Unit III
1. Introduction to Life Tables
Meaning
For the actuary working in the life insurance field, a major objective is to estimate the
mortality pattern which will be exhibited by a group of individuals. A basic device for
accomplishing this is known as a life table. Life table is also known as a mortality table.
A life table can be technically defined as follows:
Let 𝑙𝑥 is the number of people having age x and 𝑑𝑥 is the number of deaths of the people
between the age of x to x+1. The relationship between 𝑙𝑥 and 𝑑𝑥 can be expressed as
follows:
𝒍𝒙+𝟏 = 𝒍𝒙 − 𝒅𝒙
A life table is a tabulation of 𝑙𝑥 and 𝑑𝑥 . The following table shows an example of a life
table:
Age (x) 𝒍𝒙 𝒅𝒙
0 2,00,000 1,000
1 1,99,000 2,500
2 1,96,500 2,000
3 1,94,500 9,00
. . .
. . .
. . .
𝜔 0
The life table shall end at some age, say, 𝜔 such that 𝑙𝑥 = 0. This is obvious that 𝜔 is
the limiting age and all the people of the group will have died.
Components of Life Table/Mortality Table
Age (x)
This component shows a certain level of age x.
Number of surviving persons (𝒍𝒙 )
This component shows the number of lives surviving at age x in the beginning of a
given year. In equation form,
𝑙𝑥 = 𝑙𝑥+1 + 𝑑𝑥
Number of deaths (𝒅𝒙 )
This component shows the number of deaths during a given a year. In other words, dx
is the number deaths of those lives out of lx who could not see their x+1 age.
𝑑𝑥 = 𝑙𝑥 − 𝑙𝑥+1
1
Mortality rate (𝒒𝒙 )
It means the ratio of dx and lx. In other words, qx shows the number of people expired
out of lx. Thus,
𝑑𝑥
𝑞𝑥 =
𝑙𝑥
Since,
𝑑𝑥 = 𝑙𝑥 − 𝑙𝑥+1
Therefore,
𝑙𝑥 − 𝑙𝑥+1 𝑙𝑥+1
𝑞𝑥 = = 1−
𝑙𝑥 𝑙𝑥
Probability of survival (𝒑𝒙 )
Probability of survival means the ratio of the number of people surviving at age x+1
and x. Therefore,
𝑙𝑥+1
𝑝𝑥 =
𝑙𝑥
Infact,
𝑝𝑥 = 1 − 𝑞𝑥
Therefore,
𝑙𝑥+1
𝑝𝑥 = 1 − (1 − )
𝑙𝑥
𝑙𝑥+1
𝑝𝑥 =
𝑙𝑥
Total number of survival (Lx )
This shows the total number of survivals between age x and x+1. Symbolically, it means
𝑙𝑥 + 𝑙𝑥+1
𝐿𝑥 =
2
Since,
𝑑𝑥 = 𝑙𝑥 − 𝑙𝑥+1
⟹ 𝑙𝑥+1 = 𝑙𝑥 − 𝑑𝑥
Therefore,
𝑙𝑥 + 𝑙𝑥 − 𝑑𝑥
𝐿𝑥 =
2
2𝑙𝑥 − 𝑑𝑥
=
2
2
𝒅𝒙
𝑳 𝒙 = 𝒍𝒙 −
𝟐
Total number of life time survival (Tx)
This shows how many persons are surviving at any time who are aged x and more.
Thus,
𝑇𝑥 = 𝐿𝑥 + 𝐿𝑥+1 + 𝐿𝑥+2 + 𝐿𝑥+3 + ⋯ + 𝐿𝑤−1
𝑙𝑥 + 𝑙𝑥+1 𝑙𝑥+1 + 𝑙𝑥+2 𝑙𝑥+2 + 𝑙𝑥+3 𝑙𝑥+3 + 𝑙𝑥+4 𝑙𝑤−1 + 𝑙𝑤
𝑇𝑥 = + + + + ⋯+
2 2 2 2 2
𝑙𝑥 𝑙𝑤
= + (𝑙𝑥+1 + 𝑙𝑥+2 + ⋯ . +𝑙𝑤−1 ) +
2 2
Since, 𝑤 is limiting age and all the person would die ultimately, therefore,
𝑙𝑤 = 0
Thus,
𝑙𝑥
𝑇𝑥 = + (𝑙𝑥+1 + 𝑙𝑥+2 + ⋯ . +𝑙𝑤−1 )
2
Denoting
𝑤−1−𝑥
𝑙𝑥+1 + 𝑙𝑥+2 + ⋯ . +𝑙𝑤−1 = ∑ 𝑙𝑥+𝑖
𝑖=1
Therefore,
𝒘−𝟏−𝒙
𝒍𝒙
𝑻𝒙 = + ∑ 𝒍𝒙+𝒊
𝟐
𝒊=𝟏
Life expectancy (𝑒𝑥 )
Life expectancy means the average complete years that each person aged x will survive
after reaching age x. This is denoted by ex. Based on this definition, we can calculate a
formula of life expectancy in this way:
𝑙𝑥+1 out of 𝑙𝑥 have completed their 1 year from age x, similarly, 𝑙𝑥+2 have completed
their two years from age x, 𝑙𝑥+3 have completed their three years from age x, and so
on. Thus, total number of persons who have completed
𝑙𝑥+1 + 𝑙𝑥+2 + 𝑙𝑥+3 + ⋯ 𝑙𝑤−1 .
Thus,
𝑙𝑥+1 + 𝑙𝑥+2 + 𝑙𝑥+3 + ⋯ 𝑙𝑤−1
𝑒𝑥 =
𝑙𝑥
𝒘−𝟏−𝒙
𝟏
𝒆𝒙 = × ( ∑ 𝒍𝒙+𝒊 )
𝒍𝒙
𝒊=𝟏
3
This expression is known as curtate life expectancy, denoted by 𝒆𝒙 ,because it
considers only complete years of life and ignores fractional number of lives. When
fractional lives are also considered, then the resulting life expectation is known as
𝒐
complete life expectancy, denoted by 𝒆 𝒙 . To derive the formula of complete life
expectancy, we, first, find out total number of average deaths:
The total number of average deaths
𝑑𝑥 𝑑𝑥+1 𝑑𝑥+2 𝑑𝑤−1
= + + +⋯..+
2 2 2 2
1
= (𝑑𝑥 + 𝑑𝑥+1 + 𝑑𝑥+2 + ⋯ . . +𝑑𝑤−1 )
2
Since, total number of deaths should be equal to the number of lives at age x, therefore,
1 1
(𝑑𝑥 + 𝑑𝑥+1 + 𝑑𝑥+2 + ⋯ . . +𝑑𝑤−1 ) = 𝑙𝑥
2 2
Thus,
𝑤−1−𝑥
𝑜 1 1
𝑒𝑥 = × ( 𝑙𝑥 + ∑ 𝑙𝑥+𝑖 )
𝑙𝑥 2
𝑖=1
𝒐 𝑻𝒙
𝒆𝒙 =
𝒍𝒙
Moreover,
𝑤−1−𝑥
𝑜 1 1
𝑒𝑥 = ( × ∑ 𝑙𝑥+𝑖 ) +
𝑙𝑥 2
𝑖=1
𝒐 𝟏
𝒆 𝒙 = 𝒆𝒙 +
𝟐
Thus,
Complete life expectation = Curtate life expectancy + 0.5
2. Calculating Probabilities using Life Table
In order to calculate the probabilities using a given life table (or mortality table), we
need to understand the following actuarial notations:
.
𝑛𝑝𝑥 means the probability of a life aged x will survive n years. Symbolically,
.
𝒍𝒙+𝒏
𝒏𝒑𝒙 =
𝒍𝒙
.
𝑛𝑞𝑥 means the probability of a life aged x will not survive n years. Symbolically,
4
.
𝑛𝑞𝑥 = 1 − 𝑛.𝑝𝑥
.
𝑙𝑥+𝑛
𝑛𝑞𝑥 = 1−
𝑙𝑥
.
𝒍𝒙 − 𝒍𝒙+𝒏
𝒏𝒒 𝒙 =
𝒍𝒙
.
𝑚 | 𝑛𝑞𝑥 means the probability of a life aged x will survive m years, but die within next
n years. Symbolically,
.
𝑚 | 𝑛𝑞𝑥 = Probability of surviging 𝑚 years
× Probability of death within next 𝑛 years
.
𝑚 | 𝑛𝑞𝑥 = 𝑚.𝑝𝑥 × 𝑛.𝑞𝑥+𝑚
Since, 𝑛.𝑞𝑥 = 1 − 𝑛.𝑝𝑥 , therefore
.
𝑚 | 𝑛𝑞𝑥 = 𝑚.𝑝𝑥 × (1 − 𝑛.𝑝𝑥+𝑚 )
= 𝑚.𝑝𝑥 − 𝑚.𝑝𝑥 × 𝑛.𝑝𝑥+𝑚
Since, 𝑚.𝑝𝑥 × 𝑛.𝑝𝑥+𝑚 means the probability that a person aged x now will survive m
years and die within n years when he reaches to x+m age. This is equivalent to the
probability a person aged x now will survive m+n years. Therefore,
.
𝒎 | 𝒏𝒒𝒙 = 𝒎.𝒑𝒙 − 𝒎+𝒏.𝒑𝒙
Now, we know that
.
𝑙𝑥+𝑛
𝑛𝑝𝑥 =
𝑙𝑥
Therefore,
.
𝑙𝑥+𝑚 𝑙𝑥+𝑚+𝑛
𝑚 | 𝑛𝑞𝑥 = −
𝑙𝑥 𝑙𝑥
.
𝒍𝒙+𝒎 − 𝒍𝒙+𝒎+𝒏
𝒎 | 𝒏𝒒𝒙 =
𝒍𝒙
Special case: If we assume that m = 1, the above probabilities are denoted as follows:
.
𝒍𝒙+𝟏
𝟏𝒑𝒙 = 𝒑𝒙 =
𝒍𝒙
Example 1: From the extract of a life table given below, calculate various
probabilities in the following cases:
(a) A life aged 20 years will survive 3 years.
(b) A life aged 20 years will die within 3 years.
(c) A person aged 18 years now will die in the tenth year.
Extract of the life table:
5
Age (x) 𝒍𝒙 𝒅𝒙
18 987444 1000
19 986444 500
20 985944 1600
23 984344 100
27 984990 4990
28 980000 2000
29 878000 8000
30 870000 20000
31 850000 21000
Solution:
Part (1) : Probability of survival of a life aged 20 years to 3 years
We know that the probability of survival of a life aged x years to x+n years is given by
.
𝒍𝒙+𝒏
𝒏𝒑𝒙 =
𝒍𝒙
Therefore,
.
𝑙20+3
3𝑝20 =
𝑙20
𝑙23
=
𝑙20
.
984344
3𝑝20 =
985944
.
3𝑝20 = 0.9983
Part (2) : Probability of death of a life aged 20 years will die within 3 years.
We know that the probability of death of a life aged x years within n years is given
by
.
𝑛𝑞𝑥 = 1 − 𝑛.𝑝𝑥
Therefore,
.
3𝑞20 = 1 − 3.𝑝20
.
3𝑞20 = 1 − 0.9983 = 0.0017
Part (3): Probability of a life aged 18 years now will die in the tenth year.
We know that the probability of a life aged x years now will die in the nth year is
given by
6
.
𝒍𝒙+𝒎 − 𝒍𝒙+𝒎+𝒏
𝒎 | 𝒏𝒒𝒙 =
𝒍𝒙
Where m = the number of years of survival
Therefore,
.
𝒍𝟏𝟖+𝟗 − 𝒍𝟏𝟖+𝟗+𝟏
𝟗 | 𝟏𝒒𝟏𝟖 =
𝒍𝟏𝟖
.
𝑙27 − 𝑙28
9 | 1𝑞18 =
𝑙18
.
984990 − 980000
9 | 1𝑞18 =
987444
.
984990 − 980000
9 | 1𝑞18 =
987444
.
4990
9 | 1𝑞18 = = 0.0051
987444
3. Expected Present Value (EPV)
In actuarial science, the concept of expected present value is often used to assess the
present value of future cash flows, taking into account the probability of various
outcomes. It is a key concept in the valuation of insurance products, pension plans, and
other financial instruments with uncertain future cash flows. Infact, expected present
value is the present value adjusted with probabilities.
Thus, the expected present value is calculated by taking the sum of the present values
of all possible future cash flows, each weighted by its probability of occurrence. The
formula for expected present value (EPV) can be expressed in the following cases:
Case I : Different outcomes in a given time period i.e., t is constant here equal to b.
𝑛
𝐶𝐹𝑘
𝐸𝑃𝑉 = ∑ 𝑝𝑘 ×
(1 + 𝑖)𝑡=𝑏
𝑘=1
Where, EPV = Expected present value; 𝐶𝐹𝑘 is the cash flow associated with outcome
k; i is the discount rate; 𝑝𝑘 is the probability associated with outcome k; t is the time
of occurring of the cash flow.
Example 2: Suppose you are an actuary working for an insurance company, and you
are tasked with valuing a one-year insurance policy that pays a benefit of ₹10,000 in
the event of a covered loss. There are two possible outcomes:
(a) The loss occurs with a probability of 0.2.
(b) The loss does not occur (no claim) with a probability of 0.8.
The discount rate is 5%.
Now, let us calculate the expected present value:
7
We know that
𝑛
𝐶𝐹𝑘
𝐸𝑃𝑉 = ∑ [𝑝𝑘 × ]
(1 + 𝑖)𝑡
𝑘=1
Where, 𝑝1 = 0.2 ; 𝑝2 = 0.8; 𝐶𝐹1 = ₹ 10,000 (insurance claim if the loss occurs);
𝐶𝐹2 = 0 (no claim); i = 5% = 0.05
𝐶𝐹1 𝐶𝐹2
= 𝑝1 × + 𝑝2 ×
(1 + 𝑖)𝑡 (1 + 𝑖)𝑡
10,000 0
= 0.2 × 1
+ 0.8 ×
(1 + 0.05) (1 + 0.05)1
10,000
= 0.2 × +0
1.05
𝐸𝑃𝑉 = ₹ 1904.76
Case II : Different outcomes in different time periods i.e., t is variable here.
𝑛
𝐶𝐹𝑘
𝐸𝑃𝑉 = ∑ 𝑝𝑘 ×
(1 + 𝑖)𝑡
𝑘=1
This formula essentially accounts for the fact that future cash flows are worth less in
present terms due to the time value of money. The discount rate (i) reflects the
opportunity cost of tying up money in the present rather than investing it, and it is used
to convert future cash flows into their present value. Actuaries use expected present
value calculations in various contexts, such as pricing insurance policies, valuing
pension liabilities, and assessing the financial soundness of insurance companies. The
accuracy of these calculations depends on the quality of the data and assumptions used,
including probabilities of different events and the chosen discount rate.
Example 3: Let us consider a scenario with three possible outcomes. Suppose you are
an actuary working on the valuation of a financial product with the following cash flow
scenarios and probabilities:
(a) Probability of 0.3: Cash flow of ₹ 5,000 in the first year.
(b) Probability of 0.5: Cash flow of ₹ 8,000 in the second year.
(c) Probability of 0.2: No cash flow in the third year.
The discount rate is 6%.
Now, let us calculate the expected present value using the formula:
We know that
𝑛
𝐶𝐹𝑘
𝐸𝑃𝑉 = ∑ [𝑝𝑘 × ]
(1 + 𝑖)𝑡
𝑘=1
8
Where, 𝑝1 = 0.3 ; 𝑝2 = 0.5; 𝑝3 = 0.2 𝐶𝐹1 =
₹ 5,000 (insurance claim if the loss occurs); 𝐶𝐹2 = ₹ 8,000 ; ); 𝐶𝐹3 = ₹ 0 ; i = 6%
= 0.05
𝐶𝐹1 𝐶𝐹2 𝐶𝐹3
= 𝑝1 × + 𝑝2 × + 𝑝3 ×
(1 + 𝑖)𝑡 (1 + 𝑖)𝑡 (1 + 𝑖)𝑡
5,000 8000 0
= 0.3 × 1
+ 0.5 × 2
+ 0.2 ×
(1 + 0.06) (1 + 0.06) (1 + 0.06)3
1500 4000
= + +0
1.06 (1.06)2
= 1415.09 + 3559.98 + 0
𝐸𝑃𝑉 = ₹ 4975.07
Case III : EPV of insurance claim in case of term insurance policy of ₹ A.
A term insurance policy is a type of insurance policy in which the claim (₹ A) is due
when the death of the policy holder occurs before achieving a certain level of age. Thus,
the expected present value of the insurance amount is
𝑛
𝐶𝐹𝑘
𝐸𝑃𝑉 = ∑ 𝑝𝑘 ×
(1 + 𝑖)𝑡
𝑘=1
Where , 𝑝𝑘 = probability of 𝒅𝒆𝒂𝒕𝒉 ; 𝐶𝐹𝑘 = Insurance claim = 𝑑𝑥 𝐴; 𝑑𝑥 =
Number of deaths.
Case IV : EPV of insurance claim in case of pure endowment policy
A pure endowment policy a type of insurance policy in which the claim (₹A) is due
when the policy holder survives a certain level of age. Thus, the expected present value
of the insurance amount is
𝐶𝐹𝑘
𝐸𝑃𝑉 = 𝑝𝑘 ×
(1 + 𝑖)𝑘
Where , 𝑝𝑘 = probability of 𝒔𝒖𝒓𝒗𝒊𝒗𝒂𝒍 ; 𝐶𝐹𝑘 = Insurance claim = 𝑙𝑥 A; 𝑘 =
the time when the required age is achieved.
9
Exercise
Exercise 1: From the following data, calculate the expected value of the term insurance
of ₹ 12,000. Assume that the insurance claim is paid at the end of the year in which
death occurs and discounting rate is 5%. Additionally, the insurance claim is taken at
age 20 and its term is 6 years.
Age (x) qx lx
19 0.02 5,00,000
20 0.04 -
21 0.05 -
22 0.12 -
23 0.30 -
24 0.15 -
25 0.10 -
26 0.02 -
Exercise 2: From the following data, calculate the expected value of the pure
endowment insurance of ₹ 5,000. Assume that the insurance claim is paid at the end of
the year in which the policy holder survives the age of 25. Additionally, the insurance
claim is taken at age 21.
Age (x) qx lx
20 0.04 5,00,000
21 0.05 -
22 0.12 -
23 0.30 -
24 0.15 -
25 0.10 -
Exercise 3: Complete the following mortality table or life table.
𝒐
Age(x) dx qx px lx Lx Tx 𝒆𝒙 𝒆𝒙
0 274 0.0002740 10,00,000
1 240 0.0002401
2 153 0.0001531
3 250 0.0002502
4 272 0.0002722
5 148 0.0001482
6 193 0.0001933
7 282 0.0002824
8 193 0.0001934
9 106 0.0001062
10 120 0.0001203
10
Exercise 4: From the following information, model the expected value of the insurance
of ₹ 𝛽 such that interest rate being 𝜃. Assume that the insurance is taken at the age of x
and the insurance claim will be paid if he dies before age x+n.
Age (x) lx qx
x lx ∆0
x+1 lx+1 ∆1
x+2 lx+2 ∆2
x+3 lx+3 ∆3
x+4 lx+4 ∆4
- - -
x+(n-1) lx+(n-1) ∆𝑛−1
x+n lx+n ∆𝑛
Exercise 5: If the probability of death is given by
𝑓(𝑡) = 0.0004𝑡 2 such that 0 ≤ 𝑡 ≤ 50
Where t = age
a) Find the probability of survival at age 45.
b) What is the probability of death at age 39?
c) If l25 = 4,50,000, then what is d26 ?
d) If l2 = 6,30,563, then what is the value of radix?
e) What is limiting age?
Exercise 6: The following equation is called Gompertz’s law of mortality.
𝑢𝑥 = 𝐵𝑐 𝑥
Where, B and C are some constants such that 0 < B <1 and c > 1. On the basis of this
law, we have the following survival function:
𝑥+𝑡
𝐵𝑐 𝑟 𝑑𝑟
𝑆𝑥 (𝑡) = 𝑒 − ∫𝑥
−𝐵
{ ×𝑐 𝑥 (𝑐 𝑡 −1)}
= 𝑒 𝑙𝑜𝑔 𝑐
Moreover, the probability density function 𝑓𝑥 (𝑡) is defined as follows:
𝑓𝑥 (𝑡)
𝑢𝑥+𝑡 =
𝑆𝑥 (𝑡)
You are required to find out the probability density function if x = 21 after considering
Gompertz’s law of mortality.
Exercise 7: The utility function of Barin is 𝑈(𝑦) = √𝑦 where y shows the level of
wealth. Barin is endowed with initial wealth of ₹ 10000. Further, there is 20%
probability that 36% of the wealth will be stolen by a thief; therefore, Barin is thinking
to take an insurance with premium of Rs. 500. Should he take insurance? Moreover,
what is the expected value of wealth if insurance is taken?
11
Exercise 8: The utility function of Barin is 𝑈(𝑦) = √𝑦 where y shows the level of
wealth. Barin is endowed with initial wealth of ₹ 10000. Further, there is 20%
probability that 36% of the wealth will be stolen by a thief; therefore, Barin takes an
insurance. What should be the amount of single premium?
Exercise 9: What do you mean a life table and what are its various components?
Exercise 10: What is life expectancy? Use mathematical approach.
12
