Level I - Quantitative Methods

Common Probability Distributions

www.ift.world

Graphs, charts, tables, examples, and figures are copyright 2022, CFA Institute.
Reproduced and republished with permission from CFA Institute. All rights reserved.

1
 Contents
1. Introduction and Discrete Random Variables
2. Discrete and Continuous Uniform Distribution
3. The Binomial Distribution
4. Normal Distribution
5. Applications of the Normal Distribution
6. Lognormal Distribution and Continuous Compounding
7. Student’s t-, Chi-Square, And F-Distributions
8. Monte Carlo Simulation

www.ift.world 2
 1. Introduction and Discrete Random Variables
• Investment decisions require work with random variables

• Important to understand the probability distribution of random variables

• We will focus on seven probability distributions

 Uniform
 Normal
 Binomial
 Lognormal
 Student’s t
 Chi-square
 F-distribution

www.ift.world 3
 1.1 Discrete Random Variables
• A discrete random variable is one where we can list all the
possible outcomes

• A continuous random variable is where the number of points

between the lower and upper bounds are essentially infinite

• Probability distribution specifies the probabilities of all the

possible outcomes for a random variable
 Probability function specifies probability that random variable takes on a
specific value
 Probability density function is used for continuous random variables

www.ift.world 4
 2. The Discrete Uniform Distribution
X=x Probability Function Cumulative Evaluate the following:
p(x) = P(X = x) Distribution Function
F(x) = P(X ≤ x) P(3 ≤ X < 6)
1 1/6 1/6
F(1)
2 1/6 2/6
3 1/6 3/6 F(4)
4 1/6 4/6
F(6)
5 1/6 5/6
6 1/6 6/6 F(100)

www.ift.world 5
 2.1 Continuous Uniform Distribution
The continuous uniform distribution is defined
over a range that spans between some lower
limit "a" and some upper limit "b”

©Wikipedia.org

www.ift.world 6
 Example
Random variable X follows a continuous uniform distribution between 10 and 20.
What is the probability of an outcome between 15 and 25?

Answer: 0.5

www.ift.world 7
Continuous Uniform Cumulative Distribution

www.ift.world 8
 Practice Question
The execution time for a single order is uniformly distributed between 10 and 18
minutes. If an order is placed at 2:30 PM, what is the probability that order will
execute after 2:45 PM?

Answer: 3/8 or 0.375

www.ift.world 9
 3. Binomial Distribution
Bernoulli trial is the building block of the binomial
distribution

Bernoulli random variable

P(1) = P(Y = 1) = p
P(0) = P(Y = 0) = 1 – p

Binomial random variable, X, is the number of

successes (Y=1) in a given number (n) of Bernoulli
trials

www.ift.world 10
 Probability Distribution of a Binomial Random Variable

The probability distribution of a binomial random

variable for the probability of "x" success in "n"
trials is calculated using the following formula:

p(x) = P(X = x) = nCx px (1 - p)n – x

Where p = the probability of "success" on each trial

www.ift.world 11
 Example
What is the probability that a stock will move up on 7 out of the next 10
days assuming that the probability of an up-move is 0.6?

Answers: 0.21

www.ift.world 12
Mean and Variance of a Binomial Random Variable
Random Variable Mean Variance

Bernoulli, B(1,p) p p(1-p)

Binomial, B(n,p) np np(1-p)

www.ift.world 13
 Example
Consider a 20-bond portfolio where each bond is from a different issuer. Each
issuer has an estimated annual default rate of 10 percent. What is the expected
number of defaults in the portfolio assuming a binomial model for defaults?
Estimate the variance of the number of defaults over the coming year?

Answers: 2, 1.8

www.ift.world 14
 Practice Problem
You are estimating the earnings of a company
over the next ten years. Based on past data the
probability that earnings will increase in a given
year relative to the prior year is 0.8.

1) What is the probability that earnings will

increase in exactly 5 of the next 10 years?

2) What is the expected number of yearly

earnings increases?

3) What is the standard deviation of the

number of yearly earnings increases over
then next 10 years?
Answers: 0.026, 8, sqrt(1.6) = 1.26
www.ift.world 15
 4. The Normal Distribution
The normal distribution is extensively used in quantitative work

X ~ N(µ, s2)

Skewness = 0

Kurtosis = 3

www.ift.world 16
 Units of Standard Deviation

Approximately 68% of all observations fall in the interval µ ± σ

Approximately 95% of all observations fall in the interval µ ± 2σ

Approximately 99% of all observations fall in the interval µ ± 3σ

www.ift.world 17
 Confidence Interval

The 90% confidence interval for X is µ +/- 1.65s

The 95% confidence interval for X is µ +/- 1.96s

The 99% confidence interval for X is µ +/- 2.58s

www.ift.world 18
 Standard Normal Distribution
Mean = 0 and Standard Deviation = 1

z-value

Cumulative probabilities for a

standard normal distribution

www.ift.world 19
 Transformation to Standard Normal Distribution

Why transform?

Example: The return on a stock is normally

distributed with μ = 10 and σ = 2. What is the
probability that the return will be less than 11%?

Transformation formula: Z = (X - µ) / σ

www.ift.world 20
 Practice Problem
Assume that the annual earnings per share (EPS) for a large sample of firms are
normally distributed with a mean of $5.00 and a standard deviation of $2.00. What is
the approximate probability of an observed EPS value falling between $4.00 and $5.00?

Ans: 0.19

www.ift.world 21
 Multivariate Distribution
Multivariate distribution specifies probabilities for
a group of related random variables
Using asset returns as our random variables, the
multivariate normal distribution for the returns on
n assets can be completely defined by the following
three sets of parameters:
Means of the n series of returns (µ1, µ2, …, µn)
Variances of the n series of returns (s21, s 22, …, s2n)
n(n – 1)/2 pair-wise correlations

www.ift.world 22
 5. Applications of the Normal Distribution

• Mean Variance Analysis

 Main point: consider both mean and variance when making
investment decisions

• Safety First Ratio

 Next slide

www.ift.world 23
 Safety First Rule
Roy's safety-first criterion states that the optimal portfolio
minimizes the probability that the return of the portfolio falls
below some minimum acceptable level. This minimum
acceptable level is called the "threshold" level.
To select the optimal portfolio:

1. Calculate SFRatio using: [E(Rp) - RL] / sp

2. Select portfolio with the highest SFRatio

www.ift.world 24
 Example
For a given investor the minimum acceptable return is 2%. What is the Safety First
Ratio (SF Ratio) for a portfolio with an expected return on 7% and standard
deviation of 10%? Is this preferable over a portfolio with an expected return of 5%
and a standard deviation of 6%?

www.ift.world 25
 Practice Question
An investor has a portfolio of $100,000. He will need $2,000 for his medical insurance and another
$2,000 for his rent expenses by the end of the year. The investor is considering investing in one of these
three available portfolios:

Portfolio Expected Return Standard Deviation

A 5% 10%
B 8% 13%
C 10% 17%

Using Roy’s safety-first criterion ratio, which one of these portfolios will minimize the probability of
investor’s portfolio falling below $100,000?
A. Portfolio A
B. Portfolio B
C. Portfolio C

www.ift.world 26
 Solution
Answer: C

The investor requires a minimum return of $4,000/$100,000 or 4 percent. Roy’s

safety-first model uses the excess of each portfolio’s expected return over the
minimum return and divides that excess by the standard deviation for that portfolio.
The highest safety-first ratio is associated with Portfolio C: (10% – 4%)/17% = 0.35

www.ift.world 27
 6.1 The Lognormal Distribution
The lognormal distribution is generated by the function eX, where X is normally distributed.

The properties of a
lognormal distribution:

The lognormal distribution is completely defined by • Cannot be negative

the mean and variance of the normal distribution.
• Upper end of its range
Remember the phrase: Log is normal extends to infinity

Suitable for modeling asset prices • Positively skewed

www.ift.world 28
 6.2 Continuously Compounded Rates of Return
Continuously compounded return is normally distributed  the future price will be lognormally distributed

Given the holding period return over any time period, calculate the equivalent continuously compounded
rate of return for that period as:

r = ln (HPR +1)

For continuous compounding, the EAR is given by: EAR = er – 1

Example: If the holding period return of a stock was 10% for a period of one year. What is the equivalent
continuously compounded rate of return for the year?

Solution: r = ln (0.1 +1) = 0.0953 = 9.53%

www.ift.world 29
 7.1 Student’s t-Distribution
Properties of a Student’s t-distribution:
• Symmetrical, bell-shaped and similar to a normal distribution
• Lower peak and fatter tails as compared to a normal distribution
• Defined by a single parameter, degrees of freedom (df) = n – 1
• As the df increase, the t-distribution approaches the standard normal distribution

www.ift.world 30
 7.2 Chi-Square and F-Distribution (1/2)
Properties of the chi-square distribution:
• It is asymmetrical and is defined by a single parameter, degrees of freedom (k) = n – 1
• With k degrees of freedom, the distribution is the sum of the squares of k independent standard
normally distributed random variables (distribution does not take on negative values)
• A different distribution exists for each value of k
• As k increases the shape of the distribution becomes more similar to a bell curve

www.ift.world 31
 7.2 Chi-Square and F-Distribution (2/2)
Properties of the F-distribution:
• Asymmetrical and defined by two parameters, degrees of freedom of the numerator (df1) and degrees of
freedom of the denominator (df2).
• The distribution does not take on negative values.
• As both the numerator (df1) and the denominator (df2) degrees of freedom increase, the distribution
becomes more similar to a bell curve.

The relationship between the chi-square and F-

distributions is as follows: If χ12 is one chi-square
random variable with m degrees of freedom and χ22
is another chi-square random variable with n
degrees of freedom, then F = (χ12 /m)/(χ22 /n)
follows an F-distribution with m numerator and n
denominator degrees of freedom.

www.ift.world 32
Student’s t, Chi-Square, and F-Distributions: Basis for Hypothesis
Tests of Investment Returns

Distribution Test Statistic Hypothesis Tests of Returns

Student’s t t-Statistic Tests of a single population mean, of differences between two population
means, of mean difference between paired (dependent) populations, and
of population correlation coefficient

Chi-square Chi-square statistic Test of variance of a normally distributed population

F F-statistic Test of equality of variances of two normally distributed populations from

two independent random samples

www.ift.world 33
 8. Monte Carlo Simulation
What is it? What are the major uses? What are the strengths and weaknesses?

What is it?
Use computer software to simulate a complex financial system
Generate a large number of random samples from specified
probability distributions to represent risk in the system

What are the major uses? Strength:

Valuing complex securities for which no analytic Can be used to price complex securities for which no
pricing formula is available analytic expression is available
Estimate risk and return in investment applications Weaknesses:
Sensitivity analysis Provides only statistical estimates, not exact results
Compliment analytical methods Does not provide insight into cause-and-effect
relationships

www.ift.world 34
Use Monte Carlo Simulation to Value a Contingent Claim Security

www.ift.world 35
 Monte Carlo Simulation: The Steps
Step 1: Specify the quantity of interest

Step 2: Specify a time grid: K sub-periods with Δt increment for the full-time
horizon.

Step 3: Specify distributional assumptions for the key risk factors

Step 4: Draw standard normal random numbers for each key risk factor over
each of the K sub-periods.

Step 5: Convert the standard normal random numbers to stock prices, average
stock price, and other relevant risk factors.

Step 6: Calculate the value and the present value of the contingent claim payoff.

Step 7: Repeat Steps 4-6 over the specified number of trials. Then, calculate
summary value

www.ift.world 36
 Summary
• Introduction and Discrete Random Variables

• Discrete and Continuous Uniform Distribution

• The Binomial Distribution

• Normal Distribution

• Applications of the Normal Distribution

• Lognormal Distribution and Continuous Compounding

• Student’s t-, Chi-Square, And F-Distributions

• Monte Carlo Simulation

www.ift.world 37