School of Mathematical

Sciences

MTH3251
Financial Mathematics

ETC3510
Modelling in Finance and
Insurance

Exercise Book

Semester 1, 2015
 Unit Schedule

An approximate week-by-week activity schedule is given below.

UNIT SCHEDULE

Week Activities Assessment

1 Brownian Motion No formal assessment

2 Brownian Motion No formal assessment

3 Brownian Motion Homework

4 Martingales Homework

5 Stochastic Calculus Homework

6 Stochastic Differential Equations Homework

7 Options pricing. Black-Scholes model. Homework & Assignment 1

8 Fundamental Theorems. Interest Rates Homework

9 Random Walk Homework

10 Applications in Insurance Homework

11 Simulations. Convergence Homework & Assignment 2

12 Simulation methods Homework

Swot VAC Revision No formal assessment
1 Random variables, distributions, multidi-
mensional random variables

1. Show that the mean of N (µ, σ 2 ) distribution is µ.

2. Show that the variance of N (µ, σ 2 ) distribution is σ 2 .

3. Show that for any random variable X with a finite second moments
E(X 2 ) < ∞,
V ar(X) = E(X 2 ) − (EX)2 .

4. Derive the expression for the probability density of a LogNormal dis-

tribution.
1 (ln x−µ)2
f (x) = √ e− 2σ2
2πσx
Hint: Write the Lognormal cdf in terms of Normal cdf, then differenti-
ate.

5. X has N (20, 22 ) distribution and Y has N (10, 1) distribution and they

are independent.

(a) Give the distribution of the vector (X, Y ).

(b) Give the distribution of U = 2X − 3Y , specify its mean and
variance.
(c) Using Excel, simulate 100 values of U . Give the mean and the
standard deviation of these values and compare to the numbers in
above.

6. X has N (20, 22 ) distribution and Y has N (10, 1) distribution, and X

and Y are correlated with correlation ρ = 0.5.

(a) Give the covariance between X and Y .

(b) Give the distribution of the vector (X, Y ).
(c) Give the distribution of U = 2X − 3Y , specify its mean and
variance.

1
 (d) Using Excel, simulate 100 values of U . Give the mean and the
standard deviation of these values, and compare to the numbers
in above.

7. Let X and Y be random variables with E(X) = 20, V ar(X) = 4, and

E(Y ) = 10, V ar(Y ) = 1, ρ = 0.5
Find the covariance and the correlation between X and U = 2X − 3Y .

8. Let X1 and X2 both have Lognormal LN (0, 1) and are independent.

Find the following

(a) P (X1 > 2).

(b) Distribution of X1 X2 , and P (X1 > 2/X2 ).
(c) Distribution of X1 /X2 , and P (X1 > 2X2 )
" #
1 ρ
9. X = (X, Y ) is N (0, Σ), with Σ = .
ρ 1

(a) Find P (X > Y ). Hint: write this as P (X − Y > 0) and use

Theorem 11, p. 18 in the Notes.
(b) Write an integral expression for P (X ∈ D), where D is a set on
the plane. Evaluate it for D = {(x, y) : x ≤ y}.

10. Let random variables X and Y be independent Normal with distribu-

tions N (µ1 , σ12 ) and N (µ2 , σ22 ). Show that the distribution of (X, X +Y )
is bivariate
" Normal#with mean vector (µ1 , µ1 + µ2 ) and covariance ma-
σ12 σ12
trix .
σ12 σ12 + σ22
Hint: If Z = (Z1 , Z2 ) has standard normal components, show that
" # " # " #" #
X µ1 σ1 0 Z1
= + .
X +Y µ1 + µ2 σ1 σ2 Z2
| {z } | {z } | {z }
µ A Z

11. Rule for transformation of covariances in linear transformation of random

vectors. Let the random vector (X, Y ) have covariance matrix Σ, and

2
 " # " #
U X
=A
V Y

for some non-random matrix A. Derive the formula for Σ(U,V ) , the
covariance matrix of (U, V ),

Σ(U,V ) = AΣ(X,Y ) AT .

a) By using vector-matrix multiplication and properties of the covari-

ance.
b) Assume (X, Y ) is bivariate Normal. Derive the formula for Σ(U,V )
using representations for multivariate normal and matrix multiplica-
tion.

12. Show that an average of n independent identically distributed Normal

random variables N (µ, σ 2 ) is a Normal random variable with the mean
µ and variance σ 2 /n.

(a) by using Theorem 11,

(b) by using moment generating function m(u) of N (µ, σ 2 ), m(u) =
2 2
eµu+σ u /2 .

13. Simulate two observations on the bivariate normal (X1 , X2 ) with mean
0, V ar(X1 ) = 1, V ar(X2 ) = 4 and correlation ρ = 0.9.

3
2 Brownian motion
1. Let Z denote a standard
√ Normal random variable. Give the distribution
of the process Ut = Z t. Show that for any time t, the distribution of
Ut is the same as that of the Brownian motion Bt , but the process Ut
is not a Brownian motion. Sketch a graph of Ut as a function of t.

2. Using Excel, generate the values of Brownian motion at points ti =

i/100, i = 0, . . . 100 and plot them. Hint: use the fact that incre-
ments of Brownian motion are independent and Normally distributed,
√
B(ti+1 ) − B(ti ) = Zi ti+1 − ti , where Zi is a sequence of standard
Normal variables.

3. Define Vt = 2Bt + 3t, where Bt is a Brownian motion. Find the mean

and covariance functions of the process Vt .

4. Bt is a Brownian motion.

(a) Show that the process Wt = −Bt is also a Brownian motion.

(b) Show that the process Ut = √1 B2t is also a Brownian motion.
2
(c) Show that the process Vt = Bt+1 − B1 is a Brownian motion.

5. Let B1 (t) and B2 (t) be independent

√ Brownian motions, and define
W (t) = (B1 (t) + B2 (t))/ 2.

(a) Show that W (t) is also a Brownian motion.

(b) Find the covariance between W (t) and B1 (t).

6. Show that the process Bt3 is not a Brownian motion.

4
3 Conditional Expectation. Martingales
1. Y and V are independent, E(V ) = 0.
X = c + aY + bV , where a, b, c are constants.

(a) Find E(X|Y ) by using the properties of conditional expectation.

(b) Comment on the best predictor of X based on Y .

2. Y and W are independent.

X = W g(Y ), where g(y) is some bounded function.

(a) Find E(X|Y ) by using the properties of conditional expectation.

(b) Comment on the best predictor of X based on Y .

3. Find the following conditional expectations for Brownian motion Bt

(a) of B2 given B1
(b) of B1 given B2

4. Bachelier model for stock price. For t ≤ T,

St = S0 + µt + σBt ,

where Bt is Brownian motion started at zero, and µ, σ > 0 are con-

stants.
Find

(a) the marginal distribution of St

(b) the joint distribution of St and ST
(c) the conditional distribution of ST given St
(d) the best predictor of the future price ST based on the price St .

5
5. Black-Scholes model for stock price. For t ≤ T,

St = S0 eσBt +µt ,

where Bt is Brownian motion started at zero, and µ, σ > 0 are con-

stants. Find

(a) the marginal distribution of St

(b) the conditional distribution of ST given St
(c) the best predictor of the future price ST based on the price St .

6. Show that the process e2Bt −2t is a martingale.

7. (a) It is known that the process Bt2 − h(t) is a martingale with mean
zero. Find h(t).
(b) It is known that the process eBt g(t) is a martingale with mean
one. Find g(t).

8. Show that if Mt is a martingale then aMt + b for any constants a, b is

also a martingale.

9. Show that if Mt is a martingale then Mt2 is not a martingale.

6
4 Itô integral, Itô’s formula and SDE’s


2,

 if 0 ≤ t ≤ 1,
1. (a) Let X(t) = 3, if 1 < t ≤ 3,
−5, if 3 < t ≤ 4.



or in one formula X(t) = 2I[0,1] (t) + 3I(1,3] (t) − 5I(3,4] (t). Give the
Itô integral 04 X(t)dB(t) as a sum of random variables, give its
R

distribution, specify the mean and the variance.

(b) Show that the process M (t) = 0t X(s)dB(s), 0 ≤ t ≤ 4, is a
R

Gaussian process and a martingale.

2. Give the values of α for which the stochastic integral 01 s−α dB(s) is
R

defined. In the case when the integral is defined, give its mean and
variance.

3. Using Itô’s formula, find the following stochastic differentials dXt and
give their integral representations

(a) Xt = eBt
(b) Xt = sin(Bt )
(c) Xt = sin(Bt2 )
(d) Xt = t + Bt2 .
Rt
4. Show that 0 Bs dBs , 0 ≤ t ≤ T , is a martingale

a) by using its closed form expression

b) by checking conditions for Itô integral to be a martingale

5. Show that eBt −t/2 is a martingale by using Ito’s formula for the function
ex−t/2 and properties of Ito’s integral.

6. Solve the SDE

dXt = 0.2dt + 0.4dBt , X0 = 0.
Give the distribution of Xt . Give the probability density function for
X1 .

7
7. Solve the SDE

dXt = 0.2Xt dt + 0.4Xt dBt , X0 = 1.

Give the distribution of Xt . Give the probability density function for

X1 .

8. Solve SDE
dXt = −Xt dt + dBt , X0 = 1.

9. The price of stock is given by St = 10eBt , 0 ≤ t ≤ 1, t denotes time in

years.

(a) Derive the SDE for St .

(b) Find the mean of St .
(c) Find the probability that the stock at the end of the year will
outperform its mean, i.e. find the probability P (S1 > E(S1 )).

8
5 Options pricing, Fundamental Theorems,
models for interest rates

1. (a) Write the payoff function for one share of stock.

(b) Write the payoff function for the combination of one bought call
option and one bought put option with the same exercise price.
(c) Write the payoff function for the combination of one bought call
option and one sold put option with the same exercise price.
2. A stock is currently traded at $20 per share. Price a call option on this
stock with exercise price of $20 and expiration in six months. Assume
that historical volatility is 60% per annum, and riskless rate is 5% per
annum.
Use the discount factor in continuous time e−0.05T and in discrete time
1/(1 + 0.05T ), where time T is the time to expiration expressed√in
years. The parameters of the Binomial formula are given by u = eσ T ,
d = 1/u.

(a) Find the price of the call by using the Black-Scholes formula.
(b) Find the price of the call by using one-step Binomial formula.
(c) Find the price of the call by using two-step Binomial formula.
(d) Price the put option with exercise price $20 and expiration in six
months by using one-step Binomial formula.

3. Each of the following market models consist of the savings account

βt = e0.1t and the stock St . Decide which market model is free from
arbitrage by using the First Fundamental Theorem of Asset Pricing.
(a) St = S0 e0.2Bt .
(b) St = S0 e0.2 sin t .
(c) Describe an arbitrage strategy if there is one.

9
4. Consider the one-step Binomial pricing model with interest rate 10%
per year (r = 1.1).

(a) Derive the no-arbitrage conditions by using the First Fundamental

Theo- rem defining all terms used, and specifying the EMM (risk-
neutral) prob- ability measure.
(b) Describe an arbitrage strategy when no-arbitrage conditions are
violated.
(c) Show that under the EMM (risk-neutral) probability the expected
return on stock is the same as the riskless rate r = 1.1.

5. We assume the Black-Scholes model.

(a) Describe the EMM (risk-neutral) probability measure by writing

the ex- pression for the St under this measure, and show that the
discounted price process is a martingale.
(b) By using the Second Fundamental Theorem write the expression
for the price of a call option at time t < T , where T is the expira-
tion time of the option.
(c) Explain how to obtain the Black-Scholes formula from the expres-
sion in (b).

6. Assume St is a stock price in the Bachelier model St = 10 + Bt , and

βt = 1 represent a savings account (r = 0). Consider the portfolio with
the number of shares at = 2Bt and amount bt = −t − Bt2 − 20Bt in the
savings account. Show that this portfolio is self-financing. State with
reason whether this portfolio is an arbitrage strategy.

7. Let forward rates be given by f (t, T ) = 2a − ae−b(T −t) where a, b > 0.

(a) Give the initial term structure f (0, T ) and sketch its graph (as a
function of T ).
(b) Sketch the graph of f (t, T ) as a function of T . Find P (t, T ) and
the yield curve R(t, T ) and sketch their graphs.

8. The short rate rt is modelled by the Vasicek’s model.

10
 (a) Find the conditional distribution of rT given rt .
(b) Find the forward rates f (t, T ) and express them as a function of
the short rate rt .
(c) Show that for any fixed t, f (t, T1 ) and f (t, T2 ) have correlation 1.

9. Let the term structure be given by the following model (Ho-Lee)

f (t, T ) = f (0, T ) + σ 2 t(T − t/2) + σBt .

(a) Find the short rate rt .

(b) Write the equation for forward rates in terms the short rate.
(c) Find the expression for the bond. Consider P (0, T ) as known.

11
6 Random Walk, Martingales, Optional Stop-
ping Theorem, and models in Insurance
1. Let (X1 , X2 ) have a standard bivariate Normal distribution. Show that
{X1 , ρ1 X2 } is a martingale. (Here time takes only two values n = 1, 2.)
2. Let Xn , n = 0, 1, 2, . . . denote an unbiased Normal Random Walk.
X0 = 10, and Xn+1 = Xn + Yn+1 , with {Yn } are i.i.d. N (0, 1).
(a) Show that Xn is a martingale. Give EX20 .
(b) Show that Mn = Xn2 − n is a martingale. Give EM20 .
√
2Xn −n
(c) Show that Vn = e is a martingale.
(d) Generate the values of X0 , X1 , X2 , . . . , X100 and plot them as a
function of n = 0, 1, 2, . . . 100. Also plot first 100 values of Mn
and Vn .
(e) Let τ be the first time when the Random Walk becomes greater
than 20 or less than 0. By writing the event {τ > n} in terms
of the variables of the Random Walk, show that it is a stopping
time. From the simulation above give either the value of τ or a
bound on its value.
3. The random variables Y1 , Y2 , . . . , Yn , . . . are independent and identically
distributed (i.i.d.) with LogNormal distribution LN (0, 1). The process
Xn , n = 0, 1, 2, . . . is defined as follows. X0 = 1, and for n = 0, 1, 2, . . .,
Xn+1 = Xn Yn+1 .

(a) Denote m = EY1 . Find m.

(b) Show that the process Mn = m−n Xn , n = 0, 1, 2, . . ., is a martin-
gale.

4. In a game of roulette, there are 37 squares, 18 of which are red and 18

black, 1 square is neither. Suppose you bet on a colour with $1 each
bet. You start with $50. Find the probability of doubling your money
before ruin, and the probability of ruin before doubling your money.
Formulate the model for your fortune as a Random Walk, apply the
Optional Stopping Theorem, specifying the appropriate martingale and
the stopping time.

12
5. * Let {Xn } be a simple Random Walk started at zero, and τ is the first
time the Random Walk hits the boundaries ±k, for a given integer k.
This exercise shows that P (τ < ∞) = 1, hence Xτ = k or −k with
probability 1.

(a) Let Xn be a simple unbiased Random Walk. Show that P (τ <

∞) = 1 by using the Optional Stopping Theorem with the mar-
tingale Xn2 − n.
(b) Let Xn be a simple biased Random Walk. Show that P (τ < ∞) =
1 by using the Optional Stopping Theorem with the martingale
Xn − (p − q)n.

Questions 1b and 2 to be submitted.

6. The aggregate loss (i.e. total claims payout) in a year has mean µ and
variance σ 2 , the premium collected in a year is c, c > µ, and the initial
fund is x.

(a) Assuming that the loss distribution is Normal give expressions for
the probability that ruin occurs in the first year and in the second
year. Are these probabilities exponentially small in the initial
funds x?
(b) Assuming that the loss distribution function F (y) = 1 − 1/y 4 give
expressions for the probability that ruin occurs in the first year
and in the second year. Are these probabilities exponentially small
in the initial funds x?

7. Assume the following Risk model. The losses are Xt = µt + σBt , where
Bt is Brownian motion started at zero. The collected premium is ct,
and initial capital is x.

(a) Give an expression for Ut , the capital of an insurance company at

time t.
(b) Find R > 0 such that e−RUt is a martingale. Hint: use exponential
martingale of Brownian motion.
(c) Show that the probability of ruin is bounded by e−Rx .

13
7 Simulations, Transforms and Limit Theo-
rems
RAND() returns an observation from U (0, 1) distribution (random variable)
NORMSINV(RAND()) returns an observation from N (0, 1) distribution

1. Simulate 5 values from the distribution with the density

f (x) = 2x, 0 < x < 1.

State the simulation method used.

2. Describe briefly the inverse transform method for simulation of an ob-

servation from a continuous distribution and illustrate it on the exam-
ple of an exponential random variable with the cumulative distribution
function F (x) = 1 − e−x .
√
3. Explain how to evaluate the integral I = 01 (sin x/ x)dx by simulation
R

(Monte Carlo).
R √
4. Explain how to evaluate I = 01 1 − x2 dx by Monte Carlo.

5. Explain what is meant by the pseudo-random numbers and the quasi-

random numbers.

6. (a) Find the characteristic function of a discrete Uniform distribution

on a set of two numbers {−b, b}, and show that it is real.
(b) Find the characteristic function of a continuous Uniform distribu-
tion on an interval [−a, a], and show that it is real.
(c) X is continuous Uniform [−1, 1], Y is discrete Uniform {−1, 1},
and X and Y are independent. Show that the distributions of
X + Y is continuous Uniform [−2, 2].

7. (a) Explain briefly how to simulate Brownian motion process B(t),

0 ≤ t ≤ 1, at 100 equally spaced time points. State briefly the
properties of Brownian motion on which your simulation is based.
Take tn = n/100, n = 0, 1, . . . , 100.

14
 (b) Explain briefly how to simulate solution of stochastic differential
equation
dXt = µ(Xt )dt + σ(Xt )dBt
when a closed form solution is not available.

8. Simulate an observation from Rayleigh distribution with the probabil-

2
ity density function f (x) = xe−x /2 , x ≥ 0, by using the acceptance-
rejection method with exp(1) distribution.

9. (a) Simulate Brownian motion Bt on [0, 1].

2 /2
(b) Simulate exponential martingale of Brownian motion Mt = eBt −t .

10. (a) Simulate the price of the call option with strike $20, on stock
that has the current price $20, expires 9 months from now, has
volatility 30%, and the riskless interest rate is 5%.
(b) Compare the simulated price to the exact obtained by the Black-
Scholes formula. You can use the pricing Black-Scholes macro.

11. Simulate the rate in the Vasicek’s model on the time interval [0, 2]

drt = (0.1 − rt )dt + dBt , r0 = 0.05.

12. (a) Find the moment generating function (mgf) for the Bernoulli vari-
able Y , P (Y = 1) = p and P (Y = 0) = 1 − p.
(b) The sum of n independent identically distributed Bernoulli ran-
dom variables is called the Binomial, Bin(n, p) random variable.
Find its mgf.
(c) By using the mgf found above, find the mean of the Bin(n, p)
distribution.

13. Consider the discrete time Black-Scholes model for prices.

Sn+1 = Sn Xn+1 ,

where Xn , n = 0, 1, 2 . . . are independent and identically distributed

lognormal LN (µ, σ 2 ).

(a) Find the mean and variance of Sn . Specify the distribution of Sn .

15
 (b) Find the rate of growth of Sn and of its expected value ESn .
Namely, find limits n1 log Sn and n1 log ESn

14. Suppose that at each period you can either double your capital with
probability p or lose it with probability 1 − p. Consider the strategy of
reinvesting a constant proportion c of the capital, leaving the propor-
tion 1 − c in a safe place.

(a) Find the proportion c if your strategy is to maximize the expected

value ESn .
(b) Explain why reinvesting the whole capital is not such a good strat-
egy. Hint: Consider the probability of having positive capital at
time n.

16