Professional Documents
Culture Documents
The University of New South Wales Month of Examination - Nove:Mber 2008
The University of New South Wales Month of Examination - Nove:Mber 2008
ACTL2003
STOCHASTIC MODELS FOR ACTUARIAL APPLICATIONS
INSTRUCTIONS:
1. TIME ALLOWED - 3 HOURS.
\)
.~
0 0 0 54
1
1
"3 "3 0 "3 0
0 0 21 0 "21
1
1
0 1 0
i 0"2 1. 04 0
"2
2
~
i) Classify the states, and determine whether they are recurrent or transient [1
marks].
ii) Find the expected number of visits in each transient state starting in any transient
state. [2 marks]
iii) Given that X o is equally likely to be in any state, calculate the probability that
two steps later, the process is in state 3. [2 marks]
b) Consider a matrix
x I _ y2 )
( I-x
Y
i) Find all the possible values of x and y such that the above matrix is a probability
transition matrix? [2 marks]
ii) Show in detail when the above matrix will be a probability transition matrix of a
Markov chain whose limiting probabilities exist, and when it will not. [3 marks]
Question 3 (6 marks)
a) A casino has a game that makes payouts according to a Poisson process with rate 5 per
hour. The payout amounts are 1,2,3, ... without limit, and the probability that any
given payout is equal to i is
Payouts are independent. Calculate the probability
that there are only payouts with amounts 1 or 4 in a given 10 minutes period. [2
it.
marks]
b) The time elapsed between arrivals of customers is modeled such that Vk represents the
time elapsed between the arrivals of the (k - 1)th and kth customer. (VI =the time the
first customer arrives). Suppose that: i) Vi, "'2, '" are mutually independent; ii) The
pdf of each Vk is f(t) = O.le- Olt , t > 0, where t is measured in minutes.
Determine the probability that at least two customers arrive in a ten-minute period.
[1 marks]
c) Lucky Jack finds coins on the way to work according to a Poisson process with a mean
of 22 coins per month.
5% of the time, Jack finds coins worth a total of 10.
15% of the time, Jack finds coins worth a total of 5.
80% of the time, Jack finds coins worth a total of 1.
The value of the coins he finds is independent of the number of coins.
the variance of the total amount he finds in a month. [3 marks]
Calculate
a) Give the generator matrix for transitions between the four states. [2 marks]
b) Give the Kolmogorov backward equations for this model. [2 marks]
c) If a resident is in the Health Center currently, calculate the probability that he/she
will stay there for the next 2 years. [2 marks]
d) Determine the probability that a healthy resident at the beginning of Year 1 will remain
healthy for at least 2 years, then get sick and not leave the Health Center before the end
of year 5. Clearly explain the formula you use to determine your answer. [4 marks]
;2 t ) is a martingale.
b) Define yt = cB fz' t 2:
o.
[2 marks]
0:,
dX t
(c + 2a )Xtdt + o:XtdBt .
[3 marks]
for x 2: 0
= [-
In (1 - u)
]-1
Is this method and formulae correct? If not give a corrected method and formulae. [3
marks]
(c) Random variates are generated from a gamma distribution
Aa
f(x) = f(0:)xa-1e- AX , x> 0
Twenty values are generated for each sample and the mean and standard deviation of
each of the samples, to 2 decimal places, are as follows
Sample
Mean
Standard Deviation
12.01
6.15
11.79
4.73
3
13.43
6.42
14.01
7.02
5
11.44
4.30
6
11.19
3.90
7
11.24
3.84
8
12.42
4.35
9
12.91
3.59
i) Use all of the samples to estimate the mean and variance of the gamma distribution. Determine estimates of the parameters A and 0:. [5 marks]
ii) Determine the number of simulated values required in order for the estimate of
the mean of the distribution to be within 5% of the true value with (approximate)
probability 95%. [4 marks]
7
10
12.29
4.60
Question 7 (6 marks)
a) Find a filter of the form
that passes linear trends without distortion and that eliminates seasonal components
of period 2. [3 marks]
b) Show that the filter with coefficients
passes second degree polynomials and eliminates seasonal components with period 3.
[3 marks]
rv
WN(O, (72).
a) Find the relationship between Var(Xt ) and Var(Xt-d. Comment on the stationarity
of {Xd. [2 marks]
b) Find the mean of X t and explain why /-l is referred to as the "drift". [3 marks]
c) Find formulae for Var(Xt ) and Cov(Xt , X t +k ) in terms of
(72
and t. [5 marks]
Y; = "6Y;-1 - "6Y;-2
where {Zd
rv
+ Zt,
[6 marks]
c) Give an expression for the general form of an ARlMA(p, d, q) process, defining and
explaining all of the terms in the expression. [3 marks]
d) For the ARlMA(p, d, q) process
x=
Zt
rv
N (0,1) i.i.d.
END OF EXAMINATION