2.

Continuous Random Variables

2.1 Probability and Random Variables

2.2 Continuous Random Variables, the p.d.f
2.3 The Normal Family of Distributions
2.4 The C.d.F
2.5 I.I.D. Random Variables
2.6 I.I.D. Draws from the Normal Distribution
2.7 The Histogram and I.I.D. Draws
2.8 The Normal Distribution and Data
2.9 Standardization

© Imperial College Business School

 In this section of the course we learn about
random variables and probability.

This is a very important

topic that gets used
in a variety of situations.

In order to think about

Many real world problems
we have to face the fact
that we are uncertain about
some important aspects of the situation.
© Imperial College Business School
 Examples:
0.1

What do you

canada
0.0
think the return
will be next month?
-0.1
Index 20 40 60 80 100

0.08

What do you canSmall

0.04

think the return

will be next month? 0.00

Index 5 10 15 20

© Imperial College Business School

 1.0

You make 100 parts.

defs100
0.5

1 means defective.
0 means good.
25% are defective. 0.0

Index 10 20 30 40 50 60 70 80 90 100

You make 20 parts.

1 means defective.
0 means good.
25% are defective.

What is the “true defect rate” for each process?

Why are you “less sure” in the second case?
© Imperial College Business School
 2.1 Probability and Random Variables

Example: coin toss

• We toss a coin. It comes up either a head or a tail

• Report a 1 if it comes up a head and a 0 if it is a tail

• So HHTHTH is coded up as 1,1,0,1,0,1

© Imperial College Business School

 1.0

coin
0.5

Suppose we
toss the coin 25 times.
0.0

0 5 10 15 20 25
We get a sequence time

of 25 1’s and 0’s.

1.0

coin
0.5
same
data
0.0

Index 5 10 15 20 25

© Imperial College Business School

 10

Frequency
5

We had 13 heads
and 12 tails. 0

0 1
coin

1 + 1 + 0 + 1 + L + 0 13(1) + 12(0)
= = .52
25 25

The average tells us the percentage of tosses that

resulted in a 1 (a head). 52% of the tosses resulted
in a 1.

© Imperial College Business School

 What will happen the next time we toss the coin?

Let X denote the outcome.

X will be either a 1 or a 0.

X is a numerical quantity about which we are uncertain.

X is a Random Variable. We do not know what X will
be, but we do know that it will be either 1 or 0 with
certain probabilities.

What are the probabilities?

© Imperial College Business School
They are: Pr( X = 1) = .5 Pr( X = 0) = .5
The probability of a 1 is .5
The probability of a 0 is .5.

What does that mean?

The two possible outcomes are equally likely (by the very
nature of a coin).
Over the long run, if we tossed the coin over and over again,
we expect a 1 (or, equivalently, a zero) 50% of the time.

Probability is the long-run frequency: how often it happens

© Imperial College Business School
That is, if we toss the coin n times with n
really big and

n1 is the number of 1’s

n0 is the number of 0’s

n1 n0
then, ≈ .5 ≈ .5
n n

© Imperial College Business School

 7

Of course, if we toss 5

Frequency
4
a coin 10 times we 3

do not necessarily 2

expect to get exactly 1

0
5 heads and 5 tails. 0 1
toss10

500
If we toss it 1000
400
times we expect
Frequency
300
the proportion to
200
work out in the
100
long run.
0

0 1
toss10thou
© Imperial College Business School
 50

40

Percent
30
10,000 tosses
20

10

0 1
C5

• For “all” the tosses we expect to get 50% heads

• For some, we could get something different
• The closer to “all” we get, the more likely it is
that the observed fraction will be close to .5
© Imperial College Business School
Fbefore talking of continuous random variables, here it
is the most “famous” discrete random variable:
Bernoulli distribution

(the coin example is a particular case of it!)

The situation where something happens or not and we
want to talk about the probability of it happening is our
most basic scenario.
To describe this situation we use a random variable which
is 1 if something happens and 0 otherwise and probability
(“it happens”) = p. Such a random variable is said to
have the Bernoulli distribution.

Notation: Y~Bernoulli(p) means P(Y=1)=p, P(Y=0)=1-p

Example: toss a coin. X=1 if head, 0 else.
Then, X~Bernoulli(.5)
© Imperial College Business School
The random variable X has the Bernoulli distribution
with parameter p (between 0 and 1) if

p X (1) = p p X (0) = 1 − p

In general, we think of X=1 as the thing happens

and X=0 as the thing does not happen.

© Imperial College Business School

 Suppose X 1, X 2 ,L X n are i.i.d Bernoulli(p).

Then,Y = X 1 + X 2 +L+ X n

has the Binomial distribution with parameters

n and p.
This is just a symbol for the Binomial

We write: Y ~ B(n, p)
distribution. Just like Be(p) is a symbol
for the Bernoulli distribution. The
meaning is all that matters.

Xi tells you whether it happened on the ith trial.

Y is the total number of times it happened out of n trials.

We will get back on the mean and variance of Y.

© Imperial College Business School
 Something to think about

• The word random variable refers to the outcome

before it happens
• A random variable describes what we think will
happen
• After we have an outcome (say, after we toss a
coin), the obtained value is sometimes called a
draw from the common distribution (it is a data
point or an observation from the sample)

© Imperial College Business School

 Example (Investing in an asset)

Suppose you are considering investing in an asset.

Let R denote the return next month. We think of R as a random
variable. We do not know what the return will be (it is random) but
we assume we know what the possible outcomes and probabilities
are. In other words, we are truly modeling a future event.

R: r .05 .1 .15 The probability

that the return will be
p(r) .1 .5 .4 greater than .05 is .9.

Question: Does it make sense to model (future) returns as

discrete random variables? Wouldn’t it be better to model them
as random variables that can take on a continuum of outcomes?
© Imperial College Business School
2.2 Continuous Random Variables, the p.d.f

r .05 .1 .15
Consider again the
return example: p(r) .1 .5 .4

This is unrealistic as it is unlikely that you know

that the return will be only 3 possible values.

© Imperial College Business School

• Issue: sometimes we would like our random variable
to be able to take on any value in an interval
• In this case we cannot simply list all the possible
values and give each one a probability
• We need a new way to specify probabilities
• We still want the random variable X to stand for a
numerical quantity we are uncertain about, but we
need a way to describe what it is likely to be

© Imperial College Business School

Instead of specifying the probabilities for specific
values we specify probabilities for intervals.
Old (discrete): Pr(X=4) = .7
New (continuous): Pr(X is in (-4,8)) = .7
In general, we specify Pr(X in (a,b)) for any
values a and b with a< b.
An easy way to do this is with the probability
density function (p.d.f).
© Imperial College Business School
 Again, let x denote a possible value
of the random variable X.

The p.d.f. (probability density function) denoted by f(x) is

a function of x such that the probability of any interval [a,b] is
the area under the graph of the function between a and b.

0.4

0.3
f(x)

0.2

0.1

0.0

-3 -2 -1 0 1 2 3
x
© Imperial College Business School
 0.4
area is
0.3
.477
f(x)

0.2

0.1

0.0

-3 -2 -1 0 1 2 3
x

For the random variable X the probability that it is in

the interval [0,2] is .477.
(47.7 percent of the time X will fall in this interval).
© Imperial College Business School
 Here is another p.d.f:

Note The area under the entire curve must be 1

(Why?)
0.25

0.20

f(c) 0.15

0.10

0.05

0.00
0 5 10
c

Most of the probability is concentrated in 0 to 3,

but you could get a value much bigger. This kind
of distribution is called skewed to the left.

© Imperial College Business School

For a continuous random variable X,
the probability of the interval
(a,b)
is the area under the probability density
function from a to b.

© Imperial College Business School

 Question:

Let U be the random variable which can

take on any value between –1 and 1 and
each value is “equally likely”.

What does the p.d.f of U look like ?

© Imperial College Business School

Question

Let U1= 2U. What does the p.d.f of X look like ?

Question

Let U2= .5U. What does the p.d.f of X look like ?

© Imperial College Business School

Question

Let U3= 5+2U. What does the p.d.f of X look like?

Question

Let U4= 5+.5U. What does the p.d.f of X look like?

© Imperial College Business School

 U2 U4

1.0
0.8

U
0.6
pdf
0.4
0.2
0.0

-2 0 2 4 6 8

U3
U1
© Imperial College Business School
2.3 The Normal Family of Distributions

The random variable 0.4

having this p.d.f is

0.3
very special.

f(x)
0.2

This distribution is
called the standard 0.1

normal distribution. 0.0

-3 -2 -1 0 1 2 3
x

Pr(-1<Z<1) = .68
If Z has this distribution,
then:
Pr(-1.96<Z<1.96)=.95

© Imperial College Business School

Some properties of the standard normal random variable:

P ( 0 < Z < 1) =.34

P ( − 1 < Z < 1) =.68
P( − 2 < Z < 2) = .954

P ( − 1.96 < Z < 1.96 ) = .95

P ( − 3 < Z < 3 ) = .9974
NB. In these notes I will usually act as if 1.96 = 2.
© Imperial College Business School
 The Normal Family of Distributions

We are going to use the normal distribution to describe

our uncertainty about things in the real world.
The standard normal distribution is not too exciting as it
is centered around 0, with probability .95 of being in +/-2.
We can create a family of interesting probability
distributions from the standard normal by
1. spreading it out or tightening it up
2. moving it around
© Imperial College Business School
 Z is our benchmark: the standard normal random variable

0.8
X2 = .5Z X4 = 5+.5Z

0.6
0.4

Z
X1 = 2Z
X3 = 5+2Z
0.2
0.0

-5 0 5 10

0.95 = Pr(-2<Z<2) =
Pr(1<X3<9) = Pr(5+2(-2)<X3<5+2(2)) =
Pr(-4<X1<4) =
Pr(4<X4<6) = Pr(5+.5(-2)<X4<5+.5(2))=
Pr(-1<X2<1) =
© Imperial College Business School
 σ: spread or tighten
In general, let X = µ + σZ µ: move
0.4

0.3
f(x)

0.2

0.1

0.0

µ − 2σ µ µ + 2σ x

95% chance of being in ( µ − 2σ, µ + 2σ )

68% chance of being in ( µ − σ, µ + σ )
© Imperial College Business School
The Normal Random Variable

We write, X ~ N ( µ, σ 2 )
Equivalently, X = µ + σZ
where Z is the standard normal Z ~ N (0,1)
Properties:

95% chance of being in ( µ − 2 σ, µ + 2 σ )

68% chance of being in ( µ − σ, µ + σ )

© Imperial College Business School

We have a family of distributions.
2
For each pair (µ, σ ) we get a normal distribution.
µ determines the “center” of the distribution
σ determines how spread out the probability
is around the center.

Note: σ>=0

µ : your “prediction”
σ : how “sure” are you, +/- 2σ.
© Imperial College Business School
Note: in the next section of the notes we will see that
µ is the “mean”, σ is the “standard deviation” and σ2
is the variance of the normal random variable.

I will use these names right away, but explain what

they mean later (in the next section of notes).

© Imperial College Business School

 Question
All of these normal distributions have µ=-3,0, or 3 and
σ=.5,1,or 2. Which is which?
0.8

0.7

0.6

0.5
C2

0.4

0.3

0.2

0.1

0.0

-6 -4 -2 0 2 4 6 8
x
© Imperial College Business School
Note: If we say X~N(5,4), then

µ=5
σ=2

That is, we use the mean and variance to

define a normal distribution (it is just notation,
do not get confused).
I wish it had been the mean and standard deviation.

© Imperial College Business School

 2.4 The c.d.f.

The c.d.f. (cumulative distribution function) is just

another way (besides the p.d.f.) to specify the
probability of intervals.

For a random variable X the c.d.f., which we denote

by F (we used f for the p.d.f.), is defined by

FX ( x) = P( X ≤ x)
© Imperial College Business School
 Example:

For the standard normal: F(0) = .5

F(-1) = .16
F(1) = .84

0.4

0.3
f(x)

0.2

0.1

0.0

-3 -2 -1 0 1 2 3
x
© Imperial College Business School
 The c.d.f. is handy for computing the probabilities
of intervals.
P(a < X ≤ b) = P( X ≤ b) − P( X ≤ a)
= FX (b ) − FX ( a )

Example 0.4

0.3

For Z (standard normal),

f(x)
0.2

we have: 0.1

0.0
P((-1,1)) = F(1)-F(-1) -3 -2 -1 0 1 2 3
x
= .84 - .16 = .68
.68
© Imperial College Business School
 1.0

F (b)

F(x)
0.5

F (a )
0.0

-3 -2 -1 0 1 2 3
a x b

The probability of an interval is the “jump” in the c.d.f.

over that interval.
Note: for x big enough F(x) must get close to 1.
for x small enough F(x) must get close to 0.
© Imperial College Business School
 Example:

Let R denote the return on our portfolio next month.

We do not know what R will be. Let us assume we
can describe what we think it will be by:

R~N(.01,.042)
10
p(r)

-0.1 0.0 0.1

r
© Imperial College Business School
What is the probability of a negative return?

Normal with mean = 0.0100000 and standard deviation =

0.0400000
>> mu=0.01 ;
>> sigma = 0.04;
>> NORMCDF(0,MU,SIGMA);
ans = 0.401293674317076

What is the probability of a return between 0 and .05?

>> NORMCDF(0.05,MU,SIGMA) - NORMCDF(0,MU,SIGMA)
ans = 0.440051071751467

© Imperial College Business School

 2.5 I.I.D. Random Variables
1.0

Remember our basic

coin tossing example.

coin
0.5

We think of the next

coin as
0.0

Index 5 10 15 20 25
P(X=1)=.5
P(X=0)=.5
How would you predict
(or X~Bernoulli(.5)) the next one?

© Imperial College Business School

P(X=1)=.5 Captures the idea that we expect
P(X=0)=.5 to get heads 50% of the time.

Important: there is more to coin tossing than that.

Suppose we “toss” a coin 1.0

20 times and get this

C5
0.5

How would you predict

the next one? In other
words, does the fact that 0.0

Index 5 10 15 20
we have a pattern matter?

© Imperial College Business School

With the coins we think that:

1) For each coin the probability of a head is .5

and
2) The outcome for each coin is unrelated
to that of the others.

© Imperial College Business School

 To express this formally, let us
just start with two coins.

We are about to toss two coins.

Let X1 be the random variable denoting

the outcome of the first coin.

Let X2 be the random variable denoting

the outcome of the second coin.

© Imperial College Business School

To model what we think about the coins we have:

1) Each coin has the same, or identical,

probability distribution

X1 ~ Bernoulli(.5) X 2 ~ Bernoulli(.5)
and

2) What happens to the first coin does not

affect the outcome of the second

X2 is independent of X1
© Imperial College Business School
We say that the two X’s are independent
and identically distributed, Xi~Bernoulli(.5)

They are

i.i.d.

independent identically distributed

(the first i) (the i.d. part)

© Imperial College Business School

i.i.d. random variables

In general, if we say that

X1, X2,LXn

are i.i.d., we mean that each one is independent

of all the others (i), and they all have the same
probability distribution (i.d.).

© Imperial College Business School

 Example

Again suppose you are making a part repeatedly

Xi = 1 if the part is bad

Xi = 0 if the part is good

To say that the X’s are i.i.d Bernoulli(.1), says a lot.

Two are the reasons:

© Imperial College Business School

 time series plot histogram
The numbers 1.0 900
800

we have 700
600

Frequency
already seen 500
400

look like 300

200

C1
0.5
i.i.d. draws 100
0

from the 0
C1
1

Bernoulli(.1)
distribution 0.0

Index 200 400 600 800 1000

1) It is a way to summarize what we have seen

2) It tells us what we expect to see in the future (if things

do not change of course)

We are using the idea of i.i.d random variables as a model

for something in the real world.
© Imperial College Business School
 Examples:

Does this data

look i.i.d.?

Does this data

look i.i.d.?

© Imperial College Business School

 Important

• The i.i.d. model is our most basic statistical model

(we will see more complicated models later on)

• In general, we look at the data and try to build

models that describe how the data was generated
(think about the example of the defective and good
parts) – the past

• Given the model (the data generating process) we

can predict what will happen next – the future

• This basic logic is really what this course is all about

© Imperial College Business School

2.6 I.I.D Draws from the Normal Distribution

• Remember how we used the idea of i.i.d draws

from the Bernoulli(.5) distribution to model coin
tosses and i.i.d draws from the Bernoulli(.1) to
model a manufacturing process that gives
faulty and good parts?
• Now, we want to use the normal distribution to
model data in the real world
• Surprisingly, often data looks like i.i.d draws
from a normal distribution

© Imperial College Business School

Note: We can have i.i.d draws from any distribution.

By writing

X1, X2 ,K Xn ~ N(µ, σ2 ) iid

we mean that each random variable X will be an

independent draw from the same normal distribution.
We have not formally defined independence for
continuous distributions, but our intuition is the same
as before!
What do i.i.d normal draws look like?
© Imperial College Business School
 MatLab can generate i.i.d draws from the normal distribution.

>>c1 = randn(100,1) (gives 100 i.i.d draws from

N(0,1) and puts them in c1)
3

1
C1

0
There is no pattern,
-1 they look “random”
-2

Index 10 20 30 40 50 60 70 80 90 100

© Imperial College Business School

 Same with lines drawn in at µ = 0 and +/- 2σ = +/- 2
3

1 In the long run,

95% will be in
C1

0 here (between
+2 and –2)
-1

-2

ndex 10 20 30 40 50 60 70 80 90 100

© Imperial College Business School

 Here are draws from a normal other than the standard one.

>>c1=randn(100,1);
We multiply each value in
>>c1 = 5+2*c1;
c1 by 2and add 5
>>plot(c1)

10
These are i.i.d draws from

5 N(5,4)
C1

How would you predict

the next one ?
0

Index 10 20 30 40 50 60 70 80 90 100

© Imperial College Business School

 2.7 The Histogram and I.I.D.
Draws
Here is the histogram of 1000 draws from the standard
Normal:
>>[N,BIN] = histc(randn(1000,1),(-4:0.5:4));
>> bar((-4:0.5:4),N);

The height of
each bar tells
us the percentage
of observations in
the interval.

We can see that

about 68% are
between -1 and 1.

61
 >>[N,BIN] = histc(randn(1000,1),(-4:0.5:4));

>> bar((-4:0.5:4),N/100)

It looks the same, but the vertical scale is different. Why?

© Imperial College Business School
 For a large number of
draws, the observed
percent in an interval
should be close to the
For the density the probability.
area is the
probability of the
interval.
0.4

For the histogram

0.3
the area is the Density
observed percent 0.2

in the interval.
0.1

In large samples
these are close. 0.0

-4 -3 -2 -1 0 1 2 3 4
z
© Imperial College Business School
 Example:

Frequency
histogram of 20
1
i.i.d N(0,1) draws
0

-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
C3

90
80
70

histogram of 1000 60
Frequency

50

i.i.d N(0,1) draws 40

30
20
10
0

-3 -2 -1 0 1 2 3
C4
© Imperial College Business School
 Example:

The histogram of a “large” number of i.i.d

draws from any distribution should look
like the p.d.f.
10,000 draws, uniform on (-1,1) 10,000 draws, N(0,1)

150 400

300
Frequency

100

Frequency 200

50
100

0 0

-1 0 1 -4 -3 -2 -1 0 1 2 3 4 5
C1 normal
 2.8 The Normal Distribution
and Data

We look at the return data for Canada.

We have monthly returns from Feb ‘88 to Dec ‘96.
0.1
canada

0.0 No apparent
pattern!

-0.1
Index 20 40 60 80 100

© Imperial College Business School

 35

30

25
Frequency

20 Normality
15 seems
10 reasonable!
5

-0.10 -0.07 -0.04 -0.01 0.02 0.05 0.08 0.11

canada

Conclude: The returns look like i.i.d normal draws!

© Imperial College Business School
If we think µ is about .01 and σ is about .04 (based on the
data), then our best guess at the next return is .01.

An interval which has a 95% chance of containing the next

return would be: .01 +/- .08

© Imperial College Business School

Note: we used i.i.d. Bernoulli draws to model coin
tosses and defects.

Now we are using the idea of i.i.d. normal draws

to model returns! We have a statistical model for
the real world.

This is a powerful statement about the real world.

© Imperial College Business School

Intuitively, the i.i.d normal model is meant to
describe numbers that

(i__)

have no pattern, are “random”

(_id)
but, over the long haul, the probabilities of the
distribution tell you how often certain values
(or sets of values) occur.
In our case, 95% of the values will be in µ +/- 2σ

© Imperial College Business School

 Example:

Of course, not all data looks normal.

60
Windsor return: 50

40

Frequency
30

20

10

Skewed to the 0

0 1000 2000 3000 4000 5000 6000 7000

right. Volume

© Imperial College Business School

Sometimes we can succinctly describe real
world data by saying that they “look” normal.

In this case we would really like to know the true

( µ, σ )
Similarly, if the data are i.i.d Bernoulli, we would like
to know p.

Given data, we will estimate the parameters (this

is what we call statistical inference – we will come
back to this issue – in fact your econometrics
course next term is all about this)
© Imperial College Business School
 Example:

An unusually zealous manager of a government office wanted to know

how long it takes a customer to be processed. For 100 consecutive
day, the manager sampled a customer at 10am, 11am, 12noon, 1pm,
2pm, 3pm and 4pm.
For each customer the time was obtained.
Here is the time series plot and histogram of the average time per day.

31
30
20
29
28

Frequency
aWait

27
26 10
25
24
23
0
22
Index 10 20 30 40 50 60 70 80 90 100 22.0 23.2 24.4 25.6 26.8 28.0 29.2 30.4 31.6
aWait

© Imperial College Business School

 Dow Jones
2500

Example

dji
1500

Do these look like i.i.d. draws? 500

Index 50 100 150 200

Lake Level Beer Production

12 20 8
19 5
67 567 7 56 6
11 56 5 8 6 7
4 4 7
18 34 3 8 8 3
45
10 7 8
17 3 4 3 8
3

beerprod
level

9 16 9 9 1 10 12 9
4 1 10
12 10 10 2 9
10
15 12 2 9 2
8 9
10
1 12
14 11 11 11 12
7 12 11
12 11
11
12
13 12
6
12
Index 10 20 30 40 50 60 70 80 90 Index 10 20 30 40 50 60 70

© Imperial College Business School

 2.9 Standardization
How unusual is it?

Sometimes something weird or unusual happens

and we want to quantify just how weird it is.
Typical examples are market crashes

0.1

Monthly returns 0.0

on a market index
C36

from Jan ’80 -0.1

to Oct ’87. -0.2

Index 10 20 30 40 50 60 70 80 90

© Imperial College Business School

 Question: how crazy is the crash?

Histogram of C37, with Normal Curve

The data looks
normal (from the 15

histogram). We can

Frequency
10

use a normal curve

(model) to describe 5

all the values except 0

-0.1 0.0 0.1

the last. C37

The curve (model) has µ=.0127 and σ=.0437.

We are kind of estimating the true
µ and σ based on the data. We
will be very precise in the future.

The crash month return was -.2176.

© Imperial College Business School
 N(.0127,.04372 )
9
8
7
6

The crash return was 5

p(r)
4

way out in left tail! 3

2
1
0

-0.3 -0.2 -0.1 0.0 0.1 0.2 0.3

r

We can do essentially the same thing by standardizing

the value.

We ask: if the value were from a standard normal,

what would it be?

© Imperial College Business School

We can think of our return values as:

r = .0127 + .0437z (µ + σ z)

So, the z value corresponding to a generic r value is:

r − µ r − .0127
z= =
σ .0437
The z values should look standard normal.

© Imperial College Business School

 So, how unusual is
the crash return?
>> k1 = (-.2176-.0127)/.0437
Its z value is -5.27!
k1 =
-5.270022883295193 It is like getting a
value of –5.27 from
the standard
20
normal.

No way!
Frequency

10

Here are the z

0
values for the
-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 previous months.
C50

© Imperial College Business School

Another way to say it is that the crash return is 5.3
standard deviations away from the mean.

For X~N(µ,σ2), the z value corresponding

to an x value is:
x −µ
z=
σ
z can be interpreted as the number of standard
deviations the x value is from the mean.

© Imperial College Business School

 Example:

How unusual is (the hockey player) Wayne Gretzky ?

(Recall that ESPN picked him 5th greatest athlete of the century).

This is the histogram of the total career points of the 42 players

judged by the Hockey News to be the greatest ever, not counting
goalies and Wayne Gretzky.

µ=1000 and σ = 450 look reasonable (based on the data).

Gretzky had 2855 points.

15
>> let k1 = (2855-1000)/450
10 k1 = 4.12222
Frequency

5 Grezky is like getting

4.1 from the standard
0
Normal. No way!
0 500 1000 1500 2000
C6
Normal Probabilities & Standardization
(useful for applications)

Suppose Normal with mean = 0 and

R~N(.01,.042).
standard deviation = 1.00000.
We evaluate P( X <= x).
What is the probability
of a return between 0 x=-0.25

and .05? >> NORMCDF(x,0,1)

Ans = 0.401293674317076
This is equivalent to Z x = 1
being between >> NORMCDF(x,0,1)
(0-.01)/.04= -.25 ans = 0.841344746068543
and
>> k1 = .8413-.4013
(.05-.01)/.04=1.
k1 = 0.440000000000000
© Imperial College Business School
For X~N(µ,σ2),

Pr(a < X < b) =

a−µ b−µ
Pr( <Z< ) Z ~ N(0,1)
σ σ

© Imperial College Business School