0% found this document useful (0 votes)
11 views19 pages

Random Variables in Risk Theory

This document discusses different types of random variables including discrete, continuous, and mixed random variables. It defines key concepts like probability mass functions, cumulative distribution functions, moments, and probability generating/moment generating functions. Several examples are provided to illustrate how to apply these concepts to specific random variables.

Uploaded by

2664896266
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
11 views19 pages

Random Variables in Risk Theory

This document discusses different types of random variables including discrete, continuous, and mixed random variables. It defines key concepts like probability mass functions, cumulative distribution functions, moments, and probability generating/moment generating functions. Several examples are provided to illustrate how to apply these concepts to specific random variables.

Uploaded by

2664896266
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

MTH223 Mathematical Risk Theory

Chapter 2: Random Variables

Xi’an Jiaotong-Liverpool University

Dr Jiajun Liu

Dr Jiajun Liu (XTJLU) MTH 223 1 / 13


Overview

In this chapter, we review some basic probability concepts of random


variable. These concepts are central to the characterization of any
mathematical model considered to predict insurance cost of an insurance
portfolio.

In this course, three types of random variables are encountered, namely


discrete, continuous and mixed.

Dr Jiajun Liu (XTJLU) MTH 223 2 / 13


Discrete random variables
We say that X is a discrete random variable if it only takes a countable
(including finite) number of real values {xi }i∈N .
• The probability mass function (pmf) of X is

p(xi ) = P(X = xi ), i ∈ N.
• The cumulative distribution function (cdf) of X is
X
F (x) = P(X ≤ x) = p(xi ).
xi ≤x

• The kth moment of X is


X
E[X k ] = (xi )k p(xi ).
i∈N

• Example of discrete distributions include the Poisson, binomial, and


negative binomial distributions.
Dr Jiajun Liu (XTJLU) MTH 223 3 / 13
Probability generating function
Suppose that X only takes nonnegative integers values.
• The probability generating function (pgf) of X is
h i
G (z) = E z X

X
= z k p(k)
k=0
= p(0) + zp(1) + z 2 p(2) + z 3 p(3) + · · ·,
where p(·) is the pmf of X .
• Note that
dk 1 (k)
p(k) = G (z) = G (0).
dz k z=0 k!
Thus pgf uniquely determines the distribution of a discrete random
variable.
Dr Jiajun Liu (XTJLU) MTH 223 4 / 13
Exemple: discrete random variable

Example 1
Consider a discrete random variable X with pmf

 0.5, x = 0,
p(x) = P(X = x) = 0.4, x = 1, (1)
0.1, x = 2.

Find its cdf and pgf.

Dr Jiajun Liu (XTJLU) MTH 223 5 / 13


Continuous random variable
We say that X is a continuous random variable, if there exists a
nonnegative function f called probability density function (pdf) such
that its cdf is
Z x
F (x) = f (y )dy .
−∞
0
Thus, we have F (x) = f (x).
• The kth moment of X is
Z ∞
k
E[X ] = x k f (x)dx, (2)
−∞

provided the above integral is finite.

• Examples of continuous distributions include the uniform,


exponential, gamma, Weibull, and normal distributions.
Dr Jiajun Liu (XTJLU) MTH 223 6 / 13
Moment generating function

For a continuous random variable X , its moment generating function


(mgf) is
h i Z ∞
M(t) = E etX = etx f (x)dx, t ∈ R, (3)
−∞
provided the above integral is finite.
Further,

dk
M (k) (0) = M(t) = E[X k ], k ∈ N.
dt k t=0

Dr Jiajun Liu (XTJLU) MTH 223 7 / 13


Example: continuous random variable

Example 2
Consider an exponential random variableX with pdf

f (x) = 0.1e−0.1x , x > 0.

Find its cdf and mgf.

Dr Jiajun Liu (XTJLU) MTH 223 8 / 13


Hazard rate function
• For a continuous random variable, sometimes it is convenient to use
the so-called hazard rate function (or failure rate) defined as
f (x) d
h(x) = = − ln F (x),
F (x) dx
where F (x) = 1 − F (x) is the survival function.
• Intuitively, for small ∆x > 0,
F (x + ∆x) − F (x) P(x < X ≤ x + ∆x)
h(x) = =
∆xF (x) ∆xP(X > x)
P(X ≤ x + ∆x|X > x)
=
∆x
This concept is particularly relevant to assess the heaviness of the
right tail of the loss distribution, i.e., how likely the random variable
takes large positive values.
Dr Jiajun Liu (XTJLU) MTH 223 9 / 13
Example: hazard rate function

Example 3
Suppose X ∼ Wei(θ, τ ) with pdf
τ
τ (x/θ)τ e−(x/θ)
f (x) = , x > 0,
x
where θ, τ > 0. Find its hazard rate function.

Dr Jiajun Liu (XTJLU) MTH 223 10 / 13


Mixed random variable
We call X a mixed random variable if it has both discrete and
continuous components. Such random variables play an important role in
modeling insurance claims, for example, the loss amount is usually a
continuous random variable with a probability mass at 0.

Let {x1 , x2 , · · ·} be a sequence of real numbers in an increasing order.


Suppose X has a density function f on each interval (xi , xi+1 ), and has
discrete mass points at the boundaries of these intervals, i.e.,

P(X = xi ) = p(xi ) > 0, i ∈ N.

We require that
X XZ xi+1
p(xi ) + f (x)dx = 1.
i∈N i∈N xi

Dr Jiajun Liu (XTJLU) MTH 223 11 / 13


Mixed random variable (cont’)

The cdf of a mixed random variable X is


X XZ xi+1
F (x) = P(X ≤ x) = p(xi )1{xi ≤x} + f (y )1{y ≤x} dy .
i∈N i∈N xi

The kth moment of X is


X XZ xi+1
k k
E[X ] = (xi ) p(xi ) + x k f (x)dx.
i∈N i∈N xi

The mgf of X is
h i X XZ xi+1
tX txi
M(t) = E e = e p(xi ) + etx f (x)dx.
i∈N i∈N xi

Dr Jiajun Liu (XTJLU) MTH 223 12 / 13


Example: mixed random variable

Example 4
Assume a claim amount of an insurance policy is modeled by a nonnegative
random variable X which has a probability mass of p at 0 and otherwise
continuous with a pdf f over (0, ∞). Find its cdf, kth moment, and mgf.

Dr Jiajun Liu (XTJLU) MTH 223 13 / 13

You might also like