Stochastic Calculus for Finance II

some Solutions to Chapter IV

Matthias Thul
Last Update: August 1, 2011

Exercise 4.1
This proof is fully analogous to the one of Theorem 4.2.1. We want to show that for
0stT

E[I(t)|F(s)] = I(s).
Assume again, that the s [tl , tl+1 ) and t [tk , tk+1 ) for l k. We start by splitting
up the sum into an F(s) measurable part and a part independent of F(s) and obtain

I(t) =

l1
X

(tj ) [M (tj+1 ) M (tj )] + (tl ) [M (tl+1 ) M (tl )]

j=0

k1
X

(tj ) [M (tj+1 ) M (tj )] + (tk ) [M (t) M (tk )]

j=l+1

The first sum is F(s) measurable. The conditional expectation of the second term is

E [ (tl ) [M (tl+1 ) M (tl )]| F(s)] = (tl ) [E [M (tl+1 )| F(s)] M (tl )]

= (tl ) [M (s) M (tl )] .
Here we have used that M (tl ) is F(s) measurable. Furthermore, the martingale
property of M (t) implied that E [M (tl+1 )| F(s)] = M (s). The conditional expectation of
each element of the second sum is

The author can be contacted

http://www.matthiasthul.com.

via

<<firstname>>.<<lastname>>@googlemail.com

and

E [ (tj ) [M (tj+1 ) M (tj )]| F(s)] = E [E [ (tj ) [M (tj+1 ) M (tj )]| F (tj )]| F(s)]
= E [ (tj ) [E [M (tj+1 )| F (tj )] M (tj )]| F(s)]
= E [ (tj ) [M (tj ) M (tj )]| F(s)]
= 0.
Here, we first used the tower law of iterated conditioning, then took out all F (tj )
measurable terms from the inner conditional expectation and finally used the martingale
property of M (t) again. The same steps can be used to show that

E [ (tk ) [M (t) M (tk )]| F(s)] = 0.

It follows that

E [I(t)| F(s)] =

l1
X

(tj ) [M (tj+1 ) M (tj )] + (tl ) [M (s) M (tl )]

j=0

= I(s)

(q.e.d.)

Exercise 4.4 (Stratonovich integral)

(i) Since the elements in the sum are independent, we start by computing the mean



2
and the variance of W tj W (tj ) . This is completely analogous to the proof
of Theorem 3.4.3.

2 i




E W
W (tj )
= Var W tj W (tj ) = tj tj
h
h



4 i



2 i2

2

E W tj W (tj )
= 3E W tj W (tj )
= 3 tj tj
h



2 i

2

2

2

Var W tj W (tj )
= 3 tj tj tj tj = 2 tj tj
tj

Here, we have used that the forth moment of a normal random variable is three
times its variance squared. The mean and variance of Q/2 are then given by
n1
h
X



2 i


E Q/2 =
E W tj W (tj )
j=0

n1
X

tj

tj =

n1
X
tj+1 tj
j=0

j=0

1
= T
2

n1
h
X





2 i

Var Q/2 =
Var W tj W (tj )
j=0

n1
X

tj

tj

2

j=0

||||
2

n1
X
1
j=0

n1
X
j=0

(tj+1 tj )2

1
(tj+1 tj )
2



1
lim Var Q/2 = 0 T = 0
||||0
2

(q.e.d.)

Since the expected value of Q/2 is the same for all partitions, but the variance as
limit zero when the maximum partition size gets smaller, we can conclude that Q/2
has limit 12 T .
(ii) We expand the sum in the definition of the Stratonovich integral and add the terms
necessary to obtain an Ito integral over the partition Q/2 .

W (t) dW (t)
0

lim

||||0

lim

||||0

lim

||||0

n1
X
j+0

n1 n
X

o





2


W tj W (tj+1 ) W tj W (tj ) W tj 2W (tj ) W tj W (tj )2

j+0
n1 n
n1
X
X



2 o




W tj W (tj+1 ) W tj
+ lim
W (tj ) W tj W (tj )2

||||0

||||0

j+0

+ lim
Z

W tj (W (tj+1 ) W (tj ))

j+0

n1 n
X

2 o
W (tj )2 2W (tj ) W tj + W tj

j+0

W (t)dW (t) + lim

0

1 2
W (T )
=
2

||||0

n1 
X

W (tj )2 W tj

2 2

j+0

(q.e.d.)

In the last step, we have used the result that

RT
0

W (t)dW (t) = 21 W 2 (T ) 21 T , as

e.g. shown in Example 4.3.2 and the quadratic variation of the Stratonovich integral
that we calculated in part (i).

Exercise 4.5 (Solving the generalized geometric Brownian motion

equation)
Let f (t, x) = ln x. We have
f
= 0,
t

f
1
= ,
x
x

2f
1
= 2
2
x
x

and
(dS(t))2 = 2 (t)S 2 (t)dt.
Applying Itos lemma, the differential of the log stock price d ln S(t) becomes

d ln S(t) = df (t, S(t))

1
1 1
=
dS(t)
(dS(t))2
2
S(t)
2 S (t)
1
= (t)dt + (t)dW (t) 2 (t)dt
2


1 2
=
(t) (t) dt + (t)dW (t).
2
In integral form, we get

Z t
Z t
1 2
ln S(t) = ln S(0) +
(s)dW (s).
(s) (s) ds +
2
0
0

(1)

Taking the exponential yields

Z t 


Z t
1 2
S(t) = S(0) exp
(s) (s) ds +
(s)dW (s) .
2
0
0
The distribution of S(t) will in general not be log-normal. In order to have lognormality for S(t), we require ln S(t) to be normally distributed. By Theorem 4.4.9, the
Ito integral in Equation (1) is only guaranteed to be normally distributed if (t) and (t)
are deterministic functions of time. In this case
Z t 


Z t
1 2
2
ln S(t) N
(s) (s) ds,
(s)ds .
2
0
0
Note that this includes the less general case of a Geometric Brownian Motion with
(t) =
and (t) =
, i.e the drift and diffusion coefficient being a constant as it is
assumed in the Black-Scholes model.
4

Exercise 4.6
Let f (t, x) = S(0)ex . We have
f
= 0,
t

f
= f (x),
x

2f
= f (x).
x2

Now define

X(t) =


1 2
t + W (t).
2

By the Ito formula, the differential of S(t) is given by

dS(t) = df (t, X(t))

1
= S(t)dX(t) + S(t)dX(t)dX(t)
2
1 2
1
= S(t)dt S(t)dt + S(t)dW (t) + 2 S(t)dt
2
2
= S(t)dt + S(t)dW (t).
Now let f (t, x) = xp such that
f
= 0,
t

f
= pxp1 ,
x

2f
= p(p 1)xp1 .
2
x

Using the Ito formula to compute the differential of d (S p (t)) yields

d (S p (t)) = d (f (t, S(t)))

1
= pS p1 (t)dS(t) + p(p 1)S p2 (t)dS(t)dS(t)
2
1
p
p
= pS (t)dt + pS (t)dW (t) + 2 p(p 1)S p (t)dt
2


1
=
+ (p 1) pS p (t)dt + pS p (t)dW (t).
2

Exercise 4.7
(i) Let f (t, x) = x4 . We have
f
= 0,
t

f
= 4x3 ,
x

2f
= 12x2 .
x2

Using the Ito formula to compute the differential of d (W 4 (t)) yields

d W 4 (t)

= 4W 3 (t)dW (t) + 6W 2 (t)dt

Z T
Z T
3
4
W 2 (t)dt
W (t)dW (t) + 6
W (t) = 4

Note that there is in general a constant of integration in the second equation. But
since W 4 (0) = 0, we omitted it directly.
(ii) Taking expectations on both sides and using that an Ito integral is a martingale that
starts at zero yields

Z T

Z T

 4 
3
2
E W (t) = 4E
W (t)dW (t) + 6E
W (t)dt
0
0
Z T
Z T
 2 
= 6
E W (t) dt = 6
tdt
0
0
 T
1 
= 6  t2  = 3T 2
(q.e.d.)
2 0
(iii) Let f (t, x) = x6 . We have
f
= 0,
t

f
= 6x5 ,
x

2f
= 30x4 .
2
x

Using the Ito formula to compute the differential of d (W 6 (t)) yields

d W 6 (t) = 6W 5 (t)dW (t) + 15W 4 (t)dt
Z T
Z T
6
5
W (t) = 6
W (t)dW (t) + 15
W 4 (t)dt
0

Taking expectations on both sides and using the result in part (ii) yields

Z T

Z T


5
4
E W (t) = 6E
W (t)dW (t) + 30E
W (t)dt
0
0
Z T
Z T
 4 
= 15
E W (t) dt = 15
3t2 dt
0
0
 3 T
3


= 15 t 0 = 15T


Exercise 4.8 (Solving the Vasicek equation)

(i) Let f (t, x) = et x. We have
f
= f (t, x),
t

2f
= 0.
x2

f
= et ,
x

Using the Ito formula, we can compute the differential of d et R(t) to be

d et R(t)

= df (t, R(t))
= et R(t)dt + et dR(t)
= et R(t)dt + et dt et R(t)dt + et dW (t)
= et dt + et dW (t)

(2)

(ii) Integrating Equation (2) yields

e du +
eu dW (u)
0
0


Z t
 1 u t
ebu dW (u)
= R(0) +  e  +

0
0
Z t


t
e 1 +
ebu dW (u)
= R(0) +

e R(t) = R(0) +

Thus, R(t) is given by

R(t) = e

R(0) +
1 et + et

ebu dW (u)

This is the same expression as given in Example 4.4.10.

Note that the solution strategy employed is very common for stochastic differential
equations with a geometric drift term. Let X( t) = (t)X(t)dt + . . .. Then by
computing the differential of e(t)t X( t), we can remove the geometric drift.

Exercise 4.11
When dynamically replicating an mispriced option in the Black-Scholes model, we have to
decide upon which volatility to use to compute the hedge ratio. In this case, the market
7

uses a volatility of 1 to price the option at all times t [0, T ]. Knowing that all model
assumption hold but with the actual volatility of the underlying diffusion being 2 > 1 ,
we will enter a long position in the underpriced european call option and dynamically
delta hedge it by shorting (t) =

c
S

shares of the underlying. But since the delta itself

depends upon the volatility, there are two possibilities to do this: Either we use the
market implied volatility (1 ) or the actual volatility (2 ) to compute the hedge ratio.
These two approaches will in general lead to different returns. As implicitly suggested by
the exercise, we will use the delta calculated based upon the implied volatility to hedge
the long position in the european call option.
Let f (t, x) = ert x. We have
f
f
2f
= rf (t, x),
= ert ,
= 0.
t
x
x2
Apply Itos lemma to the differential of the discounted portfolio value d (ert X(t))

d ert X(t) = df (t, X(t))
= rert X(t)dt + ert dX(t)
= ert (rX(t)dt + dX(t)) .

(3)

We furthermore note that the differential of the call option price dc(t, S(t)) evolves
according to
c
1 2c
c
dt +
dS(t) +
(dS(t))2
2
t
S
2
S


c 1 2 2 2 c
c
=
+ 2 S (t) 2 dt +
dS(t).
t 2
S
S

dc(t, S(t)) =

(4)

Plugging the given differential of the portfolio value and the differential of the call
option price in Equation (4) back into Equation (3) yields

rt

d e

X(t)

c 1 2 2 2 c
+ S (t) 2
t 2 2
S

c
dS(t)
S
c
c
dS(t) + rX(t)dt rc(t, S(t))dt + rS(t) dt
S
S

2


c
1
22 12 S 2 (t) 2 dt
2
S


1 2 2 2c
c
rt c
= e
+ S (t) 2 + rS(t)
rc(t, S(t)) dt.
t 2 1
S
S
rt

= e

rX(t)dt +

dt +

We note that the call price c(t, S(t)) is a specific solution to the Black-Scholes PDE
where the volatility is equal to 1
c
2c
c
1
+ rS(t)
+ 12 S 2 (t) 2 = rc(t, S(t))
t
S 2
S
for the payoff-specific terminal condition c(T, S(T )) = max(S(T )K, 0), S(T ) 0 and



the boundary conditions c(t, 0) = 0, 0 t T and limS(t) c(t, S(t)) S(t) Ker(T t) =
0, 0 t T . Using this result, it directly follows that d (ert X(t)) = 0.

Exercise 4.13 (Decomposition of correlated Brownian motions

into independent Brownian motions)
The differential form of B1 (t) and B2 (t) are given by

dB1 (t) = dW1 (t)

dB2 (t) = (t)dW1 (t) +

p
1 2 (t)dW2 (t).

This allows us to express the differentials of W1 (t) and W2 (t) in terms of the changes
in the correlated Brownian motions B1 (t) and B2 (t)

dW1 (t) = dB1 (t)

1
(t)
dB1 (t) + p
dB2 (t).
dW2 (t) = p
2
1 (t)
1 2 (t)

(5)

We have to show that these two processes are independent Brownian motions. While it
is obvious that W1 (t) is a Brownian motion since it is just equal to B1 (t), we have to check
if this also holds for W2 (t). According to Levys theorem, any continuous martingale M (t)
with M (0) = 0 and quadratic variation equal to [M, M ](t) = t (i.e. dM (t)dM (t) = dt) is a
Brownian motion. Furthermore, according to the extended Levy theorem, two Brownian
motions, with zero cross variation (i.e. dW1 (t)dW2 (t) = 0) are independent.
(i) Continuity:

The continuity of W2 (t) follows from the continuity of the two Brownian motions

B1 (t) and B2 (t). To see this explicitly, we write down the integral form of Equation
(5) to get
Z

W2 (t) = W2 (0)
0

(s)
p

2 (s)

Z
dB1 (s) +
0

1
p

1 2 (s)

dB2 (s).

(6)

Since an Ito integral is a continuous function of its upper limit of integration, W2 (t)
is continuous as well.
(ii) Martingale property:

The martingale property of W2 (t) directly follows from the absence of a drift term
in Equation (5) or its representation as an Ito integral in Equation (6).
(iii) Starting at Zero:

Given the information in the exercise, we cant show that W2 (0) = 0. When integrating Equation (5), we have to add a constant of integration that might take
any value in R. Similarly, we could argue that in the equation for dB2 (t), only the
differential dW2 (t) and thus its initial value does not matter.
(iv) Unit quadratic variation:

The quadratic variation of W2 (t) can be computed as

2 (t)
1
2 (t)
dt
+
dt

2
dt
1 2 (t)
1 2 (t)
1 2 (t)
1 2 (t)
=
dt = dt (q.e.d.).
1 2 (t)

dW2 (t)dW2 (t) =

and it follows that [W2 , W2 ](t) = t.

(v) Zero cross variation:

Using dB1 (t)dB1 (t) = dt and dB1 (t)dB2 (t) = (t)dt, the instantaneous cross varia-

10

tion between W1 (t) and W2 (t) can be computed as

dW1 (t)dW2 (t) = p

(t)

(t)
dt + p
dt = 0.
1 2 (t)
1 2 (t)

This establishes that W1 (t) and W2 (t) have zero cross-covariance. An alternative
proof of this result directly computes dB1 (t)dB2 (t)

dB1 (t)dB2 (t) = (t)dt +

p
1 2 (t)dW1 (t)dW2 (t).

This equation is only equal to (t)dt as given in the exercise, when

p
1 2 (t)dW1 (t)dW2 (t) = 0.
Since (1, 1), the square root term can never be equal to zero and we conclude
that dW1 (t)dW2 (t) = 0.

Exercise 4.15 (Creating correlated Brownian motions from independent ones)

(i) Similar to Exercise 4.13, we use Levys theorem to show that Bi (t) is a Brownian
motion for each i = 1, . . . , m. Since Wj (0) = 0 for each j = 1, . . . , d, it directly
follows from the definition of Bi (t) that Bi (0) = 0. Furthermore, due to its representation as a sum of Ito integrals, we obtain both the continuity and the martingale
property. It remains to show that Bi (t) has quadratic variation [Bi , Bi ](t) = t or
equivalently that dBi (t)dBi (t) = dt. In differential form, Bi (t) is
d

dBi (t) =

1 X
ij (t)dWj (t).
i (t) j=1

Using the independence of the d-dimensional vector of Brownian motions, we get

Pd
d
2
1 X 2
j=1 ij (t)
dBi (t)dBi (t) = 2
(t)dt = Pd
dt = dt (q.e.d.).
2
i (t) j=1 ij
j=1 ij (t)

11

(ii) Again using independence we get

d

X
1
dBi (t)dBk (t) =
ij kj (t)dt = ik (t)dt.
i (t)k (t) j=1

Exercise 4.16 (Creating independent Brownian motions to represent correlated ones)

In differential form and using matrix notation, we have dB(t) = A(t)dW (t) or dW (t) =
A1 (t)dB(t). We use the multidimensional Levy theorem (a straight forward extension
of Theorem 4.6.5) to show that W (t) is a vector of independent Brownian motions. We
need to check the following conditions:
Martingale property:
Since we can represent W (t) in terms of an integral constant plus and Ito integral

t
1

A (u)dB(u)

W (t) = W (0) +

Wi (t) = Wi (0) +

m Z
X
j=1

a1
ij (u)dBj (u)

it follows that each element Wi (t) of this vector is a martingale (Theorem 4.3.1(iv)).
Continuity:
By Theorem 4.3.1(i), Ito integrals have continuous sample paths.
Starting at zero:
Similar to Exercise 4.13(iii), we cant show that W (0) is a vector of zeros. In the
first point, we could choose the integration constant W (0) arbitrarily as only the
change dW (t) is relevant in the definition of B(t). However, the exercise is only
asking for the existence of a vector of Brownian motions W (t). Thus, by setting
W (0) = 0 we can find such a vector.
Unit quadratic variation and zero cross variation:

We need to show that each Wi (t) has unit quadratic variation but zero cross variation
with any other Wj (t) for j 6= i or equivalently that dW (t)dW T (t) = I where I is
the m m identity matrix. This can be easily done using matrix notation. We get
12

dW (t)dW T (t) = A1 (t)dB(t) A1 (t)dB(t)

T

= A1 (t)dB(t)dB T (t)AT (t)

= A1 (t)C(t)AT (t)
= A1 (t)A(t)AT (t)AT
= I

(q.e.d.).

Exercise 4.17 (Instantaneous correlation)

(i) We first note that

t0 +

B1 (t0 + ) B2 (t0 + ) = B1 (t0 ) B2 (t0 ) +

d (B1 (s)B2 (s)) .

t0

By Corollary 4.6.3, we have for the differential of the product of the two Brownian
motions

d (B1 (t)B2 (t)) = B2 (t)dB1 (t) + B1 (t)dB2 (t) + dB1 (t)dB2 (t)
= B2 (t)dB1 (t) + B1 (t)dB2 (t) + dt.
Thus,

t0 +

t0 +

B2 (s)dB1 (s)+

B1 (t0 + ) B2 (t0 + ) = B1 (t0 ) B2 (t0 )+

t0

B1 (s)dB2 (s)+
t0

and

E [B1 (t0 + ) B2 (t0 + )| F (t0 )] = B1 (t0 ) B2 (t0 ) + 

where we used that the two integrals are independent of the filtration F (t0 ) and
that by Theorem 4.3.1(iv) Ito integrals are martingales. Finally,

13

E [(B1 (t0 + ) B1 (t0 )) (B2 (t0 + ) B2 (t0 ))| F (t0 )]

= E [B1 (t0 + ) B2 (t0 + )| F (t0 )] E [B1 (t0 + ) B2 (t0 )| F (t0 )]
E [ B1 (t0 ) B2 (t0 + )| F (t0 )] + E [B1 (t0 ) B2 (t0 )| F (t0 )]
= B1 (t0 ) B2 (t0 ) +  B2 (t0 ) E [B1 (t0 + )]
B1 (t0 ) E [B2 (t0 + )] + B1 (t0 ) B2 (t0 )
= .

(q.e.d.)

(ii) The means are

Mi () = E [Xi (t0 + ) Xi (t0 )| F (t0 )]

Z t0 +
=
i du
t0

= i 

(q.e.d.).

Where we again used that the Riemann integral is deterministic and that the Ito
integral is a martingale with zero expected value.
The variances are




Vi () = E (Xi (t0 + ) Xi (t0 ))2  F (t0 ) Mi2 ()
"Z
2 #
Z t0 +
t0 +
= E
i du +
i dBi (u)
2i 2
t0

Z
=

t0

 Z
t0 +
i du E

t0

t0

"Z

i dBi (u) + E

t0 +

2 #
i dBi (u)

t0

t0 +

=
=

t0 +

t0
i2 

i2 du
(q.e.d.).

Here, we used that the increment in the Ito processes is independent of the filtration
F (t0 ) in the second step. We can take the deterministic time integral out of the
expectation in the third step and again reply on the martingale property of the Ito
integral in the fourth step to eliminate this term. We further use the Ito isometry
14

(Theorem 4.3.1(v)) to transform the squared Ito integral into a Riemann integral in
the fourth step.
The covariance is

C() = E [(X1 (t0 + ) X1 (t0 )) (X2 (t0 + ) X2 (t0 ))| F (t0 )] M1 ()M2 ()
 Z t0 +

Z t0 +
Z t0 +
Z t0 +
1 du +
1 dB1 (u)
2 du +
2 dB2 (u) 1 2 2
= E
t
t
t
t
Z t0 +0
 Z t00 +
 Z0 t0 +
 0Z t0 +

=
1 du E
2 dB2 (u) +
2 du E
1 dB1 (u)
t0
t0
t0
t0
 Z t0 +

Z t0 +
1 dB1 (u)
2 dB2 (u)
+E
t0
t0
 Z t0 +

Z t0 +
1 dB1 (u)
2 dB2 (u)
= E
t0

t0

In order to solve the remaining expectation, we first define for t t0

Z

Ii (t) =

i dBi (u)

dIi (t) = i dBi (t)

t0

By the Ito product rule, we have

d (I1 (t)I2 (t)) (t) = I2 (t)dI1 (t) + I1 (t)dI2 (t) + dI1 (t)dI2 (t)
or

I1 (t)I2 (t) = I1 (t0 ) I2 (t0 ) +

d (I1 (u)I2 (u))
t0
Z t
Z t
Z t
I1 (u)dI2 (u) +
1 2 du.
=
I2 (u)dI1 (u) +
t0

t0

t0

Thus,

t0 +

C() = E [I1 (t0 + ) I2 (t0 + )] =

1 2 du = 1 2 .

(q.e.d.).

t0

Consequently, we have for the correlation between the increments of the Ito processes
C()

1 2 
= p 2 2 = .
V1 ()V2 ()
1 2 2
15

(iii) We have

Mi () = E [Xi (t0 + ) X (t0 )| F (t0 )]



 Z t0 +
Z t0 +

i (u)dBi (u) F (t0 )
= E
i (u)du +
t0
t0


 Z t0 +


i (u)du F (t0 ) .
= E
t0

We first note that by Definition 1.2.1, the real constant M can be considered a
(trivial) random variable as well. Here {M B} = , i.e. we do not learn anything
about the particular outcome by observing M . We now apply the dominated
convergence theorem (Theorem 1.4.9) to the sequence of random variables
1
I() =


t0 +

i (u)du
t0

that is bounded in absolute value by M . We have

1
lim Mi () = lim E [I()| F (t0 )]
0
0 





= E lim I() F (t0 )
0




1
0
= E lim [I (0) + I (0)  + o ()] F (t0 )
0 
= E [i (t0 )| F (t0 )]
= i (t0 )

(q.e.d.).

Here, we used a Taylor series expansion of the integral around  = 0 where I(0) = 0
and I 0 (0) = i (t0 ).
(iv) Similar to the hint, we define for t > t0
Z

Yi (t) =

i (u)dBi (u)
t0

and get

16

Dij () = E [(Xi (t0 + ) Xi (t0 )) (Xj (t0 + ) Xj (t0 ))| F (t0 )] Mi ()Mj ()




Z t0 +
Z t0 +

i (u)du
Yj (t0 + ) +
j (u)du  F (t0 )
= E
Yi (t0 + ) +
t0

t0

Mi ()Mj ()
Z

t0 +

= E [Yi (t0 + ) Yj (t0 + )| F (t0 )] +

t0
Z t0 +

+
i (u)du E [Yj (t0 + )| F (t0 )]


j (u)du E [Yi (t0 + )| F (t0 )]

t0

= E [Yi (t0 + ) Yj (t0 + )| F (t0 )]

Here, we followed the same main steps as in the derivation in (ii). We now apply
the Ito product rule again to compute the remaining expectation and get

Z

t0 +

Dij () = E

t0 +

t0 +

Yj (u)dYi (u) +
t
 Z 0t0 +

= E
t0

Yi (u)dYj (u) +



ij (u)i (u)j (u)du F (t0 ) .
t0

t0





dYi (u)dYj (u) F (t0 )

We now apply the same limit argument as in (iii) and start by defining the sequence
of random variables
1
Iij () =


t0 +

ij (u)i (u)j (u)du

t0

that is bounded in absolute value by M 3 . We get

1
lim Dij () = lim E [ Iij ()| F (t0 )]
0 
0





= E lim I() F (t0 )
0




1
0

= E lim [I(0) + I (0) + o()] F (t0 )
0 
= E [ij (t0 ) i (t0 ) j (t0 )| F (t0 )]
= ij (t0 ) i (t0 ) j (t0 )

(q.e.d.).

(v) We showed this already implicitly in the derivation of the means and covariances.
We have
17

Dij () = E [Iij ()| F (t0 )]

= E [I(0) + I 0 (0) + o()| F (t0 )]
= E [ij (t0 ) i (t0 ) j (t0 )  + o()| F (t0 )]
= ij (t0 ) i (t0 ) j (t0 )  + o()

Vi () = 2 (t0 )  + o()

for i = j = k
k
=

C() = (t0 ) 1 (t0 ) 2 (t0 )  + o() for i 6= j

(q.e.d.).

(vi) Finally,

C()
(t0 ) 1 (t0 ) 2 (t0 )  + o()
lim p
= lim p 2
0
0
V1 ()V2 ()
(1 (t0 )  + o()) (22 (t0 )  + o())
(t0 ) 1 (t0 ) 2 (t0 )
= lim p 2
+ o()
0
1 (t0 ) 22 (t0 )
= (t0 )
(q.e.d.).

Exercise 4.18



(i) Let f (t, x) = exp x r + 21 2 t . We have


f
1 2
= r + f (t, x),
t
2

f
= f (t, x),
x

2f
= 2 f (t, x).
x2

Itos lemma yields

d(t) = df (t, W (t))



1 2
1
= r + (t)dt (t)dW (t) + 2 (t)dt
2
2
= r(t)dt (t)dW (t) (q.e.d.).
(ii) By Itos product rule (Corollary 4.6.3) we get

18

d ((t)X(t)) = (t)dX(t) + X(t)d(t) + dX(t)d(t)

= r(t)X(t)dt + (t)(t)( r)S(t)dt + (t)(t)S(t)dW (t)
(t)X(t)dW (t) r(t)X(t)dt (t)(t)S(t)dt
= (t) ((t)S(t) X(t)) dW (t).

(7)

In the last step, we have used the fact that = r. Equation (7) shows that
(t)X(t) is a martingale, since all dt-terms vanish in the differential. Equivalently, we
could express (t)X(t) by a single Ito integral. Since all Ito integrals are martingales,
so is (t)X(t)
Z

(s) ((s)S(s) X(s)) dW (s).

(t)X(t) = (0)X(0) +
0

(iii) First, note that

(0) = exp {W (0)} = 1.

Now let (t) be an adapted portfolio process such that, X(T ) = V (T ), i.e. the
final value of the trading strategy is equal to the final value of the F(T ) measurable
random variable V (T ). As has been shown in (ii), (t)X(t) is a martingale, i.e.

E [(t)X(t)|F(s)] = (s)X(s),

s t.

We obtain

X(0) = (0)X(0) = E [(T )X(T )] = E [(T )V (T )] .

Exercise 4.19
(i) By Levys theorem it is sufficient to show that B(t) is a continuous martingale starting at zero with unit quadratic variation in order to establish that it is a Brownian

19

motion. We first note that B(t) is an Ito integral with B(0) = 0. By Theorem 4.3.1,
Ito integrals are continuous martingales. The quadratic variation of B(t) is

(dB(t))2 = (sign (W (t)) dW (t))2 = sign2 (W (t)) (dW (t))2 = dt

and it follows that it is a Brownian motion.
(ii) We have

d (B(t)W (t)) = B(t)dW (t) + W (t)dB(t) + dB(t)dW (t)

= B(t)dW (t) + W (t)sign (W (t)) dW (t) + sign (W (t)) dt
and since B(0)W (0) = 0 we get

Z
B(t)W (t) =

Z
B(s)dW (s) +

Z
W (s)sign (W (s)) dW (s) +

sign (W (s)) ds
0

Since Ito integrals are martingales it follows that

Z

sign (W (s)) ds =

E [B(t)W (t)] = E
0

E [sign (W (s))] ds = 0.
0

In the last step, we have used that with probability one half, sign (W (t)) = 1 for
t > 0.
(iii) Let f (t, x) = x2 . We have
f
= 0,
t

f
= 2x,
x

2f
= 2.
x2

Using the Ito formula to compute the differential of dW 2 (t) yields

dW 2 (t) = 2W (t)dW (t) + dt

20

(q.e.d.)

(iv) We have

d B(t)W 2 (t) = B(t)dW 2 (t) + W 2 (t)dB(t) + dB(t)dW 2 (t)
= 2B(t)W (t)dW (t) + B(t)dt + W 2 (t)sign (W (t)) dW (t)
+2W (t)sign (W (t)) dt.
Integrating both sides yields

B(t)W (t) =

2B(s)W (s)dW (s) +

B(s)ds +
0
0
Z t
+
2W (s)sign (W (s)) ds.

W 2 (s)sign (W (s)) dW (s)

We take expectation on both sides and again use the martingale property of the Ito
integral to get


E B(t)W (t) = E

Z


2W (s)sign (W (s)) ds

B(s)ds +

(8)

In order to evaluate the first Riemann integral, we set f (t, x) = tx such that ft (t, x) =
x and fx (t, x) = t. We apply the Ito formula to obtain the differential of tW (t) as

d (tB(t)) = B(t)dt + tdB(t).

Integrating and rearranging yields
Z

B(t)dt = T B(T )
0

(T t)dB(t).

tdB(t) =
0

The last term is an Ito integral of a deterministic integrand and thus normally
distributed with zero mean and variance
Z
0

T

1
1
3
(T t) dt = (T t)  = T 3 .
3
3
0
2

It follows that the first term in Equation (8) is zero and we get



E B(t)W 2 (t) = E

Z

Z
0

21


2 |W (s)| ds

2W (s)sign (W (s)) ds = E
0

It is obvious, that the expected value is strictly positive for t > 0 since the integrand
is non-negative and almost surely strictly positive. If B(t) and W (t) are independent,
then by Theorem 2.2.5 any two function f (B(t)) and g (W (t)) are independent. Now
let g (B(t)) = B(t) and f (W (t)) = W 2 (t) then by Theorem 2.2.7 independence
implies that E [B(t)W 2 (t)] = E [B(t)] E [W 2 (t)]. But since E [B(t)] = 0 we would
require E [B(t)W 2 (t)] = 0. We have shown above that E [B(t)W 2 (t)] > 0 and thus
B(t) and W (t) are not independent.

Exercise 4.20 (Local time)

(i)

f 0 (x) =
undefined

0
00
f (x) =

undefined

if x > K
if x = K
x<K
if x 6= K
if x = K

(ii)
Z
f (W (T )) = f (W (0)) +

I{W (t)>K} dW (t)

(9)

Using that by Theorem 3.3.2(i), W (T ) N (0, T ), we can compute the expected

value of the right hand side as

22



E [f (W (T ))] = E (W (T ) K)+
Z 
+
T x K N 0 (x)dx
=
Z


=
T x K N 0 (x)dx
K

Z
T
=


K
xN (x)dx KN
K
T

r ZT


2
T
K
x2
xe dx KN
=
2 K
T
T

r


T x2 
K
2
e  KN
=
K
2
T

T
r


K
T K2
e 2T KN
=
2
T
0

Since the Ito integral is a martingale starting at zero, it has zero expected value and
we get for the right hand side

Z
E [f (W (0))] + E




I{W (t)>K} dW (t) = E (W (0) K)+

= (0 K)+
= 0.
This shows that Equation (9) does not hold.
(iii) The answer to this question is pretty much given in the hints already.

0
if x K 2n

1
f 0 (x) =
n(x K) + 12 if K 2n
xK+

1
1
if x K + 2n

0 if x < K 2n

1
1
f 00 (x) =
n if K 2n
< x < K + 2n

0 if x > K + 1
2n
23

1
2n

(iv) Similar to (iii), there is not too much to add to the hints. We first note that the
intervals of the piecewise function definition have limits


1
lim 0, K
= [0, K)
n
2n


1
1
lim K , K +
= K
n
2n
2n


1
= (K, ).
lim K + ,
n
2n
Thus,

lim fn (x) =

if x K

x K

if x K

= (x K)+

0 if x < K

1
lim fn0 (x) =
if x = K .
2
n

1 if x > K

(10)

(v) Let M (T ) = max0tT W (T ) be the maximum of a fixed Brownian motion path

during the time interval [0, T ]. By assumption, M (T ) < K and by definition

Z
LK (T ) =

lim n

I{W (t)(K 1 ,K+ 1 )} dt.

2n
2n

1
Now, from M (T ) < K it follows that there exists an n0 such that K 2n
> M (T ) for

all n n0 . Consequently, for all n n0 , the indicator function inside the integral
evaluates to zero for all t [0, T ] and we get LK (T ) = 0.
(vi) Taking expectation on both sides and using that by Theorem 4.3.1(iv) the Ito integral
is a martingale with zero expected value yields



E [LK (T )] = E (W (T ) K)+ > 0,

24

where the inequality follows from our result in (ii). Although it is not immediately
obvious from the final result, it is enough to note that by the normal distribution of
W (T ) values of W (T ) > K have a strictly positive probability for all finite levels of
K and thus the expectation is strictly positive as well. Now,

P {LK (T ) = 0} = 1

E [LK (T ) = 0]

and just showed that the implication does not hold. Thus, by contradiction we
cannot have that P {LK (T ) = 0} = 1.

Exercise 4.21 (Stop-loss start-gain paradox)

(i) The stop-loss start-gain strategy involves holding a long position in one share whenever the stock price strictly exceeds the strike price and holding no position otherwise. The crucial point is, that buys and sells are not made at the same price and
thus the hedger incurs a cost every time he enters and unwinds the stock position.
To make it clear, assume that we by one stock whenever the price level K +  is
crossed from below (and we had no position before) and sell one stock whenever the
price level K is crossed from above (and we were long one stock before). Building up
and subsequently unwinding the stock position thus yields to a loss of . Since the
mathematical model of a geometric Brownian motion is a process with continuous
sample paths,  can theoretically made arbitrarily small. However, smaller values of
 lead to more frequent transactions.
Real world stock prices however dont take values in R but are usually quoted in full
USD-cents. Furthermore, they dont move continuously but jump. Thus, each time
the stock price crosses the strike price the hedger encounters a friction of at least
one USD-cent. Furthermore, the strategy ignores transaction cost that can become
very significant for high number of buys and sells in the stock.
(ii) Taking expectation and using that by Theorem 4.3.1(i) the Ito integral is a martingale with zero expected value yields E [X(T )] = 0. However,

25

E (S(T ) K)




+

1 2
S(0) exp T + T x K
=
N 0 (x)dx
2




Z

1 2
=
S(0) exp T + T x K N 0 (x)dx
S(0)
ln(
1 2
K ) 2
2

> 0.
From the second equation and the properties of the normal distribution it is obvious
that the integrand is non-negative everywhere and that it takes strictly positive
values for some values of x. Thus, the expected value is strictly positive as well. In
conclusion

P X(T ) = (S(T ) K)+ < 1,
i.e. there are some paths where the stop-loss start-gain strategy does replicate the
option payoff (i.e. all paths where the maximum stock price over the interval [0, T ]
is strictly less than K). However, since the expected value of the payoff of the
strategy does not agree with the expected value of the option payoff, we do not have
X(T ) = (S(T ) K)+ almost surely. This also follows from the result on the local
time in Exercise 4.21. Since a Brownian motion spends a non-zero time at any level
with positive probability, so does it at the strike K. But once the strike is crossed,
the hedger incurs a cost and the exact replication fails.

26