FRM-I - Book Quant Analysis

Index
Sr No. Topics Page No.

01. Quantitative Analysis-I 001
02. Quantitative Analysis-II 057
03. Quantitative Analysis-III 127
04. Quantitative Analysis-IV 201
05. FMP-I 272
06. FMP-II 340
07. FMP-III 427
08. FMP-IV 504
09. Valuation and Risk Models-I 623
10. Valuation and Risk Models-II 706
11. Foundation of Risk Management-I 788
12. Foundation of Risk Management-II 863
EduPristine FRM-I (2016) 0
Quantitative Analysis - I
EduPristine www.edupristine.com
Agenda
Probability
Counting Principle
Random variable
Probability Concepts, Bayes Theorem
Basic Statistics
Mean, Median, Mode, Variance, Covariance, Correlation
Skewness, Kurtosis
BLUE
Probabilities

Counting Principle
Number of ways of selecting r objects out of n objects

nC
r
n!/(r!)*(n-r)!
Number of ways of giving r objects to n people, such that repetition is allowed
Nr
Question-Counting Principle
In how many ways 3 stocks can be chosen out of 10 stocks in a portfolio? (Combination)
Choosing 3 out of 10 stocks is basically the number of combinations of 3 objects out of 10
Therefore, the number of ways are 10 C 10!
120
3!*(10 3)!
3
In how many ways 3 stocks can be sold, if the sold stock is bought back in the portfolio before the
next stock is sold?
First stock can be sold in 10 ways
Second can be sold in again 10 ways
Third stock can again be sold in 10 ways
Therefore total number of ways become =103 = 1,000

Question
You wish to choose a portfolio of 3 bonds and 4 stocks from a list of 5 bonds and 8 stocks.
How many different 7 asset portfolio can you make from this list.
A. 80
B. 700
C. 1,716
D. 100,800
Solution
B.
5 5! 8! 5* 4 8*7 *6*5
C3 *8 C4 * * 10 * 70 700
3!(5 3)! 4!(8 4)! 2 *1 4 * 3 * 2 *1

Probability: Definitions
Let us consider a Problem: A single 6-sided die is rolled 100 times, and the outcomes are noted..
What is the probability of each outcome? What is the probability of rolling an even number? of
rolling an odd number?
An Experiment is a situation involving chance or probability that leads to results called outcomes.
Example: In the above problem, rolling the dice is the experiment
An outcome is the result of a single trial of an experiment.
Example: when we roll the dice one, the possible outcomes are 1,2,3,4,5 or 6.
An event is a set of outcomes of an experiment (a subset of total possible outcomes).
Example: One event A can be defined as getting an even number, when we roll the dice
A trial constitutes rolling the die once. The experiment can have several trials, like 100 or 200.
Probability: Definitions
A probability experiment involves performing a number of trials to enable us to measure the

chance of an event occurring in the future. A trial is a process by which an outcome is noted
Probability of an event to occur is defined as number of cases favorable for the event, over the
number of total outcomes possible in unbiased experiment.
Event A Rolling an even number
Event A possible outcomes: {2,4,6}
P(A) = # ways of to roll an even number / Total # of sides = {2,4,6}/{1,2,3,4,5,6} = 50%
Event B Rolling a 6
P(B) = # ways to roll a 6/ Total # of sides = 1/6
Event C Rolling a number less than 3, i.e. {1,2}
P(C) = 2/6

Random Variable Discrete & Continuous
A random variable, usually written as X, is a variable whose possible values are numerical
outcomes of a random phenomenon
Examples:
X No of times we receive head when a coin is flipped 5 times
X Average height of a random group of 25 people selected from Bangalore
Discrete Random Variable: A random variable that can take finite number of distinct values
Example: consider an experiment where a coin is tossed 3 times. Random variable X is defined as
X No of heads obtained when we flip a coin 3 times; X {0,1,2,3}
Continuous Random Variable: A random variable that can take an infinite number of possible
values. Usually defined over an interval of values.
Example: consider an experiment where we measure average amount of rainfall received in
Bangalore over a year. Here random variable X is defined as
X any real number between 0 to 200mm
Probability Function on discrete random variable: P(x)
Probability function, P(x), denotes the probability that the discrete random variable X is equal to a
specific value x. P(x) = P(X=x)
W
For any random variable, sum of all the probabilities must equal 1
Example: Random variable X no of heads we receive, when we toss a coin twice
3 Probable outcomes: xi {0,1,2}
P(X=0) = 25%, P(X=1) = 50%, P(X=2) = 25%
Wi) = 1

Probability Density Function: f(x)
For continuous random variable, probability of any specific value is zero. Probability is always
calculated for a range of values.
Example: Continuous Random variable X average rain in Bangalore in a year
P(X = 10mm) = 0; (i.e. rain can never be exactly 10mm, it can be 9.999mm or 10.001mm etc)
But P(9.999mm < X < 10.001mm) = 20% (probability is always calculated for a range or interval)
For a Continuous Random Variable, PDF tells us the likelihood of outcomes occurring between
any two pointes. It is used to calculate the probability that outcomes of the continuous
distribution lies within a particular range.
Wl yu) = area under the PDF curve bounded by rl and ru
P(rmin ymax) = 1
Wy
i.e. minimum rain Bangalore may receive is zero, and maximum it can receive is infinity
Cumulative Density Function: F(x)
CDF defines the probability of the continuous random variable being less than a certain value.
Usually denoted as F(x).
F(a) = -a W = area under the PDF curve to the left of point a
&
If x1 < x2, F(x1&2). i.e. CDF is a non decreasing function ( in other words as the upper range of
possible outcomes increases, the cumulative probability also increases or remains same)
F(rmaxWmax ) = 1
P(x > a) = 1 &WW
W& F(a)

Question
Define the probability density function for the price of a zero coupon bond with a notional value of
$10 as
PDF, where x is the price of the bond
Thus P(a ab f(x)dx = (b2 a2)/100
Using the above equation for Probability, calculate
(i) Probability that the price of bond is between 8 and 9? i.e. find P(8
(ii) &W
(iii) W
EduPristine
E FRM-I (2016) 14
Solution
PDF, where x is the price of the bond

Using the PDF, given: P(a ab f(x)dx = (b2 a2)/100
(i) P(8 89 f(x)dx
= (81 64)/100 = 0.17
Alternatively, we can calculate it using the area under the curve f(x) bounded by range 8 and 9
(ii) &W
= (64 0)/100 = 0.64
W
=(100 0)/100 = 1
Note: probability of all possible outcomes is always 100%
EduPristine
E FRM-I (2016) 15
Some definitions and properties of Probability
Mutually Exclusive events: If one event occurs, then other cannot occur. For a given random
variable, the probability of any of two mutually exclusive events occurring is just the sum of their
individual probabilities. P(A U B) = P(A) + P(B), if A and B are mutually exclusive
Example: Random variable X annual return on portfolio
Event A: return is less then -10%; P(A) = 10%
Event B: return is more than 10%; P(B) = 20%
WhWW
Note: here A and B are exclusive, because, return cannot be more than 10% and less than -10%
simultaneously
Exhaustive: All exhaustive events taken together form the complete sample space
For previous example, if we consider another event C : -10%
Now events A, B and C are mutually exclusive and exhaustive
Thus P(A U B U C) = 1, i.e. P(C) = 70%
Some definitions and properties of Probability
Independent Events: One event occurring has no effect on the other event. i.e. we have more
than 1 random variables.
Event A: Kohli scores a century in the next match, P(A) = 10%
Event B: Stock market is up tomorrow, P(B) = 8%
P(A and B) = P(A B) = P(A) x P(B) Also called as Joint Probability
Since the two events are independent, the probability that Kohli gets a century and the Stock
market goes up is 10% x 8% = 0.8%
What if A and B are dependent?
Event A: Kohli scores a century in the next match, P(A) = 10%
Event B: India wins the match, P(B) = 40%
P(A and B) = ?
Conditional probability comes into picture.

Conditional & Joint Probability
Joint Probability
A statistical measure where the likelihood of two events occurring together and at the same point in time are
calculated. Joint probability is the probability of event Y occurring at the same time event X occurs.
Notation for joint probability takes the form:
WyzWyzWhich reads the joint probability of X and Y
The following table shows the joint probability of different events. Lets say an economist is predicting the
market scenario and the price of IBM stock from the next year.
Next year market can be Good, Bad or Neutral
IBM stock may go up or go down
Joint Probability Table:
Market
Good Bad Neutral Total
IBM
UP 10% 30% 5% 45%
DOWN 0% 15% 40% 55%
Total 10% 45% 45% 100%
The probability of IBM stock being Up and Market being Good is 10%
Similarly, the probability of IBM stock being down and Market being neutral is 40%
Conditional & Joint Probability
Conditional Probability
Probability of an event or outcome based on the occurrence of a previous event or outcome. Conditional
probability is calculated by multiplying the probability of the preceding event by the updated probability of
the succeeding event
The probability of event A given that the event B has occurred is P(A/B), which is equal to the ratio of joint
probability of A and B, and unconditional probability of B. P( A B)
P( A / B)
P( B)
The unconditional probability of market being Neutral is 45%. Then using the table below we can
find 3 conditional probabilities.
P(Up/Neutral) = 0.05/0.45
P(Up/Good) = 0.1/0.1
P(Down/Bad) = 0.15/0.45
Joint Probability Table:
Market
Good Bad Neutral Total
IBM
UP 10% 30% 5% 45%
Unconditional Probabilities of
DOWN 0% 15% 40% 55% IBM stock being UP / Down
Total 10% 45% 45% 100%

Conditional and Unconditional Probabilities
Market
Unconditional Probability
10% 45% 45%
of Market
Good Bad Natural
IMB Up Down Up Down Up Down
10% 10% 10% 10% 10% 10%
Calculate:
Unconditional Probability of market to be good next year?
Conditional Probability of IBM stock rising when the market is neutral?
Conditional Probability of market being good when IBM stock is down?
Question
For a bond with B rating, assume 1 year probability of default for each issuer is 6%, and that
default probability of each issuer are independent. What is the probability that both issuers avoid
default during the 1st year.
A. 88.0%
B. 88.4%
C. 94.0%
D. 96.4%

Solution
B.
Both would avoid default only if None defaults
This implies that first does not default AND second does not default
= (1 PD (first)) x (1 PD (second))
= (1 0.06) x (1 0.06) = 0.884 = 88.4%
Some Properties of Probability
The probability of happening of event A or event B can be given as the sum of the three portions
defined by the figure below:
P( A) P( A B) P (B )
P(A) P(B) - P(A B)

P(A B)
P(A) P(B) if A and B are Mutually Exclusive
P( A) P (B )
P( A B) 0

Question
Jensen, a portfolio manager is managing two portfolios. One for High Net Worth Individuals (HNI)
and second for Low Net Worth Individuals (LNI)
HNI portfolio contains 5 bonds and 7 stocks and LNI contains 6 bonds and 11 stocks
One instrument from HNI is transferred to LNI portfolio
What are the probabilities that a stock and a bond is transferred from HNI to LNI?
A. 42%, 58%
B. 58%, 42%
C. 50%, 50%
D. 33%, 67%
Now Jensen selects an instrument from LNI, what is the probability that instrument selected is
stock?
A. 0.5382
B. 0.7821
C. 0.6435
D. None of these
Solution
B.
C.
Here required probability = [P(stock transferred from HNI) AND P(Stock selected from LNI)] OR
[P(bond transferred from HNI) AND P(Stock selected from LNI)]
So, the required probability = (7/12) (11/18) = 139/216 = 0.6435
Hence option C is correct

Sum Rule & Bayes Theorem
The unconditional probability of event B is equal to the sum of joint probabilities of event (A,B)
and the probability of event (A,B). Here A is the probability of not happening of A
The joint probability of events A and B is the product of conditional probability of B, given A has occurred and
the unconditional probability of event A
P( B) P ( A B ) P ( Ac B ) P( B / A) P( A) P( B / Ac ) P( Ac )
We know that P(AB) = P(B/A) * P(A)
Also P(BA)= P(A/B) * P(B)
P( B / A) * P( A)
Now equating both P(AB) and P(BA) we get: P( A / B)
P( B)
P(B) can be further broken down using sum rule defined above:
P( B / A) P( A)
P( A / B)
P( B / A) P( A) P( B / Ac ) P( Ac )
Question
Out of a group of 100 patients being treated for chronic back trouble, 25% are chosen at random
to receive a new, experimental treatment as opposed to the more usual muscle relaxant-based
therapy which the remaining patients receive. Preliminary studies suggest that the probability of a
cure with the standard treatment is 0.3, while the probability of a cure from the new treatment
is 0.6.
How many patients (on an average) out of the 100 patients selected at random would be cured?
A. 30
B. 40
C. 37.5
D. 42.5
Some time later, one of the patients returns to thank the staff for her complete recovery.
What is the probability that she was given the new treatment?
A. 0.375
B. 0.425
C. 0.4
D. 0.425

Solution
C.
25% are given new treatment =>75% are given old treatment.
WWEWEWKWK
So out of 100 patients 37.5 will get cured
C.
Apply Bayes Theorem
P (New / Cure) = P (Cure / New) * P (New) / P (Cure) = 0.6 * 0.25 / 0.375 = 0.40
Question
Calculate the probability of a subsidiary and parent company both defaulting over the next year.
Assume that the subsidiary will default if the parent defaults, but the parent will not necessarily
default if the subsidiary defaults. Assume that the parent had a 1 year PD = 0.5% and the
subsidiary has 1 year PD of 0.9%.
0.45%
0.5%
0.545%
0.55%

Solution
B.
P (S| P) = 1 = P(P & S)/P(P)
P(P & S) = P(P) = 0.5%
Basic Statistics

Descriptive & Inferential Statistics
Descriptive Statistics is used to summarize important characteristics of large data set.

Steps in descriptive statistics:
Collect data
Classify data
Summarize data
Present data
Proceed to inferential statistics if there are enough data to draw a conclusion
Inferential Statistics Making Inference about the larger group based on actually observed smaller
group
Making Forecast, Estimates etc.
Usage of probability theory for making inference
Basic Statistics
A random variable is characterized by its distribution function. Instead of having to report the
whole function, it is convenient to summarize it by a few parameters, or moments.
The first four moments about the mean:
Mean
Variance
Skewness
Kurtosis

Mean
The expected value(Mean) measures the central tendency, or the center of gravity of the
population n
It is given by: x i
P E( X ) i 1
N
A family has 4 members, father, mother and 2 kids Hemal and Rishi who are twins. The average
age of the family members is 20 years. Age of mother and father is 30 and 32 respectively.
Can you tell the age of Hemal?
The graph shows the mean of normal distributions
0.45
0.40
0.35
0.30 Standard Normal Distribution
0.25 P = 0, V = 2
0.20
0.15 P = 1, V = 1
0.10
0.05
0
-4 -2 0 2 4
Mean of Random variables (Discrete & Continuous)
For a Discrete Random variable, X, with possible values xi, and corresponding probabilities pi,
n
mean is given as
P E( X ) px
i 1
i i
When probabilities of each possible outcome is same, than pi = 1/n, and

n
P E( X ) x
i 1
i /n
For a Continuous Random Variable X, with possible values xi , and PDF f(x)
x max
P E( X ) xf ( x)dx
x min

Median
Median is the midpoint of the data set when the data is arranged in ascending or
descending order
Half of the observations lie above the median and half are below
Median is important because arithmetic mean can be affected by extremely large or small values
Example Find the median of the following data set:
3, 13, 7, 5, 21, 23, 39, 23, 40, 23, 14, 12, 56, 23, 29
Arrange the values in ascending order

3, 5, 7, 12, 13, 14, 21, 23, 23, 23, 23, 29, 39, 40, 56
Total number of observations = 15
Median = Middle observation = 8th observation
Median = 23
Mode
Mode is the value which occurs most frequently in the data set
A data may have more than one mode or no mode
Example Find the mode for the following data set
11,3,5,11,7,3,11
In the above data set
Number 11 occurs 3 times,
Number 7 occurs 1 times.
So the number with most occurrences is 11 and is the Mode of this distribution.
Mode = 11

Expected Value
N
PX pX i 1
i i
Given the following table, calculate the expected value of x.

y Wy
1 0.3
2 0.1
3 0.2
4 0.4
Expected Value
y Wy yWy
1 0.3 0.03
2 0.1 0.02
3 0.2 0.06
4 0.4 1.6
2.70

Properties of Expectation
E(cX) = E(X) x c
yzyz
yy
E(XY) = E(X) x E(Y) [if X and Y are independent]
Variance & Standard deviation
Variance is the squared dispersion around the mean.

n
(x i P )2
VAR E[( X P ) 2 ] i 1
N
Also, if rearranged
VAR E[( X P ) 2 ] E[ X 2 ] P 2 E[ X 2 ] E[ X ]2
The standard deviation, which is the square root of the Variance, is more convenient to use,
as it has the same units as the original variable X
n
(x i P )2
SD(X) = VAR (x) V i 1
N

Covariance & correlation
Covariance describes the co-movement between 2 random numbers, given as:

Cov(X1, X212
Cov ( X , Y ) E[( X P X )(Y PY )]
Cov ( X , Y ) E ( XY ) P X PY E ( XY ) E ( X ) E (Y )
Correlation coefficient is a unit-less number, which gives a measure of linear dependence
between two random variables.
y1, X2) = Cov(X1, X212
-1
A correlation of 1 means that the two variables always move in the same direction
A correlation of -1 means that the two variables always move in opposite direction
If the variables are independent, covariance and correlation are zero, but vice versa
is not true
Question
ABC Annual stock prices
2001 2002 2003 2004 2005 2006
12% 5% -7% 11% 2% 11%
Assuming that the distribution of ABC stock returns is a population, what is the population
variance and standard deviation?
05.00
06.80
45.22
80.20
Given two random variables X and Y, what is the Variance of X given Variance[Y] = 100, Variance
[4X - 3Y] = 2,700 and the correlation between X and Y is 0.5?
56.3
113.3
159.9
225.0

Solution
B.
The population variance is given by taking the mean of all squared deviations from the mean.
2 = [(12-5.67)2 -5.67)2 -7-5.67)2 -5.67)2 -5.67)2 -5.67)2] / 6 = 45.22 (%2)
The standard deviation is the square root of the variance: 6.72%
D.
Using the theorems on variance and covariance
Variance [4X-zsysz-3)*Var[X]^(1/2)* Var[Y]^(1/2)*correlation[X,Y]
Solve for Var[X] = 225.0
Some Properties of Variance
Var (aX b) a 2Var ( X ) Variance of a constant = 0
n n n
Var ( X i ) Cov( X , X i j ) Covariance between same variables is also their variance
i 1 i 1 j 1
n n
Var ( X i ) Var ( X ) i
For independent or uncorrelated variables,
i 1 i 1 covariance or correlation = 0
Var (aX bY ) a 2Var ( X ) b 2Var (Y ) 2abCov ( X , Y )

Skewness
Skewness describes departures from symmetry

n
(x
i 1
i P )3
Sk
V3
Symmetric Distribution
Skewness can be negative or positive.
Negative skewness indicates that the distribution

has a long left tail, which indicates a high probability
of observing large negative values.
Negatively Skewed Distribution
If this represents the distribution of profits and

losses for a portfolio, this is a dangerous situation.
Positively Skewed Distribution
Kurtosis
Kurtosis describes the degree of flatness of a distribution, or width of its tails

n
(x
i 1
i P )4
K
V4
Because of the fourth power, large observations in the tail will have a large weight and hence
create large kurtosis. Such a distribution is called leptokurtic, or fat tailed
0.45 Platykurtic
A kurtosis of 3 is considered average 0.4
K<3
High kurtosis indicates a higher probability 0.35

Leptokurtic
0.3
of extreme movements K>3
0.25
0.2 Mesokurtic
0.15 K=3
0.1
0.05
0
-4 -3 -2 -1 0 1 2 3 4

Coskewness and Cokurtosis
Just as the concept of variance was extended to a set of two variables as covariance, concept of
skewness and cokurtosis can be extended to the set of more variables as coskewness. It is possible
that the set of variables have same mean and variable but not the same skewness and kurtosis
because the ranking of values will be different for these two variables at different point of time
Most risk models ignore the effects of coskewness and cokurtosis
The main reason for this is that as the number of variables increases, the number of coskewness
and cokurtosis terms increase rapidly
BLUE
Unbiased parameter is one for which the expected value of the estimator is equal to the
parameter you are trying to estimate
An unbiased parameter is also efficient if its sampling distribution has minimum variance
An estimator is also consistent if the accuracy of the parameter estimate increases as the sample
size increases
If this estimator is linear, we call it the Best Linear Unbiased Estimator (BLUE)

Question 1
The random variables X and Y have variances of 2 and 3 respectively, and covariance of 0.5.
dyz
A. 13
B. 29
C. 35
D. 41
Question 2
You are given the following information about the returns of stock P and stock Q:
Variance of return of stock P = 100.0
Variance of return of stock Q = 225.0
Covariance between the return of stock P and the return of stock Q = 53.2
At the end of 1999, you are holding USD 4 million in stock P. You are considering a strategy of
shifting USD 1 million into stock Q and keeping USD 3 million in stock P. What percentage of risk,
as measured by standard deviation of return, can be reduced by this strategy?
A. 0.50%
B. 5.00%
C. 7.40%
D. 9.70%

Question 3: FRM Exam 2009
Which type of distribution produces the lowest probability for a variable to exceed a specified
extreme value X which is greater than the mean assuming the distributions all have the same
mean and variance?
A. A leptokurtic distribution with a kurtosis of 4
B. A leptokurtic distribution with a kurtosis of 8
C. A normal distribution
D. A platykurtic distribution
Question 4
Which of the statements is false

A. For a symmetric distribution (skewness = 0), mean = median = mode
B. For a positively skewed distribution: mean > median > mode
C. For a negatively skewed distribution: mean < median < mode
D. For a positively skewed distribution: mean > mode > median

Solution
1. B.
syzsysz
Var(X - zsysz-2*Cov(x,y)
Var(cX) = c^2 * Var(X)
Cov (ax,by) = abCov(x,y)
^syz2 sy2 sz
2. B.
5.00%
Solution
3. D.
By definition, a platykurtic distribution has thinner tails than both the normal distribution and any
leptokurtic distribution. Therefore, for an extreme value X, the lowest probability of exceeding it will be
found in the distribution with the thinner tails
A. Incorrect. A leptokurtic distribution has fatter tails than the normal distribution. The kurtosis indicates
the level of fatness in the tails, the higher the kurtosis, the fatter the tails. Therefore, the probability of
exceeding a specified extreme value will be higher
B. Incorrect. Since answer A. has a lower kurtosis, a distribution with a kurtosis of 8 will necessarily produce
a larger probability in the tails
C. Incorrect. By definition, a normal distribution has thinner tails than a leptokurtic distribution and larger
tails than a platykurtic distribution
4. D.
In a skewed distribution, the mean is pulled in the direction of the extreme scores or tail (same as the
direction of the skew), and the median is between the mean and the mode.

dz
help@edupristine.com
www.edupristine.com
Quantitative Analysis - II
Agenda
Probability distributions
Discrete vs. Continuous probability distribution
Discrete probability distribution: Bernoulli, Binomial, Poisson
Continuous Probability Distribution
Sampling
Central Limit Theory
Hypothesis testing - Part 1
Probability Distributions

Probability Distribution
A Random Variable is a function, which assigns unique numerical values to all possible outcomes
of a random experiment under fixed conditions. A random variable is not a variable but rather a
function that maps events to numbers
Probability distribution describes the values and probabilities that a random event can take place. The values
must cover all of the possible outcomes of the event, while the total probabilities must sum to exactly 1, or
100%
Example
Suppose you flip a coin twice.
There are four possible outcomes: HH, HT, TH, and TT.
Let the variable X represent the number of Heads that result from this experiment
It can take on the values 0, 1, or 2
X is a random variable (its value is determined by the outcome of a statistical experiment)
A probability distribution is a table or an relation that links each outcome of a statistical experiment with its
probability of occurrence
Ey WWy
0 0.25
1 0.50
2 0.25
Continuous & Discrete Probability Distributions

If a variable can take on any value between two specified values, it is called a continuous variable
Otherwise, it is called a discrete variable
If a random variable is a discrete variable, its probability distribution is called a discrete
probability
For example, tossing of a coin & noting the number of heads (random variable) can take a discrete value
Binomial probability distribution, Poisson probability distribution
If a random variable is a continuous variable, its probability distribution is called a continuous
probability distribution
The probability that a continuous random variable will assume a particular value is zero
A continuous probability distribution cannot be expressed in tabular form
An equation or formula is used to describe a continuous probability distribution (called a probability density
function or density function or PDF)
The area bounded by the curve of the density function and the x-axis is equal to 1, when computed over the
domain of the variable
Normal probability distribution, Student's t distribution are examples of continuous probability distributions

Bernoulli Distribution
Assumptions
The outcome of an experiment can either be success (i.e., 1) and failure (i.e., 0)
Pr(X=1) = p, Pr(X=0) = 1-p, or
The expected value E[X] of the event is equal to the probability of success(p)
E[X] = p
The variance of the event is the product of the probability of success and probability of failure:
Var(X) = p(1-p)
Binomial Distribution
Assumptions:
There are n trials
Each trial has two possible outcomes, success or failure
The probability of success p is the same for each trial
Each trial is independent
If we take n Bernoulli trials, and say out of those n trials we have total number of x successes,
then the probability of such an event can be given as:

P( X x) C x * p x * (1 p ) ( n x )
n
The expected number of successes y
The variance of number of successes sy-p)

Question
1. There are 10 bonds in a credit default swap basket; the probability of default for each of the
bonds is 5%. The probability of any one of the bond defaulting is completely independent of what
happens to the other bonds in the basket. What is the probability exactly one bond default?
A. 5%
B. 50%
C. 32%
D. 3%
2. Company ABC was incorporated on January 1, 2004. it has expected annual default rate of 10%.
Assuming a constant quarterly default rate, what is the probability that company ABC will not
have defaulted by April 1, 2004?
A. 2.4%
B. 2.5%
C. 97.4%
D. 97.5%
3. A corporate bond will mature in 3 years. The marginal probability of default in year one is 0.03%.
The marginal probability of default in year 2 is 0.04%. The marginal probability of default in year 3
is 0.06%. What is the cumulative probability that default will occur during the 3 year period?
A. 0.1247%
B. 0.1276%
C. 0.1299%
D. 0.1355%
Solution
1. C.
One particular bond defaults and other nine do not with the probability 0.05* (0.95)^9 can happen in 10
different ways
= 10 * 0.05^1* (0.95)^9 = 32%
2. C.
Let the probability of not defaulting in 1 quarter is (nd). Then the probability of not defaulting for a full year is
(nd)4. This implies that the probability of defaulting within 1 year time is {1-(nd)4}, which is given as 10%.
1-(nd)4 = 0.1 which implies
(nd) = 0.91/4
= 97.4%
3. C.
The cumulative probability of default= 1-{Product of marginal probabilities of not defaulting
= 1-{(1-0.0003)*(1-.0004)*(1-0.0006)}
= 0.001299
Therefore the cumulative probability of default is 0.1299%

Poisson Distribution
Assumptions:
The probability of observing a single event over a small interval is approximately proportional to the size of
that interval
The probability of an event within a certain interval does not change over different intervals
The probability of an event in one interval is independent of the probability of an event in any other interval
which is not overlapping
Poisson distribution is a special case of Binomial distribution when the probability of success (p)
becomes very small and the number of events (n) becomes very large in such a way that the
product of both gives a constant (O).
Fix the expectation O =n * p
Number of trials n of
A Binomial distribution will become a Poisson distribution
yOsyO
Ox O
e xt0
P( X x) x!
0 otherwise
Plots of Poisson Distribution
0.40
=1 0.20
=5
0.35 0.18
0.16
0.30
0.14
0.25 0.12
0.20 0.10
0.15 0.08
0.06
0.10
0.04
0.05
0.02
0.00 0.00
0 2 4 6 8 10 12 14 16 18 20 22 24 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28
0.12 =15 0.09 =25

0.08
0.10
0.07
0.08 0.06
0.05
0.06
0.04
0.04 0.03
0.02
0.02
0.01
0.00 0
0 2 4 6 8 10 12 14 16 18 20 22 24 0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54

Question
1. When can you use the Normal distribution to approximate the Poisson distribution, assuming you
have "n" independent trials each with a probability of success of "p"?
A. When the mean of the Poisson distribution is very small
B. When the variance of the Poisson distribution is very small
C. When the number of observations is very large and the success rate is close to 1
D. When the number of observations is very large and the success rate is close to 0
2. The number of false fire alarms in a suburb of Houston averages 2.1 per day. What is the
(approximate) probability that there would be 4 false alarms on 1 day?
A. 1.0
B. 0.1
C. 0.5
D. 0.0
Solution
1. C.
The Normal distribution can approximate the distribution of a Poisson random variable with a large lambda
d

INCORRECT: A, The mean of a Poisson distribution must be large to allow approximation with a Normal
distribution
INCORRECT: B, The variance of a Poisson distribution must be large to allow approximation with a Normal
distribution
INCORRECT: D, The Normal distribution can approximate the distribution of a Poisson random variable with a

2. B.
Use Poisson distribution
P(X = x) = [O x *e- O]/ O!

Some Extra Questions
Question: Bayes Theorem
There is a city which hosts two taxi-cab companies, the Blue Cab Co. and the Green Cab Co. Blue
cabs are blue and Green cabs are green; they are otherwise identical. 70 percent of the cabs in the
city are Blue cabs, and 30 percent of the cabs in the city are Green cabs. Moreover, historically
speaking, Blue cabs have been involved in 70% of all traffic accidents in the city that involved cabs,
and Green cabs have been involved in 30% of all traffic accidents in the city that involved cabs.
One night, there is a traffic accident involving a taxi-cab in the city, to which there is one witness.
Authorities perform extensive tests on the witness, and determine that his ability to recognize
cabs by their color at night is approximately 80 percent accurate and 20 percent inaccurate
(meaning that when he is wrong he does not say he doesn't know, but rather misidentifies it as
being of the other color). The witness says the taxi-cab involved in the accident was 'blue.' On
these facts, and strictly assuming the taxi-cab was not from some other city, what is the
approximate probability that the taxi-cab involved in the accident belonged to the Blue Cab Co.

Solution
Let P(R) be the probability of witness being accurate. Then P(R) = 0.8 which implies P(W) = 0.2 i.e.
probability of witness being wrong
Let P(B/R) = Probability of accident by a blue car, conditional on the fact that the statement is a
right statement
Then P(B/R) = 0.7 Also P(B/W)=0.3, Similarly P(G/R) = 0.3
Here we need to find P(R/B) i.e. If the witness has said that it was a blue car, then what's the
probability that it was actually blue
Applying Bayes Theorem now:
P(R/B) = P(B/R) * P(R)/P(B)
Here we know all except P(B):
WWZWZWtWt

Therefore: P(R/B) = 0.7*0.8/0.62 = 0.903
Understanding Bayes Theorem
0.7 0.3
Witness is wrong Witness is right Witness is wrong Witness is right
Witness

Identifies it as: =0.06
Cases when witness Total number of cases

rightly identifies it when witness identifies
as blue=0.56 accident by blue cab

Therefore the probability of the car being actually blue, when the witness identified it as blue
equals: (0.56/0.62) 0.903

Question: Sum Rule & Bayes Theorem
Jack has 3 white balls and 2 red balls in his box while his friend Andrew has 4 white and 5 red balls
Andrew took 1 white ball from jack and gave him 1 red ball in compensation. Now calculate the probability of
picking a red ball from Andrews box
After the exchange, Tom stole a ball from one of the boxes and found that its white. If you have to tell who
lost his white ball, what should be your say?
Solution
Initially Jack has 3 white balls and 2 red balls in his box while his friend Andrew has 4 white and 5
red balls
After the exchange Jack has 2 white and 3 red balls, and Andrew has 5 white and 4 red balls.
Therefore the probability of picking a red ball from Andrews box is:
P(RAndrew
Now Tom stole a white ball from one of the two boxes. To make a calculated guess about who lost
1 white ball, we need to calculate the conditional probabilities
P(Jacks box/If the balls is White)= Probability of white balls in Jacks box/(Probability of white ball
:W

2/5
18 / 43
2 / 5 5 / 9 5/9
Similarly, P(Andrews box / White) 25 / 43
2 / 5 5 / 9
Point to note here is that the white ball can come from 2 boxes only, so the sum of conditional

our case

Question: Probability Matrix
A company has issued both bonds and stocks. Bonds can be upgraded, downgraded or remain
unchanged. Whereas stocks can either outperform or underperform the broader index. The
probability matrix of bonds and stock performance is as follows:
Bonds
Upgrade No change Downgrade Total
Stocks
Outperfom 10% 30% Y C
Underperform X 15% 35% 55%
Total A 45% B 100%
A. Calculate the values for missing probabilities: X,Y,A,B and C

B. What is the probability that the Stocks outperform, given that Bonds are downgraded?
C. What is the probability that the stocks outperform?
D. What is the probability that Bonds are upgraded, given that Stocks have underperformed?
Solution
A. X = 5%, Y=5%, A=15%, B=40%, C=45%
B. P(stocks outperform/Bond downgraded)

= P(stock outperform, Bond downgraded)/P(Bond downgraded)
= 5%/40% = 12.5%
C. p(stocks outperform)
= P(stocks outperform/Bond downgraded) *P(Bond downgraded)
WW
WW

D. P(Bond upgraded/Stock underperform)

= P(Bond upgraded, stock underperform)/P(stock underperform)
= 5%/55% = 9.1%

Continuous probability distribution
Uniform distribution
The simplest continuous distribution function is the uniform distribution. This is defined over a

The density function is:

Its mean and variance are given by:
y
sy2 / 12
1
b-a
a b

Question
Assume we use continuous uniform distribution U(0,10) to generate a series of random numbers.
Which of the following statements is Correct?
A. The number 5 is likely to be observed much more often than any other number
B. Numbers between 4 and 6 are more likely to occur than the number between 6 and 10, because the first
interval is closer to the center of the distribution
C. Numbers between 1 and 3 are as likely as the number between 4 and 6
D. Numbers between 1 and 3 are less likely than the number between 4 and 6, due to skewness of
the distribution
Solution
C.

Normal (Gaussian) distribution
The normal distribution is defined by first two moments, mean () and variance (2)
The probability density function P(x) of normally distributed variable is given by:
1 ( x P )2
P( x) exp
2SV 2 2V 2
b
P ( a d X d b)
The probability of the value lying between a and b is given by:
P( x).dx
a
The expected value of a normally distributed variable: E[X]= ,

The variance of normally distributed variable: Var(X)= 2
If two variables are individually normally distributed, then the linear combination of the both is
also normally distributed
Lets take an example of two variable X1 and X2 which are normally distributed such that:
X1~N(1, 1) and X2~N(2, 2)
Then X= a.X1y2 is also normally distributed
The skewness of normal distribution is = 0 and the kurtosis is = 3
Properties of Normal Distribution
The probabilities of outcomes further above and below the mean get smaller and smaller, but do
not go to zero (extent to infinity)
It is a CDF, with total area under the curve equal to 1.
Confidence interval, is the range of values around the expected outcome, within which we expect
the actual outcome to lie in, during specified percentage of time.
Confidence Level (1 ) Confidence interval
68% -1 to 1
90% -1.65 to 1.65
95% -1.96 to 1.96
99% -2.58 to 2.58
Summation property: The property states that, assuming there is a collection of normal
distributions:
The sum of the means of all the independent normal distributions form a normal distribution
The sum of the variances of all the independent normal distributions form a normal distribution

Standard normal distribution
68% of data
95% of data
99.7% of data
To standardize a given normal distribution, z value is calculated as z = (x )/
The standard normal distribution has mean = 0 and standard deviation sigma = 1
Question: FRM Exam 2002
Consider a stock with an initial price of $100. Its price one year from now is given by
S = 100exp(r), where the rate of return r is normally distributed with a mean of 0.1 and a
standard deviation of 0.2. With 95% confidence, after rounding, S will be between
$67.57 and $147.99
$70.80 and $149.20
$74.68 and $163.56
$102.18 and $119.53

Solution
C.
Note that this is a two-
t
dsd
s 0.2) = $100 * exp(0.492) = $163.56
Question
Let X be a uniformly distributed random variable between minus one and one so that the standard
deviation of X is 0.577. What percentage of the distributions will be less than 1.96 standard
deviations above the mean:
A. 100%
B. 97.5%
C. 95%
D. Insufficient information provided

Solution
A
The answer requires understanding of distributions and standard deviation. The key is that every distribution
has a standard deviation. However the number of standard deviations associated with different probabilities
are different for each distribution. In this case 1.96 standard deviation represents a move of 1.12 or less. As
the total distribution is defined as falling between minus one and one the correct answer is A
Question
For the standard normal distribution, calculate the value of P(-W

A. 0.5683
B. 0.8794
C. 0.7831
D. 0.9145

Solution
D.
In the diagram given below, the area representing the region P(-W
d-
P(-WWW
Hence option D is correct
-1.87 -1.6 1.23 1.6
-squared, Students t, Lognormal


Students t-distribution
Symmetrical distribution centered about zero

Appropriate to use, when constructing confidence intervals for small samples (n<30) from
populations with unknown variance and a normal, or approximately normal distribution
Shape depends upon number of degrees of freedom which depends on the number of
sample observations
Flatter and thicker tails than normal distribution
Kurtosis approaches that of normal distribution with increasing number of samples
Lognormal distribution
A random variable X is said to have a lognormal distribution if its logarithm Y = ln(X) is normally
distributed
1 (ln( x) P ) 2
The lognormal density function has the following expression: P(x) exp
x 2SV 2 2V 2
(100,2) (2,0.3)
1.2
0.6
(a) 0.9 (b)

0.4
Density
0.6
0.2
0.3
0.0 0.0
0 100 200 300 400 1.0 1.5 2.0 2.5 3.0
Original Scale Log Scale +

The Chi-square distribution
If there are k independent normal variables, than the sum of their squares has a chi squared
distribution.
The chi-square distribution is a family of distributions, depending on degrees of freedom: d.f. = n
1. Mean = k, and variance = 2k
Assymmetrical, but approaches normal distribution as degrees of freedom increases. As the
variable is sum of squared values, it can only have non-negative values.
0 4 8 12 16 20 24 28 F2 0 4 8 12 16 20 24 28 F2 0 4 8 12 16 20 24 28 F2
d.f. = 1 d.f. = 5 d.f. = 15
Question FRM Exam
Assume you have empirical data showing historical returns (v) for a given financial variable
(e.g.: Forex rate), how could you perform a quick test of the validity of the power law
Prob(v > x) = K * x-a where x is large, as a good model of the tail of the distribution?
A. Plot the probability of v exceeding x standard deviations against x
B. Plot the probability of v exceeding x standard deviations against Log of x
C. Plot the Log of the probability of v exceeding x standard deviations against x
D. Plot the Log of the probability of v exceeding x standard deviations against the Log of x

Solution
D.
The mathematical relationship in the question can be rewritten (by taking the logs on both sides): Log(Prob(v
> x)) = Log(K) aLog(x), i.e. the plot of the Log of the probability of v exceeding x standard deviations against
the log of x should be a straight (decreasing) line if the relationship strictly holds.
The intercept is an estimate of Log of K and the slope of the line yields the parameter a.
Hypothesis testing and confidence intervals Part 1

Sampling
A probability sample is a sample selected such that each item or person in the population being
studied has a known likelihood of being included in the sample
The sampling distribution of the sample mean is a probability distribution consisting of all
possible sample means of a given sample size selected from a population
Need for Sampling:
The physical impossibility of checking all items in the population
The cost of studying all the items in a population
The sample results are usually adequate
Contacting the whole population would often be time-consuming
The destructive nature of certain tests
Methods of probability sampling
Methods of Probability Sampling:

Simple Random Sampling: A sample formulated so that each item or person in the population has the same
chance of being included
Systematic Random Sampling: The items or individuals of the population are arranged in some order. A
random starting point is selected and then every kth member of the population is selected for the sample
Stratified Random Sampling: A population is first divided into subgroups called strata, and a sample is
selected from each stratum
Independently and identically distributed (iid) random variables: If a sample contains random
variables such that each of them have the same probability distribution function then such
random variables are called as i.i.d
The sampling error is the difference between a sample statistic and its corresponding
population parameter

Developing sampling distributions
Suppose theres a population of 4 oldest scientists in a university: Jack, Andrew, Michelle and Tom
Random variable, X is the ages of the individuals 79
Ages of Population
78
Values of X: 78, 76, 72, 74 77
76
75
74
Summary Measure for Population Distribution 73
72
71
N 70
X
69
i Andrew Jack Michelle Tom
i 1
Average Age
N Prob. of selection
0.3
78 76 72 74
75 0.25
4 0.2
N 0.15
X
2
0.1
i
i 1
2.236 0.05
N 0
Andrew Jack Michelle Tom
All Possible samples of size n = 2
16 samples of size n = 2 each 16 Sample means

1st Obs 2nd Observation 1st Obs 2nd Observation
78 76 74 72 78 76 74 72
78 78,78 76,78 74,78 72,78 78 78 77 76 75
76 78,76 76,76 74,76 72,76 76 77 76 75 74
74 78,74 76,74 74,74 72,74 74 76 75 74 73
72 78,72 76,72 74,72 72,72 72 75 74 73 72
Sampling distribution of sample means
0.30
0.25
0.20
0.15
0.10
0.05
0
72 73 74 75 76 77 78

Summary measures for the sampling distribution
N
The mean of the sample X i
72 73 73 78
Px i 1
75
N 16
The standard deviation of the sample:
X Px
N
2
Vx i 1
i
72 752 73 752 78 752 1.58
N 16
Two important points worth noting in population and sampling distributions:

Population mean and the expected value of sample means is same which is equal to 75
Variance of the population = 2.2362=5 and Variance of the sample mean = 1.582=2.5
which is lower than the population variance
Variance of sample mean (about the population mean) decreases with increase in sample size
V2
Variance of sampling distribution of mean (sqr of standard error) s2
n
So, for the sampling distribution of mean: E(x) = , and s2 = 2/n.

What about the shape of this distribution?
Central limit theorem
&2 the sampling distribution of the means of all

possible samples of size n generated from the population will be approximately normally
distributed
d
As sample size gets large (typically > 30)

Sampling distribution becomes almost normal regardless of shape of population

Question
Suppose the standard deviation of a normal population is known to be 10 and the mean is
hypothesized to be 8. Suppose a sample size of 100 is considered. What is the range of sample
means that allows the hypothesis to be accepted at a level of significance of 0.05?
A. Between -11.60 and 27.60
B. Between 6.04 and 9.96
C. Between 6.355 and 9.645
D. Between -8.45 and 24.45
Solution
B.
d-1.96 and 1.96
y- 8) / (10 / Sqrt (100)) and -1.96 <=z<=1.96
which implies that the sample mean X must be between 6.04 and 9.96

Agenda
Review of Statistics
Statistical Inference
Hypothesis Testing
Type I Errors, Type II Errors, p-value (1 hr)
One-Tailed Test, Two-Tailed Test (2 hrs)
Statistical inference
Statistical Inference is the process of drawing a relationship between a population and a sample
drawn from that population
Estimation & Hypothesis Testing: (Branches of statistical inference)
Estimation: Finding out estimator values (mean, variance etc) of the sample
Hypothesis testing: Judge whether the hypothesis made is reliable or not based on the sample estimators

Hypothesis testing
A statistical hypothesis test is a method of making statistical decisions from and about
experimental data
Null-hypothesis testing answers the question:
How well the findings fit the possibility that chance factors alone might be responsible"
Example: Does your score of 6/10 imply that I am a good teacher???
There are five ingredients to any statistical test:

Null Hypothesis
Alternate Hypothesis
Test Statistic
Rejection / Critical Region
Conclusion
Properties of point estimators
Linearity
Unbiasedness
Minimum Variance
Efficiency
Best Linear unbiased estimator (BLUE)
Consistency
Unbiased estimator: One or more values of an estimator is equal to the true value of a parameter
Efficient estimator: Considering only the unbiased estimators of a parameter, the one which has
least variance is called the efficient estimator
Consistent estimator: The estimator which approaches the true value of its parameter as the
sample size increases

Launching a niche course for MBA students?
Christos, brand manager for a leading financial training center, wants to introduce a new niche finance
course for MBA students. He met some industry stalwarts and found that with the skills acquired by
attending such a course, the students would be very hot in job market
He meets a random sample of 100 students and discovers the following characteristics of the market
Mean household income to $20,000
Interest level in students = high
Current knowledge of students for the niche concepts = low
Christos strongly believes the course would adequately profitable in students if they have the buying
power for the course. They would be able to afford the course only if the mean household income is
greater than $19,000
Would you advice Christos to introduce the course?
What should be the hypothesis?
Hint: What is the point at which the decision changes (19,000 or 20,000)?
What about the alternate hypothesis?
What other information do you need to ensure that the right decision is arrived at?
Hint: confidence intervals / significance levels?
Hint: Is there any other factor apart from mean, which is important? How do I move from population parameters to
standard errors?
What is the risk still remaining, when you take this decision?
Hint: Type-I / II errors?
Hint: P-value
Identifying the critical sample Mean value Sampling

distribution
To reach a final decision, Christos has to make a general inference (about the population) from
the sample data
Criterion: Mean income across all households in the market area under consideration
If the mean population household income is greater than $19,000, then Christos should introduce the
product line into the new market
Christoss decision making is equivalent to either accepting or rejecting the hypothesis:
The population mean household income in the new market area is greater than $19,000
The term one-tailed signifies that all z-values that would cause Christos to reject H0, are in just
one tail of the sampling distribution
-> Population Mean
H0: d $19,000
Ha: ! $19,000

Christoss criterion for decision making
0.25
0.2
Critical Value (Xc)
0.15
0.1
0.05
$19,000
Sample mean values greater than $19,000--that is x-values on the right-hand side of the sampling
distribution centered on = $19,000--suggest that H0 may be false
More important the farther to the right x is , the stronger is the evidence against H0
Reject H0 yc
Computing the criterion value
Standard deviation for the sample of 100 households is $4,000. The standard error of the mean
(sx) is given by:
s
sx $400
n
Critical mean household income xc through the following two steps:
Determine the critical z-value, zc. For D = 0.05:
zc = 1.645
Substitute the values of zc, s, and P (under the assumption that H0 is "just" true )
xc = P + zcs = $19,658
Decision Rule:
If the sample mean household income is greater than $19,658, reject the null hypothesis and
introduce the new course

Test statistic
The value of the test statistic is simply the z-value corresponding to = $20,000
xP 0.25
Z 2.5
sx
0.2

Xc P Z c *V 0.15
0.1
In this case, since the observed sample

0.05
statistic (20,000) is greater than the
critical value (19,658), so the null 0
hypothesis is rejected => = $19,000 X = $20,000
There is a significant difference in the
hypothesized population parameter
and the observed sample statistic => Do not Reject H0 Reject H0
Mean income > 19,000 => Xc $19,658

Launch the course Zc 1.645
Errors in estimation
Please note: Actual

H0 is True H0 is False
You are inferring for a population, based only on a sample
Inference
This is no proof that your decision is correct
Correct Decision Type-II Error
Its just a hypothesis
H0 is True Confidence P(Type-II Error)
There is still a chance that your inference is wrong Level = 1-
How do I quantify the prob. of error in inference? Type-I Error
H0 is False Significance Power=1-
Type I and Type II Errors Level =
Type I error occurs if the null hypothesis is rejected

when it is true
Type II error occurs if the null hypothesis is not rejected
when it is false
Significance Level
D -> Significance level
the upper-bound probability of a Type I error
1 - D ->confidence level
the complement of significance level

P-Value Actual significance level
P-value
The probability of obtaining an observed value of 0.25
x (from the sample) is as high as $20,000 or more
when actual populations mean (P) is only 0.2
$19,000 = 0.00621

0.15
This value is sometimes called the actual
significance level, or the p-value
0.1
Calculated probability of rejecting the null
hypothesis (H0) when that hypothesis (H0) is true
0.05
(Type I error)
The actual significance level of 0.00621 0
in this case means that the odds are less than p-value = 0.00621
= $19,000
62 out of 10,000 that the sample mean
income of $20,000 would have occurred
entirely due to chance Do not Reject H0 Reject H0
(when the population mean income
is $19,000)
Decision Rule: smallest level of significance for which the null hypothesis can be rejected
Question
Which of the following statements regarding hypothesis testing is/are true?

I. If the significance level is more than the p-value, the null hypothesis is rejected
II. A decrease in the level of type I error causes a decrease in Type II error as well
III. Type I error is the error when a false null hypothesis is not rejected
IV. Systematic error is caused by non-random variations due to unknown sources
A. I only
B. I and IV
C. I and III
D. I, II and IV

Solution
A.
II is false. A decrease in the level of Type-I error increases the Type-II error
III is false. Type I error is when a true null hypothesis is rejected
Question
Consider an exam taken by 15,000 FRM candidates. Mean score for the exam was 64 for all the
6,400 candidates who studied at least 250 hours in the preparation of the exam. Assuming a
population standard deviation of 16. What would be 99% confidence interval for the mean score
'0.005 =2.575)
I. 64 0.52
II. 64 0.80
III. 64 5.15
IV. 64 8.05

Solution
A.
d
= 64 2.575 * 16/sqrt(6400)
= 64 0.515
Some variations in the Z-test I
What if Christos surveyed the market and found that the student behavior is estimated to be:
They would found the training too expensive if their household income is < US$19,000 and hence would
not have the buying power for the course?
They would perceive the training to be of inferior quality, if their household income is > US$19,000 and
hence not buy the training?
How would the decision criteria change? What should be the testing strategy?
Hint:
From the question wording infer: Two tailed testing
Appropriately modify the significance value and other parameters
h-test
Appropriate change in the decision making and testing process:
Students will not attend the course if:
The household income >$19,000 and the students perceive the course to be inferior
The household income is <$19,000
This becomes a two tailed test wherein the student will join the course only when the household lie
between a particular boundary. i.e. the household income should be neither very high neither very low

Two Tailed test
Now the test is modified to two-tailed

test, which signifies that all z-values that 0.25
would cause Christos to reject H0, are in
both the tails of the sampling distribution 0.2
P -> Population Mean = 0.025 = 0.025

0.15
H0: P = $19,000
Ha: P 0.1
Since we are checking for significance

0.05
difference on both the ends, so its a two
tailed test 0
The lower boundary =
P ZD / 2 * V 19,000 1.95 * 400 $18,216
P ZD / 2 *V 19,000 1.95 * 400 $19,784
Do not
Reject H0 Reject Reject H0
Conclusion: If the household income lies H0
between $18,216 and $19,784 then the
student will attend the course at 95%
confidence
Some variations in the Z-test II
What if the sample had not been large enough?? For example, if Christos had met only 25
students, then what?
Conduct t-Test when sample size is small
Let the sample size, n = 25, X = $20,000, s = $8,000
From the t-table tc = 1.71 for D = 0.05 and d.f. = 24
Decision rule: Reject H0 if t ! 1.7l.
Points to observe:
You could not launch the course.. Why?
Hint: Is it because of T-Test?
NO!
Its because the sample size is small =>
Less value of n =>
Higher standard error =>
Lower confidence in rejecting the hypothesis =>
Almost akin to not taking a decision (hence not launching the product)
T- Test to be conducted, when sample size (n) is small (Typically<30) Degrees of freedom = (n-1)

Some variations in the Z-test III
Christos has surveyed the market and decided to launch the course. He has two markets in mind,
where he can launch the course (and hence conducts the survey):
Chicago
Mean Income (Sample Size = 100): $19,500
Standard Deviation(s1): $300
New York:
Mean Income (Sample Size = 100): $18,500
Standard Deviation(s2): $400
What if Christos wants to launch the course in one of the markets?
What would be the decision criteria? What should be the testing strategy?
hc n
tc n = 0
The only treatment to be made different is that the standard error has to be calculated as:
2 (n1 1) s12 (n2 1) s22

s
n1 n2 2
The rest of the treatment remains the same as one mean hypothesis.
t
Hint: Linear combination of normal distributions is a normal distribution
Some variations in the Z-test III
/c n h
Then the test can be appropriately modified 0.25
as a two-h,
0.2
h -> Population Mean
H0h = $0 = 0.025
0.15
= 0.025
Hah
Since we are checking for significance 0.1
difference on both the ends, so its a two

0.05
tailed test
This test can have another variant, if we 0
check for a significant difference c-n =

between two means as a particular value, $1,000
in which the hypothesis would be Do not
modified as: Reject H0 Reject Reject H0
h -> Population Mean H0
H0h = $1,000
Hah

dz
www.edupristine.com
Quantitative Analysis - III
Agenda
Hypothesis testing Part 2

Correlation and Copulas
Linear Regression with One Regressor
Regression with a Single Regressor
Linear Regression with Multiple Regressors
Hypothesis testing Part 2

Chi-Square and F-Tests

Hypothesis tests for variances
Hypothesis Test of Variances
Test for Test for

Single Population Variance Two Population Variances
Example Hypothesis Chi-Square Test Example Hypothesis F-test Statistic

Statistic
H0: 2 = 02 H0: 12 22 = 0 s12
(n 1) s 2
FD ,ndf ,ddf
HA: 2 02 FD2 ,( n 1) HA: 12 22 s22
V 2
0
In testing for variances, there are two different tests, because sum of two chi-squares is not a
chi-square
Test for single population variance
Single population variance test has 3

different forms: D/2
Two Tailed Test: D/2
H0: 2 = 02
HA: 2 02
Lower Tail test: D

H0: 2 t 02
HA: 2 < 02
Upper Tail Test
H0: 2 02
HA: 2 > 02
D

Chi-square test statistic
The chi-squared test statistic for a Single

Population Variance is:
(n 1)s 2
F 2
Where
F2 = standardized chi-square variable
n = sample size
s2 = sample variance
2 = hypothesized variance
Finding the critical value
The critical value, F2D , is found from the chi-square table:
Upper tail test: H0: 2 02

H A : 2 > 02
F2
Do not reject H0 Reject H0

F2 D

Lower tail or two tailed Chi-square tests
Lower tail test: H0: 2 t 02 Two tail test: H0: 2 = 02

H A : 2 < 02 HA: 2 02
D D/2
D/2
F2 F2
Reject Do not reject H0 Reject Do not Reject

reject H0
F21-D F21-D/2 F2D/2
Question
A brand new air conditioner from Carrier must maintain the selected temperature with little
variation. Specifications from the company call for a standard deviation of not more than 2
degrees (or variance of 4 degrees2). A sample of 4 air conditioners is tested and yields a sample
variance of s2 = 6. Test to see whether the standard deviation specification is exceeded.
Use D = 0.05

Solution
The chi-square table to find the critical value:

F2D = 7.815 (D = 0.05 and (4 1) = 3 d.f.)
The test statistic is:

(n 1)s 2 (4 1)6
F2 4.5
2 4
Since 4.5 < 7.815, do not reject H0
D = 0.05
There is not significant evidence at

the D = 0.05 level that the F2
standard deviation specification is
exceeded Do not reject H0 Reject H0
F2D
F2D= 7.815
F-test for difference in two population variances
Two population variance test has 3 different forms:

D/2
Two Tailed Test:
D/2
H0: 12 22 = 0
HA: 12 22
Lower Tail test:

D
H0: 12 22 t 0
HA: 12 22 < 0
Upper Tail Test
H0: 12 22
HA: 12 22 > 0
D

F-test for difference in two population variances (cont.)
The F test statistic is:
s12
F
s22
Although its not a rule but always take the higher sample variance as the numerator and lower
sample variance as the denominator
s12 = Variance of Sample 1

(n1 1) = numerator degrees of freedom
s 22 = Variance of Sample 2
(n2 1) = denominator degrees of freedom
EduPristine
E FRM-I (2016) 138
The F distribution
The F critical value is found from the F table

The are two appropriate degrees of freedom
One for numerator and other for denominator
In the F table:
Numerator degrees of freedom determine the row
Denominator degrees of freedom determine the column
s12
F where df1 = n1 1 ; df2 = n2 1
s 22

Finding the critical value
H 0: 12 22 t 0 H 0: 12 22 = 0
H A : 12 22 < 0 HA: 12 22
H0: 12 22
H A : 12 22 > 0
D D/2
0 F 0 F
Do not reject H0 Reject H0 Do not reject H0 Reject H0

FD F D/2
Rejection region for s12 Rejection region for s12

a one-tail test is: F ! FD a two-tailed test is: F ! FD / 2
s 22 s 22
(When the larger sample variance in the numerator)
F-test: Question
You are a financial analyst for a brokerage firm. You want to compare dividend yields between
stocks listed on the Dow30 & EURO STOXX 50. You collect the following data:
Dow30 hZK^dKyy
Number 30 50
Mean 3.27 2.53
Std dev 1.5 1.4
Is there a difference in the variances between the Dow30 & EURO STOXX 50 at the D = 0.05 level?

F-test: Solution
Form the hypothesis test:

H021 22 = 0 (there is no difference between variances)
HA21 22
Find the F critical value for D = 0.05

Numerator:
df1 = n1 1 = 30 1 = 29
Denominator:
df2 = n2 1 = 50 1 = 49
F0.05/2, 29, 49 = 1.881
F-test: Solution (cont.)
The test statistic is: H 0: 12 22 = 0

HA: 12 22
s12 1.50 2
F 1.148
s22 1.40 2 D/2 = 0.025
0 F
Do not reject H0 Reject H0

F D/2 = 1.881
F = 1.148 is not greater than the critical F value of 1.881, so we do not reject H0
Conclusion: There is no evidence of a difference in variances at D = 0.05

Chebyshevs inequality
Chebyshev's inequality says that at least 1 - 1/k2 of the distribution's values are within k standard
deviations of the mean.
Where k is any positive real number greater than 1
Correlations & Copulas

Definition of Correlation
Monitoring Correlation
EduPristine
d i i FRM-I (2016)
(20 6) 147
Linear regression with one Regressor
Basic concept of regression
Scatter plot
Visual representation of relationship between dependent and independent variable
Dependent/Independent variables
The main goal of regression analysis is to measure how the changes in one variable known as dependent
variable can be explained by the changes in one or more other variables called the independent variables
Linear relationship
There is linear relationship between dependent and independent variables
Parameters
Slope and intercept term of the regression equation are know as parameters
Error term
Difference between expected value from the equation and actual value. This is mainly because the
dependence of the dependent variable on other factors is not taken into account

The million dollar question
Marks obtained in the test
Middle
Hours Mumbai Delhi Chennai Kolkata Bangalore Pune Hyderabad Online Singapore
East
10 20 7 5 13 10 11 14 9 7 12
20 8 24 34 24 16 19 20 20 25 12
30 19 8 16 37 62 29 33 25 36 30
40 67 31 44 43 32 19 38 27 49 35
50 36 46 78 57 36 82 55 33 53 41
60 67 54 90 58 23 45 62 67 58 78
70 56 68 93 71 76 72 68 81 70 57
80 81 89 7 6 45 68 83 58 90 9
If I study for five more hours will it actually increase my marks?
The population
Hours of study vs. Marks
120
100
80
Marks in Test
60
40
20
0
0 10 20 30 40 50 60 70 80 90
Hours of Study
Can we draw a trend line to predict this relationship?

Introduction to regression analysis
Regression analysis is used to:

Predict the value of a dependent variable based on the value of at least one independent variable
Explain the impact of changes in an independent variable on the dependent variable
Dependent variable: the variable we wish to explain
Independent variable: the variable used to explain the dependent variable
Simple linear regression model
Only one independent variable, x

Relationship between x and y is described by a linear function
Changes in y are assumed to be caused by changes in x

Types of regression models
Negative Linear Relationship Relationship NOT Linear
Positive Linear Relationship No Relationship
Population linear regression
z
Y E 0 E1 X u
Observed Value of
zyi
Slope = 1
Wsz Random Error for this
yi x value
ui
Intercept = 0
xi x
Mean marks for hours of study
Individual persons marks

Population regression function
Population Random Error

Dependent Population Slope Independent term, or
Variable y intercept Coefficient Variable residual
Y 0 1 X u
Linear component Random Error
component
But can we actually get this equation?

If yes what all information we will need?
Information that we actually have
Hours Mumbai
10 20
20 8
30 19
40 67
50 36
60 67
70 56
80 81

Sample regression function
z
y b 0 b1 x e
Observed Value of
zyi
ei
Slope = 1
Wsz Random Error for
yi this x value
Intercept = 0
xi x
Sample regression function
Estimated Estimate of the Estimate of the Independent

(or predicted) regression regression slope variable
intercept Error term
y value
yi b 0 b1x e
Notice the similarity with the Population Regression Function

Can we do something of the error term?

The error term (residual)
Represents the influence of all the variable which we have not accounted for in the equation
It represents the difference between the actual y values as compared the predicted y values
from the Sample Regression Line
Wouldnt it be good if we were able to reduce this error term?
What are we trying to achieve by Sample Regression?
Our objective
Y 0 1 X u
To Predict PRL from SRL
y i b 0 b1x

One method to find b0 and b1
Method of Ordinary Least Squares (OLS)

b0 and b1 are obtained by finding the values of b0 and b1 that minimize the sum of the
squared residuals
e 2
(y y) 2
(y (b 0 b1 x)) 2
Are there any advantages of minimizing the squared errors?

Why dont we take the sum?
Why dont we take absolute values instead?
OLS regression properties
The sum of the residuals from the least squares regression line is 0
( y y ) 0
The sum of the squared residuals is a minimum

Minimize ( ( y y ) 2 )
The simple regression line always passes through the mean of the y variable and the mean of the
x variable
d0 1

The least squares equation
The formulas for b1 and b0 are:
b1
( x x )( y y )
(x x) 2
Algebraic equivalent: And
xy n
x y b0 y b1 x
b1
x
2
( x)
2
Interpretation of the Slope and the Intercept
b0 is the estimated average value of y when the value of x is zero. More often than not it does not
have a physical interpretation
b1 is the estimated change in the average value of y as a result of a one-unit change in x
y

Y b0 b1 X
slope of the line(b1)
b0
x

Assumptions underlying linear regression
The underlying relationship between the X variable and the Y variable is linear
For a given value of Xi the sum of error terms is equal to 0
The error term is uncorrelated with the explanatory variable X
Error values are normally distributed for any given value of X
The probability distribution of the errors for a given Xi is normal
The probability distribution of the errors for different Xi has constant variance (homoscedacity)
Error values u for given Xi are statistically independent, their covariance is zero
Distribution of error term for

a given value of X
X
Once we fulfill these assumptions in Linear Regression , we are able to estimate the variance and
standard errors of b0 and b1 and this has been possible because of the properties of OLS method
Explained and unexplained variation
y
yi RSS = Residual sum of squares
y
TSS = Total sum RSS = (yi - yi )2
of squares _
TSS = (yi - y)2

y _
_ ESS = (yi - y)2
_
y y
ESS = Explained Sum of squares
x
yi

Explained and unexplained variation (cont.)
TSS = Total sum of squares

Measures the variation of the yi values around their mean y
RSS = Residual sum of squares
Variation attributable to factors other than the relationship between x and y
ESS = Explained sum of squares
Explained variation attributable to the relationship between x and y
Explained and unexplained variation (cont.)
Total variation is made up of two parts:
TSS RSS ESS

Sum of Squares Sum of Squares
Total sum of Squares Error/Residual Sum of Regression/Explained Sum of
Squares Squares
TSS ( y y) 2
RSS ( y y ) 2
ESS ( y y ) 2
Where:
y = Average value of the dependent variable
y = Observed values of the dependent variable
y = Estimated value of y for the given x value

Coefficient of determination, R2
The coefficient of determination is the portion of the total variation in the dependent variable
that is explained by variation in the independent variable
The coefficient of determination is also called R-squared and is denoted as R2
2 SSR
R
SST
where
0 d R2 d1
Coefficient of determination, R2 (cont.)
Coefficient of determination
SSR sum of squares explained by regression
R2
SST total sum of squares
Note: In the single independent variable case, the coefficient of determination is
R2 r2
Where:
R2 = Coefficient of determination
r = Simple correlation coefficient

Examples of approximate R2 values
R2 = 1
Perfect linear relationship between x and y:
R2 = 1 x
100% of the variation in y is explained by variation in x
R2 = +1 x
Examples of approximate R2 values (cont.)
0 < R2 < 1
Weaker linear relationship between x and y:
x
Some but not all of the variation in y is explained by
variation in x
y

Examples of approximate R2 values (cont.)
y
R2 = 0
No linear relationship between x and y:
dz
(None of the variation in y is explained by variation in x)
R2 = 0 x
Question
Paul Graham, FRM is analyzing the sales growth of a baby product launched three years ago by a
regional company. He assesses that three factors contribute heavily towards the growth and
comes up with the following results:
zy1 y2 y3
Sum of Squared Regression [SSR] = 869.76
Sum of Squared Errors [SSE] = 22.12
Determine what proportion of sales growth is explained by the regression results.
A. 0.36
B. 0.975
C. 0.64
D. 0.55
What should be the answer if Instead of Sum of Square Errors, we had Standard Error of Estimate
(SEE)????

Question
The result of the linear regression is: Y = 0.10 - 0.50 X with a correlation coefficient R = (-0.90).
The fraction of the variance of Y attributable to X is equal to:
A. (-0.90)
B.
C.
D. (-0.50)
Solution
C.
R-squared is the square of the correlation coefficient and measures the fraction of the variance of Y that is
attributable to X.R2 = (-0.90)2 = 0.81

Correlation coefficient
The ) measures the strength of the association between

the variables
The sample correlation coefficient r
the linear relationship in the sample observations
&
Unit free
Range between -1 and 1
The closer to -1, the stronger the negative linear relationship
The closer to 1, the stronger the positive linear relationship
The closer to 0, the weaker the linear relationship
Examples of approximate r values
y y y
r = -1 x r = -0.6 x r=0 x
y y
r = +.3 x r = +1 x

Calculating the correlation coefficient
Sample correlation coefficient:
r
( x x )( y y )
[ ( x x ) ][ ( y y )
2 2
]
or the algebraic equivalent:
n xy x y
r
[n( x 2 ) ( x )2 ][n( y 2 ) ( y )2 ]
Where:
r = Sample correlation coefficient
n = Sample size
x = Value of the independent variable
y = Value of the dependent variable
Regression with a single Regressor:

Hypothesis testing and confidence intervals

Hypothesis testing: Two variable model
How do we know whether the values of b0 and b1 that we have found are actually meaningful?
It is actually possible that our sample was a random sample and it has given us a totally wrong
regression line?
We do know a lot about the sample error term e but what do we know about the error terms
u of the Population Regression Function?
How do we proceed from here?
Standard Error of Estimate
The standard deviation of the variation of observations around the regression line is estimated
by:
RSS
su
n k 1
Where:
RSS = Residual Sum of Squares (summation of e2)
n = Sample size
k = number of independent variables in the model
Standard Error of Estimate (SEE) is another name of Standard Error of regression

The Standard Deviation of the intercept
Xi su
2
s bo
n (x x) 2
su su
s b1
(x x) 2
( x) 2
x 2

n
Where:
sb1 = Estimate of the standard error of the least squares slope
RSS
su n2 = Sample standard error of the estimate
Comparing standard errors
Variation of observed y values from the Variation in the slope of regression lines from
regression line different possible samples
y y
small s u x small sb1 x
y y
large s u x large sb1 x

Inference about the slope: t-Test
t-test for a population slope

Is there a linear relationship between x and y?
Null and alternative hypotheses
,1 = 0 (no linear relationship)
,1
Test statistic
t b1 1
sb1
d.f. n2
Where:
b1 = Sample regression slope coefficient
,
sb1 = Estimator of the standard error of the slope
Question
A risk analyst performs a simple linear regression on return data comprising three variables
evolving in time and obtains, amongst others, the following statistics:
Coefficients Standard Error t-Statistic

Intercept 49.4 2.85 17.53
ys -38.79 138.93 -0.28
ys -431.75 170.50 -2.53
ys -70.40 121.06 -0.58
Based on these data at a 95% confidence level, the analyst should conclude that:
A. The intercept and X Variable 2 are statistically significant
B. X Variable 1 and X Variable 3 are statistically significant
C. X Variable 1, X Variable 2 and X Variable 3 are all statistically not significant
D. More information is required, such as the corresponding p-values, before any meaningful deductions may be
made

Solution
A.
A is correct. (Relatively) small standard errors and high t-stats are one indication of indicate statistical
significance
B is incorrect. (Relatively) large standard errors and low t-stats are one indication of indicate statistical
significance
C is incorrect. Negative t-stats are not an indication of statistical insignificance
D is incorrect. The p-values are redundant information if the t-stat is provided. That is, the p-values tell one
nothing more than the t-stats do
Homoscedasticity & Heteroscedasticity
One of the important assumptions of Linear regression is that the variance of the error term is
same across all observations. This type of regression is called homoscedastic regression
When the requirement of a constant variance is violated, we have a condition of
heteroscedasticity
Error u
Predicted y
We can diagnose heteroscedasticity by plotting the residual against the predicted y

Unconditional and Conditional Heteroscedasticity
Unconditional heteroskedasticity is when heteroskedasticity is not related with value of the

independent variable
Conditional heteroskedasticity is when it is related with a level of independent variable
Effects of Heteroscedasticity
Regression coefficients are not affected
Standard errors are usually unreliable
Hypothesis testing of regression coefficient becomes insignificant
The Guass-Markov Theorem
This theorem says that if the assumptions of linear regression are true, then the OLS estimators
have following properties:
OLS estimated coefficients have minimum variance compared to other methods of estimation
OLS estimated coefficients are based on linear functions
OLS estimated coefficients are unbiased
OLS estimate of variance of errors is unbiased

Linear Regression with Multiple Regressors
Multiple Regression
Using more than one explanatory variable in a regression model

Y = b0 1X1 2X2 3X3 I
Omitted variable bias

The biasness incurred due to omission of one or more explanatory variable from the model.
Omitted variable bias occurs when two conditions are met:

Omitted variables are correlated with the independent variable
Variables that are not accounted for in the model but affect the dependent variable

Assumptions of Multiple Regression Model
There exists a linear relationship between the dependent and independent variables
The expected value of the error term, conditional on the independent variables is zero
The error terms are homoskedastic, i.e. the variance of the error terms is constant for all the
observations
The expected value of the product of error terms is always zero, which implies that the error
terms are uncorrelated with each other
The error term is normally distributed
The independent variables doesn't have any linear relationships between each other
Multiple Regression Basics
In simple linear regression, the dependent variable was assumed to be dependent on only one
variable (independent variable)
In General Multiple Linear Regression model, the dependent variable derive sits value from two or
more than two variable
General Multiple Linear Regression model take the following form:
Yi b0 b1 X 1i b2 X 2i ......... bk X ki H i
Where:
Yi = ith observation of dependent variable Y
Xki = ith observation of kth independent variable X
b0 = intercept term
bk = slope coefficient of kth independent variable
i = error term of ith observation
n = number of observations
k = total number of independent variables

Estimated Regression Equation
As we calculated the intercept and the slope coefficient in case of simple linear regression by
minimizing the sum of squared errors, similarly we estimate the intercept and slope coefficient in
multiple linear regression
n
Sum of Squared Errors H
i 1
i
2
is minimized and the slope coefficient is estimated.
The resultant estimated equation becomes:

Yi b0 b1 X 1i b2 X 2i ......... bk X ki
Now the error in the ith observation can be written as:

Hi Yi Yi Yi b0 b1 X 1i b2 X 2i ......... bk X ki

Interpreting the Estimated Regression Equation
Intercept Term (b0): It's the value of dependent variable when the value of all independent
variables become zero
b0 Value of Y
when X 1 X2 ....... X k 0
Slope coefficient (bk): It's the change in the dependent variable from a unit change in the
corresponding independent (Xk) variable keeping all other independent variables constant
In reality when the value of the independent variable changes by one unit, the change in the dependent
variable is not equal to the slope coefficient but depends on the correlation among the independent
variables as well
Therefore, the slope coefficient are called partial slope coefficients as well

Coefficient of determination (R2) and Adjusted R2
Coefficient of determination(R2) can also be used to test the significance of the coefficients
collectively apart from using F-test
SST - SSE RSS Sum of Squares explained by regression
R2
SST SST Total Sum of Squares
The drawback of using Coefficient of determination is that the value of the coefficient of
determination always increases as the number of independent variables are increased even if the
marginal contribution of the incoming variable is statistically insignificant
To take care of the above drawback, coefficient of determination is adjusted for the number of
independent variables taken. This adjusted measure of coefficient of determination is called
adjusted R2
Adjusted R2 is given by the following formula:
Where: Ra2
n 1
1 2

u 1 R
n k 1
n = Number of Observations
k = Number of Independent Variables
Ra2= Adjusted R2
EduPristine
E FRM-I (2016) 198
Multicollinearity
Multicollinearity refers to the condition when two or more independent variables in a multiple
regression are highly correlated with each other
There are two types of multicollinearity:
Perfect multicollinearity
One independent variable has perfect correlation with other independent variables
In this case, coefficients of the model cannot be determined
Imperfect multicollinearity
One or more independent variable is correlated to some degree (not perfect) to other independent variables
In this case, coefficients can be determined but might result in errors during test of statistical significance
(chances of Type II error)
Effects of Multicollinearity
There is greater probability that we will incorrectly conclude that a variable is not statistically significant (A type II
error)

dz
www.edupristine.com
Quantitative Analysis-IV
Agenda
Hypothesis Tests and Confidence Intervals in Multiple Regression

Modeling and Forecasting Trend
Characterizing Cycles
Modeling Cycles: MA, AR and ARMA Models
Estimating volatilities and correlations
Simulation Modeling
Hypothesis Tests and Confidence Intervals in Multiple Regression

Hypothesis Testing of Coefficients
The values of the slope coefficients doesn't tell anything about their significance in explaining the
dependent variable
Even an unrelated variable when regressed would give some value of slope coefficients
To exclude the cases where the independent variables doesn't significantly explain the dependent
variable, we need the hypothesis testing of the coefficients for checking whether they contribute
in explaining the dependent variable significantly or not
The t-statistic is used to check the significance of the coefficients
The t-statistic used for the hypothesis testing is same as used in the hypothesis testing of
coefficient of simple linear regression
Following are the hypothesis and alternative hypothesis to check the statistical significance of bk:
Hypothesis H0: bk =0
Alternative Hypothesis (Ha): bk
The t-statistic
of (n-k-1) degrees of freedom for the hypothesis testing of the coefficient bk
bk bk
t
s
bj
If the value of t-statistic lies within the confidence interval, H0 can't be rejected
Confidence Interval for the Population Value
The confidence interval for a regression coefficient is given by:

b j r (tc u s )
bj
Where:
tc is the critical t-value, and

sb is the standard error
j

Predicted Dependent Variable
The regression equation can be used for making predictions about the dependent variable by
using forecasted values of the independent variables

Yi b0 b1 X 1i b2 X 2i ......... bk X ki
Where:
Yi is the predicted value for the dependent variable
b is the estimated partial slope for the ith independent variable

i
is the forecasted ith value for the nth independent variable

X ni
F-Statistic
An F-test explains how well the dependent variable is explained by the independent variables
collectively
In case of multiple independent variable, F-test tells us whether a single variable explains a
significant part of the variation in dependent variable or all the independent variables explain the
variability collectively
Suppose Edward now wants to check whether the household income and expenses collectively
explains the variation in his pocket money
The following hypothesis can be formed:
H0: HI=HE=0 versus Ha,/,
RSS
MSR k
F-statistic is given as: F
MSE SSE
Where: n k 1
MSR: Mean Regression sum of squares

MSE: Mean Squared Error
n: Number of observations
k: Number of independent variables
EduPristine
E FRM-I (2016) 207
F-Statistic (cont.)
Decision rule for F-test: Reject H0 if the F-statistic > Fc (Critical Value)
The numerator of F-statistic has degrees of freedom of "k" and the denominator has the degrees
of freedom of "n-k-1"
If H0 is rejected then at least one out of two independent variable is significantly different that
zero.
This implies that at least one out of household income(independent variable) or household
expenses(independent variable) explains the variation in the pocket money of Edward
F-test is always a single tailed test while testing the hypothesis

that the coefficients are simultaneously equal to zero
Restricted vs. Unrestricted Least Square Models
Restricted regression is one in which coefficients of some independent variable are assumed to be
zero
Therefore, any regression with an omitted variable is restricted regression (as coefficients of an
omitted variable are assumed to be zero)
Regression in which all independent variables are included is unrestricted regression

Model Misspecification
There are two broad categories of model misspecification:-

1. The functional form can be misspecified
Important variables are omitted
Variables should be transformed
Data is improperly pooled
2. Independent variables are correlated with the error term in the time series model
A lagged dependent variable is used as an independent variable
Function of dependent variable is used as an independent variable
Independent variable are measured with error.
Effects of Model Misspecification
Model Misspecification
Biased and Inconsistent Regression Coefficients
Unreliable Hypothesis Testing and Inaccurate Predictions

Modeling and Forecasting Trend
Mean Squared Error (MSE)

Mean Squared Error (MSE)
AIC & SIC
3.5
3 [SERIES
Penalty Factors
2.5 NAME],
2 [SERIES k/T is number of parameters

NAME],
1.5 estimated per sample
[SERIES
1 NAME], observation
0.5
0.05 0.1 0.15 0.2 0.25
k/T

Selecting Forecasting Models using the Akaike and Schwarz
Criteria
Which is the good model selection criteria?
Evaluate the model selection criteria in terms of a key property consistency
A model selection criteria is consistent if the following conditions are met

When the true model i.e. DGP is among the models considered, the probability of selecting the true DGP
approaches 1 as the sample size gets large
When the true model is not among the models considered, the probability of selecting the best approximation
to the true DGP approaches 1 as the sample size gets large
AIC is although inconsistent but is asymptotically efficient, whereas the SIC is NOT
Characterizing Cycles

Definitions
Covariance Stationary A series which have its mean and its variance structure stable over time
Autocovariance function It is a function used to quantify stability of the covariance structure
Covariance Stationary
The requirements for a series to covariance stationary

1. The mean of the series to be stable over time
2. The covariance structure of the series to be stable over time
3. The variance of the series the autocovariance at displacement 0, (0) be finite
Covariance stationary places no restrictions on other aspects of the distribution of the series

Autocorrelation Function
Analyse the qualitative shape of partial autocorrelations as a function of Displacement.

Autocorrelation function that
displays gradual one-sided damping

displays non damping

displays damped oscillation

displays sharp cutoff

White Noise
Application (Example)
Characterizing Canadian Employment Dynamics from 1962 to 1993 quarterly

It evolves in a slow persistent fashion high
in business cycles booms and low in
recessions
The series displays no trend and no

seasonality, as it is seasonally adjusted,
it appear highly serially correlated
(Q-statistic) When testing the null hypothesis of

white noise for values of m ranging
from 1 through 12,
Correlogram Analysis
Found that p-value is consistently 0

to four decimal places
So null hypothesis of white noise is

decisively rejected.

Application (Example)
Canadian employment index: Sample autocorrelation and partial autocorrelation functions,

- 2 standard error bands
The sample partial autocorrelations are

The sample autocorrelations are very large relative to their standard errors at
large relative to their standard errors and first but are statistically negligible
display slow one-sided decay beyond displacement 2.
Its clear that employment has a strong cyclical component; all diagnostics rejects the white
noise hypothesis and such patterns in the series shows that it might be useful for
suggesting candidate forecasting models
Modeling Cycles: MA, AR and ARMA Models

Modeling Cycle
When building forecasting models, dont pretend that the model is true. Instead, be aware that the
model is approximating a more complex reality.
There are three approximations:
Moving Averages (MA) models
Autoregressive (AR) models
Autoregressive moving average (ARMA) models
These models vary in their specifics and have different strengths in capturing different sorts of
autocorrelation behaviour.

Moving Averages (MA) models: Population autocorrelation

function
The key feature here is the sharp cut off in the autocorrelation function. All the autocorrelation are
0 beyond displacement 1, the order of the MA process.
When the MA process is inverted, it can express the current value of the series not in terms of a
current shock and lagged shock but rather in terms of a current shock and lagged values of the
series

Moving Averages (MA) models: Population Partial

Autoregressive (AR) models: Population autocorrelation

function

Autoregressive (AR) models: Population Partial Autocorrelation
Function
The partial autocorrelation function for the AR(1) process cuts off abruptly
Partial autocorrelation are just the last coefficients in a sequence of successively longer population
autoregressive. The 1st partial correlation is just the autoregressive coefficient, and coefficients on
all longer lags are 0
Autoregressive Moving Average (ARMA) models

Estimating Volatilities and Correlations
Estimation of Volatility
Let xi be the continuously compounded return during day i (between the end of day
i-1 and end of day I)
>n be the volatility of the return on day n as estimated at the end of day n-1
Variance estimate for next day is usually calculated as:
variance = average squared deviation from average return over last n days
x
n
2
i x
i 1
Variance
n 1
Mean of returns (x-bar) is usually zero, especially if returns are over short-time period
(say, daily returns). In that case, variance estimate for next day is nothing but simple average
(equally weighted average)
n of previous n days squared returns
x
2
i
i 1
Variance
n 1
What if the volatility is dependent on the values of volatility observed in the recent past?
What if they also depend on the latest returns?

EWMA Model
In an exponentially weighted moving average model, the weights assigned to the u2 decline
exponentially as we move back through time
This leads to: V n2 OV n2 1 (1 O )u n21
Apply the recursive relationship:
V n2 O >OV n2 2 (1 O )u n2 2 @ (1 O )u n21

V n2 > @
(1 O ) u n21 Ou n2 2 O2V n2 2
Hence we have
m
V n2 (1 O ) Oi 1u n2i OmV n2 m
i 1
Variance estimate for next day (n) is given by (1-
previous variance estimate
Risk-metrics (by JP Morgan) assumes a Lambda of 0.94
EWMA Model (cont.)
Since returns are squared, their direction is not considered. Only the magnitude is considered
In EWMA, we simply need to store 2 data points: latest return & latest volatility estimate
Consider the equation: V t 1 (1 0.94) P t 0.94V t
2 2 2
In this equation, variance for time t was also an estimate. So we can substitute for it as follows:
V t21 (1 0.94) Pt2 0.94>(1 0.94) Pt21 0.94V t21 @
V t21 0.06 * Pt2 0.94 * 0.06 * Pt21 (0.94 * 0.94V t21 )
What are the weights for old returns and variance?

Persistence factor or even Decay Factor. ,weigIht to older data
>
previous data impacts are not allowed to persist)
,higher persistence or lower decay
Since, (1- Reactive factor

EWMA Question
Example 1: On Tuesday, return on a stock was 4%. Volatility (Std. deviation) estimate for Tuesday
&t
Variance estimate for Wednesday = (1-2
Std. Dev. = sqrt (1.9%2)=1.378%
Tuesday volatility (Std. Dev.) estimate was 1%. Actual return on Tuesday was 4%. Therefore, volatility
estimate for Wednesday is estimated upwards than Tuesday i.e. 1.378% as compared to 1%.
Notice how the volatility estimate has been revised due to high return
EWMA Question
Example 1: Continuing the previous example, volatility estimate for Wednesday was 1.378%.
Assume that actual return on Wednesday was 0%. What is the variance estimate for Thursday?
Solution: Variance estimate for Thursday = (1-2 Std. Dev. = 1.34%
In very short-term like daily returns, estimated volatility is the expected return
Since latest return of 0% was lesser than estimated volatility (and estimated return) of 1.378%, volatility for
next day is revised downward from 1.378% to 1.34%
Notice the downward revision in the estimate due to lower return

EWMA Weights Graph
= 0.7 (faster decay)

Weight of
variance terms
= 0.9 (slower decay)
Days into the past
,&
Selection of Decay Factor

We are free to select the decay factor we use in our calculation of volatility
We can use common sense in our estimation
If we expect volatility to be very unstable then we will apply a low decay factor (giving a lot of weight to
recent observations)
If we expect volatility to be constant we would apply a high decay factor (giving a more equal weight to older
observations)
Sum of Weights
One special property of the weights used in the EWMA formula is that their sum will always equal to 1
m
S (1 O ).O
i 1
i 1
Om 1
Obviously we do not have an infinite set of historical data. We just have to make sure that our data set is
large enough so that this sum is close to 1
Or alternatively we can rescale the weights so that the sum is 1
What happens to the second term when m tends to infinity?

Question: FRM Exam
hZDtD
conditional variance, which weight will be applied to the return that is 4 days old?
A. 0.000
B. 0.043
C. 0.048
D. 0.950
Solution
B.
A. Incorrect. The wrong factor has been squared. The EWMA RiskMetrics model is defined as:
t-1 - 2t-1. For t = 4, and processing r0 through the equation three times produces a factor of (1-
0.95)3*0.95 = 0.00012 for r0 when t = 4.
B. Correct. The EWMA RiskMetrics model is defined as:
0.95)*0.95^3 = 0.043 for r0 when t = 4.
C. Incorrect. The 0.95 has not been squared. The EWMA RiskMetrics model is defined as:
0.95)*0.95 = 0.048 for r0 when t = 4.
/ddtDZD
t-1 - 2t-1. For t = 4, and processing r0 through the equation three times produces a factor of 0.95 =
0.950 for r0 when t = 4.

GARCH (1,1)
GARCH stands for Generalized Autoregressive Conditional Heteroscedasticity

Heteroscedasticity means variance is changing with time.
Conditional means variance is changing conditional on latest volatility.
Autoregressive refers to positive correlation between volatility today and volatility yesterday.
(1,1) means that only the latest values of the variables.
GARCH model recognizes that variance tends to show mean reversion i.e. it gets pulled to a
long-term Volatility rate over time.
V t21 JV L DP t2 EV t2
Long-term average Volatility
GARCH (1,1) (cont.)
V t21 Z DP t2 EV t2
'sL
Since the sum of all the weights is equal to 1 we get the following equation as well:
Z
VL
1 D E

GARCH Question
Suppose a GARCH model is estimated using MLE from daily data as follows:
V t21 .000005 0.12 P t2 0.85V t2

Suppose that on a particular day t; actual return was -1% and the volatility (std. deviation)
-term
average volatility (to which the model shows reversion over-time)
Solution
Solution: In the GARCH model, 12% is the weight given to latest squared return (reactive factor).
85% is the weight given to latest variance estimate (persistence factor). Therefore,
1-0.12-0.85 = 3% is weight given to long-term average Volatility.
Therefore, 3%*VL = 0.000005 i.e. VL = 0.017%
-
s^
&'Z,/-term volatility is
negative and the model becomes mean-fleeing

Question FRM Exam
Which of the following GARCH models will take the shortest time to revert to its mean?
A. ht 2t-1 t-1
B. ht 2t-1 t-1
C. ht 2t-1 t-1
D. ht 2t-1 t-1
Solution
B.
A. Incorrect. The model that will take the shortest time to revert to its mean is the model with the lowest
/

B. Correct. The model that will take the shortest time to revert to its mean is the model with the lowest
/

C. Incorrect. The model that will take the shortest time to revert to its mean is the model with the lowest
/

D. Incorrect. The model that will take the shortest time to revert to its mean is the model with the lowest
/


Question
Suppose the long-run variance rate is 0.0002 so that the long-run volatility per day is 1.4%
V n2 . un21 0.86V n21

0.000002 013
Suppose that the current estimate of the volatility is 1.6% per day and the most recent
percentage change in the market variable is 1%. What is the new variance rate?
Solution
The new volatility is 1.53% per day

0.000002 0.13 u 0.0001 0.86 u 0.000256 0.00023516

Simulation Methods
Agenda
Monte Carlo Simulation

Variance reduction techniques
Antithetic variates
Control variates
Random number re-usage
Bootstrapping
Random number generation
Disadvantage of Simulation approach

Monte Carlo Simulation (MCS)
MCS is used to model complex problems or estimate variables for where sample size is small. It is
used in finance in situation such as:
The pricing of exotic options, where an analytical pricing formula is unavailable
Determining the effect on financial markets of substantial changes in the macroeconomic
environment
Stress-testing risk management models to determine whether they generate capital requirements
sufficient to cover losses in all situations.
Steps required to conduct MCS:

1. Generate the data according to the desired data generating process (DGP), with the errors being
drawn from some given distribution
2. Do the regression and calculate the test statistic
3. Save the test statistic or whatever parameter is of interest
4. Go back to stage 1 and repeat N times.
Cont...
In the first step, a model needs to be specified to generate the data. The model can be pure time
series or a structural model. Pure time series model are simpler whereas structural model requires a
DGP for the explanatory variables and hence is more complex.
In the second stage, estimation about the parameter of interest is made. Here various scenarios can
also be developed about the outcome of the variable.
In the last two stages, data analysis is carried out and the steps are repeated N times.

Variance reduction technique
The sampling variation in Monte Carlo is estimated by the standard error estimate (Sx)
var( x)
Sx
N
It is evident from the formula that to reduce the error by a factor of 10, N must be increased by 100.
meaning, in order to improve the efficiency, the N must be significantly higher which is both time
consuming and costly. The alternative way to reduce the sampling error is to use variance reduction
techniques. Two most widely used variance reduction techniques are
1. Antithetic variates
2. Control variates
Antithetic variates
Monte Carlo requires a lot of repetition to adequately cover the entire probability space. By their
very nature, the values of the random draws are random, and so after a given number of
replications, it may be the case that not the whole range of possible outcomes has actually occurred.
What is really required is for successive replications to cover different parts of the probability space
that is, for the random draws from different replications to generate outcomes that span the entire
spectrum of possibilities. This may take a long time to achieve naturally. Hence, a an alternative
approach , the antithetic variate technique is used which reduces the sampling error by returning a
simulation using a complement set of original set of random variables.
For example, if the driving stochastic force is a set of TN(0, 1) draws, denoted u1, for each
replication, an additional replication with errors given by u1 is also used. It can be shown that the
Monte Carlo standard error is reduced when antithetic variates are used.
Suppose the average value of parameter of interest across two set of Monte carlo replication is given
by
x ( x1 x 2 ) / 2
Where x1 and x2 are the average output parameter values for simulation sets 1 & 2, respectively.
The variance of x is given by,
var( x)
1
var( x1 ) var( x2 ) 2 cov( x1 , x2 )
4

Cont..
If no antithetic variates are used, the two sets of Monte carlo replications are independent and there
co-variance will be zero leading to variance of :
var( x)
1
var( x1 ) var( x2 )
4
The use of antithetic variate results in negative covariance between the original random draws and
their complements which will further reduce the Monte Carlo sampling error.
Control variates
The application of control variates involves employing a variable similar to that used in the
simulation, but whose properties are known prior to the simulation. Denote the variable whose
properties are known by y, and that whose properties are under simulation by x. The simulation is
conducted on x and also on y, with the same sets of random number draws being employed in both
cases. Denoting the simulation estimates of x and y by x and y, respectively, a new estimate of x
can be derived from:
x* y ( x y )
The new x* variable estimate will have a smaller sampling error than the original x variable if the
control statistic and statistic of interest are highly co-related.
To illustrate, take the variance of both sides as:

var( x * ) var y ( x y )
The variance of y would be zero as it is control variable and its properties are known. So, the
variance would be:261
var( x * ) var( x ) var( y ) 2 cov( x , y )

Cont...
Var (x*) < Var (X^) for control variate method to reduce the sampling error in MCS
The same can be expressed as:
var( y ) 2 cov( x , y ) 0
The same can be simplified as
1
cov( x , y ) ! var( y )
2
By dividing both the side of this inequality by product of Standard deviations , co-relation can be
obtained:
1 var( y )
corr ( x , y ) !
2 var( x )
Cont...
Example of usage control variate:
For pricing an arithmetic Asian option using simulation. An arithmetic Asian option is one whose
payoff depends on the arithmetic average value of the underlying asset over the lifetime of the
averaging; at the time of writing, an analytical (closed-form) model is not yet available for pricing
such options. In this context, a control variate price could be obtained by finding the price via
simulation of a similar derivative whose value is known analytically e.g. a vanilla European option.
Thus, the Asian and vanilla options would be priced using simulation with the simulated price given
by P(A) and P(BS), respectively. The price of the vanilla option, P(BS) is also calculated using an
analytical formula, such as BlackScholes. The new estimate of the Asian option price, P*A, would
then be given by
PA* ( PA PBS ) PBS*

Random number re-usage
Under re-usage of random numbers, the same set of draws are used across experiments which can
greatly reduce the variability of the difference in the estimates across those experiments.
Two uses of re-using the random numbers are:
1. For testing the power of Dickey Fuller test
2. For different experiments with options using time series data
Bootstrapping
Unlike simulation, bootstrapping approach draws random return data from a historical data. It
involves sampling repeatedly with replacements from the actual data.
Suppose a sample of data, y= y1, y2, . . . , yT are available and it is desired to estimate some
parameter . An approximation to the statistical properties of d can be obtained by studying a
sample of bootstrap estimators. This is done by taking N samples of size T with replacement from y
and re-calculating with each new sample. A series of estimates is then obtained, and their
distribution can be considered.
The advantage of bootstrapping: No assumption are made regarding the true distribution of the
parameter estimate,
How bootstrapping is used:
1. Generate a sample of size T from the original data by sampling with replacement from the whole
rows taken together (that is, if observation 25 is selected, take y25 and all values of the
explanatory variables for observation 25).
2. Calculate , the coefficient matrix for this bootstrap sample.
3. Go back to stage 1 and generate another sample of size T. Repeat these stages a total of N times.
A set of N coefficient vectors, , will thus be obtained and in general they will all be different, so
that a distribution of estimates for each coefficient will result.

Situations where Bootstrapping is ineffective
Outlier data: If there are outliers in the data, the conclusions of the bootstrap may be affected. If
outliers are taken too frequently (remember it is sampling with replacement) then the representing
distribution will have fatter tails and if outliers are not sampled then the distribution will not be a
true representative.
Non-Independent data: Bootstarp assumes that the data are independent of one another. However,
if there is autocorrelation in the data, then it is not useful. To overcome the same, a method called
moving block bootstap can be used wherein blocks of data are examined at one time in order to
preserve the original data dependency.
Random number generation

The simplest class of number generator are from uniform (0,1) distribution. A unifrom distribution is
one where values between o and 1 are only chosen with each one having equal chance of being
selected. A number generator can be discrete or uniform.
Recursive process for generating the random number will require the user to specify an initial value
to get the process started. The choice of this value will, undesirably, affect the properties of the
generated series. This effect will be strongest for y1, y2, . . . , but will gradually die away. For
example, if a set of random draws is used to construct a time series that follows a GARCH process,
early observations on this series will behave less like the GARCH process required than subsequent
data points. Consequently, a good simulation design will allow for this phenomenon by generating
more data than are required and then dropping the first few observations. For example, if 1,000
observations are required, 1,200 observations might be generated, with observations 1 to 200
subsequently deleted and 201 to 1,200 used to conduct the analysis.
These computer-generated random number draws are known as pseudo-random numbers, since
they are in fact not random at all, but entirely deterministic, since they have been derived from an
exact formula! By carefully choosing the values of the user-adjustable parameters, it is possible to
get the pseudo-random number generator to meet all the statistical properties of true random
numbers. Eventually, the random number sequences will start to repeat, but this should take a long
time to happen.

Disadvantages of Simulation Approaches
Computationally intensive
Results are not precise: The simulation process is based on some assumption about the
parameters. However, if there is any error in the assumption in data generating process, it can lead
to in accurate results. For example, in the context of option pricing, the option valuations obtained
from a simulation will not be accurate if the data generating process assumed normally distributed
errors while the actual underlying returns series is fat-tailed.
Results are hard to replicate: Practically, MCS is not conducted documenting each sequence of
random draw and hence results will be specific with each draw. If number of replication is small, a
repeat of the experiment would involve different sets of random draws and therefore would be
likely to yield different results.
Results are experiment specific: Simulation process in finance is based on specific assumptions
related to data generating process and set of equations. If the assumptions are changed in data
generating process or set of equation, the results can be significantly different.
Exercise
1. A risk manager was trying to understand on how to reduce the sampling error in MCS. His senior
told him about the two methods and asked him to use the most suitable method for his problem.
After going through the research provided by the seniors, the manager made following comment:
A. The manager concluded that the antithetic method will reduce the error due to perfectly negative co-
relation between the parameter and its complement while using Asian option valuation.
B. Under control variate technique, the controlled variable has no sampling error.
Which of the following option is correct?
1. A only
2. B only
3. A & B both
4. Neither A nor B
2. One can not change the assumptions of data generating process or set of equation while
implementing MCS . If the above condition is violated, which one of the limitation of MCS it will
fall in?
1. High computation cost
2. Results are imprecise
3. Results are difficult to replicate
4. Results are experiment specific

Answers
1. Option 1
2. Option 4
dz
www.edupristine.com
FMP-I
Forward and Futures
AGENDA
Introduction Options, Futures and other derivatives

Mechanics of Futures markets
Hedging strategies using Futures
Determination of Forward and Future Prices
Commodity Forwards and Futures
Fundamentals of Commodity Spot and Futures Markets

Introduction to Derivatives
EduPristine FRM-I (2016)
Introduction Derivatives
Derivatives are financial instruments which derive their value from an underlying asset and some
other variables such as interest rates, volatilities etc.
Futures, forwards, options and swaps are some of the most common examples of derivatives.
The underlying asset: It is a more basic financial instrument. Example: stocks, bonds, interest rate,
commodity etc.
Example of a derivative: Option

An investor owns a call option (which is a derivative) whose underlying asset is the common stock of a
company A. This option gives the investor, the right to buy the stock at a certain predefined price on or
before a future date.

Introduction Markets
Exchange traded Markets

Market where individuals trade standardized
contracts that have been defined by the exchange
themselves.
An Exchange acts as an intermediary which
facilitates a regulatory oversight and hence makes
the markets a safer place for trading.
Chicago Board of Trade and Chicago Mercantile
Exchange are two examples.
Open outcry system and Electronic trading.
Over the counter markets
There is no intermediary and no standardized
contracts; parties create their own T&C with each
other.
Much larger than the exchange traded market in
terms of value of underlying assets (more than 4
times larger).
Trades done between financial institutions or
between financial institutions and clients. Financial
institutions act as a market maker (quotes both bid
and ask).
Source of Graph: The Economist
Types of Investments and Purposes
Asset types
Financial Assets: Equity, Debt securities.
Commodities: Gold, Copper, Crude Oil.
Real Estate
Lets take an example of a financial asset (stock)
We can buy the stock through the broker by paying the stock price.
We can either hold the bought asset or sell it at the current market price.
During the holding period of the stock, the dividends received goes to your pocket as the income from
the asset.
After selling the asset, we earn a profit or loss on the asset, depending on the selling price of the
asset (stock).
Purpose of Assets
Investment Asset and Consumption Asset.
Market Maker
An individual or an institution which keeps an inventory of financial instruments or commodities who could
be asked for trading those assets. The individual or the institution then quotes a bid and an ask price on the
option.

Forward and Futures Contracts
Futures Contracts: Agreement to buy or sell an asset for a certain price at a certain time. A futures
contract is traded on an exchange.
Forward Contracts: Forward contracts are similar to futures except that they trade in the over-the-
counter market.
Notation for Valuing Futures and Forward Contracts
S0: Spot price of the asset underlying today.
F0: Futures or forward price today.
T: Time until delivery date (in years).
R: Risk-free interest rate per annum, expressed in continuous compounding, for maturity T.
Payoff of forwards and futures:
Long Short
P P
X S X S
In both Forward and Futures contracts there is an obligation to buy or sell an asset
Options
(to be covered in detail in later slides)
Traded both on exchanges and over the counter markets.

Call option gives the holder the right to buy the underlying asset by a specified time at a certain
price.
Put option gives the holder a right to sell the underlying asset by a specified time at a certain
price.
European options can be exercised on the specified date only, unlike American options which can
be exercised anytime up to the expiration date.
One option contract is to buy/sell 100 shares in the US.
No obligation to exercise the right
Payoff of a Call Payoff of a Put
100 100
Asset Asset
Price Price
100 100

Types of traders
Hedgers
Uses derivatives to hedge the risks they face from volatility in the asset prices.
Example: a company is getting a cash inflow in 3 months time in a foreign currency. It hedges its currency risk
by taking a short position in a currency forward at a particular price.
Speculators
Uses derivatives to bet on a particular direction of movement of the asset price.
If a speculator believes that the spot rate of a foreign currency will be higher in 3 months than its present 3
month forward rate, he goes long on the forward. After 3 months if he is correct, he receives foreign currency
at lower rate and immediately resells it at the higher spot rate.
Arbitrageurs
Take offsetting positions in 2 or more instruments to lock a profit.
Suppose that:
The spot price of gold is US$390.
The quoted 1-year forward price of gold is US$425.
The 1-year US$ interest rate is 5% per annum.
No income or storage costs for gold. Is there an arbitrage opportunity?
&Z^d&
Arbitrage = Buy Low Sell High at no risk
Mechanics of Futures Market

Mechanics of future markets
Investor Broker Trader Exchange
Clearing House Member Clearing House
Specifications of a Contract Margins Clearing House Margin

Asset: If asset is a commodity, Margin account: Investor deposits If not the broker a clearing house
exchange specifies the asset in a certain amount of money with member has to be a member of
complete detail: grade, quality, the broker in the margin account. the clearing house.
size, shape, color, etc. Initial margin: the initial amount Clearing margin: Just like a margin
Contract size: The amount of the deposited in the margin account. account with a broker, members
asset to be delivered. Maintenance margin: Is have an account with the clearing
Delivery arrangement: somewhat below the initial house.
place of delivery margin. The minimum amount No maintenance margin
Delivery month after which a margin call is sent to Account balance to be
the investor. After margin call maintained at all times = number
investor has to top his margin of contracts *
account to the initial margin. original margin
Question: Margin Calculation (Important)
An investor bought 1000 shares of ABC company each priced at $50. The initial margin
requirement were 60%.and the maintenance margin requirement is 25%. At what price would the
investor be getting a margin call?

Solution
Total investment = 1000x 50 = 50,000

Initial Margin = 60% x 50,000 = 30,000
Maintenance margin = 25% x 50,000 = 12,500
The investor gets a call when he/she loses 30,000 12,500 = 17,500
Price of share after this loss = 50 17.5 = $32.50
Hence the investor will get the margin call when the price falls to $32.50
Question: Margin Calculation (Important)
What would be the variation margin if the stock price reduced to $10 from $50?

Solution
When stock goes down to $10, the loss = 1000 x (50 10) = 40,000
Hence margin account becomes 30,000 40,000 = -10,000
Hence the investor will need to pay [(30,000 (-10,000)] = 40,000 as variation margin
Daily settlement An exmaple
Trade Price $ 100 / barrel

contract size 50 barrels
traded future contracts: 4 Long position
Initial Margin $ 2500 / contract i.e. $ 10000
Maintenance margin $ 2200 / contract i.e. $ 8800
Day Initial price Settlement price Daily Gain/Loss Cumulative Gain/Loss Margin account Margin call
1-Mar $100.00 $10,000.00
1-Mar $100.00 $101.00 $1.00 $200.00 $10,200.00
2-Mar $101.00 $101.50 $0.50 $100.00 $10,300.00
3-Mar $101.50 $98.50 -$3.00 -$600.00 $9,700.00
4-Mar $98.50 $96.50 -$2.00 -$400.00 $9,300.00
5-Mar $96.50 $92.00 -$4.50 -$900.00 $8,400.00 $1,600.00
6-Mar $92.00 $91.00 -$1.00 -$200.00 $9,800.00
7-Mar $91.00 $90.20 -$0.80 -$160.00 $9,640.00
8-Mar $90.20 $93.80 $3.60 $720.00 $10,360.00
9-Mar $93.80 $80.50 -$13.30 -$2,660.00 $7,700.00 $2,300.00
10-Mar $80.50 $90.90 $10.40 $2,080.00 $12,080.00

Delivery
Very few futures contract lead to a delivery of the underlying asset. Most are closed out early
The period of delivery is decided by the exchange, but the exact date of delivery is specified by the
short contract holder
Notification to deliver is given by the broker to the clearing house. The number of contracts and
the specifics of the delivery (what grade, type, quality, location, etc) is mentioned
Exchange identifies a party with a long position to accept delivery. Typically one with the oldest
outstanding long position
Party with the long position has to accept the delivery
Whole delivery process from the issuance of intention to deliver to delivery takes 23 days
Cash settlement is also possible. Settlement price is the spot price of the underlying on the day
opening/day close
Types of orders
Market order: Order placed at current market rate

Limit order: A buy limit order can only be executed at the limit price or lower, and a sell limit order can
only be executed at the limit price or higher
Stop Order: A stop order activates a market order to either buy or sell a stock once the price of the stock
reaches a specified price, known as the stop price. As the stop price is surpassed, the order becomes a
market order i.e. as soon as the stock surpasses the stop price, it is bought or sold at the market price
Stop Limit order: A stop limit order activates a limit order to either buy or sell a stock once the price of
the stock reaches a specified price. It has the same difference with stop order as the difference between
a limit order and a market order. Example: A stock is trading at $40, an investor places a stop limit order
with stop price $42 and limit price $42.80. As the stock crosses $42 the limit order is activated
Market-if-touched order: It is similar to stop order but the buy and sell happens the opposite way. In
case of MIT buy order, the market order is activated when a lower level price is met. While in stop order
the market order is activated when a higher level price is met
Discretionary order: A market order whose execution may be delayed at the brokers discretion in order
to get a better price
Time-of-day order: An order which is executed at a particular period of day
Open order: An order which is in effect until executed or until the end of the trading in the
particular contract
Fill-or-kill order: An order that must be executed immediately or not at all

Z
Precious Metals Settlement price: this price is used for margin call
Gold Comex (100 Troy oz; $/troy oz) calculations at end of trading day
Sett Days
High Low
Vol 0 int Change: -$8.5 per troy ounce. Hence, total of
price chge 000s 000s 100*$8.5=$850 reduction in margin account balance
Dec 738.7 -8.5 748.5 736.7 111.8 300.3 Open interest: total number of contracts outstanding
Feb 745.0 -8.6 754.8 743.5 6.00 31.86 with the exchange. It is the total number of long
Total 123.7 449.5 positions / the total number of short positions. It is
one trading day older than the prices day
Time Trading Activity Open Interest
Jan 1 A buys 1 option and B sells 1 option contract 1
Jan 2 C buys 5 option and D sells 5 option contract 6
Jan 3 A sells his 1 option and D buys 1 options contract 5
Jan 4 E buys 5 options from C who sells 5 options contracts 5
On January 1, A buys an option, which leaves an open interest and also creates trading volume of 1
On January 2, C and D create trading volume of 5 and there are also five more options left open
On January 3, A and D take offsetting positions, open interest is reduced by 1 and trading volume is 1
On January 4, E simply replaces C and open interest does not change, trading volume increased by 5
Hedging Strategies Using Futures

Futures vs. Cash (spot) position
Generally, there is a high correlation in price movements between the futures market and the
cash market.
Futures position acts as a substitute for later cash transaction.
Hedgers generally take equal and opposite position in cash and futures.
Hedging using futures
Hedge is a position established to minimize the exposure to the unwanted risks.

Hedge can be either a short hedge or a long hedge.
Short hedge (selling hedge): is a hedge in which an investors takes a short position in a contract
Example: A farmer expects to harvest 1000 bushels of corn in October. Currently in July the price of corn is
$2.5 per bushel. The farmer faces the risk of price of corn decreasing in 3 months time (i.e. October).
Available 3 months future position in corn trades at $2.3 per bushel.
In the above example, the cash price (spot) is $2.5, futures price is $2.3.
Farmer has a naturally long position in the cash market (i.e. in October the farmer would be having 1000
bushels of corn).
To hedge his position, the farmer places a short hedge (selling hedge), to sell corn bushels in at $2.3 per
bushel in the month of October.
Long hedge (buying hedge) Is a hedge in which an investors takes a long position in a contract
Example: A corn flakes producer, Kellogs, needs 1000 bushels of corn in August, to produce corn flakes for
future demand. Currently, in July, the price of corn is $2.5 per bushel. Kellogs is concerned about the
increase in price of corn in one months time. Available 1 month corn futures position trades at $2.55 per
bushel.
The cash price is $2.5/bushel, and the 1 month futures price is $2.55
Kellogs has a naturally short position in the cash market, and thus places a long hedhe (buying hedge).

Is hedging always good?
Hedging and shareholders: Shareholders can hedge the risk themselves. Companies dont need
to. But do the shareholders have as much information as the companies? What about transaction
costs and commissions? Companies carry out high volume transactions hence cost of hedge
is lower.
Hedging and competitors: What if the price of hen food was reduced as the hen producers union
pressed the suppliers to reduce their prices. HPs profits would rise as he had locked his selling
price and the raw material prices went down. For others the change in profits would be 0. What if
the raw material prices went up for some reason and the union decided to raise their selling
prices in the market proportionately. HPs profit would reduce while others profit remain the
same. Hedging actually is causing fluctuation in profits!!
Any other reasons you can think of?
Hedging in a practical world (Basis Risk)
Not always does the futures contract date be the same as the
date the asset is to be bought or sold. Future
What if the farmer didnt know when his corn produce would be Price
ready for sale? Spot Price

What if he doesnt get a long contract that will close his position
just one day before the closing of his short contract?
Time
What if there is no contract for the type/grade of corn the
farmer is selling?
This is basis risk
Basis = spot price of asset futures price contract Future
Basis = 0 when spot price = futures price Price
b1 = S1 - F1 and b2 = S2 - F2
Future
Farmer pay off when he sells his corn: S2 F1 - F2 or F1 b2 Price
In a typical transaction, F1 is known but b2 is not known at time
t1 b2 is the basis Time
Expect 2 questions directly based on the 4 factors that incorporate basis risk

Question
Imagine a stack-and-roll hedge of monthly commodity deliveries that you continue for the next
five years. Assume the hedge ratio is adjusted to take into effect the mistiming of cash flows but is
not adjusted for the basis risk of the hedge. In which of the following situations is your calendar
basis risk likely to be greatest? (FRM 2008 Sample Paper)
A. Stack and roll in the front month in oil futures
B. Stack and roll in the 12-month contract in natural gas futures
C. Stack and roll in the 3-year contract in gold futures
D. All four situations will have the same basis risk
Solution
A.
The oil term structure is highly volatile at the short end, making a front-month stack-and roll hedge heavily
exposed to basis fluctuations. In natural gas, much of the movement occurs at the front end, so the 12-month
contract wont move much. In gold, the term structure rarely moves and wont begin to compare with oil and
gas.

Choice of contracts
Choose a delivery month that is as close as possible to, but later than end of life of the hedge
because:
The futures prices are quite volatile during the delivery month.
When there is no futures contract on the asset being hedged, choose the contract whose futures
price is most highly correlated with the asset price.
In such cases the proportion of the exposure that should optimally be cross hedged
V
Optimal Hedge Ratio: h U G S
V GF
Where
S is the standard deviation of ^ the change in the spot price during the hedging period.
F is the standard deviation of & the change in the futures price during the hedging period.
is the coefficient of correlation between ^ and &.
Optimal number of contracts

The optimal number of contracts (N*) to hedge a portfolio consisting of NA number of units and
where Qf is the total number of futures being used for hedging
h * NA
N*
Qf
In the case of a stock index the similar logic follows. If P is the value of the portfolio of stocks held
by an investor and A is the current value of the stocks lying under one futures contract then the
optimal hedge ratio, N*, should be equal to P / A.
In practical cases investors dont typically have portfolios that trace the index. Hence the concept
of comes into play. Beta is a measure of a stock's volatility in relation to the market.
In order to change the beta of the portfolio to we need to long or short the (N*) number
of contracts depending on the sign of (N*)
P
N*
A
P
N * ( E * -E )
A
Negative sign of (N*) indicates shorting the contracts

Question
If you have a portfolio of $500,000 which mirrors S&P 500. Each S&P 500 contract is $250 times
the index when the index is at 500. Calculate the number of contracts to be hedged?
Solution
A = 250 * 500 = 125,000. Then N* = 500,000/125,000 = 4 contracts should be shorted for

the hedge

Question
The current value of the S&P 500 index is 1,457, and each S&P futures contract is for delivery of
US$250 times the index. A long-only equity portfolio with market value of US$300,100,000 has
beta of 1.1. To reduce the portfolio beta to 0.75, how many S&P futures contract should you sell?
(FRM Sample Paper 2009)
Solution
[(0.75 1.1)/ 1] * [300,100,000 / {250 * 1,457}] = -288.36 sell 288 contracts

Rolling forward a hedge
We can use a series of futures contracts to increase the life of a hedge

Each time we switch from 1 futures contract to another we incur a type of basis risk
Strip Hedge
Stack and Roll Hedge
Expect 2 questions based on Stack and Roll hedge or MG case.

Please refer original case study for Stack and Roll Hedge
Determination of Forward and Futures Prices

Consumption vs. Investment Assets
Investment assets
Assets held for investment purposes by significant numbers of investors.
(examples: stocks, bonds, gold, silver)
Consumption assets:
Assets held primarily for consumption (examples: copper, oil and pork bellies)
Gambling Short Selling an example
Short selling involves selling securities that are not owned
Suppose an investor short sells 500 IBM shares, the broker will borrow the securities from another client and sells
them in the market in the usual way
At some stage the investor will close the position by purchasing 500 IBM shares. The investor takes the profit if the
stock prices have declined , else vice versa
Short Squeezed: If anytime the broker runs out of shares to borrow, the investor is short squeezed and forced to
close his position immediately
Forwards vs. Futures contracts
Forwards Futures
Not traded on exchanges Traded on exchanges
Are private agreements between two parties Standard contracts
and are not as rigid in their stated terms and Clearing house and daily mark to market
conditions reduces credit risk
Credit risk is high Settlement can occur over a range of dates
High customization Usually closed out before maturity and hardly
Settlement at the end of contract and on a any deliveries happen
specific date
Mostly used by hedgers that want to remove the
volatility of the underlying, hence delivery/ cash
settlement usually takes place

Determination of Forward Price
The price of a forwards contract is given by the equation below:

F0 = S0ert in the case of continuously compounded risk free interest rate, r
F0 = S0(1+r )t in the case of annual risk free interest rate, r
Where:
F0: forward price
S0: Spot price
t: time of the contract
Known income from underlying
If the underlying asset on which the forward contract is entered into provides an income with a present
value, I, then the forward contract would be valued as:
F0 = (S0 I )ert
Known yield from underlying
If the underlying asset on which the forward contract is entered into provides a continuously compounded
yield, q, then the forward contract would be valued as:
F0 = S0e(r-q)t
q: continuously % of return on the asset divided by the total asset price
Value of forward contracts
At the time of entering into a forward contract, long or short, the value of the forward is zero
This is because the delivery price (K) of the asset and the forward price today (F0) remains the
same
The value of the forward is basically the present value of the difference in the delivery price and
the forward price
Value of a long forward, f, is given by the PV of the pay off at time T:
= (F0 K )erT
K is fixed in the contract, while F0 keeps changing on an everyday basis
For continuous dividend yielding underlying

f = S0e-qt Ke-rt
For discrete dividend paying stock
f = S0 I Ke-rt
Index futures: A stock index can be considered as an asset that pays dividends and the dividends
paid are the dividends from the underlying stocks in the index
If q is the dividend yield rate then the futures price is given as:
F0= S0e(r-q)t
Index Arbitrage
When F0 > S0e(r-q)T an arbitrageur buys the stocks underlying the index and sells futures
When F0 < S0e(r-q)T an arbitrageur buys futures and shorts or sells the stocks underlying the index

Forward vs. Futures Prices
Forward and futures prices are usually assumed to be the same. When interest rates are uncertain,
they are slightly different in theory
A strong positive correlation between interest rates and the asset price implies the futures price is
slightly higher than the forward price
A strong negative correlation implies the reverse
Futures and Forwards on Currencies
Interest rate Parity

( rbc r fc )T
F0 S0e
Formula to remember:
If Spot rate is given in USD/INR terms then take American Risk-free rate as the first rate
In other words, individual who is interested in USD/INR rates would be an American (Indian will always think
in Rupees not dollars!!), which implies foreign currency (rf) in his case would be rINR
FUSD S USD e ( rUSD rINR )T

INR INR

Question
The forward rate of a 3-month EUR/USD foreign exchange contract is 1.1565 USD per EUR. USD
LIBOR is 4% and EUR LIBOR is 2%. The spot USD per EUR exchange rate is?
Solution
F0 = S0 e(r-rf)t
1.1565 e-(.04- .02).25 = 1.1507

Question
Assume that the current 1-year forward exchange rate is 1.200 USD per EUR. An American bank
pays 2.4% annual interest rate on a 1-year deposit and a 4.0% annual interest rate on a 3-year USD
deposit. A European bank pays a 1.5% annual interest rate for a 1-year deposit and a 2.0% annual
interest rate for a 3-year EUR deposit. The forward exchange rate of USD per EUR for exchange
three years from today is closest to:
Solution
The 2 year forward rate in US = [(1.04)3 / 1.024] 1 = 4.81%

The 2 year forward rate in Europe = [(1.02)3 / 1.015] 1 = 2.25%
The forward exchange rate of USD per EUR for exchange three years from today:
1.2 *(1.04812) / (1.02252) = 1.261

Question
The two-year risk-free rate in the United Kingdom is 8% per annum, continuously compounded.
The two-year risk-free rate in France is 5% per annum, continuously compounded. The current
French Franc to the GBP currency exchange rate is 1GBP = 0.75 French Franc.
What is the two-year forward price of one unit of the GBP in terms of the French Franc so that no
arbitrage opportunity exists?
A. 0.578
B. 0.706
C. 0.796
D. 0.973
Solution
B. Ans = 0.75*e(0.05-0.08)*2 = 0.706

Question
A bank has a USD 50,000,000 portfolio available for investing. The cost of funds for the USD
50,000,000 is 4.5%. The bank lends 50% of the assets to domestic customers at an average loan
rate of 6.25%. The rest of the portfolio is lent to UK clients at 7%. The current exchange rate is
USD1.642/GBP. At the same time, the bank sells a forward contract equal to the expected receipts
one year from now. The forward rate is USD1.58/GBP. The weighted average return to the bank on
its investments is closest to?
Solution
The return from UK customers, $25,000,000/1.642 = GBP 15,225,335* 1.07 = GBP 16,291,108
The bank sells a forward contract: GBP 16,291,108*1.58 = USD 25,739,951
Earnings (USD 25,739,951 25,000,000) / 25,000,000 = 2.96%
t

Question
Given the following:

Current spot CHF/USD rate: 1.3680 (CHF1.3680 = USD1)
3-month USD interest rates: 1.05% ; 3-month Swiss interest rates: 0.35%
A currency trader notices that the 3-month forward price is USD / CHF 0.7350. In order to
arbitrage, the trader should?
Solution
The spot is quoted in terms of Swiss Francs per USD. To convert this into USD per Swiss Franc, we
get: 1/1.3680 = 0.7310. The theoretical futures price = 0.7310 * exp((0.0105 0.0035) * 0.25) =
0.7323. Therefore, the quoted futures price is too high. Thus, one should sell the overvalued CHF
futures contract.
In order to arbitrage, one would do the following:
Borrow USD 0.7310 * exp((-0.0035)*0.25) = USD 0.7304 for 3 months
Buy spot exp((-0.0035)*0.25) = CHF0.9991, invest at 0.35% for 3 months
Short a futures contract on CHF1
At maturity,
Pay back 0.7304 * exp((0.0105) * 0.25) = USD 0.7323
Receive 0.9991 * exp((0.0035) * 0.25) = CHF 1
Delivers CHF 1 on the futures contract, receives USD 0.7350
An arbitrage profit of USD0.7350 USD0.7323 = USD 0.0027 would be realized in 3 months time

Commodity Futures
Futures on Consumption Assets

F0 ^0 ed
Where u is the storage cost per unit time as a percent of the asset value
Alternatively, F0 ^0 hrT
Where U is the present value of the storage costs

Question
The current spot price for cotton is $0.325 per pound. The annual risk-free rate is 3.0%, and the
cost to store and insure cotton is $0.002 per pound per month. A 3 month futures contract for
cotton is trading at $0.3368 per pound. Is there an arbitrage opportunity available, and if so, how
should an investor take advantage of it?
Solution
The Forward price = 0.325e0.03*0.25 (0.002 0.002*1.0025 0.002*1.0025^2) = 0.3335; Yes; the
investor should sell the futures contract, borrow at the risk-free rate, and buy the
spot asset.

Question
The spot rate for a commodity is $19. The annual lease rate for the commodity is 5%. The
appropriate continuously compounding annual risk-free rate is 6.5%. What is the 3-month
commodity forward price?
Solution
F0 = S0 e(r- G)t = $19.07

The Cost of Carry
The cost of carry, c, is the storage cost plus the interest costs less the income earned
For an investment asset F0 = S0ecT
For a consumption asset F0 ^0ecT
The convenience yield on the consumption asset, y, is defined so that: F0 = S0 e(cy )T
Futures Prices

Futures price
In a treasury futures contract the price of the bond is not easy to deliver because the cheapest to
deliver bond is not known
However, if the cheapest to deliver bond and its delivery date is known we can call upon the
equation which considers discreet payouts from an underlying and can be given as below:
F0 = (S0-I) ert
Eurodollar Futures:
A Eurodollar is a dollar deposited in a foreign bank/US bank outside the United States
Eurodollar futures are futures on the 3-month Eurodollar deposit rate (same as 3-month LIBOR
rate)
Long position => agrees to give a loan at the determined price
One contract is on the rate earned on $1 million
A change of one basis point or 0.01 in a Eurodollar futures quote corresponds to a contract price
change of $25 (1mm * 0.01% * 90/360)
When it expires (on the third Wednesday of the delivery month), final settlement price is 100
minus actual three month deposit rate.
Contract Price = 10,000*[100 0.25*( 100 Q)]
Q = Quoted Price
Question
Suppose you buy (take a long position in) a Eurodollar futures contract on November -1
The contract expires on December-21 Date Quote
The prices are as shown Nov 1 97.12
How much do you gain or lose Nov 2 97.23
On the first day Nov 3 96.98
On the second day

Over the whole time until expiration?
Dec 21 97.42

Solution
Day 1: increase by 11 basis points, hence gain = 11*25 = $275

Date Quote
Day 2: decrease by 25 basis points, hence loss = 25*25= $625
Nov 1 97.12
Until expiration: increase by 30 basis points, gain = 30*25 = $750
Nov 2 97.23
Nov 3 96.98

Dec 21 97.42
Eurodollar futures and forwards

In a Eurodollar futures contract that locks in an interest rate between times T1 and T2 the interest
rate is locked in at time T1 and the settlement is made at time T1
In an FRA which also locks in an interest rate between times T1 and T2, the final settlement is
made at time T2
Difference between Eurodollar futures and FRA
In an FRA the payoff is equal to the difference in the forward interest rate and the realized interest rate
The settlement is at time T1 for the E-futures contract while its at time T2 for the forward contract
Analysts adjust forward rates with the following equation:
1
Forward rate futures rate V 2T1T2
2
V is the standard deviation of the change in the short term interest rate in 1 year

Normal Backwardation and Contango
When the futures contracts have substantial time to maturity then the futures prices are different
from the expected future spot prices
When futures prices are greater than the expected future spot prices then the scenario is termed
as contango
When futures prices are lower than the expected future spot prices then the scenario is termed as
normal backwardation
Normal futures curve: When futures prices are greater for greater maturity
Inverted futures curve: When futures prices are lower for greater maturity. (example: orange juice,
because its value depreciates with time)
Commodity Spreads

Commodity Spreads
Commodity Spread: is a result of a commodity that is used in the production process. Lets take an
example of mustard seeds which can be used to prepare mustard oil which sells at a higher price
than mustard seeds. This difference of prices between the raw and processed commodity is the
commodity spread
Commonly used Commodity spreads:
Commodity Spreads
Crack Spread Crush Spread

->Long (short) position in ->Long (Short) position in
Crude oil and short (long) Soybeans and short (long) position in
position in heating oil and gasoline soybean meal and soybean oil
Some Important Commodities and their Properties
Commodity Demand Production Property Futures Price

Constant
Seasonal Interest and Storage Cost Increases until harvest time
Corn throughout
Production determines the price and then drops sharply
the year
Produced
Too expensive to store / Futures price rise steadily
Natural Gas Seasonal Demand throughout
Demand peaks in winter in fall months
the year
Produced Oil prices are stable in absence
Constant world- Can be cheaply
Oil throughout of short-run
wide demand transported
the year supply and demand
Futures prices of electricity is
Price is determined by the demand and
Electricity Non-storable more volatile than financial
supply at a given point in time
futures

Five Minute Recap
Key Terms to Remember: Eurodollar Futures: Cost of Carry Model:

1
Initial, Maintenance and variation Margin Futures rate ( FRA) V 2T1T2 Investment asset F0 = S0ecT
2
Contango and Backwardation Consumption asset F0 ^0ecT
Normal Contango and Normal Backwardation Basis Risk: Convenience yield F0 = S0 e(cy )T
Commodity Spreads Basis = spot price of asset futures price
contract
Forwards on Currencies: Index Futures: Commodity Spreads: Rolling forward a hedge:

( rbc r fc )T F0= S0e(r-q)t Crack Spread Strip Hedge
F0 S0e
Crush Spread Stack and Roll Hedge
Optimal Hedge Ratio: Optimal number of contracts: Future Price: Interest Rate Parity:
V S USD e ( rUSD rINR )T
h
U GS N* ( E * -E )
P F0 = S0erT FUSD
V GF A F0 = S0 (1+r )T INR INR
Payoff of a Call Payoff of a Put Optimal contract to hedge a Types of Orders:

portfolio: Market order Discretionary
h * NA order
N* Limit order
Qf Stop Order Time-of-day
100 100 order
Stop Limit
Types of traders: order: Open order
Hedgers Arbitrageurs Market-if- Fill-or-kill order
Speculators touched order
100 Asset 100 Asset
Price Price Margins: Maintains Margin
Initial Margin Variation Margin
F4,5(forward rate) = (R5T5 R4T4)/(T5 T4)
dz
www.edupristine.com
FMP-II
Interest Rates
Agenda
This reading also covers the following readings from Valuation and Risk Models
Spot, forward and Par Rates
Returns, Spreads and Yields

Interest Rates
Types of interest rates
Interest rate is the amount of money a borrower promises to pay to the lender over and above the
principal amount
Treasury Rates: This is the rate an investor receives when he invests in Treasury bills and Treasury bonds.
Treasury bills are short term while Treasury bonds are longer term (> 1 year)
Corporate bond rates: These are rates on long term bonds issued by a corporate
LIBOR: This is the London Interbank Offer Rate (LIBOR) and the rate at which banks make a large wholesale
deposit or loan with/to another bank
1 month, 3 months, 6 months and 12 month LIBORs
Opportunity cost for AA rated banks
Not entirely risk free
Repo rates and Reverse Repo: Repo rate is the rate at which banks borrow money from the central bank.
Reverse Repo rate is the rate at which the central bank borrows money from banks
A Repurchase agreement (also known as a repo or Sale and Repurchase Agreement) allows a borrower to use
a financial security as collateral for a cash loan at a fixed rate of interest
A repo is equivalent to a cash transaction combined with a forward contract

Calculation of interest rates
There are many ways to calculate interest rates annual, semi annual, quarterly, continuously
compounding and so on
Each rate can be expressed in the form of another rate. For example an interest rate of 10%
compounded semi-annually would fetch (1 10% / 2) * (1 10% / 2) = 1.1025 (remember
6months rate is 10% / 2) on $1 after one year. This is equivalent of 10.25% annual rate
Amount compounded annually would be given by:
A = P (1+ r)t
A terminal amount
P principal amount
r annual rate of interest
t number of years for which the principal is invested
If amount compounded n times a year then:
A = P ( 1 r/n )nt
When n then we call it continuous compounding:
A = Pert (this formula is derived using limits and continuity)
Question
If the interest rate is 10% per annum compounded continuously, then what is the effective annual
interest rate?

Solution
At continuous compounding, $1 after an year will become 1.ert = e0.1x1 = 1.10517

Had it been just annual compounding, then the interest rate required for $1 to rise up to $1.10517
would have been 1.10517 1 = 10.517% which is the effective annual rate.
Question
If the interest rate is 10% per annum compounded semi-annually then what is the equivalent
continuously compounded interest rate.

Solution
2 = 1e(rx1)
=> 1.1025 = er
=> r = 0.09758 = 9.758%
Alternatively, following formulae can be used to calculate the interest rates:

Rc = Continuous compounding interest rate
Rm = Periodic compounding interest rate with m periods per year
Rc = m.ln(1+ Rm/m)
Rm = m[e^(Rc/m)
Try solving the above problems using these formulae
Introduction Bonds
A bond is a debt security usually issued by a company or the government to raise funds
Example: A company ABC issues bonds of worth $100. An investor X buys the bond by paying
$100 to the company ABC. ABC promises to repay the money back to X after 5 years and also pay
5% of the $100 principle every year, semi-annually
In the above example:
Face Value: $100
Coupon rate: 5%
Time to maturity: 5 years
W
C C C C

Bond pricing
Bonds are either zero coupon bonds (having no interest payments) or coupon bonds
(with periodic interest payments)
The price of a bond is the present value of all the coupon payment and the final principal payment
received at the end of its life
1
T
1
B Ce Pe
rt rT
(1 YTM) n
1
t 1 B Iu Fu
B the bond price YTM (1 YTM) n
C coupon payment
r zero interest rate at time t
P bond principal
T time to maturity
The yield of a bond is the discount rate (applied to all future cash flows) at which the present
value of the bond is equal to its market price
z to Maturity = Investors Required Rate of Return
The par yield is the coupon rate at which the present value of the cash flows equal to the par
value (principal value) of the bond
If we are looking at a semi-annual 5 year coupon bond with a par value of $100 then the coupon
payment would be solved using the following equation:
5
100 (C / 2)e
t 1
rt
100e 5 r
Question
A 10 year bond has a yield of 12% with a 7% coupon payment annually

What is the bond price?
What will happen if the yield of the bond increases by 1%
1
1 (1 YTM) n 1
B Iu Fu
Use : YTM (1 YTM) n

Or
z 1 2 3 4 5 6 7 8 9 10
z 12%
Coupon payments 7 7 7 7 7 7 7 7 7 7
Principal payment 100
PV factor 0.892857 0.797193878 0.711780248 0.635518078 0.567427 0.506631 0.452349 0.403883 0.36061 0.321973
Total PVs 6.25 5.580357143 4.982461735 4.448626549 3.971988 3.546418 3.166445 2.827183 2.52427 34.45114
Bond price 71.74888
z 1 2 3 4 5 6 7 8 9 10
z 13%
Coupon payments 7 7 7 7 7 7 7 7 7 7
Principal payment 100
PV factor 0.884956 0.783146683 0.693050162 0.613318728 0.54276 0.480319 0.425061 0.37616 0.332885 0.294588
Total PVs 6.19469 5.482026784 4.851351136 4.293231094 3.79932 3.36223 2.975425 2.633119 2.330194 31.52095
Bond price 67.44254

t,z
Default risk The higher the default risk, the higher the required YTM
Liquidity The less liquid the bond, the higher the required YTM
Call features Increase required YTM
A bond that can be redeemed by the issuer prior to its maturity
Extendible feature Reduce required YTM
An extendible bond gives its holder the right to "extend" its initial maturity at a specific date
or dates
Retractable feature Reduce required YTM
A bond that features an option for the holder to force the issuer to redeem the bond before maturity at par
value
Treasury zero rates
In the case of treasury rates there are some key facts to know:
Treasury bills are issued at a discount from face value and are paid at their par (face amount) at maturity.
The purchase price is expressed as a price per hundred dollars
Bills are sold at a discount. The discount rate is determined at auction
Bills pay interest only at maturity. The interest is equal to the face value minus the purchase price
Bills are sold in increments of $100. The minimum purchase is $100
Boot Strap Method to determine zero rates
Consider the bond prices of Treasury bonds given below in column 4. Calculate the continuously
compounded zero rates for 6 months, 12 months, 18 months and 24months
Continuously
Bond Principal Time to Maturity Annual Coupon Bond Price Compounded 0-
rate
100 0.5 10 99.5 10.76
100 1.0 10 98.4 11.43
100 1.5 10 96.5 12.31
100 2.0 10 94.3 13.01

Forward rate agreements (FRAs)
In general:
R 2 T2 R 1T1
Ft1, t2
T2 T1
A forward rate agreement (FRA) is an over the counter agreement where the forward interest rate,
Ft1,t2 ,is fixed for a certain principal between times T1 and T2
The payer of the fixed interest rate is also known as the borrower or the buyer. The buyer hedges against
the risk of rising interest rates, while the seller hedges against the risk of falling interest rates
Payment to the long at settlement = Notional Principal X (Rate at settlement FRA Rate) (days/360)
----------------------------------------------------------
Z
Question
An FRA settles in 30 days

Has $1mn notional
Is Based on 90-day LIBOR
Forward rate of 5%, Actual 90-day LIBOR at settlement is 6%
What is the value of FRA at settlement?

Solution
Value at the end of agreement = (6% 5%) * (90/360)* $1mn = $2,500

s
Duration
Duration it is the measure of how long on an average the holder of the bond has to wait before he
receives his payments on the bond
A coupon paying bonds duration would be lower than n as the holder gets some of his payments
in the form of coupons before n years
Macaulays duration: is the weighted average of the times when the payments are made. And the
weights are a ratio of the coupon paid at time t to the present bond price
n
t *C n*M
(1 y)
t 1

(1 y ) n
t
Macaluay Duration
Current bond price
Where:
t = Respective time period
C = Periodic coupon payment
y = Periodic yield
n = Total no of periods
M = Maturity value

Duration (Cont.)
Macaulay duration is also used to measure how sensitive a bond or a bond portfolio's price is to
changes in
interest rates
A bonds interest rate risk is affected by:
Yield to maturity
Term to maturity
Size of coupon
From Macaulays equation we get a key relationship:
'B
D'Y
B
In the case of a continuously compounded yield the duration used is modified duration given as:
Macaulay Duration
D*
r
1
n
Consider a bond trading at 96.54 with duration of 4.5 years. In this case:
- 96.54* 4.5 -434.43
/) in the yield then the bond price would change by:
-434.43 * ( 0.001)
-0.43443
Hence, B = 96.54 0.43443 = 96.10
Convexity
Duration is a good measure when the changes in yield are small

However if the yield changes are high then we use the measure of convexity along with duration
Convexity is a measure of the curvature of the price / yield relationship
1 d 2B
C
B dy 2
Note that this is the second partial derivative of the bond valuation equation w.r.t. the yield
Hence, convexity is the rate of change of duration with respect to the change in yield
Bond price ($)
P* Actual bond price
Tangent
Y* Yield

Convexity
The convexity of the price / YTM graph reveals two important insights:
The price rise due to a fall in YTM is greater than the price decline due to a rise in YTM, given an identical
change in the YTM
For a given change in YTM, bond prices will change more when interest rates are low than when they are high
To make the convexity of a semi-annual bond comparable to that of an annual bond, we can divide
the convexity by 4
In general, to convert convexity to an annual figure, divide by m2, where m is the number of
payments per year
Calculating Bond Price Changes
We can approximate the change in a bonds price for a given change in yield by using duration and
convexity:
'VB D Mod
u 'i u VB 0.5 u C u VB u 'i
2


Question
If yields rise by 1% per period, then by what price will the bond fall by? Assume C = 16.65.
Solution
'VB . u 0.01 u 964.54 0.5 u 16.75 u 964.54 u 0.01

359 34.63 0.81 3382
2
.
Note:
The 1st term -D x x P , can either be a positive or a negative amount, indicating an increase or
decrease in price respectively. Depending on the sign of .
The 2nd 2 x P, will always be a positive amount.

Theories of the Term Structure
(1 ilt ) n (1 ist )(1 ist )...(1 ist

year1 year 2 yearn
Three theories are used to explain the shape of the )
term structure
Expectations theory
The long rate is the geometric mean of expected
future short interest rates.
(1 ilt ) n rpn (1 ist )(1 ist )...(1 ist
year1 year 2 yearn
Liquidity preference theory )
Investors must be paid a liquidity premium to hold Where rpn is the risk premium associated with an n
less liquid, long-term debt. year bond
Market segmentation theory
Investors decide in advance whether they want to
invest in short term or the long term.
Distinct markets exist for securities of short term
bonds and long term bonds.
Supply demand conditions decide the prices.
Term structures
The term structure of interest rates is graphed as though each coupon payment of a non callable
fixed-income security were a zero-coupon bond that matures on the coupon payment date
The yield curve describes the yield differential among treasury issues of differing maturities
The Yield Curve is the graph created by putting term to maturity on the X axis, YTM on the Y axis
and then plotting the yield at each maturity
Upward sloping: This is the most persistent shape historically when short-term interest rates and inflation
are low
Downward sloping (Declining): This occurs at peaks in the short-term interest rate cycle, when inflation is
expected to decrease in the future
Flat: This shape is evident during periods of interest rates transitions
Humped: This occurs when rates are transitioning or perhaps market participants are attracted in large
numbers to particular maturity segment of the market, thereby creating the hump
Rising Declining Flat Humped

zZZWz
Spreads)
16
14
12
z^
10
Percent
0
1 mth 3 mths 6 mths 1 yr 2yrs 5 yrs 7 yrs 10 yrs 30 yrs
Term Left to Maturity

BBB Corporates Government Bonds
Questions
Discuss the yield curve below and the economic impacts it conveys:
Yield
Maturity

Questions
What is the 1.5 year par yield of a $100 bond when the zero rates (Continuous Compounded) on
6months, 12 months and 18 months are 4%, 4.5% and 5%?
Sol: Let it be C
100 = C*e-0.5*.04 -1*.045 -1.5*.05

C = 7.225651/2.863939
C = 2.523
Coupon = 2.523*2 = 5.05%
Summary of Bond Features

Bond price is inversely related to yield
When yield > coupon, bond trades at discount; and when yield < coupon, bond trades at a
premium.
Bonds that do not pay any coupon and are issued at a price below its face
value.
Treasury zero rates the yield obtained from treasury zero coupon bonds
Treasury yield curve (term structure) the curve representing treasury zero rates vis--vis the
'
Forward rate rates implied by zero rates for a period of time in the future.
Duration Linear measure of interest rate risk of a bond.
Convexity rate of change of duration w.r.t. to interest rate changes.

Prices, Discounts Factors and Arbitrage
(This Reading is From Valuation & Risk Models)
Day count conventions
Day count defines the way in which interest is accrued over time. Day count conventions normally
used in US are:
Actual / actual treasury bonds
30 / 360 corporate bonds
Actual/360 money market instruments
The interest earned between two dates
(Number of days between dates)*(Interest earned in reference period)

=
(Number of days in reference period)

Examples
Actual / 360
The interest price of a 91-day T-bill is 9%. Find the dollar amount of interest paid over the 91 day period and
the corresponding rate of interest
Dollar interest is $100*0.09*91/360 = $2.275
Rate of interest = 2.275/(100-2.275) = 2.328 %
Actual / Actual
A treasury bond with face value $100 pays a semi-annual coupon of 12%. Coupon payment dates are Mar 1
and Sept 1. Find the interest earned between Mar 1 & July 3
Reference period Mar 1 to Sept 1 is 184 days
Desired period is Mar 1 and July 3, is 124 days
Interest earned is 124/184*6 = $4.043
30 / 360
A corporate bond with face value $100 pays a semi-annual coupon of 12%. Coupon payment dates are Mar 1
and Sept 1. Find the interest earned between Mar 1 & July 3
Reference period Mar 1 to Sept 1 is 6 months with each month @30 days =180 days
D:
Interest earned is 122/180*6 = $4.0666
Treasury bonds
The prices for treasury bonds are quoted in dollars and 1/32nd of a dollar
$8227 is equivalent to $82.84375
Cash price / dirty price is the price at which the investor buys a bond from the market
Cash price = Quoted price + accrued interest
Accrued interest is the interest which the nearest coupon that is due generates

Treasury bond futures
Treasury bond futures are the most commonly traded futures contract on the Chicago board of
trade (CBOT)
When the Treasury bond futures contract expires, any government bond with maturity more than
15 years on the first day of the delivery month and is not callable for the next 15 years from that
day can be delivered
Conversion price
When a bond is delivered the party with the short position, the amount transacted is:
Quoted futures price + accrued interest
Where, Quoted futures price = settlement price conversion factor
The conversion factor is equal to the quoted price the bond would pay per dollar or principal on
the first day of the delivery month on the assumption that the interest rate for all maturities
equals 6% per annum (with semi annual compounding)
Question
The last coupon payment of $10 was paid on a treasury bond on June 19, 2009. The next coupon
is due on December 19, 2009 and we are currently on September 1, 2009. If the quoted price is
$8227 then the cash price would be?

Solution
The number of days for which the December coupon has accrued interest is the time period
between June 19 to September 1 (74 days). The actual time period between June 19 and
December 19 is 183 days. Hence the interest the December coupon accrues is:
$10 * (74/183) = 4.04
,
Cheapest to deliver bond
At any given time during the delivery month there are many bonds that can be delivered in the
CBOT futures contract
The party with the short position can chose to deliver the cheapest bond when it comes to
delivery, hence he would chose the cheapest to deliver bond
Net pay out for delivery (he has to buy a bond and deliver it):
Quoted bond price
Consider an example in the table below where the short position holder has 3 options for delivery.
His cheapest to deliver bond is Bond 2
Cheapest to Deliver Bond (All figures in $)
Settlement Future Price: 94.23
Bond Quoted Bond Price Conversion Factor
1 99.6 1.033
2 135.67 1.432
3 122.45 1.257
Cost of Delivering ($)

Bond 1 99.6 - (94.23*1.033) = 2.26041
Bond 2 135.67 - 0.73264
Bond 3 122.45 - (94.23*1.257) = 4.00289

One Factor Risk Metrics and Hedges
(This Reading is From Valuation and Risk Models)
DV01
DV01: The price value of basis point (PVBP) or dollar value of basis point (DV01) change is the
absolute change in the bond price from one basis point change in yield.
DV01 = ~ price at YTM0 price at YTM1 ~
YTM0 = the initial YTM
YTM1 = the YTM 1 basis point above or below YTM0 (YTM1 = YTM0 0.0001)
DVBP [ D * u P0 ] u 0.0001

DV01 Application to hedging
Hedge ratio is calculated using DV01 with the help of following relation
HR = DV01 (per $100 of initial position)

DV01 (per $100 of hedging instrument)
Duration based hedging strategies
Considering a situation where an asset that is interest rate dependant is hedged using an interest
rate futures contract
In such cases the number of contracts to hedge is given by the equation below:
PDP
N*
FC DF
FC Contract price for interest rate futures

DF Duration of asset underlying futures at maturity
P Value of portfolio being hedged
DP Duration of portfolio at hedge maturity
When hedges are constructed using interest rates it is important to note that interest rate and
futures prices move in opposite directions . So if one is expecting to lose money when the interest
rate falls, one should long futures contracts so that they can hedge their losses by gains in futures
prices

Question
An investor has invested $10m in government bonds and is expecting the interest rates to rise in
the next 6 months so he decides to hedge himself by interest rate futures. It is currently June and
he decides to use the December T-bond futures contract for the hedge. If the current futures price
is 97.2345 and the duration of the portfolio of government bonds at the end of 6 months is 7.1
years. The duration of the cheapest to deliver T-bond in December is given as 9.121 years. What
position should the investor take in the futures contract? How many futures contract should long /
short for the hedge if each contract is for the delivery of $100,000 face value?
Solution
The number of contracts that should be shorted is:
10,000,000 7.1
u 80
(97.2345 u100,000 / 100) 9.121

Portfolio Duration and Convexity
Portfolio duration is the weighted sum of durations of individual securities.

Weight of each security is the value of the security divided by the value of the portfolio.
Portfolio convexity is calculated in the same way portfolio duration. It is the weighted sum of
durations of individual securities.
Negative Convexity
Callable bonds exhibit negative convexity when yields fall below certain level.
At lower yield, there is incentive for the issuer to call the bond.
Price curve of the bond bends away from the normal curve thereby exhibiting negative convexity.
Barbell and Bullet Strategy
In barbell strategy, investor uses the bonds of short and long maturities and does not invest in the
bonds of intermediate maturity.
In bullet strategy, investor uses the bonds concentrated in intermediate maturity range.
In volatile rate environment, barbell strategy is preferred over bullet strategy.

Multi Factor Risk Metrics and Hedges
(This Reading is From Valuation and Risk Models)
Weakness of Single Factor Approach
Single factor approach assumes that all the future rate changes are driven by single factor.
The same change in interest rate is assumed for the entire yield curve.
In practice, change in short term interest rate might be different from the change in long term
interest rate.
Same hedging instrument cannot be used for hedging the change in short term interest rate and
long term interest rate.

Key Rate Exposures
Key Rates are the rates selected at key point on the yield curve. These are usually 2, 5, 10 and 30
year rates.
Key rate exposures hedge risk by using rates from a small number of available liquid bonds.
Partial 01 is used to measure the risk of the bond or swap portfolio in terms of liquid money
markets and swap instruments.
Forward 01 is used to measure the risk of the bond or swap portfolio in terms of shifts in the
forward rates.
Key Rate Shift
Key Rate Shift technique is a approach to nonparallel shift in the yield curve.
This technique allows to determine changes in all the rates due to the changes in key rates.
Choice has to be made as to which key rates shifts and how the key rate movement relate to prior
or subsequent maturity key rates.

Key Rate 01 and Key Rate Durations
Key Rate 01 measures the dollar change in the value of the bond for every basis point shift in the
key rate
Key Rate 01 = (-1/10,000) (Change in Bond Value/0.01%)
Key rate duration provides the approximate percentage change in the value of the bond
Key Rate Duration = (-1/BV) (Change in Bond Value/Change in Key rate)
Hedging with Key Rate Exposures
Hedging positions can be created in response to key rate shifts.

This can be done by equating individual key rate exposures next to key rate shifts to the overall key
rate exposure for that particular key rate change.
The above step indicates either a long position or a short position in securities to protect against
the changes in interest rate surrounding key rate shifts.

Mechanics of Options Markets
Agenda
Introduction to Options
What are Options
Intrinsic Value of Options
Returns to Option buyers and sellers
Put Call Parity
Bounds and Option Values
Determinants of Option Values
Some special cases
Summary

What are Options?
Options are contracts that give its buyer the right to buy or sell a particular asset
In future
At a pre-decided price (i.e. exercise or strike price)
Without any obligations
The seller of the option collects a payment (Premium) from the buyer for providing the option
Types of options:
Call or Put Options
Call Option: Gives option holder the right to buy the asset at an agreed price
Put Options: Gives option holder the right to sell the asset at an agreed price
European or American Options
European options: Are those that can only be exercised on expiration
American options: May be exercised on any trading day on or before expiration
Question
Early exercise of an option is more likely for:

A. European calls options on stocks paying large dividends.
B. American call options on stocks paying small dividends.
C. American put options deep in the money and close to maturity.
D. American put options on stocks paying large dividends.

Solution
C.
A. European options cannot be exercised early
B. Small dividends will not make much of a difference in the price fall in the stock
C. A deep in the money put option should always be exercised early because it is likely that the stock might
recover from the fall
D. Though this might be profitable if the stock prices significantly fall after the ex-dividend date but the third
option is likely to provide more profit
Question
Assuming the stock price and all other variables remain the same what will be the impact of an
increase in the risk-free interest rate on the price of an American put option?
A. No impact
B. Negative
C. Positive
D. Cannot be determined

Solution
B. (Negative)

Intrinsic value: is the maximum of zero and the value of the option if the option were
exercised immediately
At the money:
When the price of the underlying is the same as the strike price of the option, the option is termed at the money
and exercising it carries a nil pay-off
In the money:
When the price of the underlying is greater than the strike price carried by a call option, the call option is termed in
the money, as exercising it results in a positive pay off
When the price of the underlying is less than the strike price carried by a put option, the put option is termed in
the money, as exercising it results in a positive pay off
Out of the money:
When the price of the underlying is less than the strike price carried by a call option, the call option is termed out
of the money, as exercising it will result in a nil pay off
When the price of the underlying is greater than the strike price carried by a put option, the put option is termed
out of the money, as exercising it will result in a nil pay off

Illustration: Pay-offs from buying a Call
yz
whose price is expected to range from 010 at the time of expiration.
If share price is less than 5, then the pay off to the option buyer is nil.
If the price is more than 5, the pay-off moves upward linearly with the share price.
Stock Strike
Call Value Put Value
Price Price
Call Value
0 0 5 5
1 0 4 5
6
2 0 3 5
Call-Pay off
3 0 2 5 4
4 0 1 5
5 0 0 5
2
6 1 0 5
7 2 0 5
8 3 0 5 0
0 4 8 12
9 4 0 5
Stock Price
10 5 0 5

Illustration: Pay-offs from selling a Call
yz
If share price is less than 5, then the pay off to the option seller is nil.
If the price is more than 5, the pay-off moves downward linearly with the share price.
Stock Strike
Call Value Put Value Short Call-Pay off
Price Price
0 4 8 12
0 0 -5 5 0
1 0 -4 5
2 0 -3 5
Short Call-Pay off

3 0 -2 5 -2
4 0 -1 5
5 0 0 5 -4
6 -1 0 5
7 -2 0 5
-6
8 -3 0 5
9 -4 0 5 Stock Price
10 -5 0 5

Illustration: Pay-offs from buying a Put
Wyz
If share price is more than 5, then the pay off to the option buyer is nil.
If the price is less than 5, the pay-off moves linearly with the share price.
Stock Strike
Call Value Put Value
Price Price Put Value
0 0 5 5 6
1 0 4 5
Put-Pay off
2 0 3 5
3 0 2 5 4
4 0 1 5
5 0 0 5 2
6 1 0 5
7 2 0 5
0
8 3 0 5 0 4 8 12
9 4 0 5 Stock Price
10 5 0 5

Illustration: Pay-offs from selling a Put
Wyz
If share price is more than 5, then the pay off to the option seller is nil.
If the price is less than 5, the pay-off moves downward linearly with the share price.
Stock Strike
Call Value Put Value Short Put-Pay off
Price Price
0 4 8 12
0 0 -5 5 0
1 0 -4 5
2 0 -3 5
Short Put-Pay off

3 0 -2 5 -2
4 0 -1 5
5 0 0 5 -4
6 -1 0 5
7 -2 0 5
-6
8 -3 0 5
9 -4 0 5 Stock Price
10 -5 0 5
Payoffs from Options
Long call payoff = Max (ST yK >Dy ST,O)
0 0
y ST y ST
y y
0 ST 0 ST
Short call Short put

Returns to Option Buyers and Sellers
Returns to Option Sellers
Returns to Option sellers:

The price that the option writer gets for underwriting the contract is called premium.
If the option is not exercised, the option writer makes profit from the premium.
If the option is exercised, the option writer may make profit or loss depending on the spot price of the
underlying asset at the time.
Example: A Call option writer gets premium of 1 for an option with strike price of 5
He makes a profit if:
The option is not exercised when spot price is less than 5. The profit is 1 (i.e. premium).
The option is exercised and spot price is more than 5 but less than 6.
The profit to the call writer is less than 1.
If the spot price is 6, the writer has no profit and no loss.
For all spot prices more than 6, the call writer makes losses, which increase linearly with increase in spot prices.

Returns to Option Sellers
Short put payoff with premium

0 2 4 6 8 10 12
2
1
Short put payoff

Stock Option Short-Call Put Strike 0
Price Premium Value Value Price -1
-2
0 1 1 -4 5
-3
1 1 1 -3 5 -4
2 1 1 -2 5 -5
Stock price
3 1 1 -1 5
4 1 1 0 5
5 1 1 1 5 Short call payoff with premium
6 1 0 1 5 0 2 4 6 8 10 12
2
7 1 -1 1 5 1
Short call payoff
8 1 -2 1 5 0
9 1 -3 1 5 -1
-2
10 1 -4 1 5
-3
-4
-5
Stock price
Returns to Option Buyers
Profit to Option buyers:

The pay-off are distinct from the profit (or loss) to Long call payoff with premium
the option holder. 5
To estimate the profit, the premium (price of 4
Long call payoff
option) is to be subtracted from the pay-off. 3

Illustration: In continuation to above, further 2
consider options which carries a premium of 1. 1
0
Stock Option Call Put Strike -1
Price Premium Value Value Price -2
0 1 -1 4 5 0 2 4 6 8 10 12
Stock price
1 1 -1 3 5
2 1 -1 2 5 Long put payoff with premium
3 1 -1 1 5
5
4 1 -1 0 5 4
Long put payoff
5 1 -1 -1 5 3
6 1 0 -1 5 2
1
7 1 1 -1 5
0
8 1 2 -1 5 -1
9 1 3 -1 5 -2
10 1 4 -1 5 0 2 4 6 8 10 12
Stock price
Put Call Parity
Put Call parity
Consider the Pay-off of a trader who has the following position:

A Call Option with a Strike Price of 5 and
A Bond with a maturity value of 5.
Share Price Call Bond Value

Strike Price Bond + Call
at Expiration Pay-Off at Maturity
05 0 5 5 5
6 1 5 5 6
7 2 5 5 7
8 3 5 5 8
9 4 5 5 9
10 5 5 5 10

Put Call parity
Consider, now, the Pay-off of a trader who has:

A Put Option with a Strike Price of 5 and
An equivalent unit of the underlying asset.
Share Price Put Pay-Off (Exercise Stock Stock+

at Expiration Price 5) Pay-off Put
0 5 0 5
1 4 1 5
2 3 2 5
3 2 3 5
4 1 4 5
510 0 510 510
Put Call parity
The Pay-offs are exactly the same
12
10
Total payoff
8
6
4
2
0
0 2 4 6 8 10 12
Share price

Question: Put Call parity
According to Put Call parity for European options, purchasing a put option on ABC stock will be
equivalent to
A.
B.
C. ^
D.
Solution: Put Call parity
B: p + S0 = c + Ke-rT

Put Call parity
Put Call parity provides an equivalence relationship between the Put and Call options of a
common underlying and carrying the same strike price.
It can be expressed as:
Value of call Present value of strike price = value of put share price
If value of put is not available, it can be derived as:
Value of put = Value of call present value of strike price - share price
Put-call parity relationship, assumes that the options are not exercised before expiration day, i.e. it
follows European options.
This holds true for American options only if they are not exercised early.
In case of dividend-paying stocks, either the amount of dividend paid should be known in advance
or it is assumed that the strike price factors the future dividend payment.
The mathematical representation of Put Call Parity is:
X d
Pr emium(C ) PV of strike price PV of dividends
1 rt 1 rt
= Initial stock price (S) Put premium (P)
Put Call Parity is valid only for European options, for American Options this relationship turns into an inequality
Question: Put Call parity
Consider a 1-year European call option with a strike price of $27.50 that is currently valued at
$4.10 on a $25 stock. The 1-year risk-free rate is 6%.What is the value of the corresponding put
option?
A. 4.1
B. 5
C. 6
D. 25

Solution: Put Call parity
p + S0 = c + D + y-rt

The value of an option changes over its life.

Consider the earlier illustration of the call
If the share price of is below 5 on the exercise date, the call will be worthless.
If the stock price is above 5, the call will be worth 5 less than the value of the stock.
Even before maturity of the option, its value can never remain below this lower-bound line.
For options that still have some time to run, the heavy lower line is thus the lower-bound limit on the
market price of the option.
The diagonal line in the plot is the upper bound limit to the option price, because the stock gives a higher
ultimate pay-off than the option.
4
Call payoff
0
0 1 2 3 4 5 6 7 8 9 10 11
Share price

The value of an option changes over its life.
Consider the earlier illustration of the call
If at the options expiration, stock price > exercise price, the option is worth the stock price minus the
exercise price.
If the stock price < exercise price, the option is worthless. But the share owners still have a valuable financial
asset in the form of stock of ABC Corporation.
The value of the option would lie between these two bounds throughout the options life.
4
Call payoff
0
0 1 2 3 4 5 6 7 8 9 10 11
Share price

Determinants of Option Values
Option Minimum Value Maximum Value
European call (c) ct D^t-yZ&Zt) St
American Call (C) Ct D^t-yZ&Zt) St
European put (p) pt DyZ&Zt)-St) yZ&Zt
American put (P) Pt Dy-St)) X
Where t is the time to expiration

Question
If a European call option is written on a dividend paying stock, an increase in which of the
following will not automatically result in an increased option price?
A. The stock price
B. The risk-free rate
C. The time to expiration
D. The volatility of the stock price
Solution
C.
The time to expiration

Thank z
www.edupristine.com
FMP-III
Options Valuation
Agenda
Complications in Valuing Options
Binomial Method of Valuing Options
An Example
Replicating Call Option
Replicating Put Option
Risk Neutral Valuation
Change in Future Stock Price
Generalizing Binomial Method
Black Scholes Model
Historical Volatility and Implied Volatility
Limitations of Black Scholes Model
Summary
Expect around 6-8 questions in the exam from todays lecture

Standard approach for valuing any asset:
Figure out expected cash flows and
Discount them at risk adjusted cost of capital reflecting the opportunity cost of capital.
Complications that arise in valuing Options
Impossible to quantify risks associated with the Option cash flows
Risks associated with the Option change every-time there is change in the price of the underlying
Binomial Method of Valuing Options

Binomial Method
Binomial method entails

Assuming the price of the underlying asset can take only two values in any given interval of time
Determining Option pay-offs at these prices
Replicating the same pay-offs in a package consisting of assets that can be valued
Alternatively, determining probability of each pay-off to arrive at a certainty equivalent expected cash-flow
and discounting it to the present value at the risk-free rate
Risk Neutral Method
Su2 IV1 = Max[(Su2-X), 0]

p
Su
p 1-p
S0 Sud IV2
p
1-p
Su
1-p IV3
Sd2
An Example

Question
Consider a six-month European call and put option on non-dividend paying stock with identical
exercise prices of $85. This option is at the money. The short-term, risk-free interest rate was a bit
less than 4 percent per year, or about 2 percent for six months. The stock either falls to $63.75 or
rises to $113.33 after six months. Determine their pay-offs at expiration:
Solution
The pay-offs are as follows:
Stock Price = $63.75 Stock Price = $113.33

1 Call option $0 $28.33
1 Put option $21.25 $0

Question Replicating Call Option

Determine the value of the call option in previous question by replicating the call option
Solution: Lets look at the pay-offs from a package consisting of 0.5714 stocks and borrowing a
principal of $35.71 from the bank. The total amount to be repaid is $36.42 (including interest)

0.5714 Shares $36.42 $64.75
Z -36.42 -36.42
Total Payoff 0 $28.33
The pay-offs are exactly the same as in the previous example for the call option. It follows that
the value of the call today should be equal to the value of 0.5714 shares less Present Value of
$36.43
Thus, value of Call = $12.86

Two questions remain, how did we determine the number of stocks i.e. 0.5714 and how did we
determine the amount to be borrowed?
The number of shares to be held is give by the option delta, given by:
Spread ofPossible option price 28.33 0

0.5714
Spread of possible share prices 113.33 63.75
The amount to be borrowed is equal to the present value of the difference between the pay-offs
from the option and pay offs from the delta shares, i.e. 0.5714 share. In our example:
Scenario 1 Scenario 2
Stock Price 63.75 113.33
Option Value 0 28.33

s^ 36.43 64.76
Payoff from Option 0 -28.33
Portfolio Value 36.43 36.43
The amount to be borrowed equals Present Value or PV of 36.43

The Pay-offs from a put option can be replicated by selling delta share and setting aside a sum of
money in a risk-free investment
In our example, the delta for the put option is given by:
Spread ofPossible option price 0 21.25

0.4286
Spread of possible share prices 113.33 63.75
The amount to be placed in risk-free investment is PV(48.57). Calculated as shown below:
Scenario 1 Scenario 2
Stock Price 63.75 113.33
Option Value 21.25 0
s^ 27.32 48.57
Payoff from Option 21.25 0
Portfolio Value 48.57 48.57

The put can be replicated as shown below:

Sale of 0.4284 Shares $-27.32 $-48.57
Z 48.57 48.57
Total Payoff 21.25 $0
The value of put therefore is:

Value of put = -Ws
= -

Risk Neutral Valuation
Risk Neutral Method

The assumption is that investors are indifferent to risk.
Step 1: Determine the probabilities associated with the different pay-offs
Step 1: Determine expected cash flow under the assumption that investors are indifferent to risk
Step 2: Discount the expected cash flow at the risk-free rate to arrive at the present value
In our example, since the risk-free rate for six months is 2%, and investors are indifferent to risk, it
follows that:
- probability of rise) * (-25)] = 2.0 percent
Therefore the probability of rise, p, = 0.463 or 46.3%
Expected future value of the call option after six months is given by
W- probability of rise) * 0]

= $13.16
The value today therefore is PV(13.16) = $12.86

Question
Currently, shares of ABC Corp. trade at USD 100. The monthly risk neutral probability of the price
increasing by USD 10 is 30%, and the probability of the price decreasing by USD 10 is 70%. What
are the mean and standard deviation of the price after 2 months if price changes on consecutive
months are independent? (FRM 2010 Sample Paper)
Solution
Develop a 2 step tree.
D
Variance = 9% (120 92)2 92)2 92)2 = 168
Thus, standard deviation = 12.96


The following formula that relates the up and down changes to the standard deviation of stock
returns:
1 upside change = u = e
1 downside change = d = 1/ u
Where, e = base of natural logarithms = 2.718
= standard deviation of (continuously compounded) stock returns
h = time interval as fraction of year
In our example, standard deviation of stock returns, = 40.69%, h = 0.5
u= e 0.40690.5 = 1.3333, => upside change = 33%
d= 1/u = 1/1.3333 = 0.75, => downside change = 25%
Thus stock price takes the following two values
$85 x 1.3333 = $113.33
$85 x 0.75 = 63.75

Generalizing Binomial Method
Generalizing the Binomial Method

One step Binomial Method is simplistic
Assumes just two values for the asset price is possible in the future
More realism can be added by shortening the time intervals so that the calculations can allow for
greater number of values for the asset price at expiration
In our example if we allowed the stock to take values at the end of three months, we would have three values
at the end of six months.
To work out the equivalent upside and downside changes when we divide the period into two
three-month intervals (h = 0.25), we use the same formula:
= 1.226,=> upside change = 22.6%

We get the following tree:
85
69.36 104.21
3 Months
-18.6%
56.6 85 127.76
6 Months
-18.6% -18.6%

Generalizing the Binomial Method
If the time intervals could be made extremely small, we would be able to account for a large
number of changes in the share price
With the help of computer programs available today the binomial method can be used with very
small time intervals
If one can think of infinitesimally small time intervals, one could have a continuum of stock prices
as reflected in the plot below:
0.1 Lognormal Distribution
0.09
0.08
0.07
Probability
0.06
0.05
0.04
0.03
0.02
0.01
0
0 10 20 30 40 50 60
Stock Price
The plot above reflects the log-normal distribution of stock prices

It can take any value between 0 and infinity
The fact that the stock price can never fall by more than 100 percent, but that there is a small chance that it
could rise by much more than 100 percent is captured in this distribution
Black Scholes Model

Black and Scholes Model
Black and Scholes formula allows for infinitesimally small intervals as well as the need to revise
leverage for European options on Non Dividend paying stocks
The formula is:
R f uT
[ N (d1) u P ] [ N (d 2) u X u e ]
Where,
P
ln[ ] [ R f (0.5 u V 2 )] u T
d1 X
V T
d 2 d1 V T
Log is the natural log with base e

N (d) = cumulative normal probability density function
X = exercise price option;
T = number of periods to exercise date
P =present price of stock
= standard deviation per period of (continuously compounded) rate of return on stock
Value of Put = R f uT
[X ue u {1 N (d 2}] [{1 N (d1)} u P]
Question: Black and Scholes Model

Calculation of the value of call option:
Price of stock now (P) 85

Exercise price (EX) 85
Standard deviation of continuously compounded annual returns () 0.4069
Year to maturity (t) 0.5
Risk-free interest rate per annum, rf 4%
Log [P/PV (EX)] 0.02
Log [P/PV (EX)] / 0.07
0.14
d1 = log [P/PV (EX)] / 0.2134
d2 = d1 -0.0743
N(d1) Can be calculated by using NORMSDIST (d1) in excel 0.5845
N(d2) Can be calculated by using NORMSDIST (d2) in excel 0.4704
-4%/2
PV (EX) = 85 * e 83.3169
Value of call 10.49

Question: Black and Scholes Model
For European Options on dividend paying stocks, the present value of expected dividends during
the life of the option needs to be reduced from the present price of the stock:
Without Dividend With Dividend
Price of stock now 85 85
Present value of dividend 0 1.99
Price of stock adjustment for dividend (P) 85 83.01
Exercise price (EX) 85 85
Standard deviation of continuously compounded annual returns () 0.4069 0.4069
Year to maturity (t) 0.5 0.5
Risk-free interest rate per annum, rf 4% 4%
Log [P/PV (EX)] 0.02 -0.004
Log [P/PV (EX)] / 0.07 -0.01
0.14 0.14
d1 = log [P/PV (EX)] / 0.2134 0.1309
d2 = d1 -0.0743 -0.1568
N(d1) 0.5845 0.5521
N(d2) 0.4704 0.704
-4%/2
PV (EX) = 85 * e 83.316 83.316
Value of call 10.49 9.36

Historical Volatility (HV)
Historical variability in stock price returns
Backward looking
Calculated by using historical share prices
Implied Volatility (IV)
Model implied variability
Black Scholes model can be used to determine Implied Volatility
Forward looking
Reflects market opinion in the likely volatility of a stock
Limitations of Black Scholes Model

Limitations of Black and Scholes Model
Limitations:
The model does not allow for early exercise
Not suitable for valuing American Options that can be exercised any time during their life
The stepwise binomial method is superior for valuing American Options, particularly American Puts and
American Calls on stocks that pay dividends
Not suitable for valuing warrants as warrants are long term options and it is quite likely that the underlying
stock will pay dividends during the life of the warrant
Also, when exercised warrants increase the total number of shares which adds another level of complication
in valuing warrants using Black and Scholes formula
Summary

Summary
Complications arise in valuing options because its impossible to quantify risks associated with
options
Options can be valued using the binomial method
Replicating options
Risk neutral method
European options on non dividend paying stocks can be valued using the Black Scholes method
Option Delta is defined as:
Spread of Possible option price

Spread of possible share prices
Summary (cont.)
Replicating a call option
Construct a package containing
Buy delta stocks and
Borrow a sum of money which is equal to the difference between the pay-offs from the option and pay offs
from the delta shares
This package has the same pay-off as a call option
The value of the package is the value of the call option
Replicating a put option

Construct a package containing
Sell delta stocks and
Deposit a sum of money which is equal to the difference between the pay-offs from the option and pay offs
from the delta shares
This package has the same pay-off as of a put option
The value of the package is the value of the put option

Summary (cont.)
Risk Neutral Method
Determine the probability of upside and downside changes in stock price
Assume investors are risk neutral
Discount the future expected pay-off at the risk-free rate to derive the option value
Summary (cont.)
Black Scholes Model
Assumes log normal distribution of stock prices
Provides a model for valuing European options on non dividend paying stocks:
R f uT
[ N (d1) u P ] [ N (d 2) u X u e ]
Where, P
ln[ ] [ R f (0.5 u V 2 )] u T
d1 X
V T
d 2 d1 V T
Log is the natural log with base e

N (d) = cumulative normal probability density function
X = exercise price option;
T = number of periods to exercise date
P =present price of stock
= standard deviation per period of (continuously compounded) rate of return on stock
R f uT
Value of Put = [ X u e u {1 N (d 2}] [{1 N (d1)} u P]

Summary (cont.)
Black Scholes model can be used to derive Implied Volatility
Reflects market opinion in the likely volatility of a stock
Question
Company X owns a property with a book value of 80,000. There is a buyer willing to pay
200,000 for the property. However, Company X must also provide the buyer with a put option to
sell the property back to Company X for 200,000 at the end of 2years. Moreover, Company X
agrees to pay the buyer 40,000 for a call option to repurchase the property for 200,000 at the
end of 2 years. In effect, with this transaction Company X borrows money from the buyer. What
is the annually compounded interest rate per year on this implied loan
A. 11.80%
B. 25.00%
C. 41.40%
D. Cannot be determined

Solution
A.
(11.80% )
FMP-III
Options Trading Strategies
Agenda
Trading Strategies
Covered Call
Protective Put
Spread Strategies
Combination Strategies
Trading Strategies
Traders may create positions using different kinds of strategies depending on the:
Expectations regarding the movement of the price of the underlying
Risk appetite
Availability of contracts

Covered Call
Covered Call
Involves selling call options of stocks already owned or simultaneously bought
Motivation
Earning a return from the underlying that is already owned
Lowering the cost of acquisition of the underlying asset
Expectation`
Moderate rise in the price of the underlying
Profit Potential
Maximum Profits when the options are exercised by the buyer
W^W Spot Price
If the options are not exercised the trader gets to keep the premium, thus lowering the cost of acquiring
the asset
More conservative than buying the stock only

Covered Call (cont.)
If My Company (MC) trades at $33 and $35 calls are priced at $1, then an investor can purchase
100 shares of MC for $3300 and sell one (100-share) call option for $100, for a net cost of only
$3200. The $100 premium received for the call will cover a $1 decline in stock price. The break-
even point of the transaction is $32/share. Upside potential is limited to $300, but this amounts
to a return of almost 10%. (If the stock price rises to $35 or more, the call option holder will
exercise his option and the investor's profit will be $35-$32 = $3). If the stock price at expiry is
below $35 but above $32, the call option will be allowed to expire, but the investor can still profit
by selling his shares. Only if the price is below $32/share will the investor experience a loss.
Stock Price Comparison to

Net Profit / Loss
at Expiration Simple Stock Purchase
$30 (200) (300)
$32 0 (100)
$33 100 0
$35 300 200
$37 300 400
Protective Put

Protective Put
Involves buying put options of stocks already owned or simultaneously bought
Motivation
Protection against loss in the value of stocks owned
Expectation
Rise in the price of the underlying
Advantage
Trader profits from the rise in price of the underlying albeit the amount of profit is reduced by the premium
paid to purchase the put
In case the price of the underlying goes down, the trader is still able to sell the underlying at the strike price,
thus insuring her profit
Spread Strategies

Spread Strategies
Bull Call Spread
Involves purchase of Call options at a particular strike price and selling Call options for the same
underlying and carrying the same maturity but having a higher strike price.
A vertical spread
Motivation
Downside protection by agreeing to a limit to the upside profits
Expectation
Moderate rise in the price of the underlying
Profit Potential
Maximum Profits when the trader is able to exercise the option purchased, i.e. the spot price is greater than
the strike price of the option written
^WWZ Premium Paid
Loss is limited to:
Premium Paid Premium Received, when the options expire unexercised
Spread Strategies (cont.)
Bull Call Spread
0
95 100 105 110 115 120 125
-2
-4
Share Price

Bear Spread
Involves purchase of put options at a particular strike price and the sale of the same put options at
a lower strike price.
A vertical spread
Motivation
Downside protection by agreeing to a limit to the upside profits
Expectation
Moderate fall in the price of the underlying
Profit Potential
Maximum Profits when the trader is able to exercise the option purchased, i.e. the spot price is lower than
the strike price of the option written
^WWZ Premium Paid
Loss is limited to:
Bear Spread
4
Total Profit
0
95 100 105 110 115 120 125
-2
-4
Share Price

Butterfly Spread
Involves sale and purchase of two calls (or puts) of the same underlying carrying the same
maturity. A Long Call Butterfly can be established by selling two at the money calls with strike price
WWy
say P-X
Motivation
To profit even when the price of the underlying is range bound and limit losses in case it moves beyond the
expected bound
Expectation
Not much change in the price of the underlying
Profit Potential
Profits translate when the stock price remains within the bounds indicated by the purchased calls. Profits are
maximized when the price of the underlying remains unchanged
Loss is limited to:
Butterfly Spread
4
Total Profit
0
100 105 110 115 120 125 130
-2
-4
Share Price

Combination Strategies (cont.)
Calendar Spread
Involves selling a European call option with a certain strike price and buying a longer maturity
European call option with the same strike price.
Short Call
Profit
Long Call
Diagonal Spreads
Created when both expiration date and strike price of the calls are different for Bull, Bear and
Calendar spreads. Can increase profit maximization possibilities.

Combination Strategy: Positions taken in both the call as well as the put options of the same
underlying stocks
Straddle
Involves purchasing same quantity of at the money call and put options carrying the same strike
price and same maturity
Motivation
To profit from wide variations in the price of the underlying, even though the direction of the movement in
price is uncertain
Expectation
Large change in the price of the underlying
Profit Potential
Profits translate when any one of the options can be exercised and the resultant pay-off is greater than the
premium paid for establishing the position. Profit potential is unlimited
Loss is limited to:
Premium Paid
Straddle
Straddle Profits
2.5
2.0
1.5
1.0
Total Profits
0.5
0
-0.5 0 2 4 6 8 10
-1.0
-1.5
-2.0
-2.5
Share Price

Strap
Involves purchasing two at the money calls for every one at the money put purchased. Both put
and call options carry the same strike price and same maturity. More bullish version of the straddle
Motivation
price is uncertain
Expectation
Large change in the price of the underlying, price expected to increase more than decrease
Profit Potential
premium paid for establishing the position. Profit potential is unlimited although the rise profit is steeper in
case the underlying price increases more than the call strike price
Loss is limited to:
Premium Paid
Strap
Strap Profits
5.0
4.0
3.0
2.0
Total Profits
1.0
0
-1.0 0 2 4 6 8 10
-2.0
-3.0
-4.0
-5.0
Share Price

Strip
Involves purchasing two at the money puts for every one at the money call purchased. Both put
and call options carry the same strike price and same maturity. Bearish version of the straddle
Motivation
price is uncertain
Expectation
Large change in the price of the underlying, price expected to decrease more than increase
Profit Potential
premium paid for establishing the position. Profit potential is unlimited although the rise profit is steeper in
case the underlying price decreases more than the put strike price
Loss is limited to:
Premium Paid
Strip
Strip Profits
5.0
4.0
3.0
2.0
Total Profits
1.0
0
-1.0 0 2 4 6 8 10
-2.0
-3.0
-4.0
-5.0
Share Price

Long Strangle
Involves purchasing slightly out of money calls and puts of the same underlying carrying the same
maturity
Motivation
price is uncertain
Expectation
Large change in the price of the underlying
Profit Potential
premium paid for establishing the position
Loss is limited to:
Premium Paid
Long Strangle
Strangle Profits
2.5
2.0
1.5
1.0
0.5
0
0 2 4 6 8 10
-0.5
-1.0
-1.5

Box Spread
Involves purchasing a Bull Call Spread and one Bear Put Spread. Box Spread yields us
Risk-free rate
Bull Call Spread
Rf Box Spread
Bear Put Spread
Stock Price
Question
The S&P March 2005 index futures contract is trading at 280. The associated American 260 call
option is at 16 and the associated 260 American put option is at 3. Which of the following
strategies would you select to lock in a profit?
A. No strategy would result in a risk-free profit.
B. Buy the put, sell the call and buy the futures contract.
C. Buy and exercise the put and buy the futures contract.
D. Buy and exercise the call and sell the futures contract.

Solution
D.
Buy and exercise the call and sell the futures contract.
Question
An investor sells a June 2008 call of ABC Limited with a strike price of USD 45 for USD 3 and buys a
June 2008 call of ABC Limited with a strike price of USD 40 for USD 5. What is the name of this
strategy and the maximum profit and loss the investor could incur?
A. Bear Spread, Maximum Loss USD 2, Maximum Profit USD 3.
B. Bull Spread, Maximum Loss Unlimited, Maximum Profit USD 3.
C. Bear Spread, Maximum Loss USD 2, Maximum Profit Unlimited.
D. Bull Spread, Maximum Loss USD 2, Maximum Profit USD 3.

Solution
D.
Bull Spread, Maximum Loss USD 2, Maximum Profit USD 3.
Exotic Options

Non-standard features
Bermuda Option: Early Exercise may be restricted to certain dates
Lock out Period: Early Exercise allowed during only a part of the life of the option
Strike Price may change during the life of the option
Incentives for trading in exotic options:

Genuine hedging need
Tax, Accounting, Legal, Regulatory reasons
Position taking on potential future price movement
Exotic Options
GAP Options: Occurs when there is a gap between the stated strike price and the strike price on
which pay-off will take place.
Eg: You buy a put option from an insurance company on your asset with a strike price of $10.0 million.
However you will incur $1.0 million cost in asset transfer, if you exercise the option. So, effectively, you would
exercise the option at <$9.0 million and not at <$10.0 million.
Forward Start Option: Options which will start at some time in the future
Eg: Employee Stock Options (ESOPs)
Cliquet Options: Consists of a series of call/ put options, where the first option is initially active
and subsequent option becomes active upon expiry of previous option.
Eg: First Option with K strike price and T time to expiry. At T first option is settled and next option becomes
active with K1 strike price (may be mutually decided at T) and T1 time to maturity.
Compound Options: Option on another option as underlying
Eg: A call option on another call option, put option on another put, a call on another put and a put on
another call.
Chooser Options: Option holder can choose whether option is call or put after a specified period
of time.

Exotic Options
Barrier Options: Options where pay-off is dependent on the underlying assets price meeting a
certain hurdle during a specified period of time.
Eg: Knock-out options ceases to exist when barrier is reached, Knock-in options comes into existence when
barrier is reached.
Binary Option: Options which have only two stated outcomes.
Eg: cash-or-nothing call will pay-off nothing if asset price is lower than strike price or would pay a fixed cash
amount if asset price goes above the strike price.
Look-back Options: Payoff depends on the maximum or minimum asset price reached during the
life of the option.
Eg: In Floating look-back call option payoff is the amount of final price exceeding the minimum asset price
achieved during the life of the option. For floating look-back put option, the amount would be maximum
asset price exceeding the final price.
Shout Options: A European option where option holder can shout (lock the price) to the writer at
any one time during the life of the option.
Eg: The strike price is $100 and the option holder shouts when underlying asset price goes up to $110. Now
at maturity, if final price is less than $110, the holder will still get payoff of $10 and if the final price is more
than $110, then the holder will receive the pay-off as per the final price.
Asian Options: Options where pay-off is determined by arithmetic average of price of the
underlying asset during the life of the option.
Exotic Options
Exchange Options: Option to exchange one asset for another.
Eg: Option to give up an Asset worth P at time T1 and receive in return another asset worth Q.
Rainbow Options: Options involving two or more risky assets.

Eg: Cheapest-to-deliver bonds.
Volatility/ Variance swaps: Exchange of cash flows over pre-specified time intervals based on
actual realized volatility/ variance and pre-specified volatility/ variance over a notional principal.
Static Option Replication

Deals with hedging of exotic option position
Involves searching for a portfolio of actively traded regular options whose value matches the value
of the exotic option on some boundary
Shorting this portfolio will hedge the exotic option position

Five Minute Recap
Binomial Method of Option Pricing Option Trading Strategies: Strangle Profits

Su 2 Covered Call 2.5
p IV1 = Max[(Su2-X), 0] Protective Put 2.0
Su 1.5
Combination Strategy
p 1-p 1.0
Black Scholes Model: 0.5
S0 Sud IV2
p [ N ( d1) * P ] [ N ( d 2) * PV ( EX )] 0
1-p 0 2 4 6 8 10
log[ p / PV ( EX )] V t -0.5
Su d1
1-p IV3 V t 2 -1.0
S d2 -1.5
d2 d1 V t Share Price
Bull Call Spread Bear Spread Butterfly Spread

8 8 8
6 6 6
Total Profit
Total Profit
Total Profit
4 4 4
2 2 2
0 0 0
95 100 105 110 115 120 125 95 100 105 110 115 120 125 100 105 110 115 120 125 130
-2 -2 -2
-4 -4 -4
Share Price Share Price Share Price
Straddle Profits Strap Profits Strip Profits

2.5 5.0 5.0
2.0 4.0 4.0
1.5 3.0 3.0
Total Profits
Total Profits
Total Profits
1.0 2.0 2.0

0.5 1.0 1.0
0 0 0
-0.5 0 2 4 6 8 10 -1.0 0 2 4 6 8 10 -1.0 0 2 4 6 8 10
-1.0 -2.0 -2.0
-1.5 -3.0 -3.0
-2.0 -4.0 -4.0
-2.5 -5.0 -5.0
Share Price Share Price Share Price
Thank z
www.edupristine.com
FMP-IV
Swaps, Commodities, Foreign Exchange, Central
Counterparties, Corporate Bonds and Mortgage
Backed Securities
Agenda
Introduction
The Comparative Advantage Argument
Interest rate swaps
Valuation of swaps
Currency swaps
Credit risk
Other types of swaps

Introduction
Introduction
A swap is an agreement between two parties to swap cash flows in the future
The arrangement covers swaps on multiple dates
Futures or forwards can be considered as a simple example of a swap where there is a cash flow
exchange on one particular date
Most common swaps are Interest Rate Swaps (IRS) and currency swaps
The legal agreement in which the two parties enter is called a confirmation, which covers the
termination date, calendar used, rates of payment, day count conventions etc.


Take the example of two firms X and Y where:
X wants to borrow floating
Y wants to borrow fixed
Company Fixed Borrowing Floating Borrowing

X 5% LIBOR
Y 7% >/KZ
Table tells us that X can borrow fixed at 5% and Y can borrow fixed at 7%
Also X can borrow floating at LIBOR and Y can borrow floating at (LIBOR 100bps)
This implies that X has absolute advantage in borrowing over Y
The point to note here is that the difference in fixed borrowing rates for X and Y (is not the same
for the floating borrowing rates)
Combined benefit to both X and Y by using swap is (' Fixed ' Floating ) which is 100 bps for X and Y

The Comparative Advantage Argument (Cont.)
To reduce the borrowing rates X and Y enter into a swap shown below through the intermediary
which is usually an Investment Bank (IB)
Assuming zero transaction charges for IB, X borrows at 5% and lends that money at 5.5% to Y
through an investment banker
^z>/KZy>/KZ
Therefore the net borrowing rate for X becomes (LIBOR 50bps) which is lower than the original
rate at of LIBOR
Similarly the net borrowing rate Y becomes 6.5% which is 50bps less than the original rate of 7%
5.50% 5.50%
5%
X IB Y
Libor Libor >
100bps
Question Comparative Advantage

&yzmarket
Fixed Rate Floating Rate

ABC 11% >/KZ
yz 10% >/KZ
How can they benefit from Interest Rate SWAP?

Solution Comparative Advantage

10.00% 10.50%
10%
yz IB ABC
>/KZ >/KZ >
250bps 100bps 100bps
150bps
yz/
Interest Rate Swaps

Interest rate swaps
In the case of fixed-for-floating interest rate swaps two parties get into an agreement where one
pays interest on a floating rate to the other, while the other pays a fixed rate of interest on the
same amount
LIBOR is the most common reference rate of floating interest
Notional principal is exchanged or basically no principal is exchanged
Floating Rate
Party 1 Party 2
Fixed Rate
IRS can be used for

Changing a liability
Transforming a liability
Changing a liability
8%
10% >/KZ0.3%
Party 1 Party 2
LIBOR
Party 1 avails a loan of 10% while party 2 avails a floating rate loan at LIBOR 0.3%
Party 1 is receiving a fixed rate of 8% from party 2
Party 1 pays floating interest rate
Party 1s effective cash flow:
Net cash outflow is (LIBOR 2% )
Party 2s effective cash flow:
Net cash outflow is 8.3 %
Point to note here is that Party 1s fixed liability is changed to floating liability after the swap.
Party 2s liability is changed from a floating liability to a fixed liability of 8.3%

Financial intermediaries
In the practical world most swaps are traded in the OTC market where financial institutions act as
market makers
LIBOR LIBOR
10% Financial
Party 1 Party 2
Institution >/KZ
9.985% 10.015%
0.3%
In the diagram above you can see that the financial institution is making a 3 basis point spread on
the fixed payment of the transaction
In such cases the bank has separate contract with party 1 and party 2
Party 1 and party 2 might not even know that they are on the other sides of the same swap
Bank creates a market by creating both bid and offer positions so that it can seek clients on either
side of the swap
It is exposed to certain credit risks in case it is unable to find a counter party for a swap
Swap rates
The swap rate is the average of:

The fixed rate a market maker is prepared to pay in exchange for a receiving LIBOR (its bid rate)
The fixed rate it is willing to receive in return for a payment of a floating rate (its offer rate)
Like LIBOR swap rates are not risk free rates but close to risk free rates

Valuation of Swaps
Valuation of swaps
There are 2 ways to value a swap.
Considering it as a difference of two bonds
Considering it as a portfolio of FRAs
Value using bonds
Consider an example in which the swap lasts for n years. If the payments are made at the end of each year
then :
If the principal is exchanged between the 2 parties at the end of the swap, then Party 1s cash flow suggests
that its long a fixed rate bond and short a floating rate bond.
Party 2 is short a fixed bond and long a floating rate bond.
We can value the swap by looking at the pay offs of either party.
Hence the value of the swap can be given as:
V = Bfix Bfl
Where:
Bfix = PV of payments
Bfl W/-rt
Value of a floating bond is equal to the par value at coupon reset dates and equals to the Present Value of
Par values (P) and Accrued Interest (AI)

Valuation of swaps
Value using portfolio of FRAs
In this case we assume that each payment at a future date is a forward rate agreement.
For payment at time t, the rate used is the rate for the period between t-1 and t. This rate would
be FRA at t-1
Currency Swaps

Currency swaps
Currency swap involves exchanging principal and interest payment in one currency with the
principal and interest payments in other currency
In this case the principal needs to be specified and it is exchanged in the beginning as well as the
end of the swap
Consider a currency swap between party 1 and 2. In this case Party 1 is in US and can borrow in
USD and party 2 is in Australia and can borrow competitively in AUDs at 6%. Party 1 borrows
$385,000 at 4% and exchanges the principal with Party 2 for 350,000 AUDs (which it borrows in
Australia). The principal is exchanged back at the end of the life of the swap and the life of the
swap is 5 years
5.1% AUD 5% AUD

4% USD Financial 6% AUD
Party 1 Party 2
Institution
4% USD 4.1% USD
What is the net payout for party 1 and party 2?
Questions
Which of the following statements is correct when comparing an Interest rate Swap with a
Currency Swap?
A. At maturity there is no exchange of principal between the counterparties in IRS and there is an exchange of
principal in Currency Swaps.
B. At maturity there is no exchange of principal between the counterparties in Currency Swaps and there is an
exchange of principal in IRS.
C. The counterparty in an IRS needs to consider fluctuation in exchange rates, while currency swap
counterparties are only exposed to fluctuations in interest rates.
D. Currency swaps counterparties are exposed to less counterparty credit risk due to offsetting effect of
currency risk and interest rate risk embedded within the transaction.

Valuation of currency swaps
Just like IRS this swap can be valued using bonds approach and FRA approach
Valuation using bonds
Party 1 is receiving payments in Rupees while paying in AUDs. Hence we can say that he is long a rupee bond
and short an AUD bond
The value of the swap will be the difference in the PV of the bonds
Vswap = BRs S0BAUD
Where:
S0 is the current spot exchange rate between Rs and AUDs
Valuation as a portfolio of forward contracts
In this case we determine the forward exchange rate at each point when the swap payments occur
The foreign currency is converted using the forward exchange rate
In the example above the 1 year, 2 year, 3 year, 4 year forward rate for USD-AUD exchange is used for
converting AUD cash flows to USD every year
This is then discounted back to the present value to give the value of the swap
Question
The USD interest rate is 4% per annum and the AUD rate is 6% per annum. Assume that the term
structure of interest rates is flat in the US and Australia. Assume current value of AUD to be $0.91.
Company ABC, under the terms of a swap agreement, pays 7% per annum in AUD and receives 3%
per annum in US$. The principal in the US is 10million USD and that in Australia is 11million AUD.
Payments are exchanged each year and the swap will last for 3 more years. Determine the value of
swap assuming continuous compounding in all interest rates.

Solution
Valuation of currency swap in terms of bonds (millions):
Time Cash Flow ($) Present Value Cash Flow (AUD) Present Value
1 0.3 0.2885 0.77 0.7264
2 0.3 0.2774 0.77 0.6853
3 0.3 0.2667 0.77 0.6465
3 10.0 8.8900 11 9.2358
Total 9.7225 Total 11.2940
Value of swap in million $ = 11.294*0.91 9.7225= $0.5448 million
Credit Risk

Credit risk (Covered in detail in VaR later)
A financial institution has a credit risk exposure from a swap only when the value of the swap is
greater than zero
Potential losses from a swap are much less than losses from defaults on a loan with the
same principal
Potential losses from a currency swap are much higher than losses from an interest
rate swap
Other Types of Swaps

Other types of swaps
LIBOR is the most common floating rate in IRS; however there can be other floating rates like,
commercial paper (CP) rates
In floating for floating swaps: rates of one type (LIBOR) can be swapped with floating rates of
another type (CP)
In an amortizing swap the principal amount reduces in a predetermined amortization rate
In a step up swap the principal increases in a predetermined way
In Credit Default Swaps (CDS) the buyer of the swaps pays premium to the seller of the swap till
the time the underlying does not default. If the underlying defaults then the seller of the swap
makes a payment to the buyer and the CDS is terminated
In a compounding swap the interest on one or both sides is compounded forward to the end of
the life of the swap and there is only one payment at the end of the contract
In a fixed for floating currency swap the fixed rate of interest in one currency is swapped for a
floating rate of interest in another currency
An equity swap is an agreement to exchange the total returns (dividends and capital gains) from
an equity index for a fixed/floating rate of interest
In a puttable swap one party has the option of terminating the contract early
Swaptions are options on swaps which provide one party with the right at a future time to enter
into a swap where a predetermined fixed rate is exchanged for floating
Question
Which of the following achievable swap positions could be used to transform a floating-rate asset
into a fixed-rate asset?
A. Receive the floating-rate leg and receive the fixed-rate leg of a plain vanilla interest-rate swap
B. Pay the fixed-rate leg and receive the floating-rate leg of a plain vanilla interest-rate swap
C. Pay the floating-rate leg and pay the fixed-rate leg of a plain vanilla interest-rate swap
D. Pay the floating-rate leg and receive the fixed-rate leg of a plain vanilla interest-rate swap

Answer
D. Pay the floating-rate leg and receive the fixed-rate leg of a plain vanilla interest-rate swap
FMP-IV
Backed Securities
Agenda
Introduction
Commodity futures and forwards
The commodity Lease rate
Storage Costs and Forward Prices
Pricing with convenience
Hedging oil costs
Hedging production costs
Strip and stack hedges
Introduction commodity spot and futures markets

Bill of lading is a document that mentions the commodity owner and acknowledges that the
goods have been received as cargo and are ready for delivery
The major risks involved with commodity transactions are:
Price risk: Risk of downward movement in price. Futures/Forward contracts reduce this risk
Transportation risk: Consists of two risks:
1. Ordinary: Deterioration, spoilage, accident etc.
2. Extraordinary: wars, riots, strike etc.
Delivery risk: Parties may withdraw from delivery. This risk has been greatly decreased by robust
practises by clearing houses
Credit risk: Counter party risk which is mainly an issue in spot market
Commodity markets also, like financial markets consists Hedgers, Speculators and Arbitrageurs
Hedgers are generally farmers/ranchers who want to lock in a price

Basis risk in commodity futures
Basis is the difference between spot price and the price of commoditys future contract at any
given time
Changes in basis is due to changes in cost of carry of the asset. Basis risk is generally represented
by the volatility / variance of the basis over time
2S(t)-F(t) = 2S(t) + 2f(t) - 2S(t)f(t) s,f
V s2( t ) f ( t )
Hedge effectiveness = 1
V s2( t )
Commodity Forwards
Commodity forward prices can be described using the same formula as used for financial
forward prices
F0,T S0 e(r G )T
For financial assets, G is the dividend yield
For commodities, G is the commodity lease rate
The lease rate is the return that makes an investor willing to buy and lend a commodity
Some commodities (metals) have an active leasing market
Lease rates can typically only be estimated by observing forward prices

Futures term structure
The set of prices for different expiration dates for a given commodity is called the forward curve
(or the forward strip) for that date
If on a given date the forward curve is upward-sloping, then the market is in contango
If the forward curve is downward sloping, the market is in backwardation
Note that forward curves can have portions in backwardation and portions in contango
F0,T S0 e(r G )T
Since r is always positive, assets with G =0 display upward sloping (contango) futures term structure
With G >0, term structures could be upward or downward sloping
A commodity loan
If you loan a commodity, you are giving up S0 today, and will get back St
If loan is fairly priced, its NPV = 0
NPV = E0(St)e-d S0
Where is required return on the commodity
Now, suppose commodity price grows at rate g, E0(ST)= S0egT
Then, NPV = S0e(g-d S
If NPV<0
this is common for commodities (supply with near-perfect elasticity)
Therefore, to make loan feasible, you would require lender to pay you the -g difference
This would get NPV back to zero
If 1 unit is loaned, you receive a lease payment of e-g)T units, and NPV = 0

The Commodity Lease Rate
The lease rate (G) is the difference between the commodity discount rate, D, and the expected
growth rate of the commodity price, g
For a commodity owner who lends the commodity, the lease rate is like a dividend
With the stock, the dividend yield, G, is an observable characteristic of the stock
With a commodity, the lease rate, G l, is income earned only if the commodity is loaned. It is not
directly observable unless there is an active lease market
Gl Dg
Commodity loan (cont.)

With the addition of the lease payment, NPV of loaning the commodity is 0
The lease payment is like the dividend payment that has to be paid by the person who borrowed
a stock
Therefore:
F0,T S0 e(r G )T
t

Forward Prices and the Lease Rate
The lease rate has to be consistent with the forward price
Therefore, when we observe the forward price, we can infer what the lease rate would have to be
if a lease market existed
The annualized lease rate
The effective annual lease rate
1
Gl r In (F0,T / S)
T
(1 r)
Gl 1
(F0,T / S)1/T
Storage and Carry Markets

A commodity that may be stored is said to be in a carry market
Reasons for storage
There is seasonal variation in either supply or demand (e.g., some agricultural products)
There is a constant rate of production, but there are seasonal fluctuations in demand
(e.g., natural gas)

Storage Costs and Forward Prices
One will only store a commodity if the PV of selling it at time T is at least as great as that of
selling it today
Whether a commodity is stored is peculiar to each commodity
If storage is to occur, the forward price is at least
Where O(0,T) is the future value of storage costs for one unit of the commodity from time 0 to T
F0,T t S0 erT O (0,T )
Storage Costs and Forward Prices (contd)

When there are storage costs, the forward price is higher. Why?
The forward price must compensate a commodity holder for both the financial cost of carry
(interest) and the physical cost of carry (storage)
With storage costs, the forward term structure can be steeper than the interest term structure
Convenience z
Some holders of a commodity receive benefits from physical ownership (e.g., a commercial user)
This benefit is called the commoditys convenience yield
The convenience yield creates different returns to ownership for different investors, and may or may not be
reflected in the forward price
Convenience and leasing
If someone lends the commodity they save storage costs, but lose the convenience
Stated as (O c)
Therefore, commodity borrower pays a lease rate that covers the lost convenience less the storage costs:
G=cO

Pricing with convenience
So, if:
F0,T S0 e(r G )T
And if, G = c O
Then, F0,T = S0e(r+ O -c)T
No-Arbitrage with Convenience

From the perspective of an arbitrageur, the price range within which there is no arbitrage is:
S0 e(r O c)T d F0,T d S0 e(r O )T

Where c is the continuously compounded convenience yield
The convenience yield produces a no-arbitrage range rather than a no-arbitrage price. Why?
There may be no way for an average investor to earn the convenience yield when engaging
in arbitrage

Question
Suppose that the price of corn is $2.20/bushel, the effective annual interest rate is 4.6%, and
effective annual priced storage costs are 10% of the current price/bushel. What is the 6-month
forward price?
Solution
&
Now suppose the holder of the asset realizes a convenience yield of 2%. What is the price?
&- 0.02) 0.5 = $2.34
The futures price dropped because the cost of carrying corn dropped

Hedging oil costs?
Suppose we are scheduled to purchase 15,000 bbls of oil in July 2008. The current futures price is
$105/bbl, and each contract covers 1,000 bbls. If we hedge, what is our cost of oil if the spot price
of oil in July 2008 is $70/bbl or $120/bbl?
Our natural exposure is short, therefore hedge long
E
Payoff = -15,000ST ^T 105)
Payoff = -$1,575,000, at any future oil ST
Hedging production costs

Suppose oil is a major component of our total production costs which equal $40 million, but it is
not the only component. In general, our production costs rise/fall with sensitivity of 0.72
(beta=0.72) to oil. Each crude oil contract is on 1,000bbls. Suppose S0=108 and F=105.
Now, how many contracts do we use to hedge?
E
Suppose oil goes up by 10% from 108 to 118
Increase in production costs = 0.072*40 = $2.88 million
Payoff from forwards = 274.28*1,000 (118-105) = $2.88 million
Note, we now have basis risk the basis for our hedge does not match the hedging instrument
perfectly what if our relation is not 0.72?

Strip and Stack Hedges
In the last example, we bought 450K bbls forward
This might be one component of a strip hedge if we are selling forward in other periods as well
In a stack hedge, we enter near-term contracts sufficient to cover the present value of future
obligations
We then roll the hedge into new contracts as the near-term contracts expire
FMP-IV
Backed Securities
Agenda
Introduction
Sources of profits and losses
Unhedged foreign assets and liabilities
Balance sheet hedging
Interest rate parity
Introduction to Foreign Exchange Risk

Sources of FE Risk:
Large financial institutions hold significant foreign currency assets and liabilities and also buy/sell
significant amount of foreign currency.
An institutions actual exposure to any given currency is its net exposure to the currency.
For example, a banks net Great Britain Pound (GBP) exposure is given by:
net GBP exposure = (GBP assets GBP liabilities) (GBP bought GBP sold)
i.e., net GBP exposure = net GBP assets net GBP bought
A positive net exposure (net long) is subject to the risk that the foreign currency will fall while a
Negative net exposure (net short) is subject to the risk that the foreign currency will rise.
Thus if the institution fails to maintain a position where net assets matches net liabilities in the
foreign currency, its exposed to the risks due to fluctuations in that currency

Sources of profits and losses on foreign exchange trading
A financial institution derives profits from differences between income and costs of funds.
In foreign exchange markets, an extra dimension foreign exchange rate comes into play
increasing the volatility of net returns of the bank if unhedged
In foreign exchange markets, hedging is of two types:

Off balance sheet hedging
Balance sheet hedging is achieved when the financial institution matches maturity and currency in
the foreign asset-liability book
Off balance sheet hedging is done by taking a position in the forward market
Example: Unhedged foreign asset and liabilities

Consider the balance sheet of a US bank:
Assets Liabilities
USD 10 million 7% US loans, maturity 1 year USD 20 million 5% CDs maturity 1 year
USD 10 million equivalent 12% Euro loans,
1-year maturity
What is the return for the bank if:
1. Exchange rate is unchanged
2. If exchange rate of euro has fallen from 1.2 dollars/euro to 1.05 dollars/euro
3. If exchange rate of euro has risen from 1.2 dollars/euro to 1.35 dollars/euro

Solution
1. Average return on assets = (10(710(12) )/2 = 9.5%.

Cost of funds = 5%
Therefore, net return = 4.5%
2. Issued euro loans= 10,000,000/1.2= 8333,333 euros;

End of maturity at 12% interest, 8,333,333*1.12=9,333,333
Dollar value=9,333,333*1.05=9,800,000.. i.e 2% loss
Average return = (-27)/2= 2.5%
Cost of funds = 5 %
Net return= -2.5% (loss)
3.Euros received at the end of maturity = 9,333,333*1.35 = 12,560,000 i.e. 25.6%

Average return = (25.67)/2 = 16.3%
Cost of funds = 5%
Net return = 11.3%
In the previous example, matching the maturity duration but not the currency composition made
the returns very unpredictable. One way to minimize this risk is through balance sheet hedging.
Consider the modified balance sheet of the previous example after hedging
Assets Liabilities
USD 10 million 7% US loans, maturity 1 year USD 10 million 5% CDs, 1 year maturity
USD 10 million equivalent 12% Euro loans, USD 10 million 9% euro CDs, 1 year maturity
1-year maturity

Now even if the euro falls from 1.2 USD/euro to 1.05 USD/euro, the bank can lock in a
positive return
Steps:
1. The bank borrows USD 10 million equivalent of Euros for a year at 9%. i.e., 10/1.2 = 8,333,333 Euros
2. Pays back the Euro CD holders at the end of maturity i.e., 8,333,333*1.09=9,083,332 Euros. Euro depreciated
to 1.05 USD i.e., 9083,332*1.05 = 9,537,499 dollars. i.e., cost of funds = -4.6%
3. As calculated in previous problem, it receives an USD equivalent of 9,800,000 from the 12% euro loans
granted. i.e., -2%
Average return on assets = (-
Average cost of funds = (-
Net return = 2.3%
Interest rate parity
In the previous examples, the Euro loans have better return than US loans and lead to arbitrage
argument
As more banks move to euro loans, the spot exchange rate for buying euro will rise because of
excess demand of Euro
In equilibrium, the forward exchange rate falls to completely eliminate the attractiveness of Euro
investments. This is called Interest Rate Parity (IRP)
T
1 rDC
Forward Spot
1 rFC
Where; rDC = Domestic currency rate

rFC = Foreign currency rate

Example
A japanese investor can invest in Japanese Yen at 4.5% or in GBP at 4.67%. Current spot rate is
0.01 JPY/ GBP. Calculate 1 year forward rate in JPY/GBP.
Solution
&ZdcZfc) = 0.01 * (1.045/1.0467) = 0.00998 Japanese yen per pound

FMP-IV
Backed Securities
Agenda
Introduction to Central Counterparties (CCPs)

Exchanges, OTC Derivatives, Derivative Product Companies (DPCs), Monoline s and Credit
Derivative Product Companies (CDPCs) and SPVs
Basic Principles of Central Clearing
Risk caused by CCPs

What is a CCP (Central Counterparties)?
CCP is a financial intermediary which helps in trade clearing process. Clearing is a process that occurs
after execution of a trade in which a CCP may step in between counterparties to guarantee
performance.
Main function of CCP: To interpose itself directly or indirectly between counterparties to assume
their rights and obligations by acting as buyer to every seller and vice versa. This means that the
original counterparty to a trade no longer represents a direct risk, as the CCP to all intents and
purposes becomes the new counterparty.
The general role and mechanics of a CCP are:
Sets standard for its clearing members
Takes responsibility for closing out all the positions of a defaulting clearing members
Maintains financial resources to cover losses in the event of a clearing member default in the form
of:
1. Variation Margin: To closely track market movements
2. Initial Margin: To cover the worst case liquidation or close out costs above the variation margin
3. Default Fund: To mutualise (risk sharing) losses in the event of a severe default
What is a CCP (Central Counterparties)? Cont

Mechanism to withstand extreme situation by way of :
1. Additional calls to the default fund.
2. Variation margin gains haircutting.
3. Selective tear-up of positions.
Benefits of Central clearing and CCPs

1. Netting: Reduces counterparty risk unlike bilateral market.
2. Margining: Protects against the adverse movements of the market and helps uphold the settlement of the
trade.
3. Loss mutualisation: Risk sharing mechanism helps protect the financial systems from contagion of default
risk originating from a loss of one or more counterparty.
4. Liquidity: Orderly close out of the position by way of proper auctioning mechanism along with netting
reduces price impact of unwinding of a large loss making position and enhances liquidity.

Possible Drawbacks of CCPs
Moral Hazard: As OTC derivatives mandatorily getting cleared by CCPs, their importance in the financial
system increases and they become systematically important institutions which , in the event of a
default, are likely to be bailed out by the government to limit the contagion it can create to the entire
financial market and eventually to the real economy. The systematically important status of CCP may
make it take undue risk leading to moral hazard problem.
Imposing cost and potential instability: The margin requirement can significantly add cost of the
participant and initially may impact the profitability of the participant financial institution and eventually
impact the economic growth in general. Also, in the adverse market condition, margin calls can put
further downward pressure on already depressed prices and lead to further volatility in the market and
create instability in the market.
Problem of loss mutualisation: Risk sharing mechanism of homogenising the credit risk among all the
members may significantly disadvantage the more credit worthy market participant. The most credit
worthy market participant may see less advantage of their stronger credit quality with CCP clearing.
Hence, this kind of system leads to adverse selection problem wherein the firms that trade OTC
derivatives know more about the risk of a particular cleared product than the CCPs. The informationally
advantaged party will over trade the product for which CCP underestimates the risk and under trade the
product where CCP over estimates the risk. Loss mutualisation leads to over trading by the party who
knows that in the event of his default, losses would be shared by all the members of CCP.
Possible Drawbacks of CCPs Contd

Increase in systematic risk: One of the major advantage of CCP was that it reduces systematic risk and
protects the financial system from contagion effect. However, being the single point counterparty to all
trades, CCP also create risk concentration. And any significant market movement, may lead to failure of
CCP (although rare) which in turn leads to huge amount of systematic risk for the markets.
Characteristics of bilateral OTC derivatives trading, their role in recent financial crisis and regulatory
changes implemented after the financial crisis:
OTC derivatives are private contracts and often contains exotic features. The underlying asset on the
derivative can be interest rate, currency etc. OTC derivatives have counterparty credit risk that is
typically pronounced during the time of default of the counterparty as unwinding of positions put
downward pressure on the already falling prices. The same may lead increased volatility and illiquidity.
This counterparty risk has been historically managed by bilateral clearing process.
The risk with bilateral clearing process is the default of a counterparty can lead to significant systematic
risk which was evident in the financial crisis of 2007 and failure of AIG which was eventually bailed out
by the government.

Possible Drawbacks of CCPs Contd
Regulatory changes implemented after the financial crisis:
The regulator focus on improving transparency, reducing market interconnectedness and greater bank
capital requirement to shift the risk from large global financial institution and bilateral trading to
centrally cleared trading securities. The Dodd Frank Act in USA and EMIR in Europe proposed central
clearing through CCPs of all the standardized derivative contracts. Subsequently, they introduced margin
requirement also for the non-cleared OTC derivatives as well.
Exchanges, OTC Derivatives, DPCs, CDPCs and SPVs

Definition of an exchange: An exchange is a central financial centre where parties can trade
standardized contract at a specified price. An exchange promotes market efficiently and enhances
liquidity by centralizing trading in a single price. Exchanges have evolved from a normal trading
forum to a sophisticated financial centre with settlement and counterparty risk management
function.
An exchange performs a number of a function:

Product Standardization
Trading Venue
Reporting Services

Clearing and Forms of Clearing
Clearing: Clearing is the process of reconciling and matching contracts between counterparties from
the time the commitments are made until settlement.
Forms of Clearing:
1. Direct Clearing: Means bilateral reconciliation of commitments between the original two counterparties. For
example, if party A is in contract with party B where it is required to make a payment of USD 5000 and in an
another transaction between the same party, party B is require to make a payment of USD 4500 to party A.
Under the direct clearing mechanism, rather than exchanging two cash flows of USD 5000 and USD 4500
each, party A will make a payment of USD 500 as a final settlement for both the transaction.
2. Clearing Ring: Under such mechanism of clearing, counterparty exposure is reduced between three or more
parties. It is a voluntary mechanism but once the member joins it, they have to follow the rules of the
exchange and must accept each others contracts. For example, here D owns USD 100 to B and is expected
to receive the same amount from party C. By using clearing ring, party D can be removed and party c and B
can be made counterparty. Also, party A remains unaffected in the process. Clearing ring can mitigate
counterparty risk and simplify the dependencies of a members open position and allow them to close out
contracts more easily whereby increasing liquidity.
Clearing and Forms of Clearing Contd
Complete Clearing: Refers to clearing through CCP. All exchange traded contracts are currently
subject to central clearing. The CCP function can be operated either by exchange or provided to
the exchange by a third party.
CCP assumes all contractual responsibilities as counterparty to all contracts: Here party A owes
USD 125 to Party C and B owes USD 50 to party A. When CCP is introduced, party A will be
required to pay USD 75 to CCP being a central counterparty. It is evident from the below figure
that CCP clearly reduces the counterparty risk and facilitates settlement.

OTC Vs Exchange traded Derivatives
Comparison between exchange traded and OTC derivatives: Exchange traded derivatives are
standardized contracts with greater liquidity and regulation as compared to OTC contracts. OTC are
bilateral customized contracts which increases the hedging utility of the product by reducing basis
risk and risk of term mismatch between what is to be hedged and hedging instrument. Clearing of
OTC derivative is challenging due to its long term nature of contracts and late settlement.
Exchange traded contracts are settled through CCP reducing counterparty risk whereby OTC
contracts are generally settled bilaterally.
Exchange-traded Over-the-counter (OTC)
Terms of contract Standardised (maturity, size, Flexible and negotiable
strike, etc.)
Maturity Standard maturities, typically Negotiable and non-standard
at most a few months Often many years
Liquidity Very good Limited and sometimes very poor
for non-standard or complex
products
Credit risk Guaranteed by CCP Bilateral
Classes of OTC Derivatives
There are five classes of OTC derivatives:

1. Interest rate derivatives
2. Foreign Exchange derivatives
3. Equity derivatives
4. Commodity derivatives
5. Credit derivatives
Comparison of total notional o/s and market value of OTC derivative (in $ trillions) as on June13
Gross notional outstanding 's Ratio

Interest Rate 561.3 15.2 2.7%
Foreign exchange 73.1 2.4 3.3%
Credit default swaps 24.3 0.7 3.0%
Equity 6.8 0.7 10.2%
Commodity 2.4 0.4 15.7%

Classes of OTC Derivatives (Contd.)
Though interest rate derivatives market dominate the majority of the OTC market, counterparty risk
is a major concern for foreign exchange derivative like cross currency swap where principal is
exchanged at the beginning as well as at the end of the contract. Furthermore, volatility and wrong-
way risk is a major concern for credit default swaps.
It is worth noting that only notional value should not be considered to determine dominating
product as many contracts like Fixed for Floating interest rate swap never requires exchange of
principle amount which is considered notional. Hence, gross market value is a more useful indicator
to understand the derivatives market product mix. Hence, a ratio of Gross Market value to Gross
Notional Value is useful and the ratio should be relatively small and close to 3% for interest rate,
foreign exchange and credit default swap.
Counterparty risk mitigation in OTC markets

Counterparty risk mitigation technique : Capital requirement, regulation, netting, margining etc.
Other mechanism of controlling counterparty credit risk:
1. SPVs: SPV is a legal entity (e.g. a company or limited partnership) created typically to isolate a
firm from financial risk. SPVs have been used in the OTC derivatives market to protect from
counterparty risk. A company will transfer assets to the SPV for management or use the SPV to
finance a large project without putting the entire firm or a counterparty at risk. Jurisdictions may
require that an SPV is not owned by the entity on whose behalf it is being set up. An SPV
transforms counterparty risk into legal risk. The obvious legal risk is that of consolidation, which
is the power of a bankruptcy court to combine the SPV assets with those of the originator.
2. DPCs: The bilaterally cleared dealer-dominated OTC market were perceived inherently more
vulnerable to counterparty risk than the exchange-traded market. The DPCs evolved as a means
for OTC derivative markets to mitigate counterparty risk. DPCs are generally triple-A rated entities
set up by one or more banks as a bankruptcy-remote subsidiary of a major dealer, which, unlike
an SPV, is separately capitalised to obtain a triple-A credit rating. The DPC structure provides
external counterparties with a degree of protection against counterparty risk by protecting
against the failure of the DPC parent.
Advantage of DPC: Provides some of the benefits of the exchange based system while preserving
the flexibility and decentralisation of the OTC market

Counterparty risk mitigation in OTC markets (Contd.)
The Triple A (AAA) rating of DPCs typically depends on:
1. Minimizing Market risk: In terms of market risk, DPCs can attempt to be close to market- neutral
via trading offsetting contracts. Ideally, they would be on both sides of every trade as these
mirror trades lead to an overall matched book.
2. Support from a parent.
3. Credit risk management and operational guidelines: Restrictions are also imposed on
counterparty credit quality and activities (position limits, margin, etc.). The management of
counterparty risk is achieved by having daily mark-to-market and margin posting.
DPCs used defined triggers for their own failure through a pre packaged bankruptcy process,
which outlines the bankruptcy process and is simpler alternative to standard bankruptcy process.
Once is it is into bankruptcy, it either continues as a part of another firm or is terminated.
The question on their AAA rated status, their link with their parents (whose credit rating is worse
than itself), and advent of alternative AAA rated entities followed by global financial crisis
reduced the importance of DPCs.
Counterparty risk mitigation in OTC markets (Contd.)

Monoline: Monoline insurance companies were financial guarantee companies with strong credit
ratings that they utilised to provide credit wraps.
Monolines are well-capitalized entities with their AAA ratings supported by capitalization
requirement based on possible losses and related to the assets for which they provide guarantee.
They are generally highly leveraged and do not have to post margin. Many monoline companies fell
during financial crisis of 2007.

Lessons for Central Clearing
The history of SPVs, DPCs, Monolines provide the following valuable lessons for CCP:
Shifting priorities from one party to other will really help the system as a whole? CCPs give priority
to OTC derivatives counterparties but that will make other parties (bond holders) worse off and
may increase risk in other markets.
Reliance on precise sound legal system exposes it a flow in such a framework. Example of
bankruptcy ruling by court.
Unlike monoline and CDPCs, CCPs do not take up residual risk as they generally have matched
book for trades. Hence, CCPs do not have one-way market exposure.
Unlike Monolines and CDCPs, CCPs require initial and variation margin which reduces risk.
Basic Principles of Central Clearing

What is clearing: Clearing represents the period between execution and settlement of a
transaction. This period is short for Non-OTC derivatives while OTC derivatives can be for a period
of years to decades.
Functions of CCP: The primary role of CCP is to standardise and simplify operational processes .
CCP can reduce interconnectedness within financial markets which may lessen the impact of the
insolvency of a participant. Also, CCP being at the heart of the clearing increases transparency on
the positions of the members.
Concept of Novation: A legal process whereby the CCP is positioned between buyers and sellers.
Novation is the replacement of one contract with one or more other contracts. Novation means
that the CCP essentially steps in between parties to a transaction and therefore acts as an insurer
of counterparty risk in both directions. The viability of novation depends on the legal
enforceability of the new contracts and the certainty that the original parties are not legally
obligated to each other once the novation is completed. Assuming this viability, novation means
that the contract between the original parties ceases to exist and they therefore do not have
counterparty risk to one another

Basic Principles of Central Clearing (Contd.)
Margining: To cover the market risk of the trade they cover, CCPs require margin from its
members. CCPs charge two types of margin: Variation margin covers the net change in market
value of the members positions. Initial margin is an additional amount, which is charged at trade
inception, and is designed to cover the worst-case close out costs (due to the need to find
replacement transactions) in the event a member defaults.
CCPs generally set margin levels solely on the risks of the transactions held in each members
portfolio. Initial margin does not depend significantly on the credit quality of the institution
posting it: the most creditworthy institution may need to post just as much initial margin as others
more likely to default. Two members clearing the same portfolio may have the same margin
requirements even if their total balance sheet risks are quite different.
Auctions: CCP absorbs the Domino effect of a counterparty default by acting as a central shock
absorber and swiftly terminates all financial relations with the defaulting counterparty without
suffering any losses. CCP guarantees the performance of the trade of the surviving members by
replacement of the defaulted counterparty with one of the other clearing members for each trade.
This is typically achieved via the CCP auctioning the defaulted members positions amongst the
other members.

Loss Mutualisation: Under this, losses above the resources contributed by the defaulter are shared
between CCP members. The most obvious way in which this occurs is that CCP members all
contribute into a CCP default fund which is typically used after the defaulters own resources to
cover losses. Since all members pay into this default fund, they all contribute to absorbing an
extreme default loss.
What can be cleared?

The OTC derivatives markets have wide range of products including standardized, non-standardized
products and exotic derivatives
There are four stages of central clearing history:
1. Long history of central clearing (IRS)
2. Short history of central clearing (Index CDS)
3. May soon be centrally cleared (Interest rate swaptions, CDS)
4. Products that will be never centrally cleared (Exotic Derivatives)

Conditions for a transaction to be centrally cleared:

1. Standardization
2. Complexity: Only vanilla (or non-exotic) transactions can be cleared as they need to be relatively
easily and robustly valued on a timely basis to support variation margin calculation.
3. Liquidity: Liquidity of a product is important so that risk assessments can be made to determine
how much initial margin and default fund contribution should be charged. In addition, illiquid
products may be difficult to replace in an auction in the event of the default of a clearing
member. Finally, if a product is not widely traded then it may not be worthwhile for a CCP to
invest in developing the underlying clearing capability because they do not stand to clear enough
trades to make the venture profitable.
Who can Clear?
There are requirements to be a member who in turn can only transact with a CCP.
The requirements fall into the following category:
1. Admission criteria
2. Financial commitment
3. Operational

Number of CCPs: A large number of CCPs will maximise competition but could lead to a race to the
bottom in terms of cost, leading to a much more risky CCP landscape while having a small number of
CCPs is beneficial in terms of offsetting benefits and economies of scale. A single global CCP is
optimal but not feasible due to following reason:
1. Regional
2. Product
Should CCPs be utilities or profit making organizations?
Due to a very significant systematic role played by CCPs, they need to be resilient . Hence, a utility
led long term stable business model is preferred over short term profit maximizing goal. However, it
can also be argued that it could also be argued CCPs will need to have the best personnel and
systems to be able to develop the advanced risk management and operational capabilities.
Moreover, competition between CCPs will benefit users and provide choice. Expertise and
competition implies that CCPs should be profit-making organisations.

Advantages of CCPs
Transparency
Offsetting
Loss Mutualisation
Legal and Operational efficiency
Liquidity
Default Management
Disadvantages of CCPs
Moral hazard
Adverse Selection
Bifurcation: The requirement to clear standard products may create unfortunate bifurcations
between cleared and non-cleared trades. This can result in highly volatile cash-flows for
customers, and mismatches (of margin requirements) for even hedged positions.
Procyclicality: CCPs may create procyclicality effects by increasing margins (or haircuts) in volatile
markets or crisis periods. The greater frequency and liquidity of margin requirements under a CCP
(compared with less uniform and more flexible margin practices in bilateral OTC markets) could
also aggravate procyclicality.

Comparing OTC derivatives market with CCP and exchange-
traded market
OTC CCP Exchange

Trading Bilateral Bilateral Centralised
Counterparty Original CCP
Products All Must be standard, vanilla, liquid, etc.
Participants All Clearing members are usually large
dealers. Other margin posting entities can
clear through clearing members.
Margining Bilateral, bespoke arrangements Full margining, including initial margin
dependent on credit quality and enforced by CCP
open to disputes.
Loss buffers Regulatory capital and margin Initial margins, default funds and CCP
(where provided) own capital
Risk caused by CCPs

Default Risk
1. Default or distress of other clearing members
2. Failed Auctions
3. Resignations
4. Reputational

Risk caused by CCPs (Contd.)
Non default loss events:
1. Fraud
2. Operational
3. Legal
4. Investment

Model Risk: Model risk arises due to valuing complex derivative products to calculate initial and
variable margin requirements
Liquidity Risk
Operational and legal risk
Other risk:
1. Settlement and Payment
2. Forex Risk
3. Custody Risk
4. Concentration Risk
5. Sovereign Risk
6. Wrong-way Risk

How to keep CCPs safe?
Being a central clearer, CCP creates huge concentration of risk. CCPs should not become overly
competitive during buoyant market and increase the likelihood of falling during volatile market and
crashes.
CCPs are systematically important and their failure lead to cross border impact due to global nature
of derivatives markets. The failure of CCP can be termed a bigger failure than a bank.
If the failure of a single large participant in the OTC derivatives market is capable of endangering the
entire financial system then so is the failure of a CCP that clears OTC derivatives. Therefore,
governments are likely to have little choice as to whether or not to support a failing OTC CCP. This
realisation is problematic since taxpayers bailing out a CCP is no better than bailing out other
financial institutions such as banks. Indeed, a CCP bailout represents a bank bailout of sorts since it
protects the banks that are CCP members (that may be viewed as having taken an excessively large
exposure to the CCP.
FMP-IV
Swaps, Commodities, Foreign Exchange, Corporate
Bonds and Mortgage Backed Securities
Agenda
Introduction to corporate bonds
Interest payment classification
Retiring of bonds before maturity
Credit risk
Default rates
Corporate Bonds
A corporate bond is a debt instrument that obligates the issuer to pay an indicated percentage of
the bonds face value on designated dates and repay the bonds face value at maturity
In the event that either the interest or principal payments are not paid, the bond is in default
Bondholders have a higher priority for the issuing companys income over preferred and common
shareholders
In the United States, corporate bonds are issued in denominations of $1,000 and multiples thereof
Indenture is a contract that states the promises of the corporate bond issuer and the rights of the
bond holder
As the indenture is hard to interpret, a third party called the Trustee is introduced
The basic functions of a trustee are:
To authenticate the bonds issued; the trustees keep a record of all the bonds sold and ensure they do not
exceed the principal amount stated in the indenture
The trustees ensure that the issuing firm adheres to all the covenants of the bonds indenture

Interest Payment Classifications
There are a variety of bonds based on the interest payment characteristics:
Straight Coupon Bonds: These are also known as fixed-rate bonds. The interest rate received on
these bonds is called the coupon. Most straight coupon bonds pay interest semiannually. For
example, for a 6% coupon rate bond with a face value of $1000 would pay $30 every six months
Some bonds, known as participating bonds, receive payments that are greater than the coupon.
These payments depend on the profits of the issuer
Income bonds pay coupon interest if earnings of the issuing company are sufficient. It is not
mandatory. Failures to pay interest or the principal of these bonds do not indicate a default
Zero Coupon bonds are bonds without coupon payments. They only have a principal payment at
maturity. These bonds are issued at a discount to par. The difference between the face value and
the issue price of the bond is known as the Original-Issue Discount (OID)
Deferred Interest Bonds (DIBs), generally issued by non-investment grade companies, dont need
to pay interest for the first several years and then pay semi-annually till maturity
Pay In Kind (PIK) Bonds are similar to DIBs but rather than accreting the original discount,
additional pieces of the same security are issued
Interest Payment Classifications

Floating Rate Bonds: The interest payments by these bonds are computed in reference to a
reference rate such as the 6- month LIBOR. The issuer may add a spread to the LIBOR
Collateralized Corporate Bonds
Mortgage Bonds: A mortgage bond grants the holder a first-mortgage lien on its property. This
means, if the mortgage payments are not made timely, the bond holder has the right to sell the
property. The underlying real estate property is the collateral for the bond
Collateral Trust Bonds: In case a company wants to issue bonds but does not have a fixed asset or
property base, it can pledge securities of other companies which it owns
Equipment Trust Certificates: A Bond issued by railway companies where the collateral is cars and
locomotives is an example of Equipment Trust Certificates (ETCs). This method of financing is
called rolling stock
Debenture Bonds: Debenture bonds are unsecured bonds. Most corporate bond issues are
debentures. They are traded at higher yields than secured debt
Debenture bonds however, have a general claim on the assets of the issuer that are not pledged
specifically to secure other debt. Debenture bonds are issued by companies who have strong
credit ratings
Convertible Debentures give the bondholder the right to convert the debenture into
common stock

Retiring of Corporate Bonds before Maturity
Retiring of Corporate Bonds before Maturity
Corporate bonds can broadly be retired before maturity in two ways namely those mechanisms
that are included in the bonds indenture and those that are not included in the bonds indenture.
The methods included in the bonds indenture are:
Call and refunding provision
Sinking Funds
Maintenance and Replacement Funds
Redemption through sale of assets
A method not indicated in the bonds indenture is the fixed-spread tender offers.
Call and Refunding Provisions
The right that the issuer has to buy back the bonds in whole or part before maturity is known as a
call provision. A bondholder would demand a higher yield to buy a callable bond, all else equal.
Callable Bond Cost = Option Free Bond Cost + Value of Embedded Option
Methods for retiring bonds

Fixed Price Call Provision
The fixed price a bond can be called at its call price. Normally, a bonds indenture would contain a
call price schedule that typically starts at a high premium above par and converges to par at
maturity
Bondholders can be protected from a call in one of two ways:
Some bonds cannot be callable for the first few years
Some bonds prohibit the issue to be refunded (at a lower cost) for the first few years non-refundable bonds
Make-Whole Call Provision
The call price is not fixed in this provision. The call price is determined as the present value of cash
flows at a particular discount rate. This discount rate would be the yield on a comparable maturity
Treasury bond with a spread. This spread is known as the make-whole call premium
There are two ways to arrive at the Treasury bond yield:
Using the Constant-Maturity Treasury rate, published weekly by the FED, for a maturity closest to that of the
issue or a linear interpolation of these rates. This is a more common method
The yield of a treasury that has similar maturity as that of the issue. The selection of this bond would be a
primary US Treasury dealer mentioned in the bond indenture

Methods for retiring bonds
Sinking Fund Provision: Bonds are retired periodically rather than retiring the entire issue at
maturity. The terms for the sinking fund would be mentioned in the indenture
For example, if $20 Million is the notional principal for a bond with a 20 year maturity, the sinking
fund provision may state to retire $5 Million every 5 years
From a bondholders perspective there are two advantages of a sinking fund provision:
Reduced default risk
Bond price increases near retirement dates as a result of excess demand because of the issuer buying bonds
in the open market
However the sinking fund provision can be a disadvantage to the bondholder if their bond is to be
retired when bonds are selling above par in the market
An accelerated sinking fund provision is when the issuer is granted the right to retire more bonds
than indicated. This reduces the bondholders call protection
Maintenance and Replacement Fund is mainly used by electric utilities companies. These
companies retire their bonds for the maintenance and repair of the pledged collateral
The tender offer method of retiring bonds is one that is not specified in the bonds indenture. In
this method, a firm sends a tender offer and announces its aim of buying back its debt issue. The
firm sends a circular to all the bond holders stating the price at which it is willing to pay for the
security
Credit Risk
Credit Default Risk : The uncertainty whether the issuer can pay interest and principal payments
in a timely manner
Credit Spread Risk: A credit spread is the difference between a bonds yield and the yield of a
comparable maturity benchmark of a Treasury security
Credit spread risk is the risk of a loss in the value of a bond from changes in the level of credit spreads used in
the marking to market
Factors affecting credit spread risk:
Macroeconomic factors
Level and slope of treasury yield curve
Business cycle
Consumer confidence
Issue-Specific factors
Corporations financial position
Future prospects of the firm and its industry
A measure of credit spread risk is spread duration. Spread duration is the change in the value of a bond for
1% change in credit spread, assuming the underlying treasury securitys yield is constant

Default rates
Number of issuers that default

Issuer default rate =
Total number of issuers at the beginning of issue
Cumulative dollar value of all defaulted bonds

Dollar default rate =
Cumulative $ value of all issuance *
Weighted Avg. number of years outstanding
Cumulative dollar value of all defaulted bonds

Cumulative annual default rate =
Cumulative dollar value of issue
FMP-IV
Swaps, Commodities, Foreign Exchange, Corporate
Bonds and Mortgage Backed Securities
Agenda
Residential Mortgage Products and Types
Mortgage Payments: Fixed rate, Level-Payment
Securitization: Mortgage Pass-through Securities
Payment/Pre-payment rates
Prepayment
Dollar Role Transaction
Prepayment Modeling
Dynamic Valuation: Monte Carlo Stimulation
Dynamic Valuation: Option Adjusted Spread
Residential Mortgage Loans

Mortgage Loan:
Mortgage is a loan secured by some sort of real estate property as a collateral
Borrower is obliged to make a predetermined series of payments or EMI in future.
Prior to 1970 mortgage were restricted solely to primary market.
MBS (Mortgage-backed security):
Mortgages are polled and packages to investor in Secondary Market through Securitization
Payments follow pass through structure.

Residential Mortgage Loan Types
Residential Mortgage Loan Types can be classifies based on various properties of loans:
Lien Status: Lien status of impacts the lenders ability to recover the balance owed in the event of
default.
First Lien is senior than other subsequent i.e. In event of default lender of first lien loan will have first
right to receive proceeds.
Original Loan Term: Loan term commonly varies from 10-30 years with 30 year loan being most
common. Medium term loan of 10-20 year maturity recently gaining popularity.
Residential Mortgage Loan Types (Contd.)

Credit Classification:
Prime (A-grade) loans: FICO score 660 or greater, Low loan-to-value ratios (<95%).
Subprime (B-grade) loans: FICO score less than 660, High loan-to-value ratios (>95%)
Alternative-A loans: Between Prime and Subprime, essentially prime loans but certain characteristics
makes then riskier (incomplete documentations, LTV on higher end)
Interest Rate Type:
Fixed-rate mortgages: Both interest and payments are constant.
Adjustable-rate mortgages (ARM): Variable interest payments, mostly linked to market interest rate
(LIBOR, OTC etc)
Prepayments and Prepayment Penalties:
Credit Guarantees:

Mortgage Payments: Fixed rate, Level-Payment
Example: To buy a home an individual borrows from a bank,$100,000 which is secured by that
home. To repay the loan the individual agrees to pay the bank $804.62 per month for 30 years.
Interest =9%
The payments are called level because the monthly payment is same every month
360
1
$804.62 $100,000
1 (1 y 12)
n
n
Interest rate on a mortgage is defined as the monthly compounded yield-to-maturity of the
mortgage
360
X
1
$100,000
1 (1 0.09 12)
n
n
Every monthly mortgage payment or EMI is due on the first of each month, and consists of interest
on the outstanding mortgage balance and repayment of portion of outstanding mortgage loan B(n)
y
d B(n) u
12
y
The principal component of the payment is the remainder: EMI - B(n) u
12
Mortgage Payments: Fixed rate, Level-Payment (Contd.)

The portion of EMI towards interest payment decreases over the tenure of a loan, whereas portion of
EMI towards principal component increases over the life of a outstanding mortgage loan
Payment Months Interest Payment Principal Payment Ending Balance Interest Principal EMI
900
1 750.00 54.62 99,945.38
750
60 719.74 84.88 95,880.14
600
120 671.72 132.9 89,429.74
450
180 596.54 208.08 79,330.49
300
240 478.83 325.79 63,518.27
150
300 294.54 510.09 38,761.39
0
360 5.99 798.63 0.00
1
19
37
55
73
91
109
127
145
163
181
199
217
235
253
271
289
307
325
343
Payment Allocation Between Principal and Interest:

Crossover point is the point where principle and interest allocation is same. Post this point more amount is
allocated to principal.
Mortgages with shorter amortization period result in less interest paid and more of the payment applied
toward reducing the principal balance sooner.
Servicing fees is the fees charged by servicer or originators for servicing of mortgage i.e.
administrative work which involves collecting EMIs, initiating foreclosures etc.
Servicing fee forms the part of each payment and reduces over period.

Securitization
Securitization Process
Mortgage Banks
Borrower Mortgage Backed
Security
Mortgage Pool of Mortgages Banks (Investment/
Borrower (collecting principal Commercial), Mutual
repayments and Funds, Pension House,
Mortgage
interest payments, Institutional/ HNI investors,
Borrower
also prepayments) Insurance Firms, etc
Mortgage
Borrower
Securitization can be structured in numerous ways to suit the Investors requirement
Mortgage pools can be segregated into Interest-Only (IO) tranche and Principal-Only (PO) tranche
Mortgage pools can have various classes to adjust prepayments (lowest class to receive highest
prepayment, etc)
Mortgage 1 Investor 1
Mortgage 2 Investor 2
Pool

Mortgage N Investor N
Passthrough securities backed by

the pool are issues to investors

A mortgage pass-through security is an asset-backed security or debt obligation that represents a
claim on the cash flows from the pool of underlying mortgage loans,held by the SPV.
E.g.. Securitization process: Servicing
25-50 bps
Mortgages are pooled by FNMA
Principal and Interest (P&I) from the underlying
assets (mortgages) is the security's cash inflow
$ 100Mn,
Servicing fee for collecting and dispersing 9% P&I FNMA P&I
Investors
payments is removed (25-50 basis points) 30 year
fixed
Default fee paid to credit enhancer is removed
Remaining cash is dispersed to investors.
Cash flow to a pass-through security investor depends on Default fee
the cash flow from pool of mortgage loans

Not all mortgage loans carries the same interest rate & same maturity and hence weighted
average coupon rate (WAC) & weighted average maturity (WAM) is calculated
Ginnie Mae is backed by U.S. federal government and carries the full faith & guarantee of U.S.
Govt.
Freddie Mac & Fannie Mae are U.S. government sponsored corporate entities and does not carry
full faith & guarantee of government
Payment / Pre-payment Rate Measurement

Calculate the dollar value of a pass-through security if the security is trading at $92 for a $1mn of
a pass-through with a pool factor of 80%
Pool factor of 80% indicates that 80% of original mortgage pool is still outstanding
Price u par value u pool factor 0.92 u $1,000,000 u 0.80 736,000
Single monthly mortality rate (SMM) is the ratio of the prepayment in a month and total
outstanding amount available to prepayment:
Prepayment in month(t)
SMM
Begining mortgage balance for month(t) - scheduled principal payment in month(t)
e.g. calculate the prepayment for a month, if a remaining mortgage balance is $250 mn with a
next month SMM of 0.55% and the scheduled principal payment for that month is $5 mn
Prepayment in month t 250,000,000 - 5,000,000 u 0.0055 $1,347,500

Conditional Prepayment Rate (CPR) is a annualized SMM rate which assumes that some amount
of prepayment is bound to happen apart from scheduled principal repayment
CPR 1 1 SMM 1 (1 0.0055)12 6.404%
12
Best predictor of CPR is past prepayment rates

It's called conditional since it's conditional on the remaining mortgage balance
Benchmark CPR = 100% PSA (Public Securities Association) is discussed in more detail on the next slide

Prepayment rate: Public Securities Association (PSA)
PSA prepayment benchmark is the monthly series of CPR
PSA benchmark assumes that prepayment is low for a new mortgages, but it will speed up until 30
months, remaining constant thereafter.
100 PSA assumes CPR of 0.2% for the first month and thereafter increases by 0.2% per month for the next
30 months
If t< 30 months, CPR = 6% x (month/30)
If t>30 months, CPR = 6%
Different prepayment characteristics are quoted as a percentage of PSA (e.g. 165% PSA has a
higher pre-payment rate. CPR is 65% greater than the average CPR)
The percent PSA increases as the yield decreases, which further increases prepayment risk (low
interest rate)
165 PSA 100 PSA
12
10
0
30
0
9
18
27
36
45
54
63
72
81
90
99
108
117
126
135
144
153
162
171
180
Mortgage Age in months
Prepayment rate: Example

Compute the prepayment for the 22nd month on a 150 PSA mortgage loans of $250mn and
scheduled repayment of $5mn
CPR 6% u 22 30
4.4% 150 PSA 1.5 u 4.4% 6.6%
1 1 6.6%
1
SMM 12 0.005674
Prepayment in 22 nd month 250,000,000 - 5,000,000u 0.005674 $1,390,130

Prepayment
Average life of a pass-through security depends on the prepayment assumptions:
T
t u Projected principal received at time(t)
Average Life
t 1 12 u Total Principal
Types of Mortgage Prepayments:
Increasing frequency or amount of payments.
Repaying/refinancing the entire outstanding balance.
Impact of prepayment on lender:
Loss on higher interest rate of Mortgage loan.
Reinvestment has to be generally made on lower market rate.
Factors influencing Prepayment:

Seasonality
Age of Mortgage pool
Personal
Housing Prices
Refinancing burnout
Prepayment ..Cont
$1,250.00
$1,200.00
$1,150.00
$1,100.00
Prepayment Mortgage
$1,050.00
Security
$1,000.00
$950.00
$900.00
$850.00 Standard Bond
$800.00
$750.00
7.0% 8.0% 9.0% 10.0% 11.0% 12.0%
Contraction risk refers to consequences resulting from a decline in interest rates for a pass-
through security. It results in faster prepayments leading to shortening of life
Limited upside potential for a pass-through security
Reinvestment risk that received cash flow have to be invested at a much lower interest rate
Extension risk refers to consequences resulting from a increase in interest rates.
It results in slower prepayments leading to lengthening of average life
Unlimited downside potential with a limited upside makes pass-through security unattractive

Dollar Roll Transaction
A dollar roll transaction: Buying position for one settlement months and selling those same positions
for another month at the same time.
Valuing dollar role:

The process involves calculating the income and expenses over the holding period.
Price drop between the two settlements makes purchase of security of back month more attractive.
Factors impacting Dollar role valuations:

Securitys coupon age and WAC
Holding period
Assumed prepayment speed
Funding cost in repo market
Factors Causing Dollar role to trade special: Decrease in back month price (Due to access sale by
originator) and Increase in front month prices (shortage of securities due to increase in demand of
certain maturity)
Prepayment Modeling
There are four major components of prepayment modeling:
Refinancing: Using proceeds of new mortgage to pay of f principal for existing mortgage.
Factors impacting refinancing :
Interest rate fall. Also known as Media effect as large declines in rate likely gain media attention.
Cash-out Refinancing: Increase in property value allowing borrowers to get more cash for new mortgage on same
property.
Refinancing burnout: In cyclic change in interest rate, most of the refinancing occurs on first dip with less people
opting to refinancing on subsequent dips.
Incentive function: Modeling any dollar gain that borrower will refinance.
Turnover: Refinancing caused due to sale of property.

Defaults: Modeling default requires an analysis of LVT and FICO scores and overall analysis of housing
market.
Curtailments: Prepayment due to curtailment depend on age of mortgage as partial payments tend
to occur mortgage is older and has relatively low balance.

Dynamic Valuation: Monte Carlo Stimulation
Monte Carlo Simulation: Its more process of steps than a specific model. Using various parameters
we try to determine various possible paths that can be taken by underlying variables and based on
that try to determine probability distribution of MBS values.
Steps for MCS valuation:

Stimulate interest rate and refinancing path.
Project cash-flows for each interest rate path.
Calculate PV of each cash flow path
Calculate theoretical value of the mortgage security
Dynamic Valuation: OAS

Option Adjusted Spread: Constant spread K when added to spot rate curve and resultant curve used
to discount all cash-flows give market value.
OAS is determined using Monte Carlo Stimulation to get various possible cash-flow path which are
then discounted using Spot rate OAS. OAS is thus determined iteratively.
Market Price = PV[path1] PV[path2] PV[path3] PV[path4] . PV[pathN] / N

Where N is number of paths
Option Cost = spread OAS

Here Option Cost represents value of prepayment risk.
Challenges with OAS

Modeling Risk
Adjustment required in interest rate path
Assumption of constant OAS over time.
Dependency on underlying prepayment model

Thank z
www.edupristine.com
Valuations and Risk Models-I
VaR Methods
Introduction to Risk
Risk can be broadly defined as the degree of uncertainty about future net returns
Credit risk relates to the potential loss due to the inability of a counterpart to meet its obligation
Operational risk takes into account the errors that can be made in instructing payments or settling
transactions
Liquidity risk is caused by an unexpected large and stressful negative cash flow over a short period
Market risk estimates the uncertainty of future earnings, due to the changes in market conditions
Broadly the standard deviation of the variable measures the degree of risk inherent in the variable
Say the standard deviation of returns from the assets owned by you is 50% and the standard
deviation of returns from assets I own is 0%. We can say that risk of my assets is zero
I own risk-less assets as My assets are very risky

the standard deviation of returns as the standard deviation of
of my assets is 0% returns of my assets is 50%.

Value at Risk (VaR)
Value at Risk (VaR) has become the standard measure that financial analysts use to quantify
this risk.
VAR represents maximum potential loss in value of a portfolio of financial instruments with a given
probability over a certain time horizon.
In simpler words, it is a number that indicates how much a financial institution can lose with
probability (p) over a given time horizon (T).
Say the 95% daily VAR of your assets is $120, then it means that out of those 100 days there would
be 95 days when your daily loss would be less than $120. This implies that during 5 days you may
lose more than $120 daily.
There may be a day out of 100 when your loss is $5000,

which means VAR doesnt tell anything about the extent to which we can lose
Visualizing VAR
Confidence (x%) Zy
0.45 Probability
90% 1.28
0.4
95% 1.65
0.35
97.5% 1.96
0.3
99% 2.32
0.25
95% daily-VAR
0.2
0.15
0.1
0.05
0
-4 -3 -2 -1 0 1 2 3 4
Mean = 0
The colored area of the normal curve constitutes 5% of the total area under the curve.
There is 5% probability that the losses will lie in the colored area i.e. more than the VAR number.

Measuring Value-at-Risk (VAR)
0.45
0.4
0.35
0.3
0.25
0.2
VARX % (in %) Z X % *V
0.15
0.1
0.05
0
-4 -2 0 2 4
Mean = 0
y
0 and standard deviation as 1)

VAR in absolute terms is given as the product of VAR in % and Asset Value:
VAR VARX % (in %) * Asset Value
This can also be written as:
VAR Z X % * V * Asset Value
Measuring Value-at-Risk (VAR)

VAR for n days can be calculated from daily VAR as:
VaR (n days) (in %) VaR (daily VaR) (in %) * n
This comes from the known fact that the n-period volatility equals 1-period volatility multiplied by
the square root of number of periods(n).
VaR (n days) (in %) Z X% * V * Asset Value * n
As the volatility of the portfolio can be calculated from the following expression:
V portfolio wa2 V a2 w 2b V b2 2w a w b * V a * V b * U ab
The above written expression can also be extended to the calculation of VAR:
VaR portfolio (in %) wa2 (%VAR a ) 2 w 2b (%VAR b ) 2 2w a w b * (%VAR a ) * (%VAR b ) * U ab

Question 1
Asset daily standard deviation is 1.6%
Market Value is USD 10 mn
What is VaR (%) at 99% confidence?
Solution
Daily VaR = 0.016 x 10 x 2.33 = 0.3728 mn

Question 2
What is the VaR value for 10 day VaR in the earlier case?
Solution
10 day VaR = 0.3728 x (10)^0.5 = 1.1789

Question 3
What is the daily portfolio VaR at 97.5% confidence level?

Investment in asset A is $40 mn
Investment in asset B is $60 mn
Volatility of asset A is 5.5% and asset B is 4.25%
Portfolio VaR if correlation between A and B is 20% ?
Solution
VaR(A)(in %) = 5.5 x 1.96 = 10.78%; VaR(B)(in %) = 4.25 x 1.96 = 8.33%;
Portfolio VaR = [(40 x 0.1078)2 60 x 0.0833)2 2x0.1078x0.833x40x60x0.20]0.5 = 7.22 mn

Extended Question 3.1
Portfolio VaR if
/
What if correlation is 1?
Or if -1?
What are the implications?
Question 4
Market Value of asset $10 mn

Daily variance is 0.0005
What is the annual VaR at 95% confidence with 250 trading days in a year?

Solution
Daily VaR = 10 x (0.0005)0.5 x 1.65 = 0.36895 mn
Annual VaR = 0.36895 x (250)0.5 = 5.834 mn
Question 5
For an uncorrelated portfolio what is the VaR if:

VaR asset A is $10 mn
VaR asset B is $20 mn

Solution
The VaR comes out to be 22.36 mn
Value-at-Risk Measurement Methods
VaR
Linear Valuation Method Full Revaluation Method
Delta Normal Historical Monte Carlo

Method Simulation Simulation

Delta Normal VAR: Linear and Non-Linear Assets
Linear: When the value of the delta is constant for any change in the underlying
Primarily in the case of forwards and futures we have linear assets
The method to calculate VAR for linear assets is called Delta Normal method
Delta Normal method assumes that the variables are normally distributed
Non Linear: When the value of the delta keeps on changing with the change in the underlying
asset
Options are non-linear assets, where delta-normal method cannot be used as they assume the linear payoff
of the assets
To calculate the VAR for non-linear assets, full revaluation of the portfolio needs to be done
Monte Carlo methods or Historical Simulation are commonly used to fully revaluate the portfolio
Delta Normal VAR: VaR for Linear Derivatives
Linear Derivatives: Payoff diagrams that are linear or almost linear:

Forwards, futures
linear
Payoff
0
k Share/asset price
Long position
Delta of Derivative: Change in price of Derivative to change in underlying asset
For example:
The permitted lot size of S&P CNX Nifty futures contracts is 200 and multiples thereof.
So VAR of Nifty Futures contract is 200 * VAR of Nifty Index
VARLinear Derivative ' *VARUnderlying Risk Factor

Delta Normal VAR: VaR for Non-Linear Derivatives
Main reason for difference is the shape of the
payoff curve Option
Price value
For Delta Normal VAR
A linear approximation is created
Approximation is an imperfect proxy for the portfolio
Computationally easy but may be less accurate.
The delta-normal approach (generally) does not work for Slope = delta
B
portfolios of nonlinear securities.
KsZKsZ A Asset/Stock price
Consider a portfolio of options dependent on a
single stock price, S. Define: f ' ' ( x 0)( x x 0) 2
f ( x) f ( x 0) f ' ( x 0)( x x 0)
2
'S 3
'x Delta (G )
S 6
Approximately: 'P G 'S SG 'x
For Many Underlying variables: 'P S G

i
i i 'xi
Question 6 (Linear Assets)
If the daily VaR at 5%of Nikkei is USD 0.8 mn and you have 100 lots of Nikkei contract.
Calculate annual VaR at 95% confidence for your portfolio assuming 250 days?

Solution
Solution = 0.8 x 100 x (250)0.5 = USD 1264.911 mn
Here, delta = 100, because for every 1 unit change in the Index Nikkei, the futures price will change
by 100 units because the lot size is 100
Question 7 (Non-Linear Assets)

If the value of stock is 100 and the value of the put option at 110 is 20. 10 units change in the
underlying brings in change of 4 units change in the option premium. If the annual volatility is 0.25.
Calculate daily VaR at 97.5% assuming 250 days?

Solution
Solution: Daily volatility = 0.25/(250)0.5 = 0.0158; Daily VaR = 100 x 0.0158 x 1.96 = 3.099;
Daily VaR of option = 0.4 x 3.099 = 1.239
Quantifying Volatility in VaR Model

Quantifying Volatility in VAR Model: Fat Tails
There are two explanations for the Fat Tails
Conditional Mean: Mean changing over the period of time 0.45
0.4
Conditional Volatility: Volatility changing over the period of 0.35
time 0.3
95%
daily-VAR
Market conditions may cause the mean and variances 0.25
0.2
to change over the period of time, which leads to 0.15
fat-tailed distributions 0.1
0.05
0
-4 -2 0 2 4
The fat-tailed unconditional distribution can be broken down into two conditional distributions,
either with similar means and different variances or similar variances and different means.
Many a times when we observe marked differences between the estimated and actual volatilities,
its a result of regime switching which means that the average volatility in the market has now
changed too much when compared to the previous estimate
Risk Management is all about understanding tails of distribution
Regime Switching
High Vol Estimate
1 n
V2 (R i E ( R)) 2
n i1 Actual volatility
30bp/day
18bp/day
Low Vol Estimate
Unconditional
1 n
volatility V2 (R i E ( R)) 2
n i1
10bp/day

Alternative Measure
In this case we observe distinct periods of high and low volatility.

A more accurate measure of risk would be to measure the risk or volatility in these high and low
risk areas separately, and then base our forecast depending on whether or not we are currently in
a high or low volatility period.
This implies, we would have the problem of assessing whether we were currently in a high or low
risk period.
Cyclical Volatility Of Financial Markets

Volatility
Time Varying
Sticky
Predictable
Magnitude of recent changes is informative
More information in recent past than distant data

Estimation of Volatility
Let xi be the continuously compounded return during day i (between the end of day i-1 and end
of day I).
Let be the volatility of the return on day n as estimated at the end of day n-1.
Variance estimate for next day is usually calculated as:
Variance = average squared deviation from average return over last n days
x
n
2
i x
i 1
Variance
n 1
Mean of returns (x-bar) is usually zero, especially if returns are over short-time period (say, daily returns). In
that case, variance estimate for next day is nothing but simple average (equally weighted average) of
previous n days squared returns.
n
x
2
i
i 1
Variance
n 1
What if the volatility is dependent on the values of volatility observed in the recent past?
What if they also depend on the latest returns?
EWMA & GARCH
This topic has already been covered in Quants Lectures

EWMA Model
In an exponentially weighted moving average model, the weights assigned to the u2 decline
exponentially as we move back through time.
This leads to: V n2 OV n2 1 (1 O )u n21
Apply the recursive relationship:
V n2 O >OV n2 2 (1 O )u n2 2 @ (1 O )u n21

V n2 > @
(1 O ) u n21 Ou n2 2 O2V n2 2
Hence we have:
m i 1 2
V 2
n (1 O ) O u n i OmV n2 m
i 1
Variance estimate for next day (n) is given by (1-
previous variance estimate.
Risk-metrics (by JP Morgan) assumes a Lambda of 0.94.
EWMA Model (cont.)

Since returns are squared, their direction is not considered. Only the magnitude is considered.
In EWMA, we simply need to store 2 data points: latest return & latest volatility estimate.
Consider the equation: V t21 (1 0.94) P t2 0.94V t2
In this equation, variance for time t was also an estimate. So we can substitute for it as follows:
V t21 (1 0.94) Pt2 0.94>(1 0.94) Pt21 0.94V t21 @
V t21 0.06 * Pt2 0.94 * 0.06 * Pt21 (0.94 * 0.94V t21 )

What are the weights for old returns and variance?
is called Persistence factor or even Decay Factor. Higher gives more weight to older
data (impact of older data is allowed to persist). Lower gives higher weight to recent data
(i.e. previous data impacts are not allowed to persist).
Higher means higher persistence or lower decay.
Since, (1- is weight given to latest square return, it is called Reactive factor.

EWMA Question
Example 1: On Tuesday, return on a stock was 4%. Volatility (Std. deviation) estimate for Tuesday
was 1%. Find volatility estimate for Wednesday using of 0.94
EWMA Solution
Variance estimate for Wednesday = (1-
Std. Dev. = sqrt (1.9%) = 1.378%
Tuesday volatility (Std. Dev.) estimate was 1%. Actual return on Tuesday was 4%. Therefore,
volatility estimate for Wednesday is estimated upwards than Tuesday i.e. 1.378% as compared to
1%.
Notice how the volatility estimate has been revised due to high return

EWMA Question
Example 2: Continuing the previous example, volatility estimate for Wednesday was 1.378%.
Assume that actual return on Wednesday was 0%. What is the variance estimate for Thursday?
EWMA Solution
Variance estimate for Thursday = (1-0.94)*(0%)^2 0.94*(1.378%)^2 = 1.78%2

Std. Dev. = 1.34%
In very short-term like daily returns, estimated volatility is the expected return
Since latest return of 0% was lesser than estimated volatility (and estimated return) of 1.378%,
volatility for next day is revised downward from 1.378% to 1.34%
Notice the downward revision in the estimate due to lower return

EWMA Weights Graph
= 0.7 (faster decay)

Weight of
variance terms
= 0.9 (slower decay)
Days into the past
,^&
Selection of Decay Factor

We are free to select the decay factor we use in our calculation of volatility
We can use common sense in our estimation
If we expect volatility to be very unstable then we will apply a low decay factor (giving a lot of weight to
recent observations)
If we expect volatility to be constant we would apply a high decay factor (giving a more equal weight to older
observations)
Sum of Weights
One special property of the weights used in the EWMA formula is that their sum will always equal to 1
m
S (1 O ).O
i 1
i 1
Om 1
Obviously we do not have an infinite set of historical data. We just have to make sure that our data set is
large enough so that this sum is close to 1
Or alternatively we can rescale the weights so that the sum is 1
What happens to the second term when m tends to infinity?

Question: FRM Exam
hZDtD
conditional variance, which weight will be applied to the return that is 4 days old?
A. 0.000
B. 0.043
C. 0.048
D. 0.950
Solution
B.
A. Incorrect. The wrong factor has been squared. The EWMA RiskMetrics model is defined as:
ht t-1 - 2t-1. For t = 4, and processing r0 through the equation three times produces a factor of (1-
0.95)3*0.95 = 0.00012 for r0 when t = 4.
B. Correct. The EWMA RiskMetrics model is defined as:
0.95)*0.95^3 = 0.043 for r0 when t = 4.
C. Incorrect. The 0.95 has not been squared. The EWMA RiskMetrics model is defined as:
0.95)*0.95 = 0.048 for r0 when t = 4.
/ddtDZD
ht t-1 - 2t-1. For t = 4, and processing r0 through the equation three times produces a factor of 0.95 =
0.950 for r0 when t = 4.

GARCH (1,1)
GARCH stands for Generalized Autoregressive Conditional Heteroscedasticity
Heteroscedasticity means variance is changing with time.
Conditional means variance is changing conditional on latest volatility.
Autoregressive refers to positive correlation between volatility today and volatility yesterday.
(1,1) means that only the latest values of the variables.
GARCH model recognizes that variance tends to show mean reversion i.e. it gets pulled to a
long-term Volatility rate over time.
V t21 JV L DP t2 EV t2
Long-term average Volatility
GARCH (1,1) (cont.)
V t21 Z DP t2 EV t2
's>
Since the sum of all the weights is equal to 1 we get the following equation as well:
Z
VL
1 D E

GARCH Question
Suppose a GARCH model is estimated using MLE from daily data as follows:
V t21 .000005 0.12 P t2 0.85V t2

Suppose that on a particular day t; actual return was -1% and the volatility (std. deviation)
estimate for that was 1.81) and long-term
average volatility (to which the model shows reversion over-time).
GARCH Solution
Solution: In the GARCH model, 12% is the weight given to latest squared return (reactive factor).
85% is the weight given to latest variance estimate (persistence factor). Therefore,
1-0.12-0.85 = 3% is weight given to long-term average Volatility.
Therefore, 3%*VL = 0.000005 i.e. VL = 0.017%
Also, variance estimate for 1 = .000005 0.12*(-1%)^2 0.85*(1.88%)^2 = 0.0317%
Volatility (Std. Dev.) estimate for 1 = sqrt (0.0317%) = 1.782%
For a stable GARCH model, alpha Beta <=1. If alpha Beta>1, then weight given to long-term volatility is
negative and the model becomes mean-fleeing

Question: FRM Exam
Which of the following GARCH models will take the shortest time to revert to its mean?
A. ht 2t-1 t-1
B. ht 2t-1 t-1
C. ht 2t-1 t-1
D. ht 2t-1 t-1
Solution
B.
A. Incorrect. The model that will take the shortest time to revert to its mean is the model with the lowest
1 /
1 0.03 0.96 = 0.99.
B. Correct. The model that will take the shortest time to revert to its mean is the model with the lowest
1 /
1 0.02 0.95 = 0.97.
C. Incorrect. The model that will take the shortest time to revert to its mean is the model with the lowest
1 /
1 0.01 0.97 = 0.98.
D. Incorrect. The model that will take the shortest time to revert to its mean is the model with the lowest
1 /
1 0.01 0.98 = 0.99.

Question
Suppose the long-run variance rate is 0.0002 so that the long-run volatility per day is 1.4%
V n2 . un21 0.86V n21

0.000002 013
Suppose that the current estimate of the volatility is 1.6% per day and the most recent percentage
change in the market variable is 1%. What is the new variance rate?
Solution
The new volatility is 1.53% per day
0.000002 0.13 u 0.0001 0.86 u 0.000256 0.00023516

VAR Methods for Estimating Risk
Historical based Approach
Parametric Approach
Exponential smoothing
Risk Metrics (EWMA model with = 0.94)
Typical example of parametric approach is the delta-normal VAR
Non-Parametric Approach
Historical Simulation
Multivariate Density Estimation
Hybrid Approach
Implied Volatility based Approach
Hybrid approach Steps

Estimating returns percentile directly as in HS

Exponential smoothing as in Risk Metrics
Step 1:
Denote by r (t 1 ,t ) the realized returns from (t-1) to t
For the most recent k returns
Choose lamda based on weight requirements (L)
Assign a weight (1 L) / (1 L ^K) for t = 1 , (1 L) / (1 L ^K) * L for t = 2
Step 2 :
Order returns in ascending order
Step 3 :
Keep accumulating the weights till you reach x% VaR
Linear interpolation is used to reach exactly x % of the distribution

Question 14 (Hybrid approach)
Hybrid Approach Problem Worksheet
Lamda 0.98 K 100
Return Periods Ago Hybrid Wt Cum Hybrid Wt HS Wt Cum HS Wt
-3.3 3 0.022144839 0.022144839 0.01 0.01
-2.9 2 0.022596774 0.044741613 0.01 0.02
-2.7 65 0.006328331 0.051069944 0.01 0.03
-2.5 45 0.009479113 0.060549057 0.01 0.04
-2.4 5 0.021267903 0.08181696 0.01 0.05
-2.3 30 0.012834429 0.09465139 0.01 0.06
Hybrid Interpolation 0.16618552
Ans -2733814418
HS Ans -2.35
Portfolio Volatility
Variance Covariance Approach

Extend the parametric approach to entire portfolio
Disadvantage: Correlations tend to increase in periods of stress. This approach might underestimate the
portfolio Volatility
Extend Historical Simulation Approach

Apply current portfolio weight to old period returns
Third Approach
Aggregate the simulated returns and then apply the parametric approach to aggregated portfolio

Implied Volatilities
Forward looking, predictive
Use specific Derivative pricing model
Black Scholes pricing model for ATM options
The implied volatility calculated from a European call option should be the same as that calculated
from a European put option when both have the same strike price and maturity
/d^
market price
Truly Predictive in nature and is not a historical simulation but it assumes that Black Scholes
(or any other model) is correct model for calculating market Value of Option
Discussion (VaR Methodology)

Three value at risk (VaR) methods are reviewed: Delta-normal, historical simulation and Monte
Carlo. Which are true of the following? Among the VaR methods, which:
i. Efficient in terms of data use?
ii. Requires a (parametric) distributional assumption?
iii. Is LEAST appropriate for a portfolio that contains many embedded derivatives (e.g., options)?
iv. Is computationally fast?
v. Handles fat tails?
vi. Suffers from sampling variation?

Question 15 (Historical Simulations)
Suppose we have an asset with ordered simulated price returns as below for sample of 500 days
and is trading at 70. What is the VaR at 99% confidence if the returns for the last 500 days are:
-7%, -6.7%,-6.6%, -6.5%,-.6.1%,-5.9% 4%, 4.75%,5.1%, 5.2%,5.3%
Solution
99% of 500 = 495th i.e. -5.9% = 0.059 x 70 = 4.13

Question 16 (HS)
We have an asset with ordered simulated price returns as below for sample of 400 days and is
trading at 100. What is the VaR at 99% confidence if the returns for the last 400 days are:
-7%, -6.7%,-6.6%, -6.5%,-.6.1%,-5.9% 4%, 4.75%,5.1%, 5.2%,5.3%
What is the expected shortfall?
Solution
VaR = 396th = -6.1% of 100 = -6.1
Expected shortfall = - -6.7%
Expect atleast one direct question on the calculation of Expected Shortfall

Putting VaR to Work
Full Valuation Method
Full Valuation method is the process of measurement of risk of a portfolio by fully re-pricing it
under a set of scenarios over a time period. It can be used to cover a large range of values of the
portfolio returns in order to provide more accurate results. It generally provides more accurate
results compared to delta normal approach but is a complicated process.
Advantages over delta normal:

It accounts for non-linearities of derivatives whereas delta normal assumes a linear approximation.
It accounts for extreme fluctuations.
Two popular methods under full revaluation approach have been explained in the subsequent
slides.

Historical Simulation
No assumptions is required about the distribution of returns.
VaR is estimated directly from the past data which includes all the correlations among
historical data.
Historical data has fat tails and skewness already accommodated in itself.
The most important parameter in historical simulation is the Look back window.
For example, in 250 observations window the 5th percentile is between 12th and 13th observation.
Historical Simulation method is not exposed to any model risk.
The biggest drawback with the Historical Simulation method is that the changes in volatility and
correlation from structural changes are not recognized.
We have an asset with ordered simulated price returns as below for sample of 100 days and is
trading at 100. What is the VaR at 99% confidence if the returns for the last 100 days are:
-7.5%, -6.7%,-6.6%, -6.51%,-.6.12%,-5.92% 4.34%, 4.5%,5.1%, 5.2%,5.3%

Straddles: Long position in call and put with same exercise price.
Straddle is a non Linear derivative whose payoff increases with the increase in the volatility.
Also delta normal VAR increases with the increase in the volatility.
Going by the delta normal VAR for straddles, as the volatility increases, VAR should increase but in
reality the payoff is becoming positive.
Payoff
Gains !
Price
Risk !

The Monte Carlo approach assumes that there is a known probability distribution for the
risk factors.
The usual implementation of Monte Carlo assumes a stable, Joint-Normal distribution for the risk
factors.
This is the same assumption used for Parametric VaR.
The analysis calculates the covariance matrix for the risk factors in the same way as
Parametric VaR.
Unlike Parametric VaR Monte Carlo Simulation:
Decomposes the covariance matrix and ensures that the risk factors are correlated in each scenario.
The scenarios start from today's market condition and go one day forward to give possible values at the end
of the day.
Full, nonlinear pricing models are then used to value the portfolio under each of the end-of-day scenarios.
For bonds, nonlinear pricing means using the bond-pricing formula rather than duration.
For options, it means using a pricing formula such as Black-Scholes rather than Greeks.

Advantages:
Unlike Parametric VaR, it uses full pricing models and can therefore capture the effects of nonlinearities.
Unlike Historical VaR, it can generate an infinite number of scenarios and therefore test many possible future
outcomes.
Disadvantages:
The calculation of Monte Carlo VaR can take 1,000 times longer than Parametric VaR because the potential
price of the portfolio has to be calculated thousands of times.
Unlike Historical VaR, it typically requires the assumption that the risk factors have a Normal distribution.

Comparison between Methods
The best methods depend on speed required and whether the portfolio has non-linear elements.
For large portfolios where non linearity is not a major factor, the delta normal method is fast
and efficient.
For portfolios with substantial non-linear elements full valuation may be necessary.
Visualizing WCS
The worst case scenario (WCS) answers:
What is the worst loss that can happen over a period of time?
The probability of a worst loss is certain (100%); the timing is uncertain.
Focuses on distribution of loss during the worse trading period:
A worst case scenario can be simulated only using Monet Carlo Simulations

Visualizing WCS
The area under the normal curve for confidence value is:
Confidence (x%) Zy
90% 1.28
95% 1.65
97.5% 1.96
0.016 99% 2.32

Distribution of WCS
0.011
0.005 Mean = 0
- stdev
Approx. Normal curve representing Value at Risk
Question (WCS)
H = 100
-2.33 ( 99% VaR ) 1 out of 100
Distribution of WCS mean = -2.51 and its 1st and 5th percentile are -3.1 and -3.9 resp.
What does this mean to a financial risk manager?

Solution
Mean of distribution of WCS = -2.51

It means that if the loss exceeds the VAR number(2.33), the average value of the loss would be 2.51 which is
nothing but the expected shortfall (ES)
The 5th and 1st percentile of WCS is -3.1 and -3.9
It means that if the loss exceed the VAR, then probability of loss exceeding 3.9 is 1%
Similarly the probability of loss exceeding 3.1 would be 5%
Stress Testing

Stress Testing
Limitations of VaR
Tells that n number of times in 100 days, the loss is not going to exceed N$
But it cannot predict the loss when it exceeds!
Does not focus on large losses (Tails of distribution)
Stress Testing:
Supplement to VaR
"VaR should always be supplemented with stress testing" has been one of the recommendations
of the supervisor
Testing how well a portfolio performs under some of the most extreme market moves seen in the last 10 to
20 years
Stress Testing
Advantages:
Can take a large number of risk factors into consideration
Can specifically focus on the tails (extreme losses)
Disadvantages:
Highly subjective and can become overcautious
Requires complete top management support
Two independent sections to the risk report:

VAR-based
Top-down identification of the relevant risk generators for the trading portfolio
Stress testing-based risk report proceeds in one of two ways
It examines a series of historical stress events and
It analyzes a list of predetermined stress scenarios

Stress Testing
Capital Allocation
Exposure
Ride out Turmoil
The one significant shortcoming of VaR that stress testing does address is sudden changes in
historical correlations
If two currencies have been pegged to one another, they will exhibit a high historical correlation. A
VaR analysis based on that historical correlation will not address the risk that one of the currencies
may be devalued relative to the other. If this is a scenario that concerns management, a simple
stress test will offer more insights than would, say, a VaR analysis performed with a modified
correlation assumption
Scenario Analysis
Oct19, 1987 : The Black Monday, a -20 sigma event

Sep 1992, breakup of fixed exchange system
The 99% VAR would have totally missed the magnitude of actual loss
Goals
Identify scenarios that would not occur under standard VAR models
Simulating shocks that have never occurred
Simulating shocks that reflect permanent structural breaks or temporally changed statistical patterns
Principles of Scenario Analysis

Identify risk factors
Forecast and Change Risk factors
Revalue Portfolio using VaR system in place
Portfolio Vs Event Driven

Scenario Analysis
Parallel shift in yield curve

Changes in steepness of yield curve
Parallel shift in yield curve along with the changes in the steepness of yield curve
Changes in yield volatilities
Changes in the values of equity indices
Changes in equity index volatilities
Changes in the values of key currencies with respect to US Dollar
Changes in foreign exchange rate volatilities
Changes in Swap spreads in G7 countries plus Switzerland
Standard Portfolio Analysis of Risk (SPAN)
Example of scenario based method for measuring portfolio risk

SPAN uses full valuation methods that makes it a good tool for analyzing portfolio
including options
Margin is set to the worst portfolio loss after considering all scenarios
Useful when considering only two risk factors. Becomes very complex when risk factors increase.
Drawback:
Ignores correlation between risk factors
With more number of risk factors, alternative scenarios could become unmanageable

Unidimensional and Multidimensional Scenarios
Uni-dimensional scenario analysis:

Identifies important risk factors
Shocks the factor by a large amount
Measures the impact on portfolio value
Multi-dimensional Scenario Analysis

Incorporates correlation between risk factors.
This increases the complexity of the analysis
This analysis can be backward or forward looking
Various Approaches to Scenario Analysis
Prospective Scenarios Factor Push Method and Conditional Scenario Method.
Factor push method shifts each risk factor in the direction that would have an adverse impact on
the portfolio.
Conditional Scenario Method incorporates the correlation between various key risk factors.

Improving Stress Tests
Stress tests can be improved by adopting one or more of the following methods:
Buy protection through insurance or other derivative products
Change portfolio composition to decrease the exposure or diversify
Change business strategy to suit the changing business environment
Develop back up plans for unforeseen events
Secure alternative funding in stressed environment
Five Minute Recap
Key Terms to Remember: Volatility calculation: EWMA: Confidence (x%) Zy

Delta Normal VAR 1 n
Implied Volatility V 2
n
(R
i 1
i E ( R )) 2
V n2 OV n2 1 (1 O )u n21 90% 1.28
Regime Switching 95% 1.65
The worst case scenario
GARCH (1,1): 97.5% 1.96
Stress Testing
VAR Z X % * V * Asset Value
V 2
t 1 JV L DP EVt
2
t
2
99% 2.32
VaR (n days) (in %) VaR (daily VaR) (in %) * n VARLinear Derivative ' *VARUnderlying Risk Factor
VaR portfolio (in %) wa2 (%VAR a ) 2 w 2b (%VAR b ) 2 2w a w b * (%VAR a ) * (%VAR b ) * U ab
Value at Risk: Value-at-Risk Measurement Methods: Monte Carlo Simulation:

0.5 Probability
Payoff
0.4 VaR Gains !
0.3
0.2 95% daily-VAR
Linear Valuation Full Revaluation
0.1 Method Method
0 Price
-4 -2 0 2 4
Mean = 0 Delta Normal Historical Monte Carlo
Method Simulation Simulation Risk !
VAR X % (in %) Z X % *V

Thank z
www.edupristine.com
Valuations and Risk Models-II
Greeks
Agenda
Naked and Covered Position

Stoploss Strategy
GREEKS
Delta
Theta
Gamma
Vega
Rho
Summary
Expect around 68 questions in the exam from these slides

If a trader sells call contracts standalone, he has a naked position exposure

He will gain from the premium received if the price of the underlying falls and the option is not exercised
Potential for unlimited loss in case stock prices rise and the trader has to cover his short position
If a trader sells call contract and has equivalent long position in the underlying he has cover
He will gain if the stock price rises and the option is exercised
Potential for large losses in case there is a rapid decline in the stock price

Stoploss Strategy
Stoploss Strategy
Trader can hope to construct a stop-loss strategy by combining the two strategies
Suppose the trader is able to ensure that he has a covered call position each time the price of the underlying
exceeds the strike price on the option and
A naked call position whenever the stock price goes down from the exercise price. If a trader sells call
contracts standalone, he has a naked position exposure
If this was possible in the real world, traders would be able to earn riskless profits as the cost of
setting the hedge would always be less than the theoretical price of the option
However, it is impossible to execute this in reality because of the following reasons:
Purchase of stock will always be slightly above the strike price and sale slightly below it
This difference would increase the cost of this stop-loss strategy.
In case one wants to decrease this difference to a really small amount, frequent buying and selling will be
required, which in turn would again increase the cost of this trade
The difference in time of the associated cash flows and the time value of money also result in increased costs

GREEKS Delta
Delta
Options delta is defined as the rate of change in an option's price relative to a one-unit change in
the price of the underlying asset
The delta for European Call option is given by N(d1) and for the Put option it is given by N(d1)-1
Holding delta share and selling one call lead to a risk neutral outcome, other things remaining
same
This is an example of delta hedging
Such a package of option and share is called a delta neutral portfolio
Delta hedging strategy is an improvement over the simple stop-loss strategy
However, the options delta changes with changes in stock price
Requires traders to frequently readjust their positions in order to remain delta neutral. This process is called
rebalancing
Delta hedging therefore is a dynamic hedging strategy

Delta (cont.)
Delta (Call / Put)
0.8 2.0
0.6 1.5
0.4 1.0
0.2 0.5
0 0
0.2 0.4 0.6 0.8 1 1.2 1471013161922252831343740434649
-0.2 -0.5
-0.4 -1.0
-0.6 -1.5
Delta (ATM Call) Delta (Call)

Delta (ATM Put) Delta (Put)
Delta (cont.)
Properties of Options delta

Call has a positive delta and Put has a negative delta
Option delta will lie between -1 1
The delta increases in absolute terms as the option goes further in-the-money and decreases as the option
goes out-of-the-money
At-the-money call and put options would have a delta around 0.50 and -0.50 respectively
The delta of a Put option is negative reflecting an inverse relationship with the price of the underlying
Deep in-the-1.00 as they are most likely to be exercised.
Similarly, deep in-the-money put options would have a delta tending towards -1.00
Deep out-of-the-money calls and puts have deltas that approach zero as the probability that they will be
exercised is nearly zero
The delta of the underlying asset is always 1.00

Delta (cont.)
The delta of a portfolio of derivatives (such as options) with the same underlying asset, can be
found out if the deltas of each of these derivatives are known
n
' portfolio W ' i i
Example: i 1
Lets find the delta of the following portfolio constructed from the options of the same underlying
asset and also estimate the trade required in order to turn the portfolio delta neutral
The delta of the portfolio is given by:
50,000 * 0.53 100,000 * 0.47 25,000 * (-0.51) = -7750
It follows from the calculations above that 7,750 stocks need to be bought in order to get a delta
neutral portfolio
Option (Strike Price, Expiration in Months) Quantity Position Delta of Option

Call (55, 3 months) 50,000 Long 0.53
Call (56, 5 months) 100,000 Short 0.47
Put (56, 2 months) 25,000 Short -0.51
Delta (cont.)
Cost involved in Delta Hedging

Although effective as a hedging strategy, delta hedging carries a cost
Every time a portfolio is rebalanced, the difference between the price paid for a stock and the price received
for it creates this cost
This is so because this hedging strategy calls for buying when stock prices rise and selling when they fall

GREEKS Theta
Theta
The theta of an option is the rate of change in its value with the passage of time, assuming that
other things remain the same
For a portfolio, the theta is the rate of change in the value of the portfolio as time passes, given
that other things are constant. A positive theta implies that the portfolio value will increase as
the time passes, while a negative theta implies that the value will decrease with the passage of
time, if there is little move in the stock price and the implied volatility
Normally expect the theta of an option to be negative as with the passage of time, an option
loses value
Exception can be an in-the money European put option on a non dividend paying stock
For at-the-money option, theta increases as the expiration date nears
Theta decreases as an option which is either out of money or in the money approaches expiration

Theta (cont.)
Theta (ATM) vs. Time Call / Put
0 0.2 0.4 0.6 0.8 1.0 1.2 2.5

0 2.0
1.5
1.0
-5
0.5
0
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49
-10 -0.5
-1.0
-1.5
-15
-2.0
-2.5
-20 -3.0
Theta (Call)
Theta (Put)
Theta (cont.)
We have theta of call given by: Where:

S0 = Stock price at time 0, i.e. present price of the
S 0 N ' ( d1 )V stock
4(Call ) rKe rT N ( d 2 )
2 T d1 and d2 are as defined in the Black-Scholes
Pricing formula earlier
e ( x ^ 2) / 2 ^
N ' ( x)
2S K = Strike price
Where: T = Time of maturity of the option measured in

years, so that 6 months will be 0.5 years
S 0 N ' ( d1 )V
4( Put ) rKe rT N ( d 2 ) r = Risk neutral rate of interest
2 T
For a put option, theta is given by:

GREEKS Gamma
Gamma
Gamma is the rate of change in delta for unit change in the price of the stock
For a portfolio of options with the same underlying asset, the gamma is the change in the value of
the portfolios delta with the change in the portfolio value:
A large gamma suggests that delta will change rapidly as the price of the underlying stock changes
The following table presents and compares the sign of Gamma and Delta
Option Position Delta Gamma

Call Long
Put Long - +
Stock Long 1 0
Call Short - -
Put Short -
Stock Short -1 0
Gamma reaches its maximum absolute value when a stock is trading at the money or near the
money
It reduces in value as the security moves further out of the money or further in the money

Gamma (cont.)
Calculation of Gamma
Gamma for European options can be calculated using the following formula:
N ' ( d1)
*
S 0V T
Where symbols have their usual meaning
Gamma (ATM) vs. Time Gamma (Call / Put)

0.45 0.07
0.40 0.06
0.35
0.05
0.30
0.25 0.04
0.20 0.03
0.15
0.02
0.10
0.05 0.01
0 0
0 0.2 0.4 0.6 0.8 1.0 1.2 1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49
Gamma (cont.)
Gamma Neutral Portfolio

Suppose we have a delta neutral portfolio of options that has a positive gamma
In order to make the portfolio gamma neutral one would need to undertake trades in options that reduce the
gamma to zero
However, these trades will lead to a new portfolio, which will not be delta neutral anymore
Therefore as a second step one would need to trade in stocks so that the portfolios delta neutrality is
restored
This will not impact the gamma neutrality of the portfolio as stocks carry a zero gamma
Relation between Gamma, Delta and Theta
For a delta neutral portfolio, when theta is large and negative, gamma will tend to be large and positive. The
converse of this holds too. Thus, if delta is zero and theta is large and positive, gamma is likely to be large in
magnitude and negative in sign.

GREEKS Vega
Vega
The Vega of a derivative portfolio is the rate of change of the value of the portfolio with the
change in the volatility of the underlying assets. It can be expressed as:
G3
s
GV
For European options on a stock that does not pay dividends, Vega can be found by:
s^dE^dE
given by:
e ( d 1^ 2 ) / 2
N ' ( d1) Vega
2S 16
The Vega of a long position is always positive 14
12
A position in the underlying asset has a zero Vega
10
Thus its behavior is similar to gamma 8
Vega is maximum for options that are at the money 6
4
2
0
1 4 7 1013161922252831343740434649

GREEKS Rho
Rho
Rho of a portfolio of options is the rate of change of its value with respect to changes in the
interest rate
Rho = G 3 , where is the value of the portfolio, and r is the rate of interest
Gr
For European options on non dividend paying stocks, we have;
Rho (call) = KTe-rTN(d2), where the symbols carry their usual meanings
Also, Rho (put) = -KTe-rTN(-d2), the symbols carrying their usual meanings
Rho (Call / Put)

30
20
10
Rho (Call)
0
Rho (Put)
-10
-20
-30
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49

Summary
Summary
Traders may combine naked and covered positions to evolve a stop-loss strategy
Stop loss strategy is impractical given the realities of trade and transaction costs involved
Delta hedging is an improvement
Involves that delta of the portfolio is maintained at zero
Requires frequent rebalancing as delta change
Dynamic hedging strategy
Hedging can also be attempted with respect to changes in:
Time (theta)
Option delta itself (Gamma)
Volatility (Vega)
Interest rate (Rho)
In addition to option Greeks, traders also rely on scenario analysis
Involves evaluating the option value for simultaneous changes in:
Time (Theta)
Volatility (Vega)
Interest rate (Rho)
Other factors

Question
1. A stock trading at $20 has call options available on it with exercise prices $18 and $20. For $1
increase in the stock price how will the delta of the two options change? Choose the most
appropriate answer. Change in deltas for the two options are denoted by d '18 and d '20.
A. d'18 < d'20
B. d'18 > d'20
C. d'18 = d'20
D. d'18 > d'20 and d '20 = 0
2. Which of the following statements is true regarding options Greeks?

A. Theta tends to be large and positive for at-the-money options
B. Gamma is greatest for in-the-money options with long times remaining to expiration
C. Vega is greatest for at-the-money options with long times remaining to expiration
D. Delta of deep in-the-money put options tends towards 1
Question
3. Which of the following statements is false?

A. European-styled call and put options are most affected by changes in Vega when they are at the money
B. The delta of a European-styled put option on an underlying stock would move towards zero as the price of
the underlying stock rises
C. The gamma of an at-the-money European-styled option tends to increase as the remaining maturity of the
option decreases
D. Compared to an at-the-money European-styled call option, an out-of-the money European option with the
same strike price and remaining maturity would have a greater negative value for theta

Solution
1. A.
2. C.
Vega is the rate of change in the price of an option with respect to changes in the volatility of the underlying
asset. Vega is greatest for at-the-money options with long times remaining to expiration
3. D.
Theta is large and negative for an atthe-money European-styled option, whilst theta is close to zero when
the price for the underlying stock is very low. Therefore the theta for an out-of themoney European styled
call option would have a lower negative value compared to that of an at-the-money European-styled call
option
Operational Risk

Operational Risk
Some companies define operational risk as all the risk that is not market or credit risk. But this is a
very broad definition of operational risk
The Basel definition of operational risk is the risk of direct or indirect loss resulting from
inadequate or failed internal processes, people and systems or from external events
Operational Risk definition includes legal risk but does not include reputational or strategic risk
because they can be difficult to quantify
Approaches for Determining Operational Risk
There are three main approaches for calculating operational risk capital requirement:
The basic indicator approach
In this approach, operational risk capital is based on 15% of the banks annual gross income over the period
of 3 years
Includes both interest and non interest income
The standardized approach
In this approach, bank uses eight business lines with different beta factors to calculate the capital
requirement
Beta factor for each business line is multiplied with the annual gross income of the business line over the
period of 3 years
The results are then added to arrive at total operational risk capital charge
The advanced measurement approach (AMA)
Large banks are encouraged to move from standardized approach to AMA because with AMA, banks can
reduce their capital requirement by investing in risk management infrastructure

Advanced Measurement Approach
The operational risk capital requirement currently proposed by the Basel Committee is equal to
the unexpected loss in a total loss distribution that corresponds to the confidence level of 99.9%
over a 1 year time horizon
Expected Unexpected
Loss (EL) Loss (UL)
Probability
of loss
Potential Loss
Operational Risk Categories
The Basel committee on Banking Supervision divides operational risk into seven types:
1. Clients, products and business practices
2. Internal Fraud
3. External Fraud
4. Damage to physical assets
5. Execution, delivery and process management
6. Business Disruption and System failures
7. Employment Practices and Workplace safety

Loss Frequency and Loss Severity
Loss frequency is defined as number of losses over a specific period of time.

Loss frequency is often modeled with Poisson distribution.
In Poisson distribution, mean and standard deviation is equal to the single parameter, lambda.
Loss severity is defined as the value of financial loss suffered.

Loss severity is modeled using lognormal distribution.
Loss frequency and loss distribution are combined to simulate the expected loss distribution.
This is known as convolution.
Monte Carlo Simulation process is used to run this simulation.
Data Limitations
The historical operational risk loss data is currently inadequate to run these simulations.
Banks should use internal data to estimate the frequency of losses.
Banks should use both internal and external data to estimate the severity of losses.
There are two sources of external data:

Data sharing agreement with other banks
Public Data

Scenario Analysis in Scarce Data
In scenario analysis, different scenarios are created to incorporate the events that have not
yet occurred.
Regulators have encouraged the use of scenario analysis.
Scenario analysis act as a tool for the management to immunize against potential operational risk.
The drawback of scenario analysis is the excessive amount of time spent by the management to
create such scenarios.
Forward Looking Approaches
In addition to scenario analysis, forward looking approaches are also used to discover potential
operational risk events. There are three forward looking methods:
1. Causal Relationships
2. Risk and Control Assessment
3. Key Risk Indicators

Scorecard Data
Manager of each business unit allocates operational risk capital to manage operational risk.
Less capital will be allocated to those business units which are able to reduce the frequency and
severity of losses.
One of the approaches to allocate capital is scorecard approach.
Each manager is surveyed regarding the key features of each type of risk.
Answers of the managers are given a score to quantify the responses.
The Power Law
Power law is very useful in evaluating the nature of the tails of a given distribution in extreme
value theory.
This law is very useful in analysis of operational risk because most of the operational risk losses are
likely to occur in the tails.
Insurance
There are many insurance companies which provide insurance for operational risk.
There are two main issues faced by such insurance companies:
Adverse Selection
Moral Hazard

External and Internal Ratings
External Credit Ratings
Issue Specific Credit Rating Conveys information about specific instrument.
Issuer Credit Rating Information about the entity that has issued the instruments.
Investment Grade Bonds.
Non Investment Grade Bonds.

Equivalent Credit Ratings
Credit Risk Moody's* Standard & Poor's* Fitch IBCA** Duff & Phelps**
Investment Grade
Highest quality Aaa AAA AAA AAA Triple A = Credit risk almost zero
High quality (very strong) Aa AA AA AA Safe investment, low risk of failure
Safe investment, unless unforeseen

Upper medium grade (strong) A A A A events should occur in the economy at
large or in that particular field of business
Medium safe investment. Occurs often

Medium grade Baa BBB BBB BBB when economy has deteriorated.
Problems may arise
Not Investment Grade

Speculative investment. Occurs often in
Lower medium grade
Ba BB BB BB deteriorated circumstances, usually
(somewhat speculative)
problematic to predict future development
Speculative investment. Deteriorating
Low grade (speculative) B B B B situation expected
Poor quality (may default) Caa CCC CCC CCC
Most speculative Ca CC CC CC
No interest being paid or
C C C C
bankruptcy petition filed Bankruptcy or lasting inability to make
In default C D D D payments most likely
* The ratings from Aa to Ca by Moody's may be modified by the addition of a 1, 2 or 3 to show relative standing within the category.
**The ratings from AA to CC by Standard & Poor's, Fitch IBCA and Duff & Phelps may be modified by the addition of a plus or minus sign to show relative.
The Credit Rating Process
1. Qualitative Analysis
2. Quantitative Analysis
3. Meeting the firms management
4. Committee in the credit rating agency meets to assign the rating
5. Assigned rating is notified to the firm
6. Opportunity for the firm to appeal or offer new information
7. Disseminating the rating to the public via the news media

Transition Matrices
Original Probability of Migrating to Rating by Year End (%)

Rating AAA AA A BBB BB B CCC Default
AAA 93.66 5.83 0.40 0.08 0.03 0.00 0.00 0.00
AA 0.66 91.72 6.94 0.49 0.06 0.09 0.02 0.01
A 0.07 2.25 91.76 5.19 0.49 0.20 0.01 0.04
BBB 0.03 0.25 4.83 89.26 4.44 0.81 0.16 0.22
BB 0.03 0.07 0.44 6.67 83.31 7.47 1.05 0.98
B 0.00 0.10 0.33 0.46 5.77 84.19 3.87 5.30
CCC 0.16 0.00 0.31 0.93 2.00 10.74 63.96 21.94
Default 0.00 0.00 0.00 0.00 0.00 0.00 0.00 100.00
One-year ratings migration probabilities based upon bond rating data from 1981-2000. Data is
adjusted for rating withdrawals. Numbers in each row should sum to 100%. Due to round-off
error, they may not do so exactly. Source: Standard & Poor's
Impact of Time Horizon and Economic Cycle
The chances of default given any rating at the beginning of a cycle increases with the time horizon.
External ratings are designed to be relatively stable over the business cycle. There is a possibility
that there are errors in severe cycles.
But external ratings have fairly good records in indicating relative defaults rates.

Impact on Rating Changes on Bond and Stock Prices
A rating downgrade is likely to decrease the price of the bond.
A rating upgrade is likely to increase the price of the bond.
A rating downgrade is likely to decrease the price of the stock.
A rating upgrade is somewhat likely to increase the price of the stock.
Internal Credit Ratings
At-The-Point (ATP) Assesses for one year e.g. scoring models, KMV etc.
Through-The-Cycle (TTC) Captures the credit-worthiness over a larger period of time, including
the impact of normal cycles in the economy, these ratings are more stable.
Some banks use TTC for large corporate and ATP for SME.
One view is that TTC is the combination of various ATP at various time.

Internal Ratings
Building Internal Ratings

Identify most meaningful factors.
Assign weights to these factors.
Calibrate these weights.
Calibrating and back-testing.
Issues in Back-Testing
There is no database when an internal rating system is designed.
Generally a time of 10 to 18 years required to validate an internal rating system.
APT will have pro-cyclicality if they stick to PD but will have unstable transition matrices and no long term
guidance if they follow internal ratings.
Biases in a Rating System
Time horizon bias

Homogeneity bias
Principal agent bias
Information bias
Criteria bias
Scale Bias
Back testing Bias
Distribution Bias

Country Risk: Determinants, Measures and Implications
Agenda
What is country risk?
Sources of Country risk
Measuring Country risk
Sovereign defaults
Consequences of sovereign default
Factors affecting sovereign default risk
Sovereign Rating
Criticism of ratings agencies
Sovereign Default Spreads

Country Risk
As the businesses are increasingly becoming global with globalization and rise of non-domestic investing
activities have lead to a greater assessment and understanding of country risk.
The need to understand, analyze and incorporate country risk has become a priority at most large
corporations, as they have globalized and become more dependent upon growth in foreign markets for
their success. Thus, a chemical company based in the United States now has to decide whether the hurdle
rate that it uses for a new investment should be different for a new plant that it is considering building in
India, as opposed to the United States, and if so, how best to estimate these country-specific hurdle rates.
Governments often have a direct effect on country risk, with increased country risk often translating into
less foreign investment in the country, leading to lower economic growth and potentially political turmoil,
which feeds back into more country risk.
Sources of Country Risk

Economic growth life cycle: countries that are in early growth, with few established business and
small markets, being more exposed to risk than larger, more mature countries. By this, a small,
emerging market is greatly exposed to global recession in terms of economic growth and
unemployment. A country that is still in the early stages of economic growth will generally have
more risk exposure than a mature country, even it is well governed and has a solid legal system.
Political Risk:
1. Whether democracy leads greater growth or dictatorship
2. Corruption and Side Cost
3. Physical violence
4. Nationalization/Expropriation risk
Legal risk
Economic Structure: Well diversified economy or reliance of economy on single commodity

Measuring Country Risk
Political Risk Services (PRS): Provides numerical measures of country risk using 22 variables consisting of
three dimensions, political, financial and economic. PRS also provides forecast of country risk scores.
Euromoney: Euromoney has country risk scores, based on surveys of 400 economists that range from
zero to one hundred. It updates these scores, by country and region, at regular intervals.
The Economist: The Economist developed its own variant on country risk scores that are developed
internally, based upon currency risk, sovereign debt risk and banking risk.
The World Bank: The world bank measures risk by considering governance factors like corruption,
government effectiveness, political stability, regulatory quality, rule of law and voice/accountability, with
a scaling around zero, with negative numbers indicating more risk and positive numbers less risk.
Limitations of various services in measuring country risk:
Measurement model/methods
No Standardization
More ranking than scores: The country risk score are more useful for ranking country on risk rather than
measuring risk. A risk score of 60, compared with a risk score of 30 is definitely more risky but we can not
conclude that it is double risky.
Sovereign Default
Risk that government issued bonds will fail in their contractual repayment of coupon payments or maturity
value or both in timely manner to the investors.
Foreign Currency defaults

Local Currency defaults
In a study of sovereign defaults between 1975 and 2004, Standard and Poors notes the following facts
about the phenomenon:
1. Countries have been more likely to default on bank debt owed than on sovereign bonds issued.
2. In dollar value terms, Latin American countries have accounted for much of sovereign defaulted debt in
the last 50 years

Reasons explaining local currency defaults
Gold Standard: Till 1971, countries following the gold standard, had to be backed up with gold
reserves. As a consequence, the extent of these reserves put a limit on how much currency could
be printed.
Shared Currency: When the Euro was adopted as the common currency for the Euro zone, the
countries involved accepted a trade off. In return for a common market and the convenience of a
common currency, they gave up the power to control how much of the currency they could print.
Thus, in July 2015, the Greek government cannot print more Euros to pay off outstanding debt.
Inflationary concerns
Consequences of Sovereign Default

Reputation loss
Capital Market Turmoil
Real Output
Political Instability

Factors Affecting Sovereign Default Risk
Degree of Indebtedness
Pension/Social Services commitments
Revenues/Inflow to the government
Stability of revenue
Political Risk
Implicit Backing from the other entities
Sovereign Ratings
Bond rating agencies have initiated sovereign rating and they have two big advantage in gaining acceptance
in the market:
1. Rating agencies have been assessing default risk in corporations for a hundred years or more and
presumably can transfer some of their skills to assessing sovereign risk.
2. It is easy for investors to relate the assessment of risk conveyed by the rating agencies as they are
familiar with the ratings measures, from investing in corporate bonds. The same leads to a AAA rated
country to be viewed as close to riskless whereas a C rated country is very risky.
Rating agencies generally provide both local currency and foreign currency rating and generally
local currency ratings are at least as high or higher than foreign currency rating. The same is due to
more power in the hands of the government to print its own currency and pay-off the debt.
Sovereign ratings do change over time but less frequently than corporate rating. The same is
evident from the rating transition matrix.

Sovereign Ratings (Cont.)
Fitch sovereign Transition rates across the major rating categories: 1995-2008 (%, Annual)
The rating agencies have been criticized d for having a regional bias means they under rate entire regions
of the world (Latin America and Africa). However, the argument of rating agency for the same is past
default history is a good predictor of future default and the regions rated lower have a bad history to
overcome default.
Rating measure: Sovereign ratings focus on credit risk to private creditors and not to official creditors like
IMF, World Bank etc. Default means either a failure to pay interest or principal on a debt instrument on
the due date (outright default) or a rescheduling, exchange or other restructuring of the debt
(restructuring default).
Determinants of Rating
Political Risk:
Stability and legitimacy of political institutes
Participation of population in political process
Leadership Succession
Transparency in economic policy decisions and objectives
Public security
Geopolitical risk
Economic Risk:
Market orientation of economy
Income disparities
Financial intermediation effectiveness: Availability of credit
Labour flexibility

Determinants of Rating (Cont.)
Pattern and rate of economic growth.
Size and composition of savings and investment in the economy.
Fiscal Flexibility: Revenue raising flexibility and efficiency, General Govt. Expenditure, revenue and
surplus/deficit trends, pension obligation, timeliness, coverage, and transparency in reporting.
General Govt burden: General govt gross and net debt, currency composition and maturity profile,
depth and breadth of local capital market.
Robustness of financial sector.
Monetary Flexibility: Behavior of prices in economic cycles, compatibility of exchange rate and
monetary goals, central bank independence (institutional factors), Efficiency of monetary policy
tools.
External liquidity: Structure of the current account, composition of capital flows, reserve
adequacy.
External Debt burden: Access to concessional funding, debt service burden etc.
Rating Process: The analyst with primary responsibility for the sovereign rating prepares a ratings
recommendation with a draft report, which is then assessed by a ratings committee composed of 5-
10 analysts, who debate each analytical category and vote on a score. Following closing arguments,
the ratings are decided by a vote of the committee.
Local versus Foreign currency rating: The ratings agencies usually assign two ratings for each
sovereign a local currency rating and a foreign currency rating. There are two approaches used by
ratings agencies to differentiate between these ratings. In the first, called the notch-up approach,
the foreign currency rating is viewed as the primary measure of sovereign credit risk and the local
currency rating is notched up, based upon domestic debt market factors. In the notch down
approach, it is the local currency rating that is the anchor, with the foreign currency rating notched
down, reflecting foreign exchange constraints.
The differential between foreign and local currency ratings is primarily a function of monetary policy
independence. Countries that maintain floating rate exchange regimes and fund borrowing from
deep domestic markets will have the largest differences between local and foreign currency ratings,
whereas countries that have given up monetary policy independence, either through dollarization or
joining a monetary union, will see local currency ratings converge on foreign currency ratings.

Rating review and updates: Sovereign ratings are reviewed and updated by the ratings agencies and
these reviews can be both at regular periods and also triggered by news items. Thus, news of a
political coup or an economic disaster can lead to a ratings review not just for the country in
question but for surrounding countries (that may face a contagion effect).
Criticism of Rating Agencies

Ratings are upward biased
Herd behavior
Tool little, too late
Vicious cycle
Ratings failures
Possible reason for rating failure:

Information problem
Limited resources
Revenue bias
Incentive problems

Sovereign Default Spread
More countries have increased their reliance on bank loans to bonds and the market prices commanded by
these bonds have yielded an alternate measure of sovereign default risk.
When a government issues bonds, denominated in a foreign currency, the interest rate on the bond can be
compared to a rate on a riskless investment in that currency to get a market measure of the default spread
for that country. To illustrate, the Brazilian government had a 10-year dollar denominated bond
outstanding in July 2015, with a market interest rate of 4.5%. At the same time, the 10-year US treasury
bond rate was 2.47%. If we assume that the US treasury is default free, the difference between the two
rates can be attributed (2.03%) can be viewed as the markets assessment of the default spread for Brazil.
Even though there is strong co relationship between sovereign ratings and market default spreads, there
are advantages for using default spread for assessment of default risk. The primary advantages are:
The first is that the market differentiation for risk is more granular than the ratings agencies; thus, Peru
and Brazil have the same Moodys rating (Baa2) but the market sees more default risk in Brazil than in
Peru.
The second is that the market-based spreads are more dynamic than ratings, with changes occurring in
real time.
Disadvantages of Default Risk spreads

Need for a risk free security:
Difficult to compare local currency bonds: Local currency bonds issued by governments cannot be
compared to each other, since the differences in rates can be due to differences in expected
inflation.
Greater volatility: Default spreads are more volatile than ratings and can be affected by variables
that have nothing to do with default. Liquidity and investor demand can sometimes cause shifts in
spreads that have little or nothing to do with default risk.

Results of the studies between default spread and ratings
Sovereign bonds with low rating tend to trade at higher interest rate indicating higher default
spread and are more likely to default.
The sovereign bond market leads ratings agencies, with default spreads usually climbing ahead of
a rating downgrade and dropping before an upgrade.
Despite having lead lag relationship, a change in sovereign ratings is still an informational event
that creates a price impact at the time that it occurs
Exercise
Q.1 An analyst makes following comment with regard to the measurement of country risk. Identify
which one of the following is correct.
1. The various agencies measurement techniques of country risk is standardized.
2. Country risk score of 80 in comparison of a score of 40 can be concluded as double risky country.
3. Political Risk Service (PRS) uses three dimensions to assess country risk.
4. Under the economists country scoring system, low score means higher risk and high score means low
risk.

Exercise
Q.2 A country is having a sovereign rating of BBB and BBB-. The analyst concludes the following
from these ratings
A. BBB is countrys local currency rating and BBB- is its foreign currency rating.
B. Local currency rating is generally higher or equal to foreign currency rating due to flexibility of
country to print its own currency.
Which one of the following option is correct?
1. Both A & B are correct
2. Neither A nor B is correct
3. Either of A or B is correct
4. B is correct but A is not correct
Exercise
Q.3 Sovereign default will not lead to
1. GDP growth decline
2. Rating downgrade
3. Stable bilateral trade
4. Political change

Exercise
Q.4 Mr. y, a portfolio analyst, is considering to invest in an emerging country. He is going
through various reports and also ask his associate analyst to get the rating agencies report on
the same country. He , after going through the reports, asks the associate to identify any
potential limitation in the rating agencies assessment. The associate responds the following:
A. Rating agencies may walk in tandem with action of other rating agency compromising
independence.
B. Rating actions by rating agencies can come with a lag .
C. Rating agencies do face conflict of interest in sovereign rating.
Which are the associates statements are correct?
1. A, B and C
2. A and C only
3. C only
4. A and B only
Answers
A.1: Option 3
A.2: Option 1
A.3: Option 3
A.4: Option 4

Credit Risk
Credit Risk
Credit risk arises out of:

Non-payment/ re-scheduling/ delayed payment of promised payments
Unfavorable credit migration (rating downgrades)
Credit losses leads to economic losses, consisting Expected Loss and Unexpected Loss
Deriving Economic Capital for credit risk mitigation:

Expected Losses (EL)
Unexpected Losses (UL)
Unexpected Loss Contribution (ULC)
Economic Capital for Credit Risk

Expected Loss (EL)
EL = (Assured payment at Maturity Time T )* Loss Given Default * (Probability that the default
occurs before maturity Time T)
The term Assured payment is more relevant for bonds than loans
For Bank Loans the terms Assured Payment is better replaced by Exposure
EL = Exposure * LGD*PD
EL is the amount the bank can lose on an average over a period of time
However, during the same period because of macroeconomics conditions, or market trends there
could be some deviation from the expected loss which is known as Unexpected Loss (UL). Banks
need capital to take care of this loss
Unexpected Loss
UL is the estimated volatility of the potential loss in value of the asset around its EL
UL is the standard deviation of the unconditional value of the asset at the time horizon
UL = s.d. of expected asset value
h>&>'>'&
Underlying assumption that EDF is independent of LGD. In case it is not so then correlation between LGD
and EDF terms will come into picture. Though it has been found that they will affect the result only slightly
What if all the variance were zero?

Unexpected Loss Contribution/ Economic Capital
Unexpected Loss Contribution

Adding risky (credit) assets to existing portfolio diversifies the overall credit risk
ULC is function of Expected Loss, Exposure Amount and Correlation of the exposure to the rest of
the portfolio
Economic Capital for Credit Risk

The required capital reserve which acts as a buffer against insolvency for the bank in the event of
the default by an obligor is known as economic capital
With the fluctuation in the unexpected loss the economic capital will also fluctuate
Economic Capital (EC) = ULC x CM (Capital Multiplier)
Capital Multiplier is distance between expected outcome and chosen confidence interval for unexpected
outcome
Thank z
www.edupristine.com
Foundation of Risk Management - I
Agenda
Risk Management: A Helicopter View
Corporate Risk Management: A Primer
Corporate Governance and Risk Management
What is ERM?
Risk-Taking and Risk Management by Banks
Financial Disasters
The Credit Crisis of 2007
Risk Management Failures: What are they and When do they Happen?

Risk Management: A Helicopter View
What is Risk?
Risk, in traditional terms, is viewed as a negative. The dictionary defines risk as exposing to danger
or hazard.
You cannot have opportunity without risk and that offers that look too good to be true (offering
opportunity with little or no risk) are generally not true. By emphasizing the upside potential as well
as the downside dangers, this definition also serves the useful purpose of reminding us of an
important truth about risk.

Risk Management Process
Identify Risk Exposures
Measure & Estimate Risk

sk Find
nd instruments
i & facilities to
Exposure shift or trade risks
Assess costs
t & benefits of
Assess effects of Exposures
Instruments
Form a Risk Mitigation Strategy:

Avoid
Transfer
Mitigate
Keep
Evaluate Performance
Source: Risk Management: A Helicopter View
Expected Loss vs Unexpected Loss

Expected Loss
Some of the costs that is expected to incur in our daily life are very large in terms of annual budgets, like
food, fixed mortgage payments, college fees, etc. These cost are big but they are not a threat to ones
ambitions because they are reasonably predictable and are already accounted for in our annual plans.
Unexpected Loss
A catastrophic loss of a magnitude well beyond losses seen in the course of normal daily life. The real risk is
that the daily life costs will suddenly rise in an entirely unexpected way, or some other cost will appear from
nowhere and steal the money that was kept aside for our annual budget.

Classification of Risks
Equity Price Risk
Commodity Price General Market

Risk Risk
Market Risk Trading Risk
Interest Price Risk Specific Risk
Gap Risk
Financial Risks Foreign Ex. Risk
Portfolio
Operational Risk Issue Risk
Concentration
Credit Risk
Legal & Regulatory
Transaction Risk Issuer Risk
Risk
Risks
Business Risk Counterparty Risk

Funding Liquidity
Risk
Liquidity Risk
Trading Liquidity
Risk
Strategic Risk
Reputation Risk
Source: Risk Management: A Helicopter View
Corporate Risk Management: A Primer

Disadvantages of Managing Risk
Under the assumption that capital markets are perfect, M&M reasoned that whatever the firm can
accomplish in financial markets, the individual investor in the firm can also accomplish or unwind
on the same terms and conditions.
William Sharpe establishes that firms should not worry about the risks that are specific to them
because all the specific risks are diversified away in a large investment portfolio, under the perfect
capital markets assumption.
Hedging is a zero-sum game and cannot increase earnings or cash flow.
Confidential information might be revealed by the means of forward transaction.(The scale of

sales envisaged in certain currency)
Advantages of Managing Risk

Management have an interest in reducing risks because the results of a firm provide signals to all
the stakeholders concerning the management competencies.
With the effect of progressive tax rates, volatile earnings induce higher taxation than stable
earnings.
Risk management activities allow management better control over the firms natural economic
performance.
Risk reduction activity offers synergies with the operations of the firm (Hedging the price of an
input of a firm can stabilize its cost and pricing policy, which in turn offer a competitive advantage
in the marketplace that could not be replicated by any outside investors)

Hedging Operations vs Hedging Financial Positions
It is important to make a clear distinction between hedging activities related to the operations of
the firm and hedging related to the balance sheet.
If a company chooses to hedge activities related to its operations, such as hedging the cost of raw
materials, this clearly has implications for its ability to compete in the marketplace.
For example When an American manufacturing company buys components from French company in euros,
the American company can opt to avoid the foreign currency risk by hedging the exposure.
This hedge has both a size and a price effect-i.e., it might affect both the price and the amounts of products
sold.
Hedging Operations vs Hedging Financial Positions

When a company chooses to hedge activities related to the balance sheet, there are three
different scenarios for this.
If the capital markets are perfect, it is suggested that the company should not hedge
If the financial markets are in some sense perfect, it is suggested that the company can do hedging of balance
sheet provided it is properly controlled and the firms policy is fully transparent and disclosed to all the
investors.
If the financial markets are not perfect, then the company may get some advantage from hedging its balance
sheet. It may have a tax advantage, benefits from economies of scale.
Two conclusions
Firms should risk-manage their operations
Firms may also hedge their assets and liabilities, so long as they disclose their hedging policy

Financial Products used in Hedging
Over The Counter (OTC) Derivatives

Non-deliverable forwards
Currency Swaps
Interest Rate Swaps
Exchange Traded Derivatives
Currency Options
Interest Rate Futures
Commodity Futures
Securitizations
CDO
MBS
Exotic Options, Structured Products and Non-Traditional Derivatives
Weather Options
Electricity Options
Asian Options
Steps of Risk Management into Practice
Board determines the Risk Appetite

Determining the Objective Whether to hedge accounting profit or economic profit
Short-term profits or long-term profits
Map the relevant risks

Mapping the Risks Estimate current & future magnitudes
Identify relevant Instruments to mitigate the risks

Instruments for Risk Compare ways to hedge risks as transferable or insurable
Management Evaluate the likely costs and benefits
Create/select certain pricing & hedging models

Constructing and Implementing Form specific strategies
Strategy Deep understanding of tools is must
Evaluate the performance periodically

Performance Evaluation Assess the overall goals

Corporate Governance and Risk Management
Corporate Governance & Risk Management

Primary roles & responsibilities of board
Board is to look after the interests of shareholders
Board needs to be sensitive to the concerns of other stakeholders such as debt holders
Board needs to be on the alert for any conflict such as agency risk
Board needs to ensure that it develops a clear understanding of the banks business strategy and
the fundamental risks & rewards
Board must characterize an appropriate risk appetite
Board must ensure that the bank has put in place an effective risk management program that is
consistent with the fundamental strategic and risk appetite choices
Board should ensure that business and risk management strategies are directed at economic
rather than accounting performance
Board must ensure that the bank has put in place an effective risk management program that is
consistent with the fundamental strategic and risk appetite choices
Board should ensure that business and risk management strategies are directed at economic
rather than accounting performance
An effective board should establish ethical standards
Board should ensure that the information it obtains about risk management is accurate and
reliable
Board must make sure that all the mechanisms used to delegate and drive risk management
decisions are functioning properly

Committees & Risk Limits
Special focus on corporate governance in the Banking Industry
At most banks, the board charges its main committees with ratifying the key policies and associated
procedures of the banks risk management activities. The various committees are:
Responsible for the accuracy of the banks financial & regulatory
reporting and also ensure that the bank complies with best-
Audit Committee
practice standards in key activities such as regulatory, legal,
compliance and risk management
Improve the overall efficiency and effectiveness of the senior risk

Risk Advisory Director committees and the audit committee, as well as the
independence and quality of risk oversight by the main board
Responsible for independently reviewing the identification,

Risk Management measurement, monitoring, and controlling of credit, market, and
Committee liquidity risks, including the adequacy of policy guidelines and
system
Compensation Determine the compensation of top executives

Committee
Roles & Responsibilities in Practice
Risk Committee of the Board Approves market risk tolerance each year
Step 1: Approve market risk tolerance, stress and

Senior Risk Committee perform limits each year, reviews business unit
mandates and new business initiatives
Step 2: Delegates authority to the CRO and hold

Senior Risk Committee addition authority in reserves approved by the risk
committee of the board
Responsible for Independent monitoring of limits;

CRO may order positions reduced for market, credit, or
operational concerns
Heads of Business Share responsibility for risk of all trading activities
Responsible for risk and performance of the

Business Unit Manager business. Must ensure limits are delegated to
traders
Source: Corporate Governance and Risk Management

Role of the Audit Committee
Provide an independent assessment of the design and implementation of the banks risk
management.
What is ERM?

Definition of ERM
The Committee of Sponsoring organization of the Treadway Commission (COSO) in 2004 defined
ERM as
ERM is a process, effected by an entitys board of directors, management, and other personnel,
applied in strategy setting and across the enterprise, designed to identify potential events that may
affect the entity, and manage risk to be within its appetite, to provide reasonable assurance
regarding the achievement of entity objectives.
International Organization of Standardization (ISO 31000) defined ERM as
Risk is the effect of uncertainty on objectives and risk management refers to coordinate activities
to direct and control an organization with regard to risk.
ERM is about integration in three ways:

1. ERM requires an integrated risk organization
2. ERM requires the integration of risk transfer strategies
3. ERM requires the integration of risk management into the business processes of a company
Benefits of ERM
Three major benefits of ERM are:
Increased
Improved business
organizational Better risk reporting
performance
effectiveness

Examples of benefits of ERM
Benefits Company Actual Results

Market value Top money centre bank Outperformed S&P 500 banks by 58% in stock
improvement price performance
Early warning of Large commercial bank Assessment of top risks identified over 80% of
risks future losses; global risk limits cut by one-third
prior to Russia crisis
Loss reduction Top asset management 30% reduction in the loss ratio enterprise-wide: up
company to 80% reduction in losses at specific business
units
Regulatory capital Large international $1 Billion reduction of regulatory capital
relief commercial and requirements, or about 8-10%
investment bank
Risk transfer Large property and $40 million in cost saving, or 13% of annual
rationalisation casualty insurance reinsurance premium
company
Insurance premium Large manufacturing 20-25% reduction in annual insurance premium
reduction company
Role of Chief Risk Officer

Providing the overall leadership, vision and direction for ERM
Establishing an integrated risk management framework for all aspects of risks across the
organization
Developing risk management policies, including the quantification of the firms risk appetite
through specific risk limits
Implementing a set of risk indicators and reports, including losses and incidents, key risk
exposures, and early warning indicators
Allocating economic capital to business activities based on risk, and optimizing the companys risk
portfolio through business activities and risk transfer strategies
Communicating the companys risk profile to key stakeholders
Developing the analytical, systems, and data management capabilities to support the risk
management program

Components of ERM
Corporate Governance
Establish top-down risk management
Line Management
Business strategy alignment
Portfolio Management
Think and act like a fund manager
Risk Transfer
Transfer out concentration or inefficient risks
Risk Analytics
Develop advanced analytical tools
Data and Technology

Resources Integrate data and system capabilities
Stakeholders
Management Improve risk transparency for key stakeholders
ERM and decision making
ERM is an integral part of organizations culture and is embedded in its corporate strategy,
planning and day to day decision making
As part of ERM, an organization must consider its risk appetite, alongside its goals and operational
tactics. Management should take three steps w.r.t. risk appetite:
1. Develop risk appetite
2. Communicate risk appetite
3. Monitor and update risk appetite

Risk Appetite
As per COSOs Enterprise Risk Management Integrated Framework definition, Risk Appetite:
Is the amount of risk, on a broad level, an entity is willing to accept in pursuit of value. It reflects
he entitys risk management philosophy, and in turn influences the entitys culture and operating
style.
Is strategic and is related to the pursuit of organizational objectives
Forms an integral part of corporate governance
Guides resource allocation.
Assists the organization in aligning the organization, people and processes in designing the
infrastructure necessary to effectively respond to and monitor risks.
Influences the organizations attitude towards risk
Is multidimensional
Requires effective monitoring of the risk itself and of the organizations continuing risk appetite
Risk Appetite
Considerations affecting risk appetite
1. Existing Risk Profile current level and distribution of risks across the entity and across various
risk categories
2. Risk Capacity The amount of risk the entity is able to support in pursuit of its objectives
3. Risk Tolerance Acceptable level of variation an entity is willing to accept regarding the pursuit of
its objectives
4. Attitudes towards Risk The attitudes towards growth, risk and return
Steps in adopting Risk Appetite
1. Develop Management develops, with broad review and concurrence, a view of the
organizations overall risk appetite
2. Communicate The view of risk appetite is translated into a written or oral form that can be
shared across the organization
3. Monitor Management monitors the risk appetite over time adjusting how it is expressed as
business and operational conditions warrant

Risk Appetite
Risk Appetite Statement

Effective way to communicate across an organization, the sense of acceptable risks
Should be clear and descriptive enough to guide actions of personnel across the organization
Serves as a basis for evaluating and monitoring the amount of risk an organization faces vis--vis
acceptable range
Risk Appetite and Risk Tolerance

Risk tolerance is defined as the acceptable level of variation relative to achievement of a specific
objective, and often is best measured in the same units as those used to measure the rated
objective. In setting risk tolerance, management considers the relative importance of the related
objective and aligns risk tolerances with risk appetite. Operating within risk tolerances helps
ensure that the entity remains within its risk appetite and in turn that the entity will achieve its
objectives.
Risk tolerances are always related to risk appetite and objectives. Together they guide the
organizations actions
Risk-Taking and Risk Management by Banks

What is Good Risk and Bad Risk from Banks shareholders
perspective?
Good Risk: Provides ex-ante reward on standalone basis i.e. Value creating risk opportunities
Bad Risk: Leads to loss of value without providing an opportunity for value creation. This type of risk
are danger and needs to be avoided.
While assuming good risk, bank will increase its overall risk as bank will invest in assets and projects
that are valuable but can also lead to a loss in value because of an adverse impact which will
increase its risk of financial distress and its ability to create value through liabilities. Hence, a well
governed bank will have to identify an optimal level of risk from its shareholder perspective. Bank
should try not to deviate much from its optimal level of risk amount.
To make it simple, bank should take up the project that increases its value while taking into
consideration the cost associated with the project on total risk of the bank. However, practically, risk
taking decisions are taken throughout various departments of the bank almost continuously and
hence, each risky decision can not be considered in isolation but must be assessed in terms of its
impact on overall risk of the bank.
Determining the risk appetite

Targeting a credit rating: Credit rating corresponds to probability of default. For example, the
annual average default rate for A-rated credits is 0.08% according to S&P. Hence, by targeting credit
rating, bank will, in turn, target a specific probability of default and achieves its desired level of risk.
So, should banks always target a AAA rating? The answer is No. Because, by targeting the highest
credit rating bank will also limit its opportunity in taking up projects which can increase shareholder
wealth. A target of optimal credit rating for a bank is a function of composition of its business
activities.
For example, A bank with more of a deposit franchise and with more relationship lending is likely to
prefer a higher rating than an institution that is engaged in more transactional activities. Similarly, a
bank that enters in long-term derivatives contracts might find a higher rating more valuable than
one that does not because the counter partys assessment of credit risk is paramount in such type of
long term contracts.
Hence, the ratings that maximizes the value of the bank differs across banks.

Bank value as a function of bank risk measure by the banks
credit rating
1. The relation between ratings and bank value is concave.
2. Bank Safes value falls steeply if the bank is riskier than its target rating and increases only
moderately as it increases its risk towards the target rating. Bank Risky has a substantially
different relation between value and its rating. Its target rating is BBB and its value increases
substantially as it increases its risk towards its target and falls sharply if it exceeds it. For both
banks, having too much risk is extremely costly in terms of their value. However, for one bank
having too little risk has little cost while for the other it has a large cost.
Determining Risk Appetite (Cont..)

Tailoring risk: Banks may also want to determine risk in a more complex way by specifying how they
are affected by specific shocks. For example, a bank might choose to set a level of risk such that it
can survive a major recession with only a one-notch downgrade in its credit rating.
Flexible V/s Inflexible risk appetite: Bank can not have an inflexible risk appetite as its opportunity
set is changing constantly. However, at the same time, a small change in opportunity set can not
change banks risk appetite.

Ways in which Risk Management can add or destroy value
If deviation from a target risk leads to greater loss than the cost incurred to have the risk
management department, then the risk management departments fixed cost is justified. Merely
paying a very high salary to CRO in comparison with CEO does not guarantee a good risk
management practise.
The risk management function of the bank should be evaluated with the value it generates in terms
of protecting the bank from taking excessive risk.
Also, flexibility of risk management function is paramount so as not to merely control the risk but
also provides framework for exploiting value creating opportunities.
Risk management function should take the perspective of the entire bank level risk rather than
individual level department risk. Individual business units are generally not capable of determining
that whether their risk taking will add or destroy value at the firm level. Hence, risk management
function can add value by requiring all the business units to take a firm wide perspective while taking
risk.
Structural challenges and limitations to effective risk

management
If top management could know exactly what is the risk faced by the bank and if it can hedge risk at
zero cost, the risk management is at bank will be straight forward. As long as risk-takers in the bank
only took projects that create value regardless of their risk, top management would have no reason
to monitor the risk in the sense of assessing risk decisions made by employees. All that the bank
would have to do is measure the risk taken within the bank and control it through hedging.
Limitations of the above method for managing risk:
1. Risk Management technology limitations
2. Hedging limitations
3. Risk-taker incentives limitations
Due to the above limitations, risk has to be managed and monitored through out the organization.

Risk Management within the bank
Risk management within the firm can not be merely a verification function. Also, the risk manager
and the manager of business line should have independence for proper monitoring of risk.
Use of VaR for setting limits of risk:
There are many different units in a bank and a separate VaR can be calculated for each unit and then
an aggregate VaR can be calculated on a firm wide basis. For instance, a VaR can be estimated for the
book of a trader as well as for the unit that the trader belongs to. Starting from the smallest units for
which VaR is estimated, the VaRs can be aggregated to the bank level keeping the correlations in
risks across these units in perspective. Further, using the VaRs of the smallest units and the
correlations, it is possible to assess how each unit contributes to the risk of the bank. For instance, a
bank could estimate how much of the risk of the bank as measured by VaR is accounted for by a
specific trader.
The firms risk appetite specifies firm level VaR limit and VaR limit should depend on the profitability
of the risk taking unit in relation of its own VaR.
As profit opportunities for various departments change, it follows that limits cannot be
unchangeable. When profit opportunities appear for a sector of the bank, it makes sense for limits
to be adjusted. However, if the banks risk appetite has not changed, VaR limits cannot be increased
in one sector of the bank without being decreased elsewhere. Of course, if profit opportunities
change for the bank as a whole, so that the expected return from risk-taking increases, it can be
optimal for the bank to change its risk appetite and, as a consequence, its firm-wide VaR limit as
well.
Limitations of VaR as a risk management tool

1. Aggregating VaR to obtain a firm wide risk VaR is difficult
2. VaR does not capture all the risks
3. VaR has substantial model risk
Bank divides its risk into category of Market, Credit and Operational risks. A firm wide VaR,
aggregating all the categories of risk typically does not capture all the risks .i.e. business risk
Business risk, non-interest income of bank (processing fees etc) unexpected changes in interest rates
and unexpected changes in credit spread needs to be modelled separately.
VaR also does not capture funding liquidity risk. A shock to funding can force the bank to sell assets
at a loss. Further, shocks to funding are more likely to happen in periods when markets for securities
are themselves less liquid, so that selling assets quickly will be costly as they are sold at a discount.
Hence, the same needs to be modelled separately.

Also, the market, credit and operational risks aggregation should consider the correlation among
them. The higher the correlation among various risk factors leads to greater firm wide VaR and lower
correlation leads to lower firm wide VaR. However, there is insufficient data available to make
correlation estimates. Also, errors in computing such estimates may result in bank holding
insufficient capital exceeding its targeted risk appetite.
Different types of risk will lead to different statistical distributions and typically credit and
operational risk have non-normal distributions and is difficult to model them.
Risk that are unknown, black swan or unknown unknowns are important only when VaR is estimated
at extremely low probability levels like 0.06% VaR. A focus on historical data and the use of
sophisticated statistical technique is not much useful for developing forecast for this extremely rare
events. A stress testing technique can be applied to find whether bank will withstand the
occurrence of those rare events.
Impact of banks governance , incentive structure and risk

culture on its risk profile
There is a growing interest between firms risk governance and its risk outcome and risk
performance. However the challenges to these are:
1. There is limited data available about how the risk function is organized in firms.
2. Banks risk appetite has also greater influence on banks risk function apart from governance
3. At a firm level, poor ex-post performance can be consistent with very good risk management

Incentive Structure
Incentive structure should not reward managers only for performance of their business unit but also
for the risk taken overall value creation of the firm.
The concept of risk capital needs to be considered while designing compensation for the managers.
Risk Culture:
Research concludes that: Companies employing honest and trustworthy managers are more
profitable and were given high valuation.
Shareholder governance improvement would change a firms culture from focusing on employee
integrity and customer service to focusing on end results.
Examples of impact of corporate culture on risk management

1. A loan officer who can decide whether a loan is granted makes a decision to take a risk has
information about that risk that nobody else has in the organization and being so close to the
information, he would have ensured that nobody may ever know whether the decision was right
from the perspective of the firm for a number of reasons.
First, it may not be possible for the loan officer to communicate credibly information that she had.
Second, the loan officer may have incentives to grant loans that she knows should not be granted.
Third, loan outcomes are of limited use since expected defaults are not zero.
Solution for the bank : Minimize the discretion of the loan officer by relying on statistical models for the
decision. However, blindly following the results of model reduces flexibility and subjectivity which is not
captured by the model. Bank should rather create a underwriting culture .A bank with an underwriting
culture that is highly focused on the interests of the bank will make it harder for a loan officer to deviate
from the social norms within the bank as employees who are in contact with the loan officer might be able
to assess when the officer deviates from the banks norms in away that neither risk managers nor executives
could.
2. A corporate culture that has traders in confrontation with risk managers, would limit the risk
managers ability to assess correctly the risk of positions and how to mitigate this risk. In this case,
the energy of risk managers has to be devoted to fighting with traders and figuring out what they
might be hiding.

Examples of impact of corporate culture on risk management (Cont..)
3. A trader taking positions that cant be expected to be profitable for the firm but might be very
valuable to the trader if it pays off, if observed by the supervisor, needs to be understood and a
detail dialogue could have been warranted in a firm having a risk culture. Traders position is not
breaching the risk level does not mean he cant be questioned for the unusual trades. In some
firms, supervisor would say nothing. In other firms, he would start a dialogue with the trader. In
the latter firms, one would expect risk taking to be more likely to increase value.
Financial Disasters

Case Studies Metallgesellschaft (MG)
It used stack and roll hedging strategy

In 1991, it offered fixed price contract for supplying gasoline for 5 to 10 years. In order to hedge
MG took long positions in near month futures and rolled the stack into next near month
contract every time by decreasing the trade size gradually so as to match the stack with
pending short position (in long term supply contracts)
MG bought futures on NYMEX to offset its forward commitments exposure with hedge ratio of
one (every barrel was hedged)
As these derivatives were short-term thus MG had to roll them forward every month-end or
term-end till 5-10 years or the contracts end
Company was exposed to rising spot prices. It eventually lost more than USD 1.5bn in 1993
It had various risk exposures .such as Basis Risk, Market Risk, Funding Liquidity Risk
Chase Manhattan Bank / Drysdale Securities
Drysdale Chase Manhattan Bank

Investors
Inexperienced Managers:
Capital: $20 Mn 1. Thought they were just
Borrowed middlemen
2. Didnt realize contract
Debt Market: $300 Mn indicated Chase taking
(Unsecured Loan) full responsibility of
debt

Union Bank of Switzerland (UBS)
Drysdale Chase Manhattan Bank
Losses between 1.1Bn and 1.4Bn from 1997-8

UBS held a large position in LTCM (40% direct, 60% Options)
Equity Derivatives team not scrutinized by Corporate Risk Team
Head of Analytics compensation was in line with fund performance
Equity Derivative Losses due to:
Change in British Tax laws regarding valuation of long dated stock options
Large position in Japanese Bank Warrants (were not adequately hedged)
Correlation assumptions on long dated equity options was not in line with the rest of the market
Modeling Deficiencies
Case studies Sumitomo
Yasuo Hamanaka copper trader at Sumitomo manipulated copper prices on London

Metal Exchange
Fall in copper prices in June 1996 after revelation of Hamanakas unfair dealings led to ~US$2.6 Bn
loss for Sumitomo
Positions were so large that company could not liquidate them completely
Hamanaka used his independence to trade in the market on behalf of the company and
manipulated the copper prices by buying physical copper in large quantities and storing in the
warehouse thereby creating lack of copper in the market
He sold put options to collect the premiums as he thought he can push the prices up and thus
writing put options was not risky for him
Though, he never imagined that he could be susceptible to steep decline of copper prices
It had various risk exposures .such as Operational Risk, Employee / People Risk, Liquidity Funding
Risk, Market Risk

Case studies LTCM
LTCM was a hedge fund using highly leveraged arbitrage trading activities in fixed income
in addition to pairs trading. Before failing in 1998, it had given spectacular returns in 1995-97
periods (up to 40% post-fees). Post Russian default on its ruble denominated debt, LTCM lost more
than USD 4bn in 4 months
LTCM used proprietary mathematical models to engage in arbitrage trading in US, Danish, Russian,
European and Japanese Govt. bonds. In 1998, LTCMs positions were highly leveraged (1:28) with
~US$5: 130 Bn of equity and assets
LTCMs model assumed maximum volatility of 20% annually. Based on its models, it was expected
to losses more than ~US$500 Mn in once in 20 months
It had its bet on convergence of Russian and American G-sec yield, which however diverged
after Russian default. Its failure led to a huge bailout by large commercial and merchant banks
under the guidance of Federal Reserve
It had various risk exposures .such as Model Risk, Funding liquidity risk, Sovereign Risk,
Market Risk
Case studies Barings Bank

Nick Leeson, a trader at Barings-Singapore, led to failure of one of the oldest banks in UK: Barings
Bank
Modus operandi: Leeson was an employee of Barings-Singapore and was responsible to profit
from arbitrage opportunities between Osaka exchange and Nikkei 225 futures prices. Without
knowledge of anyone at London HO and Barings Singapore, he started speculative positions and
incurred heavy losses. He hid the losses in a fictitious account: 88888 and continued taking more
risks to recover earlier losses. Finally his activities were uncovered in 1995 when accumulated
losses were US$1 Bn
Reported profits: Leeson was considered as a star-performer and reported record-profits for many
years. Instead he was hiding all losses in a fictitious account 88888. Trading transactions
reported by Leeson were in nature of Switching activities (arbitrage) that are not supposed to
generate as high profits as reported by Leeson. This should have raised some suspicion amongst
the senior management , but it is clear that they were clueless about true profit generating
potential of Leesons trading activities. Leeson also maintained over-night open positions &
position in options that he was not authorized to do
Funding of the losses: London HO: Barings-London funded the position of Barings-Singapore
thinking that they are funding the clients (as loans to clients). They were actually funding losses on
proprietary book (i.e. trading by Leeson). In 1995, when funding the losses was becoming difficult,
Leeson used artificial trades to reduce margin calls from SIMEX

Case studies Barings Bank (cont.)
Settlements Department was also unable to reconcile the funding transactions but failed to initiate
sufficient action
Credit department: Since the funding was being recorded as loans to clients, credit department should
have paid attention to growth of advances to clients, but it failed to do so
Financial control system was inadequate in terms of understanding the requirement of funding by Baring
London to Barings Singapore. Surprisingly, even though funding by Barings London to Singapore
operations
was recorded as advances to clients, it was not reported to regulators as part of large exposure to
particular clients
Leeson was allowed to be in charge of both front office and back office, helping him conceal his activities
for long. Even though internal audit suggested separation of front and back office function, their
recommendations were never implemented. Leeson faced negligible supervision in his back office
activities
Barings Bank had a Matrix structure of organization, but without proper controls 7 communication
channels, the organization structure added to confusion as to who was reporting to whom
External auditors C&L Singapore once pointed a spurious receivable of Euro 50 Mn from an entity SLK.
Instead of going into details of this transaction, the management requested the auditor that no
reference to this transaction be made in their report and attributed the spurious receivable to an
operational error. External auditors also failed in examining the nature of large funding done by London
HO to Singapore operations that was reported as advances to clients but never reconciled to individual
accounts of clients
Case studies Barings Bank (cont.)

Role of Bank of England (regulator for Barings London): BOE knew the extent of profits reported by
Barings Singapore and the large exposure of Barings London to Singapore operations. BOE allowed
concession to Barings Bank that exposure to Osaka exchange and SIMEX can exceed 25% of its
capital base, but failed to put a limit on the extent of this exposure. Consequently, Barings
exposure reached 73% of capital base to Osaka exchange and 40% on SIMEX at a point in time
In 1995: Market rumors suggested market concerns about Barings positions on Osaka Stock
exchange. SIMEX (Singapore exchange) sought an assurance from Barings-Singapore if they would
be able to fund margins calls on short notice, if market moves against the bank
Important Lessons:
Internal controls & systems failed in case of Barings
Unauthorized & hidden trading activities by a single trader led to huge losses who was responsible for both
front-office and back-office operations
Internal departments like credit & settlement dept., external auditor & supervisors failed to detect the true
position prior to the Banks collapse

Case studies Bankers Trust (BT)
BT was sued by four major derivatives clients including P&G for misselling (classified as
operational risk) i.e. selling derivatives and structured products without fully explaining the risks
involved
BT suffered huge reputational loss. It also had to settle with four customers leading to monetary
losses exceeding US$170 Mn
P&G had invested in heavily leveraged IRS that lost substantially after US Fed raised interest rates
in 1994
(i.e. same time when Orange county lost money on its inverse floaters)
Good stakeholder management practices are required to safeguard interests of all parties involved
including clients, to prevent such losses
Case studies Enron, JPMorgan and Citigroup

Enron, for years, had been engaging in dubious accounting practices to hide the size of its
borrowings from investors and lenders. Enron disguised these borrowings as transactions part of
oil future contracts.
Being a major player in energy markets, Enron was expected to participate in future contracts on
oil. But the futures contract undertaken by Enron, did not involve taking any position on oil price
movements. Enron sold oil for future delivery, while getting upfront cash payments. And than
agreed to buy back that oil for a fixed price in future. Effectively resulting in a contract which does
not include delivery or taking position on oil price.
The above transaction was equivalent to borrowing cash and paying it back later i.e. a loan
arrangement. Enron did not report it transparently to its investors and regulators. It projected the
loan as oil futures contracts. The major accounting firm Arthur Andersen had also helped in
accounting for these transaction in its books.
Eventually Enron filed for Bankruptcy in 2001.
Its bankers and principal counterparties: JPMorgan and Citigroup however had treated these
transaction as loan in their respective account, and have reported to investors and regulators.
They had not deceived their investors or lenders.
But JPMorgan and Citigroup, had played a part by helping committing the fraud. In the end, they
had to pay a combined $286 million.

Case studies Greece and Goldman Sachs
Following the Enron episode, Investment banks and regulators agreed to put new controls in place,
to make sure that their clients were accounting derivative transactions with them, in ways
transparent to investors.
However the controls were not followed strictly. Goldman Sachs helped Greece in hiding its level
of indebtedness from its EU partners, by disguising debt as an interest rate swap.
The swap was done deliberately at an off market rate, creating an upfront payment to Greece, that
was scheduled to be paid back by Greece with suitable interest over the course of Swaps life.
However, effectively the transaction was a loan arrangement, which was disguised as a derivative
transaction.
Case studies Kidder Peabody
In 1991 Kidder, Peabody & Co. hired Joseph Jett to their Strips trading desk
In his first 6 months the desk did not make money, then reported profits grew quickly ($32 Mn in
1992,
$151 Mn in 1993, and $81 Mn in the first quarter of 1994)
Jett was Kidders man of the year in 1993 and was awarded a $9 million bonus (his boss received
a $20 Mn bonus in 1993, primarily for overseeing Jetts activities)
It was later reported that Jetts desk lost $85 Mn during that period and a $358 Mn cumulative
write-down. The false profits were the result of an error in their internal accounting whereby a
zero-coupon bond was treated like a coupon bond. Thus a forward reconstitution, which
exchanges a zero for an identical coupon bond, created instant profits

Case studies Kidder Peabody (lessons learned)
Beware of increasing trading volume. Kidders balance sheet was $25 Bn in 1991, $273 Bn in 1992,
and $1,567 billion in 1993. Any time notional amount grows dramatically it is wise to double check
the business model
Most value-destroying activity is not illegal. Kidder and GE shareholders brought suit against Jett
for fraud, but it was impossible to prove because Jett was openly engaged in his trading fraud
Compensation is an imperfect signal of competence
Credentials are imperfect signals of competence. Joseph Jett had an MIT undergraduate degree
and a Harvard MBA. Jetts boss, Ed Cerullo, had a respectable career on Wall Street, and was
considered street smart and technically savvy
The back office and auditors should not be fearful of traders, and it is managements job to make
this so
Case studies Allied Irish Bank (AIB)
1993: John Rusnak, who had been working for Chemical Bank in New York, joins First Maryland
Bancorp as a foreign exchange trader
1999: Allfirst is formed from the merger of First Maryland Bancorp (in which AIB first took a stake
in 1983) and Dauphin Deposit Corporation (which AIB acquired in 1997). Susan Keating becomes
Allfirst chief executive
June 2001: John Rusnak is promoted to managing director in charge of foreign exchange trading, in
the global trading division of the treasury funds management section, or front office
Late December, 2001: Allfirst officials start to become suspicious about the sums being demanded
by Rusnak to cover his trading
January 10, 2002: Keating is appointed to the AIB chief executive committee, the group
responsible for developing corporate strategy and overseeing management of AIB group
February 4, 2002: Rusnak fails to show up for work on Monday morning
February 6: AIB says it is investigating a suspected $750 Mn fraud at Allfirsts Baltimore HQ, and
warns that it will take a one-off charge of E596 Mn ($520 Mn) to cover the resulting losses
February 8: Eugene Ludwig, a former US Comptroller of the Currency, is hired to compile a report for AIB on
the affair

Case studies Allied Irish Bank (AIB) (cont.)
February 19: AIB chief executive Michael Buckley says that the origins of the scandal might stretch
back to 1997, and gives the final figure for losses as $691 Mn
March 12: Buckley and AIB chairman Lochlann Quinn offer their resignations to the AIB board, but
neither resignation is accepted
March 13: The Ludwig Report is published jointly by Ludwigs Promontory Financial Group and law
firm Wachtell, Lipton, Rosen & Katz
March 14: Allfirst and AIB announce that six executives who were responsible for oversight of
Rusnaks activities are to be dismissed. A number of organizational and structural changes are also
announced, including the appointment of an individual to oversee risk management across the AIB
group
March 17: AIB denies rumors that Allfirst CEO Keating is about to step down. Later newspaper
reports claim that Keating has been given one year to get Allfirst back on track
Case studies Allied Irish Bank (lessons learned)
Allied Irish Bank lacked clear reporting lines and structure

There was inadequate supervision of employees and failure to control the business that an
overseas office was engaged in
Proprietary trading is very risky and it is not just a question of market risk. A relatively small
outfit without access to the information, expertise and economies of scale of much larger financial
institutions may find it difficult to manage and control a proprietary trading business effectively.
The potential operational risks may outweigh the potential market returns, perhaps greatly
Risk management architecture is crucial The Ludwig report concluded that risk management
structure and practices within Allfirst's currency trading operations were seriously flawed. The
relationship between parent company and overseas units needs to be clear
In some areas, it was not clear who was accountable to whom, and the reporting lines within
Allfirst and between Allfirst and AIB were blurred
Strong and enforceable back-office controls are essential Unlike Barings Singapore unit, there
were independent back-office staff overseeing Rusnak's activities. But the Ludwig Report says that
Rusnak was able to persuade back-office staff to let normal procedures slip. Back-office staff must
be empowered to stand by their guns if they have concerns about trading activities

The Credit Crisis of 2007
The U.S. Housing Market

In about the year 2000, house prices started to rise much faster in the U.S.
Low interest rate between 2002 to 2005
House prices increased mainly for mortgage lending practices
During 2000 to 2006 Subprime mortgage lending increased significantly
Mortgage lenders relaxed the lending standards in 2000
Many families purchased house that were previously not considered to qualify for a mortgage
As a result the demand for real estate and prices rise.
The result of the relaxation of lending standards created a bubble in house prices.
In the 2nd half of 2006, house prices started to come down
The demand of the house decreased as the prices soared
Some borrowers with teaser rates could not afford their mortgages when the teaser rates ended
Supply of house increased
The amount owing on the mortgage was greater than the value of the house
Some of the borrowers choose to default
Further increase in the supply of the houses and further decrease in house prices

Securitization
The originators of mortgages did not in many cases keep the mortgages themselves. They sold
portfolios of mortgages to companies that created products for investors from them. This process
is called securitization.
The structure of Asset-Backed Securities
Assets
Principal: $100 Company ABS
million
Senior Tranche
Principal: $75 million
Return = 6%
Asset 1 Mezzanine Tranche

Asset 2 SPV Principal: $20 million
Asset 3, etc Return = 10%
Equity Tranche
Principal: $5 million
Return = 30%
Cash flow for ABS
Assets Cash
Flows
Senior Tranche
Mezzanine
Tranche
Equity Tranche

CDOs and COOs in Practice
Senior AAA 88%
Junior AAA 5%
AAA 81%
AA 3%
AA 11%
Subprime Mortgages
A 2%
A 4%
BBB 1%
NR 1%
Senior AAA 60%
Senior AAA 62%
Junior AAA 27%
Junior AAA 14%
AA 4%
AA 8%
BBB 3% A 3%
A 6%
BB/NR 1% BBB 3%
BBB 6%
NR 2%
NR 4%
The Crisis
The defaults on mortgages in the U.S. had a number of consequences

Financial Institutions & investors who bought the tranches of ABSs and ABS CDOs lost money
The tranches were downgraded by rating agencies in 2007
Credit spread increased sharply
Interbank lending rates increased sharply
The tranches of ABS became very illiquid
Investors realized the complexity of tranche products during that period
Banks such as Citigroup, UBS, and Merrill Lynch suffered huge losses
Government bailed out many financial institutions
Lehman Brothers was allowed to fail
Unemployment increased

What went wrong?
Mortgage lenders, the investors in tranches of ABSs from residential mortgages, and the companies
that sold protection on the tranches assumed that the good times would last forever.
Mortgage originators used relaxed lending standards

Products were developed to enable mortgage originators to profitably transfer risk to investors
Rating agencies moved from their traditional business of rating bonds
Regulatory Arbitrage
Agency costs
Lessons for Risk Managers
Watching for situations where there is irrational exuberance

Correlations always increase in stressed markets
Recovery rates decline when default rates increase
Incentives of traders and other personnel should be linked to the interest of the organisation
Investors should not only rely on ratings
Risks in the tranches that are not considered by rating agencies
Transparency is important in financial markets
There is nothing to be gained from further securitization

Risk Management Failures:
What are they and When do they Happen?
Role of risk management
Assess the risks faced by the firm

Communicate these risks to those who make risk-taking decisions for the firm
Manage and monitor those risks to make sure that the firm only bears the risks its management
and board of directors want it to bear
In general, a firm will specify a risk measure that it focuses on together with additional risk metrics
(Example for risk metric is VaR)
When that risk measure exceeds the firms tolerance for risk, risk is reduced
Alternatively, when the risk measure is too low for the firms risk tolerance, the firm increases its
risk

Large financial loss is not necessarily a failure of risk
The managers took risks they should not have, but that is not a risk management issue as long as
the risks were properly understood
Rather, it is an issue of assessing the costs of losses versus the gains from making large profits.
Example: Failure of LTCM
The decision depends on the risk appetite of an institution

Defining the risk appetite is a decision for the board and top management
With good risk management, large losses can occur when those making the risk-taking decisions
conclude that taking large, well understood risks creates value for their organization
Create value by taking calculated risk
Mistakes in risk management
Mistakes that can be made in measuring risk:

Known risks can be mis-measured
Some risks can be ignored
Failing to identify all the risks
A failure in communicating risk to management is a risk management failure
When risks are known, statistical techniques are generally brought to bear to estimate the
distribution of risks
Such approaches work well when there is a lot of data and when it is reasonable to believe that
the returns will have the same statistical distribution in the future as they had in the past
The unknown risks that represent risk management failures are risks that, had the firms managers
known about them, their actions would have been different

Types of risk management failures
1. Risk metrics failure. Example: MG and LTCM

2. Incorrect measurement of known risks. Example: MRM and LTCM
3. Ineffective risk monitoring. Example: Barrings and Sumitomo
4. Ineffective risk communication
5. Ignorance of significant known risks. Example: MG and LTCM
6. Unknown risk
Role of risk metrics
Risk Metrics like VaR, help measure and quantify risk and help managers to take decisions in line
with the targeted risk measures
Problems with VaR:
Frequency of Reporting: Daily losses below VaR, but accumulating on an annual basis
Does not capture extremely large losses which have a very low probability of occurrence
VaR assumes distribution of losses is not correlated. Correlation among various variables increases in times
of crisis

dz
www.edupristine.com
Foundation of Risk Management - II
Agenda
The Standard Capital Asset Pricing Model
Applying CAPM to Performance Measurement: Single Index Performance Measurement Indicators
Arbitrage Pricing Theory & Multifactor Models of Risk and Return
Information Risk and Data Quality Management
Principles for Effective Risk Data Aggregation and Risk Reporting
GARP Code of Conduct
The Standard Capital Asset Pricing Model

Assumptions of CAPM
All assets are infinitely divisible, traded and are collectively held and there always exists a
risk-free asset
Borrower and lenders are easy to find and they do so at risk-free rate (no counterparty risk)
Investors consider only mean and variance to choose a portfolio and Risk-Return distribution
follows normal distribution
All the investors have similar investment period.
Markets are perfect with no tax, transaction cost etc and market participants are price takers, i.e.,
they cannot influence prices
There is no information asymmetry in the market
Unlimited short selling is allowed
All the assets are marketable including human capital
Security Market Line: Basics

Diversifiable Risk and Systematic Risk
Investors are compensated only for bearing Systematic Risk
Market risk premium: Additional return over the risk-free rate, required to compensate investors
for assuming an average amount of risk. Its size depends on the perceived risk of the stock market
and investors degree of risk aversion
A measure of market risk: Beta

The tendency of a stock to move up or down with the market is reflected in its beta coefficient
Indicates how risky a stock is if the stock is held in a well-diversified portfolio

Capital Asset Pricing Model (CAPM)
As per CAPM, stocks required rate of return = risk-
Rs R f R m R f
Rm- Rf: Risk Premium

YZ
covR i , R m
i
Var R m
Security Market Line: Application
Return
SML
Rf
Risk- less rate of return
Beta
Linear relation of risk and return plotted on a graph is called Security Market Line (SML)
SML determines - if an asset is expected to offer commensurate return w.r.t. potential risk
Single stock return is plotted versus its security's risk
If it is above SML, stock is undervalued because market expects higher return
If it is below SML, stock is overvalued because market expects lesser return

Applying CAPM to Performance Measurement:
Single Index Performance Measurement Indicators
Beta
Sensitivity of the return of the asset to the market return is known as Beta
Beta is calculated as follows:-
covR i , R m
i
Var R m
Portfolio Beta
Beta can also be calculated for portfolio
Portfolio Beta is the weighted average of the betas of individual assets in the portfolio

Beta
Sharpe ratio:
R p Rf
Sharpe ratio p
Rp = portfolio return, Rf = risk free return
The higher the Sharpe measure, the better the portfolio
Treynor ratio:
R p Rf
Treynor ratio Beta
Rp = portfolio return, Rf = risk free return
The higher the Treynor measure, the better the portfolio
However, this measure should be used only for well-diversified portfolio
Jensons alpha:
Jensons alpha R p R c
Rp = portfolio return, Rc = return predicted by CAPM
Positive alpha (portfolio with positive excess return) is always preferred over negative alpha
Measures of performance
Tracking Error
V R p R b
Information Ratio
E ( R p ) E ( Rb )
V R p R b
Sortino Ratio
E ( R p ) MAR
1 T

Nt0
R p MAR
2
Where MAR is the Minimum Acceptable Return

Class exercise (Question): Treynor ratio
For a given portfolio, the expected return is 10% with a standard deviation of 15%. The beta of the
portfolio is 0.75. The expected return of the market is 11% with a standard deviation of 18%.
The risk-free rate is 4%. The portfolio's Treynor measure is:
A. 0.060
B. 0.012
C. 0.040
D. 0.080
Treynor ratio: Solution
D.
TR
0.10 0.04 0.08
0.75

Measures of performance
Tracking Error (TE): TE V E P
(Std. dev. of portfolios excess return over Benchmark index)
Where Ep = RP RB
RP = portfolio return, RB = benchmark return
Lower the tracking error lesser the risk differential between portfolio and the benchmark index
Information Ratio (IR):
Measure of risk-adjusted return for a portfolio, defined as expected active return per unit of tracking error
IR
R p Rb
TE
Higher IR indicates higher active return of portfolio at a given risk level
Sortino Ratio (SR): SR

R p MAR
SSD
SSD 1/t R p MAR 2 ,

MAR is Minimum Accepted Return. SSD is standard deviation of returns below MAR. (Or) SSD is the Semi
Standard Deviation from MAR where Rp<MAR
Higher the Sortino Ratio, lower is the risk of large losses
Class exercise (Question): Sortino ratio
For the past four years, the returns on a portfolio were 6%, 9%, 4%, and 12%. The corresponding
returns of the benchmark were 7%, 10%, 4%, and 10%. The minimum acceptable return is 7%.
The portfolio's Sortino ratio is:
A. 0.4743
B. 0.2143
C. 0.5303
D. 0.6700

Sortino ratio: Solution
A.
Average Return 6 9 4 12 7.75%
4
SR
0.0775 0.07
SSD
SSD
1 0 3 0
2 2
10
1.58%
4 4
0.75%
SR 0.4743
1.58%
Class exercise (Question): Sortino ratio
An analyst has compiled the following information on a portfolio:

Sortino Ratio: 0.82
Beta: 1.15
Portfolio return: 12.2%
Standard deviation: 16.4%
Benchmark return: 11.9%
Risk-free rate: 4.75%
Calculate the semi-standard deviation of the portfolio
A. 0.4000%
B. 0.3658%
C. 0.1338%
D. 0.9080%

Sortino ratio: Solution
B.
0.82
0.122 0.119
SSD
Semi Standard Deviation = SSD = 0.3658%
Semi Variance = SSD2 = 0.1338%
Class exercise (Question): Information Ratio
A portfolio has an average return over the last year of 13.2%. Its benchmark has provided an
average return over the same period of 12.3%. The portfolios standard deviation is 15.3%, its beta
is 1.15, its tracking error volatility is 6.5% and its semi-standard deviation is 9.4%. Lastly the risk
free rate is 4.5%. Calculate the portfolios Information Ratio (IR)
A. 0.569
B. 0.076
C. 0.138
D. 0.096

Information ratio: Solution
C.
IR
0.132 0.123 0.138
0.065
Class exercise (Question): Sharpe ratio / IR
Average Volatility Performance

Risk free 3% 0%
Portfolio -6% 25% Calculate SR of Portfolio
Benchmark -10% 20% Calculate SR of Benchmark
Tracking error 4% 8% Calculate IR of Portfolio
Correlation PB = 0.961

Sharpe ratio / IR: Solution
Sharpe ratio for Portfolio
SR
> 6% 3%@0.36
25%
Absolute performance is poor
Sharpe ratio for benchmark
SR
> 10% 3% @ 0.65
20%
Absolute performance is even poorer than portfolio
Information ratio
Absolute performance is even poorer than portfolio
IR
> 6% 10%@ 0.50
8%
Sharpe ratio / IR: Summary
Average Volatility Performance

Cash 3% 0%
Portfolio -6% 25% SR = -0.36
Benchmark -10% 20% SR = -0.65
Deviation 4% 8% IR = 0.50
IR Positive Signifies better relative performance

Recent risk adjusted performance measures
There are three recent risk adjusted performance measures which are covered in FRM
Part 1 curriculum:-
Morning Star Rating System
VaR Based Risk Adjusted Measures
Management Related Risk Adjusted Measures
Morning star rating system
Risk adjusted performance of the investment funds is ranked by the firm Morning Star
This ranking is done within a peer group
Top 10% of the funds in the peer group get 5 star rating whereas bottom 10% of the funds get 1
start rating
Middle 35% of the funds are assigned 3 star rating

VaR based risk adjusted measures
VaR states the maximum loss that the asset can sustain given a stated confidence level and
time period
By modifying the Sharpe Ratio, risk-return performance based on VaR can be analyzed
R p Rf
VaR
/Vp
p
Where
VaRP = Portfolio VaR
VP = Initial Portfolio Value
VaR can also be used for investment decision making process. If the fund manager wants to add a
new security in the portfolio, then the new VaR of the portfolio will be compared with the VaR of
the portfolio before adding the security. Change in the VaR of the portfolio from the addition of
the new security is known as Incremental VaR.
Management related risk adjusted measures
In management related risk adjusted measures, capital market line is modified to capture the
appropriate investment style of the fund
This measure is also known as style risk adjusted performance measure (SRAP)
There is another management related risk adjusted measure known as correlation adjusted
performance measures
This measure has been developed to compare managers within a peer group
This measure helps the investors to allocate their funds optimally across fund managers that
maximizes the desired return and minimizes risk

Arbitrage Pricing Theory
&
Multifactor Models of Risk and Return
Factor model a deviation from Index Model

A simple Index Model (analogous to CAPM and SML)
Stock variability (risk) constitutes
Market Risk (systematic) proxy of risk from macroeconomic events
Firm specific or Idiosyncratic risk (unsystematic) diversifiable in large portfolios
Stock return is dependent on market portfolio return
Here, Market portfolio return (Rm) sums up the impact of all macro factors on the stock return
What if, rather than using a single market proxy, we consider each macro factor as a separate
source of risk?
Factor Model
The systematic factor summarized by market return can arise from a number of separate factors such as
uncertainty of business cycle, interest rate, inflation rate etc.
Can be used to describe and quantify each factor that has an impact on the security return. And building
model of equilibrium security pricing
Useful in risk management. Factor model can be used to measure exposure of the stock to various risk
factors.

Single Factor Model
ri E (ri ) E i F ei
E (ri ) E i ( Factual Fexp ected ) e i
ri = Return on the stock i

E(ri) = Expected return on the stock I
F = deviation of the actual value of factor from its expected value
i = sensitivity of the firm i to the factor, factor sensitivity/factor loading
ei = firm specific disturbance
Note:
i. F is a random variable with an expected value of zero, i.e. if there are no macro surprises, (F = 0), than

ii. Ei is uncorrelated with F
Multifactor Model
ri E (ri ) E i1 F1 E i2 F2 ei
An example
ri E ( ri ) E iGDP GDP E iIR IR e i
here
E iGDP is the factor beta for unexpected changes in GDP
E iIR is the factor beta for Interest Rate

Multifactor SML
In multifactor model, how do we arrive at E(r)? Using a SML (CAPM)
For the single factor model, the single factor SML states
E(r) = rf (rm rf) CAPM model
For the two factor model, the multi factor SML states, that the expected return on the security is
sum of
i. Risk free rate of return
ii. Sensitivity to GDP risk time the risk premium for bearing GDP risk
iii. Sensitivity to interest rate risk times the risk premium for bearing interest rate risk
E(r) = rf + GDP(RPGDP) + IR(RPIR)

Here: GDP and IR denotes sensitivity of security returns to unexpected changes in GDP growth
and interest rates respectively
RPGDP is the risk premium of GDP associated with one unit of GDP exposure

Assumptions of APT
1. Security returns can be described by a factor model
2. There are sufficient number of securities to form a well diversified portfolio
3. Well functioning security markets do not have persistent arbitrage opportunities
Arbitrage
An arbitrage opportunity arises when an investor can earn riskless profit by constructing a zero
net investment portfolio. Example shares of stocks sold for different prices in different
exchanges
The law of one price states that if two assets are equivalent in all economically relevant respects,
than they should have the same market price. As per the law of one price, when arbitrage
opportunities arise, large amount of trade will be generated (buying cheap and selling expensive),
that puts pressure on security prices.
The arbitrage trades will increase the price, where it is low and decrease it where it is high, till
there is no more arbitrage opportunity available
Well diversified portfolios

If a portfolio is well diversified, its firm-specific risk becomes negligible, and only the systematic
risk remains
Non systematic risk across firms cancel out in a well diversified portfolio

Betas and Expected returns
Well diversified portfolios (with no unsystematic risks) having equal betas must have equal
expected returns, or arbitrage opportunities will exist
Well diversified portfolios having different betas must have expected returns proportional to their
corresponding betas. i.e. the risk premiums are proportional to portfolio betas.
To rule out arbitrage opportunity, the expected return on all well diversified portfolios must lie on
a straight line drawn from the risk free asset. (similar to SML)
(graph: expected return of portfolio8 vs beta of portfolio)
7
6
5 Market portfolio,
4 Rm = 6%, =1
3
Risk free asset,
2
Rf = 4%, =0
1
0
-1.5 -1 -0.50 0.5 1 1.5
The market portfolio will also lie on the line, corresponding to 2beta = 1 equivalent to one
factor SML
If a portfolio lies above or below the SML, an arbitrage opportunity exists
As per APT, the prices of portfolios are restored to equilibrium due to arbitrage activities
Individual Assets APT

For well diversified portfolios, the expected return is proportional to portfolio beta. Similarly, for
individual stocks, the expected return is proportional to its individual beta.
APT vs CAPM
APT and CAPM serve similar purpose, i.e. they provide the benchmark for determining the
expected rate of returns for security valuation
APT does not have the restrictive assumption of CAPM i.e. as per CAPM, the benchmark
portfolio in the SML should only be the (unobservable) market portfolio.
In APT, any portfolio on the SML line can serve as the benchmark portfolio. For example, even if
the index portfolio is not an exact proxy of market portfolio, it cannot be used for SML (by CAPM).
But as per APT, this index portfolio can be used for SML.
Multifactor APT
Multifactor APT generalizes the single factor model to accommodate several sources of systematic
risk.
The multifactor SML predicts that exposure to each risk factor adds to securitys total risk
premium

Shortcoming of APT
No guidance regarding the determination of relevant risk factors.
For determining a reasonable list of factors choose a limited number of systematic factors which
seem likely to be the important risk factors.
Fama French Three-Factor model

rit = i + iM(RPMt) + iSMB(SMBt) + iHML(HMLt) + eit
SMB small minus big; i.e. return of portfolio of small stock minus return on portfolio fo large
stocks
HML High minus low; i.e. return of portfolio of stock with high BV ratio, minus return of
portfolio of stocks with low BV ratio
Information Risk and Data Quality

How Poor Data can Affect Business
Poor Data can impact your financials: increase operating costs, increase penalties, delays in
cash flows
Loss of Confidence of management reporting, low confidence in forecasting
Customer Satisfaction goes down
Impacts Credit Assessment
Impacts Productivity negatively
Compliance is jeopardized
Common Issues Resulting in Data Errors
Data Entry Errors

Duplicate Records
Missing Data
Nonstandard Formats
Inconsistent Data
Failed identity management processes
Complex data transformations
Undocumented, incorrect or misleading metadata

Key Dimensions Characterizing Acceptable Data
Accuracy
Completeness
Consistency
Reasonableness
Currency
Uniqueness
Other Dimensions of Data Quality
Operational Data Governance
Operational Data Governance is the manifestation of the processes and protocols necessary to
ensure that an acceptable level of confidence in the data effectively satisfies the organizations
business needs
It governs the roles, responsibilities associated with the quality of data
Data Quality and Data Validation are two different things:

Data Validation
Process reviews and measures conformance of data with a set of defined business rules
Data Inspection
Reduce the number of errors to a reasonable level
Identify errors and have a systematic approach to correct them
Set time bounds to mitigate the cause of the error

Data Metrics
Base-Level Metrics Measures against defined dimensions of data

Complex Metrics using functions of sums or weights
Viewpoints where data metrics can be reported:

By Issue
By Business Process
By Business Impact
Scoreboards
A quality scoreboard:
Hierarchical rollup of metrics should reflect in the scorecards
Definitions
Appropriate level of presentation depending on the intended audience

Principles for Effective Risk Data Aggregation and Risk Reporting
What is Risk Data Aggregation

Defining, gathering, and processing risk data according to the banks risk reporting requirements
enable the bank to measure its performance against its risk appetite.

Potential Benefits
Enhance the infrastructure for reporting key information used by senior management to identity,
monitor and manage risks
Improve the decision-making process throughout the banking organisation
Enhance the management of information across legal entities
Reduce the probability and severity of losses resulting from risk management weaknesses
Improve the speed at which information is available and hence decision can be made
Improve the organisations quality of strategic planning and the ability to manage the risk of new
products and services
Principles
The principles are initially addressed to SIBs and at both the banking group and on a solo basis.
The principles cover four closely related topics
Overarching Risk Data Risk Reporting Supervisory

Governance and Aggregation Practices review, tools and
Infrastructure capabilities Accuracy cooperation
Governance Accuracy & Comprehensiveness Review
Data architecture Integrity Clarity and Remedial actions
and IT Completeness Usefulness and supervisory
infrastructure Timeliness Frequency measures
Adaptability Distribution Home/host
cooperation

Code of Conduct
Principals Professional Standards
1. Fundamental
1. Professional Integrity
Responsibilities
and Ethical Conduct
2. Adherence to generally
2. Conflicts of Interest
accepted practices of
3. Confidentiality
risk management

Violations of Code of Conduct
Violations of this code:

Temporary or Permanent Removal from GARP Membership roles
Temporary or Permanent Removal of the right to use FRM
dz
www.edupristine.com

FRM-I - Book Quant Analysis

Uploaded by

Copyright:

Available Formats

You might also like

FRM-I - Book Quant Analysis

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

FRM-I - Book Quant Analysis

Uploaded by

Copyright:

Available Formats

Index

Sr No. Topics Page No.

EduPristine FRM-I (2016) 0

EduPristine FRM-I (2016) 2

EduPristine FRM-I (2016) 3

Number of ways of selecting r objects out of n objects

EduPristine FRM-I (2016) 4

EduPristine FRM-I (2016) 5

EduPristine FRM-I (2016) 6

EduPristine FRM-I (2016) 7

EduPristine FRM-I (2016) 8

A probability experiment involves performing a number of trials to enable us to measure the

EduPristine FRM-I (2016) 9

EduPristine FRM-I (2016) 10

Probability Function on discrete random variable: P(x)

EduPristine FRM-I (2016) 11

EduPristine FRM-I (2016) 12

Cumulative Density Function: F(x)

EduPristine FRM-I (2016) 13

PDF, where x is the price of the bond

EduPristine FRM-I (2016) 16

Some definitions and properties of Probability

EduPristine FRM-I (2016) 17

EduPristine FRM-I (2016) 18

Conditional & Joint Probability

EduPristine FRM-I (2016) 19

Good Bad Natural

IMB Up Down Up Down Up Down

10% 10% 10% 10% 10% 10%

EduPristine FRM-I (2016) 20

EduPristine FRM-I (2016) 21

EduPristine FRM-I (2016) 22

Some Properties of Probability

P(A)  P(B) - P(A B)

EduPristine FRM-I (2016) 23

EduPristine FRM-I (2016) 24

EduPristine FRM-I (2016) 25

EduPristine FRM-I (2016) 26

EduPristine FRM-I (2016) 27

EduPristine FRM-I (2016) 28

EduPristine FRM-I (2016) 29

EduPristine FRM-I (2016) 30

EduPristine FRM-I (2016) 31

Descriptive Statistics is used to summarize important characteristics of large data set.

EduPristine FRM-I (2016) 32

EduPristine FRM-I (2016) 33

EduPristine FRM-I (2016) 34

Mean of Random variables (Discrete & Continuous)

When probabilities of each possible outcome is same, than pi = 1/n, and

EduPristine FRM-I (2016) 35

Arrange the values in ascending order

EduPristine FRM-I (2016) 36

EduPristine FRM-I (2016) 37

Given the following table, calculate the expected value of x.

EduPristine FRM-I (2016) 38

EduPristine FRM-I (2016) 39

EduPristine FRM-I (2016) 40

Variance & Standard deviation

P(A) P(B) - P(A B)

Var (aX b) a 2Var ( X ) Variance of a constant = 0

Var (aX bY ) a 2Var ( X ) b 2Var (Y ) 2abCov ( X , Y )

The expected number of successes y