1)et/2 )
Z ln(2+(y−1)et/2 )
1 x2
= √ e− 2t dx.
2πt −∞
 189

E[Xt ] = 2(1 − e−t/2 ), V ar(Xt ) = V ar(e−t/2 eWt ) = e−t V ar(eWt ) = et − 1.

7.5.1 (a)
dXt = (2Wt dWt + dt) + Wt dt + tdWt
= d(Wt2 ) + d(tWt ) = d(tWt + Wt2 )
so Xt = tWt + Wt2 . (b) We have
1 1
dXt = (2t − Wt )dt + dWt
t2 t
1 1
= 2tdt + d( Wt ) = d(t2 + Wt ),
t t
so Xt = t2 + 1t Wt − 1 − W1 .
1
(c) dXt = et/2 Wt dt + et/2 dWt = d(et/2 Wt ), so Xt = et/2 Wt . (d) We have
2
= t(2Wt dWt + dt) − tdt + Wt2 dt
dXt
1
= td(Wt2 ) + Wt2 dt − d(t2 )
2
2
t
= d(tWt2 − ),
2
t2
√ √ √
so Xt = tWt − 2 . (e) dXt = dt + d( tWt ) = d(t + tWt ), so Xt = t + tWt − W1 .
Z t
1
7.6.1 (a) Xt = X0 e4t + (1 − e4t ) + 2 e4(t−s) dWs ;
4 0
2
(b) Xt = X0 e3t + (1 − e3t ) + e3t Wt ;
3
1 t
(c) Xt = et (X0 + 1 + Wt2 − ) − 1;
2 2
t 1
(d) Xt = X0 e4t − − (1 − e4t ) + e4t Wt ;
4 16
(e) Xt = X0 et/2 − 2t − 4 + 5et/2 − et cos Wt ;
(f ) Xt = X0 e−t + e−t Wt .
Rt Rt α2
α dWs + 21 α2 ds
7.8.1 (a) The integrating factor is ρt = e− 0 0 = e−αWt + 2 t
, which transforms
α2
the equation in the exact form d(ρt Xt ) = 0. Then ρt Xt = X0 and hence Xt = X0 eαWt − 2 t
.
2 2
−αWt + α2 t t t (1− α2 )t+αWt
(b) ρt = e , d(ρt Xt ) = ρt Xt dt, dYt = Yt dt, Yt = Y0 e , ρt Xt = X0 e , Xt = X0 e .
Rt X2
7.8.2 σt dAt = dXt − σAt dt, E[At ] = 0, V ar(At ) = E[A2t ] = t12 0 E[Xs ]2 ds = t0 .
Chapter 8
1 1
Pn 2
8.1.2 A = 2∆ = 2 k=1 ∂k .
8.1.4
X1 (t) = x01 + W1 (t)
Z t
X2 (t) = x02 + X1 (s) dW2 (s)
0
Z t
0 0
= x2 + x1 W2 (t) + W1 (s) dW2 (s).
0
190

8.3.3 E[τ ] is maximum if b − x0 = x0 − a, i.e. when x0 = (a + b)/2. The maximum value is

(b − a)2 /4.
8.3.4 Let τk = min(k, τ ) ր τ and k → ∞. Apply Dynkin’s formula for τk to show that

1 2
E[τk ] ≤ (R − |a|2 ),
n
and take k → ∞.
8.3.6 x0 and x2−n .
RT
8.4.1 (a) We have a(t, x) = x, c(t, x) = x, ϕ(s) = xes−t and u(t, x) = t xes−t ds = x(eT −t −1).
2 2 RT 2 2

(b) a(t, x) = tx, c(t, x) = − ln x, ϕ(s) = xe(s −t )/2 and u(t, x) = − t ln xe(s −t )/2 ds =
2
−(T − t)[ln x + T6 (T + t) − t3 ].
3

8.4.2 (a) u(t, x) = x(T −t)+ 12 (T −t)2 . (b) u(t, x) = 32 ex e 2 (T −t) −1 . (c) We have a(t, x) = µx,
b(t, x) = σx, c(t, x) = x. The associated diffusion is dXs = µXs ds + σXs dWs , Xt = x, which
is the geometric Brownian motion
1 2
Xs = xe(µ− 2 σ )(s−t)+σ(Ws −Wt )
, s ≥ t.

The solution is
hZ T
1 2
i
u(t, x) = E xe(µ− 2 σ )(s−t)+σ(Ws −Wt )
ds
t
Z T
1
= x e(µ− 2 (s−t))(s−t) E[eσ(s−t) ] ds
t
Z T
1 2
= x e(µ− 2 (s−t))(s−t) eσ (s−t)/2
ds
t
Z T
x  µ(T −t) 
= x eµ(s−t) ds = e −1 .
t µ

Chapter 9
9.1.7 Apply Example 9.1.6 with u = 1.
Rt Rt 1
9.1.8 Xt = 0 h(s) dWs ∼ N (0, 0 h2 (s) ds). Then eXt is log-normal with E[eXt ] = e 2 V ar(Xt ) =
1
R t 2
e 2 0 h(s) ds .
9.1.9 (a) Using Exercise 9.1.8 we have
Rt Rt
u(s) dWs − 21 u(s)2 ds
E[Mt ] = E[e− 0 e 0 ]
Rt 2
Rt Rt Rt
− 21 u(s) ds 1
u(s)2 ds 1
u(s)2 ds
= e 0 E[e− 0
u(s) dWs
] = e− 2 0 e2 0 = 1.

(b) Similar computation with (a).

9.1.10 (a) Applying the product and Ito’s formulas we get

d(et/2 cos Wt ) = −e−t/2 sin Wt dWt .

 191

Integrating yields Z t
t/2
e cos Wt = 1 − e−s/2 sin Ws dWs ,
0

which is an Ito integral, and hence a martingale; (b) Similarly.

9.1.12 Use that the function f (x1 , x2 ) = ex1 cos x2 satisfies ∆f = 0.
9.1.14 (a) f (x) = x2 ; (b) f (x) = x3 ; (c) f (x) = xn /(n(n − 1)); (d) f (x) = ecx ; (e) f (x) =
sin(cx).