FRM Part 1

Book 2 – Quantitative Analysis

FUNDAMENTALS OF PROBABILITY
 Learning Objectives
After completing this reading you should be able
to:
 Describe an event and an event space.
 Describe independent events and mutually exclusive events.
 Explain the difference between independent events and conditionally
independent events.
 Calculate the probability of an event for a discrete probability function.
 Define and calculate a conditional probability.
 Distinguish between conditional and unconditional probabilities.
 Explain and apply Bayes’ rule.
 Mutually Exclusive Events
 Two events, A and B, are said to be mutually exclusive if the occurrence
of A rules out the occurrence of B, and vice versa.
 For example, a car cannot turn left and turn right at the same time.

Turn right Turn left

P[B] = 0.5 P[A] = 0.5

 For a given random variable, the probability of any of two mutually

exclusive events occurring is just the sum of their individual
probabilities.
 In statistics notation, we can write: 𝑷[𝑨 Ս 𝑩] = P[A] + P[B]
o where [A Ս B] is the union of A and B.
 This is the probability of either A or B occurring.
 This is true only with mutually exclusive events.
 Independent Events
 For two variables to be independent, the outcome of one random variable
should never influence the outcome of the other variable.
 For example, a day’s weather and the stock market happen to be two
random variables that are independent. Suppose we want to establish
the probability that tomorrow, the market will be down and rain will
fall: we should multiply the likelihoods of the two events happening
independently.

1st Variable: Weather 2nd Variable: Stock market

Tomorrow: It will rain The market will be down
P [Rain] x P [Market Down]

𝐏 𝐈𝐭 𝐫𝐚𝐢𝐧𝐬 𝐚𝐧𝐝 𝐭𝐡𝐞 𝐦𝐚𝐫𝐤𝐞𝐭 𝐢𝐬 𝐝𝐨𝐰𝐧 = 𝐏 [𝐑𝐚𝐢𝐧 ∩ Market down] =

= P [Rain] * P [Market Down]
 The joint probability is the likelihood that it will rain and at the same
time, the markets will be down.
 Conditional Probabilities
 Conditional probability is the probability of one event occurring with some
relationship to one or more other events.
Example
 Event A is that GDP will grow, and there’s a probability of 40% that this will
happen.
 Event B is that interest rates will rise, and that has a probability of 0.5
(50%).
 A conditional probability would look at these two events in relationship with
one another, such as the probability that there will be GDP growth given
that interest rates rise.

GDP growth Interest Rates Rise

P(A | B)
“given”
 Conditional Probabilities
The formula for conditional probability is:
𝑷 𝑨 𝒂𝒏𝒅 𝑩
𝑷 𝑨𝑩 =
𝑷 𝑩
Example
 In a group of 100 investors,
o 40 buy stocks,
o 30 purchase bonds, and
o 20 purchase stocks and bonds.
 If an investor chosen at random bought bonds, what is the probability they
also bought stocks?
 Conditional Probabilities
Solution
Event Notation Probability
Buys stocks A 0.4 (=40/100)
Buys bonds B 0.3 (=30/100)
Buys stocks and A and B 0.2 (= 20/100)
bonds

 We want the probability of an investor buying stocks given that they have
already bought bonds.
 This is P(A | B)
 P(A|B) = P(A and B) / P(B) = 0.2/0.3 = 67%
 Summary
P(A∩B)
P(A|B) = P(A∩B) = P(A|B)P(B)
P(B)

 For independent events, however,

P(A|B) = P(A) and P(A∩B) = P(A and B) = P(A) P(B)

 For mutually exclusive events, however,

P (A or B) = P (A) + P(B)

 Given two events, A and B that are not mutually exclusive, the
probability that at least one of the events will occur is given by:
P (A or B) = P (A) + P (B) - P(A∩B)
 A Description of Bayes’ Theorem
 Bayes theorem is named after the Reverend T Bayes and is used
extensively in Bayesian methods of statistical inference.

 The theorem helps us use a known outcome to predict the

sequence of events leading up to that outcome.

 In essence, it allows us to “turn around” probabilities in light of

new information.
 A Description of Bayes’ Theorem
Example 1
 Assume you don’t care enough about politics and you haven’t
bothered to find out who won the last presidential election, but you
do know that there were only two candidates: a republican and a
democrat.
 However, a tax cut was recently enacted and you wonder what
kind of president would enact that policy. Assume that you must
guess from which camp the sitting president comes.
 From analyzing historical data, you find out that the probability of a
republican president enacting a tax cut is 80%. On the same note,
the probability of a democrat president enacting a tax cut is 20%.
 Can these probabilities help you work out the probability that
the sitting president is republican or democrat?
o The answer is yes, thanks to Bayes theorem.
 A Description of Bayes’ Theorem
Example 1
 What do we have?
o The probability of a tax cut given that the president is a
republican/democrat.

 What do we want?
o The probability that the president is a republican/democrat given
that there has been a tax cut.
 A Description of Bayes’ Theorem
Example 2
 Assume you know that the probability of a certified hedge fund
manager earning a return in excess of the market is X%.
 You also know that the probability of a non-certified manager
earning a return in excess of the market is Y%.
 If it is known that a manager has actually earned an excess return,
can these probabilities help you work out the probability that the
manager is certified or not certified?
o Again yes, thanks to Bayes theorem.
 A Description of Bayes’ Theorem
Example 2
 What do we have?
o The probability of an excess return given that the manager is
certified/not certified.

 What do we want?
o The probability that the manager is certified/not certified given that
there has been an excess return.
 Bayes’ Theorem – The Simple Case
 Using conditional probabilities, we know that:
P A∩B P B∩A
P AB = and, P BA =
P B P A
 But intuitively, “A and B” and “B and A” both have the same
meaning, that is,
P(A ∩ B) = P(B ∩ A)
 Thus, we can deduce that,
P AB P B =P BA P A
 Making P(A|B) the subject, we have:
P(B|A)P(A)
P AB =
P(B)
 This equation represents the theorem in its simplest case.
 Bayes’ Theorem – The General Case
 More generally,

P Ei P(A|Ei )
P Ei A = σn
j=1 P Ej P(A|Ej )
i = 1, 2, 3, …, n

 Where
o The values P(Ej) are known as prior probabilities;
o The event A is some event which is known to have occurred;
o The conditional probability P(Ei | A) is known as the posterior
probability.
 Prior vs. Posterior
 Prior probability is the probability of an event before new data is
collected. It is the best rational assessment of the probability of an
outcome based on existing knowledge before an experiment is
performed.
 A posterior probability is the revised or updated probability of
an event occurring after taking into consideration new information.

Application of
Prior New Posterior
Bayes’
Probabilities Information Probabilities
Theorem
 Prior vs. Posterior
In the tax cut scenario from earlier,
 A priori probability would be the probability that the president
elected is republican, or the probability that he is a democrat.

 A posterior probability would be the probability of a republican

president given that there has been a tax cut.

 In summary, Bayes theorem helps us to find:

P B | A P A | B 
having
already
known
 Applying Bayes’ Theorem
Example 3
 A consulting firm submitted a bid for a large risk management
consulting contract.
o The firm’s management felt it had a 50% chance of landing the
project.
 However, the client to whom the bid was submitted subsequently
asked for additional information.
 From its records, the consultancy firm knows that the client asked
for additional information for:
o 70% of successful bids; and
o 40% of unsuccessful bids.
 What is the posterior probability that the bid will be successful given
a request for additional information?
 Applying Bayes’ Theorem
Solution
 Let S1 denote the event of a successful bid.
 S2 is the event of an unsuccessful bid (not obtaining the contract).
 A is the event of being asked for additional information about a bid.
Event Probability of Probability of
event additional info
Successful 0.5 0.7
Unsuccessful 0.5 0.4
 We have:
P(S1) = 0.5 and P(S2) = 0.5 (these are priori
probabilities)
P(A|S1) = 0.7 and P(A|S2) = 0.4
 We want P(S1|A):
𝑃 𝑆1 ×𝑃 𝐴 𝑆1 0.5×0.7
𝑃 𝑆1 𝐴 = = =
𝑃 𝑆1 ×𝑃 𝐴 𝑆1 +𝑃 𝑆2 ×𝑃 𝐴 𝑆2 0.5×0.7+0.5×0.4
0.63
 Applying Bayes’ Theorem
Example 4
 Suppose you are an equity analyst for XYZ investment bank. You
use historical data to categorize the managers as excellent or
average.
o Excellent managers outperform the market 70% of the time;
and
o Average managers outperform the market only 40% of the time.
o 20% of all fund managers are excellent managers; and
o 80% are simply being average.
o The probability of a manager outperforming the market in any
given year is independent of their performance in any other
year.
 A new fund manager started three years ago and outperformed the
market all three years.
 What’s the probability that the manager is excellent?
 Applying Bayes’ Theorem
Solution
Kind of manager Probability Probability of
beating market
Excellent 0.2 0.7
Average 0.8 0.4

 Let E be the event of an excellent manager, and A represent the event of an

average manager.
o P(E) = 0.2 and P(A) = 0.8 Power of
 Further, let O be the event of outperforming the market. three to
 We know that: indicate three
o P(O|E) = 0.7 and P(O|A) = 0.4 consecutive
 We want P(E|O) years
P O E P(E) 0.73 ×0.2 0.0686
o P EO = = 3 = = 0.5726 or
P O E P E +P O A P(A) 0.7 ×0.2+0.4 3 ×0.8 0.1198
57.26%
 Book 2 – Quantitative Analysis

FUNDAMENTALS OF PROBABILITY
MULTIVARIATE RANDOM VARIABLES

Learning Objectives Recap

 Describe an event and an event space.
 Describe independent events and mutually exclusive events.
 Explain the difference between independent events and conditionally
independent events.
 Calculate the probability of an event for a discrete probability function.
 Define and calculate a conditional probability.
 Distinguish between conditional and unconditional probabilities.
 Explain and apply Bayes’ rule.