Nov2012mlc Ko
Nov2012mlc Ko
Solution.
The probability sought is
2 p80:90 − 3 p80:90 = ( 2 p80 + 2 p90 − 2 p80:90 ) − ( 3 p80 + 3 p90 − 3 p80:90 ) =
Using Poohsticks
Solution.
Under constant force over each year of age, for 0 ≤ k ≤ 1, and x an integer,
k
⎛l ⎞
= k px = e− kµx = ( e− µx ) = ⎜ x+1 ⎟ ,
lx+k k
lx ⎝ lx ⎠
so that
lx+k = ( lx ) (lx+1 )k .
1−k
We have
l[ 60 ]+2.75 − l[ 60 ]+5.75 l([ 60 ]+2 )+0.75 − l65.75
1000 2 3 q[ 60 ]+0.75 = 1000 ⋅ = 1000 ⋅ =
l[ 60 ]+0.75 l[ 60 ]+0.75
(l[ ] ) ⋅l − l
0.25
60 +2
0.75
63
0.25
65 ⋅l66
0.75
= 1000 ⋅ =
(l[ ] ) ⋅ (l[ ] )
0.25 0.75
60 60 +1
77000 0.25
⋅ 74000 0.75 − 67000 0.25 ⋅ 65000 0.75
= 1000 ⋅ ≈ 116.70.
80000 0.25 ⋅ 79000 0.75
Answer B.
Solution.
We have
⎛ 1
⎞
⎛ t ⎞4
µt = − ( ln S0 ( t )) = − ⎜ ln ⎜ 1− ⎟ ⎟ =
d d
dt dt ⎜ ⎝ ω ⎠ ⎟
⎝ ⎠
1 d⎛ ⎛ t ⎞⎞ 1 1 ⎛ 1⎞ 1 1
=− ⎜ ln ⎜ 1− ⎟ ⎟ = − ⋅ ⋅⎜ − ⎟ = ⋅ .
4 dt ⎝ ⎝ ω ⎠ ⎠ 4 1− t ⎝ ω ⎠ 4 ω − t
ω
Therefore,
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
1 1 1
= µ65 = ⋅ ,
180 4 ω − 65
so that
1 1
= ,
45 ω − 65
and ω = 110. Based on this
1
⎛ 106 + t ⎞ 4
⎜⎝ 1− ⎟
1
S (106 + t ) 110 ⎠ ⎛ 4−t⎞4
= 0 = =⎜
⎝ 4 ⎟⎠
p106 .
S0 (106 )
t 1
⎛ 106 ⎞ 4
⎜⎝ 1− ⎟
110 ⎠
Therefore,
1 1 1
⎛ 4 − 1⎞ 4 ⎛ 4 − 2 ⎞ 4 ⎛ 4 − 3 ⎞ 4
= 1 p106 + 2 p106 + 3 p106 +
4 p106 + ... = ⎜ +⎜ +⎜ ≈ 2.4786.
⎝ 4 ⎟⎠ ⎝ 4 ⎟⎠ ⎝ 4 ⎟⎠
e106
=0
Answer B.
Solution.
Prospectively, we calculate the reserve as
(
10V = 50,000A50:10 + 100,000 10 A50 − 1116
1
a50:10 = )
= 50,000 ( A50 + 10 E50 ⋅ A60 ) − 1116 ( a50 − 10 E50 ⋅ a60 ) =
= 50,000 ⋅ ( 0.24905 + 0.51081⋅ 0.36913) − 1116 ⋅ (13.2668 − 0.5108 ⋅11.1451) =
≈ 13, 428.
Answer D.
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
(ii) qx = 0.150 and qx+1 = 0.165.
Calculate the level annual benefit premium.
Solution.
The initial reserve at time 0 is 0V = 0. The terminal reserve at time 2 is 2V = 2000, as the
policy endows at that time. Let us write P for the level annual benefit premium sought.
We use the recursive reserve formula twice to find it:
( 0V + P )(1+ i ) = qx ( 2000 + 1V ) + (1− qx ) 1V,
( 1V + P )(1+ i ) = qx+1 ( 2000 + 2V ) + (1− qx+1 ) 2V.
Substituting known values, we obtain
( 0 + P ) ⋅1.1 = 0.150 ⋅ ( 2000 + 1V ) + 0.850 ⋅ 1V,
( 1V + P ) ⋅1.1 = 0.165 ⋅ ( 2000 + 2000 ) + 0.835 ⋅ 2000.
From the first equation, 1V = 1.1P − 300, and after we put this in the second equation
(1.1P − 300 + P ) ⋅1.1 = 0.165 ⋅ 4000 + 0.835 ⋅ 2000,
or
2660
P= ≈ 1151.52.
2.31
Answer C.
Solution.
Using Euler’s method with step size 0.2 and the derivative of tV at time 9.6, we obtain
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
d tV
V ≈ 9.8V − 0.2 ⋅
9.6 .
dt t=9.6
Recall the Thiele’s differential equation when expenses are included:
d
V = δ t tV + Pt − et − ( St + Et − tV ) µ x+t .
dt t
In this case, t = 9.6, δ 9.6 = δ = 0.05, tV is denoted by 9.6V and it is unknown, Pt is a gross
premium with the annual rate of 450, et are premium related expenses at 2% of premium,
St is constant with S9.6 = 106,000, E9.6 = 200, so that
d tV
= 0.05 ⋅ 9.6V + 450 − 0.02 ⋅ 450 − (106,000 + 200 − 9.6V ) ⋅ 0.01 = −621+ 0.06 ⋅ 9.6V.
dt t=9.6
Substituting into the Euler’s formula, we obtain
d tV
9.6V ≈ 9.8V − 0.2 ⋅ = 126.68 − 0.2 ⋅ ( −612 + 0.06 ⋅ 9.6V ) = 250.88 − 0.012 ⋅ 9.6V.
dt t=9.6
This results in
250.88
9.6V ≈ ≈ 247.91.
1.012
Answer B.
A. 39 B. 40 C. 41 D. 42 E. 43
Solution.
The account value (a.k.a., reserve) at the end of year 5 is
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
⎛ ⎞
⎜ ⎟
⎜ 30
+ 20
− 9 ⎟ ⋅ 1.06 ≈ 43.46.
⎜ Account value at Premium Cost of insurance of 2 ⎟ Accumulated at
⎝ the end of year 4 ⎠ credited
plus annual expense the interest rate
charge of 7 (i.e., 6%)
The surrender value at the end of year 5 is therefore this account value of 43.46 reduced
by surrender charge of 20, i.e., 43.46 – 20 = 23.46. The asset share at the end of year 5 is
⎛ AS4 ⎛ Premium Expenses incurred ⎞ ⎞ at asset share rate
Accumulated Expected payout
Expected death benefit payout
at withdrawal
⎜ 20 + ⎜ 20 − 2 ⎟ ⎟ ⋅ 1.08 − 0.001⋅1000 − 0.05 ⋅ 23.46
⎜ ⎜ ⎟⎟
⎜⎝ ⎜⎝ ⎟⎠ ⎟⎠
≈ 40.96.
1− 0.001−
0.05
Probability of policy continuing
Answer C.
Solution.
Let us write CSVk for the cash surrender value at time k, with time counted in months,
and SCk for the surrender charge at time k. Let be COI k be the cost of insurance for
month k. Since i (12 ) = 0.054, the effective monthly interest rate is
i (12 ) 0.054
j= = = 0.0045 = 0.45%.
12 12
Note that the cash surrender value equals the account value (i.e., reserve) minus the
surrender charge. We are given that CSV13 = 1802.94. Based on this,
1802.94 = CSV13 = AV13 − SC13 = AV13 − 125,
so that
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
AV13 = 1802.94 + 125 = 1927.94.
Similarly,
AV11 = CSV11 + SC11 = 1200 + 500 = 1700.
Now we use the recursive formula for the account value (reserve) for universal life with
increasing total benefit and fixed ADB
( )
AVt−1 + Gt (1− rt ) − et (1+ it ) = qx+t ⋅ ( B + AVt+1 ) + px+t ⋅ AVt = qx+t ⋅ B + AVt .
The cost of insurance in this formula is COI t = qx+t ⋅ B, so that
( )
AVt = AVt−1 + Gt (1− rt ) − et (1+ it ) − COI t .
In this case
( )
AV13 = AV12 + G13 (1− r13 ) − e13 (1+ i13 ) − COI13 ,
AV12 = ( AV 11
+ G12 (1− r12 ) − e12 )(1+ i ) − COI
12 12
.
Substituting known values (note that the cost of insurance figures are given per thousand),
we obtain
(
1927.94 = AV12 + 300 (1− 0.15) − 10 ⋅1.0045 − 50 ⋅3, )
(
AV12 = 1700 + 300 (1− W ) − 10 ⋅1.0045 − 50 ⋅ 2. )
From the first equation
1927.94 + 50 ⋅3
AV12 = + 10 − 300 (1− 0.15) ≈ 1823.63116,
1.0045
and substituting this into the second equation, we obtain
1823.63116 + 50 ⋅ 2
− 1700 + 10
W ≈ 1− 1.0045 ≈ 0.25.
300
Answer A.
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
Calculate the present value at issue of the insurer’s expected surrender benefits paid in
the second year.
Solution.
The account value at time 0, just before the first premium is paid is zero. Then, since
death benefit equals fixed ADB plus account value,
( )
AV1 = AV0 + G1 (1− r1 ) − e1 (1+ i1 ) − q60 B = ( 0 + 5000 ⋅1− 100 ) ⋅1.06 − 200 ⋅5.40 = 4114.
Following into the second year,
(
AV2 = AV1 + G2 (1− r2 ) − e2 (1+ i2 ) − q61 B = )
= ( 4114 + 5000 ⋅1− 100 ) ⋅1.06 − 200 ⋅6.00 ≈ 8354.84.
If a policy is surrendered in year 2 (and for a policy in existence at the end of year 1 this
happens with probability 0.06), then it was not surrendered in year 1 (and for a policy
issued at time 0 this happens with probability 1 – 0.06 = 0.94), so that the present value at
time 0 of the expected surrender benefits in year 2 is
⋅ ( 1− 0.06 ) ⋅ (1− 0.0034 ) ⋅ (1− 0.0038) ⋅ 1.07 −2
8354.84
⋅ 0.93
⋅ 0.06 Present value ≈
Account value Adjusted for Fraction that
available for surrenders surrender charge surrenders in Fraction that did 1−q60 1−q61 factor
year 2 not surrender
in year 1
≈ 380.01.
Answer A.
Solution.
The death benefit of $100,000 is paid at the end of year 1 if both (x) and (y) die during
the first year, or it is paid at the end of year 2 if both (x) and (y) are alive at time 1 and
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
both die during the second year, or it is paid at the end of year 3 if both (x) and (y) are
alive at time 2 and both die during the third year. Let us note the following, true for k = 0,
1, or 2:
(
qx+k:y+k = 1− px+k:y+k = 1− px+k + py+k − px+k:y+k =
Poohsticks
)
= 1− ( 0.84366 + 0.86936 − 0.77105 ) = 0.05803.
Also,
px:y = px:y ⋅ px+1:y+1 = 0.77105 ⋅ 0.77105 = 0.77105 2.
2
Solution.
It seems like it will be a lot of calculations to go from age 70 to the limiting age if we do
every year calculation separately. But the Illustrative Life Table of course gives us
actuarial present values of whole life insurance death benefits, except that it does so for
the valuation interest rate of 6%. Here, the spot rates for the first two years are well
below 6%, but beyond that all one-year forward rates are 6% (so that all rates for all
periods starting at time 2 or later are also 6%). Thus starting from age 72, we can
determine actuarial present values from the Illustrative Life Table. We have
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
⎛ q q p ⋅p ⎞
1000A70 = 1000 ⎜ 70 + p70 ⋅ 71 2 + 70 71 ⋅ A72 ⎟ ,
⎝ 1.016 1.026 1.026 2
⎠
where all values on the right-hand side can now be looked up in the Illustrative Life
Table. From the table: q70 = 0.03318, q71 = 0.03626, A72 = 0.54560, and therefore
p70 = 0.96682, p71 = 0.96374, so that
⎛ 0.03318 0.03626 0.96682 ⋅ 0.96374 ⎞
1000A70 = 1000 ⎜ + 0.96682 ⋅ + ⋅ 0.54560 ⎟ ≈
⎝ 1.016 1.026 2
1.026 2
⎠
≈ 548.89.
Answer C.
Solution.
Note that because it is impossible to return to state 0, the probability we are looking for,
00 00
1 p0 is the same as 1 p0 . We calculate
1 1 1
− ∫ ( µt + µt02 dt ) − ∫ ( µt )
+0.5 µt01 dt ∫ ( )
− 1.5 0.01+0.02⋅2t dt
01 01
1 p0 = 1 p0 = e =e =e =
00 00 0 0 0
t=1
⎛ 0.03 t ⎞ ⎛ 0.06 0.03 ⎞ 0.03
−⎜ 0.015t+ ⋅2 ⎟ −⎜ 0.015+ −0− ⎟ −0.015−
⎝ ln 2 ⎠ t=0
=e =e ⎝ ln 2 ln 2 ⎠
=e ln 2
≈ 0.94338496.
We could also note that
( ) (
µt01 + µt02 = 0.01+ 0.02 ⋅ 2 t + 0.5 0.01+ 0.02 ⋅ 2 t = 0.015 + 0.03⋅ 2 t , )
and this is the Makeham’s Law with A = 0.015, B = 0.03, c = 2. Recall that under
( ) , where g = e− ln c and s = e− A . Here, g = e− ln 2 ,
B 0.03
c x c n −1
Makeham’s Law, n px = s n g
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
(
2 x ⋅ 2 n −1 )
⎛ − 0.03 ⎞
s = e−0.015 , and n px = e−0.015n ⋅ ⎜ e ln 2 ⎟ , so that the quantity sought, i.e., 1 p000 = 1 p000 ,
⎝ ⎠
is the same as 1 p0 under this Makeham model, i.e.,
−0.03
(21−1)
−0.015⋅1
1 p0 = 1 p0 = e ⋅ e ln 2 ≈ 0.94338496.
00 00
Answer B.
Solution.
Recall the key formula for UDD over each year of age in the multiple decrement table:
q(x j )
p′x ( j)
( )
= px (τ ) q(xτ ) .
We have
(τ )
(τ ) l51 90, 365
p50 = (τ ) = = 0.90365,
l50 100,000
( 3)
( 3) d50 1,100
q50 = (τ ) = = 0.011,
l50 100,000
(τ ) (τ )
q50 = 1− p50 = 1− 0.90365 = 0.09635.
We calculate therefore
( )
3
( )
q50 0.011
( 3) ( 3) ( 3) (τ ) ( )
′ = q50
′ = 1− p50
′ = 1− p50 = 1− 0.90365 0.09635 ≈ 0.0115.
τ
q51 q50
′( 3) ≈ 0.9885 and
This means that p51
( )
3
( )
3 q51
( ) ⎛ 80,000 ⎞
q51 80,000
′( 3) = p51
(τ ) ( )
0.9855 ≈ p51 =⎜
τ 1−
⎝ 90, 365 ⎟⎠
q51 90,365 .
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
⎛ 80,000 ⎞
⎜ 1− ⎟ ⋅ ln 0.9855
( 3) ⎝ 90, 365 ⎠
q51 ≈ ≈ 0.010889815.
⎛ 80,000 ⎞
ln ⎜
⎝ 90, 365 ⎟⎠
This means that
( 3) ( 3) (τ )
d51 = q51 ⋅l51 ≈ 0.010889815 ⋅ 90, 365 ≈ 984.0581,
(1)
d51 (τ )
= d51 (2)
− d51 ( 3)
− d51 (τ )
= l50 ( (τ )
− l51 ) (2)
− d51 ( 3)
− d51 ≈
≈ ( 90, 365 − 80,000 ) − 8200 − 984 = 1,181,
(1)
(1) d51 1,181
q51 = (τ ) ≈ ≈ 0.01306922,
l51 90, 365
and we also have
(τ ) 80,000
p51 = ≈ 0.88529851,
90, 365
(τ ) (τ ) 80,000
q51 = 1− p51 = 1− ≈ 0.11470149.
90, 365
We therefore conclude that
()
1
( )
q51 0.01306922
(1) (τ ) ( )
′ = p51 ≈ 0.885298510.11470149 ≈ 0.98621441.
τ
p51 q51
Solution.
The present value of death benefit random variable is
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
⎧⎪ 2v K +1 , K < 20,
Z = ⎨ K +1
⎩⎪ v , K ≥ 20.
Therefore,
E ( Z ) = 2A40 − 20 E40 ⋅ A60 = 2 ⋅ 0.36987 − 0.51276 ⋅ 0.62567 ≈ 0.41892.
We conclude that the variance of Z is
Var ( Z ) = E ( Z 2 ) − ( E ( Z )) ≈ 0.24954 − 0.41892 2 ≈ 0.07405,
2
(v) Deaths are assumed to follow a constant force of mortality between integral ages.
(vi) Z is the present value random variable for this insurance.
Calculate Pr(Z > 277,000).
Solution.
Note that the effective half a year interest rate is 9%. Also, under the constant force of
mortality between integral ages assumption, if we write µ x for the constant force
between ages x and x + 1, and µ x+1 for the constant force between ages x +1 and x + 2,
then
−0.5 µ x
0.5 q x = 1− 0.5 p x = 1− e = 1− px0.5 = 1− 1− 0.16 ≈ 0.083485,
0.5 qx+0.5 = 1− 0.5 px+0.5 = 1− px0.5 = 1− 1− 0.16 ≈ 0.083485,
0.5 qx+1 = 1− 0.5 px+1 = 1− e−0.5 µx+1 = 1− px+1
0.5
= 1− 1− 0.23 ≈ 0.122504,
qx+1.5 = 1− 0.5 px+1.5 = 1− e−0.5 µx+1 = 1− px+1
0.5
0.5
= 1− 1− 0.23 ≈ 0.122504.
Based on the above, probabilities of the insured dying in each of the four half-years are:
q = 0.5 qx ≈ 0.083485,
0 0.5 x
0.5 0.5
qx = 0.5 px ⋅ 0.5 qx+0.5 ≈ 0.916515 ⋅ 0.083485 ≈ 0.076515,
1 0.5
qx = 1 px ⋅ 0.5 qx+1 ≈ (1− 0.16 ) ⋅ 0.122504 ≈ 0.102903,
qx = 1 px ⋅ 0.5 px+1 ⋅ 0.5 qx+1.5 ≈ (1− 0.16 ) ⋅ 0.877496 ⋅ 0.122504 ≈ 0.090297.
1 0.5
The random present value of the death benefit is
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
⎧ 300,000
⎪ ≈ 275,229.36 with probability 0.5 qx ≈ 0.083485,
⎪ 1.9
⎪ 330,000
⎪⎪ 1.9 2 ≈ 277, 754.40 with probability 0.5 0.5 qx ≈ 0.076515,
Z=⎨
⎪ 360,000 ≈ 277,986.05 with probability q ≈ 0.102903,
⎪ 1.9 3 1 0.5 x
⎪ 390,000
⎪ 4
≈ 276,285.83 with probability 1 0.5 qx ≈ 0.090297.
⎪⎩ 1.9
Also, Z = 0 otherwise. We conclude that
Pr ( Z > 277,000 ) = 0.5 0.5 qx + 1 0.5 qx ≈ 0.076515 + 0.102903 ≈ 0.179418.
Answer D.
November
You are evaluating 2012 Course
the financial strengthMLC Examination,
of companies Problem
based on No. 16 multiple
the following
You are evaluating the financial strength of companies based on the following multiple
state model:
state model:
State 0 State 1
Solvent Bankrupt
State 2
Liquidated
For each company, you assume the following constant transition intensities:
For each company,
(i) µyou
01
= assume
0.02. the following constant transition intensities:
µ 01 µ = 0.06.
10
(i) (ii) = 0.02
(iii) µ = 0.10.
12
(ii) µ 10 = 0.06
Using
1
Kolmogorov’s forward equations with step h = , calculate the probability that a
2
(iii) company
µ = 0.10
12 currently Bankrupt will be Solvent at the end of one year.
(D) 0.057
λ10 ( t ) = λ10 = µ10 = 0.06,
λ12 ( t ) = λ12 = µ12 = 0.10,
λ1 ( t ) = λ1 = λ10 + λ12 = 0.06 + 0.10 = 0.16,
λ20 ( t ) = λ20 = µ 20 = 0.00,
λ21 ( t ) = λ21 = µ 21 = 0.00,
λ2 ( t ) = λ2 = λ20 + λ21 = 0 + 0 = 0.
This tells us that the generator matrix is:
⎡ − λ (t + t ) λ (t + r ) λ (t + r ) ⎤
⎢ 0 01 02
⎥ ⎡ −0.02 0.02 0 ⎤
Qt+r ⎢ λ10 ( t + r ) − λ1 ( t + t ) λ12 ( t + r ) ⎥ = 0.06 −0.16 0.1 ⎥ . ⎢
⎢ ⎥ ⎢ 0 ⎥
λ
⎢⎣ 20 ( t + r ) λ 21 ( t + r ) − λ 2 ( t + t ) ⎥⎦ ⎢
⎣ 0 0 ⎥⎦
For
⎡ p(t ) p(t ) p(t ) ⎤
⎢ r 00 r 01 r 02 ⎥
r Pt =
⎢ r p10 (t )
r p11
(t ) (t ) ⎥ ,
r p12
⎢ ⎥
(t ) (t ) (t )
⎢ r p20 r p21 r p22
⎥
⎣ ⎦
the resulting Kolmogorov Forward Equation, in matrix form, is
⎡ d (t ) d (t ) d (t )
⎤
⎡ p(t ) (t ) (t ) ⎤
⎢ r p00 r p01 r p02 ⎥
dr dr dr
d d ⎢⎢
r 00 r p01 r p02
⎥ ⎢⎢ ⎥
⎥
(t ) (t ) (t ) ⎥ = d (t ) d (t ) d (t )
r Pt = r p10 r p11 r p12 ⎢ r p10 r p11 r p12 ⎥=
dr dr ⎢ ⎥ ⎢ dr dr dr
⎢ r p20 (t ) (t ) (t )
⎥ ⎥
p p
⎣ r 21 r 22
⎦ ⎢ d p(t ) d p(t ) d p(t ) ⎥
⎢ dr r 20 dr r 21 dr r 22 ⎥
⎣ ⎦
⎡ (t )
p00 (t )
p01 (t ) ⎤
p02
⎢ r r r
⎥ ⎡ −0.02 0.02 0 ⎤
=⎢ (t ) (t ) (t ) ⎥ ⋅ ⎢ ⎥
p10 p11 r p12 ⎢ 0.06 −0.16 0.1 ⎥ =
⎢ r r
⎥
⎢ (t )
p20 (t )
p21 p (t )
⎥ ⎢⎣ 0 0 0 ⎥⎦
⎣ r r r 22
⎦
⎡ −0.02 p(t ) + 0.06 p(t ) 0.02 p(t ) − 0.16 p(t ) 0.1 p(t ) ⎤
⎢ r 00 r 01 r 00 r 01 r 01
⎥
= ⎢ −0.02 r p10 + 0.06 r p11 0.02 r p10 − 0.16 r p11 0.1 r p11
( t ) ( t ) ( t ) ( t ) (t ) ⎥ .
⎢ ⎥
(t ) (t ) (t ) (t ) (t )
⎢ −0.02 r p20 + 0.06 r p21 0.02 r p20 − 0.16 r p21 0.1 r p21 ⎥
⎣ ⎦
(0)
We are looking for an approximate value of the probability 1 p10 . We can actually remove
the superscript related to the starting point in time, because due to constant intensities of
(0)
transition, the starting time does not affect probabilities, so 1 p10 = 1 p10 , and similarly we
will drop the superscript for other probabilities. The equation
d (t ) (t ) (t )
r p10 = −0.02 r p10 + 0.06 r p11
dr
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
gives us this approximation
1 p10 − 0.5 p10
≈ −0.02 0.5 p10 + 0.06 0.5 p11 ,
0.5
or
1 p10 ≈ 0.99 0.5 p10 + 0.03 0.5 p11 .
Additionally
d (t ) (t ) (t ) d (t ) (t ) (t )
r p10 = −0.02 r p10 + 0.06 r p11 r p11 = 0.02 r p10 − 0.16 r p11
dr dr
give us approximations
0.5 p10 − 0 p10 0.5 p11 − 0 p11
≈ −0.02 0 p10 + 0.06 0 p11 , ≈ 0.02 0 p10 − 0.16 0 p11 ,
0.5 0.5
or
0.5 p10 ≈ 0.99 0 p10 + 0.03 0 p11 , 0.5 p11 ≈ 0.01 0 p10 + 0.92 0 p11 .
But 0 p10 = 0, 0 p11 = 1, so that
0.5 p10 ≈ 0.99 0 p10 + 0.03 0 p11 = 0.99 ⋅ 0 + 0.03⋅1 = 0.03,
Calculate 10.25V.
Solution.
We can use the recursive reserve formula twice, in quarterly steps. Note that the effective
quarterly interest rate is 2%, and that when going from policy duration 10.25 to 10.75, we
only have a premium payment at policy duration 10.50. Thus
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
( 10.25V + 0 ) ⋅1.02 = 0.25 q90.25 ⋅1000 + 0.25 p90.25 ⋅ 10.50V,
( V + 60 (1− 0.10 )) ⋅1.02 = q ⋅1000 + p
10.50 0.25 90.50 0.25 90.50 ⋅ 10.75V.
Now we substitute the following known data:
10.75V = 753.72,
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
Solution.
We begin by finding Π. The actuarial present value of benefits at issue is
1000A40 + 4 ⋅ 11 E40 ⋅1000A51 = 161.32 + 4 ⋅ 0.50330 ⋅ 259.61 ≈ 683.9669.
The actuarial present value of expenses at issue is
100 + 10a40 = 100 + 10 ( a40 − 1) = 100 + 10 ⋅ (14.8166 − 1) = 238.17.
All the values used in calculations above came from the Illustrative Life Table. The
premium is paid as a life annuity due on (40), hence
683.97 + 238.17 683.97 + 238.17
Π≈ = ≈ 62.2365.
a40 14.8166
Therefore, G = 1.02Π ≈ 63.4812. If we calculated the reserve retrospectively, we would
obtain
G − 1000q40 ⋅1.06 −1 − 100 63.4812 − 2.78 ⋅1.06 −1 − 100
≈ ≈ −41.6056.
1 E 40 (1− 0.00278 ) ⋅1.06 −1
But prospectively, we obtain
(
1000A41 + 4 ⋅ 10 E41 ⋅1000A51 ) + 10
a41 − 63.4812
a41 =
Actuarial present value of future benefits Actuarial present Actuarial present
value of future value of future
expenses gross premiums
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
AAV (Past Gross Premiums up to time k) +
+ APV (Future Gross Premiums from time k on) =
= AAV (Past Benefits and Expenses up to time k) +
+ APV (Future Benefits and Expenses from time k on) +
+ Actuarial Value (Past and Future Profits).
Therefore, at policy duration k,
Retrospective Reserve =
AAV (Past Gross Premiums) – AAV (Past Benefits and Expenses) =
= APV (Future Benefits and Expenses) – APV (Future Gross Premiums) +
+ Actuarial Value (Past and Future Profits) =
= Prospective Reserve + Actuarial Value (Past and Future Profits).
If you recall that the purpose of the reserve is to provide funds for benefits and expenses
payment that are not paid by remaining premiums, you now should be able to clearly see
that the prospective reserve calculation is the correct one, and if you do the calculation
retrospectively, you must subtraction the actuarial value of all profits at the moment of
calculation from the retrospective calculation result. In this problem, the reserve at policy
duration 1 is:
–61.20 = –41.60 – 19.60.
Interestingly, –40 was one of the answer choices, nice little trap waiting for the
candidates taking the exam. The correct answer choice is –60, the one closest to –61.20.
Answer B.
Solution.
Let X1 , X2 , …, X 200 , be the random present values of life annuities-due considered.
These are independent identically distributed random variables, and
S = X1 + X2 + ...+X200 .
This implies that S is approximately normal and
E ( S ) = E ( X1 + X2 + ...+X200 ) = 200E ( X1 )
as well as
Var ( S ) = Var ( X1 + X2 + ...+X200 ) = 200Var ( X1 ) .
⎛ − ⎞
1
Recall that d (12 ) = 12 ⎜ 1− (1+ i ) 12 ⎟ . We have
⎝ ⎠
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
(12 )
1− A62 1− 0.4075
E ( X1 ) = 15 ⋅12 ⋅ a62
(12 )
= 15 ⋅12 ⋅ = 15 ⋅12 ⋅ ≈ 1834.7545,
d (12 ) ⎛ − ⎞
1
12 ⎜ 1− 1.06 12 ⎟
⎝ ⎠
and
( )
2
(12 ) (12 )
2
A62 − A62 0.2105 − 0.4075 2
Var ( X1 ) = (15 ⋅12 ) ⋅
2
= 180 ⋅
2
≈ 426,176.9086.
(d( ) )
2 2
12
⎛ ⎛ − ⎞⎞
1
⎜ 12 ⎜⎝ 1− 1.06 ⎟⎠ ⎟
12
⎝ ⎠
Based on this,
E ( S ) = 200E ( X1 ) ≈ 200 ⋅1834.7545 ≈ 366,936.12,
and
Var ( S ) = 200Var ( X1 ) ≈ 200 ⋅ 426,176.9086 ≈ 85,235, 381.71.
We have
⎛ 200Π − E ( S ) S − E ( S ) ⎞
0.90 = Pr ( 200Π > S ) = Pr ⎜ > ⎟.
⎝ Var ( S ) Var ( S ) ⎠
S − E (S ) 200Π − E ( S )
Since is approximately standard normal, must be the 90-th
Var ( S ) Var ( S )
percentile of the standard normal distribution, i.e., 1.282. This gives the equation
200Π − E ( S ) 200Π − 366,936.12
1.282 = = ,
Var ( S ) 85,235, 381.71
resulting in
366,936.12 + 1.282 ⋅ 85,235, 381.71
Π= ≈ 1893.8597.
200
Answer E.
Solution.
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
Let us write P for the annual premium that Stuart has paid. Since it was calculated using
the equivalence principle, we have
E ⋅ a
P = 20 45 65 .
a45:20
Let us write K 65 for the curtate future lifetime of (65) (we could also write K(65), but we
need to get used to the new notation, so this is a good time to practice). The probability
we are looking for is (note that a65 = 9.8969 from the Illustrative Life Table)
⎛ P
a ⎞ ⎛ E45 ⋅ a65 a45:20 ⎞
Pr ⎜ 45:20 > aK +1 ⎟ = Pr ⎜ 20
⋅ > aK +1 ⎟ =
⎝ 20 E45 65
⎠ ⎝ a45:20 20 E 45
65
⎠
(
= Pr a65 > aK
65 +1
) = Pr (9.8969 > a ).K 65 +1
If we use BA II Plus Pro, set BGN mode by pushing 2ND, BGN if BGN is not displayed,
and enter 0 FV, –1 PMT, 9.8969 PV, 6 I/Y, CPT N, we obtain 14.0973895. This means
that the probability sought is
Pr ( K 65 + 1 < 14.0973895 ) = Pr ( K 65 + 1 ≤ 14 ) = 1− 14 p65 =
l79 4225163
= 1−
= 1− ≈ 0.43917133.
l65 7533964
For a moment, when you see the formula for P you might worry about working with
mortality from before age 65, when we are told that mortality follows the Illustrative Life
Table only from age 65 on, but as you can see, items related to mortality from before age
65 cancel nicely in the work involved in finding the probability sought.
Answer C.
Solution.
Note that the future lifetime of (x) is ruled by constant force of mortality of 1%, and the
future lifetime of (y) is rules by constant force of mortality of 2%. Let us write P for the
annual benefit premium rate sought. We have
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
100Ax:y
P= .
ax:y
Also,
+∞
⎛1 3 ⎞
∫e
−0.05t
ax:y = ⋅ ⎜ e−0.01t + e−0.03t ⎟ dt =
0
⎝4 4 ⎠
+∞ +∞
1 3 1 1 3 1
∫ e dt + ∫e
−0.06t −0.08t
= dt = ⋅ + ⋅ ≈ 13.5416667,
4 0
4 0
4 0.06 4 0.08
Ax:y =
Poohsticks
Ax + Ay − Ax:y = Ax + Ay − 1− δ ax:y = ( )
0.01 0.02
= + − (1− 0.05 ⋅13.5416667 ) ≈ 0.12946429.
0.01+ 0.05 0.02 + 0.05
Therefore,
100Ax:y 100 ⋅ 0.12946429
P= ≈ ≈ 0.95604396.
ax:y 13.5416667
Answer A.
A. 82 B. 86 C. 90 D. 94 E. 98
Solution.
Let us write P for the level annual premium sought. Based on the equivalence principle,
we write
+ P ⋅ ( IA )80:2 ,
1
a80:2 = 1000A80:2
P 1
so that
1
1000A80:2
P= .
a80:2 − ( IA )80:2
1
We calculate
1− q80 1− 0.0803
a80:2 = 1+ vp80 = 1+ = 1+ ≈ 1.90388206,
1.0175 1.0175
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
⎛ q p q ⎞
1
1000A80:2 = 1000 ⎜ 80 + 80 81 2 ⎟ =
⎝ 1.0175 1.0175 ⎠
⎛ 0.0803 0.9197 ⋅ 0.08764 ⎞
= 1000 ⎜ + ⎟⎠ ≈ 156.772702,
⎝ 1.0175 1.0175 2
and
q80 2 p80 q81 0.0803 2 ⋅ 0.9197 ⋅ 0.08764
( IA )180:2 = + = + ≈ 0.23462648.
1.0175 1.0175 2 1.0175 1.0175 2
Therefore,
1
1000A80:2 156.772702
P= ≈ ≈ 93.9177342.
a80:2 − ( IA )80:2 1.90388206 − 0.23462648
1
Answer D.
Solution.
The mortality gain equals
( Expected deaths − Actual deaths ) ⋅ Net Amount at Risk.
In this case, the net amount at risk is 1000 – 3V = 1000 – 12.18 = 987.82. Expected
deaths are q47 from the Illustrative Life Table, i.e., 0.00466, times the number of alive
policyholders at the beginning of 2011, which is 10,000 – 30 = 9970, which gives
0.00466 ⋅ 9970 ≈ 46.4602.
The actual number of deaths in 2011 is 18. Hence, the mortality gain equals
( 46.4602 − 18 ) ⋅ 987.82 ≈ 28,113.5548.
Answer A.
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.
(H) Healthy
(Z) infected with virus “Zebra”
(L) infected with virus “Lion”
(D) Death
The annual transition probability matrix is given by:
H Z L D
H 0.90 0.05 0.04 0.01
Z 0.10 0.20 0.00 0.70
L 0.20 0.00 0.20 0.60
D 0.00 0.00 0.00 1.00
You are given:
(i) Transitions occur only once per year.
(ii) 250 is payable at the end of the year in which you become infected with either virus.
(iii) For lives infected with either virus, 1000 is payable at the end of the year of death.
(iv) The policy is issued only on healthy lives.
(v) i = 0.05.
Calculate the actuarial present value of the benefits at policy issue.
A. 66 B. 75 C. 84 D. 93 E. 102
Solution.
We list all possible cases of payouts, with their probabilities and actuarial present values
(calculated as products of probabilities and discounted benefits)
Possible Probability Discounted Actuarial
transition benefit present
value
250
H à Z 0.05 11.9047619
1.05
250
H à L 0.04 9.52380952
1.05
1000
H à Z à D 0.05 ⋅ 0.7 31.7460317
1.05 2
1000
H à L à D 0.04 ⋅ 0.6 21.7687075
1.05 2
250
H à H à Z 0.9 ⋅ 0.05 10.2040816
1.05 2
250
H à H à Z 0.9 ⋅ 0.04 8.16326531
1.05 2
The total actuarial present value of the benefits at policy issue is the sum of all numbers
in the last column, which is approximately 93.3106576.
Answer D.
Solution.
Let us write G for the gross premium sought. Using the equivalence principle, we write:
+ G ⋅ ( IA )x:10 + 0.45G + 0.05G
1
ax:10 = 100,000Ax:10
G 1
ax:10 + 200
ax:10 .
APV of death benefit APV of return of % of premium expenses Per policy expenses
premium benefit
Copyright © 2012 by Krzysztof Ostaszewski. All rights reserved. No reproduction in any form is permitted without
explicit permission of the copyright owner.