Financial Modelling

Extra Questions - Solutions

October 2016

Solution 1. A random variable random which takes the value 1 with success
probability of p and the value 0 with failure probability of q = 1 − p can be
represented by using the Bernoulli distribution. The expected value of a
Bernoulli random variable is

E(X) = 1 × p + 0 × (1 − p)
= p = 0.75
while its variance is

Var(X) = E(X 2 ) − E(X)2

= (1)2 × p + (0)2 × (1 − p) − E(X)2
= p − p2 = p(1 − p)
= 0.75 × 0.25 = 0.1875
The correct answer is (III).

Solution 2. The Binomial random variable can be used to describe the

number of success in n independent trials. Therefore, the number X of
successes in n independent and identical Bernoulli trials with probability p
of success is a Binomial(n, p) random variable.
The correct answer is (II)

Solution 3. We are in the situation described in the Figure 1. A standard

Normal random variable has a symmetric distribution around 0. Therefore,

1
 Figure 1: Standard Normal Random Variable

the area under the curve below 0 is exactly equal to 0.5. We need to find the
value of x such that the area under the curve is exactly equal to 0.25.
Looking at the table of the Standard Normal distribution, the correct answer
is (III).

Solution 4. Any normal random variable has a symmetric distribution

around its mean. It implies that 50% of values are less than the mean and
the rest 50% are greater than the mean. Therefore, any normal random
variable has the same value of both mean and median.
The correct answer is (IV).

Solution 5. Let X ∼ (0, 4). We need to calculate the P r(X > 2). Let
us first standardize X. If X ∼ N (0, 4), then X−0
√
4
∼ N (0, 1). Denoting by
Z ≡ N (0, 1), we have

2
 ( )
X −0 2−0
P r(X > 2) = P r √ > √
4 4
= P r (Z > 1)

= 1 − P r(Z ≤ 1)

= 1 − 0.8413 = 0.1587

The correct answer is (III).

Solution 6. We need to derive the P r(X̄ > 2/3), where X̄ is the sample
mean of X. Using the properties of Normal random variable, we know that

( 2
)
σX
if X ∼ 2
N (µX , σX ), then X̄ ∼ N µX ,
n

That is, the sample mean X̄ is still normally distributed with the same mean
of X and a variance equal to the variance of X divided by the sample size.
Therefore, in our case, we have that X̄ ∼ N (0, 4/9). Having derived the
distribution of X̄, we can now derive the required probability. That is

( )
X̄ − 0 2
− 0
P r(X̄ > 2/3) = P r √ >√ 3
4/9 4/9
( )
2/3
= Pr Z > = P r (Z > 1)
2/3
= 1 − P r(Z ≤ 1)
= 1 − 0.8413 = 0.1587

The correct answer is (III).

Solution 7. X ∼ N (0, σ 2 ) such that P r(X > 6) = 0.0013. Let us first

standardize our information. Since Xσ ∼ N (0, 1), we have

3
 ( )
X −0 6−0
P r(X > 6) = 0.0013 ⇒ Pr √ > √ = 0.0013
σ2 σ2
( )
6
⇒ Pr Z > = 0.0013
σ
( )
6
⇒ 1 − Pr Z ≤ = 0.0013
σ
( )
6
⇒ Pr Z ≤ = 1 − 0.0013
σ
⇒ P r (Z ≤ x) = 0.9987

where we denote x = σ6 . Let us now look at the table of the standard Normal
distribution (follow the same idea of Question 3): we need to find the value
of x such that the area under the curve below x is equal to 0.9987. The table
indicates that x = 3. It implies that
6
x=3 ⇒ =3
σ
6
⇒ σ= = 2.
3
The correct answer is (I).

Solution 8. Both X and Y are normally distributed random variables.

Since Z is just a linear combination of two Normal random variables, the
properties of Normal distribution imply that Z will be normally distributed
too. Let us calculate its mean and variances. Remember the following rules
regarding the sum (difference) between two random variables, say A and B:

Var(A + B) = Var(A) + Var(B) + 2Cov(A, B)

Var(A − B) = Var(A) + Var(B) − 2Cov(A, B)

Var(cA) = c2 Var(A), for any constant c.

4
Therefore:

E(Z) = E(X − 2Y )
= E(X) − 2E(Y )
= 4−4=0

and

Var(Z) = Var(X − 2Y )
= Var(X) + 4Var(Y ) − 4Cov(X, Y )
= 2 + 16 + 0 = 18

Hence, the correct answer is (I).

Solution 9. Let X ∼ N (8, 6) be the random variable describing the time

between the graduation and the first job of a student. We need to derive the
P r(X > 12). Using the fact that X−8
√
6
∼ N (0, 1), we have

( )
X −8 12 − 8
P r(X > 12) = P r √ > √
6 6
= P r(Z > 1.632)

= 1 − P r(Z < 1.632) = 0.0516

The correct answer is (I).

Solution 10. We need to derive the confidence interval at 95% level for µ
of a Normal random variable with σ known. First notice that the 95% level
implies that

α
1 − α = 0.95 ⇒ α = 0.05 ⇒ = 0.025 ⇒ zα/2 = 1.96.
2
Moreover, we know that if X ∼ N (µ, σ 2 ), then X̄ ∼ N (µ, σ 2 /n) or, equiva-
lently

5
 X̄ − µ
√ ∼ N (0, 1).
σ/ n
Therefore

( )
X̄ − µ
P r −1.96 ≤ √ ≤ 1.96 = 0.95
σ/ n
( )
σ σ
⇒ P r X̄ − 1.96 √ ≤ µ ≤ X̄ + 1.96 √ = 0.95,
n n

and the 95% confidence interval for µ is

[ ]
σ σ
CI95% = x̄ − 1.96 √ , x̄ + 1.96 √
n n
[ ]
2 2
= 12 − 1.96 √ , 12 + 1.96 √
100 100
= [11.6, 12.4]

The correct answer is (III).

Solution 11. Since the sample size is large (n = 100), the Central Limit
Theorem ensures that the confidence interval derived in Question 10 is still
valid even without the normality assumption. The correct answer is (II).

Solution 12. Consider, as an example, the confidence interval for the mean
µ with known variance. Then, in this case the confidence interval is

[ ]
σ σ
CI = x̄ − zα/2 √ , x̄ + zα/2 √
n n
As the confidence level 1−α decreases, the quantity zα/2 decreases too. Hence
the length of the confidence interval reduces. For example, when 1−α = 0.95,
then zα/2 = 1.96. When 1 − α = 0.90, then zα/2 = 1.64. Consider the same
information of Question 10, with x̄ = 12, σ = 2 and n = 100. Then

6
 CI95% = [11.6, 12.4] ⇒ Length CI95% = 12.4 − 11.6 = 0.8

while

CI90% = [11.7, 12.3] ⇒ Length CI90% = 12.3 − 11.7 = 0.6

The correct answer is therefore (I).

Solution 13. Let X ∼ N (3, 0.1) a Normal random variable describing Firm
A’s daily income. We need to calculate the P r(X ≤ 3.5). Using the fact
that X−3
√
0.1
∼ N (0, 1), we have

( )
X −3 3.5 − 3
P r(X ≤ 3.5) = P r √ ≤ √
0.1 0.1
= P r(Z ≤ 1.582) = 0.9429

The correct answer is (IV).

Solution 14. We need to derive the P r (2.9 ≤ X ≤ 3.5). Using again the
fact that X−3
√
0.1
∼ N (0, 1), we have

P r (2.9 ≤ X ≤ 3.5) = P r(X ≤ 3.5) − P r(X ≤ 2.9)

( ) ( )
3.5 − 3 2.9 − 3
= Pr Z ≤ √ − Pr Z ≤ √
0.1 0.1
= 0.5684

The correct answer is (I).

Solution 15. We know that Firm A’s daily income is a Normal random
variable with mean 3 and variance 0.1. Therefore, Firm A’s monthly income
corresponds to the sum of 30 independent random variables with the same

7
distribution each day. Formally, let Xi ∼ N (3, 0.1) be the random variable
describing the daily income. Then we can think of monthly income as

∑
30
Xmonthly = Xi
i=1

Using the properties of Normal random variables, we know that the sum of
n independent Normal random variables is itself a Normal random variable
with mean equal to

( )
∑
30 ∑
30
E(Xmonthly ) = E Xi = E (Xi )
i=1 i=1

= 30 × E(Xi ) = 30 × 3 = 90

and variance equal to

( 30 )
∑ ∑
30
Var(Xmonthly ) = Var Xi = Var (Xi )
i=1 i=1

= 30 × Var(Xi ) = 30 × 0.1 = 3

Hence, Xmonthly ∼ N (90, 3). We can now easily derive the P r(Xmonthly ≤ 95)
as

( )
95 − 90
P r(Xmonthly ≤ 95) = P r Z ≤ √ = 0.998
3
The correct answer is (III).

Solution 16. We need to derive the distribution of the average income, say
X̄, using a sample of n = 30 days. Using the properties of Normal random
variables, we know that

if X ∼ N (µX , σX
2
), then X̄ ∼ N (µX , σX
2
/n)

8
Hence, in our case

X ∼ N (3, 0.1), then X̄ ∼ N (3, 0.1/30)

and the correct answer is (III).

Solution 17. Using the t-test, we have that we will reject the null hypothesis
at 5% level if

x̄ − µ0
√ > tn−1,0.025
s/ n

Using our data, we have that x̄−µ

√ 0 = 0.79, while t9,0.025 = 2.262. Therefore,
s/ n
we cannot find evidence to reject H0 . The correct answer is (III).

Solution 18. We know that we cannot reject the null hypothesis at 5%. It
means that, at 5% level, the observed quantity x̄−µ √ 0 is not in the rejection
s/ n
region. Using the same data, the rejection region at 1% level will be even
smaller, implying that, for sure, x̄−µ
√ 0 will not be in the rejection region at
s/ n
1% level. The correct answer is therefore (I).