MAF Johnson ExamMLC3L StudySupplement J2014

SOA EXAM MLC & CAS EXAM LC STUDY
SUPPLEMENT
by Paul H. Johnson, Jr., PhD.
Last Modified: January 2014
A document prepared by the author as study materials for the

Midwestern Actuarial Forum’s Exam Preparation Seminars
This document is not to be redistributed or reproduced in any form

without permission from the author.
This document is not in any way officially affiliated with

the Society of Actuaries or the Casualty Actuarial Society.
i
Contents
PREFACE iv
AUTHOR BACKGROUND vi
1 SURVIVAL MODELS AND LIFE TABLES 1

1.1 Key Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.3 Past Exam Questions . . . . . . . . . . . . . . . . . . . . . 27
2 SELECTION (Exam MLC Only) 28

2.1 Key Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3 INSURANCE BENEFITS I 36
3.1 Key Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4 INSURANCE BENEFITS II (Exam MLC Only) 67

4.1 Key Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . 67
4.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
5 ANNUITIES I 81
5.1 Key Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . 81
5.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
6 ANNUITIES II (Exam MLC Only) 104

6.1 Key Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . 104
6.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
ii
7 PREMIUM CALCULATION I 123

7.1 Key Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . 123
7.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
8 PREMIUM CALCULATION II (Exam MLC Only)141

8.1 Key Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . 141
8.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
9 RESERVES I 158
9.1 Key Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . 158
9.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170
10 RESERVES II (Exam MLC Only) 180

10.1 Key Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . 180
10.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
11 MULTIPLE STATE MODELS (Exam MLC Only)208

11.1 Key Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . 208
11.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
12 MULTIPLE DECREMENTS I 227

12.1 Key Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . 227
12.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
13 MULTIPLE DECREMENTS II (Exam MLC Only)241

13.1 Key Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . 241
13.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250
iii
14 MULTIPLE LIVES I 261

14.1 Key Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . 261
14.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271
15 MULTIPLE LIVES II (Exam MLC) 283

15.1 Key Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . 283
15.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
16 OTHER TOPICS (Exam MLC Only) 296

16.1 Key Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . 296
16.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
17 REFERENCES 320
18 APPENDIX 321
18.1 APPENDIX A: EXAM SYLLABI, EXAM TABLES,
AND PAST EXAM QUESTIONS . . . . . . . . . . . . . 321
18.2 APPENDIX B: MULTI-STATE MODEL EXAMPLES
(Exam MLC Only) . . . . . . . . . . . . . . . . . . . . . . . 321
iv
PREFACE
This study supplement is intended to be an aid to candidates studying for

either Exam MLC (Models for Life Contingencies) of the Society of Actuar-
ies or Exam LC (Models for Life Contingencies) of the Casualty Actuarial
Society. Each section of this study supplement contains key life contingencies
concepts and major life contingencies formulas with brief interpretations for
specific Exam MLC/LC topics. This study supplement is NOT meant to
take the place of an exam’s required text book or a proper study guide. For
example, there are no detailed explanations of topics that would be found in
a text book and no worked examples of the material that would be found in
a study guide.
This study supplement is meant to outline key life contingencies concepts

and major life contingencies formulas associated with Exam MLC or Exam
LC, and to provide exercises that are meant to test a candidate’s basic knowl-
edge of the material and to improve the candidate’s speed in working with
certain concepts and formulas. If the candidate can master the exercises in
this study supplement, the candidate should be in a good position to tackle
actual past Exam MLC and/or Exam 3L questions. Many of the exercises in
this study supplement are based on past homework and midterm questions
that I used in teaching life contingencies classes at the University of Illinois
at Urbana-Champaign. In some cases, the exercise answer choices are
rounded. Choose the letter of the answer choice that is closest to
your own answer. At present, this supplement does not contain the type
of written answer problems that will be part of Exam MLC starting in 2014.
Exam MLC candidates should refer to Appendix A of this supplement, for a
link to the SOA’s sample written exercises exam.
Once the exercises in this study supplement have been mastered, the can-
didate can refer to the listing of past Exam MLC and Exam 3L questions
at the end of each section for a more advanced test of the candidate’s un-
derstanding of the material. Web links to past Exam MLC and Exam 3L
questions can be found in Appendix A of this study supplement. Also, some
Exam MLC exercises in this study supplement will utilize “Multi-State Model
Examples”from Dickson et al; these figures are provided in Appendix B of
this study supplement.
v
For candidates enrolled in the Midwestern Actuarial Forum’s Exam MLC

& LC Seminar, it is expected that each candidate will have read the required
study materials for either Exam MLC or Exam LC mentioned in the exam’s
syllabus. The candidate is also expected to have read the relevant sections
of this study supplement and also have worked its relevant exercises PRIOR
to sitting for the seminar. Furthermore, it is expected that each candidate
will have worked as many of the past Exam MLC and Exam LC questions
as possible. Exam LC candidates should attempt life contingencies problems
from the prior 3L exams first before attempting any problems from prior MLC
exams. Also, some sections are marked “Exam MLC Only;” as you may have
guessed, candidates preparing for Exam LC can skip those sections. I will
rely heavily on candidate’s asking questions regarding these study supplement
exercises and past Exam MLC and Exam 3L questions DURING the seminar;
often, working exercises and past exam questions identifies issues that once
resolved will enhance the candidate’s understanding of the material. I would
also contend that candidates will only get the most out of this seminar if they
have sufficiently studied for their exam beforehand.
Candidates should bring a printed or electronic copy of this study

supplement with them to the seminar. Candidates should also print
copies of all tables provided during the exam they will be writing
(Exam MLC or Exam LC) - please refer to Appendix A of this study
supplement. Candidates should also bring copies of any past Exam MLC
and/or Exam LC questions that they wish to discuss during the seminar.
Despite proofreading, it is possible that this study supplement may contain

typos. If there are any questions regarding possible typos, or anything else,
please e-mail me at: pjohnson@illinois.edu.
Best of luck in your exam preparation!
Regards:
-Paul H. Johnson, Jr., PhD

vi
AUTHOR BACKGROUND
Paul H. Johnson, Jr. earned a PhD in Actuarial Science, Risk Management

and Insurance from the University of Wisconsin-Madison in 2008. He is
currently an Assistant Professor of Actuarial Science in the Department of
Mathematics at the University of Illinois at Urbana-Champaign. Paul has
passed all of the preliminary examinations of the Society of Actuaries (the
modern day equivalents of Exams P, FM, M, and C). Since 2003, Paul has
taught undergraduate courses meant to prepare students for each the pre-
liminary exams of the Society of Actuaries, and has received various honors
and awards for teaching excellence. Most recently, Paul was honored with
the N. Tenney Peck Teaching Award in Mathematics for 2011. He has also
published in actuarial and risk management journals, including the North
American Actuarial Journal and the Risk Management and Insurance Re-
view.
1
1 SURVIVAL MODELS AND LIFE TABLES
1.1 Key Concepts

Let (x) denote a life aged x.
Future Lifetime:
• Tx = time-until-death for (x), a continuous random variable (in years).

Tx is also called the future lifetime random variable. Tx may also be
written as T (x) or T .
Special case: T0 = age-at-death for (0), where (0) denotes a newborn life.
Note: T0 = x + Tx .
• Fx (t) = t qx = P r(Tx ≤ t)
This is the cumulative distribution function of Tx , “the probability that
(x) dies within t years.” The q-notation will be used most of the time.
F0 (t) can also be written more simply as F (t).
• Sx (t) = t px = P r(Tx > t)

This is the survival function function of Tx ,“the probability that (x) sur-
vives for at least t years.” The words “at least” are often omitted. The
p-notation will be used most of the time.
S0 (t) can also be written more simply as S(t) or s(t).
• From above: t qx + t px = 1.
“(x) will either survive or die within t years.”
• S0 (x + t) = S0 (x)t px
“The probability that (0) survives x + t years is equivalent to (0) first
surviving x years to age x, and then surviving t additional years to age
x + t.”
Paul H. Johnson, Jr.

1.1 Key Concepts 2
• u+t px = (u px )(t px+u )

“The probability that (x) survives u + t years is equivalent to (x) first
surviving u years to age x + u, and then surviving t additional years to
age x + u + t.”
• Be careful: u+t qx 6= (u qx )(t qx+u ).

“The right-hand side implies that it is possible for (x) to die within u
years, then somehow come back to life at age x + u in order to die again
within t years. This, of course, is not possible and cannot be equal to
the left-hand side which is the probability (x) dies within u + t years.”
• u|t qx = u+t qx - u qx = u px - u+t px = (u px )(t qx+u )

This is a u-year deferred probability of death, “the probability that (x)
dies between ages x + u and x + u + t.” Note: 0|t qx = t qx .
• Note: 1 qx , 1 px , and u|1 qx are written as qx , px , and u| qx , respectively.

qx may be referred to as a mortality rate at age x, and px may be referred
to as a survival rate at age x.

1.1 Key Concepts 3
Force of Mortality:
• µx = µ(x) = force of mortality at age x, given survival to age x. This is

sometimes called the “hazard rate” or “failure rate.”
d
0 [S (x)]
d
µx = − dxS0 (x) = − dx [ln S0 (x)]
“This is the instantaneous death rate for a life at age x.”
• µx+t = µx (t) = force of mortality at age x + t, given survival to x + t.

d
[p ]
µx+t = − dtt pt xx = − dtd [ln t px ]
“This is the instantaneous death rate for a life at age x + t. Here, the
variable is time after age x. You could also obtain µx+t by replacing x in
µx with x + t.”
R x+t Rt
• t px = exp[− x µs ds] = exp[− 0 µx+s ds]
• If c > 0, then µ∗x+s = cµx+s =⇒ t p∗x = (t px )c .

For constant k, then µ∗x+s = µx+s + k =⇒ t p∗x = (e−kt )(t px ).
The constant k should be such that µ∗x+s > 0.
• fx (t) = t px µx+t = probability density function of Tx .

This comes from the above formula for µx+t , recognizing that fx (t) =
d d
dt [t qx ] = − dt [t px ].
Rt
• t qx = 0 s px µx+s ds
R∞
• t px = t s px µx+s ds
R u+t
• u|t qx = u s px µx+s ds

1.1 Key Concepts 4
Curtate Future Lifetime:
• Kx = curtate future lifetime for (x), a discrete random variable.

Kx = bTx c = integer part of Tx . That is, Kx represents the complete num-
ber of future years survived by (x), where any fractional time survived
in the year of death is ignored. Note: Kx = 0, 1, 2...
Kx may also be written as K(x) or K.
• k| qx = P r(Kx = k) = P r(k ≤ Tx < k + 1) for k = 0, 1, 2...

This is the probability mass function of Kx , “the probability that (x) dies
in the (k + 1)st year, between ages x + k and x + k + 1.”
• k+1 qx = 0| qx + 1| qx + ... + k| qx
“The probability that (x) dies within k + 1 years is the sum of the
probabilities that (x) dies in the first year, the second year, ..., the (k +
1)st year.”

1.1 Key Concepts 5
Other Features of Tx and Kx Distributions:

R∞ R∞
• e̊x = E(Tx ) = 0 t(t px )(µx+t )dt = 0 t px dt
This is the complete expectation of life for (x), the average time-until-
death for (x). That is, (x) is expected to die at age x + e̊x .
R∞
• V ar(Tx ) = E(Tx2 ) - [E(Tx )]2 = 2 0 tt px dt - [e̊x ]2
Rn
• e̊x:n = E[min(Tx , n)] = 0 t px dt
This is the n-year temporary complete life expectancy for (x), the aver-
age number of years out of the next n years that (x) survives.
This expectation helps define the recursion: e̊x = e̊x:n + n px e̊x+n .
“The average number of future years that (x) survives is the average
number of years out of the first n years that (x) survives plus the average
number of years (x) survives after the first n years (accounting for the
probability that (x) survives the first n years).”
Similarly: e̊x:m+n = e̊x:m + m px e̊x+m:n .
• The 100α-th percentile of the distribution of Tx , πα , is such that:
πα q x = α for 0 ≤ α ≤ 1.
Special case: α = 0.50; π.50 is called the median future lifetime for (x).

1.1 Key Concepts 6
P∞ P∞
• ex = E(Kx ) = k=0 k(k| qx ) = k=1 k px
This is the curtate expectation of life for (x), the average curtate future
lifetime for (x).
1
• e̊x ≈ ex + 2
P∞
• V ar(Kx ) = E(Kx2 ) - [E(Kx )]2 = k=1 (2k − 1)k px - [ex ]2
Pn
• ex:n = E[min(Kx , n)] = k=1 k px
This is the n-year temporary curtate life expectancy for (x).
This expectation helps define the recursions: ex = ex:n + n px ex+n and

ex:m+n = ex:m + m px ex+m:n .

1.1 Key Concepts 7
Special Mortality Laws:
de Moivre’s Law: Tx has a continuous uniform distribution.

The limiting age is ω such that 0 ≤ x ≤ x + t ≤ ω.
1
• µx = ω−x (Note: x 6= ω)
ω−x
• S0 (x) = ω
x
• F0 (x) = ω
ω−x−t
• t px = ω−x
t
• t qx = ω−x
t
• u|t qx = ω−x
1
• fx (t) = t px µx+t = ω−x (Note: x 6= ω)
ω−x
• e̊x = 2
n
• e̊x:n = nn px + 2 n qx
“(x) can either survive n years with probability n px , or die within n years
with probability n qx . Surviving n years contributes n to the expectation.
Dying within n years contributes n2 to the expectation as future lifetime
has a uniform distribution - (x), on average, would die halfway through
the n-year period.”
(ω−x)2
• V ar(Tx ) = 12
ω−x−1
• ex = 2
(ω−x)2 1
• V ar(Kx ) = 12 - 12

1.1 Key Concepts 8
Modified/Generalized de Moivre’s Law: Tx has a beta distribution.

The limiting age is ω such that 0 ≤ x ≤ x + t ≤ ω. Also, there is a
parameter α > 0.
α
• µx = ω−x (Note: x 6= ω)
¡ ω−x ¢α
• S0 (x) = ω
¡ ¢α
• F0 (x) = 1 - ω−x ω
¡ ω−x−t ¢α
• t px = ω−x
¡ ¢α
• t qx = 1 - ω−x−t
ω−x
¡ t ¢α
Note: t qx 6= ω−x .
ω−x
• e̊x = α+1
α(ω−x)2
• V ar(Tx ) = (α+1)2 (α+2)
Note: α = 1 results in uniform distribution/de Moivre’s Law.

1.1 Key Concepts 9
Constant Force of Mortality: Tx has an exponential distribution, x ≥

0. There is another parameter that denotes the force of mortality: µ > 0.
• µx = µ
• S0 (x) = e−µx
• F0 (x) = 1 - e−µx
• t px = e−µt = (px )t
• t qx = 1 - e−µt
1
• e̊x = µ
1−e−µn
• e̊x:n = µ
1
• V ar(Tx ) = µ2
px
• ex = qx
px
• V ar(Kx ) = (qx )2
Note1: A constant force of mortality implies that “age does not matter.”
This can easily be seen from t px = e−µt ; x does not appear on the right-
hand side.
Note2: Tx has an exponential distribution implies that Kx has a geometric

distribution.

1.1 Key Concepts 10
Gompertz’s Law:
• µx = Bcx for x ≥ 0, B > 0, c > 1
• S0 (x) = exp[− lnBc (cx − 1)]
• F0 (x) = 1 - exp[− lnBc (cx − 1)]
• t px = exp[− lnBc cx (ct − 1)]
• t qx = 1 - exp[− lnBc cx (ct − 1)]
Note: c = 1 results in a constant force of mortality.
Makeham’s Law:
• µx = A + Bcx for x ≥ 0, A ≥ -B, B > 0, c > 1

B x
• S0 (x) = exp[−Ax − ln c (c − 1)]
B x
• F0 (x) = 1 - exp[−Ax − ln c (c − 1)]
B x t
• t px = exp[−At − ln c c (c − 1)]
B x t
• t qx = 1 - exp[−At − ln c c (c − 1)]
Note1: A = 0 results in Gompertz’s Law.

Note2: c = 1 results in a constant force of mortality.
Weibull’s Law: Tx has a Weibull distribution.
• µx = kxn for x ≥ 0, k > 0, n > 0

k
• S0 (x) = exp[− n+1 xn+1 ]
k
• F0 (x) = 1 - exp[− n+1 xn+1 ]
k
• t px = exp[− n+1 ((x + t)n+1 − xn+1 )]
k
• t qx = 1 - exp[− n+1 ((x + t)n+1 − xn+1 )]

1.1 Key Concepts 11
Life Tables:
Given a survival model with survival probabilities t px , one can construct a

life table, also called a mortality table, from some initial age x0 (usually age
0) to a maximum age ω (a limiting age).
• Let lx0 , the radix of the life table, represent the number of lives age x0 .
lx0 is an arbitrary positive number.
• lω = 0.
• lx+t = (lx )(t px ) for x0 ≤ x ≤ x + t ≤ ω.

lx+t represents the expected number of survivors to age x + t out of lx
individuals aged x.
• t dx = lx - lx+t = (lx )(t qx ) for x0 ≤ x ≤ x + t ≤ ω.

represents the expected number of deaths between ages x and x + t
t dx
out of lx lives aged x.
Note 1: 1 dx is written as dx .
Note 2: If n = 1, 2..., then n dx = dx + dx+1 + ... + dx+n−1 .
• t dx+u = lx+u - lx+u+t = (lx )(u|t qx ).
The Illustrative Life Table is the life table that is provided to the candidate
taking Exam MLC or Exam LC. Some questions from either exam will involve
Illustrative Life Table calculations. A web link to this table (and ALL exam
tables) is provided for each exam in Appendix A of this study supplement.

1.1 Key Concepts 12
Fractional Age Assumptions:
Life Tables are usually defined for integer ages x and integer times t. For
a quantity that involves fractional ages and/or fractional times, one has to
make an assumption about the survival distribution between integer ages;
that is, one has to interpolate the value of the quantity within each year of
age. Two common interpolation assumptions follow.
Uniform Distribution of Deaths (UDD):
One linearly interpolates within each year of age. For integer age x and 0
≤ s ≤ s + t ≤ 1:
• lx+s = lx - sdx = (1 − s)lx + (s)lx+1 . This is a linear function of s.
• s qx = sqx
• s px = 1 - sqx
qx
• µx+s = 1−sqx (does not hold at s = 1)
• fx (s) = s px µx+s = qx (does not hold at s = 1)

sqx
• s qx+t = 1−tqx
1
• e̊x = ex + 2
1
• V ar(Tx ) = V ar(Kx ) + 12
• Note: uniform distribution/de Moivre’s Law has the property of UDD

across all ages up to the limiting age ω.
Furthermore, uniform distribution/de Moivre’s Law may be expressed as
lx = k(ω − x) for 0 ≤ x ≤ ω where k > 0.

1.1 Key Concepts 13
Constant Force of Mortality (Exam MLC Only):

One exponentially interpolates within each year of age. For integer age x
and 0 ≤ s ≤ s + t ≤ 1:
• lx+s = lx psx =⇒ ln[lx+s ] = (1−s) ln[lx ] + s ln[lx+1 ]. This is an exponential

function of s.
• s px = psx
• s qx = 1 - psx
• µx+s = -ln px = µx (does not hold at s = 1)
• fx (s) = s px µx+s = -ln px (psx ) (does not hold at s = 1)
• s qx+t = 1 - psx

1.2 Exercises 14
1.2 Exercises
1.1. Suppose: F0 (t) = 1 - (1 + 0.00026t2 )−1 for t ≥ 0.
Calculate the probability that (30) dies between ages 35 and 40.
(A) 0.056 (B) 0.058 (C) 0.060 (D) 0.062 (E) 0.064
10,000−x2
1.2. You are given: s(x) = 10,000 for 0 ≤ x ≤ 100.
Calculate: q49 .
(A) 0.009 (B) 0.011 (C) 0.013 (D) 0.015 (E) 0.017
1.3. Consider a population of newborns (lives aged 0). Each newborn has
mortality such that:
x2 2x
S0 (x) = ω2 - ω + 1 for 0 ≤ x ≤ ω.
It is assumed that ω varies among newborns, and is a random variable
with a uniform distribution between ages 95 and 105.
Calculate the probability that a random newborn survives to age 18.
(A) 0.66 (B) 0.67 (C) 0.68 (D) 0.69 (E) 0.70
2
t
1.4. Suppose: S0 (t) = exp[− 2500 ] for t ≥ 0.
Calculate the force of mortality at age 45.
(A) 0.036 (B) 0.039 (C) 0.042 (D) 0.045 (E) 0.048
1.5. The probability density function of the future lifetime of a brand new
3
machine is: f (x) = 4x
27c for 0 ≤ x ≤ c.
Calculate: µ(1.1).
(A) 0.06 (B) 0.07 (C) 0.08 (D) 0.09 (E) 0.10

1.2 Exercises 15
1.6. You are given:
(i) The probability that (30) will die within 30 years is 0.10.
(ii) The probability that (40) will survive to at least age 45 and that another
(45) will die by age 60 is 0.077638.
(iii) The probability that two lives age 30 will both die within 10 years is
0.000096.
(iv) All lives are independent and have the same expected mortality.
Calculate the probability that (45) will survive 15 years.
(A) 0.90 (B) 0.91 (C) 0.92 (D) 0.93 (E) 0.94
1.7. You are given:
(i) e50 = 20 and e52 = 19.33

(ii) q51 = 0.035
Calculate: q50 .
(A) 0.028 (B) 0.030 (C) 0.032 (D) 0.034 (E) 0.036
1.8. You are given:

³ ´
1+0.005(1.1)x −0.2098
S0 (x) = 1.005 for x > 0.
Calculate the force of mortality at age 30.
(A) 0.0012 (B) 0.0016 (C) 0.0020 (D) 0.0024 (E) 0.0028

1.2 Exercises 16
1.9. For a population of smokers and non-smokers:
(i) Non-smokers have a force of mortality that is equal to one-half the force
of mortality for smokers at each age.
(ii) For non-smokers, mortality follows a uniform distribution with ω = 90.
Calculate the difference between the probability that a 55 year old smoker
dies within 10 years and the probability that a 55 year old non-smoker dies
within 10 years.
(A) 0.20 (B) 0.22 (C) 0.24 (D) 0.26 (E) 0.28
1.10. You are given:
(i) The standard probability that (40) will die prior to age 41 is 0.01.
(ii) (40) is now subject to an extra risk during the year between ages 40
and 41.
(iii) To account for the extra risk, a revised force of mortality is defined
for the year between ages 40 and 41.
(iv) The revised force of mortality is equal to the standard force of mortality
plus a term that decreases linearly from 0.05 at age 40 to 0 at age 41.
Calculate the revised probability that (40) will die prior to age 41.
(A) 0.030 (B) 0.032 (C) 0.034 (D) 0.036 (E) 0.038

1.2 Exercises 17
1.11. An actuary assumes that Jed, aged 40, has the force of mortality:
x2
µx = c3 −x3 for 0 ≤ x < c.
Using µx , the actuary calculates the probability that Jed dies within 20
years as 0.06844. However, µx is only appropriate for a life with standard
mortality. Jed is actually a substandard life with force of mortality:
3x2
µ∗x = 3µx = c3 −x3 for 0 ≤ x < c.
Using µ∗x , calculate the correct value of the probability that Jed dies within
20 years.
(A) 0.16 (B) 0.17 (C) 0.18 (D) 0.19 (E) 0.20
(i) Tx denotes the time-until-death random variable for (x).

(ii) Mortality follows de Moivre’s Law with limiting age ω.
(iii) The variance of T25 is equal to 352.0833.
Calculate: e̊40:10 .
(A) 7.5 (B) 8.0 (C) 8.5 (D) 9.0 (E) 9.5
1.13. You are given: µx = √ 1 for 0 ≤ x < 80.

80−x
Calculate the median future lifetime for (40).
(A) 4.0 (B) 4.3 (C) 4.6 (D) 4.9 (E) 5.2
½
0.04 for 0 ≤ x < 40
µx =
0.05 for 40 ≤ x
Calculate: e̊25 .
(A) 22 (B) 23 (C) 24 (D) 25 (E) 26

1.2 Exercises 18
1.15. You are given: k| q0 = 0.10 for k = 0, 1, ..., 9.

Calculate: 5 p2 .
(A) 0.275 (B) 0.325 (C) 0.375 (D) 0.425 (E) 0.475
1.16. For the current model of Zingbot:
ω−x
(i) s(x) = ω for 0 ≤ x ≤ ω
(ii) var[T (5)] = 102.083333.
For the proposed model of Zingbot, with the same ω as the current model:
(1) s∗ (x) = ( ω−x α

ω ) for 0 ≤ x ≤ ω, α > 0
(2) µ∗10 = 0.0166667.
Calculate the difference between the complete expectation of life for a

brand new proposed model of Zingbot and the complete expectation of life
for a brand new current model of Zingbot.
(A) 5.9 (B) 6.1 (C) 6.3 (D) 6.5 (E) 6.7
1.17. Mortality for Frodo, age 33, is usually such that:

¡ 110−x−t ¢2
p
t x = 110−x for 0 ≤ t ≤ 110 − x.
However, Frodo has decided to embark on a dangerous quest that will

last for the next three years (starting today). During these three years only,
Frodo’s mortality will be revised so that he will have a constant force of
mortality of 0.2 for each year. After the quest, Frodo’s mortality will once
again follow the above expression for t px .
Calculate Frodo’s revised complete expectation of life.
(A) 15.2 (B) 15.4 (C) 15.6 (D) 15.8 (E) 16.0

1.2 Exercises 19

 0.020 for 20 ≤ x < 30
µx = 0.025 for 30 ≤ x < 42

0.030 for 42 ≤ x < 60
Calculate the probability that (26) dies in the 19th year.
(A) 0.015 (B) 0.017 (C) 0.019 (D) 0.021 (E) 0.023
1.19. An actuary has developed a survival model for a widget, denoted by

A, such that:
(10−x)2
S0A (x) = 100 for 0 ≤ x ≤ 10.
The actuary’s supervisor notes that the above survival model is incorrect.
The correct survival model for a widget, denoted by B, is such that:

(10−x)2
 100 for 0 ≤ x < 5
S0B (x) =

e−0.2 ln(4)x for x ≥ 5
Calculate: e̊B A
2 - e̊2 .
(A) 0.70 (B) 0.72 (C) 0.74 (D) 0.76 (E) 0.78
1.20. For a group of lives aged 40, consisting of 30% smokers and 70%
non-smokers, you are given:
(i) For non-smokers, µN (x) = 0.05 for x ≥ 40.

(ii) For smokers, µS (x) = 0.10 for x ≥ 40.
Calculate the 90th percentile of the distribution of the future lifetime of a

randomly selected member from this population.
(A) 40 (B) 42 (C) 44 (D) 46 (E) 48

1.2 Exercises 20
(i) Tx is the time-until-death for (x) random variable.

(ii) The force of mortality is constant.
(iii) ex = 15.63
Calculate the variance of Tx .
(A) 240 (B) 250 (C) 260 (D) 270 (E) 280
1.22. Originally, mortality for Daniel, currently aged 30, is such that:
(i) e30 = 40.78

(ii) e30:15 = 14.07
(iii) 15 p30 = 0.8764 and 16 p30 = 0.8664
(iv) The limiting age is 100.
Now, it is believed that in the year of age between ages 45 and 46, Daniel
will be subject to an additional risk such that the constant 0.15 will be added
to the force of mortality µ45 (t) for 0 ≤ t < 1.
Calculate the revised value of e30 for Daniel, accounting for the additional
risk in the year of age between ages 45 and 46.
(A) 36 (B) 37 (C) 38 (D) 39 (E) 40
1.23. For (x):
(i) K is the curtate future lifetime random variable.

(ii) ½
0.20 for k = 0, 1, 2
k| qx =
0.40 for k = 3
Calculate the standard deviation of K.
(A) 1.1 (B) 1.2 (C) 1.3 (D) 1.4 (E) 1.5

1.2 Exercises 21
(i) µ(x) = B(1.05)x for x ≥ 0, B > 0.

(ii) p51 = 0.9877
Calculate: B.
(A) 0.001 (B) 0.002 (C) 0.003 (D) 0.004 (E) 0.005
(i) The force of mortality for Vivian is µVx = µ for x ≥ 0, µ > 0.

1
(ii) The force of mortality for Augustine is µA
x = 90−x for 0 ≤ x < 90.
Calculate µ so that 10 p30 is the same for Vivian and Augustine.
(A) 0.016 (B) 0.018 (C) 0.020 (D) 0.022 (E) 0.024
1.26. Consider the following life table, where missing entries are denoted
by “—”:
x qx lx dx
48 — 90,522 —
49 0.007453 89,900.9286 —
Calculate the expected number of deaths between ages 48 and 50.
(A) 1280 (B) 1290 (C) 1300 (D) 1310 (E) 1320

1.2 Exercises 22
1.27. You are given the following life table, where missing entries are
denoted by “—”:
x lx qx ex
65 79,354 0.0172 —
66 — 0.0186 —
67 — — —
68 74,993 — 14.89
69 — — 14.22
Calculate the expected number of deaths between ages 67 and 69.
(A) 2970 (B) 3020 (C) 3070 (D) 3120 (E) 3170
(i) lx = 1000(ω - x) for 0 ≤ x ≤ ω

(ii) µ30 = 0.0125
(A) 17.1 (B) 17.6 (C) 18.1 (D) 18.6 (E) 19.1
1.29. You are given the following life table, where missing values are indi-
cated by “—”:
x lx dx px
0 1000.0 — 0.875
1 — 125.0 —
2 — — —
3 — — 0.680
4 — 182.5 —
5 200.0 — —
Calculate: 2| q0 .
(A) 0.16 (B) 0.17 (C) 0.18 (D) 0.19 (E) 0.20

1.2 Exercises 23
1.30. Woolhouse is currently age 40. Woolhouse’s mortality follows 130%

of the Illustrative Life Table; that is, the probability that Woolhouse dies
between ages x and x + 1 is 130% of the probability of death between ages
x and x + 1 in the Illustrative Life Table for x = 40, 41, ..., 110.
Calculate Woolhouse’s 4-year temporary curtate life expectancy.
(A) 3.950 (B) 3.955 (C) 3.960 (D) 3.965 (E) 3.970
1.31. Suppose mortality follows the Illustrative Life Table, and deaths are
uniformly distributed within each year of age.
Calculate: 4.5 q40.3 .
(A) 0.0141 (B) 0.0142 (C) 0.0143 (D) 0.0144 (E) 0.0145
1.32. Suppose mortality follows the Illustrative Life Table, where deaths
are assumed to be uniformly distributed between integer ages.
Calculate the median future lifetime for (32).
(A) 44.7 (B) 45.0 (C) 45.3 (D) 45.6 (E) 45.9
1.33. Suppose mortality follows the Illustrative Life Table with the as-
sumption that deaths are uniformly distributed between integer ages.
Calculate: 0.9 q60.6 .
(A) 0.0130 (B) 0.0131 (C) 0.0132 (D) 0.0133 (E) 0.0134
1.34. You are given the mortality rates:

q30 = 0.020, q31 = 0.019, q32 = 0.018.
Assume deaths are uniformly distributed over each year of age.
Calculate the 1.4-year temporary complete life expectancy for (30).
(A) 1.377 (B) 1.379 (C) 1.381 (D) 1.383 (E) 1.385

1.2 Exercises 24
1.35. Using the Illustrative Life Table, calculate: 11|17 q42 .
(A) 0.20 (B) 0.21 (C) 0.23 (D) 0.24 (E) 0.25
1.36. Consider two survival models A and B:
(i) For Model A: lx = 1000(ωA - x) for 0 ≤ x ≤ ωA

(ii) For Model B: lx = 500(ωB − x)α for 0 ≤ x ≤ ωB , α > 0
Furthermore:
(1) For Model B, the force of mortality at age 55 is 0.046.

(2) The complete expectation of life for (40) under Model A is 39.615%
higher than the complete expectation of life for (40) under Model B.
(3) For Model A, the probability that (45) survives the first 20 years and
dies in the subsequent 10 years is 0.20.
For Model B, calculate the probability that (45) dies between ages 65 and
75.
(A) 0.16 (B) 0.17 (C) 0.18 (D) 0.19 (E) 0.20
1.37. Consider a population that consists of 600 lives aged 50 and 520 lives
aged 60.
Each life has mortality that follows the Illustrative Life Table, and all lives
have independent future lifetime random variables.
Calculate the standard deviation of the total number of survivors to age
80.
(A) 14.7 (B) 15.2 (C) 15.7 (D) 16.2 (E) 16.7

1.2 Exercises 25
1.38. (Exam MLC Only:) Suppose:
(i) q70 = 0.04 and q71 = 0.05.

(ii) Let UDD denote a uniform distribution of deaths assumption within
each year of age, and let CF denote a constant force of mortality within each
year of age.
Calculate the probability that (70.6) will die within the next 0.5 years
under UDD minus the probability that (70.6) will die within the next 0.5
years under CF.
(A) 0.00008 (B) 0.00010 (C) 0.00012 (D) 0.00014 (E) 0.00016
1.39. (Exam MLC Only:) You are given:
(i) The force of mortality is constant between integer ages.

(ii) 0.3 qx+0.7 = 0.10
Calculate: qx .
(A) 0.24 (B) 0.26 (C) 0.28 (D) 0.30 (E) 0.32
1.40. (Exam MLC Only:) A life insurer issues Roderick, aged 40, a pol-
icy that will pay 10,000 upon survival of a number of years equal to Roderick’s
median future lifetime. You are given:
(i) d = 0.04
(ii) For Roderick: q40+k = 0.05(1 + k) for k = 0, 1, ..., 19. (Roderick is a
very unfortunate individual with respect to his future lifetime distribution.)
(iii) The force of mortality is constant between integer ages.
Calculate the expected present value of the 10,000 payment; that is, cal-
culate the present value of 10,000 times the probability that the 10,000 will
be paid to Roderick.
(A) 4100 (B) 4130 (C) 4160 (D) 4190 (E) 4220

1.2 Exercises 26
Answers to Exercises
1.1. E 1.26. B
1.2. C 1.27. E
1.3. B 1.28. A
1.4. A 1.29. D
1.5. B 1.30. C
1.6. C 1.31. E
1.7. B 1.32. C
1.8. B 1.33. A
1.9. A 1.34. C
1.10. C 1.35. C
1.11. D 1.36. D
1.12. D 1.37. E
1.13. B 1.38. A
1.14. A 1.39. D
1.15. C 1.40. C
1.16. E
1.17. D
1.18. C
1.19. D
1.20. A
1.21. C
1.22. B
1.23. B
1.24. A
1.25. B

1.3 Past Exam Questions 27
1.3 Past Exam Questions

• Exam MLC, Fall 2013: #24, 25
• Exam 3L, Fall 2013: #1, 2, 3
• Exam MLC, Spring 2013: #20
• Exam 3L, Spring 2013: #1, 2, 3
• Exam MLC, Fall 2012: #3
• Exam 3L, Fall 2012: #1, 2, 3
• Exam MLC, Spring 2012: #2 (MLC Only)
• Exam 3L, Spring 2012: #1, 2, 3
• Exam MLC, Sample Questions: #13, 21, 22, 28, 32, 59, 65, 98, 106, 116,
120, 131, 145, 155, 161, 171, 188, 189, 200, 201, 207, 219, 223, 267 (MLC
Only), 276
• Exam 3L, Fall 2011: #1, 2
• Exam 3L, Spring 2011: #1, 2, 3
• Exam 3L, Fall 2010: #1, 2, 3
• Exam 3L, Spring 2010: #1, 2, 3, 4
• Exam 3L, Fall 2009: #1, 2, 3
• Exam 3L, Spring 2009: #1, 3
• Exam 3L, Fall 2008: #12, 13, 14
• Exam 3L, Spring 2008: #13, 14, 15, 16
• Exam MLC, Spring 2007: #1, 21

28
2 SELECTION (Exam MLC Only)
2.1 Key Concepts

For a life table based on an insured population, one must consider for each
individual both (i) the age of policy issue and (ii) the time that has elapsed
since policy issue. This is because the insurer typically underwrites individ-
uals that purchase a policy. Through the underwriting, the insurer learns
additional information about the individual’s survival distribution that the
insurer would not know for a life randomly drawn from the general popula-
tion. This additional information must be accounted for in the calculation
of various quantities such as survival probabilities for the individual and the
value of the individual’s policy.
• Consider an individual now aged x + t who purchased a policy at age x.

We say that the individual was selected, or select, at age x (and time t
= 0).
• The additional information gained from underwriting the above individ-

ual, obtained by surveys and/or a medical examination, is assumed to ap-
ply for a certain number of years after policy issue called the select period.
• Say the select period is d years. For t < d, one accounts for the initial
selection of the above individual at age x; the individual’s current age
would be written as [x] + t (the select brackets [ ] denote the initial age
of selection). For t ≥ d, one no longer accounts for the initial selection of
the individual at age x, and the individual’s age would be written simply
as x + t (with no select brackets [ ]).
• An individual has select mortality for ages/times within the select period
that differs from the mortality of the general population. An individual
has ultimate mortality for ages/times beyond the select period where
their mortality is assumed to be the same as a life from the general
population.
• A life table that accounts for both select and ultimate mortality is called
a select-and-ultimate life table.

2.1 Key Concepts 29
• A life table that ignores selection completely is called an aggregate life table.
• The previous formulas for the quantities considered so far, such as sur-
vival probabilities, are still valid in the event of selection. One simply
has to use information from the select part of the select-and-ultimate life
table for ages/times within the select period.
• For example, with a select period of 3 years, 2 p[x] = (p[x] )(p[x]+1 ) and 5 p[x]
= (p[x] )(p[x]+1) (p[x]+2 )(px+3 )(px+4 ). The p’s with select brackets would
come from the select part of the select-and-ultimate life table, and the
p’s without select brackets would come from the ultimate part.
• An illustrative select-and-ultimate table is the

Standard Select and Ultimate Survival Model. This table is provided in
Appendix D of Dickson et al. For brevity, I will refer to this table as the
Standard Select Survival Model.

2.2 Exercises 30
2.2 Exercises
2.1. Mortality follows the select-and-ultimate life table:
x l[x] l[x]+1 lx+2 x+2

30 9,906 9,904 9,901 32
31 9,902 9,900 9,897 33
32 9,898 9,896 9,892 34
33 9,894 9,891 9,887 35
34 9,889 9,886 9,882 36
Calculate: 10,0001| q[30] .
(A) 1 (B) 2 (C) 3 (D) 4 (E) 5
2.2. Suppose mortality follows the Standard Select Survival Model.

Calculate: 1|2 q[70]+1 .
(A) 0.025 (B) 0.027 (C) 0.029 (D) 0.031 (E) 0.033
2.3. Consider the following select-and-ultimate life table:

x q[x] q[x]+1 q[x]+2 qx+3 x+3
60 0.09 0.11 0.13 0.15 63
61 0.10 0.12 0.14 0.16 64
62 0.11 0.13 0.15 0.17 65
63 0.12 0.14 0.16 0.18 66
64 0.13 0.15 0.17 0.19 67
Assume that deaths follow the uniform distribution of deaths assumption

between integer ages.
Calculate: 1.6 q[61]+0.75 .
(A) 0.1855 (B) 0.1856 (C) 0.1857 (D) 0.1858 (E) 0.1859

2.2 Exercises 31
2.4. A select-and-ultimate life table with a select period of 2 years is based

on probabilities that satisfy the following relationship:
3
q[x−i]+i = 5−i × qx for i = 0, 1.
You are given that l68 = 10,000, q66 = 0.026, and q67 = 0.028.
Calculate: l[65]+1 .
(A) 10,414 (B) 10,451 (C) 10,479 (D) 10,493 (E) 11,069
2.5. Suppose mortality follows the Standard Select Survival Model.

Calculate: e[60]:5 .
(A) 4.928 (B) 4.932 (C) 4.936 (D) 4.940 (E) 4.944
2.6. Consider a select-and-ultimate life table with a 2-year select period.

You are given:
(i) l[35] = 1500

(ii) l36 = 1472.31
(iii) q[35] = 0.0240
(iv) q[35]+1 = 0.0255
Calculate: l35 (1| q35 ).
(A) 42 (B) 44 (C) 46 (D) 48 (E) 50

2.2 Exercises 32
2.7. Quinn is currently age 60. He was selected by the PlzDntDie Life
Insurance Company one year ago. Quinn has mortality that follows a select-
and-ultimate life table with a 2-year select period:
(i) The ultimate part of the model is such that:

µx = 0.0002(1.1)x for x ≥ 0.
(ii) The select part of the model is such that:

µ[x]+s = (0.8)2−s µx+s for x ≥ 0, 0 ≤ s ≤ 2.
Calculate the probability that Quinn survives to age 61.
(A) 0.938 (B) 0.944 (C) 0.950 (D) 0.956 (E) 0.962
2.8. You are given:
(i) Mortality follows the Standard Select Survival Model.

(ii) Deaths are uniformly distributed over each year of age.
Calculate: e̊[75]+1:1.3 .
(A) 1.277 (B) 1.280 (C) 1.283 (D) 1.287 (E) 1.290
2.9. For a select and ultimate life table with a 1-year select period:
(i) µ[55]+t = 0.5µ55+t for 0 ≤ t ≤ 1

(ii) e55 = 18.02
(iii) e[55] = 18.33
Calculate: e56 .
(A) 17.60 (B) 17.65 (C) 17.70 (D) 17.75 (E) 17.80

2.2 Exercises 33
2.10. Consider a population of lives each age 55 and selected at that age,
where 70% are non-smokers and 30% are smokers.
The force of mortality is:
µ[55]+t = 0.009t(0.4 + 0.1S) for t ≥ 0,
where S = 0 for a non-smoker and S = 1 for a smoker.
Calculate the probability that a randomly chosen life from the above pop-
ulation will die before age 75.
(A) 0.51 (B) 0.52 (C) 0.53 (D) 0.54 (E) 0.55

2.2 Exercises 34
2.1. C
2.2. B
2.3. C
2.4. D
2.5. E
2.6. C
2.7. B
2.8. C
2.9. B
2.10. D


• Exam MLC, Sample Questions: #66, 73, 136, 168

36
3 INSURANCE BENEFITS I
3.1 Key Concepts

A life insurance policy provides a lump sum benefit typically upon the death
of the policyholder. The benefit may also be called the face amount or the
sum insured of the policy.
For insurances with death benefits, the insurance can be described as either
(i) continuous: the death benefit is payable at the moment of death of the
policyholder or (ii) discrete: the death benefit is payable at the end of the
period of death of the policyholder, where “period” can be a year or a fraction
of a year such as a month.
Types of Life Insurance:
• Level Benefit Insurance

– Whole Life Insurance: Provides a benefit upon the death of the
policyholder. This is also called permanent life insurance.
– Term Life Insurance: Provides a benefit upon the death of the
policyholder if the policyholder dies within an n-year period after
policy issue. This is also called term insurance.
– Pure Endowment: Provides a benefit at the end of n years if the
policyholder survives the n years after policy issue.
– Endowment Insurance: A term insurance plus a pure endowment.
This insurance will provide a benefit upon the death of the policy-
holder if the policyholder dies within an n-year period after policy
issue. If the policyholder survives the n-year period, then a benefit is
payable at the end of the n years.

3.1 Key Concepts 37
– Deferred Insurance: The policyholder must survive a u-year pe-

riod after policy issue, called the deferred period or deferral period, in
order for any benefits to be payable. A deferred whole life insurance
provides a benefit upon the death of the policyholder if the policy-
holder survives the u-year deferral period. A deferred term insurance
provides a benefit upon the death of the policyholder if the policy-
holder dies within an n-year period after surviving the u-year deferral
period; that is, the benefit is payable upon death of the policyholder
if death occurs between u years and u + n years after policy issue.
Varying Benefit Insurance

– Annually Increasing Insurance: A life insurance where the death
benefit is increased arithmetically for each year the policyholder sur-
vives. For example, an annually increasing term insurance pays S if
the policyholder dies in the first year after policy issue, 2S if the
policyholder dies in the second year after policy issue, ..., nS if the
policyholder dies in the n-th year after policy issue.
– Annually Decreasing Insurance: A life insurance where the death
benefit is decreased arithmetically for each year the policyholder sur-
vives. For example, an annually decreasing term insurance pays nS
if the policyholder dies in the first year after policy issue, (n − 1)S if
the policyholder dies in the second year after policy issue, ..., S if the
policyholder dies in the n-th year after policy issue.
For each of these insurances, the following will be considered:
• The present value of the benefit, Z, is the benefit discounted for interest
between policy issue and the benefit payment date. This is a random
variable, as the benefit payment date is a function of the future lifetime
of the policyholder.
• The expected present value of the benefit, E(Z), is the benefit discounted
for both interest and mortality between policy issue and all potential
benefit payment dates. E(Z) is with respect to the distribution of the
policyholder’s future lifetime.

3.1 Key Concepts 38
E(Z) will be written differently for each type of life insurance consid-
ered. In addition, E(Z) can also be called the actuarial present value
of the insurance, the single premium, the net single premium,
or the single benefit premium.
• Secondary characteristics of the distribution of Z that will be of interest

include the variance of Z and percentiles of the distribution of Z.
Interest Theory Concepts:
• Let i denote the effective annual interest rate.
1
• The discount factor: v = 1+i .
i
• The discount rate, d, is such that: d = 1+i = iv = 1 - v.
• The force of interest, δ, is such that: δ = ln(1 + i).
1
Also: eδ = 1 + i = 1−d =⇒ e−δ = v = 1 - d.
1−v n 1−v n 1−v n

• an = i , än = d , ān = δ
än −nv n än −nv n än −nv n ¯ n = ān −nv n

• (Ia)n = i , (Iä)n = d , (Iā)n = δ , (Iā) δ
n−an n−an n−an n−ān

• (Da)n = i , (Dä)n = d , (Dā)n = δ , (D̄ā)n = δ

3.1 Key Concepts 39
Life Insurance Formulas
This section provides key formulas for different life insurances. Note:
• Most life insurances have continuous and discrete versions. In the con-
tinuous case, the death benefit is payable at the moment of death of
the policyholder (Tx years after policy issue). The discrete case can be:
(i) annual, where the death benefit is payable at the end of the year of
death of the policyholder (Kx + 1 years after policy issue) and (ii) m-thly
(Exam MLC only).
It is possible that the death benefits could be subject to a finite term

and/or a deferral period. It is also possible that there may be a survival
benefit.
• The general formula for the present value of a continuous life insurance
on (x) that pays bt at time t (> 0) is:
Z = bTx v Tx = bTx e−δTx .
The general formula for the expected present value of a continuous life
insurance on (x) that pays bt at time t (> 0) is:
R∞ R∞
E(Z) = 0 bt v t t px µx+t dt = 0 bt e−δt t px µx+t dt. Furthermore:
“Say the moment of death is at time t. Then the present value of this ben-
efit is bt v t , and the expected present value of this benefit is bt v t t px µx+t dt
((x) has to survive t years and then immediately die for bt to be paid at
time t). Integrating over all possible times of death provides the overall
expected present value.”
R∞
E(Z j ) = 0 (bt v t )j t px µx+t dt for j = 1, 2, ...

3.1 Key Concepts 40
• The general formula for the present value of an annual life insurance on
(x) that pays bk+1 at time k + 1 (k = 0, 1, 2, ...) is:
Z = bKx +1 v Kx +1 .
The general formula for the expected present value of an annual life
insurance on (x) that pays bk+1 at time k + 1 (k = 0, 1, 2, ...) is:
P∞ k+1
E(Z) = k=0 bk+1 v k| qx . Furthermore:
“Say death occurs in year (k + 1). Then the present value of this
benefit is bk+1 v k+1 , and the expected present value of this benefit is
bk+1 v k+1 k| qx . Summing over all possible values of k provides the over-
all expected present value.”
P∞
E(Z j ) = k=0 (bk+1 v
k+1 j
) k| qx for j = 1, 2, ...

3.1 Key Concepts 41
• The International Actuarial Notation for E(Z) often contains an A, which

indicates that the expected present value is for a life insurance. For
example, Ax denotes the expected present value of a whole life insurance
of 1 on (x) payable at the end of the year of death.
• 2 A is A evaluated at double the force of interest. Doubling the force of

interest, from δ to 2δ, affects other interest rates as follows:
– i changes to i2 + 2i
– (1 + i) changes to (1 + i)2
– v changes to v 2
– d changes to 2d - d2
Formulas assuming a force of interest of δ can be easily modified for a

force of interest of 2δ. Consider the following equation involving a whole
life insurance of 1 on (x): Ax = vqx + vpx Ax+1 . Doubling the force of
interest results in the equation: 2 Ax = v 2 qx + v 2 px (2 Ax+1 ).
• 2 A is NOT the notation for E(Z 2 ).
• For most insurance, a basic benefit of 1 is assumed. If the basic benefit

is S, notation and formulas are adjusted. For example, the expected
present Rvalue of a continuous whole life insurance of
R ∞S on (x) is E(Z) =
∞ t 2 2 2 2 2t
S Āx = 0 Sv t px µx+t dt, and E(Z ) = (S ) Āx = 0 (S )v t px µx+t dt.
• There is only one version of the pure endowment as there is no death

benefit.

3.1 Key Concepts 42
• A common problem involves a portfolio of N independent and identically

distributed life insurance policies. N is assumed to be “large” (≥ 30).
The present value random variable for policy j is Zj for j = 1, 2, .... N .
P
Let S = N j=1 Zj = present value of all policies in the portfolio.
Then: E(S) = N E(Z) and V ar(S) = N V ar(Z).

Assume that a policyholder pays for a policy with a single premium at
issue.
The minimum amount that the insurer should collect from the N lives
at issue, h, such that the probability is approximately α (between 0 and
1) that the insurer will have sufficient funds to pay the future benefits
solves:
h−E(S) h−E(S)
P r(S ≤ h) = Φ( √ ) = α, where √ = zα , the 100αth percentile
V ar(S) V ar(S)
of the standard normal distribution.
• Recursion formulas take an expected present value and decomposes it

into the sum of two expected present values: the expected present value
of the benefit if the policyholder dies in the first period plus the expected
present value of the benefit if the policyholder survives the first period.
• For formulas with a constant force of mortality, we will sometimes use:
– px = e−µ = p
– qx = 1 - e−µ = 1 - p = q

3.1 Key Concepts 43
Level Benefit Insurance
Whole Life Insurance of 1 on (x):

Continuous Whole Life Insurance:
• Z = v Tx
R∞
• E(Z) = Āx = 0 v t t px µx+t dt
1
– For de Moivre’s Law (Uniform Distribution): Āx = ω−x āω−x at force
of interest δ
µ
– With a constant force of mortality: Āx = µ+δ
R∞
• E(Z 2 ) = 2 Āx = 0 v 2t t px µx+t dt
1
– For de Moivre’s Law (Uniform Distribution): 2 Āx = ω−x āω−x at force
of interest 2δ
µ
– With a constant force of mortality: 2 Āx = µ+2δ
• V ar(Z) = 2 Āx - [Āx ]2
• FZ (z) = P r[Z ≤ z] = P r[v Tx ≤ z] = c px where c = - ln(z)

δ
ln(z)
– For de Moivre’s Law (Uniform Distribution): FZ (z) = 1 + δ(ω−x) for
v ω−x ≤ z ≤ 1. If the death benefit is S, replace z with Sz .
µ
– With a constant force of mortality: FZ (z) = z δ for 0 ≤ z ≤ 1. If the
death benefit is S, replace z with Sz .
Note: The 100α-th percentile of the distribution of Z, zα , solves:

FZ (zα ) = α for 0 ≤ α ≤ 1.

3.1 Key Concepts 44
Annual Whole Life Insurance:
• Z = v Kx +1
P∞ k+1
• E(Z) = Ax = k=0 v k| qx
1
– For de Moivre’s Law (Uniform Distribution): Ax = ω−x aω−x at force
of interest δ
q
– With a constant force of mortality: Ax = q+i
P∞
• E(Z 2 ) = 2 Ax = k=0 v
2(k+1)
k| qx
1
– For de Moivre’s Law (Uniform Distribution): 2 Ax = ω−x aω−x at force
of interest 2δ
q
– With a constant force of mortality: 2 Ax = q+i2 +2i
• V ar(Z) = 2 Ax - [Ax ]2
• Recursion: Ax = vqx + vpx Ax+1

3.1 Key Concepts 45
Term Insurance of 1 on (x):

Continuous Term Insurance:
• ½
v Tx for Tx ≤ n
Z=
0 for Tx > n
1
Rn
• E(Z) = Āx:n = 0 v t t px µx+t dt
1 1
– For de Moivre’s Law (Uniform Distribution): Āx:n = ω−x ān at force
of interest δ
1 µ
– With a constant force of mortality: Āx:n = µ+δ [1 − exp[−(µ + δ)n]]
Rn
• E(Z 2 ) = 2 Āx:n
1
= 0 v 2t t px µx+t dt
1
– For de Moivre’s Law (Uniform Distribution): 2 Āx:n
1
= ω−x ān at force
of interest 2δ
µ
– With a constant force of mortality: 2 Āx:n
1
= µ+2δ [1 − exp[−(µ + 2δ)n]]
• V ar(Z) = 2 Āx:n
1 1
- [Āx:n ]2

3.1 Key Concepts 46
Annual Term Insurance:
• ½
v Kx +1 for Kx = 0, 1, ..., n - 1
Z=
0 for Kx = n, n + 1, ...
1
Pn−1
• E(Z) = Ax:n = k=0 v k+1 k| qx
1 1
– For de Moivre’s Law (Uniform Distribution): Ax:n = ω−x an at force
of interest δ
1 q
– With a constant force of mortality: Ax:n = q+i [1 − (vp)n ]
Pn−1
• E(Z 2 ) = 2 Ax:n
1
= k=0 v 2(k+1) k| qx
1
– For de Moivre’s Law (Uniform Distribution): 2 Ax:n
1
= ω−x an at force
of interest 2δ
q
– With a constant force of mortality: 2 Ax:n
1
= q+i2 +2i [1 − (v 2 p)n ]
• V ar(Z) = 2 Ax:n
1 1
- [Ax:n ]2
1 1
• Recursion: Ax:n = vqx + vpx Ax+1:n−1

3.1 Key Concepts 47
Pure Endowment of 1 on (x):

Pure Endowment:
• ½
0 for Tx ≤ n
Z=
vn for Tx > n
• E(Z) = Ax:n1 = n Ex = v n n px
Note: There are two notations for the expected present value of a pure
endowment: the A-notation and the E-notation.
– For de Moivre’s Law (Uniform Distribution): Ax:n1 = (e−δn ) ω−x−n

ω−x
– With a constant force of mortality: Ax:n1 = exp[−(µ + δ)n]
• E(Z 2 ) = 2 Ax:n1 = v 2n n px
– For de Moivre’s Law (Uniform Distribution): 2 Ax:n1 = (e−2δn ) ω−x−n

ω−x
– With a constant force of mortality: 2 Ax:n1 = exp[−(µ + 2δ)n]
• V ar(Z) = 2 Ax:n1 - [Ax:n1 ]2 = v 2n (n px )(n qx )
• Recursion: Ax:n1 = 0 + vpx Ax+1:n−1

1

3.1 Key Concepts 48
Endowment Insurance of 1 on (x):

Continuous Endowment Insurance:
• ½
v Tx for Tx ≤ n
Z=
vn for Tx > n
1
• E(Z) = Āx:n = Āx:n + Ax:n1
• E(Z 2 ) = 2 Āx:n = 2 Āx:n

1
+ 2 Ax:n1
• V ar(Z) = 2 Āx:n - [Āx:n ]2
Annual Endowment Insurance:
• ½
v Kx +1 for Kx = 0, 1, ..., n - 1
Z=
vn for Kx = n, n + 1, ...
1
• E(Z) = Ax:n = Ax:n + Ax:n1
• E(Z 2 ) = 2 Ax:n = 2 Ax:n

1
+ 2 Ax:n1
• V ar(Z) = 2 Ax:n - [Ax:n ]2
• Recursion: Ax:n = vqx + vpx Ax+1:n−1

3.1 Key Concepts 49
Deferred Whole Life Insurance of 1 on (x):

Continuous Deferred Whole Life Insurance:
• ½
0 for Tx ≤ u
Z=
v Tx for Tx > u
R∞
• E(Z) = u| Āx = u v t t px µx+t dt = Āx - Āx:u
1
= u Ex Āx+u
R∞
• E(Z 2 ) = 2u| Āx = u v 2t t px µx+t dt = 2 Āx - 2 Āx:u
1
= (2 Ax:u1 )(2 Āx+u )
• V ar(Z) = 2u| Āx - [u| Āx ]2
Annual Deferred Whole Life Insurance:
• ½
0 for Kx = 0, 1, ..., u - 1
Z=
v Kx +1 for Kx = u, u + 1, ...
P∞ k+1 1
• E(Z) = u| Ax = k=u v k| qx = Ax - Ax:u = u Ex Ax+u
P∞
• E(Z 2 ) = 2u| Ax = k=u v
2(k+1)
k| qx = 2 Ax - 2 Ax:u
1
= (2 Ax:u1 )(2 Ax+u )
• V ar(Z) = 2u| Ax - [u| Ax ]2
• Recursion: u| Ax = 0 + vpx (u−1| Ax+1 )

3.1 Key Concepts 50
Deferred Term Insurance of 1 on (x):

Continuous Deferred Term Insurance:
• ½
v Tx for u < Tx ≤ u + n
Z=
0 otherwise
1
R u+n
• E(Z) = u|n Āx = u| Āx:n = u v t t px µx+t dt = Āx:u+n
1 1
- Āx:u 1
= u Ex Āx+u:n
R u+n
• E(Z 2 ) = 2u|n Āx = u v 2t t px µx+t dt = 2 Āx:u+n
1
- 2 Āx:u
1
= (2 Ax:u1 )(2 Āx+u:n
1
)
2
• V ar(Z) = u|n Āx - [u|n Āx ]2
Annual Deferred Term Insurance:
• ½
v Kx +1 for Kx = u, u + 1, ..., u + n - 1
Z=
0 otherwise
1
Pu+n−1
• E(Z) = u|n Ax = u| Ax:n = k=u v k+1 k| qx = Ax:u+n
1 1
- Ax:u 1
= u Ex Ax+u:n
Pu+n−1
• E(Z 2 ) = 2u|n Ax = k=u v 2(k+1) k| qx = 2 Ax:u+n
1
- 2 Ax:u
1
= (2 Ax:u1 )(2 Ax+u:n
1
)
2
• V ar(Z) = u|n Ax - [u|n Ax ]2
• u|n Ax = 0 + vpx (u−1|n Ax+1 )

3.1 Key Concepts 51
Varying Benefit Insurance
Annually Increasing Whole Life Insurance on (x):

Continuous Annually Increasing Whole Life Insurance:
• Z = (bTx c + 1)v Tx
R∞
• E(Z) = (I Ā)x = 0 (btc + 1)v t t px µx+t dt
1
– For de Moivre’s Law (Uniform Distribution): (I Ā)x = ω−x (Iā)ω−x at
force of interest δ
Continuously Increasing Whole Life Insurance:
• Z = Tx v Tx
R∞
• E(Z) = (I¯Ā)x = 0 tv t t px µx+t dt
– For de Moivre’s Law (Uniform Distribution): (I¯Ā)x = 1 ¯

ω−x (Iā)ω−x at
– With a constant force of mortality: (I¯Ā)x = µ

(µ+δ)2
Annual Annually Increasing Whole Life Insurance:
• Z = (Kx + 1)v Kx +1
P∞
• E(Z) = (IA)x = k=0 (k + 1)v k+1 k| qx
1
– For de Moivre’s Law (Uniform Distribution): (IA)x = ω−x (Ia)ω−x at
• Recursion: (IA)x = Ax + vpx (IA)x+1

3.1 Key Concepts 52
Annually Increasing Term Insurance on (x):

Continuous Annually Increasing Term Insurance:
• ½
(bTx c + 1)v Tx for Tx ≤ n
Z=
0 for Tx > n
1
Rn
• E(Z) = (I Ā)x:n = 0 (btc + 1)v t t px µx+t dt
1 1
– For de Moivre’s Law (Uniform Distribution): (I Ā)x:n = ω−x (Iā)n at
Continuously Increasing Term Insurance:
• ½
Tx v Tx for Tx ≤ n
Z=
0 for Tx > n
Rn
• E(Z) = (I¯Ā)x:n
1
= 0 tv t t px µx+t dt
– For de Moivre’s Law (Uniform Distribution): (I¯Ā)x:n

1
= 1 ¯
ω−x (Iā)n at
Annual Annually Increasing Term Insurance:
• ½
(Kx + 1)v Kx +1 for Kx = 0, 1, ..., n - 1
Z=
0 for Kx = n, n + 1, ...
1
Pn−1
• E(Z) = (IA)x:n = k=0 (k + 1)v k+1 k| qx
1 1
– For de Moivre’s Law (Uniform Distribution): (IA)x:n = ω−x (Ia)n at
1 1 1
• Recursion: (IA)x:n = Ax:n + vpx (IA)x+1:n−1

3.1 Key Concepts 53
Annually Decreasing Term Insurance on (x):

Continuous Annually Decreasing Term Insurance:
• ½
(n − bTx c)v Tx for Tx ≤ n
Z=
0 for Tx > n
1
Rn
• E(Z) = (DĀ)x:n = 0 (n − btc)v t t px µx+t dt
1 1 1
Also: (I Ā)x:n + (DĀ)x:n = (n + 1)Āx:n .
Continuously Decreasing Term Insurance:
• ½
(n − Tx )v Tx for Tx ≤ n
Z=
0 for Tx > n
1
Rn
• E(Z) = (D̄Ā)x:n = 0 (n − t)v t t px µx+t dt
Also: (I¯Ā)1 + (D̄Ā)x:n
x:n
1 1
= nĀx:n .
Annual Annually Decreasing Term Insurance:
• ½
(n − Kx )v Kx +1 for Kx = 0, 1, ..., n - 1
Z=
0 for Kx = n, n + 1, ...
1
Pn−1
• E(Z) = (DA)x:n = k=0 (n − k)v k+1 k| qx
1 1 1
Also: (IA)x:n + (DA)x:n = (n + 1)Ax:n .
1 1
• Recursion: (DA)x:n = nvqx + vpx (DA)x+1:n−1

3.2 Exercises 54
3.2 Exercises
3.1. Suppose:
(i) Mortality follows lx = 110 - x for 0 ≤ x ≤ 110.

(ii) i = 0.08.
1
Calculate: 1000Ā25:30 .
(A) 110 (B) 120 (C) 130 (D) 140 (E) 150
3.2. For a 20-year term insurance of 1 on (x):
(i) The death benefit is payable at the moment of death.

(ii) µx (t) = 0.04 for t ≥ 0
(iii) δ = 0.06
(iv) Z is the present value random variable for this insurance.
Calculate: V ar(Z).
(A) 0.10 (B) 0.11 (C) 0.12 (D) 0.13 (E) 0.14
3.3. Consider a life insurance on (40) that pays 3000 at the end of the
year of death if death occurs between ages 45 and 55, and pays 5000 at the
end of the year of death if death occurs after age 55. Mortality follows the
Illustrative Life Table, and i = 0.06.
Calculate the expected present value of the benefits.
(A) 670 (B) 675 (C) 680 (D) 685 (E) 690
3.4. You are given: lx = 500(90 - x) for 0 ≤ x ≤ 90 and i = 0.06.

Calculate the variance of the present value random variable for a whole life
insurance of 1000 on (35) payable at the end of the year of death.
(A) 62,330 (B) 63,245 (C) 64,660 (D) 65,500 (E) 66,135

3.2 Exercises 55
3.5. Consider a whole life insurance of 1 on (x) payable at the moment of

death:
(i) µx (t) = 0.03 for t ≥ 0, and δ = 0.05

(ii) Z is the present value random variable for this insurance.
(iii) g is the standard deviation of Z.
Calculate: P r[Z > Āx + g].
(A) 0.21 (B) 0.22 (C) 0.46 (D) 0.78 (E) 0.79
3.6. You are given:
½
0.060 for 0 ≤ t < 5
δt =
0.065 for t ≥ 5
½
0.020 for 0 ≤ t < 5
µx (t) =
0.025 for t ≥ 5
Calculate the actuarial present value of a whole life insurance of 1 on (x)

payable at the moment of death.
(A) 0.19 (B) 0.21 (C) 0.23 (D) 0.25 (E) 0.27
3.7. For a special 3-year term insurance on (x) payable at the end of the
year of death:
(i) Z is the present value random variable for this insurance.

(ii) qx+k = 0.02(k + 1) for k = 0, 1, 2
(iii) bk+1 = 5000(k + 1) for k = 0, 1, 2
(iv) i = 0.06
Calculate the standard deviation of Z.
(A) 3340 (B) 3350 (C) 3360 (D) 3370 (E) 3380

3.2 Exercises 56
3.8. Consider a 20-year endowment insurance of 25,000 with death benefit

payable at the end of the year of death on a standard life aged x. The
mortality rates for the first two years for a standard life aged x are such that:
qx = 0.010 and qx+1 = 0.011.
The actuarial present value of this endowment insurance on a standard life

aged x, using i = 0.05, is 10,145.
Sheldon, age x, wants to purchase the above endowment insurance. How-

ever, Sheldon plans on conducting a dangerous two-year physics experiment.
His mortality rates for just the first two years will be different than those of
a standard life aged x; for Sheldon: qx = 0.020 and qx+1 = 0.022.
Calculate the actuarial present value of Sheldon’s endowment insurance,
using i = 0.05.
(A) 10,283 (B) 10,421 (C) 10,527 (D) 10,877 (E) 10,911
3.9. You are given:
(i) A35:1 = 0.9434, A36 = 0.1347, and (IA)35 = 3.7100

(ii) p35 = 0.9964
Calculate: (IA)36 .
(A) 3.77 (B) 3.78 (C) 3.79 (D) 3.80 (E) 3.81
3.10. For a special insurance on (45):
(i) 1000 is payable at the moment of death if death occurs between ages
45 and 65.
(ii) 500 is payable at the moment of death if death occurs between ages 65
and 75.
(iii) Mortality follows: lx = 500(95 - x) for 0 ≤ x ≤ 95, and δ = 0.06.
Calculate the single benefit premium for this insurance.
(A) 240 (B) 245 (C) 250 (D) 255 (E) 260

3.2 Exercises 57
3.11. Let Z be the present value random variable for a whole life insurance
on (x) with a benefit of 10,000 payable at the moment of death.
Assume µx (t) = 0.03 and δt = 0.06 for t ≥ 0.
Calculate the 65th percentile of the distribution of Z.
(A) 3502 (B) 3760 (C) 4030 (D) 4225 (E) 4550
3.12. Consider a special whole life insurance of 10,000 on (30):
(i) Mortality follows the Illustrative Life Table, i = 0.06

(ii) If (30) dies during the first 10 years, the benefit is payable at the end
of the first 10 years. Otherwise, the benefit is payable at the end of the year
of death.
Calculate the actuarial present value of this insurance.
(A) 672 (B) 976 (C) 994 (D) 1025 (E) 6356
3.13. Each of 100 independent lives age 30 purchase a single-premium

continuous 35-year endowment insurance of 1000, where:
(i) Mortality follows de Moivre’s Law with ω = 100, and δ = 0.06

(ii) h is the aggregate amount the insurer receives from the 100 lives.
Using the normal approximation, calculate h such that the probability the
insurer has sufficient funds to pay all claims is 0.99.
(A) 32,000 (B) 32,300 (C) 32,600 (D) 32,900 (E) 33,200

3.2 Exercises 58
3.14. Let Z be the present value random variable for a special continuous
whole life insurance on (x), where for t ≥ 0:
(i) bt = 1000e0.05t
(ii) µx (t) = 0.01 and δt = 0.06
Calculate the expected value of Z.
(A) 420 (B) 440 (C) 460 (D) 480 (E) 500
3.15. Each of 100 independent lives age 30 purchase a single-premium 5-

year deferred whole life insurance of 10 payable at the end of the year of
death, where:
(i) Mortality follows the Illustrative Life Table, and i = 0.06.

(ii) h is the aggregate amount the insurer receives from the 100 lives.
Using the normal approximation, calculate h such that the probability the
insurer has sufficient funds to pay all claims is 0.95.
(A) 105 (B) 112 (C) 116 (D) 121 (E) 128
3.16. A special term insurance policy on (40) pays 1000 at the end of the
year of death for the first ten years and 2000 at the end of the year of death
for the next 10 years. Mortality follows the Illustrative Life Table, i = 0.06.
Calculate the single benefit premium for this policy.
(A) 90 (B) 93 (C) 96 (D) 100 (E) 104
(i) Ax = 0.20, Ax+20 = 0.35

(ii) Ax:20 = 0.50
Calculate: 20| Ax .
(A) 0.08 (B) 0.10 (C) 0.12 (D) 0.14 (E) 0.16

3.2 Exercises 59
(i) qx is a constant for all integer ages x.

(ii) The expected present value of a 2-year term insurance of 1000 on (35)
payable at the end of the year of death is 63.96.
(iii) The probability that the 1000 death benefit for the 2-year term insur-
ance on (35) will be paid is 0.068775.
Calculate the expected present value of a 10-year deferred whole life insur-
ance of 1000 on (35) payable at the end of the year of death.
(A) 150 (B) 165 (C) 170 (D) 175 (E) 180
3.19. For a group of individuals all age 26, you are given:
(i) 20% are smokers and 80% are non-smokers.

(ii) Mortality rates for non-smokers (qxN ) follows the Illustrative Life Table.
(iii) Mortality rates for smokers (qxS ) are double the mortality rates for
non-smokers.
(iv) i = 0.03
Calculate the single benefit premium for a 3-year term insurance of 5000
payable at the end of the year of death on a 26 year old chosen at random
from this group of smokers and non-smokers.
(A) 21.00 (B) 21.50 (C) 22.00 (D) 22.50 (E) 23.00

3.2 Exercises 60
3.20. A group of 100 lives each age 45 set up a fund to pay 10,000 at the
end of the year of death of each member. They each pay into the fund, at
inception, an amount equal to the single benefit premium for a whole life
insurance of 10,000 on (45) payable at the end of the year of death assuming
mortality follows the Illustrative Life Table and i = 0.06.
The actual experience of the fund is one death in the first year and two
deaths in the fourth year; the interest rate is 0.06 in the first and second
years, 0.065 in the third and fourth years, and 0.07 in the fifth year.
Calculate the difference between the expected size of the fund and the
actual size of the fund at the end of the first five years.
(A) 3160 (B) 3200 (C) 3240 (D) 3280 (E) 3320
(i) The covariance of two random variables Z1 and Z2 is equal to the

expected value of the product of Z1 and Z2 , minus the product of the expected
value of Z1 and the expected value of Z2 .
(ii) Z1 denotes the present value random variable for a whole life insurance
of 1000 on (30) payable at the end of the year of death.
(iii) Z2 denotes the present value random variable for a 10-year deferred
whole life insurance of 1000 on (30) payable at the end of the year of death.
(iv) Mortality follows the Illustrative Life Table and i = 0.06.
Calculate the covariance of Z1 and Z2 .
(A) 5700 (B) 5750 (C) 5800 (D) 5850 (E) 5900

3.2 Exercises 61
(i) Mortality follows the Illustrative Life Table, and i = 0.06

(ii) Z is the present value random variable for a whole life insurance of
5000 on (30) payable at the end of the year of death.
Calculate the probability that Z exceeds E(Z).
(A) 0.26 (B) 0.27 (C) 0.28 (D) 0.29 (E) 0.30
(i) For a standard life: k| qx = (0.8)k (0.2) for k = 0, 1, 2, ...

(ii) δ = 0.05
(iii) A certain life has the same mortality as a standard life except that the
probability of death within the first year after policy issue for the certain life
is 50% higher than the probability of death within the first year after policy
issue for a standard life of the same age at policy issue.
Assuming the certain life is issued a whole life insurance of 1 at age x

payable at the end of the year of death, calculate Ax for the certain life.
(A) 0.800 (B) 0.815 (C) 0.830 (D) 0.845 (E) 0.860
(i) The force of mortality is a constant.

(ii) 2 Āx = 0.2414.
(iii) The present value of a dollar to be paid, with certainty, at the end of
10 years is 0.5770.
Calculate the probability that (x) survives at least 15 years but no more
than 20 years.
(A) 0.080 (B) 0.085 (C) 0.090 (D) 0.095 (E) 0.100

3.2 Exercises 62
3.25. For a 10-year deferred 10-year term insurance of 1000 on (x) with
benefits payable at the moment of death:
(i) ½
0.06 for t ≤ 6
δt =
0.07 for t > 6
(ii) ½
0.025 for t ≤ 6
µx (t) =
0.035 for t > 6
Calculate the single benefit premium for this insurance.
(A) 82.50 (B) 83.50 (C) 84.50 (D) 85.50 (E) 86.50
(i) µx = 0.0002(1.1)x for x > 0

(ii) d = 0.06
Calculate: 3|2 A50 .
(A) 0.040 (B) 0.046 (C) 0.052 (D) 0.058 (E) 0.064
3.27. Consider a whole life insurance of 1000 on (x) payable at the end of
the year of death:
(i) v = 0.965
(ii) 10 px = 0.920
(iii) Ax+11 = 0.425
Suppose qx+10 is increased to qx+10 + 0.100.

Calculate the increase in the expected present value of the insurance.
(A) 32 (B) 34 (C) 36 (D) 38 (E) 40

3.2 Exercises 63
3.28. An insurer issues a special life insurance of 100,000 on (x).

If (x) dies during the first year, 100,000 is payable at the end of the first
year. However, if (x) survives the first year, the insurer flips a coin at the end
of the first year. If the coin lands on “heads,” the insurer extends the policy
for a second year so that if (x) dies in the second year, 100,000 is payable at
the end of the second year. But, if the coin lands on “tails,” the policy is
immediately terminated.
If the policy is still in force at the end of the second year, the same coin de-
scribed above is flipped at the end of the second year. The policy is extended
for a third year if the coin lands on heads, and the policy is immediately
terminated if the coin lands on tails. The coin flip continues with the same
results for heads and tails at the end of each year that the policy remains in
force.
Furthermore:
(i) δ = 0.05
(ii) px+k = 0.98 for k = 0, 1, 2, ...
(iii) Let t denote the probability that the coin lands on tails at the end of
the year between times k and k + 1, assuming that the policy is in force at
time k, where k = 0, 1, 2, ... The probability that the coin lands on heads
at the end of the year between times k and k + 1 is 0.5t.
(iv) The results of the coin flip for each year are independent of the mor-
tality of (x).
Calculate the actuarial present value of this special insurance.
(A) 2700 (B) 2760 (C) 2820 (D) 2880 (E) 2940

3.2 Exercises 64
(i) K is the curtate future lifetime random variable for a life aged 50.
(ii) The following present value random variable for an insurance on a life
aged 50:
½
1000v K+1 for K < 5
Z=
1000v 5 for K ≥ 5
(iii) Mortality follows the Illustrative Life Table.

(iv) i = 0.06
Calculate the probability that Z will be greater than 800.
(A) 0.010 (B) 0.012 (C) 0.016 (D) 0.019 (E) 0.027
3.30. A computer store sells a specific brand new laptop with a purchase
price of 1000. This laptop comes with a 5-year warranty such that if the
laptop breaks down in year k + 1, (80 - 20k)% of the purchase price will be
refunded at the end of that year; for k = 0, 1, 2, 3, 4.
Also:
(i) d = 0.03
(ii) The probability that the laptop breaks down in each of the first three
years is 0.05, and the probability that the laptop breaks down in each of the
fourth and fifth years is 0.10.
Calculate the expected present value of the warranty.
(A) 83 (B) 93 (C) 103 (D) 113 (E) 123

3.2 Exercises 65
3.1. D 3.26. B
3.2. C 3.27. C
3.3. C 3.28. B
3.4. A 3.29. D
3.5. A 3.30. C
3.6. E
3.7. A
3.8. B
3.9. E
3.10. D
3.11. D
3.12. C
3.13. B
3.14. E
3.15. B
3.16. B
3.17. E
3.18. D
3.19. D
3.20. A
3.21. C
3.22. C
3.23. B
3.24. D
3.25. D


• Exam MLC, Fall 2013: #12, 13
• Exam 3L, Spring 2012: #13
121, 141, 148, 158, 175, 176, 197, 215, 226, 286, 308
• Exam 3L, Fall 2010: #13
• Exam 3L, Fall 2009: #13, 15
• Exam 3L, Spring 2008: #21, 22
• Exam MLC, Spring 2007: #10, 22, 27

67
4 INSURANCE BENEFITS II (Exam MLC Only)
4.1 Key Concepts

Here, we consider a discrete life insurance where the death benefit is provided
at the end of the m-th of a year of death. The value of m is typically equal
to 2 (half-year), 4 (quarter of a year), or 12 (month).
Additional Interest Theory Concepts:
• The nominal interest rate convertible m-thly, i(m) , is such that:

i(m) m
(1 + m ) = 1 + i.
• The nominal discount rate convertible m-thly, d(m) , is such that:

d(m) m
(1 − m ) = 1 - d = v.
i(m) m d(p) −p
• eδ = 1 + i = (1 + m ) = (1 − d)−1 = (1 − p )
(m) 1−v n (m) 1−v n

• an = i(m)
, än = d(m)

4.1 Key Concepts 68
Additional Life Insurance Formulas
• The m-thly life insurance is such that the death benefit is payable at the
(m)
end of the m-th of a year of death of the policyholder (Kx + m1 years
after policy issue).
(m)
• Kx , measured in years, is the beginning of the m-th of a year after
(m)
policy issue in which the policyholder dies. Kx = 0, m1 , m2 , ...
• The general exact formula for the expected present value of an m-thly
life insurance that pays b k+1 at time k+1
m years (k = 0, 1, 2, ...) is:
m
P∞ k+1
E(Z) = k=0 b k+1 v
m k 1 qx . Furthermore:
m m|m
“Say death occurs in the (k + 1)st m-th of a year. Then the present
k+1
value of this benefit is b k+1 v m , and the expected present value of this
m
k+1
benefit is b k+1 v qx . Summing over all possible values of k provides
m k 1
m m|m
the overall expected present value.”
P∞ k+1
E(Z j ) = k=0 (b k+1 v
m )j k | 1 qx for j = 1, 2, ...
m m m
• The International Actuarial Notation for E(Z) above often contains an

A(m) , which indicates that the expected present value is for an m-thly
(m)
life insurance. For example, Ax denotes the expected present value of a
whole life insurance of 1 on (x) payable at the end of the m-th of a year
of death.
• 2 A(m) is A(m) evaluated at double the force of interest. In addition to the

rates discussed in Insurance Benefits I, doubling the force of interest,
from δ to 2δ, affects other interest rates as follows:
2
– i(m) changes to 12[(1 + i) m - 1]
2
– d(m) changes to 12[1 - v m ]

4.1 Key Concepts 69
• Often, we do not use the exact formulas to calculate expected present

values for m-thly life insurance. Rather, we approximate these expected
present values from the corresponding annual life insurance expected
present values using one of two assumptions:
– UDD: deaths are uniformly distributed within each year of age. In

i
most cases, A(m) = i(m) A. The major exception is the expected present
value for m-thly endowment insurance (see below).
– Claims acceleration: assumes that the death benefit is paid, on aver-
age, at the average of the times it could be paid within the year. In
m−1
most cases, A(m) = (1 + i) 2m A. The major exception is the expected
present value for m-thly endowment insurance (see below).
• Let m approach infinity. Under UDD: Ā = δi A. Under claims accelera-

1
tion: Ā = (1 + i) 2 A.
• Note: The formulas discussed below reduce to the corresponding annual

life insurance formulas in Insurance Benefits I when m = 1. This can
be an aid in memorizing and recalling these formulas.
• In this section, we also consider geometrically increasing annual life insurance.

The death benefit is increased by a constant factor each year to account
for inflation or profit sharing with the policyholder.

4.1 Key Concepts 70
Level Benefit Insurance
m-thly Whole Life Insurance of 1 on (x):

(m) 1
• Z = v Kx +m
(m) P∞ k+1
• E(Z) = Ax = k=0 v
m k 1 qx
m|m
(m) 1 (m)
– For de Moivre’s Law (Uniform Distribution): Ax = ω−x aω−x at force
of interest δ
(m) P∞ 2 k+1
• E(Z 2 ) = 2 Ax = k=0 v
m k 1 qx
m|m
2 (m) 1 (m)
– For de Moivre’s Law (Uniform Distribution): Ax = ω−x aω−x at
force of interest 2δ
(m) (m)
• V ar(Z) = 2 Ax - [Ax ]2
(m) 1 1 (m)
• Recursion: Ax = v m m1 qx + v m m1 px Ax+ 1
m
i
• UDD: A(m) x = A
i(m) x
and Āx = δi Ax
m−1 1
• Claims acceleration: A(m) x = (1 + i) 2m Ax and Āx = (1 + i) 2 Ax

4.1 Key Concepts 71
m-thly Term Insurance of 1 on (x):
• ( (m) 1 (m)
v Kx +m
for Kx = 0, m1 , ..., n - m1
Z= (m)
0 for Kx = n, n + m1 , ...
Pmn−1 k+1
• E(Z) = A(m)1x:n = v m k 1 qx
k=0 m|m
1 (m)
– For de Moivre’s Law (Uniform Distribution): A(m)1x:n = ω−x an at
Pmn−1 k+1
• E(Z 2 ) = 2 A(m)1x:n = v2 m k 1 qx
k=0 m|m
1 (m)
– For de Moivre’s Law (Uniform Distribution): 2 A(m)1x:n = ω−x an at
force of interest 2δ
• V ar(Z) = 2 A(m)1x:n - [A(m)1x:n ]2
1 1
• Recursion: A(m)1x:n = v m m1 qx + v m m1 px A(m) 1
1 1
x+ m :n− m
i
• UDD: A(m)1x:n = A1
i(m) x:n
1
and Āx:n = δi Ax:n
1
m−1 1
• Claims acceleration: A(m)1x:n = (1 + i) 2m 1
Ax:n 1
and Āx:n 1
= (1 + i) 2 Ax:n

4.1 Key Concepts 72
m-thly Endowment Insurance of 1 on (x):
• ( (m) 1 (m)
v Kx +m
for Kx = 0, m1 , ..., n - m1
Z= (m)
vn for Kx = n, n + m1 , ...
(m)
• E(Z) = Ax:n = A(m)1x:n + Ax:n1
(m)
• E(Z 2 ) = 2 Ax:n = 2 A(m)1x:n + 2 Ax:n1
(m) (m)
• V ar(Z) = 2 Ax:n - [Ax:n ]2
(m) 1 1 (m)
• Recursion: Ax:n = v m m1 qx + v m m1 px A 1 1
x+ m :n− m
(m) i
• UDD: Ax:n = A1
i(m) x:n
+ Ax:n1 and Āx:n = δi Ax:n
1
+ Ax:n1
• Claims acceleration:
(m) m−1
1 1
Ax:n = (1 + i) 2m Ax:n 1
+ Ax:n1 and Āx:n = (1 + i) 2 Ax:n + Ax:n1

4.1 Key Concepts 73
m-thly Deferred Whole Life Insurance of 1 on (x):
• (
(m)
0 for Kx = 0, m1 , ..., u - m1
Z= (m) 1 (m)
v Kx + m for Kx = u, u + m1 , ...
(m) P∞ k+1 (m) (m)

• E(Z) = u| Ax = k=mu v
m k 1 qx = Ax - A(m)1x:u = u Ex Ax+u
m|m
(m) P∞ 2 k+1 (m) (m)

• E(Z 2 ) = 2u| Ax = k=mu v
m k 1 qx = 2 Ax - 2 A(m)1x:u = (2 Ax:u1 )(2 Ax+u )
m|m
(m) (m)
• V ar(Z) = 2u| Ax - [u| Ax ]2
(m) 1 (m)
• u| Ax = 0 + v m m1 px (u− m1 | Ax+ 1 )
m
(m) i i
• UDD: u| Ax = A
i(m) u| x
and u| Āx = δ u| Ax
(m) m−1 1
• Claims acceleration: u| Ax = (1 + i) 2m
u| Ax and u| Āx = (1 + i) 2 u| Ax

4.1 Key Concepts 74
m-thly Deferred Term Insurance of 1 on (x):
• ( (m) 1 (m) 1 1
v Kx +m
for Kx = u, u + m, ..., u + n -
Z= m
0 otherwise
(m) Pm(u+n)−1 k+1

• E(Z) = u|n Ax = u| A(m)1x:n = v m k 1 qx = A(m)1x:u+n - A(m)1x:u
k=mu m|m
= u Ex A(m) x+u:n
1
(m) Pm(u+n)−1 k+1

• E(Z 2 ) = 2
u|n Ax = v2 m k 1 qx
k=mu m|m
= 2 A(m)1x:u+n - 2 A(m)1x:u = (2 Ax:u1 )(2 A(m) x+u:n

1
)
2 (m) (m)
• V ar(Z) = u|n Ax - [u|n Ax ]2
(m) 1 (m)
• u|n Ax = 0 + v m m1 px (u− m1 |n Ax+ 1 )
m
(m) i i
• UDD: u|n Ax = A
i(m) u|n x
and u|n Āx = δ u|n Ax
(m) m−1 1
• Claims acceleration: u|n Ax = (1 + i) 2m u|n Ax and u|n Āx = (1 + i) 2 u|n Ax

4.1 Key Concepts 75
Geometrically Increasing Life Insurance on (x):

Suppose the basic death benefit is 1. The death benefit is compounded at
a rate of j per year. There is an effective annual interest rate of i.
1+i
Define the modified interest rate: ib = 1+j - 1. Then:
• The expected present value of an annual whole life insurance, with Z =

1
(1 + j)Kx v Kx +1 , is 1+j Ax at ib .
• The expected present value of an annual term insurance, with:
½
(1 + j)Kx v Kx +1 for Kx = 0, 1, ..., n - 1
Z= ..
0 for Kx = n, n + 1, ...
1 1
is 1+j Ax:n at ib .

4.2 Exercises 76
4.2 Exercises
4.1. For a whole life insurance of 1000 on (30):
(i) The benefit is payable at the end of the month of death.

(ii) lx = 95 - x for 0 ≤ x ≤ 95, and δ = 0.07
Calculate the single benefit premium.
(A) 215 (B) 216 (C) 217 (D) 218 (E) 219
4.2. A fund is set up to provide benefits to 500 independent lives age 35:
(i) On January 1, 2012, each life is issued a single premium 15-year deferred
whole life insurance of 1000, payable at the end of the quarter of a year of
death.
(ii) Each life has mortality that follows the Illustrative Life Table, and i =
0.06
(iv) The claims acceleration approach is used to calculate quarterly insur-
ance expected present values.
Calculate the amount needed in the fund on January 1, 2012, so that the
probability, as determined by the normal approximation, is 0.95 that the fund
will be sufficient to provide these benefits.
(A) 52,000 (B) 52,500 (C) 53,000 (D) 53,500 (E) 54,000
4.3. Consider a life insurance on (45). 5000 is payable at the moment of

death if death occurs before age 55; 15,000 is payable at the moment of death
if death occurs between ages 55 and 65; 30,000 is payable at the moment of
death if death occurs after age 65.
Mortality follows the Illustrative Life Table, and i = 0.06. Deaths are
assumed to be uniformly distributed within each year of age.
(A) 4420 (B) 4430 (C) 4440 (D) 4450 (E) 4460

4.2 Exercises 77
4.4. Using the Illustrative Life Table, the claims acceleration approach, and
i = 0.06, calculate the actuarial present value of a 30-year deferred whole life
insurance of 10,000 on (35) payable at the end of the quarter of a year of
death.
(A) 615 (B) 620 (C) 625 (D) 630 (E) 635
4.5. Using the Illustrative Life Table, the uniform distribution of deaths
assumption over each year of age, and i = 0.06, calculate: 1000Ā50:15 .
(A) 449 (B) 451 (C) 454 (D) 457 (E) 459
4.6. A 20-year term insurance on a select life aged 25 payable at the end
of the year of death has a basic sum insured of 100,000. The insurer assumes
compound reversionary bonuses at the rate of 2% will vest at the end of each
policy year.
You are given:

(ii) i = 0.071
Calculate the expected present value of this insurance.
(A) 437 (B) 445 (C) 455 (D) 467 (E) 475
4.7. Suppose mortality follows the Standard Select Survival Model, and i
= 0.05.
Calculate the single benefit premium for a 1-year deferred, 2-year term
insurance of 10,000 on [60] that is payable at the end of the year of death.
(A) 66 (B) 69 (C) 72 (D) 76 (E) 79

4.2 Exercises 78
4.8. Consider a special 5-year deferred whole life insurance on (35) payable
at the end of the year of death. You are given:
(i) ½
0 for k = 0, 1, ..., 4
bk+1 =
5000(1.02)k for k = 5, 6, ...
(ii) Mortality follows de Moivre’s Law with limiting age 95.

(iii) i = 0.05
Calculate the expected present value of this insurance.
(A) 1905 (B) 1915 (C) 1925 (D) 1935 (E) 1945
4.9. Consider a life insurance on (40). 20,000 is payable at the end of the
quarter of a year of death if death occurs before age 55; 10,000 is payable at
end of the quarter of a year of death if death occurs between ages 55 and 65;
0 is payable if death occurs after age 65.
Mortality follows the Illustrative Life Table, and i = 0.06. Deaths are
assumed to be uniformly distributed within each year of age.
Calculate the expected present value.
(A) 1220 (B) 1230 (C) 1240 (D) 1250 (E) 1260
4.10. You are given a 20-year endowment insurance of 10,000 on a select

life aged 40 with death benefit payable at the end of the month of death. You
are given:

(ii) i = 0.05
(iii) Deaths are uniformly distributed over each year of age.
Calculate the standard deviation of the present value of the insurance

random variable.
(A) 350 (B) 360 (C) 370 (D) 380 (E) 390

4.2 Exercises 79
4.1. C
4.2. D
4.3. B
4.4. C
4.5. A
4.6. D
4.7. B
4.8. B
4.9. C
4.10. D


• Exam MLC, Sample Questions: #34, 231

81
5 ANNUITIES I
5.1 Key Concepts

A life annuity policy provides payments to an annuitant each period while
that person survives.
Annuities can be described as either (i) continuous: the payments are
made continuously each year while the annuitant survives or (ii) discrete:
the payments are made at the beginning or the end of each period while
the annuitant survives. If the payments are made at the beginning of each
period, the policy is an annuity-due; if the payments are made at the end of
each period, the policy is an annuity-immediate.
Types of Life Annuities:
• Whole Life Annuity: Provides payments each period while the annu-
itant survives.
• Temporary Life Annuity: Provides payments each period while the

annuitant survives for at most n years after policy issue. This is also
called a term annuity.
• Deferred Life Annuity: The annuitant must survive a u-year defer-

ral period after policy issue in order for any payments to be made. A
deferred whole life annuity provides payments each period while the an-
nuitant survives if the annuitant first survives the u-year deferral period;
that is, payments are made after u years while the annuitant survives.
A deferred temporary life annuity provides payments each period while
the annuitant survives for at most n years after first surviving the u-year
deferral period; that is, payments are made between u years and u + n
years after policy issue while the annuitant survives.
• Certain and Life Annuity: Is guaranteed to provide payments for

the first n years after policy issue, regardless of whether the annuitant
survives or dies within the n-year period. If the annuitant survives the
first n years, the annuity continues to provide payments each period for as

5.1 Key Concepts 82
long as the annuitant survives. This is also called a guaranteed annuity.
For each of these annuities, the following will be considered:
– The present value of the annuity, Y , are the payments discounted

for interest between policy issue and each payment date. This is a
random variable, as the number of payments is a function of the future
lifetime of the annuitant.
– The expected present value of the annuity, E(Y ), are the pay-
ments discounted for both interest and survival between policy issue
and each potential payment date. E(Y ) is with respect to the distri-
bution of the annuitant’s future lifetime.
E(Y ) will be written differently for each type of life annuity consid-
ered. In addition, E(Y ) can also be called the actuarial present value
of the annuity, the single premium, the net single premium, or the
single benefit premium.
– Secondary characteristics of the distribution of Y that will be of in-
terest include the variance of Y and percentiles of the distribution of
Y.

5.1 Key Concepts 83
Life Annuity Formulas

This section provides key formulas for different life annuities. Note:
• Life annuities have continuous and discrete versions. In the continuous

case, the payments are made continuously each year up until the moment
of death of the annuitant (Tx years after policy issue). The discrete case
can be : (i) annual or (ii) m-thly (Exam MLC only). In (i), the payments
are made at either the beginning of each year (for a total of Kx + 1
payments), or payments are made at the end of each year (for a total of
Kx payments).
It is possible that the number of payments made could be subject to a

finite term and/or a deferral period. It is also possible that there may be
a guaranteed number of payments.
• The general formula for the present value of a continuous life annuity on
(x) with payment rate πt at time t (> 0) is:
R Tx R Tx
Y = 0 πt v t dt = 0 πt e−δt dt.
The general formula for the expected present value of a continuous life
annuity on (x) with payment rate πt at time t (> 0) is:
R∞ R∞ R∞
E(Y ) = 0 πt v t t px dt = 0 πt e−δt t px dt = 0 πt (t Ex )dt. Furthermore:
“Say the payment rate is πt at time t. Then the present value of this
benefit at time t is πt v t dt, and the expected present value of this benefit
is πt v t t px dt ((x) has to survive to time t in order for πt dt to be made
at that time). Integrating over all possible payment times provides the
overall expected present value. This is the current payment approach.”

5.1 Key Concepts 84
• The general formula for the present value of an annual life annuity-due
that pays πk at time k (k = 0, 1, 2, ...) is:
P Kx k
Y = k=0 πk v .
The general formula for the expected present value of an annual life
annuity-due that pays πk at time k (k = 0, 1, 2, ...) is:
P∞ k
P∞
E(Y ) = k=0 πk v k px = k=0 πk (k Ex ). Furthermore:
“Say the payment is πk at time k. Then the present value of this benefit is
πk v k , and the expected present value of this benefit is πk v k k px . Summing
over all possible payment times provides the overall expected present
value. This is the current payment approach.”
Note: The general formulas for the present value and expected present
value of an annual life annuity-immediate on (x) are similar, expect there
would be no payment at k = 0.

5.1 Key Concepts 85
• The International Actuarial Notation for E(Y ) often contains an a, which

indicates that the expected present value is for a life annuity. For exam-
ple, äx denotes the expected present value of a whole life annuity-due of
1 per year on (x).
• 2 a is a evaluated at double the force of interest.
• 2 a is NOT equal to E(Y 2 ). This implies that V ar(Y ) is NOT equal to

2
a - a2 (which will usually be negative).
• For each annuity, a (total) payment rate of 1 per year is assumed. If the
payment rate is R, notation and formulas are adjusted. For example, the
expected present value ofRa continuous whole life annuity of R per year
∞
on (x) is E(Y ) = Rāx = 0 R(t Ex ).
• Recursion formulas take an expected present value and decomposes it

into the sum of two expected present values: the expected present value
of the annuity during the first period plus the expected present value of
the annuity if the annuitant survives the first period.
For example: äx = 1 + vpx äx+1 . “The right-hand side breaks up the
expected present value of a whole life annuity-due of 1 per year on (x)
into the very first payment of 1 at issue plus the expected present value
of all of the remaining payments of 1 at issue.”
• For many annuities, V ar(Y ) is too difficult to calculate. Therefore,

V ar(Y ) is omitted for several of the following annuities.
• The actuarial accumulated value is the expected present value of a pay-

ment or payments divided by a discount factor. For example:
1
s̈x:n = n Ex
äx:n
= the actuarial accumulated value at time n of an n-year temporary life
annuity-due of 1 per year on (x).

5.1 Key Concepts 86
Level Annuities
Whole Life Annuity of 1 per Year on (x):

Continuous Whole Life Annuity:
1−v Tx
• Y = āTx = δ
1−Z
So: Y = δ , where Z is the present value random variable for a contin-
uous whole life insurance of 1 on (x).
R∞ 1−Āx
• E(Y ) = āx = 0 t Ex dt = δ
1
– With a constant force of mortality: āx = µ+δ
R∞
• It is also true that: āx = 0 (āt )t px µx+t dt
“Say the moment of death of (x) occurs at time t. Then the present value
of the payments is āt , and the expected present value of the payments
is (āt )t px µx+t dt ((x) has to survive t years and then immediately die for
the present value to be āt ). Integrating over all times of death provides
the overall expected present value.”
2
Āx −[Āx ]2
• V ar(Y ) = δ2
• FY (y) = P r[Y ≤ y] = P r[āTx ≤ y] = c qx where c = - ln(1−δy)

δ
– For de Moivre’s Law (Uniform Distribution):

FY (y) = - ln(1−δy)
δ(ω−x) for 0 ≤ y ≤ āω−x . If the annual payment is R,
replace y with Ry .
µ
– Constant force of mortality: FY (y) = 1 - (1 − δy) δ for 0 ≤ y ≤ 1δ . If
the annual payment is R, replace y with Ry .
Note: The 100α-th percentile of the distribution of Y , yα , solves:

FY (yα ) = α for 0 ≤ α ≤ 1.

5.1 Key Concepts 87
Annual Whole Life Annuity:

Whole Life Annuity-Due:
1−v Kx +1
• Yd = äKx +1 = d
So: Yd = 1−Zd , where Z is the present value random variable for an annual
whole life insurance of 1 on (x).
P∞ 1−Ax
• E(Yd ) = äx = k=0 k Ex = d
1+i
– With a constant force of mortality: äx = q+i
P∞
• It is also true that: äx = k=0 (äk+1 )k| qx
2
Ax −[Ax ]2
• V ar(Yd ) = d2
• Recursion: äx = 1 + vpx äx+1
Whole Life Annuity-Immediate:

1−v Kx
• Yi = aKx = i
So: Yi = 1−(1+i)Z
i , where Z is the present value random variable for an
annual whole life insurance of 1 on (x).
P∞
• E(Yi ) = ax = k=1 k Ex = äx - 1
1−(1+i)Ax
• E(Yi ) = ax = i
2
Ax −[Ax ]2
• V ar(Yi ) = d2 (same as an annual whole life annuity-due)
• Recursion: ax = vpx (1 + ax+1 )

5.1 Key Concepts 88
Temporary Life Annuity of 1 per Year on (x):

Continuous Temporary Life Annuity:
• ½
āTx for Tx ≤ n
Y =
ān for Tx > n
1−Z
= δ ,
where Z is the present value random variable for a continuous n-year
endowment insurance of 1 on (x).
Rn 1−Āx:n
• E(Y ) = āx:n = 0 t Ex dt = δ
1
– With a constant force of mortality: āx:n = µ+δ [1 − exp[−(µ + δ)n]]
2
Āx:n −[Āx:n ]2
• V ar(Y ) = δ2

5.1 Key Concepts 89
Annual Temporary Life Annuity:

Temporary Life Annuity-Due:
• ½
äKx +1 for Kx = 0, 1, ..., n - 1
Yd =
än for Kx = n, n + 1, ...
1−Z
= d ,
where Z is the present value random variable for an annual n-year en-
dowment insurance of 1 on (x).
Pn−1 1−Ax:n
• E(Yd ) = äx:n = k=0 k Ex = d
2
Ax:n −[Ax:n ]2
• V ar(Yd ) = d2
• Recursion: äx:n = 1 + vpx äx+1:n−1
Temporary Life Annuity-Immediate:
• ½
aKx for Kx = 0, 1, ..., n - 1
Yi =
an for Kx = n, n + 1, ...
Pn
• E(Yi ) = ax:n = k=1 k Ex = äx:n - 1 + n Ex
• Note: äx:n = 1 + ax:n−1

2
Ax:n+1 −[Ax:n+1 ]2
• V ar(Yi ) = d2
• Recursion: ax:n = vpx (1 + ax+1:n−1 )

5.1 Key Concepts 90
Deferred Whole Life Annuity of 1 per Year on (x):

Continuous Deferred Whole Life Annuity:
• ½
0 for Tx ≤ u
Y =
v u āTx −u for Tx > u
R∞
• E(Y ) = u| āx = u t Ex = āx - āx:u = u Ex āx+u
Annual Deferred Whole Life Annuity:

Deferred Whole Life Annuity-Due:
• ½
0 for Kx = 0, 1, ..., u - 1
Yd =
v u äKx +1−u for Kx = u, u + 1, ...
P∞
• E(Yd ) = u| äx = k=u k Ex = äx - äx:u = u Ex äx+u
• Recursion: u| äx = 0 + vpx (u−1| äx+1 )
Deferred Whole Life Annuity-Immediate:
• ½
0 for Kx = 0, 1, ..., u - 1
Yi =
v u aKx −u for Kx = u, u + 1, ...
P∞
• E(Yi ) = u| ax = k=u+1 k Ex = ax - ax:u = u Ex ax+u
• Note: u| äx = u| ax + u Ex
• Recursion: u| ax = 0 + vpx (u−1| ax+1 )

5.1 Key Concepts 91
Deferred Temporary Life Annuity of 1 per Year on (x):
Continuous Deferred Temporary Life Annuity:
• 
0 for Tx ≤ u
Y = v u āTx −u for u < Tx ≤ u + n
 u
v ān for Tx > u + n
R u+n
• E(Y ) = u|n āx = u| āx:n = u t Ex dt = āx:u+n - āx:u = u Ex āx+u:n
Annual Deferred Temporary Life Annuity:

Deferred Temporary Life Annuity-Due:
• 
0 for Kx = 0, 1, ..., u - 1
Yd = v u äKx +1−u for Kx = u, u + 1, ..., u + n - 1
 u
v än for Kx = u + n, u + n + 1, ...
Pu+n−1
• E(Yd ) = u|n äx = u| äx:n = k=u k Ex = äx:u+n - äx:u = u Ex äx+u:n
• Recursion: u|n äx = 0 + vpx (u−1|n äx+1 )
Deferred Temporary Life Annuity-Immediate:
• 
0 for Kx = 0, 1, ..., u - 1
Yi = v u aKx −u for Kx = u, u + 1, ..., u + n - 1
 u
v an for Kx = u + n, u + n + 1, ...
Pu+n
• E(Yi ) = u|n ax = u| ax:n = k=u+1 k Ex = ax:u+n - ax:u = u Ex ax+u:n
• Recursion: u|n ax = 0 + vpx (u−1|n ax+1 )

5.1 Key Concepts 92
Certain and Life Annuity of 1 per Year on (x):

Continuous Certain and Life Annuity:
• ½
ān for Tx ≤ n
Y =
āTx for Tx > n
• E(Y ) = āx:n = ān + n| āx
Annual Certain and Life Annuity:

Certain and Life Annuity-Due:
• ½
än for Kx = 0, 1, ..., n - 1
Yd =
äKx +1 for Kx = n, n + 1, ...
• E(Yd ) = äx:n = än + n| äx
• Recursion: äx:n = 1 + vqx än−1 + vpx äx+1:n−1
Certain and Life Annuity-Immediate:
• ½
an for Kx = 0, 1, ..., n - 1
Yi =
aKx for Kx = n, n + 1, ...
• E(Yi ) = ax:n = an + n| ax
• Recursion: ax:n = v + vqx an−1 + vpx ax+1:n−1

5.2 Exercises 93
5.2 Exercises
5.1. Assume: µx (t) = 0.02 for t > 0 and δ = 0.05.
Calculate: āx:15 .
(A) 8.70 (B) 8.90 (C) 9.10 (D) 9.30 (E) 9.50
5.2. Assume mortality follows: lx = 100(110 - x) for 0 ≤ x ≤ 110, and d

= 0.05.
Calculate: 15| ä45 .
(A) 4.40 (B) 4.60 (C) 4.80 (D) 5.00 (E) 5.20
5.3. You are given:
(i) µ = δ = c, where c is a positive constant.

25
(ii) 2 āx = 3
Calculate: V ar(āT (x) ).
(A) 50 (B) 52 (C) 54 (D) 56 (E) 58
5.4. You are given:
(i) k| q35 = 0.005(k + 1) for k = 0, 1, 2, 3.

(ii) i = 0.05
Calculate the actuarial present value of a 4-year temporary life annuity-due

of 100 per year on (35).
(A) 350 (B) 360 (C) 370 (D) 380 (E) 390

5.2 Exercises 94
5.5. You are given:
(i) δ = 0.05
(ii) ½
0.05 for 0 ≤ x < 50
µx =
0.08 for x ≥ 50
Calculate the actuarial present value of a continuous whole life annuity of

1 per year on (30).
(A) 9.10 (B) 9.30 (C) 9.50 (D) 9.70 (E) 9.90
5.6. You are given:
½
0.05 for 0 ≤ t < 10
δt =
0.07 for t ≥ 10
½
0.02 for 0 ≤ t < 10
µx (t) =
0.03 for t ≥ 10
Calculate the expected present value of a continuous 10-year certain and

life annuity on (x) of 1 per year.
(A) 11.60 (B) 11.90 (C) 12.20 (D) 12.50 (E) 12.80
5.7. You are given:
(i) µx (t) = 0.02 for t > 0

(ii) δ = 0.06
Calculate: Pr[āTx > āx ].
(A) 0.42 (B) 0.55 (C) 0.63 (D) 0.84 (E) 0.91

5.2 Exercises 95
5.8. You are given:
(i) The force of mortality is constant.

(ii) Āx = 0.428571
1
(iii) Āx:10 = 0.215749
(iv) Y is the present value random variable for a continuous 10-year tem-
porary life annuity of 500 per year on (x).
Calculate: E(Y ).
(A) 3550 (B) 3600 (C) 3650 (D) 3700 (E) 3750
5.9. You are given:
(i) Mortality for a standard life aged 40, denoted as S, is such that:
qxS = 0.032 for x = 40, 41, 42, ...
(ii) Mortality for a certain life aged 40, denoted as C, is such that:
C
q40 = 0.048 and qxC = 0.032 for x = 41, 42, 43, ...
(iii) d = 0.05
Calculate: äS40 - äC40 .
(A) 0.16 (B) 0.17 (C) 0.18 (D) 0.19 (E) 0.20
(i) µx (t) is the force of mortality associated with the Illustrative Life Table.
(ii) i = 0.05
Calculate the single benefit premium for a 3-year temporary life annuity-
immediate of 1000 per year on (30) payable annually, assuming that the force
of mortality used is equal to µ30 (t) + 0.20 for 0 ≤ t ≤ 3.
(A) 1860 (B) 1900 (C) 1940 (D) 1980 (E) 2020

5.2 Exercises 96
5.11. A fund is created such that:
(i) There are 60 lives each age 30.

(ii) Each life receives payments of 100 per year for life, payable annually,
beginning immediately.
(iv) The lifetimes are independent.
(v) i = 0.06
(vi) The amount of the fund is determined, using the normal approxima-
tion, such that the probability that the fund is sufficient to make all payments
is 99%.
Calculate the initial amount of the fund.
(A) 98,000 (B) 98,500 (C) 99,000 (D) 99,500 (E) 100,000
5.12. For a group of individuals all age x:
(i) 35% are smokers and 65% are non-smokers.

(ii) The constant force of mortality for smokers is 0.08.
(iii) The constant force of mortality for non-smokers is 0.04.
(iv) δ = 0.06
Calculate V ar[āT (x) ] for an individual chosen at random from this group.
(A) 25 (B) 26 (C) 27 (D) 28 (E) 29
5.13. Suppose Z is the present value random variable for a 2-year pure
endowment of 1 on (x). You are given:
(i) v = 0.95 and px = 0.98

(ii) Ax = 0.45 and äx+2 = 10.68
Calculate: V ar(Z).
(A) 0.045 (B) 0.050 (C) 0.055 (D) 0.060 (E) 0.065

5.2 Exercises 97
5.14. Cody, age 25, and Ted, age 30, have each won the actuarial lottery:
(i) Cody has decided to collect his winnings via a 20-year temporary life
annuity-due, which pays 400,000 each year.
(ii) Ted has decided to collect his winnings via a 20-year certain and life
annuity-due, which pays K each year.
(iii) Mortality for both Cody and Ted follows the Illustrative Life Table,
and i = 0.06.
The expected present values of Cody’s annuity and Ted’s annuity are both
equal. Calculate: K.
(A) 281,000 (B) 286,000 (C) 291,000 (D) 295,000 (E) 299,000
5.15. Consider a special whole life annuity on (x) which pays R at the
beginning of the first year, 2R at the beginning of the second year, and 3R
at the beginning of each year thereafter. You are also given:
(i) The actuarial present value of this annuity is 3333.

(ii) i = 0.05
(iii) px = 0.98 and px+1 = 0.97
(iv) äx+2 = 31.105
Calculate: R.
(A) 30 (B) 35 (C) 40 (D) 45 (E) 50

5.2 Exercises 98
5.16. You are given the following portfolio of mutually independent lives:
(i) 50 lives age 65 purchase a whole life annuity-immediate with an annual

payment of 30,000.
(ii) 20 lives age 75 purchase a whole life annuity-immediate with an annual
payment of 20,000.
Mortality follows the Illustrative Life Table, and i = 0.06.

Let S be the present value for the total payments on the portfolio.
Calculate the 95th percentile of the distribution of S, in millions, using the
normal approximation.
(A) 16.9 (B) 17.2 (C) 17.5 (D) 17.8 (E) 18.1

1
(i) µx = 90−x for 0 ≤ x < 90
(ii) i = 0
Calculate: ä30 .
(A) 29.0 (B) 29.5 (C) 30.0 (D) 30.5 (E) 31.0
5.18. Paul, aged 35, has just taken out a home mortgage loan where he
will pay 12,000 at the end of each year for 25 years.
Paul was also required to purchase a life insurance policy that will pay any
remaining payments should he die within the 25-year period.
Paul has mortality that follows the Illustrative Life Table. The effective
annual interest rate is 6%.
Calculate the expected present value of the life insurance policy.
(A) 5150 (B) 5250 (C) 5350 (D) 5450 (E) 5550

5.2 Exercises 99
(i) Mortality follows:
x 30 31 32 33 34 35 36 37
qx 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80
(ii) d = 0.03
Calculate the probability that the present value of a 5-year temporary life
annuity-due of 500 per year on (30) exceeds its actuarial present value.
(A) 0.45 (B) 0.50 (C) 0.55 (D) 0.60 (E) 0.65
(i) δ = 0.04
(ii) µx = 0.0003(1.05)x for x ≥ 0
Calculate the expected present value of a 2-year deferred 2-year temporary

life annuity-immediate of 100 per year on (34).
(A) 167 (B) 169 (C) 171 (D) 173 (E) 175
5.21. For a 10-year deferred 10-year continuous temporary life annuity of

1000 per year on (x):
(i) ½
0.06 for t ≤ 6
δt =
0.07 for t > 6
(ii) ½
0.025 for t ≤ 6
µx (t) =
0.035 for t > 6
Calculate the single benefit premium for this annuity.
(A) 2320 (B) 2360 (C) 2400 (D) 2440 (E) 2480

5.2 Exercises 100
5.22. Consider a policy on (40) that provides the following benefits:
(i) A whole life annuity-due of 500 per year payable annually.

(ii) A death benefit of 5000 payable at the end of the year of death.
Furthermore:
(iii) i = 0.06
(iv) Mortality follows the Illustrative Life Table.
Calculate the standard deviation of the present value random variable for
this policy.
(A) 515 (B) 530 (C) 545 (D) 560 (E) 575

(i)
x 70 71 72 73 74 75
lx 50 45 40 35 30 25
(ii) i = 0.05
Calculate the actuarial accumulated value at the end of the fifth year of a
5-year temporary life annuity-immediate of 100 per year payable annually on
(70).
(A) 310 (B) 425 (C) 540 (D) 665 (E) 785

5.2 Exercises 101
5.24. Consider a whole life annuity-due of 1 per year payable annually on

(x):
(i) v = 0.965
(ii) 10 px = 0.920
(iii) äx+11 = 11.36
Suppose qx+10 is increased to qx+10 + 0.100.

Calculate the change in the expected present value of the annuity.
(A) - 0.71 (B) - 0.41 (C) - 0.26 (D) 0.26 (E) 0.58
5.25. Consider the following special annuity on (x) payable annually:
(i) The payment at time k, πk , is such that:

½
1000vqx+k for k = 0, 1, ..., 19
πk =
0 for k = 20, 21, ...
(ii) i = 0.045
(iii) 20 px = 0.945
(iv) ax = 18.23
(v) ax+20 = 13.94
Calculate the expected present value of this annuity.
(A) 30 (B) 32 (C) 34 (D) 36 (E) 38

5.2 Exercises 102
5.1. D
5.2. B
5.3. B
5.4. C
5.5. D
5.6. E
5.7. C
5.8. B
5.9. D
5.10. A
5.11. C
5.12. A
5.13. A
5.14. E
5.15. C
5.16. B
5.17. D
5.18. C
5.19. B
5.20. D
5.21. D
5.22. E
5.23. E
5.24. A
5.25. B


• Exam MLC, Fall 2013: #1, 5
• Exam 3L, Fall 2013: #12, 13
• Exam 3L, Fall 2012: #12
113, 114, 126, 130, 140, 146, 166, 186, 192, 196, 209, 210, 229, 285
• Exam 3L, Fall 2011: #12
• Exam 3L, Fall 2009: #12
• Exam 3L, Spring 2009: #12, 13
• Exam 3L, Fall 2008: #20, 21
• Exam MLC, Spring 2007: #2, 17, 24, 29

104
6 ANNUITIES II (Exam MLC Only)
6.1 Key Concepts

m-thly Annuities-Due
Here, we consider a discrete life annuity where each payment is provided

at the beginning of the m-th of a year, conditional on survival. The value of
m is typically equal to 2 (half-year), 4 (quarter of a year), or 12 (month).
Note:
• An m-thly life annuity-due is such that each payment is made at the

(m)
beginning of each m-th of a year (for a total of m(Kx + m1 ) payments,
in years).
• The general exact formula for the expected present value of an m-thly
life annuity-due on (x) that pays π k at time mk years (k = 0, 1, 2, ...) is:
m
P∞ k P∞
E(Yd ) = k=0 π m
k v m k px = k=0 π m
k ( k Ex ). Furthermore:
m m
k
“Say the payment is π k at time m. Then the present value of this benefit
m
k k
is π k v , and the expected present value of this benefit is π k v m k px ((x)
m
m m m
k
has to survive to time m in order for π k to be made at that time).
m
Summing over all possible payment times provides the overall expected
present value.”
• The International Actuarial Notation for E(Yd ) often contains an ä(m) ,

which indicates that the expected present value is for an m-thly life
(m)
annuity-due. For example, äx denotes the expected present value of
a whole life annuity-due of 1 per year on (x), payable in equal install-
ments of m1 at the beginning of each m-th of the year.
• 2 ä(m) is ä(m) evaluated at double the force of interest.

6.1 Key Concepts 105
• Often, we do not use the exact formulas to calculate expected present

values for an m-thly life annuity-due. Rather, we approximate these
expected present values from the corresponding annual life annuity-due
expected present values using one of two assumptions:
– UDD: deaths are uniformly distributed within each year of age. In

UDD formulas that approximate m-thly life annuities-due, we use the
following functions:
id i−i(m)
α(m) = i(m) d(m)
and β(m) = i(m) d(m)
Let m approach infinity. In UDD formulas that approximate contin-

uous life annuities, we use the following functions:
id i−δ
α(∞) = δ2 and β(∞) = δ2 .
You are provided a table of α(m) and β(m) for various values of m
at i = 0.06 during Exam MLC. Please refer to the web link to Exam
MLC tables provided in Appendix A of this study supplement.
– Woolhouse’s Formula: based on series expansions. For example, Wool-

house’s Formula with three terms for an m-thly whole life annuity-due
of 1 per year on (x) is:
(m) m−1 m2 −1
äx = äx - 2m - 12m2 (δ + µx ), and
Woolhouse’s Formula with two terms for an m-thly whole life annuity-
due of 1 per year on (x) is:
(m) m−1
äx = äx - 2m (this approximates the UDD formula).
Furthermore, µx can be approximated as - 12 (ln px−1 + ln px ).
• Note: The formulas discussed below reduce to the corresponding annual

life annuity-due formulas in Annuities I when m = 1.

• Note: While unlikely, it is possible that m-thly life annuities-immediate

could be tested; each payment is made at the end of each m-th of a year
(m)
(for a total of m(Kx ) payments, in years).
You can use the following relations:
(m) (m) 1
– ax = äx - m
(m) (m) 1
– ax:n = äx:n - m (1 - n Ex )
(m) (m) (m)
– n| ax = ax - ax:n
(m) (m) (m)
– ax:n = an + n| ax

m-thly Whole Life Annuity-Due of 1 per Year on (x):
(m) 1
(m) 1−v Kx + m
• Yd = ä (m) 1
= d(m)
Kx +m
So: Yd = 1−Z
d(m)
, where Z is the present value random variable for an m-thly
whole life insurance of 1 on (x).
(m) P∞ 1 1−Ax
(m)
• E(Yd ) = äx = k Ex
k=0 m m = d(m)
2 (m) (m)
Ax −[Ax ]2
• V ar(Yd ) = [d(m) ]2
(m) 1 1 (m)
• Recursion: äx = m + v m m1 px äx+ 1
m
(m)
• UDD: äx = α(m)äx - β(m); āx = α(∞)äx - β(∞)
(m) m−1 m2 −1
• Woolhouse’s Formula with 3 terms: äx = äx - 2m - 12m2 (δ + µx );
1 1
āx = äx - 2 - 12 (δ + µx )

m-thly Temporary Life Annuity-Due of 1 per Year on

(x):
• 
 ä(m) (m)
for Kx = 0, 1 1
(m) 1 m, ..., n - m
Yd = Kx +m
 ä(m) (m)
for Kx = n, n + 1
n m, ...
1−Z
= d(m)
where Z is the present value random variable for an m-thly n-year en-
dowment insurance of 1 on (x).
(m) Pmn−1 1
(m)
1−Ax:n
• E(Yd ) = äx:n = k=0 m m Ex
k = d(m)
2 (m) (m)
Ax:n −[Ax:n ]2
• V ar(Yd ) = [d(m) ]2
(m) 1 1 (m)
• Recursion: äx:n = m + v m m1 px ä 1 1
x+ m :n− m
(m)
• UDD: äx:n = α(m)äx:n - β(m)(1 − n Ex );
āx:n = α(∞)äx:n - β(∞)(1 − n Ex )
• Woolhouse’s Formula with 3 terms:

(m) m−1 m2 −1
äx:n = äx:n - 2m (1 − n Ex ) - 12m2 (δ + µx − n Ex (δ + µx+n ));
āx:n = äx:n - 12 (1 − n Ex ) - 1
12 (δ + µx − n Ex (δ + µx+n ))
I would not memorize these UDD and Woolhouse’s formulas. Just know
the m-thly whole life results and use:
(m) (m) (m)
äx:n = äx - n Ex äx+n .

m-thly Deferred Whole Life Annuity-Due of 1 per

Year on (x):
• 
0 (m)
for Kx = 0, m1 , ..., u - m1
Yd = (m) (m)
 v u ä (m) 1
for Kx = u, u + m1 , ...
Kx +m −u
(m) P∞ 1 (m) (m) (m)

• E(Yd ) = u| äx = k Ex
k=mu m m = äx - äx:u = u Ex äx+u
(m) 1 (m)
• Recursion: u| äx = 0 + v m m1 px (u− m1 | äx+ 1 )
m
(m)
• UDD: u| äx = α(m)u| äx - β(m)u Ex ; u| āx = α(∞)u| äx - β(∞)u Ex
(m) m−1 m2 −1
• Woolhouse’s Formula with 3 terms: u| äx = u| äx - 2m u Ex - 12m2 u Ex (δ +
µx+u );
1 1
u| āx = u| äx - 2 u Ex - 12 u Ex (δ + µx+u )
I would not memorize these UDD and Woolhouse’s formulas. Just know
(m) (m)
the m-thly whole life results and use: n| äx = n Ex äx+n .
m-thly Deferred Temporary Life Annuity-Due of 1 per

Year on (x):
• 
(m)
 for Kx = 0, m1 , ..., u - m1
 0 (m)

(m)
Yd = v u ä (m) 1 for Kx = u, u + m1 , ..., u + n - 1
m

 Kx + m −u
 u (m) (m) 1
v än for Kx = u + n, u + n + m, ...
(m) Pm(u+n)−1 1 (m) (m) (m)

• E(Yd ) = u|n äx = k=mu m m Ex
k = äx:u+n - äx:u = u Ex äx+u:n
(m) 1 (m)
• Recursion: u|n äx = 0 + v m m1 px (u− m1 |n äx+ 1 )
m

m-thly Certain and Life Annuity-Due of 1 per Year on

(x):
• 
 ä(m) (m)
for Kx = 0, m1 , ..., n - m1
n
Yd = (m) (m)
 ä (m) 1
for Kx = n, n + m1 , ...
Kx +m
(m) (m) (m)

• E(Yd ) = äx:n = än + n| äx
(m) 1 1 (m) 1 (m)

• Recursion: äx:n = m + v m m1 qx ä 1
+ v m m1 px ä
n− m 1
x+ m 1
:n− m

VARYING ANNUITIES
• Annually Increasing Annuity: A life annuity where the annual pay-

ment is increased arithmetically for each year the annuitant survives. For
example, an annually increasing temporary life annuity-due pays R at the
beginning of the first year, 2R at the beginning of the second year given
survival of the annuitant,..., nR at the beginning of year n given survival
of the annuitant.
• Annually Decreasing Annuity: A life annuity where the annual pay-

ment is decreased arithmetically for each year the annuitant survives. For
example, an annually decreasing temporary life annuity-due pays nR at
the beginning of the first year, (n − 1)R at the beginning of the second
year given survival of the annuitant,..., R at the beginning of year n given
survival of the annuitant.
• Geometrically Increasing Annuity: A life annuity where the annual

payment is increased at a compound rate of j per year for each year the
annuitant survives. For example, a geometrically increasing whole life
annuity-due pays R at the beginning of the first year, R(1 + j) at the
beginning of the second year given survival of the annuitant, R(1 + j)2
at the beginning of the third year given survival of the annuitant, etc.

Annually Increasing Whole Life Annuity on (x):

Continuous Annually Increasing Whole Life Annuity:
• Provides 1 continuously during the first year, 2 continuously during the

second year, 3 continuously during the third year, etc.
P
• E(Y ) = (Iā)x = ∞ k=0 (k + 1)k| āx:1
Continuously Increasing Whole Life Annuity:
• The payment rate at time t is t.

R
¯ x = ∞ tt Ex dt
• E(Y ) = (Iā) 0
¯ x=
– With a constant force of mortality: (Iā) 1
(µ+δ)2 .
Annually Increasing Whole Life Annuity-Due:
• Provides 1 at the beginning of the first year, 2 at the beginning of the

second year, 3 at the beginning of the third year, etc.
P
• E(Yd ) = (Iä)x = ∞ k=0 (k + 1)k Ex
(1+i)2
– With a constant force of mortality: (Iä)x = (q+i)2
• Recursion: (Iä)x = äx + vpx (Iä)x+1
Annually Increasing Whole Life Annuity-Immediate:
• Provides 1 at the end of the first year, 2 at the end of the second year, 3
at the end of the third year, etc.
P
• E(Yi ) = (Ia)x = ∞ k=1 k k Ex
• Recursion: (Ia)x = ax + vpx (Ia)x+1

Annually Increasing Temporary Life Annuity on (x):

Continuous Annually Increasing Temporary Life Annuity:
• Provides 1 continuously during the first year, 2 continuously during the

second year, ..., n continuously during year n.
P
• E(Y ) = (Iā)x:n = n−1 k=0 (k + 1)k| āx:1
Continuously Increasing Temporary Life Annuity:
• The payment rate at time t is t during the first n years.

R
¯ x:n = n tt Ex dt
• E(Y ) = (Iā) 0
Annually Increasing Temporary Life Annuity-Due:
• Provides 1 at the beginning of the first year, 2 at the beginning of the

second year, ..., n at the beginning of year n.
P
• E(Yd ) = (Iä)x:n = n−1 k=0 (k + 1)k Ex
• Recursion: (Iä)x:n = äx:n + vpx (Iä)x+1:n−1
Annually Increasing Temporary Life Annuity-Immediate:
• Provides 1 at the end of the first year, 2 at the end of the second year,
..., n at the end of year n.
P
• E(Yi ) = (Ia)x:n = nk=1 k k Ex
• Recursion: (Ia)x:n = ax:n + vpx (Ia)x+1:n−1

Annually Decreasing Temporary Life Annuity on (x):
Continuous Annually Decreasing Temporary Life Annuity:
• Provides n continuously during the first year, n - 1 continuously during

the second year, ..., 1 continuously during year n.
P
• E(Y ) = (Dā)x:n = n−1 k=0 (n − k)k| āx:1
Also: (Iā)x:n + (Dā)x:n = (n + 1)āx:n .
Continuously Decreasing Temporary Life Annuity:
• The payment rate at time t is n - t during the first n years.

Rn
• E(Y ) = (D̄ā)x:n = 0 (n − t)t Ex dt
¯ x:n + (D̄ā)x:n = nāx:n .
Also: (Iā)
Annually Decreasing Temporary Life Annuity-Due:
• Provides n at the beginning of the first year, n - 1 at the beginning of

the second year, ..., 1 at the beginning of year n.
P
• E(Yd ) = (Dä)x:n = n−1 k=0 (n − k)k Ex
Also: (Iä)x:n + (Dä)x:n = (n + 1)äx:n .
• Recursion: (Dä)x:n = n + vpx (Dä)x+1:n−1
Annually Decreasing Temporary Life Annuity-Immediate:
• Provides n at the end of the first year, n - 1 at the end of the second
year, ..., 1 at the end of year n.
P
• E(Yi ) = (Da)x:n = nk=1 (n − k + 1)k Ex
Also: (Ia)x:n + (Da)x:n = (n + 1)ax:n .
• Recursion: (Da)x:n = nvpx + vpx (Da)x+1:n−1

Geometrically Increasing Life Annuity on (x):
• Consider an annual whole life annuity-due where the payment at time 0

is 1, the payment at time 1 is (1 + j), the payment at time 2 is (1 + j)2 ,
etc. There is an effective annual interest rate of i.
1+i
Define the modified interest rate: iπ = 1+j - 1. Then:
The expected present value of the above annual whole life annuity-due is
äx at iπ .
• Consider an annual n-year temporary life annuity-due where the payment

at time 0 is 1, the payment at time 1 is (1 + j), the payment at time 2 is
(1 + j)2 , ..., the payment at time n - 1 is (1 + j)n−1 . There is an effective
annual interest rate of i.
1+i
Define the modified interest rate: iπ = 1+j - 1. Then:
The expected present value of the above annual n-year temporary life
annuity-due is äx:n at iπ .

6.2 Exercises 116
6.2 Exercises
6.1. You are given:
(i) Mortality follows the Illustrative Life Table.

(iii) i = 0.06
(12)
Calculate: ä35:20 .
(A) 11.0 (B) 11.2 (C) 11.5 (D) 11.8 (E) 12.0
6.2. You are given:
(i) Mortality follows a select-and-ultimate table, 3-year select period.

(ii) ä[40]+1 = 19.2297
(iii) i = 0.045
(iv) p[40] = 0.9987
(v) µ[40]+1 = 0.001321
(4)
Using Woolhouse’s formula with three terms, calculate: 1| ä[40] .
(A) 18.0 (B) 18.1 (C) 18.2 (D) 18.3 (E) 18.4
6.3. Consider a special 20-year temporary life annuity-due on (30) with

annual payments:
(i) The payment for the beginning of year (k + 1) is: πk = (1.04)k for k =
0, 1, 2, ..., 19.
(ii) i = 0.06
(iii) lx = 100 - x for 0 ≤ x ≤ 100
(A) 14.6 (B) 14.7 (C) 14.8 (D) 14.9 (E) 15.0

6.2 Exercises 117
6.4. You are given:
(i) l62 = 8,982,404, l63 = 8,915,575, and l64 = 8,842,735

(ii) v = 0.9569
(iii) (Iä)62 = 158.94
(iv) (Iä)64 = 145.55
Calculate: ä63 .
(A) 13.2 (B) 13.4 (C) 13.6 (D) 13.8 (E) 14.0
6.5. A fund is established to provide annuity benefits to 500 independent

lives all age 35.
You are given:
(i) On January 1, 2012, each life is issued a single premium whole life
annuity. The total payment for each year is 12,000, which is payable in equal
monthly installments in advance.
(ii) Each life has mortality that follows the Illustrative Life Table.
(iii) i = 0.06
(iv) Deaths are uniformly distributed within each year of age.
Calculate the amount needed in the fund on January 1, 2012, in millions,

so that the probability, as determined by the normal approximation, is 0.99
that the fund will be sufficient to provide these benefits.
(A) 90.0 (B) 90.5 (C) 91.0 (D) 91.5 (E) 92.0

6.2 Exercises 118
6.6. Consider a special life annuity issued to Jenn, aged 37:
(i) There is a deferral period of 10 years. If Jenn dies during the deferral
period, 80% of the net single premium is refunded without interest at the
end of the year of death.
(ii) During the 15-year period starting at the end of the deferral period,
1000 is payable at the beginning of each month while Jenn is alive. If Jenn is
still alive 25 years after issue, 3000 is payable at the beginning of each month
for life.
(iv) Deaths are uniformly distributed over each year of age.
(v) i = 0.06
Calculate the net single premium.
(A) 135,500 (B) 136,100 (C) 136,700 (D) 137,300 (E) 137,900
6.7. Consider a 15-year certain and life annuity-due of 24,000 per year on
(65) payable monthly (actual payments are 2000 per month):

(ii) i = 0.06
(A) 267,900 (B) 268,400 (C) 268,900 (D) 269,400 (E) 269,900

6.2 Exercises 119
6.8. You are given:
(i) Y1 is the present value random variable for a 10-year temporary life
annuity-due of 1 per year on a select life aged 40 payable quarterly.
(ii) Y2 is the present value random variable for a 10-year certain and life
annuity-due of 1 per year on a select life aged 40 payable quarterly.
(iii) i = 0.05
(iv) Mortality follows the Standard Select Survival Model.
(v) Woolhouse’s formula with three terms is used to approximate quarterly
expected present values.
Calculate the variance of the sum of Y1 and Y2 .
(A) 3.8 (B) 3.9 (C) 4.0 (D) 4.1 (E) 4.2
6.9. Consider a special increasing 3-year temporary life annuity-due payable

annually on (x):
(i) The payment for the first year is 1000, the payment for the second year
is 3000, and the payment for the third year is 7000.
(ii) k px = (0.97)k for k = 0, 1, 2.
(iii) i = 0.04
(iv) Y is the present value random variable for this annuity.
Calculate the standard deviation of Y .
(A) 1904 (B) 1920 (C) 1936 (D) 1952 (E) 1968

6.2 Exercises 120
6.10. You are given a life annuity-due on (55) payable monthly. 100 is
payable each month during the first 10 years; 300 is payable each month
after the first 10 years.
Mortality follows the Illustrative Life Table, and i = 0.06. Woolhouse’s
formula with two terms is used to approximate monthly expected present
values.
(A) 23,710 (B) 24,210 (C) 24,710 (D) 25,210 (E) 25,710

6.2 Exercises 121
6.1. C
6.2. A
6.3. B
6.4. E
6.5. C
6.6. D
6.7. C
6.8. B
6.9. A
6.10. D


• Exam MLC, Sample Questions: #7, 284

123
7 PREMIUM CALCULATION I
7.1 Key Concepts

The policyholder often pays for a life insurance or life annuity with multiple
payments to the insurer over time called premiums.
Terminology:
• Fully Continuous Insurance: a continuous insurance that is funded

with a continuous annuity of premiums.
• Fully Discrete Insurance: an annual insurance that is funded by an

annual annuity-due of premiums.
• Semi-Continuous Insurance: a continuous insurance that is funded

by an annual annuity-due of premiums.
Net Loss-at-Issue:
• The first step to determining the premiums that the policyholder should
pay to fund the benefits of a particular policy is to determine the appro-
priate net loss-at-issue random variable, ignoring policy expenses:
L0 = 0 L = L =
Present value of future benefits at issue - Present value of future premiums
at issue
= P V F B@0 - P V F P @0
• The net loss-at-issue may be written with a superscript, as Ln0 . “Net”

will often be omitted if there is no expense information provided.
• Loss is random because P V F B@0 and P V F P @0 each depend on the

future lifetime of the policyholder.
• There will be a loss on a policy if the amount the insurer pays out in
benefits is higher than the amount the insurer collects in premiums; L0
> 0 if P V F B@0 > P V F P @0. There will be a profit on a policy if the
amount the insurer pays out in benefits is smaller than the amount the
insurer collects in premiums; L0 < 0 if P V F B@0 < P V F P @0.

Premium Principles
• A premium principle is a rule that manipulates the loss-at-issue random

variable in some way to generate premiums.
• Premiums calculated using net loss-at-issue random variable are called

net premiums.
• A common premium principle is the Equivalence Principle:
– Premiums are determined such that: E(L0 ) = 0.

– Under this principle, the insurer charges premiums so that, on aver-
age, there will be neither a loss or a profit on the policy. Clearly, this
is not the case in practice...
– Using the formula for L0 , the equivalence principle can also be stated
as: E(P V F B@0) = E(P V F P @0). This should be your starting
point for complicated problems involving the equivalence principle.
– On Exam LC, and Exam MLC prior to 2014, net premiums de-
termined via the equivalence principle are called benefit premiums.
Starting in 2014, the term “net premium” will be equivalent to “ben-
efit premium” on Exam MLC unless otherwise indicated; that is, a
net premium is a premium calculated using the equivalence principle
without expenses. In this supplement, both “benefit premium” and
“net premium” will be used interchangeably.

EQUIVALENCE PRINCIPLE
Benefit Premiums
Fully Continuous Insurance of 1 on (x):
PVFP@0
Continuous PVFB@0 BenefitPremium Benefit
Insurance Premium
Whole
Āx
Life v Tx āTx P̄ (Āx ) = āx
Tx
n-year v , Tx ≤ n āTx , Tx ≤ n
1
Āx:
1 n
Term 0, Tx > n ān , Tx > n P̄ (Āx:n ) = āx:n
n-year 0, Tx ≤ n āTx , Tx ≤ n
Ax:n1
Pure Endowment v n , Tx > n ān , Tx > n P̄ (Ax:n1 ) = āx:n
n-year v Tx , Tx ≤ n āTx , Tx ≤ n
Āx:n
Endowment v n , Tx > n ān , Tx > n P̄ (Āx:n ) = āx:n
h-Payment āTx , Tx ≤ h
Tx Āx
Whole Life v āh , Tx > h h P̄ (Āx ) = ā
x:h
h-Payment
n-year v Tx , Tx ≤ n āTx , Tx ≤ h
Āx:n
Endowment v n , Tx > n āh , Tx > h h P̄ (Āx:n ) = ā
x:h
• For each fully continuous insurance,

P V F P @0
Ln0 = P V F B@0 - (Benefit Premium) Benefit Premium , using the appropriate
entries for a value of Tx .
• In the “Benefit Premium” column, the left hand side of the equals sign
gives the actuarial notation for the benefit premium. Those taking Exam
MLC do not have to know this notation, and can denote the benefit
premium in each row as P .
• Furthermore, if the face amount is S, both sides of the equation in the

“Benefit Premium” column should be multiplied by S.

• For each fully continuous insurance, the benefit premium was determined
E(P V F P @0)
by E(Ln0 ) = E(P V F B@0) - (Benefit Premium) Benefit Premium = 0.
Āx δ Āx 1
• P̄ (Āx ) = āx = 1−Āx
= āx - δ.
Āx:n δ Āx:n 1
• P̄ (Āx:n ) = āx:n = 1−Āx:n
= āx:n - δ.
1
• With a constant force of mortality: P̄ (Āx ) = P̄ (Āx:n ) = µ.
• For a couple of the insurances in the table, there are analytic formulas
for the variance of the net loss-at-issue.
– For a fully continuous whole life insurance of 1 on (x):

P̄ (Āx ) 2 2
V ar(Ln0 ) = (1 + δ ) ( Āx − [Āx ]2 ).
µ
∗ With a constant force of mortality: V ar(Ln0 ) = µ+2δ .
– For a fully continuous n-year endowment insurance of 1 on (x):

P̄ (Āx:n ) 2 2
V ar(Ln0 ) = (1 + δ ) ( Āx:n − [Āx:n ]2 ).
– If the benefit is S, multiply each of the above V ar(Ln0 ) formulas by

S 2.
– These formulas for V ar(Ln0 ) are true for any type of premium, not
just a benefit premium, except for the constant force of mortality
formula.
– For any other type of fully continuous insurance, use: V ar(Ln0 ) =

E[(Ln0 )2 ] - (E[Ln0 ])2 . If the equivalence principle is used to determine
premiums, then: V ar(Ln0 ) = E[(Ln0 )2 ].
• The equivalence principle can also determine benefit premiums for con-
tinuous annuities. For example:
n| āx
P̄ (n| āx ) = āx:n .

Fully Discrete Insurance of 1 on (x):
PVFP@0
Annual PVFB@0 BenefitPremium Benefit
Insurance Premium
Whole
Ax
Life v Kx +1 äKx +1 Px = äx
n-year v Kx +1 , Kx < n äKx +1 , Kx < n
1
Ax:
1 n
Term 0, Kx ≥ n än , Kx ≥ n P x:n = äx:n
n-year 0, Kx < n äKx +1 , Kx < n
A 1
Pure Endowment v n , Kx ≥ n än , Kx ≥ n P x:n1 = ä x:n
x:n
n-year v Kx +1 , Kx < n äKx +1 , Kx < n
A
Endowment v n , Kx ≥ n än , Kx ≥ n Px:n = ä x:n
x:n
h-Payment äKx +1 , Kx < h
Ax
Whole Life v Kx +1 äh , Kx ≥ h h Px = ä
x:h
h-Payment
n-year v Kx +1 , Kx < n äKx +1 , Kx < h
Ax:n
Endowment v n , Kx ≥ n äh , Kx ≥ h h Px:n = ä
x:h
• For each fully discrete insurance,

P V F P @0
Ln0 = P V F B@0 - (Benefit Premium) Benefit Premium , using the appropriate
entries for a value of Kx .
• Recall, Kx can only take on non-negative integer values. So, Kx < n

=⇒ Kx = 0, 1, ..., n - 1.
• In the “Benefit Premium” column, the left hand side of each equation
gives the actuarial notation for the benefit premium. Those taking Exam
MLC only have to know the notation for the whole life, n-year term, n-
year pure endowment, and n-year endowment rows; and can denote other
benefit premiums as P .

“Benefit Premium” column should be multiplied by S.

• For each fully discrete insurance, the benefit premium was determined
E(P V F P @0)
by E(Ln0 ) = E(P V F B@0) - (Benefit Premium) Benefit Premium = 0.
• Px (the benefit premium for a fully discrete whole life insurance of 1 on

(x)) should not be confused with px (the probability that (x) survives to
age x + 1).
Ax dAx 1
• Px = äx = 1−Ax = äx - d.
Ax:n dAx:n 1
• Px:n = äx:n = 1−Ax:n = äx:n - d.
1
• With a constant force of mortality: Px = P x:n = vq.
• Argue that the following 3-Premium equations are valid:
1
– n Px - P x:n = Ax+n P x:n1
– Px:n - n Px = [1 - Ax+n ]P x:n1
1
– P x:n + P x:n1 = Px:n

for the variance of the net loss-at-issue.
– For a fully discrete whole life insurance of 1 on (x):

Px 2 2
V ar(Ln0 ) = (1 + d ) ( Ax − [Ax ]2 ).
pq
∗ With a constant force of mortality: V ar(Ln0 ) = q+i2 +2i .
– For a fully discrete n-year endowment insurance of 1 on (x):

Px:n 2 2
V ar(Ln0 ) = (1 + d ) ( Ax:n − [Ax:n ]2 ).
– If the benefit is S, multiply each of the above V ar(Ln0 ) formulas by
S 2.
– These formulas for V ar(Ln0 ) are true for any type of premium, not
formula.
– For any other type of fully discrete insurance, use: V ar(Ln0 ) = E[(Ln0 )2 ]
- (E[Ln0 ])2 . If the equivalence principle is used to determine premiums,
then: V ar(Ln0 ) = E[(Ln0 )2 ].
• The equivalence principle can also determine benefit premiums for dis-
crete annuities. For example:
n| äx
P (n| äx ) = äx:n .

Semi-Continuous Insurance of 1 on (x):
• You can obtain this table by taking the table for Fully Continuous
Insurance of 1 on (x) and replacing the continuous premium annuity
with an annual annuity-due.
• For example, a semi-continuous n-year term insurance of 1 on (x) is:
PVFP@0
Insurance PVFB@0 BenefitPremium Benefit Premium
Tx
Semi-Continuous v , Tx ≤ n äKx +1 , Kx < n
1
Āx:
1 n
n-year Term 0, Tx > n än , Kx ≥ n P (Āx:n )= äx:n
• Those taking Exam MLC do NOT have to know the actuarial notation
for semi-continuous benefit premiums; P is sufficient.
• Exam MLC Only: With a uniform distribution of deaths (UDD) in each

year of age:
– P (Āx ) = δi Px
1
– P (Āx:n ) = δi P x:n
1
– P (Āx:n ) = δi P x:n
1
+ P x:n1

7.2 Exercises 131
7.2 Exercises
7.1. On January 1, 2010, Pat purchases a 5-year deferred whole life insurance
of 100,000 payable at the end of the year of death. Premiums of 4000 are
payable at the beginning of each year for the first 5 years, and i = 0.05.
Calculate the loss-at-issue if Pat dies on September 30, 2016.
(A) 48,370 (B) 52,884 (C) 53,756 (D) 57,209 (E) 62,187
7.2. Stefano, age 60, purchases a whole life insurance of 1,000,000:
(i) The death benefit is payable at the moment of death.

(ii) Premiums of 50,000 are payable at the beginning of each year for as
long as Stefano is alive.
(iii) i = 0.05
(iv) L is the loss-at-issue random variable.
Calculate the value of L if Stefano dies at age 61.5.
(A) 764,059 (B) 809,410 (C) 819,138 (D) 831,810 (E) 879,429
7.3. Consider a fully continuous whole life insurance of 1000 on (x).

Assume δ = 0.08 and µx (t) = 0.04 for t ≥ 0.
Calculate the level annual benefit premium.
(A) 30 (B) 35 (C) 40 (D) 45 (E) 50
7.4. Paul, age 31, purchases a fully discrete 20-year endowment insurance
of 1000. Assume mortality follows the Illustrative Life Table, and i = 0.06.
Calculate: 1000P31:20 .
(A) 21 (B) 23 (C) 25 (D) 27 (E) 29

7.2 Exercises 132
7.5. Consider a fully discrete 5-payment 10-year endowment insurance of

1000 on (70):
1
(i) µx = 105−x for 0 ≤ x < 105
(ii) v = 0.95
(A) 91 (B) 94 (C) 122 (D) 144 (E) 151
7.6. For a fully continuous whole life insurance of 5000 on (x):
(i) The force of mortality is a constant.

(ii) δ = 0.05
(iii) L is the loss-at-issue random variable based on level annual benefit
premiums.
(iv) The standard deviation of L is 2236.07
(A) 100 (B) 125 (C) 150 (D) 175 (E) 200
7.7. A fully continuous whole life insurance of 10,000 on (x) is issued with
premiums determined by the equivalence principle.
You are also given:
(i) µx (t) = 0.02 for t ≥ 0

(ii) δ = 0.05
Calculate the probability that the loss-at-issue is positive.
(A) 0.30 (B) 0.33 (C) 0.36 (D) 0.39 (E) 0.42

7.2 Exercises 133
7.8. An insurer has just issued each of 100 independent lives aged 35 a
fully discrete 20-year endowment insurance of 1000 with level annual benefit
premiums. Each life has mortality that follows the Illustrative Life Table.
The effective annual interest rate is 0.06.
Using the normal approximation, determine the fund amount at issue, h,
that is necessary so that the insurer is 99% sure that the sum of the 100 loss-
at-issue random variables associated with the endowment insurances will not
exceed h.
(A) 2100 (B) 2200 (C) 2300 (D) 2400 (E) 2500
7.9. For a special fully discrete whole life insurance on (35):
(i) The death benefit is equal to 2000 plus the return of all benefit premiums
paid in the past without interest.
(ii) ä35 = 19.93
(iii) (IA)35 = 5.58
(iv) i = 0.045
Calculate the level annual benefit premium for this insurance.
(A) 20 (B) 22 (C) 24 (D) 26 (E) 28
(i) The level annual benefit premium for a fully discrete 20-year term in-
surance of 5000 on (x) is 75.
(ii) The level annual benefit premium for a fully discrete 20-year endow-
ment insurance of 5000 on (x) is 200.
(iii) The level annual benefit premium for a fully discrete 20-payment whole
life insurance of 5000 on (x) is 150.
Calculate the actuarial present value of a fully discrete whole life insurance
of 5000 on (x + 20).
(A) 2000 (B) 2500 (C) 3000 (D) 3500 (E) 4000

7.2 Exercises 134
7.11. For a life insurance on (x):
(i) 1000 is payable at the end of the year of death if death occurs in the
first ten years; 2000 is payable at the end of the year of death if death occurs
in the next ten years; otherwise, the death benefit is 0.
(ii) Level annual benefit premiums are payable at the beginning of each
year for the first 20 years.
(iii) d = 0.10
(iv) qx = 0.03 for all integer ages x.
(A) 31 (B) 33 (C) 35 (D) 37 (E) 39
7.12. For a special fully discrete 3-year term insurance on (x):
(i) The death benefit: bk+1 = 500(k + 1) for k = 0, 1, 2

(ii) qx+k = 0.02(k + 1) for k = 0, 1, 2
(iii) i = 0.03
Use the equivalence principle to calculate the level annual premium for this
insurance.
(A) 41 (B) 42 (C) 43 (D) 44 (E) 45
(i) lx = 100(110 − x)2 for 0 ≤ x ≤ 110

(ii) i = 0
Calculate the level annual benefit premium for a fully discrete 5-payment
1
15-year term insurance of 1 on (30): 5 P 30:15 .
(A) 0.06 (B) 0.07 (C) 0.08 (D) 0.09 (E) 0.10

7.2 Exercises 135
7.14. A fully discrete 5-year endowment insurance of 1000 was just issued
to Math Mage, aged 30. In determining the level annual benefit premium, it
was assumed that i = 0.06 and that Math Mage had mortality that follows
the Illustrative Life Table.
Shortly after issuing the 5-year endowment insurance, it was discovered
that Math Mage had been cursed by Hattendorf. In calculating the level
annual benefit premium, it should have been assumed that i = 0.06 and that
Math Mage had mortality such that the actual force of mortality was µ30 (t)
+ 0.10 for 0 < t < 5, where µ30 (t) is the force of mortality associated with
the Illustrative Life Table.
Calculate the difference between the benefit premium that Math Mage
should be paying calculated using the correct mortality (based on µ30 (t) +
0.10) and the benefit premium actually payable by Math Mage calculated
using the incorrect mortality (Illustrative Life Table).
(A) 40 (B) 50 (C) 60 (D) 70 (E) 80
7.15. Consider a 5-year deferred whole life annuity-due on (60) with an

annual payment of 10,000. You are given:
(i) Benefit premiums are payable at the beginning of each year during the
first five years. The benefit premium payable in each of years one and two is
half of the benefit premium payable in each of years three, four, and five.
(ii) d = 0.04306
(iii) q60 = 0.006155 and q61 = 0.006765
(iv) 5 E60 = 0.77282
(v) ä65 = 13.4662 and ä62:3 = 2.8513
Calculate the benefit premium payable in each of years one and two.
(A) 14,300 (B) 14,600 (C) 14,900 (D) 15,200 (E) 15,500

7.2 Exercises 136
7.16. Consider a fully continuous whole life insurance of 100,000 on (30):
(i) If the level annual premium is π1 , the standard deviation of the loss-at-
issue random variable is 55,621.49.
(ii) If the level annual premium is π2 , the standard deviation of the loss-
at-issue random variable is 49,441.32.
(iii) π1 is 1.5 times π2 .
(iv) δ = 0.04
(v) Z is the present value random variable for the continuous whole life
insurance of 100,000 on (30).
Calculate the standard deviation of Z.
(A) 37,100 (B) 38,200 (C) 39,300 (D) 40,400 (E) 41,500
7.17. Bruce and Lucius, both aged x, have each just purchased a fully
discrete 3-year term insurance of 1000:
(i) Bruce pays a benefit premium of 175.72 each year. If Bruce dies in the
second year after policy issue, the loss-at-issue is 559.27.
(ii) Lucius pays non-level annual benefit premiums. The first benefit pre-
mium is 100, the second benefit premium is 175, and the third benefit pre-
mium is P .
(iii) Each life has mortality such that: k| qx = (0.3)k+1 for k = 0, 1, 2.
(iv) The effective annual interest rate is i.
Calculate: P .
(A) 285 (B) 315 (C) 345 (D) 375 (E) 405

7.2 Exercises 137
7.18. Consider a special 20-year deferred whole life annuity-due of 5000

per year on (45) payable annually:
(i) Level annual benefit premiums are payable at the beginning of the year
during the first 20 years after policy issue.
(ii) There is a death benefit during the premium-paying period, payable
at the end of the year of death, that is equal to the return of all benefit
premiums previously paid with interest at 6%.
(iii) i = 0.06
Calculate the benefit premium.
(A) 1180 (B) 1210 (C) 1240 (D) 1270 (E) 1300
(i) A fully discrete 2-year deferred, 3-year term insurance of 1000 is issued
to a life aged x.
(ii) Level annual premiums are only payable during the first two years.
(iii) The level annual premium is determined such that the average loss-
at-issue is zero.
(iv) v = 0.90
(v) k| qx = 0.05(1 + k) for k = 0, 1, 2, 3, 4; 5| qx = 0.25
Calculate the median loss-at-issue.
(A) 193 (B) 226 (C) 258 (D) 295 (E) 331

7.2 Exercises 138
7.20. Paul is attempting to determine the level annual benefit premium for
a fully discrete 20-year endowment insurance of 10,000 on (55). Paul assumes
the following:
(i) i = 0.06
(ii) Paul is not sure of the future lifetime distribution of (55). He believes
there is an 80% probability that the mortality of (55) follows the Illustrative
Life Table, and that there is a 20% probability that the mortality of (55) is
such that k| q55 = 0.05 for k = 0, 1, 2, ..., 19.
Determine the level annual benefit premium using Paul’s assumptions.
(A) 384 (B) 414 (C) 444 (D) 474 (E) 504

7.2 Exercises 139
7.1. B
7.2. D
7.3. C
7.4. D
7.5. E
7.6. B
7.7. D
7.8. E
7.9. A
7.10. C
7.11. B
7.12. D
7.13. B
7.14. A
7.15. B
7.16. A
7.17. B
7.18. D
7.19. C
7.20. B


• Exam MLC, Fall 2013: #14, 15, 16
• Exam MLC, Spring 2013: #1, 3, 18
• Exam 3L, Spring 2013: #13, 14
• Exam MLC, Fall 2012: #20, 22
• Exam MLC, Sample Questions: #6, 14, 29, 40, 47, 51, 76, 84, 92, 96, 97,
99, 111, 119, 127, 142, 154, 157, 172, 174, 184, 204, 221, 228, 309
• Exam 3L, Fall 2008: #22

141
8 PREMIUM CALCULATION II (Exam MLC Only)
8.1 Key Concepts

Premiums Payable on an m-THLY Basis
Life Insurance with Premiums Payable on an m-thly Basis: a life

insurance that is funded by an m-thly annuity-due of premiums.
Continuous Insurance of 1 on (x) with m-thly Premiums:
Insurance of 1 on (x) in Premium Calculation I and replacing the
continuous premium annuity with an m-thly annuity-due.
• For example, a continuous n-year term insurance of 1 on (x) with premi-

ums payable on an m-thly basis is:
PVFP@0
Insurance PVFB@0 BenefitPremium Benefit Premium
(m)
Continuous v Tx , Tx ≤ n ä (m) 1 , Kx < n
Kx +m
1
Āx:
(m)
n-year Term 0, Tx > n än , Kx ≥n P (m) (Āx:n
1
)= (m)
n
äx:n
with m-thly premiums
• You do NOT have to know the actuarial notation for m-thly benefit
premiums; P is sufficient.
(m) (m) 1 2 1
• Kx < n means that Kx = 0, m, m, ..., n - m.
• For the continuous n-year term insurance of 1 on (x) with premiums

payable on an m-thly basis, the premium that is actually payable at the
beginning of each m-th of a year is m1 P (m) (Āx:n
1
).
• The annual benefit premium for an annual n-year term insurance of 1 on

1
Ax:
(x) with premiums payable on an m-thly basis is P (m)1x:n = n
(m) .
äx:n

Gross Loss-at-Issue:
• The first step to determining the premiums that the policyholder should
pay to fund both the benefits and expenses associated with a particular
policy is to determine the appropriate
gross loss-at-issue random variable:
L0 = 0 L = L = Present value of future benefits at issue + Present value

of future expenses at issue - Present value of future premiums at issue.
= P V F B@0 + P V F E@0 - P V F P @0
• The gross loss-at-issue may be written with a superscript, as Lg0 .
• If expenses are ignored, the loss is a net loss. If expense information is

provided, the loss is assumed to be a gross loss.
• Loss is random because P V F B@0, P V F E@0, and P V F P @0 each de-

pend on the future lifetime of the policyholder.
• There will be a loss on a policy if the amount the insurer pays out in
benefits and expenses is higher than the amount the insurer collects in
premiums; L0 > 0 if P V F B@0 + P V F E@0 > P V F P @0. There will be
a profit on a policy if the amount the insurer pays out in benefits and
expenses is smaller than the amount the insurer collects in premiums; L0
< 0 if P V F B@0 + P V F E@0 < P V F P @0.

Gross Premium Equivalence Principle Calculation
• Premiums calculated using the gross loss-at-issue are called gross premiums.
Gross premiums are denoted as G.
• There are no special formulas to memorize for gross premiums associated

with various insurance coverages. Just solve the following equivalence principle
equation for G:
E(Lg0 ) = E(P V F B@0) + E(P V F E@0) - E(P V F P @0) = 0 =⇒
E(P V F B@0) + E(P V F E@0) = E(P V F P @0).
• Under this principle, the insurer charges premiums so that, on average,

there will be neither a loss or a profit on the policy. Clearly, this is not
the case in practice. It is possible to allow for profit by including a “profit
loading expense” in E(P V F E@0).
• The gross premium pays for both the benefits and expenses of a policy.
It makes sense that a gross premium (G) can be decomposed into the
sum of two premiums: a benefit premium that pays for just the benefits
(P ) plus an expense premium or expense loading that pays for just the
expenses (e):
G = P + e.

Expense Terminology
• Initial Year: The first policy year. Expenses incurred at the start of
the first year may be called initial expenses.
• Renewal Year: Any policy year after the first. Expenses incurred at
the start of a renewal year may be called renewal expenses.
• Per-Premium Expense: an expense that is a function of the premium

charged, such as a percent of premium expense. Examples include agent
commissions and taxes.
• Per-Policy Expense: an expense that is a fixed amount per policy,

regardless of the amount of coverage. Examples include maintenance
expenses.
• Per-Insurance Expense: an expense that is a function of the benefit of

the insurance. For example, consider a fully discrete whole life insurance
of 100,000 on (x). If there is an expense of 1 per 1000 of insurance payable
at the beginning of the year, that equates to an expense of 100 payable
at the beginning of the year.
• Acquisition Expense: an expense that is payable on the policy issue

date. This is a type of initial expense.
• Settlement Expense: an expense that is payable at the same time the

benefit of the insurance (the claim) is payable.

Single Policy Percentile Premium Principle
Percentile premiums are determined such that:

P r(L0 > 0) = α for 0 ≤ α ≤ 1.
The best way to calculate percentile premiums for a life insurance on (x)
is to use the following approach:
• First, note that P r(L0 > 0) = α =⇒ tα q x = α.

That is because L0 is a non-decreasing function of the policyholder’s
future lifetime for a life insurance. Therefore, the life insurance policy
will realize a loss 100α% of the time if the policyholder dies by the 100α-
th percentile of the policyholder’s future lifetime distribution.
Solve tα qx = α for tα .
If the life insurance is annual (the death benefit is payable at the end
of the year of death), keep only the integer part of tα to obtain the
appropriate kα . That is because an annual life insurance pays its benefit
based on the curtate future lifetime of the policyholder.
• Using the solved for tα or kα , setup the equation:

L0 = P V F B@0 - P V F P @0 = 0.
Solve this last equation for the 100α-th percentile premium.

Portfolio Percentile Premium Principle
• Consider a portfolio of N independent and identically distributed policies.
• Let L0,i denote the loss-at-issue for policy number i in the portfolio, where
i = 1, 2, ..., N .
PN
• Let S = i=1 L0,i .
• The Portfolio Percentile Premium Principle solves for the premium for
each policy via the following, where Φ(.) denotes the c.d.f. of a standard
normal distribution:
P r[S > 0] = 1 - Φ(− √E(S) ) = α for 0 ≤ α ≤ 1

V (S)
=⇒ Φ(− √E(S) ) = 1 - α
V (S)
p
• Note: Both E(S) and V (S) will be functions of the premium.
Set - √E(S) = z1−α , the 100(1 - α)-th percentile of the standard normal
V (S)
distribution, and solve for the premium.
The value of α will be close to zero.

8.2 Exercises 147
8.2 Exercises
8.1. For a special fully discrete 2-year term insurance on a select life aged 30
payable at the end of the year of death:
(i) Premiums are payable semiannually.

(ii) Premiums are payable only in the first year.
(iii) bk+1 = 500(2 - k) for k = 0, 1
(iv) Mortality follows the Standard Select Survival Model, and i = 0.05.
(v) Deaths are uniformly distributed within each year of age.
Calculate the amount of each semiannual benefit premium.
(A) 0.20 (B) 0.23 (C) 0.26 (D) 0.29 (E) 0.32
8.2. Calculate the level annual net premium for a 20-year endowment
insurance of 50,000 on (40) where the death benefit is payable at the moment
of death and premiums are payable at the beginning of each month. Assume
that mortality follows the Illustrative Life Table, i = 0.06, and deaths are
uniformly distributed over each year of age.
(A) 1430 (B) 1450 (C) 1470 (D) 1490 (E) 1510
8.3. You are given:

(12) (4)
(i) äx = 0.9äx
(4)
(ii) Px = 0.027
(12)
Calculate: Px .
(A) 0.02 (B) 0.03 (C) 0.04 (D) 0.05 (E) 0.06

8.2 Exercises 148
8.4. You are given a special fully discrete 20-year pure endowment insur-
ance of 20,000 on (30):
(i) Level gross premiums are payable at the beginning of each month during
the first ten years.
(ii) The monthly gross premium is 115% of the monthly benefit premium
payable at the beginning of each month during the first ten years.
(iii) Deaths are uniformly distributed within each year of age.
(iv) Mortality follows the Illustrative Life Table, and i = 0.06.
Calculate the expected gross loss-at-issue.
(A) - 880 (B) -640 (C) -440 (D) - 290 (E) 0
8.5. For a fully continuous whole life insurance of 10,000 on (30):
(i) Mortality follows de Moivre’s Law with ω = 100.

(ii) δ = 0.05
Calculate the 20th percentile premium.
(A) 455 (B) 465 (C) 475 (D) 485 (E) 495
8.6. Consider a fully discrete whole life insurance of 1000 on (40):

(ii) i = 0.06
Calculate the smallest premium such that the probability of a positive

loss-at-issue on the insurance is less than 0.20.
(A) 14 (B) 15 (C) 16 (D) 17 (E) 18

8.2 Exercises 149
8.7. An insurer issues whole life insurance policies to lives aged 30. The
sum insured of 100,000 is paid at the end of the year of death, and level
annual premiums are payable at the beginning of each year.
Initial expenses are 1500 plus 20% of the first premium. Renewal expenses
are 4% of the second and subsequent premiums.
Assume mortality follows the Illustrative Life Table, and i = 0.06.
Using the normal approximation, calculate the annual premium via the
portfolio percentile premium principle, such that the probability that the
loss-at-issue on the portfolio is negative is 0.90. Assume a portfolio of 1000
identical, independent policies.
(A) 800 (B) 810 (C) 820 (D) 830 (E) 840
8.8. An insurer issues single-premium whole life insurance policies of

100,000, payable at the end of the month of death, to select lives aged 40.
Initial expenses are 20% of the single premium. Assume mortality follows
the Standard Select Survival Model, and i = 0.05.
Using the normal approximation and Woolhouse’s formula with 3 terms,
calculate the single premium via the portfolio percentile premium principle,
so that the probability the loss-at-issue on the portfolio is negative is 95%.
Assume a portfolio of 15,000 identical, independent policies.
(A) 14,400 (B) 14,700 (C) 15,000 (D) 15,300 (E) 15,600

8.2 Exercises 150
8.9. On January 1, 2012, Pat purchases a 5-year deferred whole life in-
surance of 100,000 payable at the end of the year of death. You are given:
(i) Level gross premiums of 5000 are payable at the beginning of each year
for the first 5 years.
(ii) There is an acquisition expense of 100 that is payable on the policy
issue date.
(iii) There is an annual maintenance expense of 20 payable at the beginning
of each year, including the first year.
(iv) There is a first year total percent of premium expense of 15% of the
gross premium, and a renewal year total percent of premium expense of 5%
of the gross premium; all percent of premium expenses are payable at the
beginning of the year.
(v) There is a claims settlement expense of 120, payable at the time the
death benefit is paid.
(vi) i = 0.05.
Calculate the loss-at-issue if Pat dies on April 30, 2014.
(A) -12,925 (B) -9,221 (C) 26,409 (D) 43,857 (E) 73,562
8.10. Consider a fully discrete whole life insurance of 100,000 on (35).

Percent of premium expenses are 10% per year, and per policy expenses are
25 per year. All expenses are paid at the beginning of the year. The effective
annual interest rate is 0.05.
x
Assume that mortality follows: S0 (x) = 1 - 95 for 0 ≤ x ≤ 95. Calculate
the level annual gross premium using the equivalence principle.
(A) 2410 (B) 2440 (C) 2470 (D) 2500 (E) 2530

8.2 Exercises 151
8.11. For a fully discrete 5-payment 10-year endowment insurance of 1000

on (30):
(i) There are level annual gross premiums.

(ii) Mortality follows the Illustrative Life Table, and i = 0.06
(iii) The following expenses are all payable at the beginning of the year:
Year 1 Years 2+
% Premium Per Policy % Premium Per Policy
Taxes 5% — 5% —
Sales Commission 20% — 5% —
Policy Maintenance — 20 — 10
(iv) There is also a claims settlement expense of 50, payable at the time
the sum insured is paid.
Calculate the level annual gross premium using the equivalence principle.
(A) 160 (B) 165 (C) 170 (D) 175 (E) 180
8.12. Calculate the probability that the endowment insurance in Exercise

8.8 is profitable.
(A) 0.974 (B) 0.977 (C) 0.980 (D) 0.983 (E) 0.986

8.2 Exercises 152
8.13. For a fully discrete 5-year term insurance of 10,000 on (35):
(i) Mortality follows the Illustrative Life Table, and i = 0.06

(ii) The following expenses, all payable at the beginning of the year:
Year 1 Years 2-5

% Premium Per Policy % Premium Per Policy
Sales Commission 40% — 4% —
Taxes 6% — 6% —
Policy Maintenance — K — 5
(iii) There is an acquisition expense of 50 that is payable at policy issue,

and a settlement expense of 120 that is payable at the same time as the death
benefit.
(iv) The level annual gross premium, calculated using the equivalence prin-
ciple, is 47.73.
Calculate: K.
(A) 8 (B) 10 (C) 12 (D) 14 (E) 16
8.14. For a fully discrete 10-year endowment insurance of 10,000 on (50):
(i) Percent of premium expenses consist of commissions equal to 50% of

gross premium in the first year and 5% of gross premium premium in renewal
years. Each expense is at the beginning of the year.
(ii) Other expenses include an acquisition expense of 20 due at policy issue,
a settlement expense of 180 due at the time the benefit of 10,000 is paid, and
an annual level maintenance expense of 25 at the beginning of each year.
(iii) Mortality follows the Illustrative Life Table, and i = 0.06.
Using the equivalence principle, calculate the annual level expense pre-
mium.
(A) 110 (B) 120 (C) 130 (D) 140 (E) 150

8.2 Exercises 153
8.15. The level annual gross premium for a fully discrete 20-year endow-
ment insurance of 10,000 on (45) is determined using the following assump-
tions:
(i) Percent of premium expenses consist of sales commission equal to 20%

of premium in the first year followed by 5% of premium in all other years,
and taxes equal to 2.5% of premium each year.
(ii) Per-policy expenses are 50 in the first year and 25 in all other years.
(iii) All expenses are payable at the beginning of the year.
(v) i = 0.06
(vi) The level annual gross premium is calculated using the equivalence
principle.
Calculate the difference between the gross premium and the loading for
expenses in the gross premium.
(A) 300 (B) 320 (C) 340 (D) 360 (E) 380

8.2 Exercises 154
8.16. Consider a 15-year term insurance on a select life aged 40 with

benefits payable at the end of the year of death:
(i) The term insurance has a basic sum insured of 100,000; compound
reversionary bonuses at the rate of 3.5% will vest at the end of each year.
(ii) Level gross premiums are payable at the beginning of the month during
the 15 years.
(iii) The insurer has initial expenses of 2000 plus 4% of the first monthly
gross premium, and renewal expenses of 1% of the second and subsequent
monthly gross premiums.
(iv) Mortality follows the Standard Select Survival Model, and i = 0.05.
(v) A[40] and A55 , each at an effective annual interest rate of (1.05/1.035 -
1), are equal to 0.516777 and 0.634596, respectively.
(vi) Woolhouse’s formula with three terms is used to calculate monthly
annuity-due expected present values.
Calculate the monthly gross premium using the equivalence principle.
(A) 24 (B) 26 (C) 28 (D) 30 (E) 32
8.17. Consider a special fully discrete whole life insurance on a select life
aged 50:
(i) The death benefit is 120,000 during the first 15 years, and is 72,000
thereafter.
(ii) Gross premiums are G during the first 15 years, and are 0.6G thereafter.
(iii) There is a claims settlement expense of 100, payable at the same time
as the death benefit.
(iv) Expenses are 20% of the first gross premium, and 2% of each renewal
gross premium. Expenses are paid at the beginning of the year.
(v) Mortality follows the Standard Select Survival Model, and i = 0.05.
Calculate G using the equivalence principle.
(A) 900 (B) 950 (C) 1000 (D) 1050 (E) 1100

8.2 Exercises 155
8.18. Consider a 10-year term insurance of 100,000 on (40) payable at the

moment of death:
(i) Only one gross premium is payable at issue.

(ii) There is an expense of 1.00 per 1000 of insurance payable at issue.
There is an expense of 0.50 per 1000 of insurance payable at the beginning
of each year during the second and subsequent years.
(iii) Mortality follows the Illustrative Life Table, and i = 0.06
(iv) Deaths are uniformly distributed within each year of age.
Calculate the gross premium using the equivalence principle.
(A) 3100 (B) 3200 (C) 3300 (D) 3400 (E) 3500
8.19. For the insurance described in Exercise 8.18, calculate the probability
that the gross loss-at-issue is positive.
(A) 0.027 (B) 0.031 (C) 0.035 (D) 0.039 (E) 0.043
8.20. Consider a fully discrete whole life insurance of 1000 on (x):
(i) The contract premium is the level annual gross premium determined
using the equivalence principle.
(ii) The level annual net premium is 13.
(iii) There is a level annual expense that is payable at the beginning of
each year.
(iv) i = 0.03
(v) Ax = 0.20 and 2 Ax = 0.09
Calculate the standard deviation of the gross loss-at-issue random variable.
(A) 323 (B) 333 (C) 343 (D) 353 (E) 363

8.2 Exercises 156
8.1. A
8.2. C
8.3. B
8.4. A
8.5. E
8.6. C
8.7. C
8.8. E
8.9. A
8.10. C
8.11. D
8.12. D
8.13. B
8.14. D
8.15. A
8.16. B
8.17. D
8.18. C
8.19. D
8.20. A


• Exam MLC, Fall 2013: #9, 18, 19
• Exam MLC, Spring 2012: #4, 5, 7, 22, 25, 26
• Exam MLC, Sample Questions: #24, 37, 60, 129, 139, 147, 170, 190,
198, 239, 240, 245, 246, 247, 248
• Exam 3L, Fall 2010: #14

158
9 RESERVES I
9.1 Key Concepts

The policy value at time t is called the reserve at time t. The net premium reserve
represents the accumulated fund that the insurer needs to have at some future
time that in addition to future net premiums should, on average, fund the fu-
ture benefits associated with the policy. On Exam LC, and Exam MLC prior
to 2014, if premiums are calculated using the equivalence principle and ignor-
ing expenses, the reserve for the policy is called the benefit reserve. Starting
in 2014, the term “net premium reserve” is equivalent to “benefit reserve” on
Exam MLC; that is, a net premium reserve is a reserve based on premiums
calculated using the equivalence principle and ignoring expenses. Both “net
premium reserve” and “benefit reserve” will be used interchangeably in this
supplement.
To illustrate why a reserve is often necessary, consider a fully discrete whole
life insurance of S on (x) with net premiums. As t approaches infinity, the
present value of future benefits at time t >>> the present value of future
premiums at time t; this is because the policyholder is approaching death
meaning that the benefit is payable sooner and sooner (with increasing t)
and fewer future premiums will be collected.
Net Future Loss at Time t:
• The first step to determining the net premium or benefit reserve is to de-
termine the appropriate net future loss at time t random variable (con-
ditional on survival of the policyholder to time t): Lt = t L
= Present value of future benefits at time t - Present value of future
premiums at time t = P V F B@t - P V F P @t
• The net future loss at time t may be written as Lnt .
• The basic notation for a net premium or benefit reserve at time t is t V .

Superscripts may be included. There is different notation for reserves for
policies whom those taking Exam MLC do NOT have to know.
• We will focus on benefit reserves.

• There are two main approaches to calculating benefit reserves:
– Prospective Approach (looking into the future)

– Retrospective Approach (looking into the past)
Prospective Approach
• The general prospective benefit reserve formula is:
tV = E(Lnt ) = E(P V F B@t) - E(P V F P @t).
• This formula emphasizes that the benefit reserve is a balancing item in

funding a policy, as: t V + E(P V F P @t) = E(P V F B@t).
• Note: the benefit reserve at time 0, 0 V = E(L0 ), must equal zero by the
equivalence principle.
• The prospective benefit reserve formula is very useful for a policy where
no premiums are payable after time t, as t V = E(P V F B@t).

Prospective Benefit Reserves

Fully Continuous Insurance of 1 on (x):
Continuous Benefit Reserve Prospective Benefit

Insurance Notation Reserve Formula
Whole Life t V̄ (Āx ) Āx+t - P̄ (Āx )āx+t
1 1
n-year Āx+t:n−t - P̄ (Āx:n )āx+t:n−t , t < n
1
Term t V̄ (Āx:n ) 0, t = n
1
n-year Ax+t:n−t - P̄ (Ax:n1 )āx+t:n−t , t < n
Pure Endowment t V̄ (Ax:n1 ) 1, t = n
n-year Āx+t:n−t - P̄ (Āx:n )āx+t:n−t , t < n

Endowment t V̄ (Āx:n ) 1, t = n
h-Payment Āx+t - h P̄ (Āx )āx+t:h−t , t ≤ h

h
Whole Life t V̄ (Āx ) Āx+t , t > h
h-Payment Āx+t:n−t - h P̄ (Āx:n )āx+t:h−t , t ≤ h < n

h
n-year t V̄ (Āx:n ) Āx+t:n−t , h < t < n
Endowment 1, t = n
• For each fully continuous insurance, Lnt = P V F B@t - P V F P @t.

This formula can be calculated for a particular policy by appropriately
adjusting the formula for Ln0 for a time t years after issue. For example,
if we consider a fully continuous whole life insurance of 1 on (x):
Lnt = v Tx+t - P̄ (Āx )āTx+t .
• In the “Prospective Benefit Reserve Formula” column, the left hand side
of the equals sign gives the notation for the benefit reserve. For example,
1
t V̄ (Āx:n ) denotes the benefit reserve at time t for a fully continuous n-year
term insurance of 1 that was issued to a life aged x.


“Prospective Benefit Reserve Formula” column are multiplied by S.
• For each fully continuous insurance, the prospective benefit reserve was
determined by E(Lnt ) = E(P V F B@t) - E(P V F P @t).
Āx+t −Āx āx+t

• t V̄ (Āx ) = Āx+t - P̄ (Āx )āx+t = 1−Āx
=1- āx = (P̄ (Āx+t ) - P̄ (Āx ))āx+t .
Āx+t:n−t −Āx:n āx+t:n−t

• t V̄ (Āx:n ) = Āx+t:n−t - P̄ (Āx:n )āx+t:n−t = 1−Āx:n
= 1 - āx:n =
(P̄ (Āx+t:n−t ) - P̄ (Āx:n ))āx+t:n−t for t < n.
1
• With a constant force of mortality: t V̄ (Āx ) = t V̄ (Āx:n ) = 0.

for the variance of the net future loss at time t.
– For a fully continuous whole life insurance of 1 on (x):

P̄ (Āx ) 2 2
V ar(Lnt ) = (1 + δ ) ( Āx+t − [Āx+t ]2 ).
µ
∗ With a constant force of mortality: V ar(Lnt ) = µ+2δ .
– For a fully continuous n-year endowment insurance of 1 on (x):

P̄ (Āx:n ) 2 2
V ar(Lnt ) = (1 + δ ) ( Āx+t:n−t − [Āx+t:n−t ]2 ) for t < n.
– If the benefit is S, multiply each of the above V ar(Lnt ) formulas by
S 2.
– These formulas for V ar(Lnt ) are true for any type of premium, not
formula. Just use the appropriate premium in the formula.
– If t = 0, each formula becomes V ar(Ln0 ) for that life insurance.
– For any other type of fully continuous insurance, use: V ar(Lnt ) =

E[(Lnt )2 ] - (E[Lnt ])2 .
• The prospective benefit reserve formula can be applied to continuous

annuities. For example:
t V̄ (n| āx ) = n−t| āx+t - P̄ (n| āx )āx+t:n−t for t ≤ n; āx+t for t > n.

Fully Discrete Insurance of 1 on (x):
Annual Benefit Reserve Prospective Benefit

Whole Life k Vx Ax+k - Px äx+k
1 1
n-year Ax+k:n−k - P x:n äx+k:n−k , k < n
1
Term k V x:n 0, k = n
1
n-year Ax+k:n−k - P x:n1 äx+k:n−k , k < n
1
Pure Endowment k V x:n 1, k = n
n-year Ax+k:n−k - Px:n äx+k:n−k , k < n

Endowment k Vx:n 1, k = n
h-Payment Ax+k - h Px äx+k:h−k , k < h

h
Whole Life k Vx Ax+k , k ≥ h
h-Payment Ax+k:n−k - h Px:n äx+k:h−k , k < h < n

h
n-year k Vx:n Ax+k:n−k , h ≤ k < n
Endowment 1, k = n
• For each fully discrete insurance, Lnk = P V F B@k - P V F P @k.

This formula can be calculated for a particular policy by appropriately
adjusting the formula for Ln0 for a time k years after issue. For example,
if we consider a fully discrete whole life insurance of 1 on (x):
Lnk = v Kx+k +1 - Px äKx+k +1 .
• Recall, k can only take on non-negative integer values. So, k < n =⇒

k = 0, 1, ..., n - 1.

• In the “Prospective Benefit Reserve Formula” column, the left hand side
of the equals sign gives the actuarial notation for the benefit reserve. For
1
example, k V x:n denotes the benefit reserve at time k for a fully discrete
n-year term insurance of 1 that was issued to a life aged x. Those taking
Exam MLC do not have to know this notation, and can denote the benefit
reserve in each row as k V or k V n .

“Prospective Benefit Reserve Formula” column are multiplied by S.
• For each fully discrete insurance, the prospective benefit reserve was de-
termined by E(Lnk ) = E(P V F B@k) - E(P V F P @k).
Ax+k −Ax äx+k

• k Vx = Ax+k - Px äx+k = 1−Ax =1- äx = (Px+k - Px )äx+k .
A −Ax:n ä
x+k:n−k x+k:n−k
• k Vx:n = Ax+k:n−k - Px:n äx+k:n−k = 1−Ax:n =1- äx:n
= (Px+k:n−k - Px:n )äx+k:n−k for k < n.
1
• With a constant force of mortality: k Vx = k V x:n = 0.
• Argue that the following reserve equations are valid:

1
– Px - P x:n = n Vx P x:n1
– nn Vx - n V x:n
1
= Ax+n
– nn Vx - n Vx = Px äx+n
– n Vx:n - n Vx = 1 - n Vx
– hh Vx:n - hh Vx = Ax+h:n−h - Ax+h

for the variance of the net future loss at time k.
– For a fully discrete whole life insurance of 1 on (x):

Px 2 2
V ar(Lnk ) = (1 + d ) ( Ax+k − [Ax+k ]2 ).
pq
∗ With a constant force of mortality: V ar(Lnk ) = q+i2 +2i .
– For a fully discrete n-year endowment insurance of 1 on (x):

Px:n 2 2
V ar(Lnk ) = (1 + d ) ( Ax+k:n−k − [Ax+k:n−k ]2 ) for k < n.
– If the benefit is S, multiply each of the above V ar(Lnk ) formulas by

S 2.
– These formulas for V ar(Lnk ) are true for any type of premium, not
formula. Just use the appropriate premium in the formula.
– If k = 0, each formula becomes V ar(Ln0 ) for that life insurance.
– For any other type of fully discrete insurance, use: V ar(Lnk ) = E[(Lnk )2 ]
- (E[Lnk ])2 .
• The prospective benefit reserve formula can be applied to discrete annu-

ities. For example:
k V (n| äx ) = n−k| äx+k - P (n| äx )äx+k:n−k for k < n; äx+k for k ≥ n.

Semi-Continuous Insurance of 1 on (x):
Insurance of 1 on (x) and replacing the continuous premium annuity
with an annual annuity-due.
• For example, a semi-continuous n-year term insurance of 1 on (x) is:
Semi-Continuous Benefit Reserve Prospective Benefit

1 1
n-year Āx+k:n−k - P (Āx:n )äx+k:n−k , k < n
1
Term kV (Āx:n ) 0, k = n
• Those taking Exam MLC do NOT have to know the actuarial notation
for semi-continuous benefit reserves; k V or k V n is sufficient.
• Exam MLC Only: With UDD within each year of age:

i
– k V (Āx ) = δ k Vx
1 i 1
– k V (Āx:n )= δ k V x:n
i 1
– k V (Āx:n ) = δ k V x:n + k V x:n1

Retrospective Approach
• The general retrospective benefit reserve formula is:

E[PVFP on [0, t]] E[PVFB on [0, t]]
tV = t Ex
- t Ex
.
• This formula emphasizes that the benefit reserve is the accumulated fund
that the insurer needs to have by time t that along with future premiums,
on average, will fund the future benefits of a policy. The accumulated
fund is the accumulated premiums from issue up to time t ( E[PVFPt Eon
x
[0, t]]
)
less the accumulated value of any benefits paid from issue up to time t
( E[PVFBt Eon
x
[0, t]]
).
• The retrospective benefit reserve formula is very useful for a policy where
no benefits are payable before time t, as t V = E[PVFPt Eon
x
[0, t]]
.
• The prospective benefit reserve will equal the retrospective benefit reserve
for a particular policy at a specific time as long as the same assumptions
and basis (interest and mortality) are used.

Retrospective Benefit Reserves

Fully Continuous Insurance of 1 on (x)
Continuous Benefit Reserve Retrospective Benefit

P̄ (Āx )āx:t Ā 1
x:t
Whole Life t V̄ (Āx ) t Ex
- t Ex
1 )ā
P̄ (Āx: Ā 1
n x:t x:t
n-year t Ex
- t Ex
,t<n
1
Term t V̄ (Āx:n ) 0, t = n
P̄ (Ax:n1 )āx:t
n-year t Ex
,t<n
Pure Endowment t V̄ (Ax:n1 ) 1, t = n
P̄ (Āx:n )āx:t Ā 1
x:t
n-year t Ex
- t Ex
,t<n
Endowment t V̄ (Āx:n ) 1, t = n
h P̄ (Āx )āx:t Ā 1
x:t
h-Payment t Ex
- t Ex
,t≤h
h P̄ (Āx )ā Ā 1
h x:h x:t
Whole Life t V̄ (Āx ) t Ex
- t Ex
,t>h
h P̄ (Āx:n )āx:t Ā 1
x:t
h-Payment t Ex
- t Ex
,t≤h<n
h P̄ (Āx:n )ā Ā 1
h x:h x:t
n -year Endowment t V̄ (Āx:n ) t Ex
- t Ex
,h<t<n
1, t = n

Fully Discrete Insurance of 1 on (x)
Annual Benefit Reserve Retrospective Benefit

Px ä A1
x:k x:k
Whole Life k Vx k Ex
- k Ex
1 ä
P x: A1
n x:k x:k
n-year k Ex
- k Ex
,k<n
1
Term k V x:n 0, k = n
P x:n1 ä
x:k
n-year ,k<n
k Ex
1
Pure Endowment kV x:n 1, k = n
Px:n ä A1
x:k x:k
n-year k Ex
- k Ex
,k<n
Endowment k Vx:n 1, k = n
h Px ä A1
x:k x:k
h-Payment k Ex
- k Ex
,k<h
h Px ä A1
h x:h x:k
Whole Life k Vx k Ex
- k Ex
,k≥h
h Px:n ä A1
x:k x:k
h-Payment k Ex
- k Ex
,k<h<n
h Px:n ä A1
h x:h x:k
n - year Endowment k Vx:n k Ex
- k Ex
,h≤k<n
1, k = n

9.2 Exercises 170
9.2 Exercises
9.1. For a fully continuous 5-year endowment insurance of 1000 on (x):
(i) µx (t) = 0.01 for t > 0

(ii) δ = 0.05
(iii) Level benefit premiums are payable continuously each year.
Calculate the future loss at time 2 assuming (x) dies exactly 6.2 years after
issue.
(A) 355 (B) 400 (C) 445 (D) 490 (E) 535
9.2. For a fully discrete 3-year term insurance of 1000 on (30):
(i) i = 0.08
(ii) Mortality follows: lx = 50(100 - x) for 0 ≤ x ≤ 100.
(iii) Premiums are determined by the equivalence principle.
Calculate the future loss at time 1, given (30) dies in the second year after
policy issue.
(A) 900 (B) 910 (C) 920 (D) 930 (E) 940
9.3. Suppose: lx = 100 - x for 0 ≤ x ≤ 100, and i = 0.06.
Calculate the benefit reserve at time 10 for a fully continuous 15-year

endowment insurance of 1 on (30).
(A) 0.42 (B) 0.46 (C) 0.50 (D) 0.54 (E) 0.58

9.2 Exercises 171
9.4. Which of the following is not equal to the benefit reserve for a fully
discrete whole life insurance of 1 on (35) at the end of policy year 6?
(A) A41 - P35 ä41

ä41
(B) 1 - ä35
(C) [P41 − P35 ]ä41
(D) [1 − A35 ]−1 [A41 − A35 ]

P35 1
(E) 6 E35
[ä35:6 − A35:6 ]
9.5. You are given:
(i) A30:20 = 0.41949

(ii) 10 E40 = 0.63274
(iii) 10 V30:20 , the benefit reserve at time 10 for a fully discrete 20-year
endowment insurance of 1 on (30), equals 0.39040.
1
Calculate: A40:10 .
(A) 0.013 (B) 0.016 (C) 0.019 (D) 0.022 (E) 0.025
9.6. Consider two groups of lives: Group 1 and Group 2. Each group
consists of 1000 lives aged 25.
(i) Each life in Group 1 is issued a fully discrete 15-year term insurance
policy with face amount 10,000 assuming mortality follows the Illustrative
Life Table.
(ii) Each life in Group 2 is issued a fully discrete 15-year term insurance
policy with face amount 10,000 assuming qx = 0.015 for all integer ages x.
(iii) i = 0.06
Calculate the absolute value of the difference between the expected aggre-
gate reserves for Group 1 and Group 2, each at time 10.
(A) 24,900 (B) 25,000 (C) 25,100 (D) 25,200 (E) 25,300

9.2 Exercises 172
9.7. For a fully continuous 5-payment 10-year term insurance of 1000 on

(x):
(i) µx (t) = 0.03 for t > 0

(ii) δ = 0.05
Calculate the benefit reserve at time 4.
(A) 80 (B) 85 (C) 90 (D) 95 (E) 100
9.8. For a 10-year deferred whole life annuity-due of 1 per year on (35):
(i) Level benefit premiums are payable at the beginning of the year during
the first ten years.
(ii) Mortality follows the Illustrative Life Table.
(iii) i = 0.06
Calculate the benefit reserve at the end of the fifth year.
(A) 5.6 (B) 5.8 (C) 6.0 (D) 6.2 (E) 6.4
9.9. For a 10-year deferred whole life insurance of 1 on (35) with benefit
payable at the moment of death:
(i) lx = 500(90 - x) for 0 ≤ x ≤ 90

(ii) δ = 0.05
(iii) Benefit premiums of 0.0274 are payable continuously in each of the
first 10 years.
Calculate the benefit reserve at time 5.
(A) 0.12 (B) 0.16 (C) 0.20 (D) 0.24 (E) 0.28

9.2 Exercises 173
9.10. Consider a special fully discrete whole life insurance on (40):
(i) The death benefit is 1000 if death occurs within 10 years, 2000 if death
occurs after 10 years and before 20 years, and 3000 if death occurs after 20
years.
(ii) The benefit premium is P (1 + k) for year k + 1 where k = 0, 1, ...,9.
The benefit premium is zero after the tenth year.
(iv) i = 0.06
Calculate the benefit reserve at the end of 10 years.
(A) 660 (B) 670 (C) 680 (D) 690 (E) 700
9.11. Consider a special fully discrete life insurance on (x):
(i) The death benefit is 1000(k + 1) for year k + 1 where k = 0, 1, ..., 19.
The death benefit is 20,000 for year k + 1 where k = 20, 21, ...
(ii) There are level annual benefit premiums.
(iii) 20 px = 0.9566
(iv) (IA)x = 5.79
(v) (IA)x+20 = 7.43
(vi) Ax = 0.1531
(vii) Ax+20 = 0.3219
(viii) Ax:201 = 0.3966
Calculate the benefit reserve at the end of 20 years.
(A) 4160 (B) 4200 (C) 4240 (D) 4280 (E) 4320

9.2 Exercises 174
9.12. For a fully continuous whole life insurance of 2500 on (x):
(i) ½
0.05 for 0 ≤ t < 10
µx (t) =
0.08 for t ≥ 10
(ii) δ = 0.05
Calculate the expected value of the future loss random variable at time 10.
(A) 320 (B) 340 (C) 360 (D) 380 (E) 400
9.13. For a 10-payment fully continuous 20-year term insurance of 1000 on

(x):
(i) ½
0.05 for 0 ≤ t < 15
µx (t) =
0.07 for 15 ≤ t < 20
(ii) ½
0.07 for 0 ≤ t < 15
δt =
0.08 for 15 ≤ t < 20
Calculate the benefit reserve at the end of the 14th policy year.
(A) 225 (B) 245 (C) 265 (D) 285 (E) 305
9.14. Suppose mortality follows the Illustrative Life Table, and i = 0.06.
For a 30-year deferred whole life annuity-due of 1 per year on (35), with
level benefit premiums payable at the beginning of the year for each of the
first 30 years, calculate the benefit reserve at time 20.
(A) 3.8 (B) 3.9 (C) 4.0 (D) 4.1 (E) 4.2

9.2 Exercises 175
(i) 10 L is the prospective loss, at time 10, for a fully discrete whole life
insurance of 1 on (50).
(iii) i = 0.06
(iv) Annual premiums are calculated using the equivalence principle.
Calculate var[10 L].
(A) 0.058 (B) 0.063 (C) 0.068 (D) 0.073 (E) 0.078
9.16. Consider a fully discrete 10-year term insurance of 10,000 on (40):
(i) The annual level benefit premium is 35.94.

(iii) i = 0.06
(iv) L8 is the prospective loss for this insurance at time 8.
Calculate: V ar[L8 ].
(A) 840,000 (B) 850,000 (C) 860,000 (D) 870,000 (E) 880,000
9.17. You are given a fully discrete whole life insurance of 5000 on (40):
(i) The level annual benefit premium is 228.78.

(iii) e̊40 = 20.00 and e̊41 = 19.64
(iv) A40 = 0.5214
Calculate the benefit reserve at the end of the first year.
(A) 76 (B) 78 (C) 80 (D) 83 (E) 85

9.2 Exercises 176
(i) The level annual benefit premium for a fully discrete whole life insurance
of 1 on (40) is 0.03.
(ii) The expected present value of a whole life annuity-due of 1 per year
on (40) is 10.
(iii) The benefit reserve at the end of year 10 for a fully discrete whole life
insurance of 1 on (30) is 0.10.
(iv) The benefit reserve at the end of year 20 for a fully discrete whole life
insurance of 1 on (30) is 0.25.
Calculate the level annual benefit premium for a fully discrete whole life
insurance of 1 on (50).
(A) 0.04 (B) 0.05 (C) 0.06 (D) 0.07 (E) 0.08
(i) A fully discrete 2-year deferred, 3-year term insurance of 1000 is issued
to a life aged x.
(ii) The level annual benefit premium is 209.27, and is only payable during
the first two years.
(iii) v = 0.90
(iv) k| qx = 0.05(1 + k) for k = 0, 1, 2, 3, 4; 5| qx = 0.25
Calculate the second least likely value of the prospective loss at time 1.
(A) 450 (B) 520 (C) 600 (D) 690 (E) 740

9.2 Exercises 177
9.20. Each of 100 independent lives age 30 are issued a fully discrete 30-
year endowment insurance of 1000 with level annual benefit premiums:
(i) The level annual benefit premium is 13.90.

(ii) Mortality for each life follows the Illustrative Life Table.
(iii) i = 0.06
(iv) 7 of the 100 lives die during the first ten years.
Using the normal approximation, calculate the fund required at the end
of the first ten years so that there is a 95% probability of covering the total
prospective loss for the remaining policies at the end of ten years.
(A) 17,560 (B) 18,060 (C) 18,560 (D) 19,060 (E) 19,560

9.2 Exercises 178
9.1. A
9.2. B
9.3. D
9.4. E
9.5. A
9.6. A
9.7. D
9.8. C
9.9. B
9.10. D
9.11. A
9.12. E
9.13. C
9.14. D
9.15. D
9.16. D
9.17. D
9.18. B
9.19. C
9.20. A


• Exam MLC, Fall 2013: #7, 17
• Exam 3L, Fall 2013: #14, 15
• Exam 3L, Fall 2012: #14, 15
• Exam 3L, Spring 2012: #14, 15
• Exam MLC, Sample Questions: #10, 27, 68, 77, 78, 115, 132, 169, 203,
208, 222, 227, 287
• Exam 3L, Fall 2011: #14, 15
• Exam 3L, Spring 2011: #14, 15
• Exam 3L, Fall 2010: #15, 16
• Exam 3L, Fall 2009: #14
• Exam 3L, Spring 2009: #14, 15
• Exam 3L, Fall 2008: #23, 24

180
10 RESERVES II (Exam MLC Only)
10.1 Key Concepts

Continuous Insurance of 1 on (x) with Premiums Payable on an
m-thly Basis:
Insurance of 1 on (x) in RESERVES I: KEY CONCEPTS and
replacing the continuous premium annuity with an m-thly annuity-due.
• For example, a continuous n-year term insurance of 1 on (x) with premi-

ums payable on an m-thly basis is:
Continuous Benefit Reserve Prospective Benefit

with m-thly Premiums
1 (m)
n-year Āx+k:n−k - P (m) (Āx:n
1
)äx+k:n−k , k < n
(m) 1
Term kV (Āx:n ) 0, k = n
• You do NOT have to know the actuarial notation for m-thly benefit
reserves; k V or k V n is sufficient.
1 2 1
• k < n means that k = 0, m, m, ..., n - m.
• The prospective benefit reserve for an annual n-year term insurance of 1

on (x) with premiums payable on an m-thly basis is:
(
1 (m)
(m)1 Ax+k:n−k − P (m)1x:n äx+k:n−k for k < n
kV x:n =
0 for k = n

Gross Premium Reserves:
• The first step to determining the gross premium reserve is to determine

the appropriate gross future loss at time t random variable (conditional
on survival of the policyholder to time t): Lt = t L
= Present value of future benefits at time t + Present value of future
expenses at time t - Present value of future premiums at time t
= P V F B@t + P V F E@t - P V F P @t
The gross future loss at time t may be written as Lgt .
If expenses are ignored, the loss is a net future loss at time t.
The policy value at time t is called the gross premium reserve at time
t. The gross premium reserve represents the accumulated fund that the
insurer needs to have at some future time that in addition to future
gross premiums should, on average, fund the future benefits and expenses
associated with the policy.
• The general prospective gross premium reserve formula is:
tV = E(Lgt ) = E(P V F B@t) + E(P V F E@t) - E(P V F P @t).
The gross premium reserve is a balancing item in funding a policy:

tV + E(P V F P @t) = E(P V F B@t) + E(P V F E@t).

• The retrospective gross premium reserve formula is:

E[PVFP on [0, t]] E[PVFB on [0, t]]+E[PVFE on [0, t]]
tV = t Ex
- t Ex
.
This formula emphasizes that the gross premium reserve is the accumu-
lated fund that the insurer needs to have by time t that along with future
premiums, on average, will fund the future benefits and expenses of a pol-
icy. The accumulated fund is the accumulated premiums from issue up
to time t ( E[PVFPt Eon
x
[0, t]]
) less the accumulated value of any benefits and
expenses paid from issue up to time t ( E[PVFB on [0, t]]+E[PVFE
t Ex
on [0, t]]
).
• The prospective gross premium reserve will equal the retrospective gross
premium reserve as long as the equivalence principle and the same basis
(interest, mortality, expenses) are used.
• The gross premium reserve helps fund both the future benefits and ex-
penses of a policy. With the equivalence principle used to calculate pre-
miums, it makes sense that (t V g ) can be decomposed into the sum of two
reserves: a benefit reserve that helps fund just the benefits (t V n ) plus an
expense reserve that helps fund just the expenses (t V e ):
g
tV = tV n + tV e.
This decomposition mirrors the gross premium decomposition:

gross premium = benefit premium + expense premium.
Prospectively: t V e = EP V F E@t - EP V (future expense premiums@t).

Reserve Recursions:
• Now, let us consider fully discrete gross premium reserve recursions.
• Consider a fully discrete life insurance on (x). For k = 0, 1, 2, ...; let:
– bk+1 = death benefit payable at time k + 1.

– Gk = gross premium payable at time k.
– ek = total premium-related expense payable at time k.
– Ek+1 = claims settlement expense payable at the same time as the
death benefit (at time k + 1)
– ik = effective annual interest rate between times k and k + 1.
– qx+k = mortality rate during year k + 1.
– k V = gross premium reserve for a policy in force at time k.
• Then, the fully discrete gross premium reserve recursion is:

(k V + Gk − ek )(1 + ik ) = (qx+k )(bk+1 + Ek+1 ) + (px+k )(k+1 V )
• There is a very logical interpretation of the gross premium reserve recur-

sion:
“The funds the insurer has at the start of year k + 1 (k V + Gk − ek )
accumulated with interest to the end of the year should be sufficient to
provide the average benefit due at time k + 1: the death benefit and the
corresponding settlement expense if the policyholder dies during the year
(bk+1 + Ek+1 with probability qx+k ) and the gross premium reserve if the
policyholder survives the year (k+1 V with probability px+k ).”
• If the gross premium is determined using the equivalence principle:

0V = 0.

• An alternative way of expressing the fully discrete gross premium reserve

recursion in terms of the net amount at risk for year k + 1, [bk+1 + Ek+1
- k+1 V ], is:
(k V + Gk − ek )(1 + ik ) = k+1 V + qx+k [bk+1 + Ek+1 - k+1 V ]
“The funds the insurer has at the start of year k + 1 (k V + Gk − ek )
accumulated with interest to the end of the year should be sufficient
to definitely provide the gross premium reserve at time k + 1. If the
policyholder dies during the year (with probability qx+k ), the insurer also
needs to fund the net amount at risk (bk+1 + Ek+1 - k+1 V ) in order to
provide the death benefit bk+1 and settlement expense Ek+1 at time k +
1.”
• If ek = Ek+1 = 0 and Gk = Pk for k = 0, 1, 2, ..., k V is interpreted as a

net premium reserve and we obtain the fully discrete
net premium reserve recursion:
(k V + Pk )(1 + ik ) = (qx+k )(bk+1 ) + (px+k )(k+1 V )
=⇒ (k V + Pk )(1 + ik ) = k+1 V + qx+k [bk+1 - k+1 V ].
• If Pk = P , the equivalence principle is used to determine P , and ik = i,

the fully discrete benefit reserve recursion can be expressed as:
Pk−1 k−j
k V = P s̈k - j=0 vqx+j (bj+1 − j+1 V )(1 + i)
Some important special cases:

– bj+1 = j+1 V =⇒ kV = P s̈k
Pk−1
– bj+1 = S + j+1 V =⇒ k V = P s̈k - S j=0 vqx+j (1 + i)k−j
– bj+1 = S + j+1 V and qx+j = q =⇒ kV = (P − Svq)s̈k

Interim Benefit Reserves:
• Consider a fully discrete life insurance on (x).

Let k = 0, 1, 2, ... and 0 < s < 1.
• k+s V is the interim benefit reserve, the policy value at some time between
policy anniversaries.
• The exact “recursion” formula for the interim benefit reserve is:
k+s V = v 1−s (bk+1 )1−s qx+k+s + v 1−s (k+1 V )1−s px+k+s .
Incorporation of expenses is relatively straightforward, for example:
k+s V = v 1−s (bk+1 + Ek+1 )1−s qx+k+s + v 1−s (k+1 V )1−s px+k+s .
• If we assume a uniform distribution of deaths within each year of age,

the interim benefit reserve formula becomes:
v 1−s
k+s V = 1−sqx+k [(1 − s)(k V + πk )(1 + i) + s(k+1 V )px+k ].
Furthermore, if i ≈ 0 and qx+k ≈ 0, the interim benefit reserve formula

with a uniform distribution of deaths becomes the traditional actuarial
approximation:
k+s V ≈ (1 − s)(k V + πk ) + s(k+1 V ).
• Similar formulas for policies with m-thly benefits and/or premiums can
be similarly reasoned out when necessary: handle these on a case-by-case
basis.

Thiele’s Differential Equation:
• Thiele’s Differential Equation is a continuous version of the gross pre-

mium reserve recursion.
• Consider a fully continuous life insurance on (x). For t > 0; let:
– bt = death benefit payable at time t.

– Gt = gross premium rate at time t.
– et = total premium-related expense rate at time t.
– Et = claims settlement expense payable at the same time as the death
benefit (at time t)
– δt = the force of interest at time t.
– µx+t = the force of mortality at age x + t.
– t V = gross premium reserve for a policy in force at time t.
• Thiele’s Differential Equation is the following:

d
dt [t V ] = δtt V + Gt - et - (bt + Et − t V )µx+t .
• There is a very logical interpretation of Thiele’s Differential Equation:

“The change in the value of the reserve at time t is due to interest earned
(positive change), the gross premium rate (positive change), and the
benefits and expenses payable (negative change, accounting for death via
µx+t when necessary).”
• Euler’s method provides a discrete approximation:

t+h V − t V ≈ h[δtt V + Gt - et - (bt + Et − t V )µx+t ].
By picking small time intervals for h (an m-th of a year), we can approx-
imate other reserve values given one reserve value.
• The benefit reserve version is obtained by letting et = Et = 0 and Gt =

Pt for all t.

Full Preliminary Term Method:
• The full preliminary term method modifies benefit reserves so that they
can approximate gross premium reserves. Gross premium reserves tend
to be lower than benefit reserves because the expense reserve is often
negative.
The method is such that: a life insurance on (x) with a term of n years
is decomposed into two separate insurances: (i) a 1-year term insurance
on (x) that provides coverage for the first policy year plus (ii) the same
type of life insurance as the original except on (x + 1) with a term of n
- 1 years.
• Theoretically, a lower benefit premium could be charged for the first year
with higher level benefit premiums charged in the second and subsequent
years. In reality, we only care about the full preliminary term reserves
denoted as k V F P T : modified benefit reserves that lower the reserve value
to approximate gross premium reserves.
• For example, consider the full preliminary term method on a fully discrete
whole life insurance of S on (x).
– Under the full preliminary term method, this insurance is conceptual-

ized as a 1-year term insurance of S on (x) plus a whole life insurance
of S on (x + 1). Note that n = ∞ for a whole life insurance.
– Based on the decomposition, the benefit premium for the first year is
1
SP x:1 = Svqx , and each benefit premium in second and subsequent
years is SPx+1 .
– 0 V F P T = 0 (via equivalence principle) and 1 V F P T = 0 (1-year term
insurance has expired).
– k V F P T = the benefit reserve at time k - 1 for a fully discrete whole
life insurance of S on (x + 1), for k = 2, 3, ... Note the offset of one
year: under the decomposition, the whole life insurance on (x + 1)
takes effect one year after policy issue, so time k years after policy
issue corresponds to the whole life insurance on (x + 1) at time k - 1.

Policy Alterations:
• After a policy has been in force for some time, a policyholder may want
to alter the policy terms. For example, the policyholder may want to:
– Immediately cancel the policy. This will result in a lump sum payment
to the policyholder called the surrender value or cash value.
– Keep the policy but pay no more premiums; the policyholder wants a
paid-up policy. Typically, the face amount of the paid-up policy will
be smaller than the face amount of the original policy.
– Convert the policy into a different type; for example, a whole life
policy may be converted into an endowment insurance.
– Change the amount of benefits and/or premiums.
• In a general problem involving policy alterations, you will need to solve

for a parameter (such as changed benefits and premiums) of the altered
policy. Suppose the policy is altered t years after issue. The general
approach is to:
– Determine the surrender value for the original policy, Ct . This will
typically be a percentage of the reserve or asset share at time t.
– If the policy is canceled at time t, then Ct is paid to the policyholder
and we are done. If the policy is otherwise altered at time t, treat Ct
as an extra-preliminary premium that will help fund the purchase of
the altered policy.
– Let * denoted the altered policy. Use the equivalence principle at
time t to relate the benefits, expenses, and premiums (including the
extra-preliminary premium) for the altered policy:
Ct + E ∗ [P V F P @t] = E ∗ [P V F B@t] + E ∗ [P V F E@t].
You can use this equivalence relation to solve for the required param-
eter of the altered policy. Handle these problems on a case-by-case
basis.

Gain by Source:
This is best considered with an example. The following is an older version

of Problem 300 from the SOA Exam MLC Sample Questions:
Problem 300) For a fully discrete 20-year term life insurance on (40):
• The death benefit is 10,000.
• The death benefit is payable at the end of the year of death.
• Values in year 4:
Anticipated Actual
Gross annual premium 90 90
Expenses as a percent of premium 0.030 0.025
q43 0.003 0.002
Annual effective rate of interest 0.05 0.04
• Expected value of the present value of future losses random variable based
on assumed values in years 3 and later.
End of Year Expected Value

3 100
4 125
A company issued the 20-year term life insurance to 1000 lives age 40
with independent future lifetimes. At the end of the 3rd year 990 insurances
remain in force.
Calculate the total gain from mortality, interest and expenses in year 4
from the 990 insurances.

Problem 300) Solution:
• For policy year k + 1, the profit per inforce policy at the start of the year
can be calculated as:
P rk+1 = (k V + Gk − ek )(1 + ik ) - (qx+k )(bk+1 + Ek+1 ) - (px+k )(k+1 V ).
• The total gain from mortality, interest, and expenses in year k + 1 per
inforce policy is:
Actual Anticipated
P rk+1 - P rk+1 .
• So, we can calculate the total gain in year 4 per inforce policy as follows:
• P r4Anticipated
= [100 + 90 − (0.030)(90)](1.05) - (0.003)(10, 000) - (1 − 0.003)(125)
= 42.04.
• P r4Actual
= [100 + 90 − (0.025)(90)](1.04) - (0.002)(10, 000) - (1 − 0.002)(125)
= 50.51.
• The total gain in year 4 per inforce policy is:

P r4Actual - P r4Anticipated = 50.51 - 42.04 = 8.47.
• The total gain in year 4 is:
(990)(8.47) = 8385.30.

Problem 300) Gain By Source:

The contribution of each of mortality, interest, and expenses to total gain
in year 4 can be determined.
• Gain Due to Mortality in Year 4:

The gain due to mortality in year 4 per inforce policy is:
(0.003 - 0.002)(10,000 - 125) = 9.875.
Therefore, the gain due to mortality in year 4 is: (990)(9.875) = 9776.25.
• Gain Due to Interest in Year 4:

The gain due to interest in year 4 per inforce policy is:
(0.04 - 0.05)[100 + 90 - (0.03)(90)] = -1.873.
Therefore, the gain due to interest is: (990)(-1.873) = -1854.27.
Note that as expenses have not been considered yet, you should use the
anticipated expense values in this part of the calculation.
• Gain Due to Expenses in Year 4:

The gain due to expenses in year 4 per inforce policy is:
[(0.030)(90) - (0.025)(90)](1.04) = 0.468.
Therefore, the gain due to expenses in year 4 is: (990)(0.468) = 463.32.
Note that as interest has already been considered, you should use the
actual interest rate in this part of the calculation.
The total gain in year 4 per inforce policy is: 9.875 - 1.873 + 0.468 = 8.47,
and the total gain in year 4 is: 9776.25 - 1854.27 + 463.32 = 8385.30, which
both agree with the previously obtained values.
If you were to calculate each of the gains in a different order; such as

interest, mortality, and expenses; you would obtain slightly different values.

10.2 Exercises 192
10.2 Exercises
10.1. For a whole life insurance of 1000 on (35) payable at the moment of
death:
(i) Benefit premiums are payable at the beginning of each quarter of a year.
(ii) Mortality follows the Illustrative Life Table, and i = 0.06.
Calculate the benefit reserve at the end of year 15.
(A) 110 (B) 120 (C) 130 (D) 140 (E) 150
10.2. For a fully discrete two-year term insurance of 5000 on (x):
(i) i = 0.10
(ii) qx+1 = 0.25
(iii) The level annual net premium is 929.13.
Calculate: qx .
(A) 0.16 (B) 0.17 (C) 0.18 (D) 0.19 (E) 0.20
10.3. For a fully discrete whole life insurance of S on (25):
(i) There are level annual benefit premiums.

(ii) i = 0.045
(iii)
k q25+k Benefit Reserve at time k

20 0.001805 1228.42
21 0.001940 1316.34
22 0.002080 1407.22
(A) 38 (B) 42 (C) 46 (D) 50 (E) 54

10.2 Exercises 193
10.4. For a special fully discrete whole life insurance on (40):
(i) The death benefit is 100,000 if death occurs before age 65, otherwise
the death benefit is 200,000.
(ii) Gross premiums are determined using the equivalence principle.
(iii) Expenses are 20% of the first year’s gross premium, and 5% of the
second and subsequent year’s gross premium. These expenses are payable at
the beginning of the year.
(iv) ä40 = 19.20, ä41 = 19.00, ä65 = 13.40, 25| ä40 = 4.02
(v) p40 = 0.99
Calculate the gross premium reserve at the end of the 25th year.
(A) 72,400 (B) 74,600 (C) 76,800 (D) 78,200 (E) 80,400
10.5. For a fully discrete whole life insurance of 100,000 on a select life
aged 60:

(ii) i = 0.05.
(iii) Gross premiums are determined using the equivalence principle.
(iv) Commissions are 50% of the gross premium in the first year, and 5%
of the gross premium in renewal years. They are paid at the same time as
the premium.
(v) Settlement expenses are 100, and are paid at the same time as the
(vi) Other expenses are 200 in the first year, and 10 in renewal years. They
are all payable at the beginning of the year.
Calculate the gross premium reserve at the end of the tenth year.
(A) 16,600 (B) 17,100 (C) 17,600 (D) 18,100 (E) 18,600

10.2 Exercises 194
10.6. For a fully discrete 3-year term insurance of 10,000 on (x):
(i) The level annual gross premium is calculated using the equivalence
principle.
(ii) qx+k = 0.10 + 0.01k for k = 0, 1, 2.
(iii) There is a per policy expense of 15.00 at the beginning of the first
year, and 7.50 at the beginning of each renewal year. There is a percent of
premium expense of 25% of the gross premium at the beginning of the first
year, and 5% of the gross premium at the beginning of each renewal year.
(iv) i = 0.05
Calculate the difference between the gross premium reserve and the benefit
reserve at the end of policy year 1.
(A) - 312 (B) - 177 (C) - 27 (D) 0 (E) 119
10.7. Consider the setup provided in Exercise 10.6. Further assume deaths
are uniformly distributed over each year of age. Calculate the gross premium
reserve at time 0.75.
(A) - 693 (B) - 78 (C) 0 (D) 192 (E) 886
10.8. Consider a fully discrete 5-year term insurance of 3000 on (x):
(i) Level gross premiums are determined using the equivalence principle.
(ii) i = 0.05
(iii) A 20% of gross premium expense is payable at the beginning of the
first year. A 5% of gross premium expense is payable at the beginning of
each renewal year. A settlement expense of 20 is payable at the same time
as the death benefit.
(iv) qx+2 = 0.051 and qx+3 = 0.054.
(v) Let k V denote the gross premium reserve at time k. Then: 2 V = 8.20
and 3 V = 15.58.
Calculate: 4 V .
(A) 10 (B) 12 (C) 14 (D) 16 (E) 18

10.2 Exercises 195
10.9. Consider a special fully discrete 5-year term insurance on (40):

(i) The death benefit is: bk+1 = 1000(k + 1) for k = 0, 1, ..., 4.
(ii) Level gross premiums are payable.
(iii) There is a percent of premium expense of 20% of gross premium for
the first year. The percent of premium expense is 5% of gross premium in
renewal years. These expenses are payable at the beginning of the year.
(iv) There is a claims settlement expense of 50 payable at the time the
(v) The gross premium reserve at the end of the fourth year is 6.275.
(vi) Mortality follows the Illustrative Life Table, and i = 0.06.
Calculate the net amount at risk for the third year.
(A) 3030 (B) 3040 (C) 3050 (D) 3060 (E) 3070
10.10. Consider a special fully discrete 5-year term insurance on (x):
(i) The death benefit is 2500 plus the net premium reserve at the end of
the year of death.
(ii) There are level annual net premiums.
(iii) i = 0.04
(iv) qx+k = 0.01 + 0.005k for k = 0, 1, 2, 3, 4.
Calculate the net premium reserve at the end of the first year.
(A) 21 (B) 22 (C) 23 (D) 24 (E) 25

10.2 Exercises 196
10.11. For a special fully discrete 3-year endowment insurance on (65):
(i) The death benefit is 5000 plus the benefit reserve at the end of the year
of death. The pure endowment benefit is 5000.
(ii) There are level annual benefit premiums.
(iii) p65+h = 0.97 for h = 0, 1, 2
(iv) i = 0.06
Calculate the benefit reserve at the end of the first year.
(A) 1550 (B) 1570 (C) 1590 (D) 1610 (E) 1630
10.12. For a special fully discrete 3-year term insurance on [45]:
(i) The death benefit is 10,000 plus the benefit reserve at the end of the
year of death.
(ii) Mortality follows the Standard Select Survival Model, i = 0.05.
(A) 7 (B) 10 (C) 13 (D) 16 (E) 19
(i) i = 0.06
(ii) The level annual gross premium is 5.04.
(iii) The following expenses, paid at the beginning of the year:
First Year Renewal Years

Percentage of Premium 70% 10%
Per Policy 2.20 0.45
Calculate the gross future loss at time 1, given (40) dies in the third year
after policy issue.
(A) 850 (B) 860 (C) 870 (D) 880 (E) 890

10.2 Exercises 197
10.14. Consider a fully discrete 20-year term insurance of 5000 on (45):

(ii) i = 0.06
(iv) The level annual net premium is 38.21.
Calculate the net premium reserve at time 9.3.
(A) 160 (B) 170 (C) 180 (D) 190 (E) 200
10.15. Consider a 10-year term insurance of 500,000 on a select life aged

50.
(i) The death benefit is payable at the end of the month of death.
(ii) Level gross premiums of 460 are payable at the beginning of each
quarter of a year for at most five years.
(iii) There is a 10% of gross premium expense payable at the beginning of
each quarter of a year.
(iv) Mortality follows the Standard Select Survival Model, i = 0.05
(v) Deaths are uniformly distributed within each year of age.
(vi) The benefit reserve at time 2.75 is 3091.
Calculate the benefit reserve at time 2 65 .
(A) 3320 (B) 3380 (C) 3340 (D) 3400 (E) 3460

10.2 Exercises 198
10.16. For a fully continuous 5-year term insurance of 50,000 on (35):
(i) The annual rate of gross premium is 100.

(ii) Expenses are 2.5% of gross premium paid continuously each year plus
a settlement expense of 80 paid at the same time as the death benefit.
(iii) µx = 0.001(1.02x ) for x ≥ 0
(iv) δ = 0.04
Using Euler’s method with step 0.10 to numerically solve Thiele’s Differ-
ential Equation, approximate 4.80 V .
(A) 2.2 (B) 2.3 (C) 2.4 (D) 2.5 (E) 2.6
10.17. Consider a fully continuous 10-payment 20-year endowment insur-

ance of 10,000 on (45):
(i) The annual premium rate is 488.

(ii) µx = 0.00022 + 0.0000027(1.124x ) for x ≥ 0
(iii) i = 0.05
Using Euler’s method with step 0.2 to numerically solve Thiele’s Differen-
tial Equation, approximate 19.8 V .
(A) 9900 (B) 9920 (C) 9940 (D) 9960 (E) 9980
10.18. For a special fully continuous 10-year endowment insurance of 1000

on a select life aged 30:
(i) Mortality follows the Standard Select Survival Model, and i = 0.05.
(ii) The annual rate of gross premium is 150.
(iii) The annual rate of per policy expense is 50.
Estimate the gross premium reserve at time 9.8 using Euler’s method with
step size 0.1 to numerically solve Thiele’s Differential Equation.
(A) 950 (B) 960 (C) 970 (D) 980 (E) 990

10.2 Exercises 199
10.19. Consider a continuous 10-year deferred 10-year temporary life an-

nuity of 2400 per year on a select life aged 30:
(i) A level net premium of 1465.62 is payable continuously each year during
the first 10 years.
(ii) Mortality follows the Standard Select Survival Model, and i = 0.05.
Which of the following gives Thiele’s Differential Equation at t = 11.4?

d
(A) dt [t V ]|t=11.4 = ln(1.05)11.4 V + 1465.62 + 11.4 V µ41.4
d
(B) dt [t V ]|t=11.4 = ln(1.05)11.4 V + 1465.62 - 11.4 V µ41.4
d
(C) dt [t V ]|t=11.4 = ln(1.05)11.4 V - 2400 + 11.4 V µ41.4
d
(D) dt [t V ]|t=11.4 = ln(1.05)11.4 V - 2400 - 11.4 V µ41.4
d
(E) dt [t V ]|t=11.4 = ln(1.05)11.4 V - 934.38 + 11.4 V µ41.4
10.20. Consider the setup provided in Exercise 10.19. Estimate the benefit
reserve at time 9.8 using Euler’s method with step size 0.2 to numerically
solve Thiele’s Differential Equation.
(A) 18,000 (B) 18,500 (C) 19,000 (D) 19,500 (E) 20,000
(i) Mortality follows the Illustrative Life Table and i = 0.06

(ii) ä71:29 = 8.29634
Calculate the full preliminary term reserve at the end of year 10 for a fully
discrete 30-year endowment insurance of 1000 on (70).
(A) 250 (B) 260 (C) 270 (D) 280 (E) 290
10.22. Consider a fully discrete 10-year term insurance of 1000 on (40).

Mortality follows the Illustrative Life Table, and i = 0.06. Using the full
preliminary term method, calculate the benefit premium in renewal years.
(A) 3.68 (B) 3.74 (C) 3.80 (D) 3.86 (E) 3.94

10.2 Exercises 200
(i) q35 = 0.001

(ii) 20 p35 = 0.960
(iii) 10,000A35 = 1417; 10,000A36 = 1473; 10,000A55 = 3000
(iv) ä44:11 = 8.83
Calculate the full preliminary term reserve for this insurance at the end of
year 9.
(A) 3180 (B) 3220 (C) 3260 (D) 3300 (E) 3340
10.24. You are given a fully discrete whole life insurance of 5000 on (25).
Mortality follows the Illustrative Life Table, and i = 0.06.
Calculate the difference between the net premium reserve at time 5 and
the full preliminary term reserve at time 5.
(A) 16 (B) 17 (C) 18 (D) 19 (E) 20
10.25. For a fully discrete whole life insurance of 2500 on (45):
(i) The full preliminary term reserve at the end of year 2 is 40.98.
(ii) ä45 = 18.54 and ä47 = 18.00
(iii) q45 = 0.015
Calculate the first year benefit premium using the full preliminary term
method.
(A) 32 (B) 34 (C) 36 (D) 38 (E) 40

10.2 Exercises 201
10.26. Consider a fully discrete 20-year endowment insurance of 20,000 on

(45) with level annual net premiums that has been in force for 8 years:
(i) Mortality follows the Illustrative Life Table, and i = 0.06.

(ii) The surrender value is equal to the net premium reserve.
The policyholder wants to stop paying premiums after 8 years. In ex-

change, the insurer decides to reduce the benefit of the endowment insurance
that is payable during the remainder of the 20-year period.
Using the equivalence principle at time 8, calculate the reduced benefit
amount.
(A) 8240 (B) 8930 (C) 9620 (D) 10,310 (E) 11,000
10.27. For a whole life insurance of 50,000 on (30) payable at the end of
the month of death:
(i) Gross premiums of 53.90, determined using the equivalence principle,

are payable at the beginning of each month.
(ii) There is a per-policy expense of 100 payable at the beginning of the
first month. There is a per-policy expense of 25 payable at the beginning of
the second and subsequent months.
(iv) The surrender value is the gross premium reserve minus 100.
(v) Mortality follows the Illustrative Life Table, and i = 0.06.
After 10 years, the policyholder wants to immediately convert the whole life
insurance into a 10-year term insurance of S payable at the end of the month
of death. Furthermore, the policyholder wants to pay no more premiums;
that is, the policyholder wants a paid-up 10-year term insurance.
The insurer now anticipates a per-policy expense of 20 will be payable at
the beginning of each month while the 10-year term insurance is in force.
Using the equivalence principle at time 10, calculate: S.
(A) 49,000 (B) 49,700 (C) 50,400 (D) 51,100 (E) 51,800

10.2 Exercises 202
10.28. For a fully discrete 20-year endowment insurance on (35) that has
been in force for 10 years:
(i) The death benefit is 5000, and the pure endowment benefit is 7500.
(iii) i = 0.06
(iv) The surrender value is equal to 95% of the benefit reserve.
The policyholder decides to stop paying premiums after 10 years. The

death benefit remains at 5000 but the pure endowment benefit is reduced.
Using the equivalence principle at time 10, calculate the reduced pure
endowment benefit.
(A) 4400 (B) 4800 (C) 5200 (D) 5600 (E) 6000
10.29. For a fully discrete whole life insurance of 5000 on [51]:
(i) Mortality follows the Standard Select Survival Model, and i = 0.05.
(ii) The level annual gross premium, determined using the equivalence
principle at issue, is 83.16.
(iii) Expenses are 40% of gross premium plus 50.00 in the first year, and
10% of gross premium plus 12.50 in renewal years. All expenses are paid at
the beginning of the year.
(iv) The surrender value is equal to the gross premium reserve.
After 12 years, the policyholder wants the face amount of the policy dou-
bled. As a result, the level annual gross premium will be increased, including
the gross premium payable at the beginning of the thirteenth year.
Using the equivalence principle (at time 12), calculate the increased level
annual gross premium.
(A) 170 (B) 190 (C) 210 (D) 230 (E) 250

10.2 Exercises 203
10.30. Consider a portolio of fully discrete whole life insurances of 100,000

on 1000 select lives each aged 30. For each policy, the level annual gross
premium, calculated using the equivalence principle, is 474.29.
Each policy is issued assuming (i) mortality follows the Standard Select
Survival Model, (ii) an effective annual interest rate of 0.05, and (iii) expenses
at the start of each year are 10% of premium plus 30. Annual profits are
based on gross premium reserves calculated using the above assumed values
for mortality, interest, and expenses.
Per policy, what is actually experienced during the first year is (i) a mor-
tality rate of 0.0001, (ii) an effective annual interest rate of 0.07, and (iii)
expenses at the start of each year of 10% of premium plus 50.
Calculate the total gain from mortality, interest, and expenses during the
first year.
(A) 3380 (B) 3400 (C) 3420 (D) 3440 (E) 3460
10.31. Consider the setup provided in Exercise 10.30.

Calculate the gain from interest during the first year, assuming that gains
by source are calculated in the order of mortality, interest, and expenses.
(A) 7370 (B) 7480 (C) 7540 (D) 7940 (E) 8120

10.2 Exercises 204
10.32. For a special fully discrete 10-year deferred whole life insurance of
100,000 on (x):
(i) The level annual gross premium is 3600, and is payable at the beginning
of the year during the deferral period.
(ii) It is expected that: (a) i = 0.05, (b) qx+k = 0.005 + 0.001k for k = 0,
1, 2, ...,19, and (c) there is an expense of 0.80 per 1000 of insurance at the
beginning of the first year, an expense of 0.20 per 1000 of insurance at the
beginning of each renewal year, and a settlement expense of 0.25 per 1000 of
insurance payable at the same time as the death benefit.
(iii) Based on expected interest and expected mortality: Ax+9 = 0.27.
(iv) During the eleventh year, actual experience is: (a) i = 0.07, (b) qx+10
= 0.015, and (c) there is an expense of 0.25 per 1000 of insurance at the
beginning of the year and a settlement expense of 0.20 per 1000 of insurance
payable at the same time as the death benefit.
(v) Annual profits are based on prospective gross premium reserves calcu-
lated using the above expected values for mortality, interest, and expenses.
For this policy, calculate the gain from interest during the eleventh year,
assuming that gains by source are calculated in the order of interest, expenses,
and mortality.
(A) 520 (B) 550 (C) 580 (D) 610 (E) 640

10.2 Exercises 205
10.33. Your company issues 10-year deferred whole life annuities-due of

1000 per year, payable annually, to a group of lives each aged 55. For each
annuity:
(i) Level gross premiums of 700 are payable at the beginning of the year
during the first ten years.
(ii) It is expected that mortality follows the Illustrative Life Table.
(iii) It is expected that i = 0.06.
(iv) It is expected that there will be a 5% of gross premium expense,
payable upon receipt of premium.
(v) Annual profits are based on prospective gross premium reserves calcu-
lated using the above expected values for mortality, interest, and expenses.
During year 11, the actual experience is as follows:
(a) There are 480 annuities inforce at the beginning of the year.
(b) There are six deaths.
(c) Interest earned is 0.07.
(d) Expenses equal 6% of gross premium.
Calculate the total gain due to interest, mortality, and expenses for year
11.
(A) 1910 (B) 1930 (C) 1950 (D) 1970 (E) 1990
10.34. Consider the setup provided in Exercise 10.33.

Calculate the gain from mortality in year 11, assuming that gains by source
are calculated in the order of interest, mortality, and expenses.
(A) -42,000 (B) -40,800 (C) -35,560 (D) - 33,240 (E) -31,010

10.2 Exercises 206
10.1. D 10.26. D
10.2. B 10.27. B
10.3. C 10.28. A
10.4. B 10.29. C
10.5. E 10.30. C
10.6. B 10.31. D
10.7. D 10.32. B
10.8. C 10.33. A
10.9. B 10.34. B
10.10. D
10.11. B
10.12. A
10.13. D
10.14. D
10.15. E
10.16. D
10.17. A
10.18. C
10.19. C
10.20. B
10.21. E
10.22. B
10.23. A
10.24. E
10.25. C


• Exam MLC, Fall 2013: #6, 8
• Exam MLC, Spring 2013: #9, 16, 22
• Exam MLC, Fall 2012: #5, 6, 17, 18, 23
• Exam MLC, Spring 2012: #8, 9, 17, 21
108, 110, 118, 156, 162, 177, 185, 199, 214, 230, 243, 274, 275, 277, 288,
299, 300, 302, 303, 304, 305, 306, 307

208
11 MULTIPLE STATE MODELS (Exam MLC Only)
11.1 Key Concepts

Markov Representation:
Consider a multiple state model or multi-state model. This is a model with
(n + 1) states numbered 0, 1, ..., n. At any given time, an individual can be
in exactly one of the (n + 1) states. It is possible over time for an individual
to transition between states.
Examples of common multi-state models are provided in

Examples of Multiple State Models from Chapter 8 of Dickson et al.; pro-
vided in Appendix B of this study supplement.
Multiple decrement and multiple life models are special types of multiple
state models encountered in actuarial science.
Let: i, j = 0, 1, 2, ..., n:
ij
• µij
x (t) = µx+t denotes the force of transition from State i to State j at
age x + t. It is always true that µiix (t) = 0.
• t pij
x denotes the transition probability that someone in State i at age x
will be in State j at age x + t.

• Multi-state models are subject to the following assumptions (these are

called the “standard assumptions”), unless otherwise indicated:
– The state process is a Markov process; that is, t pij

x does not depend
on the states the life was in prior to age x.
– The probability of 2 or more transitions in any period of length h
years is o(h). A function g(h) is said to be o(h) if:
g(h)
h → 0 as h → 0.
– t pij
x is a differentiable function of t, and for i 6= j:
ij
h px
µij
x = limh→0+ h =⇒ ij
h px = hµij
x + o(h).
• If t pij
x does not depend on x, the model is homogeneous. Otherwise, the
model is non-homogeneous.
• 0 piix = 1 and 0 pij

x = 0 for i 6= j
“If (x) is in State i, (x) must be in State i 0 years from now.”
• t pi0 i1 in
x + t px + · · · t px = 1
“If (x) is in State i (which is one of States 0, 1, ..., n), then (x) must be
in one of the (n + 1) States t years from now.
• Transition probability calculations can be handled discretely:

ij
Pn ik kj
u+t px = k=0 (u px )(t px+u ).
• t piix denotes the probability that someone in State i at age x will continuously
be in State i up to age x + t.
R t Pn ij
• t piix = exp[− 0 j=0,j6=i µx+s ds]

• Note: t piix is not necessarily equal to t piix . t piix is the probability that (x)
in State i will be in State i at age x + t; it is possible that (x) could
leave State i and then re-enter it by age x + t which is not allowed in
the probability t piix . Thus:
ii
t px = t piix + o(t) =⇒ t piix ≤ t piix .
• Furthermore:
P
1 - h piix = h nj=0,j6=i µij
x + o(h).
Both sides of the above equation are equivalent to the probability that
the process does leave State i at some time between ages x and x + h,
possibly returning to State i prior to age x + h.
• Consider the Permanent Disability Model in Examples of Multiple State

Models.
– By conditioning on the state the life is in at age x + t:

01
t+h px = t p01 11 00 01
x h px+t + t px hµx+t + o(h)
01
=⇒ t+h px = t p01 11 00 01
x h px+t + t px hµx+t + o(h)
01
Ru
=⇒ u px = 0 t p00 01 11
x (µx+t )u−t px+t dt.
“The probability that a healthy life aged x is disabled at age x + u

can be obtained by picking a time t within the u years. (x) needs
to remain healthy until time t (with probability t p00 x ), immediately
transition from healthy to disabled at age x + t (at rate µ01 x+t ), and
then remain disabled for the remaining u − t years (until age x + u,
with probability u−t p11
x+t ). Considering all possible times t within u
years via integration yields the desired probability.”
– With constant forces of transition:

01 02 12 )u
01 01 −µ 12
u 1−e−(µ +µ −µ
u px = µ e µ01 +µ02 −µ12 , where µ01 + µ02 6= µ12 .

Kolmogorov’s Forward Equations:
• Kolmogorov’s Forward Equations are differential equations for the tran-

sition probabilities of a multi-state model:
d ij
Pn ik kj jk
dt [t px ] = k=0,k6=j [t px µx+t − t pij
x µx+t ].
“The left hand side is the change in the probability, or the rate, that (x)
in State i will be in State j at age x + t. The right hand side is the
kj
net rate of entry of (x) in State i into State j by age x + t. (1) t pikx µx+t
is the rate at which (x) in State i is in State k at age x + t and then
immediately transitions into State j; this is the entry rate of (x) in State
jk
i into State j at age x + t. (2) t pij
x µx+t is the rate at which (x) in State i
is in State j at age x + t and then immediately transitions into State k;
this is the exit rate of (x) in State i out of State j at age x + t. Then,
summing (1) - (2) over all feasible States k 6= j provides the net rate of
entry of (x) in State i into State j at age x + t (i.e., the rate of being in
State j at age x + t.
• Euler’s method provides a discrete approximation:

ij
Pn ik kj jk
t+h px − t pij
x ≈ h k=0,k6=j [t px µx+t − t pij
x µx+t ].
imate other transition probabilities given one transition probability.

Expected Present Values:
• Cash Flows While Continuously in a State: Annuities
– Consider an annuity on a life in State i at age x which provides

payments of R per year continuously each year while the life is in
State j. The expected present value of this annuity is:
R ∞ −δt ij
Rāij
x = R 0 e t px dt.
– Consider an annuity on a life in State i at age x which provides

payments of R per year at the beginning of each year while the life is
in State j. The expected present value of this annuity is:
P∞ k ij
Räij
x = R k=0 v k px .
• Cash Flows Upon Transition Between States: Insurance
– Consider an insurance on a life in State i at age x which provides a

payment of S at the moment of transition into State j. The expected
present value of this insurance is:
P R ∞ −δt ik kj
S Āij
x = S j6=k 0 e t px µx+t dt.
– Consider an insurance on a life in State i at age x which provides a

payment of S at the end of the year of transition into State j. The
expected present value of this insurance is:
P P∞ t+1 ik kj
SAij
x = S j6=k t=0 v t px px+t .
This formula requires (x) to be in State j at the end of the year in

order for S to be payable at that time. We usually further assume
that transitions between states can only occur once a year to make
the model fully discrete.

• The expected present value formulas can easily be adjusted for a finite
n-year term. For example:
R n −δt ij
Rāij
x:n = R 0 e t px dt.
• Level annual benefit premiums can be calculated using:

E[P V F B@0] = E[P V F P @0].
E[P V F B@0] and E[P V F P @0] will consider the future states in the
model, and each will be calculated using the appropriate expected value
formulas. We also usually assume that policies are issued to healthy or
active lives unless otherwise indicated.
• The benefit reserve at time t for a life in State i at that time can be
calculated using:
(i)
tV = E[P V F B@0|i] - E[P V F P @0|i].
E[P V F B@0|i] and E[P V F P @0|i] will consider the future states in the
model, and each will be calculated using the appropriate expected value
formulas. Note that we can have a different reserve at time t depending
on the assumed state of the life at that time.

Thiele’s Differential Equation - Multiple State Model Version:
• Consider a fully continuous policy on (x), who can be in one of (n + 1)

possible states at any given time. Let i, j = 0, 1, ..., n. For t > 0; let:
(ij)
– bt = benefits and expenses payable upon transition from State i to
State j at time t.
(i)
– Bt = benefits and expenses payable while in State i at time t minus
the gross premium payable while in State i at time t.
– δt = the force of interest at time t.
– µij
x+t = the force of transition between States i and j at age x + t.
– t V (i) = gross premium reserve for a policy in State i at time t.

– t V (j) = gross premium reserve for a policy in State j at time t.
• Thiele’s Differential Equation - Multiple State Model Version is the fol-

lowing:
(i) Pn (ij)
d
[
dt t V (i)
] = δ tt V (i)
- B t - j=0,j6=i (bt + t V (j) − t V (i) )µij
x+t .
• There is a very logical interpretation of Thiele’s Differential Equation -

Multiple State Model Version:
“The change in the value of the reserve in State i at time t is due to
interest earned, the net benefits payable while in State i, and the benefits
and expenses payable and change in the reserve upon transition from
State i to State j (accounting for the transition between states via µijx+t
when necessary).”
• Euler’s method, with backward recursion, provides a discrete approxima-

tion:
(i) P (ij)
tV
(i)
− t−h V (i) ≈ h[δtt V (i) − Bt − nj=0,j6=i (bt + t V (j) − t V (i) )µij
x+t ].
imate other reserve values given one reserve value.

11.2 Exercises 215
11.2 Exercises
11.1. The number of dollars Sam wins in successive games follows a 3-state
model:
(i) States 0, 1, 2 correspond to 0, 1, or 2 dollars won in a game.

(ii) For k = 0, 1, 2,...; the transition probabilities are:
i pi0
x+k pi1
x+k pi2
x+k
0 0.2 0.5 0.3
1 0.1 0.6 0.3
2 0.1 0.5 0.4
(iii) Sam won zero dollars in the first game today.
Calculate the probability that Sam will win at least three dollars in the
next two games today.
(A) 0.36 (B) 0.38 (C) 0.40 (D) 0.42 (E) 0.44
11.2. The Complicated Insurance Company classifies its policyholders with

a 3-state model based on each policyholder’s credit rating, as one of Preferred
(State 0), Standard (State 1), or Poor (State 2). You are provided the fol-
lowing annual transition probabilities, for k = 0, 1, 2,...:
i pi0
x+k pi1
x+k
0 0.95 0.04
1 0.15 0.80
2 0.00 0.25
Calculate the probability that a policyholder classified as Preferred at age

x will be classified as Standard at age x + 3.
(A) 0.06 (B) 0.07 (C) 0.08 (D) 0.09 (E) 0.10

11.2 Exercises 216
11.3. Consider the Disability Income Insurance Model in Examples of

Multiple State Models. You are given:
µ01 02 10 12
x = 0.010x, µx = 0.002x, µx = 0.006x, µx = 0.004x
11
Calculate: 10 p25 .
(A) 0.050 (B) 0.055 (C) 0.060 (D) 0.065 (E) 0.070
11.4. In a Continuing Care Retirement Community (CCRC), a resident

may be in one of four possible states: Independent Living (State 1), Tem-
porarily in the Health Center (State 2), Permanently in the Health Center
(State 3), or Gone (State 4). The probabilities of a resident transitioning
between these states are as follows for k = 0, 1, 2, ...:
(i) p11 12 13
x+k = 0.60, px+k = 0.15, px+k = 0.15
(ii) p21 22 23
x+k = 0.50, px+k = 0.20, px+k = 0.20
(iii) p33 34 44
x+k = 0.90, px+k = 0.10, px+k = 1
Calculate the probability that a resident currently in Independent Living

will be in Independent Living at the end of three years.
(A) 0.30 (B) 0.32 (C) 0.34 (D) 0.36 (E) 0.38

Multiple State Models. You are given:
µ01 = 0.06, µ02 = 0.02, µ10 = 0.04, µ12 = 0.03
Calculate the probability that a healthy life will die within 10 years, where
the healthy life will first become sick and then remain sick until death.
(A) 0.052 (B) 0.056 (C) 0.060 (D) 0.064 (E) 0.068

11.2 Exercises 217
11.6. Consider the following 3-state model for the career progression of an
actuary: Analyst (0), Associate (1), Fellow (2). Furthermore:
(i) The only possible transitions: State 0 to State 1, State 1 to State 2.

(ii) µ01 = 0.8 and µ12 = 0.6
Calculate the probability that an Analyst will not be a Fellow at the end
of 5 years.
(A) 0.10 (B) 0.12 (C) 0.14 (D) 0.16 (E) 0.18
11.7. For a 3-state model: Healthy (0), Disabled (1), and Dead (2):
(i) p00
x+k = 0.7 for k = 0, 1, ...
(ii) p01
x+k = 0.2 for k = 0, 1, ...
(iii) p10
x+k = 0.1 for k = 0, 1, ...
(iv) p11
x+k = 0.6 for k = 0, 1, ...
(v) There are 1000 Healthy independent lives at time 0.
Calculate the expected number of lives who will survive the first two years.
(A) 710 (B) 730 (C) 750 (D) 770 (E) 790

11.2 Exercises 218
11.8. A 4-state heterogeneous discrete-time Markov model is used to model

an employee’s eligibility for employee benefits, with states: Active (0), Tem-
porarily Disabled (1), Permanently Disabled (2), and Inactive (3).
(i) The annual probability transition matrices are:
i pi0
x pi1
x pi2
x pi3
x
0 0.82 0.08 0.06 0.04
1 0.17 0.63 0.14 0.06
2 0.00 0.00 0.79 0.21
3 0.00 0.00 0.00 1.00
i pi0
x+1 pi1
x+1 pi2
x+1 pi3
x+1
0 0.80 0.10 0.05 0.05
1 0.11 0.65 0.16 0.08
2 0.00 0.00 0.77 0.23
3 0.00 0.00 0.00 1.00
i pi0
x+2 pi1
x+2 pi2
x+2 pi3
x+2
0 0.79 0.11 0.05 0.05
1 0.08 0.66 0.17 0.09
2 0.00 0.00 0.75 0.25
3 0.00 0.00 0.00 1.00
(ii) Transitions between states can only occur at the end of the year.
Consider a group of 80 active employees each aged x. Their future states are
independent. Calculate the variance of the number of the 80 active employees
that will be temporarily disabled at the end of three years.
(A) 7 (B) 8 (C) 9 (D) 10 (E) 11

11.2 Exercises 219
11.9. Consider the Permanent Disability Model in Examples of Multiple

State Models. You are given the following transition intensities:
µ01 2 02 3 12
x = 0.0001x , µx = 0.00005x , µx = 0.005(1.08)
x
Transition probabilities are estimated using Kolmogorov’s forward equa-

tions solved numerically with Euler’s method. A step of h = 0.1 is used.
Estimate the probability that a healthy life aged 10 is disabled at age 10.2.
(A) 0.001 (B) 0.002 (C) 0.003 (D) 0.004 (E) 0.005

Multiple State Models. You are given the following transition intensities:
µ01
x = 0.000003e
0.1382x
, µ02
x = 0.000076e
0.0875x
, µ10 01 12 02
x = 0.1µx , µx = µx
Transition probabilities are estimated using Kolmogorov’s forward equa-

tions solved numerically with Euler’s method using a step of h = 0.1.
Find the probability a healthy life aged 45 is dead at age 45.1.
(A) 0.0001 (B) 0.0002 (C) 0.0003 (D) 0.0004 (E) 0.0005
11.11. Consider a 3-state model on (x) with States 0, 1, and 2:
(i) The only possible transitions are from State 0 to State 1 and State 0 to
State 2.
(ii) µ01
x = 0.002x + 0.00001x
2
(iii) µ02
x = 0.05
Calculate the derivative of t p00

35 with respect to t evaluated at t = 5.7.
(A) - 0.07 (B) -0.03 (C) - 0.01 (D) 0.05 (E) 0.08

11.2 Exercises 220
11.12. Professors at a university can be in:
State 0: Non-tenured professor

State 1: Tenured professor
State 2: Dismissed from the university
Paul joins the university as a non-tenured professor at age 30. You are
given:
(i) ½
0.10 for t ≤ 6
µ01
30+t =
0.00 for t > 6
(ii) ½
0.05 for t ≤ 6
µ02
30+t =
0.20 for t > 6
(iii) µ12
30+t = 0.001 for t > 0
(iv) Tenured professors can never return to the non-tenured professor state.
(v) Professors dismissed from the university can never get rehired.
(vi) Paul has mortality such that: µ30+t = 0.006(1.02)30+t for t > 0, re-
gardless of state.
Calculate the probability that Paul will be a tenured professor at the uni-
versity at age 40.
(A) 0.35 (B) 0.39 (C) 0.43 (D) 0.47 (E) 0.50

11.2 Exercises 221
11.13. Consider the following 3-state model for the career progression of
an actuary: Analyst (0), Associate (1), Fellow (2). Furthermore:
(i) The only possible transitions: State 0 to State 1, State 1 to State 2.

(ii) µ01 = 0.8 and µ12 = 0.6
Estimate the probability that an Analyst aged x will be an Associate at

age x + 0.4 using Euler’s method with step size 0.2 to numerically solve
Kolmogorov’s forward equations.
(A) 0.26 (B) 0.28 (C) 0.30 (D) 0.32 (E) 0.34
11.14. Consider a special 5-year term insurance:
(i) The multiple state model underlying this insurance is the Permanent
Disability Model in Examples of Multiple State Models, with transition in-
tensities:
µ01 = 0.03, µ02 = 0.01, and µ12 = 0.05.
(ii) The insurance is only issued to healthy individuals.
(iii) The insurance pays 5000 at the moment of death.
(iv) δ = 0.06
Calculate the net single premium for this insurance.
(A) 260 (B) 280 (C) 300 (D) 320 (E) 340

State Models. You are given: (i) µ01 = 0.07, µ02 = 0.05, µ12 = 0.12 and (ii)
δ = 0.04.
A 10-year insurance issued on a healthy life aged x pays a benefit of 5000

at the moment of transition to disabled. Benefit premiums are payable con-
tinuously while the policyholder is healthy.
Calculate the annual rate of benefit premium.
(A) 250 (B) 300 (C) 350 (D) 400 (E) 450

11.2 Exercises 222

State Models:
(i) µ01 = 0.05, µ02 = 0.08, µ12 = 0.06

(ii) δ = 0.03
An annuity issued on a healthy life aged x pays a benefit of 30,000 per year
continuously each year while the life is in the disabled state. Net premiums
are payable continuously each year while the life is in the healthy state.
Calculate the annual rate of net premium.
(A) 16,000 (B) 16,400 (C) 16,700 (D) 17,000 (E) 17,500

Multiple State Models:
The annual transition probabilities, for k = 0, 1, 2, ..., are:
p00 01 10 11
x+k = 0.7, px+k = 0.2, px+k = 0.5, px+k = 0.3
Consider a special 3-year insurance issued on a healthy individual:
(i) 1000 is payable at the end of the year of death.

(ii) Level benefit premiums of P are paid at the beginning of the year while
the individual is healthy, and level benefit premiums of 0.5P are paid at the
beginning of the year while the individual is sick.
(iii) i = 0.06
Calculate the benefit reserve at time 2 assuming the individual is healthy

at that time.
(A) - 30 (B) - 20 (C) - 10 (D) 0 (E) 10

11.2 Exercises 223
11.18. Consider a special 3-year policy on Leonard aged x:
(i) The multiple state model underlying this policy is the Disability Income
Insurance Model in Examples of Multiple State Models, with:
p00 01 10 11
x = 0.70, px = 0.20, px = 0.50, px = 0.30
p00 01 10 11
x+1 = 0.65, px+1 = 0.25, px+1 = 0.45, px+1 = 0.35
p00 01 10 11
x+2 = 0.60, px+2 = 0.30, px+2 = 0.40, px+2 = 0.40
(ii) Transitions between states occur only once per year.

(iii) 5000 is payable at the beginning of the year if Leonard is sick.
(iv) Level net premiums are payable at the beginning of the year only if
Leonard is healthy.
(v) i = 0.04
(vi) Leonard is in the healthy state at policy issue.
Calculate the net premium reserve at time 1 assuming Leonard is sick at

that time.
(A) 5100 (B) 5400 (C) 5700 (D) 6000 (E) 6300

State Models:
(i) µ01 = 0.04, µ02 = 0.02, and µ12 = 0.05

(ii) δ = 0.05
A special 15-year policy is issued on (x). This policy pays a benefit of 20,000
per year continuously each year while (x) is disabled, and pays 100,000 at the
moment of death of (x). A premium of 8000 per year is payable continuously
each year while (x) is healthy.
Calculate the reserve in the disabled state at time 14.6 using Euler’s
Method, with backward recursion and a step size of 0.2, to numerically solve
Thiele’s Differential Equation.
(A) 5000 (B) 6225 (C) 7450 (D) 8675 (E) 9900

11.2 Exercises 224
11.20. Consider the setup provided in Exercise 11.19. Calculate the reserve
in the healthy state at time 14.6 using Euler’s Method, with backward recur-
sion and a step size of 0.2, to numerically solve Thiele’s Differential Equation.
(A) -2600 (B) -2300 (C) -2000 (D) -1600 (E) -1200

11.2 Exercises 225
11.1. D
11.2. E
11.3. A
11.4. B
11.5. B
11.6. C
11.7. D
11.8. E
11.9. B
11.10. D
11.11. A
11.12. A
11.13. B
11.14. A
11.15. C
11.16. C
11.17. B
11.18. E
11.19. E
11.20. B


• Exam MLC, Fall 2013: #4, 10, 21
• Exam MLC, Spring 2013: #4, 10, 13
• Exam MLC, Fall 2012: #12, 16, 24
• Exam MLC, Spring 2012: #12, 19, 28
218, 250

227
12 MULTIPLE DECREMENTS I
12.1 Key Concepts

• A multiple decrement model has a total of n + 1 states.
• There is one starting state, State 0. Often this is interpreted as “alive”

or “active.”
• An individual can only transition from State 0 to one of n other states

(States 1, 2, ..., n), called causes of decrement or decrements. Once a
transition is made to one of States 1, 2, ..., n, it is impossible for the
individual to transition from that state to any other state.
– Sometimes the model is referred to as an n-decrement model. For

example, a 2-decrement model is a double decrement model.
– States 1, 2, ..., n could represent different types of death (such as
accidental, non-accidental death), or different states of activity (dead,
withdrawal, disability, retirement).
– In general, “transitioning to State j” can also be replaced with “failing
due to cause j” or “decrementing due to cause j,” for j = 1, 2, ..., n.
• In the multiple decrement model, all n causes of decrement are “com-

peting with each other.” All decrements are assumed to be operating
“against” an individual, each one attempting to cause the individual to
decrement.

Probability Functions:
• Let j = 1, 2, ..., n; and τ denote total decrement.
• The force of transition/decrement at age x + t due to cause j:

(j) (j)
µx (t) = µx+t
• The total force of transition/decrement at age x + t:

(τ ) P (j)
µx+t = nj=1 µx+t
• The probability that (x) stays in State 0 for at least t years (survives all
decrements for at least t years):
(τ ) R t (τ )
p
t x = exp[− 0 µx+s ds].
Note: In order to survive all decrements between ages x and x + t, one
(τ )
needs to overcome the total force of transition/decrement: µx+s .
• The probability that (x) transitions out of State 0 within t years (decre-
ments within t years):
(τ ) (τ )
t qx = 1 - t px

• The probability that (x) transitions out of State 0 to State j within t

years (decrements due to cause j within t years):
(j) R t (τ ) (j)
t qx = 0 s px µx+s ds.
(j)
Note: t qx is a dependent probability. “In order for (x) to decrement due
to cause j within t years, there has to be some time s within t years where
(x) survives all decrements for s years to age x + s and then immediately
decrements due to cause j at age x + s.”
(τ ) Pn (j)
– t qx = j=1 t qx
(τ ) Rt (τ ) (τ )
– t qx = 0 s px µx+s ds
• The probability that (x) transitions out of State 0 to State j (decrements

(j)
due to cause j): ∞ qx .
Here, there is no specific amount of years in which this should occur -
hence the ∞.
• The probability that (x) transitions out of State 0 to State j between

ages x + u and x + u + t (decrements due to cause j between ages x +
u and x + u + t):
(j) R u+t (τ ) (j) (τ ) (j)
q
u|t x = u s px µx+s ds = (u px )(t qx+u )
(j) R u+1 (τ ) (j) (τ ) (j)

An important special case: u| qx = u s px µx+s ds = u px qx+u
• The probability that (x) transitions out of State 0 between ages x + u

and x + u + t (decrements between ages x + u and x + u + t):
(τ ) R u+t (τ ) (τ ) (τ ) (τ ) (τ ) (τ ) (τ ) (τ )
u|t q x = u s px µx+s ds = (u px )(t qx+u ) = u px - u+t px = u+t qx - u qx .
An important special case:

(τ ) R u+1 (τ ) (τ ) (τ ) (τ ) (τ ) (τ ) (τ ) (τ )
q
u| x = u s px µx+s ds = u px qx+u = u px - u+1 px = u+1 qx - u qx .

Life Tables:
(τ ) (τ ) (τ )
• lx+t = lx t px
“The expected number of survivors of all decrements at age x + t is
the expected number of survivors of all decrements at age x times the
probability that each life survives all decrements for t years.”
For l0 s, the τ is sometimes omitted.
(τ ) (τ ) (τ ) (τ ) (τ )
• t dx = lx t qx = lx - lx+t
“The expected number of lives who transition out of State 0 between ages
x and x + t is (i) the expected number of survivors of all decrements at
age x times the probability that each life transitions out of State 0 within
t years and (ii) the expected number of survivors of all decrements at age
x less the expected number of survivors of all decrements at age x + t.”
(j) (τ ) (j)
• t dx = l x t q x
“The expected number of lives who transition out of State 0 to State
j between ages x and x + t is the expected number of survivors of all
decrements at age x times the probability that each life transitions out
of State 0 to State j within t years.”
(τ ) P (j)
Note: t dx = nj=1 t dx .
(j) (τ ) (j)
• t dx+u = lx u|t qx
(τ ) (τ ) (τ )
• t dx+u = lx u|t qx
An example of a multiple decrement life table is the Illustrative Service Table,

which is one of many tables provided during Exam MLC (refer to Appendix
A of this study supplement).

12.2 Exercises 231
12.2 Exercises
12.1. For a double decrement model on (40):
(1) 1
(i) µ40 (t) = 60+t for t ≥ 0
(2) 2
(ii) µ40 (t) = 60+t for t ≥ 0
Calculate the probability that (40) decrements within 10 years due to cause
2.
(A) 0.21 (B) 0.23 (C) 0.25 (D) 0.27 (E) 0.29
12.2. Consider a double decrement model on (x):

(1)
(i) µx (t) = 0.10 for t > 0
(2)
(ii) µx (t) = 0.20 for t > 0
Calculate the probability (x) decrements in the fifth year.
(A) 0.08 (B) 0.09 (C) 0.10 (D) 0.11 (E) 0.12
12.3. For a triple decrement model on (50):

(1) 1
(i) µ50 (t) = 50−t for 0 ≤ t < 50
(2) 2
(ii) µ50 (t) = 50−t for 0 ≤ t < 50
(3) 3
(iii) µ50 (t) = 50−t for 0 ≤ t < 50
Calculate the probability that (50) decrements due to cause 2.
(A) 0.31 (B) 0.33 (C) 0.35 (D) 0.37 (E) 0.39

12.2 Exercises 232
12.4. Paul, age 33, is an actuarial science professor. His career is subject
to two decrements:
(1) t
(i) Decrement 1 is mortality: µ33 (t) = 50 for t ≥ 0.
(2) t
(ii) Decrement 2 is leaving academic employment: µ33 (t) = 40 for t ≥ 0.
Calculate the probability that Paul remains an actuarial science professor

for less than five years.
(A) 0.41 (B) 0.43 (C) 0.45 (D) 0.47 (E) 0.49
12.5. For a double decrement model:

(1) c
(i) µ45 (t) = 50−t for c > 0 and 0 < t < 50
(2) 1
(ii) µ45 (t) = 50−t for 0 < t < 50
(iii) The probability that (45) fails due to cause 2 is 0.4.
Calculate: c.
(A) 0.75 (B) 1.00 (C) 1.25 (D) 1.50 (E) 1.75
12.6. (Exam MLC Only) Using the Illustrative Service Table, calculate:
(τ )
10 p52 .
(A) 0.60 (B) 0.61 (C) 0.62 (D) 0.63 (E) 0.64
(w)
8 q41 .
(A) 0.09 (B) 0.10 (C) 0.11 (D) 0.12 (E) 0.13

12.2 Exercises 233
(i)
7|4 q46 .
(A) 0.010 (B) 0.013 (C) 0.016 (D) 0.019 (E) 0.022
12.9. Consider the following triple decrement table, where “—” denotes a
missing value::
(1) (2) (3) (τ ) (τ ) (1) (2) (3)
x qx qx qx qx lx dx dx dx
30 0.005 — — 0.075 2000 — 120 —
31 0.010 0.070 0.015 — 1850 — — —
32 0.015 0.075 0.025 0.115 — 25.11 125.57 41.86
(τ )
Calculate: l32 .
(A) 1658 (B) 1662 (C) 1666 (D) 1670 (E) 1674
12.10. The following double decrement table applies to students entering

an extremely difficult 4-year college, where “—” denotes a missing value:
(i)
For a student at the beginning of that academic year:
Academic Probability of Probability of Withdrawal Probability of Surviving
Year Academic Failure for All Other Reasons the Academic Year
during the Year during the Year
1 0.19 0.29 —
2 — — 0.67
3 — 0.11 0.84
4 0.00 — 0.92
(ii) The probability that a student at the start of their second academic
year will withdraw for all other reasons during the year is twice the probability
that a student at the start of their second academic year will academically
fail during the year.
If there are 2000 entering students, calculate the expected number that
will academically fail within four years.
(A) 520 (B) 530 (C) 540 (D) 550 (E) 560

12.2 Exercises 234
12.11. For a triple decrement model on (x):

(1)
(i) qx = 0.20
(2) (1)
(ii) qx = 0.75qx
(3) (2)
(iii) qx = 0.75qx
(τ )
(iv) px = 0.5375
Out of 1000 lives each aged x, calculate the expected number of lives who
will decrement due to cause 2 prior to age (x + 1).
(A) 130 (B) 140 (C) 150 (D) 160 (E) 170
12.12. The Justice League’s home is subject to destruction from one of

the following four perils: fire, wind, collision, and super villain attack. If the
Justice League’s home is destroyed, it will not be rebuilt. An insurer has just
issued a homeowners’ policy to the Justice League covering their home from
all perils except for super villain attack.
The following table provides the probability of each peril in any year:
Peril Fire Wind Collision Super Villain Attack

Probability 0.030 0.035 0.025 0.300
Calculate the probability that a covered peril will destroy the Justice
League’s home after the issue of the homeowners’ policy.
(A) 0.23 (B) 0.24 (C) 0.25 (D) 0.26 (E) 0.27
12.13. Consider a double decrement model on (x):

(1)
(i) µx (t) = 0.04 for t > 0
(2)
(ii) µx (t) = 0.07 for t > 0
Calculate the expected time until decrement for (x) given that the cause
of decrement is (2).
(A) 8.1 (B) 8.6 (C) 9.1 (D) 9.6 (E) 10.1

12.2 Exercises 235
12.14. Consider the following triple decrement table, with missing entries
denoted as “—”:
(τ ) (1) (2) (3)

x lx dx dx dx
55 1000 — — 11
56 — — 37 —
57 — — 41 —
58 — — — —
59 689 — — —
You are given:

(τ )
(i) 1| q55 = 0.082
(1) (1)
(ii) q56 = 0.82617q57
(2) (1)
(iii) d55 = 1.44d55
(τ )
(iv) p58 = 0.89133
(1) (1)
(v) d56 = 1.16d55
(τ )
(vi) q55 = 0.072
(2) (τ )
(vii) d58 = 0.54762d58
(3)
(viii) µx = 0 for x ≥ 57
(1)
Calculate: 2 q57 .
(A) 0.075 (B) 0.077 (C) 0.079 (D) 0.081 (E) 0.083

12.2 Exercises 236
12.15. A multi-national corporation currently has 30,000 employees each

aged 56. If an employee exits employment from the corporation it can only
be due to death, withdrawal, or retirement. If an employee exits due to
withdrawal or retirement, they cannot be re-employed at the corporation.
No new employees will be hired during the next three years. For the above
30,000 employees:
(i) The probability that an employee aged 56 retires by age 57 is 0.009.

(ii) The expected number of employees aged 56 that withdraw by age 57
is 70% of the expected number of employees aged 56 that die by age 57.
(iii) The expected number of employees aged 56 that die by age 57 is equal
to 86.372% of the expected number of employees aged 57 that die by age 58.
(iv) The probability that an employee aged 56 is employed at age 57 is

0.9655.
(v) The expected number of employees aged 57 that exit employment by
age 58 due to causes other than death is 608.
(vi) The probability that an employee aged 56 exits employment between
ages 58 and 59 is 0.03723.
Calculate the expected number of employees at the corporation three years

from now.
(A) 26,120 (B) 26,720 (C) 27,320 (D) 27,920 (E) 28,520
12.16. Consider a double decrement model on (30):

(1) 1
(i) µ30 (t) = 60−t for 0 ≤ t < 60
(2)
(ii) µ30 (t) = 0.02 for t ≥ 0
Calculate the probability that (30) will not decrement due to cause 2 before
age 65.
(A) 0.54 (B) 0.57 (C) 0.60 (D) 0.63 (E) 0.66

12.2 Exercises 237
12.17. Consider a triple decrement model on (x):

(1)
(i) µx (t) = 0.02 for t > 0
(2)
(ii) µx (t) = 0.04 for t > 0
(3)
(iii) µx (t) = 0.06 for t > 0
Calculate the expected number of years that (x) will avoid decrement.
(A) 6 (B) 7 (C) 8 (D) 9 (E) 10
12.18. Consider a group of independent policyholders aged 31, each with

identical life insurance policies. Each of these life insurance policies can be
terminated due to either death (d) or withdrawal (w) of the policyholder.
You are given:
(i) Each policyholder is subject to the following double decrement table,

with missing entries denoted as “—”:
(τ ) (d) (w)
x lx dx dx
30 1000 20 5
31 — 23 —
32 946 — 8
33 — 29 9
34 875 — —
Calculate the probability that (31) dies before age 33.6.
(A) 0.055 (B) 0.061 (C) 0.067 (D) 0.073 (E) 0.079

12.2 Exercises 238
12.19. Nick, aged x, is a professional rollerblader whose career is subject

to the following decrements:
(1)
(i) Decrement 1 is injury, where: µx+t = 0.025 for t > 0.
(ii) Decrement 2 is retirement, where:
½
(2) 0.010 for 0 ≤ t < 10
µx+t =
0.020 for 10 ≤ t
(iii) Decrement 3 is all other causes, where:

½
(3) 0.035 for 0 ≤ t < 10
µx+t =
0.050 for 10 ≤ t
Calculate the probability that Nick will retire within 15 years.
(A) 0.11 (B) 0.12 (C) 0.13 (D) 0.14 (E) 0.15
12.20. For a double decrement model:

(τ )
(i) 2| qx = 0.11
(1)
(ii) qx = 0.05
(τ )
(iii) 1| qx = 0.12
(2)
(iv) qx = 0.08
(τ )
Calculate: qx+2 .
(A) 0.11 (B) 0.12 (C) 0.13 (D) 0.14 (E) 0.15

12.2 Exercises 239
12.1. C
12.2. A
12.3. B
12.4. B
12.5. D
12.6. D
12.7. C
12.8. D
12.9. E
12.10. B
12.11. C
12.12. A
12.13. C
12.14. E
12.15. B
12.16. D
12.17. C
12.18. C
12.19. A
12.20. E


• Exam 3L, Fall 2013: #6, 7
• Exam 3L, Fall 2012: #6, 7
• Exam 3L, Spring 2012: #6, 7
• Exam MLC, Sample Questions: #33, 103, 105, 144, 167
• Exam 3L, Fall 2011: #6, 7
• Exam 3L, Spring 2011: #6, 7
• Exam 3L, Fall 2010: #6, 7
• Exam 3L, Spring 2010: #8, 9
• Exam 3L, Fall 2009: #7, 8

241
13 MULTIPLE DECREMENTS II (Exam MLC Only)
13.1 Key Concepts

For an n-decrement model: Let State 0 denote the starting state, and
let each of States j = 1, 2, ..., n denote an exit or decrement state. There
is only one transition that can actually occur: a transition from State 0 to
exactly one of State 1, State 2, ..., State n. Once in exactly one of State 1,
State 2, ..., State n, the individual remains in that state forever. Refer to
the Multiple Decrement Model in Examples of Multiple State Models for a
diagram.
(j)
• µx0j = µx
Pn
• µ0•
x =
0j
j=1 µx
Rt (τ )
• t p00 00
x = t px = exp[− 0 µ0•
x+s ds] = t px
(τ )
• t p0• 00
x = 1 - t px = t q x
Rt 00 0j (j)
• t p0j
x = 0 s px µx+s ds = t qx
• Note for i = 0, 1, ..., n and i 6= j: t pjj ji

x = 1, and t px = 0

Associated Single Decrements:
• Consider cause of decrement j. If we assume that j is the only decrement

operating against an individual, we can construct the
associated single decrement model for j.
• Primes on probabilities will indicate that j is the only cause of decrement;

i.e. the associated single decrement model.
• The probability that (x) remains in State 0 for at least t years (survives
decrement j for at least t years), where j is the only cause of decrement:
0(j) Rt (j)
t px = exp[− 0 µx (s)ds]
(τ ) 0(j) 0(1) 0(2) 0(n)
Note: t px = Πnj=1 t px = t px t px · · · t px .
• The probability that (x) transitions from State 0 to State j within t years
(decrements due to cause j within t years) , where j is the only cause of
decrement:
0(j) 0(j)
t qx = 1 - t px
0(j) Rt 0(j) (j)
Note: t qx = 0 s px µx (s)ds.
(τ ) 0(j)
• It is useful to note that: t px ≤ t px .
This makes sense: the probability of surviving just one decrement is
higher than the probability of surviving all decrements.
(j) 0(j)
• It is useful to note that: t qx ≤ t qx .
This makes sense: the probability of decrement j “striking” the individual
will be higher when there are no other decrements competing to “strike”
the individual.

Relationships between Multiple and Single Decrements:
UDD in the Multiple Decrement Model:
• Assume that the cause of decrement j and the total decrement τ are
uniformly distributed within each year of age in a multiple decrement
table (the latter is satisfied automatically if all causes of decrement are
uniformly distributed within each year of age in a multiple decrement
table). Then, for 0 ≤ s ≤ 1:
(j) (j)
– s qx = sqx
(τ ) (τ )
– s qx = sqx
0(j) (τ ) (j) (τ )
– s px = [1 − sqx ]qx /qx
.
UDD in All of the Associated Single Decrement Models:
• Assume that all causes of decrement are uniformly distributed within

each year of age in the associated single decrement models. Then, for
cause of decrement j and 0 ≤ s ≤ 1:
0(j) 0(j)
• s qx = sqx
0(j) (j) 0(j)
• s px µx+s = qx
– If there are two causes of decrement (j = 1, 2):

(1) 0(1) s2 0(2)
s qx = qx (s − 2 qx ) and
(2) 0(2) s2 0(1)
s qx = qx (s − 2 qx ).
(1) (2) (τ )
It is still true that: s qx + s qx = s qx .
– If there are three causes of decrement (j = 1, 2, 3):

(1) 0(1) s2 0(2) s2 0(3) s3 0(2) 0(3)
s qx = qx (s − 2 qx − 2 qx + 3 qx qx ),

(2) 0(2) s2 0(1) s2 0(3) s3 0(1) 0(3)

s qx = qx (s − 2 qx − 2 qx + 3 qx qx ), and
(3) 0(3) s2 0(1) s2 0(2) s3 0(1) 0(2)
s qx = qx (s − 2 qx − 2 qx + 3 qx qx ).
(1) (2) (3) (τ )
It is still true that: s qx + s qx + s qx = s qx .
UDD in Some of the Associated Single Decrement Models:

There really are no general formulas, these have to be handled on a case-
by-case basis. Here are a couple of examples, each of which considers a triple
decrement model on (x):
• Suppose decrements 1 and 2 are uniformly distributed over each year of

age in the associated single decrement tables, and decrement 3 can only
occur at the end of the year.
(1) 0(1) 0(2) (2) 0(2) 0(1)
Then: qx = qx (1 − 21 qx ) and qx = qx (1 − 21 qx ).
“These make sense because decrement 3 can only occur at the end of
the year, meaning that during the year, we effectively have a double
decrement model with only decrements 1 and 2 in effect. Therefore, the
double decrement equations in UDD in All of the Associated Single
Decrement Models with s = 1 apply.”
(3) 0(1) 0(2) 0(3)

Also: qx = px px qx .
“This makes sense because the only decrements in effect during the year
are decrements 1 and 2. Therefore, to decrement due to 3 during the
year, (x) has to survive both decrements 1 and 2 in order to make it to
the end of the year, and then decrement due to 3.”

• Suppose decrements 1 and 2 are uniformly distributed over each year of

age in the associated single decrement tables, and decrement 3 can only
occur at the beginning of the year.
(1) 0(3) 0(1) 0(2) (2) 0(3) 0(2) 0(1)
Then: qx = px qx (1 − 12 qx ) and qx = px qx (1 − 21 qx ).
“These make sense because decrement 3 can only occur at the beginning
of the year, meaning that during the year, we effectively have a double
decrement model with only decrements 1 and 2 in effect. Therefore, the
double decrement equations in UDD in All of the Associated Single
Decrement Models with s = 1 apply, after first surviving decrement 3
at the beginning of the year.”
(3) 0(3)
Also: qx = qx .
“This makes sense because decrement 3 occurs at the beginning of the

year. As decrements 1 and 2 are in effect during the year, they do not
impact the probability of decrement due to 3 at the very start of the
year.”
For more complicated situations, appeal to the formula:

(j) R t (τ ) (j)
t q x = 0 s px µx (s)ds.
In fact, each of the results in both UDD in All of the Associated

Single Decrement Models and UDD in Some of the Associated
Single Decrement Models were obtained from this particular relation,
applying the appropriate assumptions regarding the distribution of each
decrement throughout the year.

Moments for Life Insurance in a Multiple Decrement Model:
• Consider a whole life insurance on (x) which pays a benefit at the moment
of transition from State 0 to a decrement state. Upon transition to State
(1)
1 at time t the benefit is bt , upon transition to State 2 at time t the
(2) (n)
benefit is bt , ..., upon transition to State n at time t the benefit is bt .
Let Z denote the present value random variable for this insurance.
Pn R ∞ (j) (τ ) (j)
– E[Z k ] = j=1 0 [v t bt ]k t px µx (t)dt for k = 1, 2, ...
– The expected present value of the whole life insurance would be E[Z].
• Consider a whole life insurance on (x) which pays a benefit at the end
of the year of transition from State 0 to a decrement state. Let i = 0,
(1)
1, ... Upon transition to State 1 in year i + 1 the benefit is bi+1 , upon
(2)
transition to State 2 in year i + 1 the benefit is bi+1 , ..., upon transition
(n)
to State n in year i + 1 the benefit is bi+1 . Let Z denote the present
value random variable for this insurance.
P P i+1 (j) k (τ ) (j)
– E[Z k ] = nj=1 ∞ i=0 [v bi+1 ] i px qx+i for k = 1, 2, ...
– The expected present value of the whole life insurance would be E[Z].
• If all forces of decrement are constant, then for a whole life insurance
that pays 1 at the moment of transition from State 0 to State j:
(j) µ(j)
– Āx = µ(τ ) +δ
• If all forces of decrement are constant, then for a whole life insurance that
pays 1 at the moment of transition out of State 0 to any other State:
(τ ) µ(τ )
– Āx = µ(τ ) +δ

• The above can easily be modified for an insurance with a finite term. As
an example, if all forces of decrement are constant, then for an n-year
term life insurance that pays 1 at the moment of transition from State 0
to State j:
µ(j)
– Ā(j)1x:n = µ(τ ) +δ
(1 − exp[−(µ(τ ) + δ)n])
• A common type of multiple decrement life insurance is a double indemnity insurance,

where there are two causes of decrement: accidental death and non-
accidental death. The benefit for accidental death is double the benefit
for non-accidental death.
For example, consider a whole life insurance on (x) with a double indem-
nity provision that pays 1 at the moment of non-accidental death (N)
and 2 at the moment of accidental death (A). Then the expected present
value can be calculated as either:
(τ ) (A) (A) (N )
Āx + Āx = 2Āx + Āx .
• Level annual benefit premiums can be calculated using:

E[P V F B@0] = E[P V F P @0].
E[P V F B@0] will consider all of the decrements in the model, and will
be calculated as discussed earlier in this section for a life insurance.
• The benefit reserve at time t can be calculated using:
tV = E[P V F B@t] - E[P V F P @t].

E[P V F B@t] will consider all of the decrements in the model, and will
be calculated as discussed earlier in this section for a life insurance.

Asset Shares:
• An asset share is the amount of a company’s assets that are allocated to

support a policy or a block of similar policies.
• Consider a fully discrete life insurance on (x). For k = 0, 1, 2, ...:
– ASk = k AS = asset share at time k
– G = level annual gross premium
– ck = fraction of premium used to determine the percent of premium

expense at time k
– ek = per policy expense at time k

1
– The effective annual interest rate is ik+1 for year k + 1; vk+1 = i+ik+1
– Assume there are two causes of decrement for (x): death (d) and
withdrawal (w). The respective probabilities of death and withdrawal
(d) (w)
between ages x + k and x + k + 1 are qx+k and qx+k .
– bk+1 = death benefit paid at time k + 1 if (x) dies in year k + 1.
– CVk+1 = k+1 CV = withdrawal benefit/cash value paid at time k + 1

if (x) withdraws in year k + 1.
• The asset share recursion; for k = 0, 1, ...; is:

(τ ) (d) (w)
vk+1 px+k ASk+1 = ASk + G - Gck - ek - vk+1 qx+k bk+1 - vk+1 qx+k CVk+1
(d) (w)
=⇒ (ASk + G - Gck - ek )(1 + ik+1 ) = qx+k bk+1 + qx+k CVk+1 +
(τ )
px+k ASk+1

• This second asset share recursion has a similar interpretation as the re-
serve recursions, as it should. An asset share is essentially a reserve
calculated using actual experienced interest rates, mortality rates, ex-
penses, etc. We are also allowing for a benefit payable upon withdrawal
(w)
of CVk+1 payable between ages x + k and x + k + 1 at rate qx+k .
• Unless otherwise stated, assume AS0 = 0.

13.2 Exercises 250
13.2 Exercises
13.1. Consider a double decrement model on (40):
(i) The decrements are mortality and lapse.

(ii) The force of mortality is given by:
µx = 0.001 + c(x - 40) for c > 0, x ≥ 40.
(iii) The transition intensity for lapses is a constant equal to 0.03.
(τ )
(iv) 5 q40 = 0.14572
0(1)
Calculate: 1000q40 .
(A) 1.1 (B) 1.4 (C) 1.7 (D) 2.0 (E) 2.3
13.2. Consider a 3-state model on (x) with States 0, 1, and 2:
(i) µ01
x (t) = 0.01t for t > 0
(ii) µ02
x (t) = 0.02t for t > 0
(iii) µ1i 2i
x (t) = µx (t) = 0 for i = 0, 1, 2
Calculate: 5 p01
x .
(A) 0.096 (B) 0.100 (C) 0.104 (D) 0.108 (E) 0.112
(i) There are two causes of decrement.

(1) 1
(ii) µ40 (t) = 50−t for 0 ≤ t < 50
(τ ) 95−2t
(iii) µ40 (t) = 2250−95t+t2 for 0 ≤ t < 45
0(2)
Calculate: 4 q40 .
(A) 0.08 (B) 0.09 (C) 0.10 (D) 0.11 (E) 0.12

13.2 Exercises 251
13.4. In a triple decrement table, lives are subject to decrement by death

(d), disability (i), and withdrawal (w):
(i) Each decrement is uniformly distributed in the triple decrement table

over each year of age.
(τ ) (τ )
(ii) lx = 25,000 and lx+1 = 23,000
(d) (i)
(iii) dx = 50 and dx = 750
0(w)
Calculate: 0.6 qx .
(A) 0.017 (B) 0.021 (C) 0.025 (D) 0.029 (E) 0.033
13.5. Consider the following triple decrement table, with missing entries
denoted as “—”:
(τ ) (1) (2) (3)
x lx dx dx dx
40 1000 30 56 13
41 — 35 58 —
42 789 — — 22
Each decrement is uniformly distributed within each year of age in the

triple decrement table.
Calculate the independent probability that (41) decrements within one

year due to cause 3.
(A) 0.015 (B) 0.018 (C) 0.020 (D) 0.022 (E) 0.25
13.6. Suppose there are two decrements, (1) and (2):

0(1) 0(2)
(i) q37 = 0.1 and q37 = 0.2
(ii) Each decrement is uniformly distributed over each year of age in the
associated single decrement tables.
(2)
Calculate: q37 .
(A) 0.15 (B) 0.16 (C) 0.17 (D) 0.18 (E) 0.19

13.2 Exercises 252
13.7. Lex owns a plane that is x years old. You are given:
(i) The plane faces three decrements: breakdown (1), sale (2), and collision
(3).
(ii) All decrements are uniformly distributed within each year of age in the
0(1) 0(2) 0(3)
(iii) qx = 0.10, qx = 0.25, qx = 0.05
Calculate the probability that the plane will be involved in a collision

within three months.
(A) 0.012 (B) 0.016 (C) 0.019 (D) 0.022 (E) 0.026
13.8. An actuary has constructed a multiple decrement table using the

following assumptions:
(i) There are 3 decrements: mortality (d), disability (i), withdrawal (w).
0(d) 0(i) 0(w)
(ii) q40 = 0.02, q40 = 0.03, and q40 = 0.05
(iii) All decrements are uniformly distributed over each year of age in the
Upon further review, the actuary has determined that assumption (iii)
is incorrect. The correct assumption (iii) is the following: mortality and
disability are uniformly distributed over each year of age in the associated
single decrement tables, and withdrawal can only occur at time t = 0.80 each
year.
(w)
Calculate the percent change in the value of q40 when the correct assump-
tion (iii) is used relative to when the incorrect assumption (iii) is used.
(A) -1.5 (B) -0.5 (C) 0 (D) 0.5 (E) 1.5

13.2 Exercises 253
13.9. For a triple decrement model:
(i) Decrement 1 is uniformly distributed over each year of age in the as-
sociated single decrement table; decrement 2 occurs only at the end of the
year; decrement 3 occurs only at the beginning of the year.
(ii)
(τ ) 0(1) 0(2) 0(3)
x lx qx qx qx
60 100,000 y 0.010 0.030
61 91,705 1.2y 0.011 —
62 80,622 — — —
Consider a group of 900 lives each aged 61. Calculate the expected number
from this group that will succumb to decrement 3 prior to age 62.
(A) 48 (B) 50 (C) 52 (D) 54 (E) 56
13.10. Consider the following double decrement table:
(τ ) (1) (2)
x lx dx dx
55 20,000 2000 600
56 17,400 2265 910
57 14,225 2545 1175
(1)
Two actuaries are trying to calculate 0.3 q56 . Actuary A assumes that each
decrement is uniformly distributed within each year of age in the double
decrement table. Actuary B assumes that each decrement has a constant
(j) (j)
force of decrement within each year of age; that is, µx+s = µx for j = 1, 2
and 0 ≤ s < 1.
(1)
Calculate the ratio of Actuary A’s value of 0.3 q56 to Actuary B’s value of
(1)
0.3 q56 .
(A) 0.86 (B) 0.93 (C) 1.00 (D) 1.08 (E) 1.16

13.2 Exercises 254
13.11. A special whole life insurance is issued on (x). The non-accidental

death benefit is 750, and the accidental death benefit is 1500. Furthermore:
(i) Benefits are payable at the moment of death.

(ad)
(ii) The force of mortality due to accidental death is: µx (t) = 0.005 for
t > 0.
(τ )
(iii) µx (t) = 0.040 for t > 0
(iv) δ = 0.05
Calculate the actuarial present value of this insurance.
(A) 333 (B) 375 (C) 486 (D) 500 (E) 675
13.12. For a special 2-year term insurance on (x):
(i) 20,000 is payable at the end of the year of decrement due to (1).
(ii) 10,000 is payable at the end of the year of decrement due to (2).
(iii) Level benefit premiums are payable at the beginning of the year.
(iv)
(1) (2)
k qx+k qx+k
0 0.010 0.100
1 0.015 0.110
(v) i = 0.05
(A) 1210 (B) 1230 (C) 1250 (D) 1270 (E) 1290

13.2 Exercises 255
13.13. Consider a special fully continuous whole life insurance on (x) with
a double indemnity provision:
(i) The benefit for non-accidental death is 25,000 in all years.

(ii) The benefit for accidental death during the first 5 years is 50,000.
(iii) The benefit for accidental death after the first 5 years is 0.
(accidentaldeath) (τ )
(iv) µx (t) = 0.01, µx (t) = 0.21, and δt = 0.10 for t ≥ 0.
Calculate the level annual net premium.
(A) 5400 (B) 5700 (C) 6000 (D) 6300 (E) 6600
13.14. For a special fully continuous whole life insurance on (x):
(i) δ = 0.05
(ii) There are two decrements: (1) = accidental death and (2) = non-
accidental death.
(iii) The benefit for accidental death is 30,000.
(iv) The benefit for non-accidental death is 10,000.
(v) Premiums are payable for at most the first 10 years.
(τ )
(vi) µx (t) = 0.010 for t > 0
(1)
(vii) µx (t) = 0.003 for t > 0
Determine the benefit reserve at time 5, 5 V .
(A) 1090 (B) 1105 (C) 1120 (D) 1135 (E) 1150

13.2 Exercises 256
13.15. For a special fully discrete 3-year term insurance on (55):
(i) There are two decrements: (1) = accidental death and (2) = all other
causes of death
(1)
(ii) qx = 0.002 +0.003(x − 55) for x = 55, 56, 57.
(2)
(iii) qx = 0.020 + 0.020(x − 55) for x = 55, 56, 57.
(iv) i = 0.06
(v) The death benefit is 2000 for accidental death and 1000 for death from
all other causes.
Calculate the net premium reserve at the end of the first year.
(A) 20 (B) 25 (C) 30 (D) 35 (E) 40
13.16. You are given a special whole life insurance on (x):
(i) There is a triple decrement model on (x) with forces of decrement:

µ(1) = 0.09, µ(2) = 0.05, µ(3) = 0.02.
(ii) The benefit payable at the moment of decrement due to cause 1 is
300,000; the benefit payable at the moment of decrement due to cause 2 is
150,000; the benefit payable at the moment of decrement due to cause 3 is
50,000.
(iii) i = 0.06
Calculate the expected present value of this whole life insurance.
(A) 159,310 (B) 160,420 (C) 161,530 (D) 162,640 (E) 163,750

13.2 Exercises 257
13.17. Consider a special fully discrete 3-year term insurance on (30) with
level annual benefit premiums:
(i) (i) There are two decrements: (1) = accidental death and (2) = all
other causes of death
(1) (τ )
(ii) qx = 0.005 and qx = 0.025 + 0.010(x - 30) for x = 30, 31, 32.
(iii) The death benefit for non-accidental death is 50,000; the death benefit
for accidental death is 100,000.
(iv) i = 0.05
Calculate the benefit reserve at the end of the second year.
(A) 400 (B) 450 (C) 500 (D) 550 (E) 600
13.18. For a special single premium whole life insurance on (x):
(i) δ = 0.05
(ii) Benefits are payable at the moment of death.
(iii) There are two decrements: (1) = accidental death and (2) = non-
accidental death.
(iv) The benefit for accidental death is 100,000.
(v) There are two possible benefits for non-accidental death: (a) the return
of half of the single premium without interest if non-accidental death occurs
in the first 20 years and (b) 50,000 if non-accidental death after the first 20
years.
(τ )
(vi) µx (t) = 0.008 for t > 0
(1)
(vii) µx (t) = 0.001 for t > 0
Calculate the single premium using the equivalence principle.
(A) 3700 (B) 3800 (C) 3900 (D) 4000 (E) 4100

13.2 Exercises 258
13.19. For a fully discrete life insurance of 1000 on (x):
(i) The asset share at the end of year 5 is 710.

(ii) The level annual gross premium is 300.
(iii) i = 0.05
(iv) The cash value (withdrawal benefit) at the end of year 5 is one quarter
of the sum of all gross premiums previously paid without interest.
(v) The percent of the gross premium paid at time 4 for expenses is 10%.
(vi) The amount of per policy expenses paid at time 4 is 10.
(vii) The probability that (x + 4) dies prior to age x + 5 is 0.045.
(viii) The probability that (x + 4) withdraws prior to age x + 5 is 0.250.
Calculate the asset share at the end of year 4.
(A) 270 (B) 290 (C) 310 (D) 330 (E) 350
13.20. Consider a fully discrete 5-year endowment insurance on (x):
(i) The level annual gross premium is G.

(ii) The endowment insurance has a death (d) benefit of 1000. The pure
endowment benefit is 1000.
(iii) The withdrawal (w) benefit payable at the end of year 2 is 100.
(iv) For all years, the percent of premium expense is 5% of G and the per
policy expense is 30. All expenses are payable at the beginning of the year.
(d) (w)
(v) You are given: qx+1 = 0.03 and qx+1 = 0.20.
(vi) The effective annual interest rate for the second year is 6%.
(vii) The asset share at the end of the first year is 275.90, and the asset
share at the end of the second year is 535.10.
Calculate: G.
(A) 200 (B) 220 (C) 240 (D) 260 (E) 280

13.2 Exercises 259
13.1. A
13.2. C
13.3. B
13.4. D
13.5. D
13.6. E
13.7. A
13.8. A
13.9. D
13.10. B
13.11. B
13.12. B
13.13. A
13.14. D
13.15. B
13.16. D
13.17. C
13.18. B
13.19. E
13.20. A


• Exam MLC, Spring 2013: #2, 11, 24
100, 117, 133, 135, 138, 159, 160, 178, 179, 187, 202, 206, 216, 224, 232,
234, 235, 236, 242, 244, 283
• Exam MLC, Spring 2007: #9, 12, 20, 23, 28, 30

261
14 MULTIPLE LIVES I
14.1 Key Concepts

• Consider two lives, (x) and (y), with future lifetime random variables Tx
and Ty , respectively.
• In almost all problems, it is assumed that (x) and (y) are independent,
which means that Tx and Ty are independent (an exception is the Com-
mon Shock Model, Exam MLC Only).
Multiple Life Functions:

Joint Distribution of Tx and Ty :
• The joint probability density function of Tx and Ty : fTx ,Ty (s, t)
• The joint survival function of Tx and Ty :

R∞R∞
STx ,Ty (s, t) = P r[Tx > s, Ty > t] = s t fTx ,Ty (u, v)dvdu
“The probability that (x) survives at least s years and (y) survives at
least t years.”
• The joint cumulative distribution function of Tx and Ty :

RsRt
FTx ,Ty (s, t) = P r[Tx ≤ s, Ty ≤ t] = 0 0 fTx ,Ty (u, v)dvdu
“The probability that (x) dies within s years and (y) dies within t years.”
∂ ∂
• Also: ∂t ∂s FTx ,Ty (s, t) = fTx ,Ty (s, t)
• STx ,Ty (s, t) + FTx ,Ty (s, t) 6= 1

“This is because we are not considering other possible events:
{Tx ≤ s, Ty > t} and {Tx > s, Ty ≤ t}.”

• If (x) and (y) are independent =⇒
STx ,Ty (s, t) = (s px )(t py )

FTx ,Ty (s, t) = (s qx )(t qy )
• In general, joint probabilities for independent lives can be expressed as

products of probabilities involving each individual life.
For example, with independent lives (x) and (y), the probability that (x)
dies within 10 years and (y) survives 10 years is: (10 qx )(10 py ).

Joint-Life Status (xy):
• Txy = T (xy) = time until the first death of (x) and (y).
• t pxy = STx ,Ty (t, t)
“The probability that both (x) and (y) survive for at least t years.”
u+t pxy = (u pxy )(t px+u:y+u )
“The probability that both (x) and (y) survive u + t years is equal to the
probability that both (x) and (y) first survive u years to ages x + u and
y + u, then survive another t years to ages x + u + t and y + u + t.”
If (x) and (y) are independent =⇒ t pxy = (t px )(t py )
• t qxy = 1 - t pxy
“The probability that the first death of (x) and (y) occurs within t years.”
If (x) and (y) are independent =⇒

t qxy = 1 - (t px )(t py ) = t qx + t qy - (t qx )(t qy )
• Note: t qxy 6= (t qx )(t qy )
• The force of failure at ages x + t and y + t:

d
[p ]
µxy (t) = µx+t:y+t = − dtt pt xyxy
Rt
t pxy = exp[− 0 µxy (s)ds]
If (x) and (y) are independent =⇒ µxy (t) = µx (t) + µy (t)
• u|t qxy = u+t qxy - u qxy = u pxy - u+t pxy = (u pxy )(t qx+u:y+u )
“The probability that the first death of (x) and (y) occurs after u years
but before u + t years.”

Last-Survivor Status (xy):
• Txy = T (xy) = time until the second death of (x) and (y).
• t qxy = FTx ,Ty (t, t)

“The probability that both (x) and (y) will die within t years.”
If (x) and (y) are independent =⇒ t qxy = (t qx )(t qy )
• t pxy = 1 - t qxy
“The probability that at least one of (x) and (y) will survive at least t
years.”
If (x) and (y) are independent =⇒
t pxy = 1 - (t qx )(t qy ) = t px + t py - (t px )(t py )
• Note: t pxy 6= (t px )(t py )
• u|t qxy = u+t qxy - u qxy = u pxy - u+t pxy
“This is the probability that the second death of (x) and (y) occurs after
u years but before u + t years.”
• Note1: u+t pxy 6= (u pxy )(t px+u:y+u )
“The right hand side is incorrect as it implies that both (x) and (y)
survive the first u years (t px+u:y+u : we start with two lives here); that
does NOT have to be the case. That is, u pxy is the probability that at
least one of (x) and (y) survives u years - it is possible that both (x) and
(y) will not survive the u-year period.”
• Note2: u|t qxy 6= (u pxy )(t qx+u:y+u )
“The right hand side is incorrect for the same reason as indicated in
Note1; the implication is that both (x) and (y) survive the first u years
(qx+u:y+u : we start with two lives here) which is not guaranteed in u pxy .”

Law of Addition:
• “Law of Addition”: Txy + Txy = Tx + Ty .

This implies the following:
– cTxy + cTxy = cTx + cTy for c > 0

– t qxy + t qxy = t qx + t qy
– t pxy + t pxy = t px + t py
– t pxy µxy (t) + t pxy µxy (t) = t px µx (t) + t py µy (t)
– Kxy + Kxy = Kx + Ky
– k| qxy + k| qxy = k| qx + k| qy
– “Law of Multiplication”: (Txy )(Txy ) = (Tx )(Ty )
• None of the Law of Addition relationships require independence of (x)

and (y): they are always true.
• The Law of Addition relationships are useful for calculating a quantity

involving (xy) from easier quantities involving (x), (y), and (xy).

Moments:
• Let (s) denote any of the statuses: (x), (xy), or (xy).
• Ts denotes the future lifetime random variable; Ks = bTs c denotes the

curtate future lifetime random variable.
R∞
• e̊s = E(Ts ) = 0 t ps dt
P
• es = E(Ks ) = ∞ k=1 k ps
Rn
• e̊s:n = E[min(Ts , n)] = 0 t ps dt
P
• es:n = E[min(Ks , n)] = nk=1 k ps
R∞
• V ar(Ts ) = 0 2tt ps dt - (e̊s )2
P
• V ar(Ks ) = ∞ k=1 (2k − 1)k ps - (es )
2
Consider two lives each aged x with mortality that follows de Moivre’s
Law/uniform distribution with limiting age ω:
ω−x
– e̊xx = 3
2(ω−x)
– e̊xx = 3
– For x < y: e̊xy = y−x px e̊yy + y−x qx e̊y
With constant forces of mortality for (x), (y), and (xy):

1 1
– e̊x = µx and e̊y = µy
1 1
– e̊xy = µxy , with independence of (x) and (y), = µx +µy
• e̊xy + e̊xy = e̊x + e̊y
• exy + exy = ex + ey
• Cov[Txy , Txy ] = cov[Tx , Ty ] + (e̊x - e̊xy )(e̊y - e̊xy )

If (x) and (y) are independent =⇒ cov[Tx , Ty ] = 0 =⇒
Cov[Txy , Txy ] = (e̊x - e̊xy )(e̊y - e̊xy )

Insurance and Annuity Benefits:

R∞
• Ās = E(v Ts ) = 0 v t t ps µs (t)dt
“This is the expected present value of a life insurance which pays 1 at
time Ts years after issue.”
– Āx is the expected present value of a whole life insurance which pays
1 at the moment of death of (x).”
– Āxy is the expected present value of a joint life insurance which pays
1 at the moment of the first death of (x) and (y).”
– Āxy is the expected present value of a last survivor insurance which
pays 1 at the moment of the second death of (x) and (y).”
R∞
• ās = E(āTs ) = 0 v t t ps dt
“This is the expected present value of a life annuity which pays 1 per
year continuously each year up until time Ts years after issue.”
– āx is the expected present value of a whole life annuity which pays 1
per year continuously each year while (x) survives.”
– āxy is the expected present value of a joint life annuity which pays 1
per year continuously each year while both (x) and (y) survive.”
– āxy is the expected present value of a last survivor annuity which pays
1 per year continuously each year while at least one of (x) and (y)
survive.”
1−Ās
• ās = δ
Ās
• P̄ (Ās ) = ās
“This is the level annual benefit premium for a fully continuous whole
life insurance of 1 on (s).”

• Āxy + Āxy = Āx + Āy
• āxy + āxy = āx + āy
With constant forces of mortality for (x), (y), and (xy):

µxy µx +µy
– Āxy = µxy +δ , with independence, = µx +µy +δ
1 1
– āxy = µxy +δ , with independence, = µx +µy +δ
• BUT: P̄ (Āxy ) + P̄ (Āxy ) 6= P̄ (Āx ) + P̄ (Āy ), as each benefit premium is

a quotient of an insurance and an annuity.
ALSO: t V̄ (Āxy ) + t V̄ (Āxy ) 6= t V̄ (Āx ) + t V̄ (Āy ), as each benefit reserve
is an additive function involving insurances and annuities.
• V ar(v Ts ) = 2 Ās − (Ās )2

“This is the variance of the present value of a continuous whole life in-
surance of 1 on (s).”
• We can also consider insurances and annuities with terms. For example:
Rn
ās:n = 0 v t t ps dt
• If (x) and (y) are independent =⇒ cov[v Txy , v Txy ] = (Āx − Āxy )(Āy − Āxy )

• If Tx < Ty , the following hold:
– e̊xy < e̊x < e̊y < e̊xy

– Āxy < Āy < Āx < Āxy
– āxy < āx < āy < āxy
The same relationships above hold if the circles are removed from the
e’s and the bars on the A’s and a’s are removed or replaced with double
dots.
• Reversionary Annuity: This is a life annuity which provides payments

while specific lives are alive and other specific lives are dead.
For example, consider an annuity that pays 1 per year continuously each
year while (u) is alive and (v) is dead. The expected present value of this
reversionary annuity is:
āv|u = āu - āuv
This formula makes sense:
– If (u) dies first, āv|u = āu - āu = 0; there are no payments.

– If (v) dies first, āv|u = āu - āv > 0; there are payments while only (u)
is alive.

• Consider a continuous life annuity on (x) and (y) that provides the fol-
lowing payments:
(i) 300 per year while both are alive.

(ii) 400 per year while only (x) is alive.
(iii) 200 per year while only (y) is alive.
There are two approaches for formulating the expected present value of
this annuity:
– Recognize this as the sum of a joint life annuity and two reversionary
annuities. Then, the expected present value is:
300āxy + 400āy|x + 200āx|y
= 300āxy + 400[āx - āxy ] + 200[āy - āxy ]
= 400āx + 200āy - 300āxy .
– Recognize this as the sum of two single life annuities; when consider-
ing just (x), 400 is payable each year and when considering just (y),
200 is payable each year. When both (x) and (y) are alive, the total
payment is 600 per year; 300 more per year than we want paid while
both are alive. Adjusting for the excess joint life payment each year
of 300, the expected present value can immediately be reasoned as:
400āx + 200āy - 300āxy .
• The formulas in this section can also be easily adjusted for a discrete
insurance or annuity. For example:
P
– As = ∞ k=0 v
k+1
k| qs
– äxy + äxy = äx + äy

– av|u = au - auv

14.2 Exercises 271
14.2 Exercises
14.1. Consider two independent lives (35) and (45). Each life has mortality
that follows the Illustrative Life Table.
Calculate the probability that (35) will survive ten years and (45) will not
survive ten years.
(A) 0.040 (B) 0.045 (C) 0.050 (D) 0.055 (E) 0.060
14.2. Consider two independent lives (x) and (y).

(x) is subject to a constant force of mortality of 0.05, and (y) is subject to
a constant force of mortality of 0.03.
Calculate the probability that exactly one of (x) and (y) dies in the fifth
year.
(A) 0.062 (B) 0.064 (C) 0.066 (D) 0.068 (E) 0.070
14.3. The time-until-death random variables T (x) and T (y) are indepen-
dent and identically distributed, each with p.d.f:
fT (t) = 3(1 + t)−4 for t > 0.
Calculate the probability that the first death of (x) and (y) will occur
during the second year from today.
(A) 0.010 (B) 0.011 (C) 0.012 (D) 0.013 (E) 0.014
14.4. Consider two independent lives (60) and (65). You are given:
x 60 61 62 63 64 65 66 67 68 69 70 71
lx 1000 982 963 943 922 900 877 853 828 802 775 747
Calculate the probability that the second death of (60) and (65) occurs
during the third year.
(A) 0.001 (B) 0.002 (C) 0.003 (D) 0.004 (E) 0.005

14.2 Exercises 272
(i) Mortality follows de Moivre’s Law with ω = 100.

(ii) (40) and (50) have independent future lifetimes.
Calculate the probability that the first death of (40) and (50) occurs in the
tenth year.
(A) 0.024 (B) 0.026 (C) 0.028 (D) 0.030 (E) 0.032
14.6. For two independent lives ages 30 and 34, you are given:
x 30 31 32 33 34 35 36 37
qx 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80
Calculate the increase in the value of 2 q30:34 if q31 is increased from 0.20 to
0.25.
(A) 0.031 (B) 0.036 (C) 0.041 (D) 0.046 (E) 0.051
14.7. Suppose both (40) and (60) have independent future lifetimes and
mortality where lx = 500(100 - x) for 0 ≤ x ≤ 100.
(A) 32 (B) 33 (C) 34 (D) 35 (E) 36
14.8. Consider (x) and (y), with independent future lifetimes and constant
forces of mortality equal to 0.03 and 0.05, respectively.
Calculate the expected time until the first death of (x) and (y).
(A) 13 (B) 20 (C) 33 (D) 40 (E) 57

14.2 Exercises 273
14.9. For independent lives (50) and (60) with identical expected mortality:
10| q50:60 = 0.011, 10 p50 = 0.78, 10 p60 = 0.73, and q70 = 0.03.
Find the probability that exactly one of (50) and (60) survives 11 years.
(A) 0.35 (B) 0.37 (C) 0.39 (D) 0.41 (E) 0.43
14.10. For a certain population, some lives are infected with disease A:
(i) For a life not infected with disease A: lx = 500(95 - x) for 0 ≤ x ≤ 95.
(ii) For a life infected with disease A: lx = 500(95 - x)α for 0 ≤ x ≤ 95 and
α > 0.
(iii) At age 40, the complete expectation of life for a life not infected with
disease A is 15% higher than the complete expectation of life for a life infected
with disease A.
(iv) The complete expectation of life for a joint life status involving a 45
year old not infected with disease A and a 40 year old infected with disease
A is 15.94.
Calculate the complete expectation of life for a last survivor status involv-
ing a 45 year old not infected with disease A and a 40 year old infected with
disease A.
(A) 31 (B) 32 (C) 33 (D) 34 (E) 35
14.11. For (60.6) and (62):
(i) Each life has mortality such that:

x 60 61 62 63 64 65 66
px 0.80 0.79 0.78 0.77 0.76 0.75 0.74
(ii) The two lives are independent.

(iii) There is a uniform distribution of deaths within each year of age.
Calculate the probability that at least one life survives 5 years.
(A) 0.46 (B) 0.47 (C) 0.48 (D) 0.49 (E) 0.50

14.2 Exercises 274
14.12. Consider two lives (40) and (50):
(i) Both lives have independent future lifetimes.

1
(ii) Each life is subject to: µx = 90−x for 0 ≤ x ≤ 90.
Calculate the probability that both lives will survive a number of years
equal to the expected time until the first death.
(A) 0.43 (B) 0.45 (C) 0.47 (D) 0.49 (E) 0.51
14.13. Consider two independent lives (35) and (35):

ω−x
(i) Each life has mortality such that: s(x) = ω for 0 ≤ x ≤ ω.
70
(ii) e̊35:35 = 3.
Calculate cov[T (35 : 35), T (35 : 35)].
(A) 124 (B) 127 (C) 130 (D) 133 (E) 136
14.14. (i) There are two independent lives, each age 90.
(ii) Each life has mortality such that: l90+t = (10 − t)2 for 0 ≤ t ≤ 10.
Calculate the average time until the second death.
(A) 3 (B) 4 (C) 5 (D) 6 (E) 7
14.15. For a population of smokers and non-smokers:
(i) Smokers have force of mortality: µx = 3(2 + x)−1 for x ≥ 0.

(ii) Non-smokers have force of mortality equal to one-third of the force of
mortality for smokers at each age.
Consider two independent lives each age 40, one a smoker and the other
a non-smoker. Calculate the probability that the first death occurs after 6
years, but before 12 years.
(A) 0.21 (B) 0.22 (C) 0.23 (D) 0.24 (E) 0.25

14.2 Exercises 275
14.16. Consider two independent lives ages 45 and 55, each with identical
expected mortality:
(i) 10 p45 = 0.8900

(ii) 15 q55 = 0.2518
(iii) 10 q45:55 = 0.0154
Calculate: 5 p65 .
(A) 0.8600 (B) 0.8650 (C) 0.8700 (D) 0.8750 (E) 0.8800
14.17. For two lives (x) and (y):
(i) (x) and (y) have independent future lifetime random variables.
(ii) µx (t) = − ln(0.90) for t ≥ 0
(iii) µy (t) = − ln(0.85) for t ≥ 0
Calculate the probability that (x) and (y) both die in the same year.
(A) 0.060 (B) 0.064 (C) 0.068 (D) 0.072 (E) 0.076

14.2 Exercises 276
14.18. For a special insurance on independent lives (x) and (y):
(i) The insurance pays 1000 at the moment of the first death and 2000 at
the moment of the second death.
(ii) Annual premiums of π are payable continuously while both (x) and (y)
are alive. After the moment of the first death, annual premiums of 0.6π are
payable continuously until the moment of the second death.
(iii) δt = 0.06, µx (t) = 0.02, µy (t) = 0.04 for t ≥ 0.
Calculate π using the equivalence principle.
(A) 65 (B) 66 (C) 67 (D) 68 (E) 69
14.19. A fully discrete life insurance on two independent lives, ages 25 and
35, pays 1000 at the end of the year of the first death and 2000 at the end of
the year of the second death. Furthermore:
(i) Level benefit premiums are payable at the beginning of the year while
both are alive.
(ii) Mortality for each life follows the Illustrative Life Table, and i = 0.06
(A) 17 (B) 18 (C) 19 (D) 20 (E) 21
14.20. For a special fully continuous last-survivor insurance of 1000 on

independent lives (x) and (y):
(i) The death benefit is payable at the moment of the second death.
(ii) The level annual benefit premium is π while both (x) and (y) are alive
and 0.75π while exactly one of (x) and (y) is alive.
(iii) δ = 0.04, µx (t) = 0.02, µy (t) = 0.03 for t > 0
Calculate: π.
(A) 12 (B) 13 (C) 14 (D) 15 (E) 16

14.2 Exercises 277
14.21. Consider a special life insurance on John and Daniel, each aged 30:
(i) John and Daniel have independent future lifetimes.

(ii) John and Daniel have identical expected mortality.
(iii) 1000 is payable at the end of the year of death of John.
(iv) 1000 is payable at the end of the year of death of Daniel.
(v) Level annual benefit premiums are payable at the beginning of each
year for as long as both John and Daniel are alive.
(vi) ä31 = 14 and ä31:31 = 12.
(vii) p30 = 0.98
(viii) d = 0.06
(A) 22 (B) 24 (C) 26 (D) 28 (E) 30
14.22. You are given a fully discrete 2-year term insurance on independent
lives (30) and (40):
(i) The death benefit is 10,000, payable at the end of the year of the last
death of (30) and (40).
(ii) The level annual premium is 50, and is payable at the beginning of the
year while at least one of (30) and (40) is alive.
(iii) Each life has mortality that follows the Illustrative Life Table.
(iv) i = 0.06
Calculate the expected loss-at-issue.
(A) - 16 (B) - 9 (C) 0 (D) 9 (E) 16

14.2 Exercises 278
14.23. A continuous annuity on (x) and (y) provides the payments:
(i) 500 per year while both are alive.

(ii) 400 per year while only (x) is alive.
(iii) 150 per year while only (y) is alive.
The expected present value of this annuity is 6650.

If āy = 9 and āxy = 6, re-calculate the expected present value of the annuity
if the 150 per year payment while only (y) is alive is reduced to 90 per year.
(A) 6110 (B) 6230 (C) 6375 (D) 6470 (E) 6540
14.24. John and Dwayne are independent lives, ages 30 and 40, respec-
tively. They want to purchase a special life annuity-due that will pay 1000
per year while either of them is alive and over age 60. Assuming that mor-
tality for each life follows the Illustrative Life Table and i = 0.06, calculate
the expected present value of this annuity.
(A) 3500 (B) 3600 (C) 3700 (D) 3800 (E) 3900
14.25. A husband, age 65, is retiring. He has the choice of one of two
pension plans (Plan I and Plan II, respectively) that will provide annual
retirement benefits to him and his wife, age 55, starting today:
(i) Plan I: Payments of: 20,000 at the beginning of each year for as long
as both the husband and wife are alive; 15,000 at the beginning of each year
if the wife is alive and the husband is dead; 7500 at the beginning of each
year if the husband is alive and the wife is dead.
(ii) Plan II: Payments of: R at the beginning of each year for as long
as the wife is alive; 0.60R at the beginning of each year for as long as the
husband is alive.
(iii) Mortality follows the Illustrative Life Table, and i = 0.06. The husband
and wife have independent future lifetimes.
Find R so that the actuarial present values of Plan I and Plan II are equal.
(A) 13,000 (B) 13,500 (C) 14,000 (D) 14,500 (E) 15,000

14.2 Exercises 279
14.26. Consider two independent lives (35) and (45):
(i) Mortality for each life follows the Illustrative Life Table.
(ii) i = 0.06
Calculate the expected present value of an insurance on (35) and (45) that
pays 5000 at the end of the year of the first death if the first death occurs
after the first 10 years.
(A) 900 (B) 910 (C) 920 (D) 930 (E) 940
14.27. Independent lives (50) and (60) purchase a special annuity-immediate

with annual payments of:
(i) 500 while both (50) and (60) are alive.

(ii) 300 while (60) is alive and (50) is dead.
(iii) B while (50) is alive and (60) is dead.
The special annuity-immediate has the same actuarial present value as a

whole life annuity-immediate on (50) with annual payments of 1.5B.
Mortality for each life follows the Illustrative Life Table, i = 0.06.
Calculate B.
(A) 260 (B) 280 (C) 300 (D) 320 (E) 340
14.28. You are pricing a special continuous life annuity on two independent
lives, (x) and (y):
(i) The annuity pays 30,000 each year while both lives survive, and 20,000
each year while only one life survives.
(ii) (x) and (y) have constant forces of mortality equal to 0.03 and 0.05,
respectively.
(iii) δ = 0.06.
(A) 331,600 (B) 332,100 (C) 332,600 (D) 333,100 (E) 333,600

14.2 Exercises 280
14.29. You are given a special annuity on two independent people each
aged x:
(i) The annuity pays 40,000 at the beginning of the year while both people
are alive and is 25,000 at the beginning of the year when only one person is
alive.
(ii) There is also a benefit of 20,000 payable at the end of the year of the
first death.
(iii) Both people have mortality: k px = (0.97)k for k = 0, 1, 2, ....
(iv) d = 0.04
Calculate the actuarial present value of this special annuity.
(A) 633,100 (B) 633,600 (C) 634,100 (D) 634,600 (E) 635,100
14.30. Consider a special continuous annuity on independent lives: a female

aged 40 and a male aged 50. You are given:
(i) The annual rate of payment is 30,000, only payable while exactly one
of (40) and (50) is alive and is older than age 65.
(ii) (40) is subject to a constant force of mortality of 0.02.
(iii) (50) is subject to a constant force of mortality of 0.03.
(iv) δ = 0.05
(A) 137,700 (B) 138,170 (C) 139,900 (D) 140,300 (E) 141,100

14.2 Exercises 281
14.1. D 14.26. B
14.2. B 14.27. D
14.3. E 14.28. C
14.4. C 14.29. E
14.5. D 14.30. B
14.6. B
14.7. C
14.8. A
14.9. C
14.10. C
14.11. A
14.12. B
14.13. E
14.14. C
14.15. B
14.16. C
14.17. B
14.18. D
14.19. B
14.20. A
14.21. D
14.22. A
14.23. D
14.24. E
14.25. A


• Exam 3L, Fall 2013: #4, 5
• Exam 3L, Spring 2013: #4, 5
• Exam 3L, Fall 2012: #4, 5, 13
• Exam 3L, Spring 2012: #4, 5
• Exam MLC, Sample Questions: #1, 31, 57, 80, 91, 94, 123, 128, 150,
163, 191, 193, 195, 220, 269, 270, 272, 273, 278, 280
• Exam 3L, Fall 2011: #3, 4, 5, 13
• Exam 3L, Spring 2011: #4, 5, 13
• Exam 3L, Fall 2010: #4, 5
• Exam 3L, Spring 2010: #5, 6, 7
• Exam 3L, Fall 2009: #4, 5, 6
• Exam 3L, Spring 2009: #4, 5, 23
• Exam 3L, Fall 2008: #15, 16, 17
• Exam 3L, Spring 2008: #17, 18
• Exam MLC, Sample Questions: #26, 46, 49, 104, 112, 173, 233, 281, 282

283
15 MULTIPLE LIVES II (Exam MLC)
15.1 Key Concepts

• The multiple life model described so far can be expressed as a 4-state

Markov Chain. Consider two lives, (x) and (y). The following 4-state
model will hereafter be referred to as the Multiple Life Model diagrammed
in Examples of Multiple State Models:
– State 0: Both (x) and (y) are alive

– State 1: (x) is alive and (y) is dead
– State 2: (x) is dead and (y) is alive
– State 3: Both (x) and (y) are dead
• The forces of transition are:
– µ13
x+t
– µ23
y+t
– µ02 13
x+t:y+t , with independence (x) and (y), = µx+t = µx (t)
– µ01 23
x+t:y+t , with independence of (x) and (y), = µy+t = µy (t)
• In the Multiple Life Model, it is assumed that µ03

x+t:y+t = 0. That is, it is
impossible for (x) and (y) to die simultaneously.

Multiple Life Functions:

The multiple life functions described in Multiple Lives I can be expressed
using the Multiple Life Model.
• Probabilities:
– t p00
xy = t pxy
– t p01 02 03
xy + t pxy + t pxy = t qxy
– t p00 01 02
xy + t pxy + t pxy = t pxy
– t p03
xy = t qxy
– µ01 02
x+t:y+t + µx+t:y+t = µxy (t)
• Insurance and Annuity Functions:

R∞
– Āx = 0 v t (t p00 02 01 13
xy µx+t:y+t + t pxy µx+t )dt
R∞
– Āy = 0 v t (t p00 01 02 23
xy µx+t:y+t + t pxy µy+t )dt
R∞
– Āxy = 0 v t t p00 01 02
xy (µx+t:y+t + µx+t:y+t )dt
R∞
– Āxy = 0 v t (t p01 13 02 23
xy µx+t + t pxy µy+t )dt
R∞
– āx = 0 v t (t p00 01
xy + t pxy )dt
R∞
– āy = 0 v t (t p00 02
xy + t pxy )dt
R∞
– āxy = 0 v t t p00
xy dt
R∞
– āxy = 0 v t (t p00 01 02
xy + t pxy + t pxy )dt
R∞
– āx|y = 0 v t t p02
xy dt

Contingent Functions:
• A contingent function is such that the function value depends on the

order in which the lives die. Assume the Multiple Life Model.
1
Rt 00 02
• t q xy = 0 s pxy µx+s:y+s ds
“The probability that (x) dies both before (y) and within t years.”
1
Rt
If (x) and (y) are independent =⇒ t q xy = 0 (s py )s px µx (s)ds
Rt
• t q xy2 = 02 23
0 s pxy µy+s ds
“The probability that (y) dies both after (x) and within t years.” This
1 1
is different than t q xy ; for t q xy the event of interest is (y) dying after (x),
but that does not necessarily have to occur within t years.
Rt
If (x) and (y) are independent =⇒ t q xy2 = 0 (s qx )s py µy (s)ds
• Furthermore:
1
– t qx = t q xy + t q yx2
1 1
– t qxy = t q xy + t q yx
– t qxy = t q yx2 + t q xy2
1
– t q xy = t q xy2 + t p02
xy
1 2
– ∞ q xy = ∞ q xy
– For independent lives (x) and (y) each with de Moivre’s Law/uniform
distribution with limiting age ω:
1
∗ t q xy = (t qx )(t/2 py ) and t q yx2 = 21 (t qx )(t qy )
– For independent lives (x) and (y) with constant forces of mortality
µx and µy , respectively:
1 µx 1 µx
∗ t q xy = µx +µy (t qxy ) and ∞ q xy = µx +µy

1
R∞
• Āxy = 0 v s s p00 02
xy µx+s:y+s ds
“This is the expected present value of a life insurance that pays 1 at the
moment of death of (x) if (x) is the first to die.”
1
R∞
If (x) and (y) are independent =⇒ Āxy = 0 v s (s py )s px µx (s)ds
R∞
• Āxy2 = 0 v s s p02 23
xy µy+s ds
“This is the expected present value of a life insurance that pays 1 at the
moment of death of (y) if (y) is the second to die.”
R∞
If (x) and (y) are independent =⇒ Āxy2 = 0 v s (s qx )s py µy (s)ds
• Furthermore:
1
– Āx = Āxy + Āyx2
1 1
– Āxy = Āxy + Āyx
– Āxy = Āxy2 + Āyx2
– For independent lives (x) and (y) with constant forces of mortality
µx and µy , respectively:
1 µx
∗ Āxy = µx +µy +δ
µy µx
∗ Āxy2 = µy +δ µx +µy +δ

Common Shock Model:
• Under the Common Shock Model, is possible to make a direct transition

from State 0 to State 3 in the Multiple Life Model, that is, µ03
x+t:y+t 6= 0.
• It is assumed that there is an event called the common shock that could
kill both (x) and (y) at the same time, such as a car accident or a natural
disaster.
Rt
• t p11
x+s = exp[− 0 µ13
x+s+u du]
“This is the probability that (x + s) survives t years ((y) has previously

died).”
Rt
• t p22
y+s = exp[− 0 µ23
y+s+v dv]
“This is the probability that (y + s) survives t years ((x) has previously

died).”
Rt
• t p00
xy = exp[−
01
0 (µx+s:y+s + µ02 03
x+s:y+s + µx+s:y+s )ds] = t pxy
“This is the probability that both (x) and (y) survive t years.”
Rt
• t p01
xy =
00 01 11
0 s pxy (µx+s:y+s )t−s px+s ds
“This is the probability that only (x) survives t years.”
Rt
• t p02
xy =
00 02 22
0 s pxy (µx+s:y+s )t−s py+s ds
“This is the probability that only (y) survives t years.”
• t p03 00 01 02
xy = 1 - t pxy - t pxy - t pxy = t qxy
“This is the probability that (x) and (y) do not survive t years.”

• Consider a typical special case of the common shock model:

(i) µ03
x+s:y+s = λ, a constant
(ii) µ02 03 13
x+s:y+s + µx+s:y+s = µx+s = µx , a constant
(iii) µ01 03 23
x+s:y+s + µx+s:y+s = µy+s = µy , a constant
Then:
– t px = exp[−µx t]
– t py = exp[−µy t]
– t pxy = exp[−(µx + µy − λ)t]
– t pxy = t px + t py - t pxy
• The probability that (x) and (y) are killed simultaneously by the common
shock is:
R ∞ 00 03
0 s pxy µx+s:y+s ds.
“In order for both (x) and (y) to die at the same time, both have to
survive for s years and then immediately die. Integration considers all
possible times s.”
– Consider the typical special case of the common shock model previ-
ously described; the probability that (x) and (y) die simultaneously
is:
λ
µx +µy −λ .

15.2 Exercises 289
15.2 Exercises
15.1. Consider the Multiple Life Model in Examples of Multiple State Models,
with a husband aged x and a wife aged y.
You are given the following transition intensities for t > 0:
(i) µ01
x+t:y+t = 0.03
(ii) µ02
x+t:y+t = 0.04
(iii) µ13
x+t = 0.07
(iv) µ23
y+t = 0.06
Calculate the probability that the wife dies within 10 years.
(A) 0.26 (B) 0.27 (C) 0.28 (D) 0.29 (E) 0.30
15.2. Consider the setup provided in Exercise 15.1. Calculate the proba-
bility that the husband dies both before the wife dies and within 10 years.
(A) 0.26 (B) 0.27 (C) 0.28 (D) 0.29 (E) 0.30
15.3. Consider the setup provided in Exercise 15.1. Given δ = 0.03, cal-
culate the expected present value of a life annuity that pays 1000 per year
continuously each year while the husband is alive.
(A) 13,000 (B) 13,500 (C) 14,000 (D) 14,500 (E) 15,000
15.4. Consider independent lives (25) and (50). Each life has mortality
such that: lx = 100(100 - x) for x ≤ 100.
Calculate the probability that (25) dies after (50) and within 8 years.
(A) 0.007 (B) 0.009 (C) 0.011 (D) 0.013 (E) 0.015

15.2 Exercises 290
15.5. Consider independent lives (25) and (50), each with force of mortality
1
µx = 100−x for 0 ≤ x < 100.
Calculate the probability that (50) dies before (25) and within 25 years.
(A) 0.08 (B) 0.17 (C) 0.25 (D) 0.33 (E) 0.42
15.6. Consider (40) and (50) with independent future lifetime random
variables:
(i) (40) has mortality that follows: lx = 95 - x for 0 ≤ x ≤ 95.

(ii) (50) has a constant force of mortality equal to 0.05.
Calculate the probability that (40) dies after (50) and within 10 years.
(A) 0.03 (B) 0.04 (C) 0.05 (D) 0.06 (E) 0.07
15.7. For independent lives Shane and Britney:
(i) Shane, aged 50, has mortality such that: t q50 = 0.030t and t q51 = 0.031t
for 0 ≤ t ≤ 1.
(ii) Britney, aged 40, has mortality that follows the Illustrative Life Table
with the assumption of a uniform distribution of deaths within each year of
age.
Calculate the probability that Britney dies both within 2 years and before
Shane.
(A) 0.003 (B) 0.004 (C) 0.005 (D) 0.006 (E) 0.007

15.2 Exercises 291
15.8. Consider independent lives (25) and (50):
(i) Each life has mortality such that: lx = 100(100 - x) for 0 ≤ x ≤ 100.
(ii) δ = 0.05
Calculate the actuarial present value of a life insurance that pays 1 at the
moment of death of (25) if (25) is the second to die.
(A) 0.07 (B) 0.08 (C) 0.09 (D) 0.10 (E) 0.11
15.9. Consider the setup provided in Exercise 15.8. Calculate the actuarial
present value of a life insurance that pays 1 at the moment of death of (25)
if (25) is the first to die.
(A) 0.16 (B) 0.17 (C) 0.18 (D) 0.19 (E) 0.20
15.10. For independent lives (30) and (35):
(i) µ30 (t) = 0.05 for t > 0

1
(ii) µ35 (t) = 55−t for 0 < t < 55
Calculate the probability that (35) dies before (30).
(A) 0.34 (B) 0.40 (C) 0.46 (D) 0.54 (E) 0.60
15.11. Two lives (x) and (y) have mortality such that:
(i) The Common Shock Model applies.

(ii) µ03
x+t:y+t = 0.005 for t > 0
(iii) µ02 03 13
x+t:y+t + µx+t:y+t = µx+t = 0.015 for t > 0
(iv) µ01 03 23
x+t:y+t + µx+t:y+t = µy+t = 0.015 for t > 0
Calculate the probability that both (x) and (y) are dead within 10 years.
(A) 0.05 (B) 0.06 (C) 0.07 (D) 0.08 (E) 0.09

15.2 Exercises 292
15.12. Two lives (x) and (y) have mortality such that:

(ii) µ01
x+t:y+t = 0.03 for t > 0
(iii) µ02
x+t:y+t = 0.06 for t > 0
(iv) µ03
x+t:y+t is a constant for t > 0
(v) µ02 03 13
x+t:y+t + µx+t:y+t = µx+t for t > 0
(vi) µ01 03 23
x+t:y+t + µx+t:y+t = µy+t for t > 0
(vii) The probability that (x) survives 5 years is 0.67.

Calculate the probability that (x) and (y) die simultaneously.
(A) 0.12 (B) 0.14 (C) 0.16 (D) 0.18 (E) 0.20
15.13. Consider the setup in Exercise 15.12. Estimate 0.4 p03

xy using Euler’s
Method with step 0.2 to numerically solve Kolmogorov’s forward equations.
(A) 0.004 (B) 0.006 (C) 0.008 (D) 0.010 (E) 0.012
15.14. For a special last-survivor insurance of 100,000 on (x) and (y):
(i) The death benefit is payable at the moment of the second death.
(ii) Level annual benefit premiums of π are payable continuously each year
only while exactly one of (x) and (y) is alive.
(iii) The Common Shock Model applies.
(iv) For t > 0: µ03

x+t:y+t = 0.02,
02
µx+t:y+t + µx+t:y+t = µ13
03
x+t = 0.05,
µx+t:y+t + µx+t:y+t = µ23
01 03
y+t = 0.07
(v) δ = 0.06
Calculate: π.
(A) 8400 (B) 8500 (C) 8600 (D) 8700 (E) 8800

15.2 Exercises 293
15.15. Consider a special last survivor insurance of 1 on (x) and (y):
(ii) δ = 0.08
(iii) µ01 02
x+t:y+t = µx+t:y+t = 0.04 for t > 0
(iv) µ13 23
x+t = µy+t = 0.06 for t > 0
(v) µ03
x+t:y+t = 0
Calculate the expected present value of the last survivor insurance.
(A) 0.21 (B) 0.23 (C) 0.25 (D) 0.27 (E) 0.29

15.2 Exercises 294
15.1. D
15.2. D
15.3. A
15.4. B
15.5. E
15.6. B
15.7. D
15.8. C
15.9. B
15.10. A
15.11. B
15.12. D
15.13. C
15.14. C
15.15. A


265, 266, 268, 271, 279

296
16 OTHER TOPICS (Exam MLC Only)
16.1 Key Concepts

PENSION MATHEMATICS:
• A defined benefit pension plan determines the future periodic pension

benefit by a formula that is a function of an employee’s earnings history
(career average salary or salary in the final working years), age, and
number of years of service. The employer is primarily responsible for
making contributions into a fund such that the accumulated fund value
can provide the predetermined periodic pension benefits.
• A defined contribution pension plan predetermines the contributions that

the employee (and often the employer as well) pay into a fund that will
accumulate interest and provide the future pension benefits. The em-
ployee is primarily responsible for the investment risk; the amount of the
future periodic pension benefit depends on the fund’s performance.
• The replacement ratio determines the benefit for a defined benefit pension
plan and the target benefit for a defined contribution pension plan:
Pension Income in the Year After Retirement
R= Salary in the Year Before Retirement
The replacement ratio tends to be between 50% and 70%.
• The salary scale function, sx , is a constant multiple of the salary earned

by an employee in the year between ages x and x + 1.
• If the employee’s salary between ages x and x + 1 is c, the employee’s

s
projected salary between ages y and y + 1 is: c sxy .
• Assume salaries are revised continuously. If the employee’s salary be-

tween ages x and x + 1 is c, the employee’s projected rate of salary at
s
age y + 1 is: c y−0.5
sx .
• Assume salaries are revised continuously. If the employee’s rate of salary

at age x + 1 is c, the employee’s projected rate of salary at age y + 1 is:
sy−0.5
c sx−0.5 .

INTEREST RATE RISK:
• v(t) = the current market price of a t-year zero-coupon bond

with a face amount of 1.
• yt = the t-year spot interest rate such that:

v(t)(1 + yt )t = 1 =⇒ v(t) = (1 + yt )−t .
• The term structure of interest rates describes the relationship between yt

and t. The yield curve is a plot of yt versus t, and is usually an increasing
function of t.
• f (t, t + k) = t-year, k-year forward interest rate such that:

(1+yt+k )t+k v(t)
(1 + f (t, t + k))k = (1+yt )t = v(t+k)
• Expected present value formulas for life insurances and annuities with
interest rates that vary over time (yt ) are analogous to the expected
present value formulas with constant interest rates (i). For example:
R∞
– Ā(x) = 0 v(t)t px µx+t dt
P
– ä(x : n ) = n−1
k=0 v(k)k px
• Consider a portfolio with N independent and identically distributed life

insurance policies. The random variable Xi equals some quantity of in-
terest for policy i (such as the future loss at issue), where i = 1, 2, ...,
N.
The risk in the portfolio, as measured by Xi , is diversifiable if:
q P
limN →∞ N V ar[ N
1
i=1 Xi ] = 0.
Let Xi depend on the random variable Y (example: an interest rate):
– E[Xi ] = E[E[Xi |Y ]]
– V ar[Xi ] = E[V ar[Xi |Y ]] + V ar[E[Xi |Y ]]

EMERGING COSTS FOR TRADITIONAL LIFE INSURANCE:
• Consider a fully discrete life insurance of S on (x) with an n-year term

and gross premiums of G per year. For k = 0, 1, 2, ..., n - 1:
– The effective annual interest rate is i.

– The only decrement is death.
– qx+k denotes the mortality rate between ages x + k and x + k + 1.
– E0 denotes acquisition costs: expenses incurred prior to issue but are
actually paid at issue.
– Ek+1 denotes expenses paid at the start of the year from k to k + 1.
– k V denotes the reserve at time k.
– P rk+1 denotes the profit emerging at the end of year k + 1 assuming
the policy was in force at the start of the year from k to k + 1.
• P r0 = - E0 .
• P rk+1 = (G + k V − Ek+1 )(1 + i) - Sqx+k - k+1 V px+k

“This is the profit emerging at the end of year k + 1 per policy in force
at the start of year k + 1; this is the expected profit per survivor. Take
the net cash flow received at the start of the year (G + k V − Ek+1 ) and
accumulate it for interest at i to the end of the year. Then, subtract the
expected benefit payable at the end of the year (Sqx+k + k+1 V px+k ).”
• Note: Different assumptions for interest, mortality, and expenses are

used to calculate the premium, reserves, and profits. Each assumption
is referred to as a basis. Therefore, a profit calculation problem will
often involve a premium basis, a reserve basis, and a profit test basis.
The profit test basis will have the most realistic assumptions for interest,
mortality, and expenses; the premium and reserve bases will be more
conservative than the profit test basis.

• P r = (P r0 , P r1 , ..., P rn )0 = profit vector
• Π = (Π0 , Π1 , Π2 , Π3 , ..., Πn )0
= (P r0 , P r1 , px (P r2 ), 2 px (P r3 )..., n−1 px (P rn ))0 = profit signature
“Πk+1 denotes the profit emerging at the end of year k + 1 per issued
policy. That is, it is only assumed that the policy was issued to a life
aged x, and does NOT assume survival of (x) to the start of year k + 1
for k = 0, 1, ..., n - 1; this is the expected profit per issued policy.”
• Let r denote the risk discount rate (or hurdle interest rate). The risk
discount rate is the rate used to discount profits.
• Consider the following profit measures: different ways of measuring profit

based on a policy’s profit signature.
– The net present value (NPV) is the expected present value of the fu-
ture profits such that:
P
N P V = nk=0 Πk vrk .
The policy will be profitable if the NPV is positive.
– The internal rate of return (IRR) is the interest rate j such that:
N P V @j = 0.
The policy will be adequately profitable if the IRR is greater than
the risk discount rate.
N P V @r
– The profit margin is such that: Profit Margin = Gäx:n @r
The profit margin is the NPV expressed as a proportion of the ex-

pected present value of the premiums at the risk discount rate.
– The discounted payback period (DPP) is the smallest value of m such

that:
Pm k
k=0 Πk vr ≥ 0.
m denotes the number of years the policy needs to remain in force

until the policy becomes profitable.

• The previous equations can easily be modified to consider a multiple

decrement model. Suppose that there are m decrements. For j = 1, 2,
..., m; let the benefit payable at the end of the year of decrement due to
cause j equal b(j) .
P r0 = - E0 .
For k = 0, 1, 2, ..., n - 1:
Pm (j) (j) (τ )
– P rk+1 = (G + k V − Ek+1 )(1 + i) - j=1 b qx+k - k+1 V px+k
(τ )
– Πk+1 = k px P rk+1
• The previous equations can be modified to consider a multiple state

model. Suppose there are m + 1 states. The policyholder is in State
0 at policy issue.
P r0 = - E0 .
– Let k = 0, 1, 2, ..., n - 1. For each non-terminal state i, calculate the

(i)
profit per policy in force at the start of year k + 1: P rk+1 :
(i) (i) P
– P rk+1 = (G(i) + k V (i) − Ek+1 )(1 + i) - mj=0 [b
(j)
+ k+1 V (j) ]pij
x+k
“First, accumulate the net cash flow at the start of the year with
interest to the end of the year and subtract the expected benefit
payable at the end of the year. However, for the net cash flow at
the start of the year: you need to consider premiums, reserves, and
expenses associated with being in State i at the start of the year. For
the expected benefits and reserves required at the end of the year,
consider all of the States j (which may equal i) that the policyholder
could be in at the end of the year.”
This interpretation assumes there are no benefits payable at the be-
ginning of the year. The net cash flow at the start of the year can
easily be adjusted to incorporate such benefits.
P (i)
• Πk+1 = i k p0ix P rk+1 .
“The profit per policy issued is a weighted average of all of the possible
(i)
P rk+1 s. When summing over i, only consider non-terminal states.

UNIVERSAL LIFE INSURANCE:
• Universal Life Insurance is a life insurance with an investment compo-

nent. Universal Life Insurance is an account into which the policyholder
pays premiums and earns interest. Each period, deductions are made for
the cost of insurance (mortality charge) and expenses. The death benefit
and premiums paid are chosen by the policyholder (within reason; for ex-
ample, sufficient funds have to be available to cover the cost of insurance
and expenses), and can often be adjusted while the insurance is in force.
Projecting Account Values:

• Fixed Death Benefit/Type A
Consider a universal life insurance on (x) that pays S at the end of the
year of death. Premiums are payable at the beginning of the year. Let k
= 0, 1, ...:
– AVk denotes the account value at the start of the year from k to k +
1. Unless you are told otherwise, the account value is the reserve
for a universal life insurance.
– Gk denotes the premium at the start of the year from k to k + 1.
– Ek denotes the expenses at the start of the year from k to k + 1.
– ic denotes the credited interest rate, used to calculate account values.
– CoIk denotes the cost of insurance at the start of the year from k to
k + 1, this is the expected present value of the net amount at risk for
year k + 1 (at time k).
iq denotes the interest rate used to calculate CoIk , vq = (1 + iq )−1 .
– qx+k denotes the mortality rate between ages x + k and x + k + 1
used in the calculation of CoIk .

The recursion for projecting the fixed death benefit account value at the
end of year k + 1:
AVk+1 = [AVk + Gk − Ek − CoIk ](1 + ic )
where CoIk = (S − AVk+1 )vq qx+k .

“The account value at the end of year k + 1 is the net cash flow at the
beginning of year k + 1 accumulated at the credited interest rate ic to the
end of the year. The net cash flow at the beginning of year k + 1, at time k,
is the account value at time k plus the premium at time k minus expenses at
time k minus the net amount at risk discounted for interest and mortality at
time k [AVk + Gk − Ek − CoIk ]. Note: the net amount at risk is the additional
funds beyond the account value at the end of year k + 1 required to fund the
death benefit of S (S − AVk+1 ).”
As the required account value AVk+1 appears on both sides of the above
fixed death benefit account value recursion, it can be solved for as follows:
AVk+1 = [AVk + Gk − Ek − Svq qx+k ](1 + ic )[1 − vq qx+k (1 + ic )]−1 .
• Variable Death Benefit/Type B

Consider a universal life insurance on (x) that pays S plus the account
value at the end of the year of death. Premiums are payable at the
beginning of the year. Let k = 0, 1, ...:
The recursion for projecting the variable death benefit account value at
the end of year k + 1:
AVk+1 = [AVk + Gk − Ek − CoIk ](1 + ic )
where CoIk = Svq qx+k .

“The interpretation of the variable death benefit account value recursion
is analogous to the fixed death benefit account value recursion. Note:
the net amount at risk is the additional funds beyond the account value
at the end of year k + 1 required to fund the death benefit of S + AVk+1
(S).”

Profit Testing:
Consider a universal life insurance on (x). Let k = 0, 1, ...:
• AVk denotes the account value at the start of the year from k to k + 1.
• Gk denotes the premium at the start of the year from k to k + 1.
• E0 denotes acquisition costs: expenses incurred prior to issue but are

actually paid at issue.
• Ek+1 denotes the expenses at the start of the year from k to k + 1. All
expenses are under the profit test basis and will likely be less conservative
(lower) than the expenses used to determine account values.
• i denotes the annual effective interest rate used for profit testing. This
rate is under the profit test basis and will likely be less conservative
(higher) than the rates used to determine account values and the cost
of insurance (ic , iq ).
• There are two decrements: mortality and withdrawal.
• bk+1 denotes the death benefit plus any claims settlement expenses payable
at the end of year k + 1, if death occurs between times k to k + 1.
• Ck+1 denotes the cash value/surrender value (account value less any sur-
render charge) plus processing costs payable at the end of year k + 1, if
withdrawal occurs between times k to k + 1.
(d)
• qx+k denotes the mortality rate between ages x + k and x + k + 1.
This rate is under the profit test basis and will likely be less conservative
(lower) than the rate used to determine account values and the cost of
insurance.
(w)
• qx+k denotes the surrender rate between ages x + k and x + k + 1.
(τ ) (d) (w)
• px+k = 1 - qx+k - qx+k , the persistency rate.

Then, the profit emerging at the end of year k + 1 per policy in force at
the start of the year from k to k + 1 is:
(d) (w) (τ )
P rk+1 = (Gk + AVk − Ek+1 )(1 + i) - bk+1 qx+k - Ck+1 qx+k - AVk+1 px+k .
“The profit emerging at the end of year k + 1 per policy in force at the
start of the year from k to k + 1 can be calculated as follows. Take the net
cash flow received at the start of the year (Gk + AVk − Ek+1 ) and accumulate
it for interest at i to the end of the year. Then, subtract the expected benefit
(d) (w) (τ )
payable at the end of the year (bk+1 qx+k + Ck+1 qx+k + AVk+1 px+k ).”
• The corresponding element of the profit signature is:

(τ )
Πk+1 = k px P rk+1 .
• Account values can be projected using the appropriate recursions that

were discussed in Projecting Account Values. Remember to use the
interest, mortality, and expenses under the account value basis which
will likely not equal the interest, mortality, and expenses under the profit
test basis.
• The typical assumptions for the decrements will be that death can occur
throughout the year and surrender can only occur at the end of the year.
Then:
(d) 0(d)
qx+k = qx+k
(w) 0(d) 0(w)

qx+k = (px+k )(qx+k )

16.2 Exercises 305
16.2 Exercises
16.1. A worker exact age 38 on January 1, 2012, earned 80,000 in 2011:
(i) In a defined benefit plan, the retirement age is 65.

(ii) The annual pension benefit is based on the average salary earned in
the three years prior to retirement.
(iii) The salary scale function is such that: sx = (1.04)x . Salaries are
increased on December 31 each year.
Calculate the projected final average salary.
(A) 200,000 (B) 220,000 (C) 240,000 (D) 260,000 (E) 280,000
16.2. A plan member exact age 45 on the valuation date earned salary at
a rate of 100,000 per year at the valuation date. The salary scale function
is: sx = (1.03)x . Assume salaries are increased continuously. Calculate the
projected salary rate at age 55.
(A) 126,000 (B) 128,000 (C) 130,000 (D) 132,000 (E) 134,000
16.3. An employee joins a defined benefit plan at exact age 40.
(i) Between ages 40 and 41, the employee’s salary is 45,000.

(ii) The salary scale is such that: s39 = 1.894, s40 = 2.005, s64 = 3.698.
(iv) i = 0.06
(v) Retirement is mandatory at exact age 65.
(vi) The plan will provide payments of X per year at the beginning each
year starting at age 65. X is determined such that the replacement ratio is
equal to 60%.
Calculate the expected present value of the plan payments at exact age 65.
(A) 490,000 (B) 491,000 (C) 492,000 (D) 493,000 (E) 494,000

16.2 Exercises 306
16.4. For a new participant entering a defined contribution plan at exact

age 45:
(i) The participant’s current salary (between ages 45 and 46) is 25,000.
(ii) The salary scale is sx = (1.05)x . sx is a step function that is constant
over each year of age.
(iii) Retirement is mandatory at exact age 65.
(iv) Contributions equal to 4% of salary are made in the middle of each
year.
(v) i = 0.05
(vi) Ignore mortality and other decrements.
Calculate the accumulated value of the contributions for this participant

at exact age 65.
(A) 48,000 (B) 49,000 (C) 50,000 (D) 51,000 (E) 52,000
16.5. A defined contribution plan has one participant currently aged 40.
(i) Contributions are equal to 2.5% of salary and are made at the middle
of the year.
(ii) The retirement age is 65.
(iii) The participant’s current salary is 29,200.
(iv) The salary scale is given by sx = 10,000(1.035)x . Salary does not
increase during each year of age.
(v) i = 0.08
(vi) Ignore mortality and other decrements.
Calculate the accumulated value of the contributions at age 65 for the

participant.
(A) 76,000 (B) 77,000 (C) 78,000 (D) 79,000 (E) 80,000

16.2 Exercises 307
16.6. An employee is exact age 62 on January 1, 2012:
(i) The employee’s annual salary for 2012 is 125,300. The salary scale
function is: sx = (1.04)x . Salary increases occur each year on December 31.
(ii) The Illustrative Service Table models employee exit from employment
due to four decrements: death while employed (d), withdrawal (w), retire-
ment due to disability (i), and retirement due to age (r).
(iii) The employee is issued a special 3-year insurance on January 1, 2012.
The insurance provides two types of benefits.
(iv) If the employee dies in employment prior to age 65, a lump sum death
benefit equal to twice the employee’s salary at death is paid at the moment
of death. Deaths are assumed to occur in the middle of the year.
(v) If the employee retires due to age prior to age 65, a lump sum retirement
benefit equal to a percentage of the employee’s salary is paid at the end of the
year of retirement. The percentage of the employee’s salary for the retirement
benefit is 20% if retirement occurs between ages 62 and 63, 40% if retirement
occurs between ages 63 and 64, and 60% if retirement occurs between ages
64 and 65.
(vi) The effective annual interest rate for all cash flows is 0.06.
Calculate the expected present value of the employee’s insurance.
(A) 23,000 (B) 24,000 (C) 25,000 (D) 26,000 (E) 27,000
√
t
16.7. You are given the yield curve: yt = 0.03 + 200 .
Calculate the forward rate for a 5-year zero coupon bond to be purchased
at the end of five years.
(A) 0.0500 (B) 0.0505 (C) 0.0510 (D) 0.0515 (E) 0.0520

16.2 Exercises 308
(i) The yield curve in Exercise 16.7.

Calculate the level annual benefit premium for a fully discrete 3-year term
insurance of 10,000 on (35) issued at time t = 0.
(A) 19 (B) 20 (C) 21 (D) 22 (E) 23

(ii) The rate of interest is based on the yield curve at time t = 0. You are
also given the following information based on the yield curve at t = 0:
t Annual Forward Interest Rate

0 0.032
1 0.035
2 0.039
Calculate the expected present value of the insurance.
(A) 57 (B) 59 (C) 61 (D) 63 (E) 65
(i) XN denotes the number of deaths between ages 45 and 50 for N inde-
pendent lives each aged 40.
(ii) Each of the N lives has mortality that follows the Illustrative Life
Table.
√ √
V ar(X30 ) V ar(XN )
Calculate: 30 - lim N →∞ N .
(A) 0.021 (B) 0.023 (C) 0.025 (D) 0.027 (E) 0.029

16.2 Exercises 309
16.11. Consider a fully discrete whole life insurance of 100,000 on (40).

The level annual benefit premium, calculated assuming an effective annual
interest rate of 6%, is 2086.74.
The effective annual interest rate is actually a random variable, i, with
distribution:

 5% with probability 0.25
i= 6% with probability 0.50

7% with probability 0.25
The benefit premium charged will be 2086.74, regardless of the effective

annual interest rate that is actually experienced over the duration of the
contract.
Furthermore:

 0.315488 if i = 5%
A40 = 0.269357 if i = 6%

0.233986 if i = 7%
and

 0.062603 if i = 5%
K40 +1
Var(v )= 0.062166 if i = 6%

0.060238 if i = 7%
Assume the curtate future lifetime of (40) is independent of the effective

annual interest rate.
Calculate the standard deviation of the present value of the net future loss
on the contract.
(A) 33,000 (B) 33,500 (C) 34,000 (D) 34,500 (E) 35,000

16.2 Exercises 310
16.12. For a fully discrete 3-year term insurance of 10,000 on (53):
(i) q53 = 0.015, q54 = 0.019, q55 = 0.022

(ii) The rate of interest is based on the yield curve at time t = 0. You are
also given the following information based on the yield curve at t = 0:
Start Time End Time Annual Forward Rate

0 1 0.035
1 2 0.040
2 3 0.046
Calculate the standard deviation of the present value of the death benefit
random variable of the insurance.
(A) 2100 (B) 2200 (C) 2300 (D) 2400 (E) 2500

(ii) The yield curve of effective annual spot interest rates is given by:
yt = 0.03 + 0.01t.
Calculate the benefit reserve at the end of the second year.
(A) 61,950 (B) 62,050 (C) 63,150 (D) 64,250 (E) 65,350

16.2 Exercises 311
16.14. Consider a fully discrete 2-year term insurance of 100,000 on (40):
(i) The level annual benefit premium, calculated assuming an effective

annual interest rate of 0.06, is 271.41.
(iii) The effective annual interest rate is a random variable (i):

 0.03 with probability 0.35
i= 0.06 with probability 0.50

0.09 with probability 0.15
(iv) The future lifetime of (40) is independent of the interest rate.

(v) Let V (L0 |i) = variance of the net future loss-at-issue random variable
given i. Then V (L0 |0.03) = 52,162,633 and V (L0 |0.09) = 44,322,379.
Calculate the standard deviation of the net future loss-at-issue random

variable for the policy. Use the 271.41 premium.
(A) 6000 (B) 7000 (C) 8000 (D) 9000 (E) 10,000

16.2 Exercises 312
16.15. Consider a fully discrete 5-year term insurance of 20,000 on (40).

The level annual gross premium is 300.
You are given the following profit test assumptions:
(i) There are initial expenses of 10% of the gross premium plus 100.
(ii) There are renewal expenses of 5% of the gross premium plus 20.
(iii) All expenses are paid at the beginning of the year.
(iv) Mortality is the only decrement, and follows the Illustrative Life Table.
(v) Cash flows are accumulated at an effective annual interest rate of 6%.
For the reserve basis: the reserve is set equal to the benefit reserve calcu-
lated using mortality rates (qx ) that are 110% of the Illustrative Life Table’s
mortality rates and an effective annual interest rate of 5%.
Calculate the expected profit per issued policy for the fifth year.
(A) 206 (B) 209 (C) 212 (D) 215 (E) 218
16.16. A special fully discrete 5-year endowment insurance on (50) has the
profit signature:
(−1000.00, 197.37, 288.81, 352.48, 499.64, 541.48)0 .
Calculate the internal rate of return.
(A) 0.20 (B) 0.21 (C) 0.22 (D) 0.23 (E) 0.24

16.2 Exercises 313
16.17. Consider a special 2-year term insurance on a select life aged 50:
(i) The level gross premium is 100 and is payable at the beginning of each
year.
(ii) There are two decrements: mortality (d) and lapse (w).
(iii) The death benefit is 10,000 and is payable at the end of the year of
death.
(iv) The lapse benefit is 14 of all gross premiums previously paid without
interest, and is payable at the end of the year of lapse.
(v) There is an acquisition expense of 100 that was incurred prior to issue;
this expense is paid at issue.
(vi) There are expenses of 8% of the gross premium that are paid immedi-
ately after the receipt of the premium.
(d) (d)
(vii) q[50] = 0.0010 and q[50]+1 = 0.0013
(w) (w)
(viii) q[50] = q[50]+1 = 0.04
(ix) i = 0.05
(x) The reserve at the end of year 1 is equal to 60.
(xi) Cash flows accumulate at an effective annual interest rate of 0.05.
(xii) Profits are discounted at an effective annual interest rate of 0.10.
Calculate the expected present value of the future profits.
(A) 35 (B) 40 (C) 45 (D) 50 (E) 55

16.2 Exercises 314
16.18. Consider a special insurance on Heath, aged x:
(i) The insurance is based on a 3-state model on Heath:

Healthy (0), Disabled (1), and Dead (2).
(ii) Transition probabilities include:

p00 01 10 11
x+8 = 0.85, px+8 = 0.13, px+8 = 0.22, and px+8 = 0.71.
(iii) A benefit of 8000 is payable at the end of the year of death.

(iv) A benefit of 1000 is payable at the end of the year if Heath is disabled
at that time.
(v) The reserve at the end of the eighth year is 1520 if Heath is healthy
and is 2550 if Heath is disabled.
(vi) The reserve at the end of the ninth year is 1790 if Heath is healthy
and is 2733 if Heath is disabled.
(vii) The level gross premium is 600 and is paid at the beginning of the
year if Heath is healthy.
(viii) There are expenses of 3% of the gross premium that are paid imme-
diately upon receipt of the premium.
(ix) i = 0.07
Calculate the expected profit at the end of the ninth year assuming Heath
is healthy at the start of the ninth year.
(A) 70 (B) 74 (C) 78 (D) 82 (E) 86

16.2 Exercises 315
16.19. For a universal life insurance on (40):
(i) A fixed death benefit of 150,000 is payable at the end of the year of
death.
(ii) The credited interest rate is 5% for each year.
(iii) Mortality rates (qx ) for calculating the cost of insurance follow 120%
of the mortality rates in the Illustrative Life Table.
(iv) The interest rate used to calculate the cost of insurance is 5% for each
year.
(v) Expenses are paid at the beginning of the year, and are 10% of the
premium in the first year and 6% of the premium in each renewal year.
(vi) The policyholder pays a premium of 3600 at the start of the fifth year.
(vii) The account value at the beginning of the fifth year is 19,220.
(viii) The surrender charge for the fifth year is 3000, payable at the end of
the year.
Calculate the cash value at the end of the fifth year.
(A) 20,000 (B) 20,500 (C) 21,000 (D) 21,500 (E) 22,000

16.2 Exercises 316
16.20. A profit test is conducted on the universal life insurance in Exercise

16.19. In addition to the information provided in that exercise, you are given
the following profit test assumptions:
(i) Expenses are paid at the beginning of the year; and are 8% of the
premium in the first year and 5% of the premium in each renewal year.
(iii) An effective annual interest rate of 5.5% is used to accumulate cash
flows.
(iv) The surrender rate is 8% for the first year and 4% in renewal years. It
is assumed that surrender can only occur at the end of the year.
(v) There is a claims settlement expense of 300 associated with payment
of the death benefit.
(vi) There is an additional expense of 100 associated with payment of the
cash value.
Calculate the expected profit for the fifth year per policy issued.
(A) 270 (B) 290 (C) 315 (D) 340 (E) 360

16.2 Exercises 317
16.21. Consider a universal life insurance on a select life aged 48:
(i) There is a fixed death benefit of 100,000 payable at the end of the month
of death.
(ii) The account value at the end of the 96th month is 68,500.
(iii) A premium of 800 is paid at the beginning of the 97th month.
(iv) A per policy expense of 40 is paid at the beginning of the 97th month.
(v) The credited interest rate is 0.5% for the 97th month.
(vi) The interest rate used to calculate the cost of insurance is 0.4% for
the 97th month.
(vii) Mortality rates (qx ) for calculating the cost of insurance follow 110%
of the mortality rates in the Standard Select Survival Model. It is assumed
that deaths are uniformly distributed over each year of age.
(viii) During the 97th month, there is a corridor factor of 1.46. That is, the
death benefit will actually be the greater of 100,000 and the account value
at the end of the 97th month times the corridor factor. If the corridor factor
provision would increase the death benefit, the cost of insurance calculation
should utilize the increased death benefit.
Calculate the cost of insurance for the 97th month.
(A) 5.00 (B) 5.50 (C) 6.00 (D) 6.50 (E) 7.00

16.2 Exercises 318
16.1. B
16.2. E
16.3. D
16.4. E
16.5. A
16.6. C
16.7. B
16.8. C
16.9. E
16.10. D
16.11. C
16.12. A
16.13. C
16.14. B
16.15. D
16.16. B
16.17. B
16.18. D
16.19. A
16.20. B
16.21. D


• Exam MLC, Fall 2013: #11, 20, 22, 23
• Exam MLC, Spring 2013: #8, 12, 14, 17, 23
• Exam MLC, Fall 2012: #7, 8, 9, 10, 11
• Exam MLC, Spring 2012: #10, 11, 18, 20, 23, 24, 27
297, 298, 301,

320
17 REFERENCES
• Batten, R.W. 2005. “Life Contingencies: A Logical Approach to Actu-
arial Mathematics.” Winsted, CT: ACTEX Publications, Inc.
• Bowers, N.L., Gerber, H.U., Hickman, J.C., Jones, D.A., and Nesbitt,
C.J. 1997. “Actuarial Mathematics, Second Edition.” Schaumburg, IL:
Society of Actuaries.
• Cunningham, R.J., Herzog, T.N., and London, R.L. 2011. “Models for
Quantifying Risk, 4th Edition.” Winsted, CT: ACTEX Publications,
Inc.
• Dickson, D.C.M, Hardy, M.R. and Waters, H.R. 2013. “Actuarial Math-
ematics for Life Contingent Risks, Second Edition.” Cambridge, UK:
Cambridge University Press.
• Weishaus, A. 2012. “ASM Study Manual for Exam MLC, 11th Edition.”
Westbury, NY: Actuarial Study Materials.
• Society of Actuaries: http://www.soa.org/.
• Casualty Actuarial Society: http://www.casact.org/.

321
18 APPENDIX
18.1 APPENDIX A: EXAM SYLLABI, EXAM TABLES,

AND PAST EXAM QUESTIONS
All relevant materials regarding SOA Exam MLC can be found at:
http://www.soa.org/education/exam-req/edu-exam-m-detail.aspx.
All relevant materials regarding CAS Exam LC can be found at:

http://www.casact.org/admissions/syllabus/index.cfm?fa=LCsyllabi
&parentID=332.
18.2 APPENDIX B: MULTI-STATE MODEL EXAM-

PLES (Exam MLC Only)

EXAMPLES OF MULTIPLE STATE MODELS
The Permanent Disability Model
Healthy (0) Disabled (1)
Dead (2)
The Disability Income Insurance Model
Healthy (0) Sick (1)
Dead (2)
The Multiple Decrement Model
Decrement 1 (1)
Decrement 2 (2)
Alive (0)
…………
Decrement n (n)
The Multiple Life Model
(x) Alive & (y) Alive (x) Alive & (y) Dead
[0] [1]
(x) Dead & (y) Alive (x) Dead & (y) Dead
[2] [3]

MAF Johnson ExamMLC3L StudySupplement J2014

Transféré par

Informations du document

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

MAF Johnson ExamMLC3L StudySupplement J2014

Transféré par

Droits d'auteur :

Formats disponibles

SOA EXAM MLC & CAS EXAM LC STUDY

A document prepared by the author as study materials for the

This document is not to be redistributed or reproduced in any form

This document is not in any way officially affiliated with

1 SURVIVAL MODELS AND LIFE TABLES 1

2 SELECTION (Exam MLC Only) 28

4 INSURANCE BENEFITS II (Exam MLC Only) 67

6 ANNUITIES II (Exam MLC Only) 104

7 PREMIUM CALCULATION I 123

8 PREMIUM CALCULATION II (Exam MLC Only)141

10 RESERVES II (Exam MLC Only) 180

11 MULTIPLE STATE MODELS (Exam MLC Only)208

12 MULTIPLE DECREMENTS I 227

13 MULTIPLE DECREMENTS II (Exam MLC Only)241

14 MULTIPLE LIVES I 261

15 MULTIPLE LIVES II (Exam MLC) 283

16 OTHER TOPICS (Exam MLC Only) 296

This study supplement is intended to be an aid to candidates studying for

This study supplement is meant to outline key life contingencies concepts

For candidates enrolled in the Midwestern Actuarial Forum’s Exam MLC

Candidates should bring a printed or electronic copy of this study

Despite proofreading, it is possible that this study supplement may contain

Best of luck in your exam preparation!

-Paul H. Johnson, Jr., PhD

Paul H. Johnson, Jr. earned a PhD in Actuarial Science, Risk Management

1 SURVIVAL MODELS AND LIFE TABLES

1.1 Key Concepts

• Tx = time-until-death for (x), a continuous random variable (in years).

• Sx (t) = t px = P r(Tx > t)

Paul H. Johnson, Jr.

• u+t px = (u px )(t px+u )

• Be careful: u+t qx 6= (u qx )(t qx+u ).

• u|t qx = u+t qx - u qx = u px - u+t px = (u px )(t qx+u )

• Note: 1 qx , 1 px , and u|1 qx are written as qx , px , and u| qx , respectively.

Paul H. Johnson, Jr.

• µx = µ(x) = force of mortality at age x, given survival to age x. This is

• µx+t = µx (t) = force of mortality at age x + t, given survival to x + t.

• If c > 0, then µ∗x+s = cµx+s =⇒ t p∗x = (t px )c .

• fx (t) = t px µx+t = probability density function of Tx .

Paul H. Johnson, Jr.

Curtate Future Lifetime:

• Kx = curtate future lifetime for (x), a discrete random variable.

• k| qx = P r(Kx = k) = P r(k ≤ Tx < k + 1) for k = 0, 1, 2...

Paul H. Johnson, Jr.

Other Features of Tx and Kx Distributions:

This expectation helps define the recursion: e̊x = e̊x:n + n px e̊x+n .

• The 100α-th percentile of the distribution of Tx , πα , is such that:

Paul H. Johnson, Jr.

This is the n-year temporary curtate life expectancy for (x).

This expectation helps define the recursions: ex = ex:n + n px ex+n and

Paul H. Johnson, Jr.

Special Mortality Laws:

de Moivre’s Law: Tx has a continuous uniform distribution.

Paul H. Johnson, Jr.

Modified/Generalized de Moivre’s Law: Tx has a beta distribution.

Note: α = 1 results in uniform distribution/de Moivre’s Law.

Paul H. Johnson, Jr.

Constant Force of Mortality: Tx has an exponential distribution, x ≥

Note2: Tx has an exponential distribution implies that Kx has a geometric

Paul H. Johnson, Jr.

• µx = Bcx for x ≥ 0, B > 0, c > 1

• S0 (x) = exp[− lnBc (cx − 1)]

• F0 (x) = 1 - exp[− lnBc (cx − 1)]

• t px = exp[− lnBc cx (ct − 1)]