FIN 6307 Mathematical Methods for Finance

Lecture 2:
Probability Theory

1
 Reading / Excel Functions / Web Links
• Readings:
―Ch. 2
Probability Theory

Web Links:
http://finance.yahoo.com/
http://data.bls.gov/cgi-bin/surveymost?cu
http://research.stlouisfed.org/fred2/
http://morningstar.com/
2
http://www.investopedia.com/markets/
 Outline
• Introduction
• Random Variables
• Laws of Probability
• Probability Density Function (PDF) &
Cumulative Distribution Function (CDF)
• Homework Assignment

3
 Introduction
• Descriptive Statistics: Describe important aspects of a
data set or a random variable. A group of summary
measures which each summarize, in one number a
certain quality of a group of data, for e.g.
– Averages, spread, symmetry, skew, peaking, etc.
– Measures of location or central tendency, measures of
dispersion, measures of skewness, kurtosis, etc.
• Statistical Inference: Make forecasts, estimates, or
judgements about the population from a small sample
(Will be discussed in later chapters).

4
 Population vs Sample
• Populations and samples

•Population: total set of members in a specific group

•Population statistics: Statistics of the population
•Parameter: a numerical summary of the population

•Sample: a subset of a population

•Sample Statistics: Statistics of the sample
•Sample statistic (or statistic): a numerical summary of
sample data

5
 Probability

• Probability:
– A measure of the possibility of an event happening
– Measured between a scale of 0 (certainly will not
happen) and 1 (will certainly happen)
– Used in finance because outcomes of all financial
decisions are uncertain

6
 Probability Distributions
• Probability Distributions:
– Are mathematical models of possibilities (probabilities) of uncertain
events happening
– Akin to frequency distributions discussed in next sessions
– Describe the way the probabilities associated with a random variable
are distributed across all the possible values of that variable
– Used extensively in financial decision-making

7
 Quant Models in Finance
• Probabilistic and Statistical Models in Finance
– used to analyze the evolution of financial asset prices or interest rates
– these are examples of random variables (RV)– variables whose values
could be observed in the present and in the past, but are unknown in the
future
– uncertainty –since values in the future are unknown
– Continuous & Discrete RV – for a continuous RV the probability that it
takes on a given value is always 0; can only make a probabilistic statement
about a range of values
– Basic laws of probability

8
 Basic Concepts
• Classical Vs. Empirical Vs. Bayesian Approaches

1. The classical, or a priori approach to probability

 Pioneered by Irving Fisher (1867-1947), identifies a probability associated

with an outcome based upon the range of possible outcomes – for e.g. tossing
a coin –

 Tossing of the coin is the experiment, the sample space refers to the 2
possible outcomes (H or T), and the event is whether the outcome is a head or
a tail

# equally likely outcomes associated with the event

P ( A) 
Total # of equally likely outcomes

9
 Basic Concepts
• Classical Vs. Empirical Vs. Bayesian Approaches

2. The empirical approach to probability

 Very often cannot depend on the exactness of the process to determine

probabilities

 This approach is called the relative frequency or empirical approach and is

based on objective information culled from experimental or empirical
observations - works well in science and although not best suited, is used in
finance

# of X occurences
P( X ) 
# of experiments

10
 Basic Concepts
• Classical Vs. Empirical Vs. Bayesian Approaches

3. Bayesian or Subjective approach to probability

 Pioneered by Thomas Bayes (1702-1761), allows for subjective assessments

of probabilities to be made based on prior beliefs – better suited for finance,
as it allows for departure from dependence entirely on historical data, but this
approach is not used

 Probability in this approach is simply defined as the strength of belief that

an event will occur: for e.g. used in forecasting company profits

Classical approach a special case of the Bayesian approach, and

is used more extensively in finance

11
 Basic Concepts
Laws of Probability
– W.r.t. abstract events, such as: 25 % < R<75%, P< $10, Volatility > 20 %
– Law-1: The probability of an event, A, is a number between 0 and 1: P( A)  0,1
– Law-2: For 2 mutually exclusive events, A & B: P ( A or B)  P( A)  P( B)
The event that A does not occur, is called the complement of A and is
denoted by A . (Sometimes also denoted by AC );
Hence: P (A)  P(A)=1 , where A and A are exhaustive
― Law-3: the joint probability of two events, A & B, occurring is:
P(A and B)  P(A B)P(B)

– P (A B) is the conditional probability of the event A occurring given that B

occurs, and is 0 if A & B are mutually exclusive.

– This is different from P(A), the stand-alone probability of A without any

information about, B or the marginal probability of A

Ref: Quantitative Methods in Finance, Carol Alexander- Vol-1 12

 Basic Concepts
– Law-2 (in general): If A & B are any 2 events, then the probability that either A
occurs or B occurs is the sum of their probabilities less the probability that they
both occur: P(A or B) = P(A) + P(B) - P(A and B)
– This is just a generalization of Law-2 to include the possibility that A& B are
not mutually exclusive, i.e.
• We need to subtract P( AandB ) because it is counted twice

– P(A and B) Can be illustrated using the Venn Diagram used in set theory

P(A and B)

A B

Ref: Quantitative Methods in Finance, Carol Alexander- Vol-1 13

 Basic Concepts
– Law-3 implies the following –

P(A and B)
P(A B) =
(a) P(B)

(b) P(A and B) = P(B A)P(A)

P(B A)P(A)
P(A B) =
(c) P(B)

– The left hand side of (c) is referred to as the posterior probability of A, i.e.
the probability of A given that we have some information about B;
– P(A) is the marginal probability of A, also called the prior probability,
before we have information on B.
– The two events A & B are independent, if and only if, P(A B)=P(A) , or
P(B A)=P(B)

– If two events A & B are independent, then the probability that

P(A andthey
B) =both
P(A)P(B)
occur is just the product of their marginal probabilities:
Ref: Quantitative Methods in Finance, Carol Alexander- Vol-1 14
 Basic Concepts
– Law-4 (total probability)
C
If there are two states S and S , the total probability of an
event A is the weighted average of conditional probabilities of
A given each state.

P( A)  P ( A | S ) * P( S )  P ( A | S c ) * P( S c )

Ref: Quantitative Methods in Finance, Carol Alexander- Vol-1 15

Mutually Exclusive Events

P[ AB ]  P[ A]  P[ B]
 Independent Events

P[rain and market up]  P[rain  market up]  P[rain]  P[market up]
 Probability Matrices

Stock

Outperform Underperform

Upgrade 15% 5% 20%

Bonds
No Change 30% 25% 55%

Downgrade 5% 20% 25%

50% 50% 100%
 Random Variables
• Random Variables

– A Random Variable (RV) is a real variable whose values are

stochastic i.e. there is uncertainty about the values that the
variable takes – a RV is distinct from a deterministic variable
whose values are completely determined by the information we
presently have

– A realization or observation on a RV X is a number that is

associated with a chance outcome – since every outcome is
determined by a chance event, every outcome has a measure of
probability associated with it – the set of all outcomes and their
associated probabilities is called a probability measure.

19
Random Variables, Density & Distribution
Functions
Definition: Let X be a random variable. The distribution function of X is defined as
F(x)=P(X<=x)
A random variable is completely characterized by its distribution function.

Discrete random variable (X): a random variable which can take on a finite or
countable number of possible values. For the distribution function of a discrete random
variable there is a formula:
F ( x)  p
xi  x
i ; xi is a RV outcome, its probability pi=P(X=xi), ∑i pi=1

Continuous random variable: a random variable is called absolutely continuous when

its distribution function can be represented as:
x
F ( x)   f (t )dt

f(t) is called a density of the distribution

20
 Discrete Random Variables

Variables can take on only a countable number

of values.
P[ X  xi ]  pi s.t. xi  {x1 , x2 ,..., xn }
n

p
i i
i 1
 Continuous Random Variables

 Variable can take on any value within a given range.

P[r1  X  r2 ]  p

• PDF:
r2

 f ( x)dx  p
r1
rmax

 f ( x)dx  1
rmin
 Density and Distribution Functions
• Density and Distribution Functions
– The way to represent a probability measure on a RV is via a probability
distribution function – a function that gives the probability that RV X takes
a value  x ; this function is also referred to as cumulative distribution
function
– Another way to represent a probability measure is via a probability density
function
– RVs can be discrete or continuous – most used in finance are continuous, real-
valued i.e. b Since P(-  < X < ) = 1 , the total area
P(a  X  b)   f ( x)dx
a
under the curve of a continuous density
function must be 1; also, since there is no area under the density function
defined by the inequality a  X  a every continuous RV has the property
that P(X=a) =0 for any real a.
– Hence,
P( X  a )  P( X  a ), P( X  a )  P( X  a ), P(a  X  b)  P(a  X  b)
Ref: Quantitative Methods in Finance, Carol Alexander- Vol-1 23
 Density & Distribution Functions
– Given a Probability Density Function f ( x ) , the corresponding Cumulative
Distribution Function is
x
F ( x )  P( X  x )   f ( x )dx


- i.e. the distribution function is the area under the density curve to the left of x.
– Hence also, P(a  X  b)  F (b)  F (a )
– Not really necessary to specify both the density function and distribution function
– For discrete RV, the distribution is the sum of the density, and the density is the
difference of the distribution
– For continuous RV, the distribution function is the integral of the density function and
the density function is the 1st derivative of the distribution function w.r.t. x –
– Hence, f ( x )  F ( x)

– For e.g.  0.1 

P(X  0.1)    P(X  x), for discrete X 
 x  
 0.1 
and =   f ( x)dx, for continuous X 
  
24
 MiniQuiz #1
• Is the following a valid PDF? What is the
probability that X is greater than 2/6?
f(x
4 )
2

x
1/ 2/ 3/
6 6 6
 MiniQuiz #1 (cont.)
• This is a legitimate PDF. The area under the
“curve” is equal to 1.
• P[x>2/6] = 4/6 = 67%

f(x
4 )
2

x
1/ 2/ 3/
6 6 6
 Homework 2
Miller’s Chapter 2 (page 26): problem 1,2,3,8,9,10
CFA Workbook Chapter 4: problem 5, 7, 10, 15, 19

27