MTH 223 Mathematical Risk Theory 2018/19 Final EXAMINATIONS
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
MTH 223
1st SEMESTER 2018/19 Final EXAMINATIONS
Mathematical Risk Theory
Formula Sheet
1. Discrete Distribution
· Poisson with parameter λ > 0 : A random variable X is said to have a Poisson distribution
denoted by X > P (λ) if X has the following probability function (p.f.):
λk e-λ
Pr(X = k) = k! , k = 0, 1, 2, ...
with 匝(X) = Var(X) = λ;
the moment generating function MX (t) = 匝(etX ) = exp eλ(et _ 1){ ; and the probability generating function PX (z) = 匝 ╱zX、= expeλ(z _ 1){.
2. Continuous Distribution
· Gamma with parameters α > 0 and β > 0 : A random variable X is said to have a gamma
distribution denoted by X > G(α, β) if X has the following probability density function (p.d.f):
xα - 1 e- 台(扌)
f (x) = Γ(α)βα , x < 0
匝(X) = αβ, Var(X) = αβ2 , and MX (t) = ╱ 、α for t < .
The gamma function Γ(α) is defined by
Γ(α) = .0 o xα -1 e-北 dx, α > 0
with Γ(α + 1) = αΓ(α) for α > 0 and Γ(n + 1) = n! for n = 0, 1, 2, ...
· Exponential with parameter θ > 0. For x > 0 and k = 1, 2, ...,
f (x) = e-北/θ , F (x) = e-a(扌) , 匝[Xk ] = θ
· Normal with parameters µ > 0 and σ 2 > 0 : A random variable X is said to have a normal distribution denoted by X > N(µ, σ2 ) if X has the following pdf:
1 ( 扌 −k)2
f (x) = ^2πσe- 2扌2 , _~ < x < ~
with 2 2
匝[X] = µ, Var(X) = σ 2 , and MX (t) = eµt+ 扌 2t PAPER CODE: MTH 223/2018/19/Final EXAM Page 2 of 8
· Pareto with parameters α > 0 and β > 0 : A random variable X is said to have a Pareto
distribution denoted by X > Pareto(α, β) if X has the following probability density function:
αβ α
f(x) =
x < 0, or equivalently, X has the following distribution function
F (x) = 1 _ ╱ 、α ,
x < 0
3. Others
· (a, b, 0) class
pk -1 = a + k , k = 1, 2, ...
· Zero-truncated Poisson with parameter λ > 0.
P1(T) =
PT =
eλ _ 1 , a = 0, b = λ,
λk
k!(eλ _ 1),
eλz _ 1
eλ _ 1 .
· Zero-modified distributions If we have a member of the (a, b, 0) class and replace p0 with
p0(M) , then the resulting distribution is a member of the (a, b, 1) class since
pk(M) b
· Panjer’s Recursion Let S be a compound random variable with primary random variable
N and secondary random variable M. Assuming that
– N has probability mass function epn {
– N is a member of the (a, b, 1) class, and
– M has probability mass function efj { , then the probability mass function egk { of S satisfies
[p1 _ (a + b)p0]fk +3j(k)=1 ╱a + 、fj gk -j
gk = 1 _ af0 , k = 1, 2, 3...
Q 1. Suppose the ground-up loss random variable X has the probability density funtion
f (x) = b , 、+ (1 _ b) , 、, x < 0,
for b o (0, 1), α > 2 and θ > 0.
(a) Show that the survival distribution of the random variable X, denoted by FX (x) or
SX (x), is
FX (x) = SX (x) = ╱ 、α – b + (1 _ b) ╱ 、α ! .
(b) Determine the mean of X.
(c) Show that the Value-at-Risk at level 100p%, namely VaRp (X), is given by VaRp (X) = θ ! (-1/α _ 1 .
(d) Define Y to be an exponential random variable of identical mean to X.
i Find an expression for VaRp (Y) in terms of b, α and θ . ii Find an expression for TVaRp (Y) in terms of b, α and θ .
(e) By determining the limiting ratio of the survival functions, find which random variable
of X and Y have the heavier tail.
Q 2. Suppose that X is a mixture with the following components:
hXlΛ(x}λ) = λx, x < 0
· Λ is a Gamma rv with parameters (2, β).
(a) Show that the survival function of X is
SX (x) = ╱ 、2 ,
for x > 0. [6 marks]
(b) Determine whether the distribution of X is a scale distribution and possesses a scale parameter. Justify your answer.
[6 marks]
(c) Determine 匝[X]. (Hint: use the fact that 匝[X] = 匝[匝[X}Λ]] and Γ(0.5) = ^π ).
[6 marks]
Q 3. Assume the ground-up loss X has a pure discrete distribution. The probability mass function
of X is given as follows:
x |
10 |
25 |
60 |
90 |
100 |
Pr(X = x) |
0.3 |
0.35 |
0.21 |
0.09 |
0.05 |
Suppose a limit of 80, an ordinary deductible 15 and a coinsurance factor of 0.8 are applied to the loss X .
(a) Determine the cumulative distribution function of YP , the amount paid per payment. [8 marks]
(b) Determine the value of 匝[YP ].
Q 4. Assume that an insurer’s portfolio has aggregate claims S where
S = (
The primary random variable M has probability generating function
ln(1 _ β *t)
PM (t) = 1 _ q + q
where
ln(1 _ β(1 _ α))
ln(1 _ β)
and
β * = αβ = 1
Also, the random variable’s eXi { form a sequence of i.i.d. random variable’s with common (geometric) probability mass function
fj = ╱ 、 j+1 , j = 0, 1, 2, ...
It is further assumed that the random variable’s eXi { are independent of M.
(a) Find 匝[S].
(b) Calculate Var(S).
(c) Use the normal approximation to approximate Pr(S ● 1).
(d) Use Panjer’s recursive algorithm to find Pr(S ● 1).
Q 5. Consider a counting random variable N with pmf
pk =
qk |
_k ln(1 _ q) , |
k = 1, 2, ...
and p0 = 0, where 0 < q < 1 is a constant.
(a) Show that N is a member of (a, b, 1) class and identify the value of the parameters a
and b.
(b) Calculate 匝[N].
(c) Consider a counting random variable M with probability mass function p0 = 0.2, p1 = 0.3, p2 = 0.3, and p3 = 0.2.
Is M is a member of (a, b, 0) class? Justify your answer.
2022-07-30