Stochastic Processes: Stat433/833 Lecture Notes

Stat433/833 Lecture Notes
Stochastic Processes
Jiahua Chen
Department of Statistics and Actuarial Science
University of Waterloo
Jiahua
c Chen
Key Words: σ-field, Brownian motion, diffusion process, ergordic, finite

dimensional distribution, Gaussian process, Kolmogorov equations, Markov
property, martingale, probability generating function, recurrent, renewal the-
orem, sample path, simple random walk, stopping time, transient,
2
Stat 433/833 Course Outline
• We use the formula for final grade:
25%assignment + 25%midterm + 50%final.
• There will be 4 assignments.
• Office hours: to be announced.
• Office: MC6007 B, Ex 35506, jhchen@uwaterloo.ca

3
Textbook and References
The text book for this course is Probability and Random Processes
by Grimmett and Stirzaker.
It is NOT essential to purchase the textbook. This course note will be
free to download to all students registered. We may not be able to cover all
the materials included.
Stat433 students will be required to work on fewer problems in assign-
ments and exams.
The lecture note reflects the instructor’s still in-mature understanding of
this general topic, which were formulated after reading pieces of the following
books. A request to reserve has been sent to the library for books with call
numbers.
The References are in Random Order.
1. Introduction to probability models. (QA273 .R84) Sheldon M. Ross.

Academic Press.
2. A Second course in stochastic processes. S. Karlin and H. M. Taylor.

Academic Press.
3. An introduction to probability theory and its applications. Volumes I

and II. W. Feller, Wiley.
4. Convergence of probability measures. Billingsley, P. Wiley.
5. Gaussian random processes. (QA274.4.I2613) I.A. Ibragimov, and Rozanov.

Springer-Verlag, New York.
6. Introduction to stochastic calculus with applications. Fima C. Klebaner.

Imperial College Press.
7. Diffusions, Markov processes and martingales.(QA274.7.W54) L. C. G.

Rogers and D. Williams. Cambridge.
8. Probability Theory, independence, interchangeability, martingale. Y. S.

Chow and H. Teicher. Springer, New York.
Contents
1 Review and More 1

1.1 Probability Space . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Random Variable . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 More σ-fields and Related Concepts . . . . . . . . . . . . . . . 6
1.4 Multivariate Normal Distribution . . . . . . . . . . . . . . . . 9
1.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2 Simple Random Walk 13

2.1 Counting sample paths . . . . . . . . . . . . . . . . . . . . . . 14
2.2 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3 Generating Functions 23
3.1 Renewal Events . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.2 Proof of the Renewal Theorem . . . . . . . . . . . . . . . . . . 27
3.3 Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
3.3.1 Quick derivation of some generating functions . . . . . 33
3.3.2 Hitting time theorem . . . . . . . . . . . . . . . . . . . 33
3.3.3 Spitzer’s Identity . . . . . . . . . . . . . . . . . . . . . 35
3.3.4 Leads for tied-down random walk . . . . . . . . . . . . 38
3.4 Branching Process . . . . . . . . . . . . . . . . . . . . . . . . 39
3.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
4 Discrete Time Markov Chain 43

4.1 Classification of States and Chains . . . . . . . . . . . . . . . 45
4.2 Class Properties . . . . . . . . . . . . . . . . . . . . . . . . . . 46
1
2 CONTENTS
4.2.1 Properties of Markov Chain with Finite Number of States 49

4.3 Stationary Distribution . . . . . . . . . . . . . . . . . . . . . . 49
4.3.1 Limiting Theorem . . . . . . . . . . . . . . . . . . . . . 55
4.4 Reversibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
5 Continuous Time Markov Chain 63

5.1 Birth Processes and the Poisson Process . . . . . . . . . . . . 63
5.1.1 Strong Markov Property . . . . . . . . . . . . . . . . . 70
5.2 Continuous time Markov chains . . . . . . . . . . . . . . . . . 71
5.3 Limiting probabilities . . . . . . . . . . . . . . . . . . . . . . . 76
5.4 Birth-death processes and imbedding . . . . . . . . . . . . . . 77
5.5 Embedding . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
5.6 Markov chain Monte Carlo . . . . . . . . . . . . . . . . . . . . 81
6 General Stochastic Process in Continuous Time 87

6.1 Finite Dimensional Distribution . . . . . . . . . . . . . . . . . 87
6.2 Sample Path . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
6.3 Gaussian Process . . . . . . . . . . . . . . . . . . . . . . . . . 91
6.4 Stationary Processes . . . . . . . . . . . . . . . . . . . . . . . 91
6.5 Stopping Times and Martingales . . . . . . . . . . . . . . . . . 93
7 Brownian Motion or Wiener Process 95

7.1 Existence of Brownian Motion . . . . . . . . . . . . . . . . . . 100
7.2 Martingales of Brownian Motion . . . . . . . . . . . . . . . . . 100
7.3 Markov Property of Brownian Motion . . . . . . . . . . . . . . 102
7.4 Exit Times and Hitting Times . . . . . . . . . . . . . . . . . . 103
7.5 Maximum and Minimum of Brownian Motion . . . . . . . . . 105
7.6 Zeros of Brownian Motion and Arcsine Law . . . . . . . . . . 107
7.7 Diffusion Processes . . . . . . . . . . . . . . . . . . . . . . . . 108
Chapter 1
Review and More
1.1 Probability Space

A probability space consists of three parts: sample space, a collection of
events, and a probability measure.
Assume an experiment is to be done. The set of all possible outcomes
is called Sample Space. Every element ω of Ω is called a sample point.
Mathematically, the sample space is merely an arbitrary set. There is no
need of a corresponding experiment.
A probability measure intends to be a function defined for all subsets of
Ω. This is not always possible when the probability measure is required to
have certain properties. Mathematically, we settle on a collection of subsets
of Ω.
Definition 1.1
A collection of sets F is a σ-field (algebra) if it satisfies

1. The empty set φ ∈ F,
2. If A1 , A2 , . . . ∈ F, then ∪∞
i=1 ∈ F (closed under countable union).
3. If A ∈ F, then its complement Ac ∈ F.

♦
Let us give a few examples.
1
2 CHAPTER 1. REVIEW AND MORE
Example 1.1
1. The simplest σ-field is {φ, Ω}.
2. A simplest non-trivial σ-field is {φ, A, Ac , Ω}.
3. A most exhaustive σ-field is F consists of all subsets of Ω.
♦
It can be shown that if A and B are two sets in a σ-field, then the resulting
sets of all commonly known operations between A and B are members of the
same σ-field.
Axioms of Probability Measure

Given a sample space Ω and a suitable σ-field F, A probability measure
P is a mapping from F to R (set of real numbers) such that:
1. 0 ≤ P (A) ≤ 1 for all A ∈ F;
2. P (Ω) = 1;
∞
3. P (∪∞ i=1 P (Ai ) for any Ai ∈ F, i = 1, 2, . . . which satisfy
P
i=1 Ai ) =
Ai Aj = φ whenever i 6= j.
Mathematically, the above definition does not rely on any hypothetical

experiment. A probability space is given by (Ω, F, P ). The above axioms
lead to restrictions on F. For example, suppose Ω = [0, 1] and F is the
σ-field that contains all possible subsets. In this case, if we require that the
probability of all closed intervals equal their lengthes, then it is impossible
to find such a probability measure satisfying Axiom 3.
When the sample space Ω contains finite number of elements, the collec-
tion of all subsets forms a σ-field F. Define P (A) as the ratio of sizes of A
and of Ω, then it is a probability measure. This is the classical definition of
probability.
The axioms for a probability space imply that the probability measure
has many other properties not explicitly stated as axioms. For example, since
P (φ ∪ φ) = P (φ) + P (φ), we must have P (φ) = 0.
1.1. PROBABILITY SPACE 3
Axioms 2 and 3 imply that
1 = P (Ω) = P (A ∪ Ac ) = P (A) + P (Ac ).
Hence, P (Ac ) = 1 − P (A).

For any two events A1 and A2 , we have
P (A1 ∪ A2 ) = P (A1 ) + P (A2 ) − P (A1 A2 ).
In general,
P (∪ni=1 Ai ) = P (Ai1 Ai2 ) + · · · , +(−1)n+1 P (∩ni=1 Ai ).

X X
P (Ai ) −
i1 <i2
Lemma 1.1 Continuity of the probability measure.
Let (Ω, F, P ) be a probability space. If A1 ⊂ A2 ⊂ A3 ⊂ · · · is an increasing

sequence of events and
A = lim ∪ni=1 Ai = lim An ,

n→∞ n→∞
then
P (A) = lim P (An ).
n→∞
Proof: Let Bn = An − An−1 , n = 1, 2, 3, . . . with A0 = φ. Then An =

∪ni=1 Bi , and A = ∪∞ i=1 Bi . Notice that B1 , B2 , . . . are mutually exclusive,
and P (Bn ) = P (An ) − P (An−1 ). Using countable additivity,
∞
X
P (A) = P (Bi )
i=1
n
X
= lim
n→∞
P (Bi )
i=1
= lim P (An ).
n→∞
This completes the proof. ♦

1.2 Random Variable

Definition 1.1
A random variable is a map X : Ω → R such that {X ≤ x} = {ω ∈ Ω :

X(ω) ≤ x} ∈ F for all x ∈ R. ♦
Notice that {X ≤ x} is more often the notation of an event than the
notation for the outcome of an experiment.
The cumulative distribution function (c.d.f.) of a random variable X
is FX (t) = P (X ≤ t). It is known that a c.d.f. is a non-decreasing, right
continuous function such that
F (−∞) = 0, F (∞) = 1.
The random behaviour of X is largely determined by its distribution

through c.d.f. At the same time, there are many different ways to characterise
the distribution of a random variable. Here is an incomplete list:
1. If X is absolutely continuous, then the probability density function

(p.d.f.) is f (x) ≥ 0 such that
Z x
F (x) = f (t)dt.
−∞
For almost all x, f (x) = dF (x)/dx.
2. If X is discrete, taking values on {x1 , x2 , . . .}, the probability mass

function of X is
f (xi ) = P (X = xi ) = F (xj ) − F (xj −)
for i = 1, 2, . . ., where F (a−) = limx→a,x<a F (x).
3. The moment generating function is defined as
M (t) = E{exp(tX)}.
1.2. RANDOM VARIABLE 5
4. The characteristic function of X is defined as
φ(t) = E{exp(itX)}
where i is such that i2 = −1. The advantage of the characteristic

function is that it exists for all X. Otherwise, the moment generating
function is simpler as it is not built on the concepts of complex numbers.
5. The probability generating function is useful when the random variable

is non-negative integer valued.
6. In survival analysis or reliability, we often use survival function
S(x) = P (X > x) = 1 − F (x)
where we assume P (X ≥ 0) = 1.
7. The hazard function of X is
P {x ≤ X < x + ∆|X ≥ x}
λ(x) = lim .
∆→0+ ∆
It is also called “instantaneous failure rate” as it is the rate at which

the individual fails in the next instance given it has survived so far.
When X is absolutely continuous,
f (x) d
λ(x) = = − log S(x).
S(x) dx
Hence, for x ≥ 0, Z x
S(x) = exp{− λ(t)dt}.
0
Let us examine a few examples to illustrate the hazard function and its
implication.
Example 1.2
(1) When X has exponential distribution with density function f (x) =

λ exp(−λx), it has constant hazard rate λ.
(2) When X has Weibull distribution with density function
f (x) = λα(λx)α−1 exp{−(λx)α }
for x ≥ 0. The hazard function is given by
λ(x) = λα(λx)α−1 .
(a) When α = 1, it reduces to exponential distribution. Constant hazard

implies the item does not age.
(b) When α > 1, the hazard increases with time. Hence the item ages.
(c) When α < 1, the hazard decreases with time. The longer the item
has survived, the more durable this item becomes. It ages negatively.
(3) Human mortality can be described by a curve with bathtub shape.
The death rate of new borns are high, then it stabilizes. After certain age,
we becomes vulnerable to diseases and the death rate increases.
1.3 More σ-fields and Related Concepts

Consider a random variable X defined on the probability space (Ω, F, P ).
√ √
For any real number such as − 2, the sample points satisfying X ≤ − 2
form a set which belongs to F. Similarly X ≥ 1.2 is also a set belongs to F.
This claim extends to the union of these two events, the intersection of these
two events, and so on.
Putting all events induced by X as above together, we obtain a collection
of events which will be denoted as σ(X). It can be easily argued that σ(X)
is the smallest sub σ-field generated by X.
Example 1.3
1. The σ-field of a constant random variable;
2. The σ-field of an indicator random variable;
3. The σ-field of a random variable takes only two possible values.

1.3. MORE σ-FIELDS AND RELATED CONCEPTS 7
If X and Y are two random variables, we may define the conditional

cumulative distribution function of X given by Y = y by
P (X ≤ x, Y = y)
P (X ≤ x|Y = y) = .
P (Y = y)
This definition works if P (Y = y) > 0. Otherwise, if X and Y are jointly
absolutely continuous, we make use of their joint density to compute the
conditional density function.
One purpose of introducing conditional density function is for the sake
of computing conditional expectation. Because we have to work with con-
ditional expectation more extensively later, we hope to define conditional
expectation for any pairs of random variables.
If Y is an indicator random variable IA for some event A, such that
0 < P (A) < 1, then E{X|Y = 0} and E{X|Y = 1} are both well defined
as above. Further, the corresponding values are the average sizes (expected
values) of X over the ranges Ac and A. Because of this, the conditional
expectation of X given Y is expected sizes of X over various ranges of Ω
partitioned according to the size of Y . When Y is as simple as an indicator
random variable, this partition is also simple and we can easily work out
these numbers conceptually. When Y is an ordinary random variable, the
partition of Ω is σ(Y ). The problem is: σ(Y ) contains so many events in
general, and also they are not mutually exclusive. Thus, there is no way to
present the conditional expectation of X given Y by enumerating them all.
The solution to this difficulty in mathematics is to define Z = E{X|Y }
as a measurable function of Y such that
E{ZIA } = E{XIA }
for any A ∈ σ(Y ).

We may realize that this definition does not provide any concrete means
for us to compute E{X|Y }. In mathematics, this definition has to be backed
up by an existence theorem that such a function is guaranteed to exist. At
the same time, when another random variable differs to Z only by a zero-
probability event, that random variable is also a conditional expectation of
X given Y . Thus, this definition is unique only up to some zero-probability
event.
Since Z is a function of Y , it is also a random variable. In particular, by

letting A = Ω, we get
E[E{X|Y }] = E{X}.
A remark here is: the conditional expectation is well defined only if the
expectation of X exists. You may remember a famous formula:
V ar(X) = V ar[E{X|Y }] + E[V ar{X|Y }].
Note that definition of the conditional expectation relies on the σ-field

generated by Y only. Thus, if G is a σ-field, we simply define Z = E{X|G}
as a G measurable function such that
E{ZIA } = E{XIA }
for all A ∈ G. Thus, the concept of conditional expectation under measure

theory is in fact built on σ-field.
Let us go over a few concepts in elementary probability theory and see
what they mean under measure theory.
Recall that if X and Y are independent, then we have
E{g1 (X)g2 (Y )} = E{g1 (X)}E{g2 (Y )}
for any functions g1 and g2 . Under measure theory, we need to add a cosmes-
tic condition that these two functions are measurable, plus these expectations
exist. More rigorously, the independence of two random variables is built on
the independence of their σ-fields. It can be easily seen that the σ-field gen-
erated by g1 (X) is a sub-σ-field of that of X. One may then see what g1 (X)
is independent of g2 (Y ).
Example 1.4 Some properties.
1. If X is G measurable, then
E{XY |G} = XE{Y |G}.
2. If G1 ⊂ G2 , then
E[E{X|G2 }|G1 ] = E{X|G1 }.
1.4. MULTIVARIATE NORMAL DISTRIBUTION 9
3. If G and σ(X) are independent of each other, then
E{X|G} = E{X}.
Most well known elementary properties of the mathematical expectation re-

main valid.
1.4 Multivariate Normal Distribution

A group of random variables X have joint normal distribution if their joint
density function is in the form of
C exp(−xAxτ − 2bxτ )
where A is positive definite, and C is a constant such that the density function
has total mass 1. Note that x and b are vectors.
Being positive definite implies that there exists an orthogonal decompo-
sition of A such that
A = BΛB τ
where Λ is diagonal matrix with all element positive and BB τ = I, the
identity matrix.
With this knowledge, it is seen that
Z
1 Z
1 (2π)n/2
exp(− xAx )dx = exp(− yΛy τ )dy =
τ
.
2 2 |A|1/2
Hence when b = 0, the density function has the form
1
f (x) = [(2π)−n |A|]1/2 exp(− xAxτ ).
2
Let X be a random vector with the density function given above, and let
Y = X + µ. Then the density function of Y is given by
1
{(2π)−n |A|}1/2 exp{− (y − µ)A(y − µ)0 }.
2
It is more convenient to use V = A−1 in most applications. Thus, we
have the definition as follows.
Definition 1.1 Multivariate normal distribution
X = (X1 , . . . , Xn ) has multivariate normal distribution (written as N (µ, V )),

if its joint density function is
1
f (x) = {(2π)n |V |}−1/2 exp{− (x − µ)V −1 (x − µ)τ }
2
where V is positive definite matrix. ♦
If X is multivariate normal N (µ, V ), then E(X) = µ and V ar(X) = V .
If X(length n) is N (µ, V ) and D is an n × m matrix of rank m ≤ n, then
Y = XD is N (µD, Dτ V D).
More general, we say a random vector X = (X1 , . . . , Xn ) has multivariate
normal distribution when Xaτ is a normally distributed random variable for
all a ∈ Rn .
A more rigorous and my favoured definition is: if X can be written in
the form AY + b for some (non-random) matrix A and vector b, and Y is
a vector of iid standard normally distributed random variables, then X is a
multinormally distributed random vector.
We now list a number of well known facts about the multivariate distri-
butions. If you find that you are not familiar with a large number of them,
it probably means some catch up work should be done
1. Suppose X has normal distribution with mean µ and variance σ 2 . Its

characteristic function is
1
φ(t) = exp{iµt − σ 2 t2 }.
2
Its moment generating function is
1
M (t) = exp{µt + σ 2 t2 }.
2
2. If X has standard normal distribution, then all its odd moments are
zero, and its even moments are
EX 2r = (2r − 1)(2r − 3) · · · 3 · 1.
1.5. SUMMARY 11
3. Let Φ(x) and φ(x) be the cumulative distribution function and the
density function of the standard normal distribution. It is known that
1 1 1

− 3 φ(x) ≤ 1 − Φ(x) ≤ φ(x)
x x x
for all positive x. In particular, when x is large, they provide a very
accurate bounds.
4. Suppose that X1 , X2 , . . . , Xn are a set of independent and identically

distributed standard normal random variables. Let X(n) = max{Xi , i =
√
1, . . . , n}. Then X(n) = Op ( log n).
5. Suppose Y1 , Y2 are two independent random variables such that Y1 + Y2

are normally distributed, then Y1 and Y2 are jointly normally dis-
tributed. That is, they are multivariate normal.
6. A set of random variables have multivariate normal distribution if and

only if all their linear combinations are normally distributed.
7. Let X1 , . . . , Xn be i.i.d. normal random variables. The sample mean

X̄n = n−1 ni=1 Xi and the sample variance Sn2 = (n − 1)−1 ni=1 (Xi −
P P
X̄)2 are independent. Further, X̄n has normal distribution and Sn2 has
chisquare distribution with n − 1 degrees of freedom.
8. Let X be a vector having multivariate normal distribution with mean µ

and covariance matrix Σ, and A be a non-negative definite symmetric
matrix. Then, XAX τ has chi-squared distribution when
AΣAΣA = AΣA.
1.5 Summary
We did not intend to give you a full account of probability theory learned in
courses preceding this one. Neither we try to lead you to the world of measure
theory based, advanced, and widely regarded as useless rigorous probability
theory. Yet, we did introduce the concept of probability space, and why a
dose of σ-field is needed in doing so.
In the later half of this course, we need the σ-field based definitions of
independence, conditional expectation. It is also very important to get fa-
miliar with multivariate normal random variables. These preparations are
not sufficient, but will serve as starting point. Also, you may be happy to
know that that the pressure is not here yet.
Chapter 2
Simple Random Walk
Simple random walk is an easy object in the family of stochastic processes.

At the same time, it shares many properties with more complex stochas-
tic processes. Understanding simple random walk helps us to understand
abstract stochastic processes.
The building block of a simple random walk is a sequence of iid random
variables X1 , . . . , Xn , . . . such that
P (X1 = 1) = p, P (X1 = −1) = 1 − p = q.
We assume that 0 < p < 1, otherwise, the simple random walk becomes
trivial.
We start with S0 = a and define Sn+1 = Sn + Xn+1 for n = 1, 2, . . .. It
mimics the situation of starting gambling with an initial capital of $a, and
the stake is $1 on each trial. The probability of winning a trial is p and the
probability of losing is q = 1 − p. Trials are assumed independent.
When p = q = 1/2, the random walk is symmetric.
Properties:
A simple random walk is
(i) time homogeneous:
P {Sn+m = j|Sm = a} = P {Sn = j|S0 = a}.
(ii) spatial homogeneous:
P {Sn+m = j + b|Sm = a + b} = P {Sn = j|S0 = a}.
13
14 CHAPTER 2. SIMPLE RANDOM WALK
(iii) Markov:
P {Sn+m = j|S0 = j0 , S1 = j1 , · · · , Sm = jm } = P {Sn+m = j|Sm = jm }.
Example 2.1 Gambler’s ruin
Assume that start with initial capital S0 = a, the game stops as soon as
Sn = 0 or Sn = N for some n, where N ≥ a ≥ 0. Under the conditions of
simple random walk, what is the probability that the game stops at Sn = N ?
Solution: A brute force solution for this problem is almost impossible.

The key is to view this probability as a function of initial capital a. By
establishing a relationship between the probabilities, we will get a difference
equation which is not hard to solve. ♦
2.1 Counting sample paths

If we plot (Si , i = 0, 1, . . . , n) against (0, 1, . . . , n), we obtain a path connect-
ing (0, S0 ) and (n, Sn ). We call it a sample path. Assume that S0 = a and
Sn = b. There might be many possible sample paths leading from (0, a) to
(n, b).
Property 2.1
The number of paths from (0, a) to (n, b) is

!
n
Nn (a, b) = n+b−a
2
when (n + a − b)/2 is a positive integer, 0 otherwise. ♦
Property 2.2
2.1. COUNTING SAMPLE PATHS 15
For a simple random walk, and when (n + a − b)/2 is a positive integer, all
sample paths from (0, a) to (n, b) have equal probability p(n+b−a)/2 q (n+a−b)/2
to occur. In addition,
!
n
P (Sn = b|S0 = a) = n+b−a p(n+b−a)/2 q (n+a−b)/2 .
2
♦
The proof is straightforward.
Example 2.2
It is seen that !
2n n n
P (S2n = 0|S0 = 0) = p q
n
for n = 0, 1, 2, . . .. This is the probability when the random walk returns to
0 at trial 2n.
When p = q = 1/2, we have
!
2n
u2n = P (S2n = 0|S0 = 0) = (1/2)2n .
n
Using Stirling’s approximation
√
n! ≈ 2πn(n/e)n
for large n. The approximation is good even for small n. We have
1
u2n ≈ √ .
nπ
Thus, n u2n = ∞. By renewal theorem to be discussed, Sn = 0 is a recurrent
P
event when p = q = 1/2. ♦

Property 2.3 Reflection principle
Let Nn0 (a, b) be the number of sample paths from (0, a) to (n, b) that touch
or cross the x-axis. Then, when a, b > 0,
Nn0 (a, b) = Nn (−a, b).
Proof: We Simply count the number of paths. The key step is to establish
the one-to-one relationship. ♦
The above property enables us to show the next interesting result.
Property 2.4 Ballot Theorem
The number of paths from (0, 0) to (n, b) that do not revisit the axis is
|b|
Nn (0, b).
n
Proof: Notice that such a sample path has to start with a transition from
(0, 0) to (1, 1). The number of paths from (1, 1) to (n, b), such that b > 0,
which do not touch x-axis is
0
Nn−1 (1, b) − Nn−1 (1, b) = Nn−1 (1, b) − Nn−1 (−1, b)
! !
n−1 n−1
= n−b − n−b−2
2 2
|b|
= Nn (0, b).
n
♦
What does this name suggest? Suppose that Micheal and George are in a
competition for some title. In the end, Micheal wins by b votes in n casts. If
the votes are counted in a random order, and A =“Micheal leads throughout
the count”, then
b
Nn (0, b) b
P (A) = n = .
Nn (0, b) n
Theorem 2.1
Assume that S0 = 0 and b 6= 0. Then

(i) P (S1 S2 · · · Sn 6= 0, Sn = b|S0 = 0) = |b|
n
P (Sn = b|S0 = 0).
−1
(ii) P (S1 S2 · · · Sn 6= 0) = n E(|Sn |). ♦
Proof: The first part can be proved by counting the number of sample
paths which do not touch x-axis.
The second part is a consequence of the first one. Just sum over the
probability of (i) for possible values of Sn . ♦
Theorem 2.2 First passage through b at trial n (Hitting time problem).

If b 6= 0 and S0 = 0, then
|b|
fn (b) = P (S1 6= b, · · · , Sn−1 6= b, Sn = b|S0 = 0) = P (Sn = b).
n
Proof: If we reverse the time, the sample paths become those who reach b
in n trials without touching the x-axis. Hence the result. ♦
The next problem of interest is how high Si have attained before it settles
at Sn = b. We assume b > 0. In general, we often encounter problems of
finding the distribution of the extreme value of a stochastic process. This
is an extremely hard problem, but we have a kind of answer for the simple
random walk.
Define Mn = max{Si , i = 1, . . . , n}.
Theorem 2.3
Suppose S0 = 0. Then for r ≥ 1,
P (Sn = b) if b ≥ r
P (Mn ≥ r, Sn = b) = { r−b
(q/p) P (Sn = 2r − b) if b < r
Hence,
r−1
(q/p)r−b P (Sn = 2r − b)
X
P (Mn ≥ r) = P (Sn ≥ r) +
c=−∞
∞
[1 + (q/p)c−r ]P (Sn = c)
X
= P (Sn = r) +
c=r+1
When p = q = 1/2, it becomes
P (Mn ≥ r) = 2P (Sn ≥ r + 1) + P (Sn = r).
Proof: When b ≥ r, the event Mn ≥ r is a subset of the event Sn = b.

Hence the first conclusion is true.
When b < r, we may draw a horizontal line y = r. For each sample path
belongs to Mn ≥ r, Sn = b, we obtain a partial mirror image: it retains
the part until the sample path touches the line of y = r, and completes it
with the mirror image from there. Thus, the number of sample paths is the
same as the number of sample paths from 0 to 2r − b. The corresponding

probability, however, is obtained by exchanging the roles of r − b pairs of p
and q.
The remaining parts of the theorem are self illustrative. ♦
Let µb be the mean number of visits of the walk to the point b before
it turns to its starting point. Suppose S0 = 0. Let Yn = I(S1 S2 · · · Sn 6=
0, Sn = b). Then ∞
P
n=1 Yn is the number of times it visits b. Notice that this
reasoning is fine even when the walk will never return to 0.
Since E(Yn ) = fb (n), we get
∞
X
µb = fb (n).
n=1
Thus, in general, the mean number of visit is less than 1. When p = q = 0.5,
all states are recurrent. Therefore µb = 1 for any b. ♦
Suppose a perfect coin is tossed until the first equalisation of the accu-
mulated numbers of heads and tails. The gambler receives one dollar every
time that the number of heads exceeds the number of tails by b. This fact
results in comments that the “fair entrance fee” equals 1 independent of b.
My remark: how many of us thinks that this is against their intuition?
Theorem 2.4 Arc sine law for last visit to the origin.
Suppose that p = q = 1/2 and S0 = 0. The probability that the last visit to
0 up to time 2n occurred at time 2k is given by
P (S2k = 0)P (S2n−2k = 0).
♦
Proof: The probability in question is
α2n (2k) = P (S2k = 0)P (S2k+1 · · · S2n 6= 0|S2k = 0)

= P (S2k = 0)P (S1 · · · S2n−2k 6= 0|S0 = 0).
We are hence asked to show that
P (S1 · · · S2n−2k 6= 0|S0 = 0) = u2n−2k .

Since S0 = 0 is part of our assumption, we may omit the conditional part.

We have
X
P (S1 · · · S2m 6= 0|S0 = 0) = P (S1 · · · S2m 6= 0, S2m = b)
b6=0
X |b|
= P (S2m = b)
b6=0 m
m
X 2b
= 2 P (S2m = 2b)
b=1 2m
m
2m X " ! !#
1 2m − 1 2m − 1
= 2 −
2 b=1 m+b−1 m+b
! 2m
2m 1
=
m 2
= u2m .
♦
The proof shows that
P (S1 S2 . . . S2m 6= 0) = P (S2m = 0).
Denote, as usual, u2n = P (S2n = 0) and α2n (2k) = P (S2k = 0, Si 6= 0, i =

2k + 1, . . . , 2n). Then the theorem says
α2n (2k) = u2k u2n−2k .
As p = q = 1/2 is a simple case, we have accurate approximation for u2k .

It turns out that u2k ≈ (πk)−1/2 and so α2n (2k) ≈ {π[k(n − k)]}−1/2 . If
we let T2n be the time of the last visit to 0 up to time 2n, the it follows that
{π[k(n − k)]}−1/2
X
P (T2n ≤ 2xn) ≈
k≤xn
Z x1
≈ du
0 π[u(1 − u)]1/2
2 √
= arcsin( x).
π
That is, the limiting distribution has a density function described by
√
arcsin( x).
(1) Since p = q = 1/2, one may think the balance of heads and tails
should occur very often. Is it true? After 2n tosses with n large, the chance
that it never touches 0 after trial n is 50% which is surprisingly large.
(2) The last time before trial 2n when the simple random walk touches
0 should be closer to the end. This result shows that it is symmetric about
midpoint n. It is more likely to be at the beginning and near the end.
Why is it more likely at the beginning? since it started at 0, it is likely
to touch 0 again soon. Once it wandered away from 0, touching 0 becomes
less and less likely.
Why is it also likely occur near the end: if for some reason the simple
random walk returned to 0 at some point, it becomes more likely to visit 0
again in the near future. Thus, it pushes the most recent visit closer and
closer to the end.
(3) If we count the number of k such that Sk > 0, what kind of distribution
it has? Should it be more likely to be close to n? It turns out that it is either
very small or very large.
Thus, if you gamble, you may win all the time, or lose all the time even
though the game is perfectly fair in the sense of probability theory.
Let us see how the result establishes (3).
Property 2.5 Arc sine law for sojourn times.
Suppose that p = 1/2 and S0 = 0. The probability that the walk spends
exactly 2k intervals of time, up to time 2n, to the right of the origin equals
u2k u2n−2k . ♦
Proof: Call this probability β2n (2k). We are asked to show that β2n (2k) =
α2n (2k).
We use mathematical induction.
The first step is to consider the case when k = n = m.
In our previous proofs, we have shown that for symmetric random walk
u2m = P (S1 · · · S2m 6= 0).
Since these sample paths can belong to one of the two possible groups: always
above 0 or always below 0. Hence,
u2m = 2P (S1 > 0, . . . , S2m > 0).

On the other hand, we have
P (S1 > 0, S2 > 0, · · · , S2m > 0)

= P (S1 = 1, S2 ≥ 1, . . . , S2m ≥ 1)
= P (S1 − 1 = 0, S2 − 1 ≥ 0, . . . , S2m − 1 ≥ 0)
= P (S1 = 1|S0 = 0)P (S2 − 1 ≥ 0, . . . , S2m − 1 ≥ 0|S1 − 1 = 0)
1
= P (S2 ≥ 0, S3 ≥ 0, . . . , S2m ≥ 0|S1 = 0)
2
1
= P (S1 ≥ 0, S2 ≥ 0, . . . , S2m−1 ≥ 0|S0 = 0)
2
1
= P (S1 ≥ 0, S2 ≥ 0, . . . , S2m−1 ≥ 0, S2m ≥ 0|S0 = 0)
2
1
= β2m (2m)
2
where the last equality comes from the fact that S2m−1 ≥ 0 implies S2m ≥ 0
as it is impossible for S2m−1 = 0. Consequently, since u0 = 1, we have shown
that
α2m (2m) = u2m = β2m (2m).
Let us now make an induction assumption that
α2n (2k) = β2n (2k)
for all n with k = 0, 1, . . . , m − 1 and k = n, n − 1, . . . , n − (m − 1). Note the

reason for the validity of the second sets of k is symmetry. Our induction
task is to show that the same is true when k = m.
By renewal equation,
m
X
u2m = u2m−2r f2r .
r=1
Using idea similar to renewal equation, we also have

m
1X 1 n−m
X
β2n (2m) = f2r β2n−2r (2m − 2r) + f2r β2n−2r (2m)
2 r=1 2 r=1
since the walk will start with either a positive move or a negative move, it
will touch 0 again at some point 2r in another 2n − 2r trials (it could happen
that 2r = 2n).
So, with renewal equation, and induction assumption,

m
1X 1 n−m
X
β2n (2m) = f2r u2m−2r u2n−2m + f2r u2m−2r u2n−2m−2r
2 r=1 2 r=1
1
= [u2n−2m u2m + u2m u2n−2m ]
2
= u2m u2n−2m .
The induction is hence completed. ♦
2.2 Summary
What have we learned in this chapter? One observation is that all sample
paths starting and ending at the same location have equal probability to
occur. Making use of this fact, some interesting properties of the simple
random work are revealed.
One such property is that the simple random work is null recurrent when
p = q and transient otherwise. Let us recall some detail; by counting sample
paths, it is possible to provide an accurate enough approximation to the
probability of entering 0.
The reflection principle allows us to count the number of paths going from
one point to another without touching 0. This result is used to establish the
Ballot Theorem. In old days, there were enough nerds who believed that this
result is intriguing.
We may not believe that the hitting time problem is a big deal. Yet when
it is linked with stock price, you may change your mind. If you find some
neat estimates of such probabilities for very general stochastic processes, you
will be famous. In this course, we provide such a result for the simple random
work. There is a similar result for Brownian motion to be discussed.
Similarly, arc sine law also have its twin in Brownian motion. Please do
not go away and stay tuned.
Chapter 3
Generating Functions and

Their Applications
The idea of generating functions is to transform the function under inves-

tigation from one functional space to another functional space. The new
functional space might be more convenient to work with.
If X is a non-negative integer valued random variable, then we call
GX (s) = E(sX )
the probability generating function of X.

Advantages? It can be seen that moments of X equal derivatives of GX (s)
at s = 1.
When X and Y are independent, then the probablity generating function
of X + Y is
GX+Y (s) = GX (s)GY (s).
Theorem 3.1
If X1 , X2 , . . . is a sequence of independent and identically distributed random

variables with common generating function GX (s) and N (≥ 0) is a random
variable which is independent of the Xi ’s and has generating function GN (s),
then
S = X1 , +X2 + · · · + XN
23
24 CHAPTER 3. GENERATING FUNCTIONS
has generating function given by
GS (s) = GN (GX (s)).
♦
The proof can be done through conditioning on N . When N = 0, we
assume that S = 0.
Definition 3.2
The joint probability generating function of two random variables X1 and

X2 taking non-negative integer values, is given by
GX1 ,X2 (s1 , s2 ) = E[sX1 X2

1 s2 ].
Theorem 3.2
X1 and X2 are independent if and only if
GX1 ,X2 (s1 , s2 ) = GX1 (s1 )GX2 (s2 ).
♦
The idea of generating function also applies to a sequence of real numbers.
If {an }∞
n=0 is a sequence of real numbers, then
an s n
X
A(s) =
n
is called the generating function of {an }∞

n=0 when it converges in a neighbor-
hood of s = 0.
Suppose that {an }∞ ∞
n=0 and {bn }n=0 are two sequence of real numbers, and
their generating functions exists in a neighborhood of s = 0. Define
n
X
cn = ai bn−i
i=0
for n = 0, 1, . . .. Then
C(s) = A(s)B(s)
where A(s), B(s) and C(s) are generating functions of {an }∞ ∞
n=0 , {bn }n=0 and
{cn }∞
n=0 respectively.
3.1. RENEWAL EVENTS 25
3.1 Renewal Events

Consider a sequence of trials with outcomes X1 , X2 , . . .. We do not require
Xi ’s be independent of each other. Let λ represent some property which, on
the basis of the outcomes of the first n trials, can be said unequivocally to
occur or not to occur at trial n. By convention, we suppose that λ has just
occurred at trial 0, and En represents the “event” that λ occurs at trial n,
n = 1, 2, . . ..
Roughly speaking, a property is a renewal event if the waiting time dis-
tribution for the next occurrence remains the same, and is independent of
the past, given that it has just occurred. Thus, the process renewals itself
each time when the renewal event occurs. It re-sets the clock back to time 0.
Let λ represent a renewal event and as before define the lifetime sequence
{fn } where f0 = 0 and
fn = P {λ occurs for the first time at trial n}, n = 1, 2, . . . .
In like manner, we define the renewal sequence un , where u0 = 1 and
un = P {λ occurs at trial n}, n = 1, 2, . . . .
Let F (s) = fn sn and U (s) = un sn be the generating functions of

P P
{fn } and {un }. Note that

X
f= fn = F (1) ≤ 1
because f has the interpretation that λ recurs at some time in the sequence.
Since the event may not occur at all, it is possible for f to be less than 1.
Clearly, 1 − f represents the probability that λ never recurs in the infinite
sequence of trials. When f < 1, the probability that λ occurs finite number
of times only is 1. Hence, we say that λ is transient. Otherwise, it it
recurrent.
For a recurrent renewal event, F (s) is a probability generating function.
The mean inter-occurrence time is
∞
µ = F 0 (1) =
X
nfn .
n=0
If µ < ∞, we say that λ is positive recurrent. If µ = ∞, we say that λ is

null recurrent.
Finally, if λ can occur only at n = t, 2t, 3t, . . . for some positive integer
t > 1, we say that λ is periodic with period t. More formally, let t =
g.c.d.{n : fn > 0}. (g.c.d. stands for the greatest common divisor). If t > 1,
the recurrent event λ is said to be periodic with period t. If t = 1, λ is said
to be aperiodic.
For a renewal event λ to occur at trial n ≥ 1, either λ occurs for the first
time at n with probability fn = fn u0 , or λ occurs for the first time at some
intermediate trial k < n and then occurs again at n. The probability of this
event is fk un−k . Notice that f0 = 1, we therefore have
un = f0 un + f1 un−1 + · · · + fn−1 u1 + fn u0 , n = 1, 2, . . . .
This equation is called renewal equation. Using the typical generating

function methodology, we get
U (s) − 1 = F (s)U (s).
Hence
1 1
U (s) = or F (s) = 1 − .
1 − F (s) U (s)
Theorem 3.3
The renewal event λ is
un = U (1) < ∞,
P
1. transient if and only if u =
2. recurrent if and only if u = ∞,
3. periodic if t = g.c.d.{n : un > 0} is greater than 1 and aperiodic if

t = 1.
un = ∞ and un → 0 as n → ∞.
P
4. null recurrent if and only if
The following is the famous renewal theorem.
Theorem 3.4 (The renewal theorem).

3.2. PROOF OF THE RENEWAL THEOREM 27
Let λ be a recurrent and aperiodic renewal event and let
nfn = F 0 (1)
X
µ=
be the mean inter-occurrence time. Then
lim un = µ−1 .
n→∞
The proof of the Renewal Theorem is rather involved. We will put it as

a section here. To the relief of many, this section will only for the sake of
those who are nerdy enough in mathematics.
3.2 Proof of the Renewal Theorem

I am using my own language to restate the results of Feller (1968, page 335).
First, the result can be stated without the renewal event background.
Equivalent result: Let f0 = 0, f1 , f2 , . . . be a sequence of non-negative
P
numbers such that fn = 1, and 1 is the greatest common divisor of these
n for which fn > 0. Let u0 = 1 and
un = f0 un + f1 un−1 + · · · + fn u0
for all n ≥ 1.
Then un → µ−1 as n → ∞ where µ = nfn (and µ−1 = 0 when µ = ∞).
P
♦
Note this definition of un does not apply to the case of n = 0. Otherwise,
the u-sequence would be a convolution of f-sequence and itself.
The implication of being aperiodic is as follows. Let A be the set of all
integers for which fn > 0, and denote by A+ the set of all positive linear
combinations
p1 a1 + p2 a2 + · · · + pr ar
of numbers a1 , . . . , ar in A.
Lemma 3.1
There exists an integer N such that A+ contains all integers n > N . ♦

In other words, if the greatest common divisor of a set of positive integers
is 1, then the set of their linear combinations contains all but finite number
positive integers. To reduce the burden in math respect, this result will not
be proved here. Some explanation will be given in class.
The next lemma has been re-stated in different words.
Lemma 3.2
Suppose we have a sequence of sequences: [ {(am,n )∞ ∞

m=1 }n=1 ] such that 0 ≤
am,n ≤ 1 for all m, n. There exists a sequence {ni : i = 1, 2, . . .} such that
limi→∞ am,ni exists for each m. ♦
Again, the proof will be omitted. This result is also used when proving
a result about the relationship between convergence in distribution and the
convergence of characteristic functions.
Lemma 3.3
Let {wn }∞
n=−∞ be a doubly infinite sequence of numbers such that 0 ≤ wn ≤ 1
and ∞ X
wn = fk wn−k
k=1
for each n. If w0 = 1 then wn = 1 for all n. ♦

To be more explicit, the sequence fn is assumed to have the same property
as the sequence introduced earlier.
Proof: Since all numbers involved are non-negative and wn ≤ 1, we have
that for all n ∞ X X
wn = fk wn−k ≤ fk = 1.
k=1
Applying this conclusion to n = 0 with the condition w0 = 1, we have

w−k = 1 whenever fk = 0.
Recall the definition of A, the above conclusion can be restated as
w−a = 0 whenever a ∈ A.
For each a ∈ A, using the same argument:

∞
X X
wn = fk wn−k ≤ fk = 1
k=1
to n = a, we find w−a−k = 1 whenever both a, k ∈ A.

It can then be strengthened to w−a = 1 for any a ∈ A+ by induction.
An implication is then: there exists an N such that
a−N = 1, a−N −1 = 1, a−N −2 = 1, . . . .
Using
∞
X X
wn = fk wn−k ≤ fk = 1
k=1
again for n = −N + 1, we get w−N +1 = 1.

Repeat it, we get w−N +2 = 1. Repeat it again and again, we find wn = 1
for all n. ♦
Finally, we prove the renewal theorem.
Proof of Renewal Theorem: Our goal is to prove the limit of un exists
and equals µ−1 . Since the items in this sequence are bounded, a subsequence
can be found such that this subsequence converges to some number η. As-
sume this η is the upper limit.
Notationally, let us denote it as u(iv ) → η as v → ∞.
Let un,v be a double sequence such that for each v, un,v = un+iv when
n + iv ≥ 0; and un,v = 0 otherwise. In other words, it shifts the original
sequence by iv positions toward left, and further adding zeroes to make a
double sequence.
For example, suppose i5 = 10, then we have
. . . , u−2,5 = u8 , u−1,5 = u9 , u0,5 = u10 , u1,5 = u11 , . . . .
If we set n = 0 and let v increases, the resulting sequence is
. . . , u(0, 0) = u(i0 ), u( 0, 1) = u(i1 ), u( 0, 2) = u(i2 ), . . . .
Hence, limv→∞ u0,v = η.

For the sequence with n = ±1, they look like
. . . , u(1, 0) = ui0 +1 , u(1, 1) = ui1 +1 , u(1, 2) = ui2 +1 , . . . ;

. . . , u(−1, 0) = ui0 −1 , u(−1, 1) = ui1 −1 , u(−1, 2) = ui2 −1 . . . ;
with the additional claus that if n < −iv , u(n, v) = 0.
In summary, for each n = 0, ±1, ±2, . . ., we have a sequence u(n, v) in v.
By the earlier lemma, it is possible to find a subsequence in v, say vj , j =
1, 2, . . . such that
lim un,vj
j→∞
exists for each n. Call this limit wn .
Since for all n > −v, we have
∞
X
u(n, v) = fk u(n − k, v).
k=1
Taking limit along the path of v = vj , we get
∞
X
wn = fk wn−k .
k=1
By the other lemma, we then get wn = η for all n.
After such a long labour, we have only achieve that the limit of each
un,vj = uvj +n , j = 1, 2, . . . are the same. What we wanted, however, is that
the limit of un exists and equal µ−1 .
To aim a bit low, let us show that η = µ−1 . This is easy when µ = ∞.
Let
ρk = fk+1 + fk+2 + · · ·
for k = 0, 1, . . ., we have sumρk = µ. We should have seen it before, but if you
do not remember, it is a good opportunity for us to redo it with generating
function concept. Another note is that
1 − ρk = f1 + f2 + · · · + fk .
Let us now list the defining relations as follows:
u1 = f1 u0 ;
u2 = f2 u0 + f1 u1 ;
u3 = f3 u0 + f2 u1 + f1 u2 ;
... = ...;
uN = fN u0 + fN −1 u1 + fN −2 u2 + · · · + f1 uN −1 .
Adding up both sides, we get

N
X
uk = (1 − ρN )u0 + (1 − ρN −1 )u1 + (1 − ρN −2 )u2 + · · · + (1 − ρ1 )uN −1 .
k=1
This is the same as
uN = (1 − ρN )u0 − ρN −1 u1 − ρN −2 u2 − · · · − ρ1 uN −1 .
Noting that u0 = 1, we have
ρN u0 + ρN −1 u1 + · · · + ρ1 uN −1 + ρ0 uN = 1. (3.1)
Recall that
lim un,vj = wn = η
j→∞
for each n. At the same time, un,vj = uvj +n is practically true for all n (or
more precisely for all large n). Let N = vj , and let j → ∞, (3.1), implies
X
η×{ ρk } = 1.
That is, either η = µ−1 when µ is finite, or 0 otherwise.

Even with this much trouble, our proof is not complete yet. By definition,
η is the upper limit of un . Is it also THE limit of un ? If its limit exists, then
the answer is yes. If µ = ∞, then answer is also yes because the lower limit
cannot be smaller than 0. Otherwise, we have more work to do.
If the limit of un does not exist, it must has a subsequence whose limit
exists and smaller than η. Let it be η0 < η. Being the upper limit of un ,
it implies that given any small positive quantity , for all large enough n,
un ≤ η + . Let us examine the relationship:
ρN u0 + ρN −1 u1 + · · · + ρ1 uN −1 + ρ0 uN = 1
again. By letting some uk replace by a larger value 1, it becomes

−r−1
NX N
X −1
ρN −k + ρN −k uk + ρ0 uN ≥ 1.
k=0 k=N −r
By including more terms in the first summation, it becomes

∞
X N
X −1
ρk + ρN −k uk + ρ0 uN ≥ 1.
k=r+1 k=N −r
Replacing uk by a larger value η + which is true for large k, we have

∞
X r
X
ρk + (η + ) ρk + ρ0 uN ≥ 1.
k=r+1 k=1
P∞
Further, since k=0 ρk = µ, we get
∞
X
ρk + (η + )(µ − ρ0 ) + ρ0 uN ≥ 1.
k=r+1
Re-organize terms slightly, we get

∞
X
ρk + (η + )µ + ρ0 (uN − η − ) ≥ 1.
k=r+1
At last, let N go to infinite along the subsequence with limit η0 , we get

∞
X
ρk + (η + )µ + ρ0 (η0 − η − ) ≥ 1.
k=r+1
Since this result applies to any r and , let r → ∞ and → 0, we get
ηµ + ρ0 (η0 − η) ≥ 1.
Because ηµ = 1, we must have η0 − η ≥ 0. Combined with the fact that η is

the upper limit, we must have η0 = η or there will be a contradiction.
Since the limit of any subsequence of un must be the same as η, the limit
of un exists and equals η.
Thus, we finally proved the Renewal Theorem.
3.3 Properties of Random Walks by Gener-

ating Functions
.
3.3. PROPERTIES 33
3.3.1 Quick derivation of some generating functions

The generating functions for certain sequences:
For “first returning to 0”: F (s) = 1 − (1 − 4pqs2 )1/2 .
For “returning to 0”: U (s) = (1 − 4pqs2 )−1/2 .
For “first passage of 1”: Λ(s) = (2qs)−1 [1 − (1 − 4pqs2 )1/2 ].
Consequences:
P (ever returns to 0) = 1 − |p − q|.
P ( ever rearches 1) = min(1, p/q).
(We will go through a lot of details in the class)
3.3.2 Hitting time theorem

.
Hitting time theorem is more general than what has been shown. Consider
the random walk such that
Sn = X1 + X2 + · · · + Xn ,
P (X ≤ 1) = 1
and P (X = 1) > 0. This is so called right-continuous random walk. The

random walk cannot skip over (up-ward) any state without stepping on it.
Let Tb be the first time when Sn = b for some b > 0 and assume S0 = 0.
It has been shown for simple random walk that
b
fb (n) = P (Tb = n) = P (Sn = n).
n
Let us see why this is still true for the new random walk we have just
defined.
For this purpose, define
Fb (z) = E(z Tb )
which is the generating function of Tb . It may happen that P (Tb = ∞) > 0,

in which case, it is more precise to define
n
z n P (Tb = n).
X
Fb (z) = lim
n→∞
i=1
We also define
G(z) = E(z 1−X1 ).
Since X1 can be negative with large absolute values, work with generating
function of 1−X1 makes sense. Since 1−X1 is non-negative, G is well defined
for all |z| ≤ 1. Please note that our G here is a bit different from G in the
textbook.
The purpose of using letter z, rather than s, is to allow z to be a complex
number. It seems that defining a function without a value at z = 0 calls for
some attention. We will pretend z is just a real number for illustration.
If we ignore the mathematical subtlety, our preparations boil down to have
defined generating functions of Tb and X1 . Due to the way Sn is defined, the
“generating function” of Sn is determined by [G(z)]n . Hence the required
hitting time theorem may be obtained by linking G(z) and Fb (z).
This step turns out to be rather simple. It is seen that
Fb (z) = [F1 (z)]b
as the random walk cannot skip a state without landing on it. We take notice
that this relationship works even when b = 0.
Further,
F1 (z) = E[E(z T1 |X1 )] = E[z 1+T1−X1 |X1 ] = zE{[F1 (z)]1−X1 } = zG(F1 (z)).
Denote w = F1 (z), this relation can then be written as

w
z=
G(w)
That is, it is known that w = F1 (z). At the same time, its relationship
with w is also completely determined by the above equation. The two versions
must be consistent with each other. The result to be proved is a consequence
of this consistency.
It turns out there is a complex analysis theorem for inversing z = w/G(w).
3.3. PROPERTIES 35
Theorem 3.5 Lagranges’ inversion formula
Let z = w/f (w) where f (w) is an analysis function (has derivative in complex
analysis sense) in a neighborhood of w = 0. (and a one-to-one relationship
between z and w in this neighborhood.). If g is infinitely differentiable, (but
I had the impression being analytic implies infinitely differentiable), then
∞
zn dn−1 0
" #
[g (u){f (u)}n ]
X
g(w(z)) = g(0) + .
n=1 n! dun−1 u=0
(See Kyrala (1972), Applied functions of a complex variable. Page 51 for

a close resemble of this result. Related to Cauchy integration over a closed
path). ♦
b
Apply this result with f (w) being our G(w) and g(w) = w , we have
∞
zn dn−1
" #
b
[bub−1 Gn (u)]
X
[F1 (z)] = .
n=1 n! dun−1 u=o
Let us not forget that [F1 (z)]b is the “probability” generating function of Tb ,
and Gn (u)/z n is the probability generating function of Sn .
Matching the coefficient of z n , we have
dn−1
" #
n!P (Tb = n) = [bub−1 Gn (u)] .
dun−1 u=0
Notice the latter is (n − 1)! times of the coefficient of un−1 in the power
expansion of bub−1 Gn (u), which in turn, is the coefficent of un−b in the power
series expansion of bGn (u), and which is the coefficient of u−b in the expansion
of bu−n Gn (u).
Now, we point out that u−n Gn (u) = Eu−Sn . That is, (n − 1)! times of
this coefficient is b(n − 1)!P (Sn = b). Hence we get the result. ♦
3.3.3 Spitzer’s Identity

Here we discuss the magical results of Spitzer’s identity. I kind of believe
that a result like this can be useful in statistics. Yet I find that it is still too
complex to be practical at the moment.
Theorem 3.6
Assume that Sn is a right-continuous random walk, and let Mn = max{Si :

0 ≤ i ≤ n} be the maximum of the walk up to time n. Then for |s|, |t| < 1,
∞ ∞
!
n Mn 1 +
tn E(sSn )
X X
log t E(S ) =
n=0 n=1 n
where Sn+ = max{0, Sn } (which is the non-negative part of Sn ). ♦

Proof: We keep the notation fb (n) = P (Tb = n) as before. To have Mn = b,
it has to land on b at some time j between 1 and n, and at the same time,
the walk does not land on b + 1 within another n − j steps. Thus,
n
X
P (Mn = b) = P (Tb = j)P (T1 > n − j)
j=0
Xn
= P (Tb = j)P (T1 > n − j).
j=0
Multiply both sides by sb tn and sum over b, n ≥ 0. The left hand side is
tn E(sMn )
X
n=0
1−F1 (t)
tn P (T1 > n) =
P
Recall 1−t
. The right hand side is, by convolution
relationship,
∞ ∞
b 1 − F1 (t)
Tb 1 − F1 (t) X
sb [F1 (t)]b
X
s E(t ) =
b=0 1−t 1 − t b=0
1 − F1 (t)
= .
(1 − t)(1 − sF1 (t))
Denote this function as D(s, t).
The rest contains mathematical manipulation: By Hitting time theorem,
n
X
nP (T1 = n) = P (Sn = 1) = P (T1 = j)P (Sn−j = 0)
j=0
and we get generating function relationship as
tF10 (t) = F1 (t)U (t).

3.3. PROPERTIES 37
(Recall U (s) is the g.f. for returning to 0).

With this relationship, we have
∂ −sF10 (t)
log[1 − sF1 (t)] =
∂t 1 − sF1 (t)
s
sk [F1 (t)]k
X
= − F1 (t)U (t)
t k
∞
sk+1
[F1 (t)]k+1 U (t)
X
= −
k=0 t
∞
sk
[F1 (t)]k U (t)
X
= −
k=1 t
∞
sk
[F1 (t)]k U (t)
X
= −
k=1 t
Notice n
X
P (Sn = k) = P (Tk = j)P (Sn−j = 0).
j=0
Thus, the generating function [F1 (t)]k U (t) is in fact a generating function for
the sequence of P (Sn = k) in n. That is,
∞
[F1 (t)]k U (t) = tn P (Sn = k).
X
n=0
Notice that the sum is over n.

It therefore implies
∞ ∞
∂
tn−1 sk P (Sn = k).
X X
log[1 − sF1 (t)] = −
∂t n=1 k=1
So, we get
∂ ∂ ∂ ∂
log D(s, t) = − log(1 − t) + log[1 − F1 (t)] − log[1 − sF1 (t)]
∂t ∂t ∂t ∂t
∞ ∞ ∞
!
n−1 k
X X X
= t 1− P (Sn = k) + s P (Sn = k)
i=1 k=1 k=1
∞ ∞
!
n−1 k
X X
= t P (Sn ≤ 0) + s P (Sn = k)
i=1 k=1
∞
+
tn−1 E(sSn ).
X
=
i=1
Now integrate with respect to t to get the final result.

♦
Remark: I do not see why this final result is more meaningful than the
intermedian result. I am wondering if any of you can offer some insight.
3.3.4 Leads for tied-down random walk

Let us now go back to the simple random walk such that S0 = 0 and P (Xi =
1) = p, P (Xi = −1) = q = 1 − p.
Define L2n as the number of steps when the random walk was above the
line of 0 (but is allowed to touch 0). We have an arc-sin law established for
this random variable already. This time, we investigate its property when
S2n = 0 is given.
Theorem 3.7 Leads for tied-down random walk.
For the simple random walk S,

1
P (L2n = 2k|S2n = 0) = ,
2n + 1
for k = 0, 1, . . . , n. ♦
Remark: we may place more possibility at L2n = n. It should divide the
time evenly above and below zero. This theorem, however, claims that a
uniform distribution is the truth. Note also that the result does not depend
on the value of p.
Proof: Define T0 be the first time when S is zero again. Given T0 = 2r
for some r, L2r can either be 0 or 2r as the simple random walk does not
cross zero. Making use of the results on the size of λ2r , and f2r , it turns out
the conditional distribution is placing half and half probabilities on 0 and 2r.
Thus,
1 1
E(sL2r |S2n = 0, T0 = 2r) = + s2r .
2 2
Now we define G2n (s) = E[S |S2n=0 and F0 (s) = E(S T0 ). Further, let
L2n
∞
t2n P (S2n = 0)G2n (s).
X
H(s, t) =
n=0
3.4. BRANCHING PROCESS 39
Conditioning on T0 , we have
n
E[S L2n |S2n = 0, T0 = 2r]P (T0 = 2r|S2n = 0)
X
G2n (s) =
r=1
n
1
( s2r )P (T0 = 2r|S2n = 0).
X
=
r=1 2
Also
P (T0 = 2r)P (S2n−2r = 0)
P (T0 = 2r|S2n = 0) = .
P (S2n = 0)
Hence,
1
H(s, t) − 1 = H(s, t)[F0 (t) + F0 (st)].
2
Recall that
F0 (t) = (1 − 4pqt2 )−1/2 ,
one obtains
2
H(s, t) = √ √
1 − t + 1 − s2 t2
2
√ √
2[ 1 − s2 t2 − 1 − t2 ]
=
t2 (1 − s2 )
∞
1 − s2n+2
!
2n
X
= t P (S2n = 0) .
n=0 (n + 1)(1 − s2 )
Compare the coefficient of t2n in two expansions of H(s, t), we deduce

n
(n + 1)−1 s2k .
X
G2n (s) =
k=0
3.4 Branching Process

We only consider a simple case where Z0 = 1 representing a species starts
from a single individual. It is assumed that each individaul will give birth of
random number of offsprings independently of each other. We call it family
size. We assume family sizes have the same distribution with probability
generating function G(s). Let µ and σ 2 be the mean and variance of the
family size.
Let the population size of the nth generation be called Zn .
It is known that
E[Zn ] = µn
σ 2 (µn − 1) n−1
V ar(Zn ) = µ .
µ−1
These relations can be derived from the identity:
Gn (t) = G(Gn−1 (t))
where Gn (t) = E(tZn ). The same identity also implies that the probability of
ultimate extinction η = lim P (Zn = 0) is the smallest non-negative solution
of the equation
x = G(x).
Because of this, it is known that:
1. if µ < 1, η = 1;
2. if µ > 1, η < 1;
3. if µ = 1 and σ 2 > 0, then η = 1;
4. if µ = 1 and σ 2 = 0, then η = 0.
Discussion will be given in class.
3.5 Summary
The biggest deal of this chapter is the proof of the Renewal Theorem. The
proof is not much related to generating function at all. One may learn a lot
from this proof. At the same time, it is okay to choose to ignore the proof
completely.
One should be able to see the beauty of generating functions when han-
dling problems related to the simple random walk and the branching process.
At the same time, none of these discussed should be new to you. You may
3.5. SUMMARY 41
realize that two very short sections on the simple random walk and brach-
ing process are rich in content. Please take the opportunity to plant these
knowledge firmly in you brain if they were not so before.
Chapter 4
Discrete Time Markov Chain
A discrete time Markov chain is a stochastic process consists of a countable

number of random variables arranged in a sequence. Usually, we name these
random variables as X0 , X1 , X2 , . . .. To qualify as a Markov chain, its state
space, S, which is the set of all possible values of these random variables,
must be countable, In addition, it must have Markov property:
P (Xn+1 = j|Xn = i, Xn−1 = in−1 , . . . , X1 = i1 , X0 = i0 ) = P (Xn+1 = j|Xn = i).
When this transition probability does not depend on n, we say that the
Markov chain is time homogeneous.
The Markov chain discussed in this course will be assumed time homoge-
neous unless otherwise specified. In this case, we use notation
pij = P (X1 = j|X0 = i)
and P for the matrix with the i, jth entry being pij .
It is known that all entries of the transition matrix are non-negative, and
its row sums are all 1.
Further, let P(m) be the m-step transition matrix defined by
(m)
pij = P (Xm = j|X0 = i).
Then they satisfy the Chapman-Kolmogorov equations:
P(m+n) = P(m) P(n)
43
44 CHAPTER 4. DISCRETE TIME MARKOV CHAIN
for all non-negative positive integers m and n. Here, we denote P(0) = I.

Let µ(m) be the row vector so that its ith entry
(m)
µi = P (Xm = i).
It can be seen that µ(m+n) = µ(m) Pn .

Both simple random walk and branching process are special cases of dis-
crete time Markov chain.
Example 4.1 Markov’s other chain.
Note that the Chapman-Kolmogorov equation is the direct consequence of

the Markov property. Suppose that a discrete time stochastic process has
countable state space, and its transition probability matrix defined in the
same way as for the Markov chain, satisfies the Chapman-Kolmogorov equa-
tion. Does it have to be a Markov chain?
In mathematics, we can offer a rigorous proof to show this is true, or offer
an example a stochastic process with this property which is not a Markov
chain. It turns out the latter is the case.
Let Y1 , Y3 , Y5 , . . . be a sequence of iid random variables such that
1
P (Y1 = 1) = P (Y1 = −1) = .
2
Let Y2k = Y2k−1 Y2k+1 for k = 1, 2, . . .. We hence have a stochastic process
Y1 , Y2 , Y3 , . . . which has state space {−1, 1}.
Note that
2
E[Y2k Y2k+1 ] = E[Y2k−1 Y2k+1 ] = E[Y2k−1 ] = 0.
Since these random variables take only two possible values, having correlation
0 implies independence. All other non-neighbouring pairs of random variables
are also independent of each other by definition.
Thus,
1
P (Xm+n = j|Xn = i) = P (Xm+n = j) =
2
for any i, j = ±1. Thus, the m-step transition matrix is
1 1
!
(m) 2 2
P =P= 1 1
2 2
4.1. CLASSIFICATION OF STATES AND CHAINS 45
which is idempotent. It implies P(m+n) = P(m) P(n) for all m and n.

That is, the Chapman-Kolmogorov equation is satisfied. However,
1
P (Y2k+1 = 1|Y2k = −1) =
2
but
P (Y2k+1 = 1|Y2k = −1, Y2k−1 = 1) = 0.
Thus, this process does not have Markov property and is not a Markov chain.
♦
Although this example shows that a stochastic process with transition
probability matrices satisfying the Chapman-Kolmogorov equations is not
necessarily a Markov chain. When we are given a set of stochastic matrices
satisfying Chapman-Kolmogorov equations, it is always possilbe to construct
a Markov chain with this set of stochastic matrices as its transition proba-
bility matrices.
4.1 Classification of States and Chains

We use i, j, k and so on to denote states in state space.
(m)
If there exists an integer m ≥ 0 such that pij > 0, we say that j is
reachable from i.
If i is reachable from j and j is reachable from i, then we say that i and
j communicate.
It is obvious that communication is an equivalence relationship. Thus, the
state space is partitioned into classes so that any pair of states in the same
class communicate with each other. Any pair of states from two different
classes do not communicate.
It is usually very simple to classify the state space. In most examples, it
can be done by examining the transition probability matrix of the Markov
chain.
Now let us consider individual classes. Let G be a class. If G consists of
all states of the Markov chain, we say that the chain is irreducible.
If for any state i ∈ G and state j 6∈ G, pij = 0. Then the Markov chain
can never leave class G once it enters. That is, once the value of Xn is in G
for some n, then the values of Xn+m ’s are all in G when m ≥ 0. A class G is
closed when it has this property. Notice that given Xn ∈ G for some n, the
Markov chain {Xn , Xn+1 , . . .} has effectively G as its state space. Thus, the
state space is reduced when G is a true subset of the state space.
In contrast to a closed class, if there exist states i ∈ G and j 6∈ G such
that pij > 0, then the class is said open.
4.2 Class Properties

It turns out that the states in the same class have some common properties.
We call them class properties.
Due to Markov property, the event Xn = i for any i is a renewal event.
Some properties that describe renewal events are positive or null recurrent,
transient, periodicity.
Definition 4.1 Property of renewal events
1. A renewal event is recurrent(persistent) if the probability for its future

occurrence is 1, given its occurrence at time 0.
2. If a renewal event is not recurrent, then it is called transient.
3. If the expected waiting time for the next occurrence of a recurrent

renewal event is finite, then the renewal event is positive recurrent.
Otherwise, it is null-recurrent.
4. Let A be a renewal event and T = {n : P (A occurs at n|A occurs at 0) >

0}. The period of A is d, the greatest common divisor of T . If d = 1,
then A is aperiodic.
♦
Due to the fact that entering a state is a renewal event, some properties
of renewal event can be translated easily here. Let
∞
sn pij (n)
X
Pij (s) =
i=1
4.2. CLASS PROPERTIES 47
as the generating function of the n-step transition probability. Define
fij (n) = P (X1 6= j, . . . , Xn−1 6= j|X0 = i)
be the probability of entering j from i for the first time at time n. Let its
corresponding generating function be Fij (s). We let
X
fij = fij (n)
n
for the probability that the chain ever enters state j starting from i. Note
that fij = Fij (1).
Lemma 4.1
(a): Pii (s) = 1 + Fii (s)Pii (s).

(b): Pij (s) = Fij (s)Pjj (s) if i 6= j. ♦
Proof: The first property is the renewal equation. The second one is the
delayed renewal equation. ♦
Based on these two equations, it is easy to show the following.
Theorem 4.1
(a) State j is recurrent if and only if j pjj (n) = ∞.

P
Consequently,
n pij (n) = ∞ when j is reachable from i
P
pjj (n) < ∞.

P
(b) State j is transient if and only if j Consequently,
n pij (n) < ∞.
P
♦
Proof: Using the generating function equations.
Theorem 4.2
. All properties in Definition 4.1 are class properties. ♦

Proof: Suppose i and j are two states in the same class. State i is recurrent,
and we want to show that state j is also recurrent.
Let pij (m) be the m-step transition probability from i to j. Hence,

m pii (m) = ∞. Since i and j communicate, there exist m1 and m2 such that
P
pij (m1 ) > 0 and pji (m2 ) > 0. Hence, pjj (m+m1 +m2 ) ≥ pji (m2 )pii (m)pij (m1 ).
Consequently,
X X
pjj (m) ≥ pjj (m + m1 + m2 )
m m
X
≥ pji (m2 ){ pii (m)}pij (m1 )
m
= ∞.
That is, j is also a recurrent state.
The above proof also implies that if i is transient, j cannot be recurrent.
Hence, j is also transient.
We delegate the positive recurrentness proof to the future.
At last, we show that i and j must have the same period.
Define Ti = {n : pii (n) > 0} and similarly for Tj . Let di and dj be the
periods of state i and j. Note that if n1 , n2 ∈ Ti , then an1 + bn2 ∈ Ti for any
positive integers a and b. Since di is the greatest common divisor of Ti , there
exist integers a1 , . . . , am and n1 , . . . , nm ∈ Ti such that
a1 n1 + a2 n2 + · · · + am nm = di .
By grouping positive and negative coefficients. it is easy to see that the
number of items m can reduced to 2. Thus, we assume that it is possible to
find a1 , a2 and n1 , n2 such that
a1 n1 + a2 n2 = di .
We can then further pick non-negative coefficients such that
a11 n1 + a12 n2 = kdi , a21 n1 + a22 n2 = (k + 1)di
for some positive integer k.
Let m1 and m2 be the number of steps the chain can go and return from
state i to state j. Then, both
m1 + m2 + kdi , m1 + m2 + (k + 1)di ∈ Tj .
Thus, we must have dj divides di . The reverse is also true by symmetry.
Hence di = dj . ♦
4.3. STATIONARY DISTRIBUTION 49
4.2.1 Properties of Markov Chain with Finite Number

of States
If a Markov chain has finite state space, then one of the states will be visited
infinite number of times over infinite time horizon. Thus, at least one of
them will be recurrent (persistent). We can prove this result rigorously.
Lemma 4.2
If the state space is finite, then at least one state is recurrent and all recurrent
states are positive recurrent. ♦
Proof: A necessary condition for a recurrent event to be transient is un → 0

as n increases. In the context of Markov chain, it implies that state j is
transient implies pij (n) → 0. Because
X
1= pij (n),
j
we cannot have pij (n) → 0 for all j when the state space is finite. Hence, at
least one of them is recurrent. ♦
Intuitively, the last conclusion implies that there is always a closed class
of recurrent states for a Markov chain with finite state space. Since the
Markov chain cannot escape from the finite class, the average waiting time
for the next visit will be finite. Thus, at least one of them is recurrent and
hence positive recurrent.
We skip the rigorous proof for now.
4.3 Stationary Distribution and the Limit The-

orem
A Markov chain consists of a sequence of discrete random variables. For
convenience, they are regarded as non-negative integers.
For each n, Xn is a random variable whose distribution may be different
from that of Xn−1 but is involved from it. When n is large, the dependence of
the distribution of Xn on that of X0 gets weaker and weaker. It is therefore
possible that the distribution stabilizes and it may hence have a limit. This
limit turns out to exist in many cases and it is related to so called stationary
distribution.
Definition 4.1
The vector π is called a stationary distribution of a Markov chain if π has

entries πj , j ∈ S such that
(a) πj ≥ 0 for all j and j πj = 1;
P
(b) π = πP where P is the transition probability matrix. ♦

It is seen that if the probability function of X0 is given by a vector called
α, then the probability function of Xn is given by
β(n) = αPn .
Consequently, if the distribution of Xn is given by π, then all distributions

of Xn+m are given by π. Hence it gains the name of stationary.
It should be noted that if the distribution of Xn is given by π, the distri-
bution of Xn−m does not have to be π.
Lemma 4.3
If the Markov chain is irreducible and recurrent, there exists a positive root
x of the equation x = xP , which is unique up to a multiplicative constant.
The chain is non-null if i xi < ∞, and null if i xi = ∞. ♦
P P
Proof: Assume the chain is recurrent and irreducible. For any states k, i ∈
S,
X
Nik = I(Xn = i, Tk ≥ n)
n
with Tk being the time of the first return to state k. Note that Tk is a well
defined random variable because p(Tk < ∞) = 1.
Define ρi (k) = E[Nik |X0 = k] which is the mean number of visits of the
chain to state i between two successive visits of state k.
By the way, if j is recurrent, then µj = E[Tj ] is called the mean recurrent
time. It is allowed to be infinity when the summation in the definition of
expectation does not converge. When j is transient, we simply define µj = ∞.

Thus, being positive recurrent is equivalent to claim that µj < ∞.
Note that ρk (k) = 1. At the same time,
X
ρi (k) = P (Xn = i, Tk ≥ n|X0 = k).
n
It will be seen that the vector ρ with ρi (k) as its kth component is a base for
finding the stationary distribution.
We first show that ρi (k) are finite for all k. Let
lki = P (Xn = i, Tk ≥ n|X0 = k),
the probability that the chain reaches i in n steps but with no intermediate
return to the starting point k.
With this definition, we see that
fkk (m + n) ≥ lki (m)fik (n)
in which the right hand side is one of the sample paths taking a roundtrip
from k back to k in m + n steps, without visiting k in intermediate steps.
Because the chain is irreducible, there exists n such that fik (n) > 0. Now
sum over m on two sides and we get
X X
fkk (m + n) ≥ fik (n) lki (m).
m m
Since m fkk (m + n) < ∞ and fik (n) > 0, we get m lki (m) < ∞. That is,
P P
ρi (k) < ∞.
Now we move to the next step. Note that lki (1) = pki , the one-step
transition probability. Further,
X
lki (n) = P (Xi , Xn−1 = j, Tk ≥ n|X0 = k)
j:j6=k
X
= lkj (n − 1)pji
j:j6=k
for n ≥ 2. Now sum over n ≥ 2, we have

 
X X 
ρi (k) = pki + 
lkj (n − 1) pji
j:j6=k n≥2
X
= ρk (k)pki + ρj (k)pji
j:j6=k
because ρk (k) = 1. We have shown that ρ is a solution to xP = x.

This proves the existence part of the lemma. We prove uniqueness next.
First, if a chain is irreducible, we cannot have a non-negative solution for
xP = x with zero-components. If so, write x = (0, x2 ) and
" #
P11 P21
P = .
P21 P22
Then, we must have x2 P21 = 0. Since all components in x2 are positive, we
must have P21 = 0. This contradicts the irreducibility assumption.
If there are two solutions to xP = x who are not multiplicative of each
other, then we can obtain another solution with non-negative components
and at least one 0 entry. This will contradicts the irreducibility assumption.
(This proof works only if there are finite number of states. The more general
case will be proved later).
Due to the uniqueness, for any solution x, we must have for any state k,
X
µk = c xi .
Thus, if k is positive recurrent, we have xi < ∞.
P
This completes the proof of the lemma. ♦

We have an immediate corollary.
Corollary 4.1
If states i and j communicate and state i is positive recurrent, then so is
state j. ♦
One intermediate results in the proof of the theorem can be summarised
as another lemma.
Lemma 4.4
For any state k of an irreducible recurrent Markov chain, the vector ρ(k)
satisfies ρi (k) < ∞ for all i, and furthermore ρ(k) = ρ(k)P . ♦
The above lemma shows why we did not pay much attention on when
a Markov chain is positive recurrent. Once we find a suitable solution to
x = xP , we know whether it is positive recurrent or not immediately.
The next theorem summary these results to give a conclusion on the
existence of stationary distribution.
Theorem 4.3
An irreducible Markov chain has a stationary distribution π if and only if

all the states are positive recurrent; in this case, π is the unique stationary
distribution and is given by πj = µ−1j for each i ∈ S, where µj is the mean
recurrence time of j. ♦
Proof. We have already shown that if the chain is positive recurrent, then
the stationary distribution exists and is unique. If the chain is recurrent
but not positive recurrent, then due to the uniqueness, there cannot be an-
other solution x such that xi < ∞. Thus, there cannot exist a stationary
P
distribution or it violates the uniqueness conclusion.

Now if the chain is transient, we have pij (n) → 0 for all ij as n → ∞. If
a stationary distribution π exists, we must have
X
πj = πi pij (n) → 0
i
by dominate converges theorem. This contradicts the assumption that π is

a stationary distribution.
Thus, all left to be shown is whether πj = µ−1
j for each i ∈ S when the
chain is positive recurrent.
Suppose X0 has π as its distribution.. Then
∞
X
π j µj = P (Tj ≥ n|X0 = j)P (X0 = j) = P (Tj ≥ n, X0 = j).
n=1
However, P (Tj ≥ 1, X0 = j) = P (X0 = j) = πj and for n ≥ 2,
P (Tj ≥ n, X0 = j)
= P (X0 = j, Xm 6= j for all 1 ≤ m ≤ n − 1)
= P (Xm 6= j for 1 ≤ m ≤ n − 1) − P (Xm 6= j for 0 ≤ m ≤ n − 1)
= P (Xm 6= j for 0 ≤ m ≤ n − 2) − P (Xm 6= j for 0 ≤ m ≤ n − 1)
= an−2 − an−1
with am defined as is.

Thus, sum over n we obtain
πj µj = P (X0 = j) + P (X0 6= j) − lim an = 1 − lim an .

Since state j is recurrent,
an = P (Xm 6= j for all 1 ≤ m 6= n)
has limit 0. Hence we have shown πj = µ−1

j . ♦.
Now we come back for the uniqueness again. We skipped its proof earlier.
It turns out that the general proof is very tricky.
Suppose {Xn } is an irreducible and recurrent Markov chain with transi-
tion probability matrix P . Let x be a solution to x = xP . We construct
another Markov chain Yn such that
xj
qij (n) = P (Yn = j|Y0 = i) = pji (n)
xi
for all i, j ∈ S and n. It is easy to verify that qij make a proper transition
probability matrix. Further, it is obvious that {Yn } is also irreducible. By
checking the sum over n, it further shows that {Yn } is recurrent.
Let, for i 6= j,
lji (n) = P (Xn = i, Xm 6= j for 1 ≤ m ≤ n − 1|X0 = j).
We show that
gij (n) = P (Yn = j, Ym 6= j, for 1 ≤ m ≤ n − 1|Y0 = i)
satisfies
xj
gij (n) = lij (n).
xi
Let us examine the expressions of gij and lji . One is the probability of all
the pathes which start from j and end up at state i before they ever visit j
again. The other is the probability of all the pathes which start from i and
end up at state j without being there in between. If we reverse the time of
the second, then we are working with the same set of sample pathes.
For each sample path corresponding to lji (n), let us denote it as
j, k1 , k2 , . . . , kn−1 , i,
its probability of occurrence is
pj,k1 pk1 ,k2 · · · pkn−1 ,i .

The reverse sample path for gij (n) is i, kn−1 , . . . , k2 , k1 , j and its probability
of occurrence is
qi,kn−1 qkn−1 ,kn−2 · · · qk2 ,k1 qk1 ,j

xk xk xk xk
= pkn−1 ,i · · · pk1 ,k2 pj,k1 × n−1 n−2 · · · 1 j
xi xkn−1 x k 2 x k1
xi
= pk ,i · · · pk1 ,k2 pj,k1
xj n−1
Since it is true for every sample path, the result is proved.
P
Note that n gij (n) corresponds to the probability that the chain will ever
P
enter state j and the chain is recurrent and irreducible, we have n gij (n) = 1.
Consequently,
xj
= [ lij (n)]−1
X
xi n
and hence the ratio is unique.
4.3.1 Limiting Theorem

We have seen that an irreducible Markov chain has a unique stationary dis-
tribution when all states are positive recurrent. It turns out that the limits
of the transition probabilities pij (n) exist when n increases, provided they
are aperiodic.
When the chain is periodic, the limit does not always exist. For example,
if S = {0, 1} and p12 = p21 = 1, then
p11 (n) = p22 (n) = I(n is even).
Obviously, the limit does not exist.

When the limits exist, the conclusion is neat.
Theorem 4.4
For an irreducible aperiodic Markov chain, we have that
pij (n) → µ−1

j
as n → ∞ for all i and j. ♦

Remarks:
1. If the chain is transient or null recurrent, then it is known that pij (n) →
0 for all i and j. Since µj = ∞ in this case, the theorem is automatically
true.
2. If the chain is positive recurrent, then according to this theorem,
pij (n) → πj = µ−1 j .
3. This theorem implies that the limit of pij (n) does not depend on n. It
further implies that
P (Xn = j) → µ−1 j
irrespective of the distribution of X0 .

4. If {Xn } is an irreducible chain with period d, then {Xnd } is an aperiodic
chain. (Not necessarily irreducible). It follows that
pjj (nd) = P (Yn = j|Y0 = j) → dµ−1

j
as n → ∞.
Proof:
Case I: the Markov chain is transient. In this case, the theorem is true
by renewal theorem.
Case II: the Markov chain is recurrent. We could use renewal theorem.
Yet let us see another line of approach.
Let {Xn } be the Markov chain with the transition probability matrix P
under consideration. Let {Yn } be an independent Markov chain with the
same state space S and transition matrix P with {Xn }.
Now we consider the stochastic process {Zn } such that Zn = (Xn , Yn ),
n = 0, 1, 2, . . .. Its state space is S × S. Its transition probabilities are simply
multiplication of original transition probabilities. That is,
p(ij)→(kl) = pik pjl .
The new chain is still irreducible. If pik (m) > 0 and pjl (n) > 0, then
pik (mn)pjl (mn) > 0 and so kl is reachable from ij.
The new chain is still aperiodic. It states that if the period is one, then
pij (n) > 0 for all sufficiently large n. Thus, pij (n)pkl (n) > 0 for all large
enough n too.
Case II.1: positive recurrent. In this case, {Xn } has stationary distribu-
tion π, and hence {Zn } has stationary distribution given by {πi πj : i, j ∈ S}.
Thus by the lemma in the last section, {Zn } is also positive recurrent.
Assume (X0 , Y0 ) = (i, j) for some (i, j). For any state k, define
T = min{n : Zn = (k, k)}.
Due to positive recurrentness, P (T < ∞) = 1.

Implication? Sooner or later, Xn and Yn will occupy the same state. If so,
from that moment and on, Xn+m and Yn+m will have the same distribution.
If Yn has π as its distribution, so will Xn .
More precisely, starting from any pair of states (i, j), we have
pik (n) = P (Xn = k)

= P (Xn = k, T ≤ n) + P (Xn = k, T > n)
= P (Yn = k, T ≤ n) + P (Xn = k, T > n)
≤ P (Yn = k) + P (T > n)
= pjk (n) + P (T > n).
Due to the symmetry, we also have
pjk (n) ≤ pik (n) + P (T > n).
Hence
|pik (n) − pjk (n)| ≤ P (T > n) → 0
as n → ∞. That is,
pik (n) − pjk (n) → 0.
or if the limit of pjk (n) exists, it does not depend on j. (Remark: P (T >
n) → 0 is a consequence of recurrentness).
For the existence, we have
X
πk − pjk (n) = πi (pik (n) − pjk (n)) → 0.
i∈S
Case II.2: null recurrent. We can no longer claim P (T > n) → 0.

In this case, {Xn , Yn } may be transient or null recurrent.
If {Xn , Yn } is transient, then
P (Xn = j, Yn = j|X0 = i, Y0 = i) = [pij (n)]2 → 0.
Hence, the conclusion pij (n) → 0 remains true.

If {Xn , Yn } is null recurrent, the problem becomes a bit touchy. However,
P (T > n) → 0 stays. Thus, we still have the conclusion
pik (n) − pjk (n) → 0,
but we do not have a stationary distribution handy.

What we hope to show is that pij (n) → 0 for all i, j ∈ S. If this is not
true, then we can find a subsequence of n such that
pij (nr ) → αj
as r → ∞, and at least one of αj is not zero.

Let F be a finite subset of S. Then
X
lim pij (nr ) ≤ 1.
αj = r→∞
j∈F
This is true for any finite subset, which implies

X
α= αj ≤ 1.
i∈S
(Recall that S is countable).

Using the idea of one-step transition, we have
X X
pik (nr )pkj ≤ pij (nr + 1) = pik pkj (nr ).
k∈F k∈S
Let r → ∞, we have
X X
αk pkj ≤ pik αj = αj .
k∈F k∈S
Let F get large, we have X

αk pkj ≤ αj .
k∈S
The equality has to be true, otherwise,

X X X
αk = αk pkj
k∈S k∈S j∈S
X
= αk pkj
k,j∈S
X X
= [ αk pkj ]
j∈S k∈S
X
< αj
j∈S
which is a contradiction. However, when the equality holds, we would have

X
αk pkj = αj
k∈S
for each j ∈ S. Thus, {αj /α, j ∈ S} is a stationary distribution of the

Markov chain. This contradicts the assumption of the null recurrentness. ♦
Theorem 4.5
For any aperiodic state j of a Markov chain, pjj (n) → µ−1 j as n → ∞.

−1
Furthermore, if i is another state, then pij (n) → fij µj as n → ∞. ♦
Corollary 4.2
Let
n
1 X
τij (n) = pij (m)
n m=1
be the mean proportion of elapsed time up to the nth step during which the
chain was in state j, starting from i. If j is aperiodic, then
τij (n) → fij /µj
as n → ∞. ♦
4.4 Reversibility
Let {Xn : n = 0, 1, 2, . . . , N } be a (part) of a Markov chain whose transition
probability matrix is P and it is irreducible and positive recurrent so that π
is its stationary distribution.
Now let us define Yn = XN −n for n = 0, 1, . . . , N . Suppose all Xn has
distribution given by the stationary distribution π. It can be shown that
{Yn } is also a Markov chain.
Theorem 4.6
The sequence Y = {Yn : n = 0, 1, . . . , N } is a Markov chain with transition
probabilities
πj
qij = P (Yn+1 = j|Yn = i) = pji .
πi
♦
Proof We need only verify the Markov property. Other conditions for a
Markov chain are obvious.
P (Yn+1 = in+1 |Yk = ik , k = 0, . . . , n)
= P (Yk = ik , k = 0, . . . , n, n + 1)/P (Yk = ik , k = 0, . . . , n)
= P (XN −k = ik , k = 0, . . . , n, n + 1)/P (XN −k = ik , k = 0, . . . , n)
P (XN −n−1 = in+1 )P (XN −n = in |XN −n−1 = in+1 )
=
P (XN −n = in )
πin+1 pin+1 ,in
= .
πin
Since this transition probability does not depend on ik , k = 0, . . . , n − 1, the
Markov property is verified. ♦
Although Y in the above theorem is a Markov chain, it is not the same
as the original Markov chain.
Definition 4.1
Let {Xn : 0 ≤ n ≤ N } be an irreducible Markov chain such that it has sta-
tionary distribution π for all n. The chain is call reversible if the transition
matrices of X and its time-reversal; Y are the same, which is to say that
πi pij = πj pji
4.4. REVERSIBILITY 61
for all i, j. ♦
Why do we define the reversibility? One advantage is when a chain is

reversible, the transitions between i and j are balanced at equilibrium. This
provides us a convenient way to solve for the stationary distribution. We
have also used the idea to prove the uniqueness of the solution to π = πP .
Theorem 4.7
Let P be the transition matrix of an irreducible chain X and suppose that

there exists a distribution π such that πi pij = πj pji for all i, j ∈ S. Then
π is a stationary distribution of the chain. Furthermore, X is reversible in
equilibrium.
Chapter 5
Continuous Time Markov

Chain
It is more realistic to consider processes which do not have to evolve at

specified epochs. However, setting up continuous time stochastic processes
properly involves a lot of effort in mathematics. We now work on two special
continuous time stochastic processes first to motivate the continuous time
Markov chain.
5.1 Birth Processes and the Poisson Process

These processes may be called counting processes in general. We have a
process {N (t) : t ≥ 0} such that
(a) N (0) = 0, and N (t) ∈ {0, 1, 2, . . .},
(b) if s < t, then N (s) ≤ N (t).
What other detailed properties should we place on it?
Definition 5.1
A Poisson process with intensity λ is a process N = {N (t) : t ≥ 0} taking

values in S = {0, 1, 2, . . .} such that
(a) N (0) = 0; if s < t, then N (s) ≤ N (t),
63
64 CHAPTER 5. CONTINUOUS TIME MARKOV CHAIN

 λh + o(h)
if m = 1,


(b)P {N (t + h) = n + m|N (t) = n} =  o(h) if m > 1,
 1 − λh + o(h) if m = 0.

(c) if s < t, the random variable N (t) − N (s) is independent of the N (s).
♦
In general, we call (b) the individuality, and call (c) the independence
property of the Poisson process. It is well known as these specifications imply
that N (t) − N (s) has Poisson distribution with mean λ(t − s) for s < t.
Theorem 5.1
N (t) has the Poisson distribution with parameter λt; that is to say
(λt)j
P (N (t) = j) = exp(−λt), j = 0, 1, 2, . . . .
j!
♦
Proof: Denote pj (t) = P (N (t) = j). The properties of the Poisson process
lead to the equation
p0j (t) = λpj−1 (t) − λpj (t)
of j 6= 0; likewise
p00 (t) = −λp0 (t).
With the boundary condition pj (0) = I(j = 0), the equations can be solved
and the conclusion proved. ♦
We may view a counting process by recording the arrival time of the nth
event. For that purpose, we define
T0 = 0, Tn = inf{t : N (t) = n}.
The inter-arrival times are the random variables X1 , X2 , . . . given by
Xn = Tn − Tn−1 .
The counting process can be easily recontructed from Xn ’s too.

The following theorem is a familiar story.
5.1. BIRTH PROCESSES AND THE POISSON PROCESS 65
Theorem 5.2
The random variables X1 , X2 , . . . are independent, each having the exponen-

tial distribution with parameter λ. ♦
Proof: It can be shown by working on
n
X
P (Xn+1 > t + ti |Xi = ti , i = 1, 2, . . . , n)
i=1
by making use of the property of independent increment. ♦

There is a bit problem with this proof as the event we conditioning on has
zero probability. It could be made more rigorous with some measure theory
results.
The Poisson process is a very satisfactory model for redioactive emissions
from a sample of uranium-235 since this isotope has a half-life of 7×108 years
and decays fairly slowly. That is, we have a constant radio-active source in
a short time. For some other kinds of radio-active substances, the rate of
emission should depend on the number of detected emission already.
It can be shown that {N (t) : t ≥ 0} is a Poisson process if and only if
X1 , X2 , . . . are independent and identically distributed exponential random
variables. If the Xi has exponential distribution with rate λi instead, then
we have a general birth process.
Definition 5.2
A birth process with intensity λ0 , λ1 , . . . is a process {N (t) : t ≥ 0} taking

values in S = {0, 1, 2, . . .} such that
(a) N (0) ≥ 0; if s < t, then N (s) ≤ N (t),

 λn h + o(h)
if m = 1,


(b)P [N (t + h) = n + m|N (t) = n) = o(h) if m > 1,

 1 − λ h + o(h) if m = 0.

n
(c) if s < t, the random variable N (t) − N (s) is independent of the N (s).
♦
Here is a list of special cases:
(a) Poisson process. λn = λ for all n.

(b) Simple birth. λn = nλ. This is the case when each living individual
gives birth independently of others, and at the constant rate.
(c) Simple birth with immigration. λn = nλ+ν. In addition to the simple
birth, there is a steady source of immigration.
The differential equation we derived for the Poisson process can be easily
generalized. We can find two basic sets of them. Define pij (t) = P (N (s+t) =
j|N (s) = i). The boundary conditions are pij (0) = δij = I(i = j).
Forward system of equations:
p0ij (t) = λj−1 pi,j−1 (t) − λj pij (t)
for j ≥ i.
Backward system of equations:
p0ij (t) = λi pi+1,j (t) − λi pij (t)
for j ≥ i.
The forward equation can be obtained by computing the probability of
N (t + h) = j conditioning on N (t) = i. The backward equation is obtained
by computing the probability of N (t + h) = j conditioning on N (h) = i.
Theorem 5.3
The forward system has a unique solution, which satisfies the backward sys-
tem. ♦
Proof: First, it is seen that pij (t) = 0 whenever j < i. When j = i,
pii (t) = exp(−λj t) is the solution. Substituting into the foward equation,
we obtain the solution for pi,i+1 (t). Repeat this procedure implies that the
forward system has a unique solutions.
Using Laplace transformation reveals the structure of the solution better.
Define Z ∞
p̂ij (θ) = exp(−θt)pij (t)dt.
0
Then, the forward system becomes
(θ + λj )p̂ij (θ) = δij + λj−1 p̂i,j−1 (θ).

The new system becomes easy to solve. We obtain

1 λi λi+1 λj
p̂ij (θ) = ···
λj θ + λi θ + λi+1 θ + λj
for j > i. The uniqueness is determined by the inversion theorem for Laplace
transforms.
If πij (t)’s solve the backword systems, their corresponding Laplace trans-
forms will satisfy
(θ + λi )π̂ij (θ) = δij + λi π̂i+1,j (θ).
It turns out that these satisfying the forward system will also satisfy the
backward system here in Laplace transforms. Thus, the solution to the for-
ward equation is also a solution to the backward equation. ♦
An implicit conclusion here is: the backward equation may have many
solutions. It turns out that if there are many solutions to the backward
equation, the solution given by the forward equation is the minimum solution.
Theorem 5.4
If {pij (t)} is the unique solution of the forward system, then any solution
{πij (t)} of the backward system satisfies pij (t) ≤ πij (t) for all i, j, t. ♦
If {pij (t)} is the transition probabilities of the specified birth and death
process, then it must solve both forward and backward systems. Thus, the
solution to the forward system must be the transition probabilities. This
could be compared to the problem related to probability of ultimate extinc-
tion in the branching process. Conversely, the solution to the foward system
can be shown to satisfy the Chapman-Kolmogorov equations. Thus, it is a
relevent solution.
The textbook fails to demonstrate the why the proof of the uniqueness
for the forward system cannot be applied to the backward system. The
key is the assumption of pij (t) = 0 when j < i. When this restrict is
removed, the backward system may have multiple solutions. This restriction
reflects the existence of some continuous time Markov chains which have the
same transition probability matrices to some degree to the birth process.
Consequently, the text book should not have made use of this restriction in
proving the uniqueness of solution to the forward system.
Intuitively, we may expect that
X
pij (t) = 1
j∈S
for any solutions. If so, no solutions can be larger than other solutions and
hence the uniqueness is automatic. The non-uniqueness is exactly built on
this observation. This constraint does not hold for some birth processes.
When the birth rate increases with the population size fast enough, the
population size may rearch infinity in finite amount of time.
When the solution to the forward equation
X
pij (t) < 1
j∈S
for some t > 0 and i, it is possible then to construct another solution which
also satisfies the backward system. See Feller for detailed constructions.
In that case, it is possible to design a new stochastics process so that its
transition probabilities are given by this solution.
What is the probabilistic interpretation when
X
pij (t) < 1
j∈S
for some finite t, and i? It implies that within the period of t, the population
size has jumped or exploded all the way to infinity. Consequently, infinite
number of transitions must be have occurred. Recall the waiting time for the
next transition when N (t) = n is exponential with rate λn . Let Tn = ni=1 Xi
P
be the waiting time for the nth transition. Define T∞ = limn→∞ Tn .
Definition 5.3 Honest/dishonest
We call the process N honest if P (T∞ < ∞) = 0 and dishonest otherwise.
Lemma 5.1
Let X1 , X2 , . . . be independent random variables, Xn having the exponential

distribution with parameter λn−1 , and let T∞ = ∞
P
n=1 Xn . We have that
0 if ∞ −1
(
n=1 λn = ∞.
P
P (T∞ < ∞) =
1 if ∞ −1
n=1 λn < ∞
P
Proof: By definition of T∞ , we have

∞
λ−1
X
E[T∞ ] = n .
n=1
Hence, when ∞ −1
n=1 λn < ∞, E[T∞ ] < ∞ and P (T∞ < ∞) = 1. This implies
P
dishonesty.
If ∞ −1
n=1 λn = ∞, it does not imply T∞ = ∞ with any positive prob-
P
ability. If, however, P (T∞ < t) > 0 for any t > 0, then E[exp(−T∞ )] ≥
exp(−t)P (T∞ < t) > 0. We show that this is impossible under the current
assumption. Note that
N
Y
E[exp(−T∞ )] = lim E exp(−Xn )
N →∞
n=1
N
Y
= lim E exp(−Xn )
N →∞
n=1
N
(1 + λ−1 −1
Y
= lim n )
N →∞
n=1
N
(1 + λ−1 −1
Y
= lim [ n )] .
N →∞
n=1
According to mathematical analysis result, the product with infinite terms

equals infinity when ∞ −1
n=1 λn = ∞. (Check its logrithm). Hence, E[exp(−T∞ )] =
P
0 which contridicts the assumption E[exp(−T∞ )] > 0 made earlier. ♦

The following theorem is an easy consequence.
Theorem 5.5
P∞
The process N is honest if and only if n=1 λ−1
n = ∞.
5.1.1 Strong Markov Property

The Markov property for discrete time stochastic process states: given the
present (Xn = i), the future (Xn+m ’s) is independent of the past (Xn−k ’s).
The time represents the present is a non-random constant. In the example
of simple random walk, we often claim that once the random walk returns
to 0, it renews itself. That is, the future behavior of the random walk does
not depend on how the random walk got into 0 nor when it returns to 0. We
may notice that this notion is a bit different from the Markov property. The
“present time” is a random variable.
This example implies that the Markov property is true for some randomly
selected time. The question is what kind of random variable can be used for
this purpose? Suppose {N (t) : t ≥ 0} is a stochastic process and T is
a random variable. If the event T ≤ t is completely determined by the
knowledge of {N (s) : s ≤ t}, then we call it a stopping time for stochastic
process {N (t) : t ≥ 0}. More rigorous definition relates to the concept of
σ-field we introduce before.
Theorem 5.6 Strong Markov Property
Let N be a birth process and let T be a stopping time for N . Let A be an

event which depends on {N (s) : s > T } and B be an event which depends
on {N (s) : s ≤ T }. Then
P (A|N (T ) = i, B) = P (A|N (T ) = i)
for all i. ♦
Proof: In fact, this is a simple case, as N (T ) is a discrete random variable.
The kind of event B which causes most trouble are those contains all infor-
mation about the history of the process before and including time T . If that
is the case, then the value of T is completely determined by B. Hence, we
may write T = T (B). Hence, it claims that
P (A|N (T ) = i, B) = P (A|N (T ) = i, T = T (B), B).
Among the three pieces of information on the right hand side, T is defined
based on {N (s) : s ≤ T (B)} and is a constant when “N (T ) = i, T = T (B)”.
5.2. CONTINUOUS TIME MARKOV CHAINS 71
Hence the Markov (weak one) property allows us to ignore B itself. At the
same time, the process is time homogeneous, it depends only on the fact that
the chain is in state i now, not when it first reached state i. Hence, the part
of T = T (B) is also not informative. Hence the conclusion.
The measure theory proof is as follows. Let H = σ{N (s) : s ≤ T } which
is the σ-field generated by these random variables. An event B containing
historic information before T is simply an event in this σ-algebra. Recall the
formular E[E(X|Y )] = E(X). Let H play the role of Y , and E(·|N (T ) =
i, B) play the role of expectation, we ahve
P (A|N (T ) = i, B) = E(I(A)|N (T ) = i, B)
= E[E(I(A)|N (T ) = i, B, H)|N (T ) = i, B).
We claim E(I(A)|N (T ) = i, B, H) = E(I(A)|N (T ) = i) since it is H-

measurable, plus T is a constant on B and the weak Markov property. Since
this function is a constant in the eyes of N (T ) = i, B, we have
E{E(I(A)|N (T ) = i)|N (T ) = i, B} = E(I(A)|N (T ) = i).
This result is easy to present for discrete time Markov chain. Hence, if
you cannot understand the above proof, work on the discrete time example
in the assignment will help. ♦
Example 5.1
Consider the birth process with N (0) = I > 0. Define pn (t) = P (N (t) = n).
Set up the forward system and solve it when λn = nλ. ♦
5.2 Continuous time Markov chains

Let X = {X(t) : t ≥ 0} be a family of random variables taking values in
some countable state space S and indexed by the half-line [0, ∞). As before,
we shall assume that S is a subset of non-negative integers.
Definition 5.1
The process X satisfies the Markov property of
P (X(tn ) = j|X(t1 ) = i1 , . . . , X(tn−1 ) = in−1 ) = P (X(tn ) = j|X(tn−1 ) = in−1 )
for all j, i1 , . . . , in−1 ∈ S and any sequence 0 < t1 < t2 < · · · < tn of times.
A continuous time stochastic process X satisfying the Markov property

is called a continuous time Markov chain.
One obvious example of continuous time Markov process is the Poisson
process. The pure birth process is not a Markov process when it is dishonest.
The reason is that to qualify as a Markov chain, we would have required
{N (t) : t ≥ 0} to be a family of random variables. If the population size can
explode to infinity at finite t, then N (t) is not always a random variable for
a given t.
One difficulty of studying the evolution of the continuous time Markov
chain is the lack of one-step transition probability matrix. No matter how
short a period of time is, we can always find a shorter period. This observa-
tion gives rise to the need of infinitesimal generator.
Definition 5.2
The transition probability pij (s, t) is defined to be
pij (s, t) = P (X(t) = j|X(s) = i)
for s < t.
The chain is called time homogeneous if pij (s, t) = pij (0, t − s) for all
i, j, s, t and we write pij (t − s) for pij (s, t). ♦
We will assume the homogeneity for all continuous time Markov chain
to be discussed unless otherwise specified. We also use notation Pt for the
corresponding matrix. It turns out that the collection of all transition prob-
ability matrices form a stochastic semi-group. Forming a group means we
can define an operation on this set such that the operation is closed and
invertable. A semi-group may not have an inverse for each member in the
same set.
Theorem 5.7
The family {Pt : t ≥ 0} is a stochastic semi-group; that is, it satisfies the

following conditions:
(a) P0 = I, the identity matrix;
(b) Pt is stochastic, that is Pt has non-negative entries and row sum 1;
(c) the Chapman-Kolmogorov equations: Ps+t = Ps Pt if s, t, ≥ 0. ♦
Proof: Obvious. ♦
Remark: a quick reference to mathematics definition reveals that only
(c), and sometimes also (a) are properties of a semi-group.
For continuous time Markov chain, there is no t0 such that Pt0 can be
used to compute Pt for all t. This is different from the discrete time Markov
chain. Is it possible to represent all Pt from a single entity?
The answer is positive if the semi-group satisfies some properties.
Definition 5.3
The semigroup {Pt : t ≥ 0} is called standard if Pt → I as t ↓ 0, which is to

say that pii (t) → 1 and pij (t) → 0. ♦
Being standard means that every entry of Pt is a continuous function of t,

not only at t = 0, but also at any t > 0 by Chapman-Kolmogorov equations.
It does not imply, though, that they are differentiable at t = 0.
Suppose that X(t) = i at the the moment t. What might happy in the
next short period (t, t + h)?
May be nothing will happen, the corresponding chance is pii (h) + o(h).
It may switched into state j with corresponding chance pij (h) + o(h).
The chance to have two or more transitions are assumed o(h) here. I take
it as an assumption, which means not all Markov chain has this property.
The book mentioned that it can be proved. I am somewhat skeptical.
It turns out that when the semi-group is standard, the following claim is
true:
pij (h) = gij h + o(h), pii (h) = 1 + gii h + o(h) (5.1)
where gij ’s are a set of constants such that gij > 0 for i 6= j and −∞ ≤ gii ≤ 0
for all i.
Unless additional retrictions are applied, some gii may take value −∞.
In which case I guess that the interpretation is
pii (h) − 1
→ −∞.
h
When all the needed conditions are in place, we set up a matrix G for gij
and call it infinitesimal generator.
Linking this expansion to the transition probability, we must have
X
gij = 0
j
for all i. Due to the fact that some gii can be negative infinity, the above rela-
tionship does not apply to all Markov chains. We will discuss the conditions
under which this result is guaranteed.
Once we admit the validity of (5.1), then we can easily obtain the forward
and backward equations.
Forward equations P0t = Pt G;
Backward equations P0t = GPt ;
Example 5.2
Birth process. Write down the infinitesimal generator here. ♦

The validity of (5.1) does not guarantee the uniqueness of the solution
to these two systems. It is known that the backward equations are satisfied
when the semi-group is standard. The forward equations are proved when
the semi-group is uniform which will be discussed a bit further.
When all entries of G are bounded by a common constant in absolute
value, then two systems share a unique solution for Pt . In fact, the solution
can be written in the form
∞ n
t
Gn .
X
Pt =
n=0 n!
We may also write it as

Pt = exp(tG).
Recall that for some Markov chain with standard semi-group transition
probability matrices, the instantanuous rates gii could be negative infinity.
In this case, state i is instantanuous. That is, the moment of the Markov
chain entering state i is also the moment of leaving the state. Barring such
possibilities by assuming gij have a common upper bound in absolute value,
then waiting times for Markov chain leaving state i will have momoryless
property. Therefore, we have the result as follows.
Theorem 5.8
Under some conditions, the random variable Ui is exponentially distributed

with parameter −gii , where Ui is the waiting time for the Markov chain to
leave state i.
Further, the probability that the chain jumps to state j from state i is
−gij /gii . ♦
The result on the exponential distribution is the product of the Markov
property. Since the Markov property implies the memoriless property, and
the later implies the exponential distribution. The only hassle is the param-
eter −gii . As long as it is finite for all i, the proof is good enough.
The proof for the second part is less rigorous. Taking a small positive
value h, we consider the probability under the assumption of x < U < x +
h. The conditional probability converges to −gij /gii when the boundedness
conditions on gii are satisfied.
Example 5.3
A continuous Markov chain with finite state space is always standard and
uniform. Thus, all the conclusions are valid. ♦
Example 5.4
A continuous Markov chain with only two possible states can have its forward
and backward equations solved easily. ♦
5.3 Limiting probabilities

Our next issue is about limiting probabilities. Similar to discrete time Markov
chain, we need to examine the issue of irreducibility. It turns out that for
any state i and j, if
pij (t) > 0
for some t > 0, then pij (t) > 0 for all t ≥ 0.
We say that a continuous time Markov chain is irreducible when pij (t) > 0
for all i, j and t > 0.
Definition 5.1
The vector π is a stationary distribution of the continuous Markov chain if
πj ≥ 0, j πj = 1, and π = πPi for all t ≥ 0. ♦
P
By Chapman-Kolmogorov equations, if X(t) has distribution π, then

X( t + s) also has distribution π for all s ≥ 0.
If π is such a distribution, it must be unique and non-negative when the
chain is irreducible. This is due to the fact that Pt is a transition probability
matrix of some discrete times Markov chain which is irreducible. Hence, its
stationary distribution is non-negative and unique.
Solving equations π = πPt may not be convenient as Pt ’s are hard to
specify. We hence seek help from the following theorem.
Theorem 5.9
If the Markov chain has uniform semi-group transition probability matrices,
then π = πPt for all t if and only if πG = 0. ♦
The proof is simple. When the Markov chain has uniform semi-group,
then we have Pt = exp(tG). Using the expansion for the exponential function
results in the conclusion. Note that when πG = 0, it is easy to show that
every term in the power expansion is zero. If πPt = 0 for all t, it implies the
function is a zero function. An analytical function is a zero function if and
only if all coefficients are zero. This implies that πG = 0.
The textbook does not specify the uniformity condition. The result is
true for more general Markov chains such as the birth and death process.
Finally, we state the limiting theorem.
5.4. BIRTH-DEATH PROCESSES AND IMBEDDING 77
Theorem 5.10
Let X be irreducible with a standard semigroup {Pt } of transition proba-

biblities.
(a) If there exists a stationary distribution π, then it is unique and
pij (t) → πj
as t → ∞ for all i and j.
(b) If there is no stationary distribution then pij (t) → 0 as t → 0 as t → ∞

for all i and j.
♦
Proof: For any h > 0, we may define Yn = X(nh). Then we have a
discrete time Markov chain which is irreducible and ergodic. If it is non-null
recurrent, then it has a unique stationary distribution π (h) . In addition,
(h)
pij (nh) = P (Yn = j|Y0 = i) → πj .
If the chain is null recurrent or transient, we would have
pij (nh) = P (Yn = j|Y0 = i) → 0
for all i, j.
For any two rational numbers h1 and h2 , applying the above result implies
(h ) (h )
that πj 1 = πj 2 .
Next, using continuity of pij (t) to fill up the gap for all real numbers.
Remark: Unlike other result, we are told that the conclusion is true when
the class of the transition matrix is a standard semi-group.
5.4 Birth-death processes and imbedding

Birth and death process is a more realistic model for population evolution.
Suppose the random variable X(t) represents the population size at time t.
(a) Its state space is S = {0, 1, . . .}.
(b) Its infinitesimal transition probbilities are given by


 λn h + o(h)
if m = 1,


P (X(t + h) = m + n|X(t) = n) = µn h + o(h) if m = −1,

 o(h) if |m| > 1,

(c) the ‘birthe rates’ λ0 , . . . , and ‘death rates’ µ0 = 0, µ1 , µ2 , . . . satisfy

λn ≥ 0 and µn ≥ 0.
Due to memoryless property, the waiting time for the next transition has
exponential distribution with rate λn + µn when X(t) = n. The probability
that the next transition is a birth is given by λn /(λn + µn ).
The infinitesimal generator G ha a nice form
 
−λ0 λ0 0 0 0 ···
µ1 −(λ1 + µ1 ) ···
 

 λ1 0 0 

G= 0 µ2 −(λ2 + µ2 ) λ2 0 ··· .
 
 

 0 0 µ3 −(λ3 + µ3 ) λ3 ··· 

... ... ... ... ...
The chain is uniform if and only if supn {λn + µn } < ∞. When λ0 = 0,

the chain is reducible. State 0 becomes absorbing. If λn = 0 for some n > 0
and X(0) = m < n, then the population size will be bounded by n. It seems
the standard semi-group condition is satisfied as long as all rates are finite.
Yet, we have to assume that the process is honest to maintain that it is a
continuous time Markov chain.
The transition probabilities pij (t) may be computed in principle, but may
not be so useful. It has a stationary distribution under some simple condi-
tions:
λ0 λ1 . . . λn−1
πn = π0
µ1 µ2 · · · µn
with ∞
λ0 λ1 . . . λn−1
π0−1 =
X
.
n=0 µ1 µ2 · · · µn
The limit exists when π0 > 0. Note that in this case, the condition of
standard semi-group is satisfied. Hence, the conclusion for the existence of
the limiting probabilities is solid.
5.4. BIRTH-DEATH PROCESSES AND IMBEDDING 79
Example 5.5 Simple death with immigration.
Consider the situation when the birth rates λn = λ for all n, and µn = nµ
for n = 1, 2, . . .. That is, there is a constant source of immigration, and each
individual in the population has the same rate of die. Using the transition
probability language, the book staes that
pij (h) = P (X(t + h) = j|X(t) = i)

(
P (j − iarrivals, 0 deaths) + o(h) if j ≥ i,
= .
P (i − jdeaths, 0 arrivals) + o(h) if j < i
Since the chance to have two or more transitions in a short period of length
h is o(h), it reduces to
pi,i+1 (h) = λh + o(h),
pi,i−1 (h) = (iµ)h + o(h)
and pij (h) = o(h) when |i − j| ≥ 2.
It is very simple to work out its limiting probabilities.
Theorem 5.11
In the limit as t → ∞, X(t) is asymptotically Poisson distributed with

parameter ρ = µλ . That is,
ρn
P (X(t) = n) → exp(−ρ), n = 0, 1, 2, . . . .
n!
The proof is very simple. Let us work out the distribution of X(t) directly.
Assume X(0) = I.
Let pj (t) = P (X(t) = j|X(0) = I). By the foward equations, we have
p0j (t) = λpj−1 (t) − (λ + jµ)pj (t) + µ(j + 1)pj+1 (t)
for j ≥ 1 and p00 (t)a = µp1 (t).

Note that the probability generating function of X(t) is given by G(s, t) =
E[sX(t) ], we have
∞
∂G X
= jsj−1 pj (t)
∂s j=0
and ∞
∂G X
= sj p0j (t).
∂t j=0
Hence, by multiplying sj on both sides of the forward system and sum

up, we have
∂G ∂G
(s, t) = λ(s − 1)G(s, t) − (µs − µ) (s, t)
∂t ∂s
∂G
= (s − 1)[λG(s, t) − µ (s, t)].
∂t
It is easy to verify that
G(s, t) = {1 + (s − 1) exp(−µt)}I exp{ρ(s − 1)(1 − e−µt )}.
It is interesting to see that if I = 0, then the distribution of the X(t) is

Poisson. Otherwise, the distribution is a convolution of a binomial distribu-
tion and a Poisson distribution. Due to the uniqueness of the solution to the
forward system, we do not have to look for other solutions anymore.
Example 5.6 Simple birth-death
When the birth and death rates are given by λn = nλ and µn = nµ, we can
work out the problem in exactly the same way.
It turns out that given X(0) = I, the generating function of X(t) is given
by
" #I
µ(1 − s) − (µ − λs) exp{−t(λ − µ)}
G(s, t) = .
λ(1 − s) − (µ − λs) exp{−t(λ − µ)
The corresponding differential equation is given by
∂G ∂G
(s, t) = (λs − µ)(s − 1) (s, t).
∂t ∂s
Does it look like a convolution of one binomial and one negative binomial
distribution?
One can find the mean and variance of X(t) from this expression.
E[X(t)] = I exp{(λ − µ)t}

5.5. EMBEDDING 81
λ+µ
V ar(X(t)) = exp{(λ − µ)t}[exp{(λ − µ)t} − 1]I.
λ−µ
The limit for the expectation is either 0 or infinity depending on whether the
ratio of λ/µ is larger than or smaller than 1.
The probability of extinction is given by the limit of η(t) = P (X(t) = 0)
which is min(ρ−1 , 1). ♦
5.5 Embedding
If we ignore the length of the time between two consecutive transitions in a
continuous time Markov chain, we obtain a discrete time Markov chain. In
the example of linear birth and death, the waiting time for the next birth or
death has exponential distribution with rate n(λ + µ) given X(t) = n. The
probability that the transition is a birth is (nλ)/[n(λ + µ)] = λ/(λ + µ).
Think of the transition as the movement of a particle from the integer
n to the new integer n + 1 or n − 1. The particle is performing a simple
random walk with probability p = λ/(λ + µ). The state 0 is an absorbing
state. The probability that the random walk will be absorbed to state 0 is
given by min(1, q/p).
5.6 Markov chain Monte Carlo

In statistical applications, we often need to compute the joint commulative
distribution functions of several random variables. For example, we might be
interested to know the distribution of s2n = (n − 1)−1 ni=1 (Xi − X̄n ) where
P
X1 , . . . , Xn are independent and identically distributed random variables,

X̄n = n−1 i Xi .
P
When X1 has normal distribution, the distribution of s2n is known to be

chisquare with n − 1 degrees of freedom. The density function is well known,
and the value of P (s2n ≤ x) can be easily obtained via numerical integration.
If X1 has exponential distribution, the density function of s2n is no longer
available in a simple analytical form. Straightforward numerical integration
becomes very difficult. If one can generate 10,000 sets of independent random
variables Y1 , Y2 , . . . , Yn such that they all have exponential distribution, then
we may compute 10,000 many s2n values. If 12% of them are smaller than
x = 5, then it can be shown that P (s2n ≤ 5) can be approximated by 12%.
The precision improves when we generate more and more sets of such random
variables.
A more general problem to compute
Z X
g(θ)π(θ)dθ, or g(θ)π(θ)
θ
over θ ∈ Θ, where π(θ) is a density function or probability mass function

on Θ. If we can generate random variables X1 , X2 , . . . such that they all
take values in Θ and have π(θ) as their density function or probability mass
function, then the law of large number in the probability theory implies
n−1
X
g(Xi ) → Eπ {g(X1 )}.
i
Hence, the value of n−1 i g(Xi ) provides an approximation to g(θ)π(θ)dθ,

P R
P
or θ g(θ)π(θ).
Before this idea becomes applicable, we need to overcome a technical
difficulty. How do we generate random variables with a specified distribution
very quickly? It is found that some seemly simple functions can produce very
unpredictable outcomes. Thus, when this operation is applied repeatedly,
the outcomes appear to be completely random. A so called pseudo random
number generator for uniform distribution is hence constructed. Starting
from this point, one can then generate random numbers from most commonly
used distributions, whether they are discreate or continuous.
In some applications, especially for Baysian analysis, it is often necessary
to generate random numbers from a not well defined density function. For
example, we may only know that the corresponding density function π(θ) is
proportional to some known function. In theory, one needs only rescale the
function so that the total probability becomes 1. In reality, to find the scale
itself is numerically impractical.
The beauty of the Markov Chain Monte Carlo (MCMC) is to construct
a discrete time Markov chain so that its stationary distribution is the same
as the given density function π(θ), without the need of completely specifying
π(θ). How this is achieved is the topic of this section.
5.6. MARKOV CHAIN MONTE CARLO 83
In the eyes of a computer, nothing is continuous. Hence we deal with

discrete π(θ). Assume the probability function π = (πi : i ∈ Θ) is our target.
We construct a discrete time Markov chain with state space given by Θ, and
its stationary distribution is given by π.
For a given π, we have many choices of such Markov chains. The one
which is reversible at equilibrium may have some advantage. Thus, we try
to find a Markov chain with its transition probabilities satisfying
πk pkj = πj pjk
for all k, j ∈ Θ. The following steps will create such a discrete time Markov
chain.
Assume we have Xn = i already. We need to generate Xn+1 according to
some transition probability.
(1) First, we pick an arbitrary stochastic matrix H = (hij : i, j ∈ Θ).
This matrix is called the ‘proposal matrix’. We generate a random number Y
according to the distribution given by (hik : k ∈ Θ). That is, P (Y = k) = hik .
Since hik are well defined, we consider it feasible.
(2) Select a matrix A = (aij : i, j ∈ Θ) be a matrix with entries between
0 and 1. The aij are called ‘acceptance probabilities’. We first generate a
uniform [0, 1] random variable Z and define
Xn+1 = Xn I(Z > aij ) + Y I(Z < aij ).
Repeat this two steps, we have created a sequence of random numbers

X0 , X1 , . . . with an arbitrary starting value X0 . Our next task is to choose
H and {A} such that the stationary distribution is what we are look for: π.
With the given A, the transition probabilities are given by
(
hij aij if i 6= j
pij =
1 − k:k6=i hik aik if i = j
P
This is because that the transition to j occurs only if both Y = j and Z < aij
when Xn 6= j.
It turns out that the balance equation is satisfied when we choose
aij = min{1, (πj hji )/(πi hij )}.

This choice results in the algorithm called the Hastings algorithm. Note also
that in this algorithm, we do not need the knowledge of πi , but πi /πj for all
i and j.
Let us try to verify this result. Consider two states i 6= j. We have πi pij =
πi hij aij = min{πi hij , πj hji }. Similarly, πj pji = πj hji aji = min{πi hij , πj hji }.
Hence, πi pij = πj pji .
When two states are the same, the balance equation is obviously satisfied.
When Θ is multi-dimensional, there is some advantage of trying to gener-
ate the random vectors one component a time. That is, if Xn is the current
random vector, we make Xn+1 equal Xn except of one of its component. The
ultimate goal can be achieved by randomly select a component or by rotating
the component.
Gibbs sampler, or heat bath algorithm Assume Θ = S V . That is,
its dimension is V . The state space S is finite and so is the dimension V .
Each state in Θ can be written as i = (iw : w = 1, 2, . . . , V ). Let
Θi,v = {j ∈ Θ : jw = iw for w 6= v}.
That is, it is the set of all states which have the same components as state i
other than the vth component. Assume Xn = i, we now choose a state from
Θi,v for Xn+1 . For this purpose, we select
pij
hij = P , j ∈ Θi,v
k∈Θi,v πk
in the first step of the Hastings algorithm. That is, the choice in Θi,v is based
on the conditional distribution given the vth component.
The next step of the Hastings algorithm is the same as before. It turns
out that aij = 1 for all j ∈ Θi,v . That is, there is no need of the second step.
We need to determine the choice of v. We may rotate the components or
randomly pick a component each time we generate the next random vector.
Metropolis algorithm If the matrix H is symmetric, then aij = min{1, πj /πi }
and hence pij = hij min{1, πj /πi }.
We find such a matrix by placing a uniform distribution on Θi,v as defined
in the last Gibbs samples algorithm. That is
hij = [|Θi,v | − 1]−1

5.6. MARKOV CHAIN MONTE CARLO 85
for i 6= j.
Finally, even the limiting distribution of Xn is π. The distribution of Xn
may be far from π until n is hugh. It is also very hard to tell how large
n has to be before the distribution of Xn well approximates the stationary
distribution.
In applications, researchers often propose to make use of a burn out period
M . That is, we throw away X1 , X2 , . . . , XM for a large M and start using
Y1 = XM +1 , Y2 = XM +2 and so on to compute various characteristics of
π. For example, we estimate Eπ g(θ) by n−1 ni=1 g(XM +i ). It is hence very
P
important to have some idea on how large this M must be before the random
vectors (numbers) can be used.
Let P be a transition probability matrix of a finite irreducible Markov
chain with period d. Then,
(a) λ1 = 1 is an eigenvalue of P ,
(b) the d complex roots of unity,
λ1 = ω 0 , λ2 = ω 1 , . . . , λd = ω d−1 ,
are eigenvalues of P .
(c) the remaining eigenvalues λd+1 , . . . , λN satisfy |λj | < 1.
When all eigenvalues are distinct, then P = B −1 ΛB for some matrix
B and diagonal matrix with entries λ1 , . . . , λN . Thus, this decomposition
allows us to compute the n step transition probability matrix easily. That
is, P n = B −1 Λn P .
When not all eigenvalues are distinct, then we can still decompose P as
−1
B M B. However, M cannot be made diagonal any more. The best we can
is block diagnonal diag(J1 , J2 , . . .) such that
 
λi 1 0 0 ···
···
 

 0 λi 1 0 

Ji =  0 0 λi 1 ··· .


 

 0 0 0 λi ··· 

... ... ... ...
Fortunately, M n still have a very simple form. Hence, once such a decompo-
sition is found, the n-step transition probability matrix is easily obtained.
Remember that the limiting probability is closed linked to the n-step

transition probability matrix. Suppose that the Markov chain in MCMC
algorithm is aperiodic. Then, it can be shown that
(n)
pij = πj + O(nm−1 |λ2 |n )
where λ2 is the eigenvalue with the second largest modulus and m is the
multiplicity of this eigenvalue. Note that the transition probability converges
to πj at exponential rate when m = 1.
Theorem 5.12
Let X be an aperiod irreducible reversible Markov chain on the finite state

space Θ, with transition probability matrix P and stationary distribution π.
Then
|pik − πk | ≤ |Θ| · |λ2 |n sup{|νr (i))| : r ∈ Θ}, i ∈ Θ, n ≥ 1,

X
where νr (i) is the ith term of the rth right-eigenvector Vr of P . ♦

Chapter 6
General Stochastic Process in

Continuous Time
The term stochastic process is generally referred to a collection of random

variables denoted as {X(t), t ∈ T }. Usually, these random variables are
linked by some index generally referred to as time. If so, T is a set of
time points. When T contains integer values only, we get a discrete time
stochastic process; while when T is an interval of the real numbers, we have
a continuous time stochastic process. Examples other than T being time
includes the situation when T represents geological location.
6.1 Finite Dimensional Distribution

Let us make a simplifying assumption that T = [0, T ] for some positive
number T . Let us recall that a random variable is a measurable function on
a probability space (Ω, F, P ). That is, for each t ∈ T , S(t) is a map in the
form
S(t) : Ω → R
where R is the set of all real numbers. That is, at any sample point ω ∈ Ω,
S(t) takes a real value: S(ω, t).
It becomes apparent that S(ω, t) is a bivariate function. One of its vari-
ables is time t, and the other variable is the sample point ω. Given t, we
87
88CHAPTER 6. GENERAL STOCHASTIC PROCESS IN CONTINUOUS TIME
get a random variable. At the same time, once we fix ω = ω0 , S(ω0 , t) is a

function of t. This function is called a sample path of S(t). We denote s(t)
as realized sample path of S(t). This is analog to use x for a realized value
of a random variable X.
We may also put it as follows: a random variable is a map from the sample
space to the space of real numbers; a stochastic process is a map from the
sample space to the space of real valued functions. Recall that a random
variable is required to be a measurable map. Thus, there must be some kind
of measurability requirement on stochastic process.
Compared to the space of real numbers, the space of real functions on
[0, T ] is much more complex. The problem of setting up a suitable σ-algebra
is beyond this course. The space of all continuous functions, denoted as
C[0, T ], has a simpler structure, and a proper σ-field is available. In appli-
cations, however, stochastic processes with non-continuous sample paths are
useful. The well known Poisson process, for example, does not have contin-
uous sample path. A slight extension of C[0, T ] is to add functions which
are right-continuous with left limits. We denote this space as D[0, T ]. These
functions are referred as regular right-continuous function or càdàg
functions. It turns out that D[0, T ] is large enough for usual applications,
yet is still simple enough to allow a useful σ-field.
To properly investigate the properties of a random variable X, we want
to be able to compute P (X ≤ x) for any real number x. We refer to
F (x) = P (X ≤ x) as the cumulative distribution function of X, or sim-
ply the distribution of X. If S(t) is a stochastic process on D[0, T ], then for
any t1 , t2 ∈ [0, T ], we must have
P (S(t1 ) ≤ s1 , S(t2 ) ≤ s2 )
well defined. That means, the set S(t1 ) ≤ s1 , S(t2 ) ≤ s2 must be a member
of F. It is easy to see that this requirement goes from two time points to any
finite number of time points, and to any countable number of time points. It
can be shown that if we can give a proper probability to all such sets (called
cylinder sets), then the probability measure can be extended uniquely to the
smallest σ-algebra that contains all cylinder sets.
As consequence of this discussion is the following theorem.
6.2. SAMPLE PATH 89
Theorem 6.1
A stochastic process taking values in D[0, T ] is uniquely determined by its

finite dimensional distributions. ♦
6.2 Sample Path

Let us have a look of the following example.
Example 6.1
Let X(t) = 0 for all t, 0 ≤ t ≤ 1, and τ is a uniformly distributed random

variable on [0, 1]. Let Y (t) = 0 for t 6= τ and Y (t) = 1 if t = τ .
Now we have two stochastic processes: {X(t) : t ∈ [0, 1]} and {Y (t) :
t ∈ [0, 1]}. It appears that they are very different. All sample paths of X
is identically 0 function, and every sample path of Y (t) has a jump point at
t = τ.
Yet we notice that for any fixed t,
P (Y (t) 6= 0) = P (τ = t) = 0.
That is, P (Y (t) = 0) = 1. Hence, for any t ∈ [0, 1], X(t), Y (t) have the
same distribution. Further, for any set of 0 ≤ t1 < t2 < · · · < tn ≤ 1,
the joint distribution of {X(t1 ), X(t2 ), . . . , X(tn )} is identically to that of
{Y (t1 ), Y (t2 ), . . . , Y (tn )}. That is, X(t) and Y (t) have the same finite di-
mensional distributions. ♦
The moral of this example is: even if two stochastic processes have the
same distribution, they can still have very different sample paths. This ob-
servation prompts the following definition.
Definition 6.1
Two stochastic processes are called a version (modification) of one another

if
P {X(t) = Y (t)} = 1 for all t, 0 < t < T.
♦
In the case when a number of versions exist for a given distribution, a

stochastic process with the smoothest sample path will be the version for
investigation.
Having P {X(t) = Y (t)} = 1 does not mean the event X(t) 6= Y (t) is an
empty set. Let this set be called Nt . While the probability of each is zero,
the probability of its union over t ∈ [0, T ] can be non-zero. In the above
example, it equals 1. When P (∪0≤t≤T Nt ) = 0, the two stochastic processes
are practically the same. They are indistinguishable.
Given a distribution of a stochastic process, can we always find a version
of it so that its sample paths are continuous or regular (with only jump
discontinuities)?
Theorem 6.2
Let S(t), 0 ≤ t ≤ T be a real valued stochastic process.

(a) A continuous version of S(t) exists if we can find positive constants
α, and C such that
E|S(t) − S(u)|α ≤ C|t − u|1+
for all 0 ≤ t, u ≤ T .
(b) A regular version of S(t) exists if we can find positive constants
α1 , α2 , and C such that
E{|S(u) − S(v)|α1 |S(t) − S(v)|α2 } ≤ C|t − u|1+
for all 0 ≤ u ≤ v ≤ t ≤ T . ♦
There is no need to memorize this theorem. What we should make out
of it? Under some continuity condition on expectations, a regular enough
version of a stochastic process exists. We hence have legitimate base to
investigate stochastic processes with this property.
Finally, let us summarize the results we discussed in this section. The
distribution of a stochastic process is determined by its finite dimensional
distributions, whatever it means. Even if two stochastic processes have the
same distribution, their sample paths may have very different properties.
However, for each given stochastic process, there often exists a continuous
6.3. GAUSSIAN PROCESS 91
version, or regular version, whether the given one itself has continuous sample
paths or not.
Ultimately, we focus on the most convenient version of the given stochastic
process in most applications.
6.3 Gaussian Process

Let us make it super simple.
Definition 6.1
Let {X(t) : 0 ≤ t ≤ T } be a stochastic process. X(t) is a Gaussian process

if all its finite dimensional distributions are multivariate normal. ♦
We are at ease when working with normal, multivariate normal random
variables. Thus, we also feel more comfortable with Gaussian processes when
we have to deal with continuous time stochastic process taking arbitrary real
values. We will be obsessed with Gaussian processes.
Example 6.2 A simple Gaussian Process
Let X and Y be two independent standard normal random variables. Let
S(t) = sin(t)X + cos(t)Y.
It is seen that S(t) is a Gaussian process with certain covariance structures.

♦
6.4 Stationary Processes

Since the random behavior of a general stochastic process is very complex, we
often first settle down with very simple ones and then timidly move toward
more complex ones. The simple random work is the first example of stochastic
process. We then introduce discrete time Markov chain and so on. We still
hope to stay in the familiar territory by moving forward a little: introducing
the concept of stationary processes here.
Definition 6.1 Strong Stationarity.
Let {X(t) : 0 ≤ t ≤ T = ∞} be a stochastic process. If for any choices of

0 ≤ t1 < t2 . . . ≤ tn , with any n and positive constant s, the joint distribution
of
X(t1 + s), X(t2 + s), . . . , X(tn + s)
does not depend on s, then we say that {X(t) : 0 ≤ t ≤ T = ∞} is strongly
stationary. ♦
Definition 6.2 Weak Stationarity.
Let {X(t) : 0 ≤ t ≤ T = ∞} be a stochastic process. If for any choices

of 0 ≤ t1 < t2 . . . ≤ tn , with any n and positive constant s, the mean and
covariance matrix of
X(t1 + s), X(t2 + s), . . . , X(tn + s)
do not depend on s, then we say that {X(t) : 0 ≤ t ≤ T = ∞} is weakly

stationary, or simply stationary. ♦
The above definition can be easily applied to the case when T < ∞, or
to discrete time stochastic processes. The obstacles are merely notational.
Being strongly stationary does not necessary imply the weak stationary,
because of the moment requirements of the latter.
Most results for stationary processes are rather involved. We only men-
tion two theorems.
Theorem 6.3 Strong and Weak Laws of Large Numbers (Ergodic Theo-
rems).
(a) Let X = {Xn : n = 1, 2, . . .} be a strongly stationary process such that

E|X1 | < ∞. There exists a variable Y with the same mean as the Xn such
that
n
−1
X
n Xi → Y
i=1
almost surely, and in mean.

6.5. STOPPING TIMES AND MARTINGALES 93
(b) Let X = {Xn : n = 1, 2, . . .} be a weakly stationary process. There

exists a variable Y with the same mean as the Xn such that
n
n−1
X
Xi → Y
i=1
in the second moment. ♦

If a continuous time stationary process can be discretised, the above two
results can also be very useful.
6.5 Stopping Times and Martingales

We are often interested in knowing how the property of a stochastic process
following the moment when some event has just occurred. This is a setting
point of a renewal process. These moments are random variables with special
properties.
Consider the example of simple random walk {Xn : n = 0, 1, . . .}, the
time of returning to zero is a renewal event. At any moment, we examine
the value of Xn to determine whether the process has renewal itself at n with
respect to this particular renewal event.
In general, we need to examine the values of {Xi : i = 0, 1, . . . , n} to
determine the occurrence of the renewal or other events. The time for the
occurrence of such events is therefore a function of {Xi : i = 0, 1, . . . , n}. In
measure theory language, if τ represents such moments, then the event
τ ≤n
is Fn = σ(X0 , . . . , Xn } measurable.
Definition 6.1 Stopping Time
Let {X(t) : 0 ≤ t ≤ T = ∞} be a stochastic process. A nonnegative random
variable τ , which is allowed to take value ∞, is called a stopping time if for
each t, the event
{τ ≤ t} ∈ σ(Xs , 0 ≤ s ≤ t}.
♦
Example 6.3 To be invented.
Chapter 7
Brownian Motion or Wiener

Process
It was botanist R. Brown who first described the irregular and random mo-
tion of a pollen particle suspected in fluid (1828). Hence, such motions are
called Brownian motion. Interestingly, the particle theory in physics was not
widely accepted until the Einstein used Brown motion to argue that such
motions are caused by bombardments of molecules in 1905. The theory is
far from complete without a mathematical model developed later by Wiener
(1931). This stochastic model used for Brown motion is hence also called
Wiener process. We will see that the mathematical model has many desir-
able properties, and a few that do not fit well with physical laws for Brownian
motion. As early as 1931, Brown motion has been used in mathematical the-
ory for stock prices. It is nowadays a fashion to use stochastic processes to
study financial market.
Definition 7.2 Defining properties of Brownian motion.
Brownian motion {B(t)} is a stochastic process with the following properties:
1. (Independent increment) The random variable B(t) − B(s) is indepen-

dent of {B(u) : 0 ≤ u ≤ s} for any s < t.
2. (Stationary Gaussian increment) The distribution of B(t) − B(s) is

normal with mean 0 and variance t − s.
95
96 CHAPTER 7. BROWNIAN MOTION OR WIENER PROCESS
3. (Continuity) The sample paths of B(t) are continuous.
♦
We should compare this definition with that of Poisson process: the in-
crement distribution is changed from Poisson to normal. The sample path
requirement is needed so that we will work on a definitive version of the
process.
Once the distribution of B(0) is given, then all the finite dimensional
distributions of B(t) are completely determined, which then determines the
distribution of B(t) itself. This notion is again analog to that of a counting
process with independent, stationary and Poisson increments.
In some cases, B(0) = 0 is part of the definition. We will more or less
adopt this convention, but will continue to spell it out.
Example 7.1
Assume that B(t) is a Brownian motion and B(0) = 0.
1. Find the marginal distributions of B(1), B(2) and B(2) − B(1).
2. Find the joint distribution of B(1) and B(2), and compute P (B(1) >
1, B(2) < 0).
♦
Recall the multivariate normal distribution is completely determined by
its mean vector and covariance matrix. Thus, the finite dimensional distribu-
tions of a Gaussian process is completely determined by the mean function
and the correlation function. As a special Gaussian process, it is useful to
calculate the mean and correlation function of the Brownian motion.
Example 7.2 Compute the correlation function of the Brownian motion.
♦
Based this calculation, it is clear that the above Brownian motion is not
stationary.
R1
Example 7.3 Distribution of 0 B(t)dt given B(t) = 0.
97
Example 7.4
Let ξ0 , ξ1 , . . . be a sequence of independent standard normally distributed

random variables. Define
∞
t 2 X sin(it)
S(t) = √ ξ0 + √ ξi .
π π i=1 i
Assume that we can work with infinite summation as if nothing is unusual.

Then we can verify the covariance structure of S(t) resembles that of the
Brownian motion with B(0) = 0 over [0, π]. Thus, S(t) is a Brownian motion
and it provides a very concrete way of represent the Brownian motion. It can
be seen that all sample paths of S(t) are continuous, yet they are nowhere
differentiable. ♦
Example 7.5 Sample paths of the Brownian motion.
Brownian motions have the following well known but very peculiar properties.
1. Almost all sample paths are continuous;
2. Almost every sample path is not monotone in any interval, no matter

how small the interval is;
3. Almost every sample path is not differentiable at any point;
4. Almost every sample path has infinite variation on any interval, not
matter how short it is;
5. Almost all sample paths have quadratic variation on [0, t] equal to t,

for any t.
Definition 7.3
Let s(x) be a function of real variable.

(a) The (total) variation of s(x) over [0, t] is
n
X
lim |s(xi ) − s(xi−1 )|
i=1
where the limit is taken over any sequence of sets of (x1 , . . . , xn ) such that
0 = x0 < x 1 < x 2 < · · · < x n = t
and max{xi − xi−1 : i = 1, . . . , n} → 0.

(b) The quadratic variation over the interval [0, t] is
n
{s(xi ) − s(xi−1 )}2
X
lim
i=1
where the limit is taken over any sequence of sets of (x1 , . . . , xn ) such that
0 = x0 < x 1 < x 2 < · · · < x n = t
and max{xi − xi−1 : i = 1, . . . , n} → 0. ♦

If s(x) is monotone over [0, t], then the total variation is simply |s(t) −
s(0)|. Having a non-zero quadratic variation implies that the function fluc-
tuates up and down very very often. It is hard to think of a function with
√
such property. One example is s(x) = x sin(1/x) over [0, 1].
Theorem 7.1
If s(x) is continuous and of finite total variation, then its quadratic variation
is zero. ♦
Property 1 is true by definition. Property 2 can be derived from Property
4. If a sample path is monotone over an interval, then its variation is finite,
which contradicts Property 4. Property 3 fits well with Property 2; if a
sample path is nowhere monotone, it is very odd to be differentiable. This
point can be made rigorous.
Finally, it is seen that Property 4 follows from Property 5 because of these
arguments. It is hence vital to show that Brownian motion has Property 5.
99
Let 0 = t0 < t1 < t2 < · · · < tn = t be a partition of the interval [0, 1],
where each of ti depends on n though not spelled out. Let
n
{B(ti−1 ) − B(ti )}2
X
Tn =
i=1
be the variation based on this partition. We find
n
E{B(ti−1 ) − B(ti )}2
X
E{Tn } =
i=1
n
X
= V ar{B(ti−1 ) − B(ti )}
i=1
Xn
= (ti − ti−1 ) = t.
i=1
That is, the expected sample path quadratic variation is t, regardless of how
the interval is partitioned.
Property 5, however, mean that lim Tn = t almost surely along any se-
quence of parititions such that δn = max |ti − ti−1 | → 0. We give a partial
proof when dn < ∞. In this case, we have
P
n
V ar{B(ti−1 ) − B(ti )}2
X
V ar{Tn } =
i=1
n
3(ti − ti−1 )2
X
=
i=1
≤ 3t max |ti − ti−1 | = 3δn t.
Therefore, V ar{Tn } = E{Tn − E(Tn )}2 < ∞. Using monotone conver-

P P
gence theorem in measure theory, it implies E {Tn − E(Tn )}2 < ∞. (Note
P
the change of order between summation and expectation). The latter implies
Tn − E(Tn ) → 0 almost surely, which is the same as Tn → t almost surely.
Despite the beauty of this result, this property also shows that the math-
ematical Brownian motion does not model the physical Brownian motion in
microscope level. If Newton’s laws hold, the acceleration cannot be instan-
taneous, and the sample paths of a diffusion particle should be smooth.
7.1 Existence of Brownian Motion

One technical consideration is whether there exist stochastic processes with
the properties prescribed by the definition of Brownian motion. This question
is usually answered by defining a sequence of stochastic processes such that
its limit exists, and having the defining properties of Brownian motion. One
rigorous proof of the existence is scratched as follows.
Let ξi , i = 1, 2, . . . be independent and normally distributed with mean
0 and variance 1. The existence of such sequence and the corresponding
probability space is well discussed in probability theory and is assumed here.
Let n
X
Yn = ξi
i=1
and define
1
Bn (t) = √ {Y[nt] + (nt − [nt])ξ[nt]+1 }
n
which is a smoothed partial sum of ξ’s. It can be verified that the covari-
ance structure of Bn (t) approaches what is assumed for Brownian motion.
All finite dimensional distributions of Bn (t) are normal. With a dose of
asymptotic theory which requires the verification of tightness, it is seen that
Bn (t) has a limit, and the limiting process has all the defining properties of a
Brownian motion. Thus, the existence is verified. See Page 62 of Billingsley
(1968).
One may replace ξi , i = 1, 2, . . . by independent and identically distributed
random variables taking ±an values. We may define the partial sum Yn in the
same way, and define a similar stochastic process. By properly scaling down
the time, and letting an → 0, the limiting process will also have the defining
properties of a Brownian motion. Thus, Brownian motion is regarded as a
limit of simple random walk. Many of other properties of Brownian motion
to be discussed have their simple random walk versions.
7.2 Martingales of Brownian Motion

The reason for Brownian motion being the centre of most textbooks and
courses in stochastic processes can be attributed to the fact that it is a good
7.2. MARTINGALES OF BROWNIAN MOTION 101
example of any important concepts in continuous time stochastic process.

This section introduces its martingale aspects.
Recall that each random variable X generates a σ-field: σ(X). This
extends to a set of random variables. If X(t) is a stochastic process on [0, T ],
then {X(u) : 0 ≤ u ≤ t} is a set of random variables and it generates a
σ-field which is usually denoted as Ft . As sets of sets, these σ-fields satisfy
Fs ⊂ Ft
for all 0 ≤ s < t ≤ T .

We may put forward a sequence of σ-field Ft not associated with any
stochastic processes. As long as {Ft , t ≥ 0} has the increasing property, it is
called a filtration.
Definition 7.1 Martingale
A stochastic process {X(t), t ≥ 0 is a martingale with respect to Ft if for

any t it is integrable, E|X(t)| ≤ ∞, and for any s > 0
E{X(t + s)|Ft } = X(t).
♦
With this definition, we have implied that X(t) is Ft measurable. In
general, Ft represents information about X(t) available to an observer up
to time t. If any decision is to be made at time t by this observer, the
decision is a Ft measurable function. Unless otherwise specified, we take
Ft = σ{X(s) : s ≤ t} when a stochastic process X(t) and a filtration Ft are
subjects of some discussion.
One of the defining properties of Brownian motion can be reworded as
the Brownian motion is a martingale.
We would like to mention two other martingales.
Theorem 7.2
Let B(t) be a Brownian motion.

(a). B 2 (t) − t is a martingale;
(b). For any θ, exp{θB(t) − 12 θ2 t} is a martingale. ♦
Proofs will be provided in class.
7.3 Markov Property of Brownian Motion

Let {X(t), t ≥ 0} be a stochastic process and Ft is the corresponding filtra-
tion. The Markov property is naturally extended as follows.
Definition 7.1 Markov Process.
If for any s, t > 0,
P {X(t + s) ≤ y|Ft } = P {X(t + s) ≤ y|Xt }
almost surely for all y, then {X(t), t ≥ 0} is a Markov process. ♦

Compared to the definition of Markov chain, this definition removes the
requirement that the state space of X(t) is countable.
Theorem 7.3 A Brownian motion is a Markov Process.
Proof: Working out the conditional moment generating function will be

sufficient. ♦
In applications, we often want to know the property following the occur-
rence of some event. A typical example is the behavior of a simple random
walk after its return to 0. Since the returning time itself is random, we must
be careful when claim that the simple random walk following that moment
will have the same property as a simple random walk starting from 0.
If we recall, this claim turns out to be true. Regardless, we must be
prepared to prove that this is indeed the case. As we mentioned before,
Brownian motion in many respects generalizes the simple random walk. They
share many similar properties. One of them is the strong Markov property.
As a preparation, we re-introduce the concept of stopping times.
Definition 7.2
Let {X(t), t ≥ 0} be a stochastic process, and Ft = σ{X(s) : s ≤ t}. A

random time T is called a stopping time for {X(t), t ≥ 0} if {T ≤ t} ∈ Ft
for all t > 0. ♦
7.4. EXIT TIMES AND HITTING TIMES 103
If we stripe the jargon of σ-field, a stopping time T is a random quantity

such that after we observed the stochastic process up to and include time t,
we can tell whether T ≤ t or not.
Let T to be the time when a Brownian motion first exceeds value 1. Then
T is a stopping time. By examining the values of X(s) for all s smaller or
equal t, the truthfulness of T ≤ t is a simple matter.
Let T be the time when the value of stock price will drop by 50% in the
next unit of time. If T were a stopping time, we would make a lot of money.
In this case, you really have to be able to see into the future. We do not
need strong mathematics to find that T is not a stopping time.
Another preparation is: for each non-random time t, Ft is the σ-field
generated by random variables X(s) including all s ≤ t. If T is a stopping
time, we hope to be able to use it as if it is a non-random time. We define
FT = {A : A ∈ F, A ∩ {T ≤ t} ∈ Ft for any t}.
Finally, let us return to the issue of strong Markov property.
Theorem 7.4 Strong Markov Property.
Let T be a stopping time associated with Brownian motion B(t). Then for
any t > 0,
P {B(T + t) ≤ y|FT } = P {B(T + t) ≤ y|B(T )}
for all y. ♦
We cannot afford to spend more time at proving this result. Let us
tentatively accept it as a fact and see what will be the consequences.
Let us define B̂(t) = B(T + t) − B(T ). When T is a constant, it can be
easily verified that B̂(t) is also a Brownian motion. When T is a stopping
time, it remains to be true due to the strong Markov property.
7.4 Exit Times and Hitting Times

Let T (x) = inf{t > 0 : B(t) = x} be a stopping time for Brownian motion
B(t). It is seen that T (x) is the first moment when the Brownian motion
hits target x.
Assume B(0) = x and a < x < b. Define T = min{T (a), T (b)}. Hence
T is the time when the Brownian motion escapes the area formed by two
horizontal lines. The question is: will it occur in finite time? If so, what is
the average time it takes?
The first question is answered by computing P (T < ∞), and the second
question is question is answered by computing E{T } under the assumption
that the first answer is affirmative. One may link this problem with simple
random walk and predict the outcome. The probability for the first passage
of “1” for a symmetric simple random walk is 1. Since the normal distribution
is symmetric, the simple random walk result seems to suggest that P (T <
∞) = 1. We do not have similar results for E{T }, but for E{T (a)} or
E{T (b)} for simple random walk, and the later are likely finite from the
same consideration. The real answers are given as follows.
Theorem 7.5
The hitting time T defined above have the properties:
(a) P (T < ∞|B(0) = x) = 1;
(b) E{T |B(0) = x} < ∞, under the assumption that a < x < b.
♦
Proof: It is seen that the event {T > 1} implies a < B(t) < b for all
t ∈ [0, 1]. Hence,
P (T > 1|B(0) = x) ≤ P {a < B(1) < b|B(0) = x}

= Φ(b − x) − Φ(a − x).
As long as a, b are finite,
α = sup{Φ(b − x) − Φ(a − x)} < 1.

x
Next, for any positive integer n,
pn = P {T > n|B(0) = x}
= P {T > n|T > n − 1, B(0) = x)pn−1
7.5. MAXIMUM AND MINIMUM OF BROWNIAN MOTION 105
= P {a < B(s) < b, for n − 1 ≤ s ≤ n|a < B(s) < b,

for 0 ≤ s ≤ n − 1, B(0) = x}pn−1
= P {a < B(s) < b, for n − 1 ≤ s ≤ n|a < B(n − 1) < b}pn−1
= P {a < B(s) < b, for 1 ≤ s ≤ 1|a < B(0) < b}pn−1
≤ αpn−1 .
Applying this relationship repeatedly, we get pn ≤ αn . Thus, we have
P (T < ∞|B(0) = x) = 1 − n→∞

lim pn = 1.
Further,
∞
X
E(T |B(0) = x) ≤ P {T > n|B(0) = x} < ∞.
n=0
♦
I worry slightly that I did not really make use of strong Markov property,
but the Markov property itself.
The following results are obvious:
P {T (b) < ∞|B(0) = a} = 1; P {T (a) < ∞|B(0) = a}.
7.5 Maximum and Minimum of Brownian Mo-

tion
Define, for a given Brownian motion,
M (t) = max{B(s) : 0 ≤ s ≤ t} and m(t) = min{B(s) : 0 ≤ s ≤ t}.
The significance of these two quantities are obvious.

Again, the following result resembles a result of simple random walk, and
I believe that a proof using some limiting approach is possible.
Theorem 7.6
For any x > 0,
P {M (t) > m|B(0) = 0} = 2P {B(t) ≤ m|B(0) = 0}.

Proof: We will omit the condition that B(0) = 0. Also, we use T (a) as the
stopping time when B(t) first hits a.
First, for any m, we have
P {M (t) ≥ m} = P {M (t) ≥ m, B(t) ≥ m}

+P {M (t) ≥ m, B(t) < m}.
Note that B(T (m)) = m as the sample paths are continuous almost surely,
plus {M (t) > m} is equivalent to T (m) ≤ t.
P {M (t) ≥ m, B(t) < m}

= P {T (m) < t, B(t) < B(T (m))}
= P {B(t) < B(T (m))|T (m) < t}P {T (m) < t}
= P {B(t) > B(T (m))|T (m) < t}P {T (m) < t}
Strong Markov Property
Reflection Principle
= P {M (t) ≥ m, B(t) ≥ m}.
Applying this back, we get
P {M (t) ≥ m} = 2P {M (t) ≥ m, B(t) ≥ m}.
Since B(t) ≥ m implies M (t) ≥ m, the above result becomes
P {M (t) ≥ m} = 2P {B(t) ≥ m}.
This completes the proof. ♦

This result can be used to find the distribution function of T (x) under
the assumption that B(0) = 0.
Theorem 7.7
The density function of T (x) is given by
|x| x2
f (t) = √ exp{− } t ≥ 0.
2πt3 2t
7.6. ZEROS OF BROWNIAN MOTION AND ARCSINE LAW 107
Proof: It is done by using the equivalence between T (x) ≤ t and M (t) ≥ x.

♦
We call T (x) the first passage time.
The reflection idea is formally stated as the next theorem.
Theorem 7.8
Let T be a stopping time. Define B̂(t) = B(t) for t ≤ T and B̂(t) =

2B(t) − B(t) for t > T . Then B̂(t) is also Brownian motion. ♦
The proof is not hard, and we should draw a picture to understand its
meaning.
7.6 Zeros of Brownian Motion and Arcsine

Law
Assume B(0) = 0. How long does it take for the motion to come back to 0?
This time, the conclusion is very different from that of simple random walk.
Theorem 7.9
The probability that B(t) has a least one zero in the time interval (a, b) is
given by
2 q
arc cos a/b.
π
Proof: Let E(a, b) be the event that B(t) = 0 for at lease some t ∈ (a, b).
P {E(a, b)} = E[P {E(a, b)|B(a)}]
For any x > 0,
P {E(a, b)|B(a) = −x} = P {T (x) < b − a|B(0) = 0} = P {T (x) < b − a}
by Markov property and so on.

Hence,
Z ∞
E[P {E(a, b)|B(a)}] = P {E(a, b)|B(a) = x}fa (x)dx
−∞
Z ∞
= 2 P {E(a, b)|B(a) = −x}fa (x)dx
Z0∞
= 2 P {T (x) < b − a}fa (x)dx
0
where fa (x) is the density function of B(a) and known to be normal with
mean 0 and variance a. The density function of T (x) is given earlier. Sub-
stituting two expressions in and finishing the job of integration, we get the
result. ♦
When a → 0, so that a/b → 0, this probability approaches 1. Hence, the
probability that B(t) = 0 in any small neighborhood of t = 0 (not including
0) is 1 given B(0) = 0. This conclusion can be further strengthened. There
are infinite number of zeros in any neighborhood of t = 0, no matter how
small this neighborhood is. This result further illustrates that the sample
paths of Brownian motion, though continuous, is nowhere smooth.
The famous Arcsine law now follows.
Theorem 7.10
The probability that Brownian

q motion has no zeros in the time interval (a, b)
is given by (2/π)arc sin (a/b). ♦
There is a similar result for simple random walk.
7.7 Diffusion Processes

To be continued.

Stochastic Processes: Stat433/833 Lecture Notes

Uploaded by

Copyright:

Available Formats

Stochastic Processes: Stat433/833 Lecture Notes

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Stochastic Processes: Stat433/833 Lecture Notes

Uploaded by

Copyright:

Available Formats

Stat433/833 Lecture Notes

Key Words: σ-field, Brownian motion, diffusion process, ergordic, finite

Stat 433/833 Course Outline

• We use the formula for final grade:

25%assignment + 25%midterm + 50%final.

• There will be 4 assignments.

• Office hours: to be announced.

• Office: MC6007 B, Ex 35506, jhchen@uwaterloo.ca

Textbook and References

1. Introduction to probability models. (QA273 .R84) Sheldon M. Ross.

2. A Second course in stochastic processes. S. Karlin and H. M. Taylor.

3. An introduction to probability theory and its applications. Volumes I

4. Convergence of probability measures. Billingsley, P. Wiley.

5. Gaussian random processes. (QA274.4.I2613) I.A. Ibragimov, and Rozanov.

6. Introduction to stochastic calculus with applications. Fima C. Klebaner.

7. Diffusions, Markov processes and martingales.(QA274.7.W54) L. C. G.

8. Probability Theory, independence, interchangeability, martingale. Y. S.

1 Review and More 1

2 Simple Random Walk 13

4 Discrete Time Markov Chain 43

4.2.1 Properties of Markov Chain with Finite Number of States 49

5 Continuous Time Markov Chain 63

6 General Stochastic Process in Continuous Time 87

7 Brownian Motion or Wiener Process 95

Review and More

1.1 Probability Space

A collection of sets F is a σ-field (algebra) if it satisfies

3. If A ∈ F, then its complement Ac ∈ F.

1. The simplest σ-field is {φ, Ω}.

2. A simplest non-trivial σ-field is {φ, A, Ac , Ω}.

3. A most exhaustive σ-field is F consists of all subsets of Ω.

Axioms of Probability Measure

1. 0 ≤ P (A) ≤ 1 for all A ∈ F;

Mathematically, the above definition does not rely on any hypothetical

Axioms 2 and 3 imply that

1 = P (Ω) = P (A ∪ Ac ) = P (A) + P (Ac ).

Hence, P (Ac ) = 1 − P (A).

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

P (∪ni=1 Ai ) = P (Ai1 Ai2 ) + · · · , +(−1)n+1 P (∩ni=1 Ai ).

Lemma 1.1 Continuity of the probability measure.

Let (Ω, F, P ) be a probability space. If A1 ⊂ A2 ⊂ A3 ⊂ · · · is an increasing

A = lim ∪ni=1 Ai = lim An ,

Proof: Let Bn = An − An−1 , n = 1, 2, 3, . . . with A0 = φ. Then An =

This completes the proof. ♦

1.2 Random Variable

A random variable is a map X : Ω → R such that {X ≤ x} = {ω ∈ Ω :

The random behaviour of X is largely determined by its distribution

1. If X is absolutely continuous, then the probability density function

For almost all x, f (x) = dF (x)/dx.

2. If X is discrete, taking values on {x1 , x2 , . . .}, the probability mass

f (xi ) = P (X = xi ) = F (xj ) − F (xj −)

for i = 1, 2, . . ., where F (a−) = limx→a,x<a F (x).

3. The moment generating function is defined as

4. The characteristic function of X is defined as

where i is such that i2 = −1. The advantage of the characteristic

5. The probability generating function is useful when the random variable

6. In survival analysis or reliability, we often use survival function

S(x) = P (X > x) = 1 − F (x)

7. The hazard function of X is