Financial Engineering & Risk Management

Introduction to Brownian Motion

M. Haugh G. Iyengar
Department of Industrial Engineering and Operations Research
Columbia University
 Brownian Motion

Definition. We say that a random process, {Xt : t Ø 0}, is a Brownian motion

with parameters (µ, ‡) if
1. For 0 < t1 < t2 < . . . < tn≠1 < tn

(Xt2 ≠ Xt1 ), (Xt3 ≠ Xt2 ), . . . , (Xtn ≠ Xtn≠1 )

are mutually independent.

2. For s > 0, Xt+s ≠ Xt ≥ N(µs, ‡ 2 s) and
3. Xt is a continuous function of t.

We say that Xt is a B(µ, ‡) Brownian motion with drift µ and volatility ‡

Property #1 is often called the independent increments property.

Remark. Bachelier (1900) and Einstein (1905) were the first to explore Brownian
motion from a mathematical viewpoint whereas Wiener (1920’s) was the first to
show that it actually exists as a well-defined mathematical entity.
2
Standard Brownian Motion

When µ = 0 and ‡ = 1 we have a standard Brownian motion (SBM).

We will use Wt to denote a SBM and we always assume that W0 = 0.
Note that if Xt ≥ B(µ, ‡) and X0 = x then we can write

Xt = x + µt + ‡Wt (8)

where Wt is an SBM. Therefore see that Xt ≥ N (x + µt, ‡ 2 t).

3
Sample Paths of Brownian Motion

4
 Information Filtrations

For any random process we will use Ft to denote the information available
at time t
- the set {Ft }tØ0 is then the information filtration

- so E[· | Ft ] denotes an expectation conditional on time t information available.

Will usually write E[· | Ft ] as Et [ · ].

Important Fact: The independent increments property of Brownian motion

implies that any function of Wt+s ≠ Wt is independent of Ft and that

(Wt+s ≠ Wt ) ≥ N(0, s).

5
 A Brownian Motion Calculation

Question: What is E0 [Wt+s Ws ]?

Answer: We can use a version of the conditional expectation identity to obtain

E0 [Wt+s Ws ] = E0 [(Wt+s ≠ Ws + Ws ) Ws ]
# $
= E0 [(Wt+s ≠ Ws ) Ws ] + E0 Ws2 . (9)

Now we know (why?) E0 [Ws2 ] = s.

To calculate first term on r.h.s. of (9) a version of the conditional expectation
identity implies

E0 [(Wt+s ≠ Ws ) Ws ] = E0 [Es [(Wt+s ≠ Ws ) Ws ]]

= E0 [Ws Es [(Wt+s ≠ Ws )]]
= E0 [Ws 0]
= 0.

Therefore obtain E0 [Wt+s Ws ] = s.

6
Financial Engineering & Risk Management
Geometric Brownian Motion

M. Haugh G. Iyengar
Department of Industrial Engineering and Operations Research
Columbia University
 Geometric Brownian Motion

Definition. We say that a random process, Xt , is a geometric Brownian motion

(GBM) if for all t Ø 0 ! ‡2 "
µ≠ 2 t + ‡Wt
Xt = e
where Wt is a standard Brownian motion.

We call µ the drift, ‡ the volatility and write Xt ≥ GBM(µ, ‡).

Note that
! 2
"
µ≠ ‡2 (t+s) + ‡Wt+s
Xt+s = X0 e
! 2
" ! 2
"
µ≠ ‡2 t + ‡Wt + µ≠ ‡2 s + ‡(Wt+s ≠Wt )
= X0 e
! ‡2 "
µ≠ 2 s + ‡(Wt+s ≠Wt )
= Xt e (10)

– a representation that is very useful for simulating security prices.

2
 Geometric Brownian Motion

Question: Suppose Xt ≥ GBM(µ, ‡). What is Et [Xt+s ]?

Answer: From (10) we have

5 ! ‡2 " 6
µ≠ 2 s + ‡(Wt+s ≠Wt )
Et [Xt+s ] = Et Xt e
! ‡2 " Ë È
µ≠ 2 s
= Xt e Et e ‡(Wt+s ≠Wt )
! ‡2 " 2
µ≠ 2 s ‡ s
= Xt e e2
= e µs Xt

– so the expected growth rate of Xt is µ.

3
Sample Paths of Geometric Brownian Motion

4
 Geometric Brownian Motion

The following properties of GBM follow immediately from the definition of BM:

Xt2 Xt3 X
1. Fix t1 , t2 , . . . , tn . Then Xt1 , Xt2 , . . . , Xt tn are mutually independent.
n≠1
(For a period of time t, consider 0<t1 <t2 <t3 <t4 …..tn < t)

2. Paths of Xt are continuous as a function of t, i.e., they do not jump.

1 2 1 2
Xt+s ‡2
3. For s > 0, log Xt ≥ N (µ ≠ 2 )s, ‡2 s .

5
 Modeling Stock Prices as GBM

Suppose Xt ≥ GBM(µ, ‡). Then clear that:

1. If Xt > 0, then Xt+s is always positive for any s > 0.
- so limited liability of stock price is not violated.

2. The distribution of Xt+s /Xt only depends on s and not on Xt

These properties suggest that GBM might be a reasonable model for stock prices.

Indeed it is the underlying model for the famous Black-Scholes option formula.

6
Financial Engineering and Risk Management
Review of vectors

Martin Haugh Garud Iyengar

Columbia University
Industrial Engineering and Operations Research
Reals numbers and vectors
We will denote the set of real numbers by R
Vectors are finite collections of real numbers
Vectors come in two varieties
# $
Row vectors: v = v1 v2 . . . vn
S T
w1
W w2 X
Column vectors w = W X
U ... V
wn
By default, vectors are column vectors

The set of all vectors with n components is denoted by Rn

2
Linear independence
A vector w is linearly dependent on v1 , v2 if

w = –1 v1 + –2 v2 for some –1 , –2 œ R
Example: S T S T S T
2 1 0
U6V = 2 U1V + 4 U1V
4 0 1
Other names: linear combination, linear span

A set V = {v1 , . . . , vm } are linearly independent if no vi is linearly

dependent on the others, {vj : j ”= i}

3
Basis
Every w œ Rn is a linear combination of the linearly independent set
YS T S T S TZ S T S T S T
_
_ 1 0 0 _ _ 1 0 0
_
]W0X W1X W0X^ _ W0X W1X W0X
W X W X W X W X W X W X
B = W.X , W.X , . . . W.X w = w1 W . X +w2 W . X + . . . + wn W . X
_
_U .. V U .. V U .. V_
_ U .. V U .. V U .. V
_
[ _
\
0 0 1 0 0 1
¸˚˙˝ ¸˚˙˝ ¸˚˙˝
e1 e2 en

Basis © any linearly independent set that spans the entire space
Any basis for Rn has exactly n elements

4
Norms
A function fl(v) of a vector v is called a norm if
fl(v) Ø 0 and fl(v) = 0 implies v = 0
fl(–v) = |–| fl(v) for all – œ R
fl(v1 + v2 ) Æ fl(v1 ) + fl(v2 ) (triangle inequality)
fl generalizes the notion of “length”

Examples:
Òq
n
¸2 norm: ÎxÎ2 = i=1
|xi |2 ... usual length
qn
¸1 norm: ÎxÎ1 = i=1
|xi |
¸Œ norm: ÎxÎŒ = max1ÆiÆn |x|i
1q 2 1p
n
¸p norm, 1 Æ p < Œ: ÎxÎp = i=1
|x|pi

5
Inner product
The inner-product or dot-product of two vector v, w œ Rn is defined as
n
ÿ
v·w= vi w i
i=1
Ô
The ¸2 norm ÎvÎ2 = v·v
The angle ◊ between two vectors v and w is given by
v·w
cos(◊) =
ÎvÎ2 ÎwÎ2

Will show later: v · w = v€ w = product of v transpose and w