Additional Materials

Short Sales
An Overview
AD.6
Short Sales
An Overview
This handout is a reproduction of section 2.4.2 (pages 25-30) of

Sharpe, William F., and Gordon J. Alexander, 1990, Investments, Prentice-Hall, New Jersey.
Introduction
An old adage from Wall Street is to buy low, sell high. Most investors hope to do just that by
buying securities rst and selling them later.6 However, with a short sale this process is reversed.
The investor sells a security rst and buys it back later. In this case the old adage about investor
aspirations might be reworded as sell high, buy low.
Short sales are accomplished by borrowing stock certicates for use in the initial trade and
then repaying the loan with certicates obtained in a later trade. Note that the loan here involves
certicates, not dollars and cents (although it is true that the certicates at any point in time have
a certain monetary value). This means the borrower must repay the lender by returning certicates,
not dollars and cents (although it is true than an equivalent monetary value, determined on the
date the loan is repaid, can be remitted instead). It also means that there are no interest payments
to be made by the borrower.
Rules Governing Short Sales

Any order for a short sale must be identied as such. The Securities and Exchange Commission
has ruled that short sales may not be made when the market price for the security is falling, on
the assumption that the short seller would exacerbate the situation, cause a panic, and prot
therefrom an assumption inappropriate for an ecient market with astute, alert traders. The
precise rule is that a short sales must be made on an up-tick (for a price higher than that of the
previous trade) or on a zero-plus tick (for a price equal to that of the previous trade but higher
than that of the last trade at a dierent price).
Within ve business days after a short sale has been made, the short-sellers broker must borrow
and deliver the appropriate securities to the purchaser. The borrowed securities may come from
the inventory of securities owned by the brokerage rm itself or the inventory of another brokerage
rm. However, they are more likely to come from the inventory of securities held in street name
by the brokerage rm for investors that have margin accounts with the rm. The life of the loan
is indenite, meaning there is no time limit on it.7 If the lender wants to sell the securities, then
the short seller will not have to repay the loan if the brokerage rm can borrow shares elsewhere,
thereby transferring the loan from one source to another. However, if the brokerage rm cannot
nd a place to borrow the shares, then the short seller will have to repay the loan immediately.
6
After purchasing a security, an investor is said to have established a long position in the security.
The New York Stock Exchange, the American Stock Exchange, and NASDAQ publish monthly list of the short
interest in their stocks (short interest refers to the number of shares of a given company that have been sold short
where, as of a given date, the loan remains outstanding). To be on the NYSE or AMEX list, either the total short
interest must be equal to or greater than 100,000 shares or the change in the short interest from the previous month
must be equal to or greater than 50,000 shares. The respective gures for NASDAQ are 50,000 and 25,000.
7
Short Sales
XYZ
An Overview
Dividends, annual reports,

voting rights
AD.7
Forwards
everything
Mr. L ane
Allows stock
to be le nt
Mr. L ane
Brock, Inc .
Figure 6: Before the short sale.

M r. Jones
Pays
purc hase
pric e
XYZ
Notifie d that Mr. Jones

now owns stock
Re ceives stock
c ertificates
Brock, Inc .
Provides
initial
margin
M s. Smith
Figure 7: The short sale.

Interestingly, the identities of the borrower and lender are known only to the brokerage rm that
is, the lender does not know who the borrower is and the borrower does not know who the lender
is.
An Example
An example of a short sale is indicated in Figures 6-8. At the start of the day, Mr. Lane owns
100 shares of the XYZ Company, which are being held for him in a street name by Brock, Inc., his
broker. During this particular day, Ms. Smith places an order with her broker at Brock to short
sell 100 shares of XYZ (Mr. Lane believes that the price of XYZ will rise in the future, whereas
Ms. Smith believes it is going to fall). In this situation, Brock takes the 100 shares of XYZ that
they are holding in street name for Mr. Lane and sells them for Ms. Smith to some other investor,
in this case Mr. Jones. At this point XYZ will receive notice that the ownership of 100 shares of
its stock has changed hands, going from Brock (remember that Mr. Lane held his stock in a street
name) to Mr. Jones. At some later date, Ms. Smith will tell her broker at Brock to purchase 100
shares of XYZ (perhaps from a Ms. Poole) and use these shares to pay o her debt to Mr. Lane.
At this point, XYZ will receive another notice that the ownership of 100 shares has changed hands,
going from Ms. Poole to Brock, restoring Brock to their original position.
What happens when XYZ declares and subsequently pays a cash dividend to it stockholders?
Before the short sale, Brock would receive a check for cash dividends on 100 shares of stock. After
depositing this check in their own account at a bank, Brock would write a check for an identical
amount and give it to Mr. Lane. Thus, neither Brock or Mr. Lane has been worse o by having
his shares held in the street name. After the short sales, XYZ will see that the owner of those 100
shares is not Brock any more but is now Mr. Jones. Thus, XYZ will now mail the dividend check
to Mr. Jones, not Brock. However, Mr. Lane will still be expecting his dividend check from Brock.
Indeed, if there was a risk that he would not receive it, he would have agreed to have his securities
held in street name. Brock would like to mail him a check for the same amount of dividends that
Short Sales
An Overview
AD.8
M r. Jones
Divide nds, a nnual re ports,
voting rights
XYZ
Annua l Report
Brock, Inc .
Annual report
dividends
Mr. L ane
Provides cash
to ma ke up
for divide nd
M s. Smith
Figure 8: After the short sale.

Mr. Jones received from XYZ that is, for the amount of dividends Mr. Lane would have received
from XYZ had he held his stock in his own name. If Brock does this, they will be losing an amount
of cash equal to the amount of the dividends paid. In order to prevent themselves from experiencing
this loss, they make Ms. Smith, the short seller, give them a check for an equivalent amount!
Consider all the parties involved in the short sale now. Mr. Lane is content, since he has received
his dividend check from his broker. Brock is content, since he has received his dividend check from
his broker. Brock is content since their net cash outow is still zero, just as it was before the short
sale. Mr. Jones is content, since he received his dividend check directly from XYZ. What about Ms.
Smith? She should not be upset with having to reimburse Brock for the dividend check given by
them to Mr. Lane, since the price of XYZ s common stock can be expected to fall by an amount
roughly equal to the amount of cash dividend, thereby reducing the dollar value of her loan from
Brock by an equivalent amount.
What about annual reports and voting rights? Before the short sale, these were sent to Brock,
who then forwarded them to Mr. Lane. After the short sale, Brock no longer received them, so
what happened? Annual reports are easily procured by brokerage rms free of charge, so Brock
probably got copies of them from XYZ and mailed a copy to Mr. Lane. However, voting rights
are dierent. These are limited to the registered stockholders (in this case, Mr. Jones) and cannot
be replicated in the manner of cash dividends by Ms. Smith, the short seller. Thus, when voting
rights are issued, the brokerage rm (Brock, Inc.) will try to nd voting rights to give to Mr. Lane
(perhaps Brock owns shares or manages a portfolio that owns shares of XYZ and will give these
voting rights to Mr. Lane). Unless he is insistent, however, there is a chance he will not get his
voting rights once his shares have been borrowed and used in a short sale. In all other matters, he
will be treated just as if he were holding the shares of XYZ in his own name.
As previously mentioned, a short sale involves a loan. Thus, there is a risk that the borrower
(in the example, Ms. Smith) will not repay the loan. In this situation the broker would be left
without the 100 shares that the short seller, Ms. Smith, owes him or her. Either the brokerage
rm, Brock, is going to lose money or else the lender, Mr. Lane, is going to lose money. To prevent
this from happening, the cash proceeds from the short sale, paid by Mr. Jones are not given to the
short seller, Ms. Smith. Instead, they are held in her account with Brock until she repays her loan.
Unfortunately this will not assure the brokerage rm that the loan will be repaid.
Short Sales
An Overview
AD.9
In the example, assume the 100 shares of XYZ were sold at a price of $100 per share. In
this case, the proceeds from the short sale of $10,000 are held in Ms. Smiths account, but she is
prohibited from withdrawing it until the loan is repaid. Now imagine that at some date after the
short sale, XYZ stock rises by $20 per share. In this situation, Ms. Smith owes Brock 100 shares of
XYZ with a current market value of 100 shares $120 per share = $12,000 but has only $10,000 in
her account. If she skips town, Brock will have collateral of $10,000 (in cash) but a loan of $12,000,
resulting in a loss of $2,000. However, Brock can use margin requirements to protect itself from
experiencing losses from short sellers who do not repay their loans. In this example, Ms. Smith
must not only leave the short-sale proceeds with her broker, but she must also give he broker initial
margin applied to the amount of the short sale. Assuming the initial margin requirement is 60%,
she must give her broker 0.6 $10,000 = $6,000 in cash.
In this example, XYZ stock would have to rise in value to a price above $160 per share in order
for Brock to be in jeopardy of not being paid. Thus, initial margin provides the brokerage rm with
a certain degree of protection. However, this protection is not complete, since it is not unheard
of for stock to rise in value by more than 60%. It is the maintenance margin that protects the
brokerage rm from losing money in such situations. In order to examine the use of maintenance
margin in short sales, the actual margin in a short sale is dened as:
Actual Margin =
Market Value of Assets Loan

.
Loan
In this example, if XYZ stock rises to $130 per share, the actual margin in Ms. Smiths account
will be
($100 100)(1 + 0.6) ($130 100)
= 23%.
$130 100
Assuming the maintenance margin requirement is 30%, the account is undermargined, and Ms.
Smith will receive a margin call. Just as in margin calls on margin purchases, she will be asked to
put up more margin, meaning she will be asked to add cash or securities to her account.
If, instead of rising, the stock price falls, then the short seller can take a bit more than the drop
in the price of the account in the form of cash, since in this case the actual margin has risen above
the initial margin requirement and the account is thus unrestricted.8
Having discussed the cases for short sales where the stock price either (1) fell and the account was
thereby unrestricted or (2) went up to such a degree that the maintenance margin requirement was
violated and the account was thereby undermargined, there is one more case left to be considered.
That is the case where the stock price goes up but not to such a degree that the maintenance
margin is violated. In this case, the initial margin requirement has been violated, which means
that the account is restricted. Here, restricted has a meaning similar to its meaning for margin
purchases. That is, any transaction that has the eect of further decreasing the actual margin in
the account will be prohibited.
An interesting question is: what happens to the cash in the short sellers account? When the
loan is repaid, the short seller will have access to the cash (actually, the cash is used to repay
the loan). Before the loan is repaid, however, it may be that the short seller can earn interest on
the portion of the cash balance that represents margin (some brokerage rms will accept certain
securities, such as Treasury bills, in lieu of cash for meeting margin requirements). In regard to
8
Alternatively, the short seller could short sell a second security and not have to put up all (or perhaps any) of
the initial margin.
Short Sales
An Overview
AD.10
the cash proceeds from the short sale, sometimes the securities may be lent only on the payment
of a premium by the short seller, meaning the short seller not only does not earn interest on the
cash proceeds but must pay a fee for the loan. At other times the lender may pay the short seller
interest on the cash proceeds. Usually, however, securities are loaned at the brokerage rm
keeps the cash proceeds from the short sale and enjoys the use of this money, and neither the short
seller nor the investor who lent the securities receives any direct compensation. In this case, the
brokerage rm makes money not only from the commission paid by the short seller but also on the
cash proceeds from the sale (they may, for example, earn interest by purchasing Treasury bills with
these proceeds.)
Mathematics and Statistics
A Reminder
AD.11

A Reminder
Problems:
1. The only logarithm that we will use in this course (and in all other nance courses as a
matter of fact) is the natural logarithm. We will denote the natural logarithm of a positive
number x by log x. The objective of this exercise is to remind you of some simple logarithm
manipulations.
(a) Are the following statements true of false.
(i)
(ii)
(iii)
(iv)
(v)
If y = log x, then x = ey .
log(x + y) = log x + log y.
log(xy) = log x + log y.
x
log(x/y) = log
log y .
log(xy ) = y log x.
(b) Solve y x = z for x (using natural logarithms only).

(c) What is log (log ee )?
2. In this course, we will sometimes be faced with quadratic equations of the form ax2 +bx+c = 0.
(a) What is the formula that gives the two roots (zeros) of this equation in terms of a, b,
and c?
(b) Solve
1
x
+1=
2
x2
for x.
3. To simplify long equations, we will also make use of summation signs.

(a) Calculate 4n=1 2n .

(b) Calculate 3n=1 3m=n mn.
4. Random variables are crucial to the study of nance. This course will often require calculations of the expected value and the variance of a random variable, as well as the covariance
between two random variables. Suppose that the joint distribution (i.e. Prob {
x = x, y = y})
of two random variables x
and y is given by:
x
y
Calculate the following:
(a) Prob {
x = 0} and Prob {
x = 1}.
(b) Prob {
y = 1} and Prob {
y = 2}.
(c) E(
x) and E(
y ).
0
0.2
1
0.4
0.3
0.1
A Reminder
AD.12
(d) Var(
x) and Var(
y ).
(e) Cov(
x, y).
Solutions:
1. (a) (i)
(ii)
(iii)
(iv)
(v)
True.
False.
True.
False.
True.
(b) Take the (natural) logarithm on both sides of y x = z to obtain:

log y x = log z.
Since the left-hand side of this last expression simplies to x log y using one of the rules
from part (a), we have
log z
.
x=
log y
(c) The (natural) logarithm of ee is simply e, and the (natural) logarithm of e is 1. So
log (log ee ) = log(e) = 1.
2. (a) The two roots (if they both exist) are given by
b b2 4ac
.
x=
2a
(b) First multiply both sides of the equation by x2 to obtain
x + x2 = 2,
which we can rewrite as
x2 + x 2 = 0.
We can now use the above formula to get

1 (1)2 4(1)(2)
= 1 or 2.
x=
2(1)
3. (a)
4
n=1 2
= 21 + 22 + 23 + 24 = 2 + 4 + 8 + 16 = 30.
(b) This one can be calculated step by step:

3
3
3

3
3

mn =
m(1) +
m(2) +
m(3)
n=1 m=n
m=1
m=2
m=3
= [1(1) + 2(1) + 3(1)] + [2(2) + 3(2)] + [3(3)]

= 6 + 10 + 9 = 25.
A Reminder
4. (a) Prob {
x = 0} = 0.2 + 0.3 = 0.5, and
Prob {
x = 1} = 0.4 + 0.1 = 0.5
(b) Prob {
y = 1} = 0.2 + 0.4 = 0.6, and
Prob {
y = 2} = 0.3 + 0.1 = 0.4
(c) E(
x) = 0 Prob {
x = 0} + 1 Prob {
x = 1} = 0.5.
E(
y ) = 1 Prob {
y = 1} + 2 Prob {
y = 2} = 1.4.
(d) Var(
x) = (0 0.5)2 Prob {
x = 1} = 0.25.
x = 0} + (1 0.5)2 Prob {
2
2
Var(
y ) = (1 1.4) Prob {
y = 1} + (2 1.4) Prob {
y = 2} = 0.24.
(e) Let us solve this one in more details:

x E(
x) y E(
y ) Prob {
x = x, y = y}
Cov(
x, y)
x
= [0 0.5][1 1.4](0.2) + [0 0.5][2 1.4](0.3)

+[1 0.5][1 1.4](0.4) + [1 0.5][2 1.4](0.1)
= 0.04 0.09 0.08 + 0.03 = 0.1
AD.13
Linear Algebra
Simple Methods
AD.14
Linear Algebra
Simple Methods
As seen in section I.2 of my lecture notes, knowing how to solve a system of linear equations can
often be useful in this course. In this handout, I will try to show you three approaches for solving
systems of linear equations. My goal is not to give you a rigorous exposition of linear algebra, but
rather to give you a few simple methods that (I hope) will help you in your homeworks and exams.
Lets start with a concrete example: suppose you want to solve the following system of linear
equations, i.e. you want to solve for x, y, z in
2x + y + z =
(8)
x + y + 4z = 19
(9)
4x + 3y + z = 15
(10)
First Method: Substitution

The idea of this rst method is to substitute one equation into the others in order to reduce the
3 3 (3 equations and 3 unknowns) system to a 2 2 system rst, and eventually to a single
equation with one unknown.
For example, lets substitute equation (8) into equations (9) and (10), using the y variable.9 By
this, I mean that we rst write an expression for y in terms of x and z using equation (8):
y = 9 2x z.
(11)
Then we use this equation to replace y in each of equations (9) and (10):
x + (9 2x z) + 4z = 19
x + 3z = 10
(12)
4x + 3(9 2x z) + z = 15
2x 2z = 12
(13)
Observe that we have reduced our original 3 3 system of equations to a 2 2 system of equations
given by (12) and (13). We can now repeat the same strategy to reduce this system to a single
equation with a single unknown. For example, lets substitute equation (12) into equation (13)
using the variable x. From (12), we have
x = 3z 10.
Replacing x with this expression in (13), we get
2(3z 10) 2z = 12
6z + 20 2z = 12
8z = 32
z = 4.
Now that we have found z, we can use it in (14) to obtain x:

x = 3(4) 10 = 2.
9
Note that we could use any of the three equations and any of the three variables.
(14)
Linear Algebra
Simple Methods
AD.15
Finally, we can use the values of x and z in (11) to obtain y:

y = 9 2(2) 4 = 1.
So the solution to the system of equations (8)-(10) is given by x = 2, y = 1, z = 4. In fact, you
can plug these values in (8), (9), (10), and verify that these values of x, y, z do indeed solve the
equations.
Second Method: Diagonal Reduction

This second method requires a little more time to get used to than the rst method. However, with
a little practice, it becomes much faster than the rst method.
To use this method, we rst need to rewrite our original system of equations (8)-(10) in matrix
form. This is done by writing in the rst column the coecients of x in the three equations.
Similarly, the second and third column contain the y and z coecients. Finally, the last column,
which should be separated from the other three, contains the right-hand sides of each equation. So
in this example, we write:

2 1 1 9
1 1 4 19

4 3 1 15
In order to solve our problem, we are allowed three kinds of operations on this matrix:
We can multiply (divide) any line by any nonzero constant.
We can interchange any two lines.
We can add (subtract) any multiple of a line to any other line.
Suppose for example that we want
2 1
1 1
4 3
Suppose that we now want
1
2
4
to interchange line 1 and

1 9
1 1
L L
4 19 1 2 2 1
1 15
4 3
line 2 above. Then we write10

4 19
1 9
1 15
to subtract 2 times line 1 to line 2 of this last matrix. We then write11

1 4 19
1
1
4
19
L 2L L
1 1 9 2 1 2 0 1 7 29
3 1 15
4
3
1
15
The general goal of this method is to eventually obtain an identity matrix,
1 0 0
0 1 0 ,
0 0 1
in the 3 3 part of the matrix. The rightmost column will then give us the values of x, y, z that
solve our system of equations. This can be achieved by the following steps:
10
The symbol means that the two systems are equivalent. The left-right arrow means that we interchange
two lines.
11
The right arrow means that the expression on the left-hand side will replace the line on the right-hand side.
Linear Algebra
Simple Methods
AD.16
1. Reduce the numbers below the diagonal to zeros.

2. Reduce the numbers above the diagonal to zeros.
The complete solution to our example is then12

2 1 1 9
1
1
4
19
1 1 4 19
L2
1 L2

1 1 4 19 L1
2 1 1 9 L2 2L
0 1 7 29

4 3 1 15
4
3
1
15
4 3 1 15

1
1
4
19
1
1
4
19
L L
L3 4L1 L3
L2 L2

0
1
7
29 3 3
0 1 7 29
0 1 15 61
0 1 15 61

19
1
1
4
1 1 0
1 1 4 19
1

L L3
L3 L2 L3
L1 4L3 L1
8 3

0 1 7 29
0 1 7
0 1 7 29
0 0 1 4
0 0 1
0 0 8 32

1 0 0 2
1 1 0 3
L1 L2 L1
L2 7L3 L2

0 1 0 1
0 1 0 1
0 0 1 4
0 0 1 4
1 1 4
0 1 7
0 1 15

3

29

4
19

29

61
The solution to our original system of equations (8)-(10) is therefore given by x = 2, y = 1, z = 4.
Third Method: Matrix Inversion

This method is the quickest of the three methods presented in this handout. However, it requires
a computer or a calculator that performs matrix inversions and multiplications.13 Also, I will only
present an outline of the method since this method requires more knowledge of linear algebra.
The systems of equations (8)-(10) can
2
1
4
be written in matrix form as

1 1
x
9
1 4 y = 19 .
3 1
z
15
The solution is then given by

1
9
x
2 1 1
y = 1 1 4 19 .
15
z
4 3 1
Therefore, the solution to our system of equations can be found by applying the following two steps:
2 1 1
1. Find the (matrix) inverse of 1 1 4 .
4 3 1
9
2. (Matrix) Multiply that inverse by 19 .
15
12
Note that, once you get used to this method, you dont really need to write anything above at every step.
However, it is a good idea to do so at the beginning so that you can trace back your mistakes.
13
It is possible to perform matrix inversions and multiplications by hand, but it is tedious and not very important
for this course. In any case, I show in the next section how to use a spreadsheet software to perform such calculations.
Linear Algebra
Simple Methods
A
1
2
3
4
5
6
7
AD.17
2
1
4
1
1
3
1
4
1
1.375
-1.875
0.125
-0.25
0.25
0.25
-0.375
0.875
-0.125
9
19
15
2
1
4
Figure 9: Matrix manipulations using Microsoft Excel.

Using either a computer (see next section) or a calculator, we rst nd that
1 11
2 1 1
8
1 1 4 = 15
8
1
4 3 1
8
14
1
4
1
4
38
7
8
18
9
Finally, by (matrix) multiplying this inverse by 19 , we obtain the solution to our system of
15
equations:

11
9
2
14 38
x
8
1
7
= 1 .
y = 15
19
8
8
4
1
1
1
15
4
z
8
4 8
Matrix Manipulations Using a Spreadsheet Software

In this section, I show how to use Microsoft Excel to manipulate matrices. The steps to perform
the same manipulations on other spreadsheet softwares should be similar. Figure 9 shows the
spreadsheet on which I did my calculations.
First in A1:C3, I enter the matrix I want to invert. Then I position the cursor in cell A5, highlight
the range A5:C7, type
=MINVERSE(A1:C3)
and press Ctrl-Shift-Enter.14 The inverted matrix appears in A5:C7.
To perform the matrix multiplication, I rst need to enter my second matrix; this is what I do in
E5:E7. Then I position the cursor in cell G5, highlight the range G5:G7, type
=MMULT(A5:C7,E5:E7)
and press Ctrl-Shift-Enter. The solution to my system of linear equations then appears in G5:G7.
14
If you only press Enter, only the top-left cell (cell A5) will appear. The Ctrl-Shift-Enter tells Excel that you
want a matrix as a result.
Linear Algebra
Simple Methods
AD.18
A Few Additional Problems

For practice, you can solve the following systems of linear equations using each of the three methods.
The solutions are also provided so that you can check your answers.
System 1:
x+ y
= 1
2x + 4y z = 6
y + 2z = 2
Solution : x = 1, y = 2, z = 0.
System 2:
x y z = 0
2x
+ 4z = 2
2y + 2z = 1
1
1
1
Solution : x = , y = , z = .
2
4
4
Formulas
A Sampler
AD.19
Formulas
A Sampler
Mathematics
ax + bx + c = 0 x =
2
b2 4ac
2a
z = ex x = log z
log z
z = yx x =
log y
y
log(x ) = y log x
Quadratic equation
Natural logarithm
log(xy) = log x + log y

log(x/y) = log(x) log y
Compounding and Discounting

r mt
FVt = C 1 +
m
rt
FVt = Ce
1
DFt =
(1 + rt )t
T
T

Ct
PV =
=
Ct DFt
(1 + rt )t
t=1
NPV = C0 +
T

t=1
t=1
Continuous compounding
Discount factor
Present value formula

Ct
Ct
=
(1 + rt )t
(1 + rt )t
Net present value
t=0
C
r
C
PV =
rg

1
C
1
PV =
r
(1 + r)T

C
1+g T
PV =
1
1+r
rg
PV =
Compounding m times a year
Perpetuity
Growing perpetuity
(1st payment: C at 1)
T -year annuity
T -year growing annuity
Note: In these last four formulas, the payments are made at the end of every year. If
instead the payments are made at the beginning of every year, multiply these present values
by (1 + r).
Formulas
A Sampler
AD.20
Bond Prices and the Term Structure

P =
P =
T

t=1
T

t=1
C
F
+
t
(1 + rt )
(1 + rT )T
Bond prices
C
F
+
t
(1 + y)
(1 + y)T
Bond yields
(1 + rt )t
DFt1
1=
1
(1 + rt1 )t1
DFt
ft =
(1 + Rt )t =
(1 + rt )t
(1 + it )t
Forward rates
Real interest rates

Capital Budgeting Under Certainty
C0 NPV
PV
=
C0
C0
T

Ct
0=
(1 + IRR)t
PI =
t=0
Protability index
Internal rate of return
Exam Material
Formulas
EX.42
Exam Material
Formulas
Statistics
=
X E(X)
xi p(xi )
Expectation
2
= E[(X
EX)
2] =
Var(X)
X
(xi X)2 p(xi )
Variance
Y ) = E[(X
EX)(
Y EY )]
XY Cov(X,
Covariance
XY
X Y
+ b) = aE(X)
+b
E(aX
Correlation coecient
XY =
Properties of expectation
+ Y ) = E(X)
+ E(Y )
E(X
+ b) = a2 Var(X)
Var(aX
Properties of variance
+ b2 Var(Y ) + 2abCov(X,
Y )
+ bY ) = a2 Var(X)
Var(aX
2
+ b2 Y2 + 2abXY X Y
= a2 X
+ b, cY + d) = acCov(X,
Y )
Cov(aX
Properties of covariance
Y ) = Cov(Y , X)
Cov(X,
+ Y , Z)
= Cov(X,
Z)
+ Cov(Y , Z)
Cov(X
2
= E(X
2 ) (EX)
Var(X)
Variance shortcut
Y ) = E(X
Y ) E(X)E(
Cov(X,
Y )
N

= 1
xi
X
N
i=1
N
2
1
2
xi X
X =
N 1
Covariance shortcut
Estimator of the mean
Estimator of the variance
i=1
Mathematics
ax2 + bx + c = 0 x =
z = ex x = log z
log z
z = yx x =
log y
y
log(x ) = y log x
log(xy) = log x + log y
log(x/y) = log(x) log y
b2 4ac
2a
Quadratic equation
Natural logarithm
Exam Material
Formulas
EX.43
Compounding and Discounting

r mt
FVt = C 1 +
m
rt
FVt = Ce
1
DFt =
(1 + rt )t
PV =
T

t=1
Continuous compounding
Discount factor

Ct
=
Ct DFt
(1 + rt )t
T
Present value formula
t=1
NPV = C0 +
T

t=1
PV =
Compounding m times a year

Ct
Ct
=
t
(1 + rt )t
(1 + rt )
T
Net present value
t=0
C
r
Perpetuity
C
rg

1
C
1
PV =
r
(1 + r)T

1+g T
C
1
PV =
1+r
rg
PV =
Growing perpetuity
T -year annuity
T -year growing annuity
Note: In these last four formulas, the payments are made at the end of every year. If
instead the payments are made at the beginning of every year, multiply these present
values by (1 + r).
Bond Prices and the Term Structure
P =
P =
T

t=1
T

t=1
ft =
C
F
+
t
(1 + rt )
(1 + rT )T
Bond prices
F
C
+
t
(1 + y)
(1 + y)T
Bond yields
(1 + rt )t
DFt1
1=
1
(1 + rt1 )t1
DFt
(1 + Rt )t =
T
D=
(1 + rt )t
(1 + it )t
tCt
t=1 (1+y)t
T
Ct
t=1 (1+y)t
Forward rates
Real interest rates
Duration
Exam Material
Formulas
EX.44
Capital Budgeting Under Certainty

C0 NPV
PV
=
C0
C0
T

Ct
0=
(1 + IRR)t
Protability index
PI =
Internal rate of return
t=0
Portfolio Selection and the CAPM

rp =
N

N =2
Portfolio rate of return
N =2
Expected rate of return
wi ri = w1 r1 + w2 r2
i=1
rp =
p2
N

wi ri = w1 r1 + w2 r2
i=1
N
N

wi wj ij =
i=1 j=1
N =2 2 2
= w1 1 +
N
N

wi wj ij i j
Variance of rate of return
i=1 j=1
w22 22 + 2w1 w2 12 1 2
N 1
1
(avg. variance) +
(avg. covariance)
N
N
i
im
i = 2 = im
m
m
N

N =2
p =
wi i = w1 1 + w2 2
p2 =
Variance if wi =
1
N
Beta of asset i
Beta of a portfolio
i=1
rm rf
p
m
ri = rf + (rm rf )i
r p = rf +
Capital Market Line (CML)

Security Market Line (SML)
p = p m
w1 =
Ecient portfolios
12 1 2
, and w2 = 1 w1
2
+ 2 212 1 2
22
12
Minimum variance portfolio
Stock Prices
P0 =

t=1
Et It
Dt
E
+ PVGO
=
=
t
t
(1 + r)
(1 + r)
r
(r r)kE
r(r r k)

1
PVGO 1
P0
=
1
E
r
P0
PVGO =
Stock prices
t=1
PV of growth opportunities
P/E ratios
Exam Material
Formulas
EX.45
Capital Budgeting Under Uncertainty

rA =
D
E
rD +
rE
D+E
D+E
Cost of capital and leverage
A =
D
E
D +
E
D+E
D+E
Betas and leverage
E = A + (A D )
D
E
Ct
CEQt
=
t
(1 + r)
(1 + rf )t
PV =
C1
Certainty equivalent formulas
rm rf
Cov(C1 , rm )
2
m
1 + rf
Dividend Policy and Debt Policy
DIVt DIVt1 = a (p EPSt DIVt1 )
Lintners dividend model
VL = VU
MMs Proposition I
rE = rA +
D
E

(rA rD )
VL = VU + tc DL
D
+ rE
WACC = (1 tc ) rD

V
D
= WACCU 1 tc
V
MMs Proposition II
E
V
(1 tE )(1 tc )
1 tD
VL = VU + t DL

D
E
+ rE
WACC = (1 tc ) rD
V

V
D
= WACCU 1 t
V
Corporate taxes
(no personal taxes)
t = 1
Corporate and
personal taxes
V = VU + PV (tax shield) PV (bankruptcy costs)
Bankruptcy costs
Interaction of Investment and Financing Decisions

r = r(1 t L)
L=
D
APV + C0
r = r t rD L
Modigliani-Miller formula
1+r
1 + rD

Miles-Ezzel formula
Exam Material
Formulas
EX.46
The Valuation of Options

max[0, St XBt ] Ct St , where Bt =
Pt = Ct St + XBt = Ct St +
1
(1 + rf )T t
X
(1 + rf )T t
qCu + (1 q)Cd
1+r
(1 + r) d
q=
ud
Cu Cd
uCd dCu
=
, B=
(u d)S0
(u d)(1 + r)
C0 =
u = e
d = 1/u,
Bounds on call price

Put-call parity
Binomial formula
r = (1 + rf )h 1
X
N (x T )
C0 = S0 N (x)
T
(1 + rf )

S0
log X/(1+r
1
)T
f
+ T
x=
2
T
Black-Scholes formula
With dividends: S0 = S0 PV (dividends)

C =
S0
S
C0
Beta of a call option

The Valuation of Corporate Liabilities
X
Pt (Vt , X, T )
(1 + rf )T t

Vt
1
nS
Wt =
Ct (Vt , nS X, T ) =
Ct
, X, T
nS + nW
nS + nW
nS

n
X
CDt = Dt + q Ct Vt , , T , where q =
q
nS + n
Dt = Vt Ct (Vt , X, T ) =
Value of risky debt

Value of warrants
Value of convertible bonds
Exam Material
Formulas
EX.47
Cumulative Density Function for a Standard Normal Variable

Each of the following cells contains the value N (x), where x is the sum of the leftmost and top cells.
x<0
-0.09
-0.08
-0.07
-0.06
-0.05
-0.04
-0.03
-0.02
-0.01
-5.0
-4.5
-4.0
-3.5
-3.0
-2.9
-2.8
-2.7
-2.6
-2.5
-2.4
-2.3
-2.2
-2.1
-2.0
-1.9
-1.8
-1.7
-1.6
-1.5
-1.4
-1.3
-1.2
-1.1
-1.0
-0.9
-0.8
-0.7
-0.6
-0.5
-0.4
-0.3
-0.2
-0.1
0.0
x0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
-3.0
-3.5
-4.0
-4.5
-5.0
-0.00
1N (5.0)
1N (4.5)
1N (4.0)
1N (3.5)
1N (3.0)
0.00139
0.00193
0.00264
0.00357
0.00480
0.00639
0.00842
0.01101
0.01426
0.01831
0.02330
0.02938
0.03673
0.04551
0.05592
0.06811
0.08226
0.09853
0.11702
0.13786
0.16109
0.18673
0.21476
0.24510
0.27760
0.31207
0.34827
0.38591
0.42465
0.46414
0.00144
0.00199
0.00272
0.00368
0.00494
0.00657
0.00866
0.01130
0.01463
0.01876
0.02385
0.03005
0.03754
0.04648
0.05705
0.06944
0.08379
0.10027
0.11900
0.14007
0.16354
0.18943
0.21770
0.24825
0.28096
0.31561
0.35197
0.38974
0.42858
0.46812
0.00149
0.00205
0.00280
0.00379
0.00508
0.00676
0.00889
0.01160
0.01500
0.01923
0.02442
0.03074
0.03836
0.04746
0.05821
0.07078
0.08534
0.10204
0.12100
0.14231
0.16602
0.19215
0.22065
0.25143
0.28434
0.31918
0.35569
0.39358
0.43251
0.47210
0.00154
0.00212
0.00289
0.00391
0.00523
0.00695
0.00914
0.01191
0.01539
0.01970
0.02500
0.03144
0.03920
0.04846
0.05938
0.07215
0.08692
0.10383
0.12302
0.14457
0.16853
0.19489
0.22363
0.25463
0.28774
0.32276
0.35942
0.39743
0.43644
0.47608
0.00159
0.00219
0.00298
0.00402
0.00539
0.00714
0.00939
0.01222
0.01578
0.02018
0.02559
0.03216
0.04006
0.04947
0.06057
0.07353
0.08851
0.10565
0.12507
0.14686
0.17106
0.19766
0.22663
0.25785
0.29116
0.32636
0.36317
0.40129
0.44038
0.48006
0.00164
0.00226
0.00307
0.00415
0.00554
0.00734
0.00964
0.01255
0.01618
0.02068
0.02619
0.03288
0.04093
0.05050
0.06178
0.07493
0.09012
0.10749
0.12714
0.14917
0.17361
0.20045
0.22965
0.26109
0.29460
0.32997
0.36693
0.40517
0.44433
0.48405
0.00169
0.00233
0.00317
0.00427
0.00570
0.00755
0.00990
0.01287
0.01659
0.02118
0.02680
0.03362
0.04182
0.05155
0.06301
0.07636
0.09176
0.10935
0.12924
0.15151
0.17619
0.20327
0.23270
0.26435
0.29806
0.33360
0.37070
0.40905
0.44828
0.48803
0.00175
0.00240
0.00326
0.00440
0.00587
0.00776
0.01017
0.01321
0.01700
0.02169
0.02743
0.03438
0.04272
0.05262
0.06426
0.07780
0.09342
0.11123
0.13136
0.15386
0.17879
0.20611
0.23576
0.26763
0.30153
0.33724
0.37448
0.41294
0.45224
0.49202
0.00181
0.00248
0.00336
0.00453
0.00604
0.00798
0.01044
0.01355
0.01743
0.02222
0.02807
0.03515
0.04363
0.05370
0.06552
0.07927
0.09510
0.11314
0.13350
0.15625
0.18141
0.20897
0.23885
0.27093
0.30503
0.34090
0.37828
0.41683
0.45620
0.49601
0.00187
0.00256
0.00347
0.00466
0.00621
0.00820
0.01072
0.01390
0.01786
0.02275
0.02872
0.03593
0.04457
0.05480
0.06681
0.08076
0.09680
0.11507
0.13567
0.15866
0.18406
0.21186
0.24196
0.27425
0.30854
0.34458
0.38209
0.42074
0.46017
0.50000
0.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.50000
0.53983
0.57926
0.61791
0.65542
0.69146
0.72575
0.75804
0.78814
0.81594
0.84134
0.86433
0.88493
0.90320
0.91924
0.93319
0.94520
0.95543
0.96407
0.97128
0.97725
0.98214
0.98610
0.98928
0.99180
0.99379
0.99534
0.99653
0.99744
0.99813
0.998650033
0.999767327
0.999968314
0.999996599
0.999999713
0.50399
0.54380
0.58317
0.62172
0.65910
0.69497
0.72907
0.76115
0.79103
0.81859
0.84375
0.86650
0.88686
0.90490
0.92073
0.93448
0.94630
0.95637
0.96485
0.97193
0.97778
0.98257
0.98645
0.98956
0.99202
0.99396
0.99547
0.99664
0.99752
0.99819
0.50798
0.54776
0.58706
0.62552
0.66276
0.69847
0.73237
0.76424
0.79389
0.82121
0.84614
0.86864
0.88877
0.90658
0.92220
0.93574
0.94738
0.95728
0.96562
0.97257
0.97831
0.98300
0.98679
0.98983
0.99224
0.99413
0.99560
0.99674
0.99760
0.99825
0.51197
0.55172
0.59095
0.62930
0.66640
0.70194
0.73565
0.76730
0.79673
0.82381
0.84849
0.87076
0.89065
0.90824
0.92364
0.93699
0.94845
0.95818
0.96638
0.97320
0.97882
0.98341
0.98713
0.99010
0.99245
0.99430
0.99573
0.99683
0.99767
0.99831
0.51595
0.55567
0.59483
0.63307
0.67003
0.70540
0.73891
0.77035
0.79955
0.82639
0.85083
0.87286
0.89251
0.90988
0.92507
0.93822
0.94950
0.95907
0.96712
0.97381
0.97932
0.98382
0.98745
0.99036
0.99266
0.99446
0.99585
0.99693
0.99774
0.99836
0.51994
0.55962
0.59871
0.63683
0.67364
0.70884
0.74215
0.77337
0.80234
0.82894
0.85314
0.87493
0.89435
0.91149
0.92647
0.93943
0.95053
0.95994
0.96784
0.97441
0.97982
0.98422
0.98778
0.99061
0.99286
0.99461
0.99598
0.99702
0.99781
0.99841
0.52392
0.56356
0.60257
0.64058
0.67724
0.71226
0.74537
0.77637
0.80511
0.83147
0.85543
0.87698
0.89617
0.91308
0.92785
0.94062
0.95154
0.96080
0.96856
0.97500
0.98030
0.98461
0.98809
0.99086
0.99305
0.99477
0.99609
0.99711
0.99788
0.99846
0.52790
0.56749
0.60642
0.64431
0.68082
0.71566
0.74857
0.77935
0.80785
0.83398
0.85769
0.87900
0.89796
0.91466
0.92922
0.94179
0.95254
0.96164
0.96926
0.97558
0.98077
0.98500
0.98840
0.99111
0.99324
0.99492
0.99621
0.99720
0.99795
0.99851
0.53188
0.57142
0.61026
0.64803
0.68439
0.71904
0.75175
0.78230
0.81057
0.83646
0.85993
0.88100
0.89973
0.91621
0.93056
0.94295
0.95352
0.96246
0.96995
0.97615
0.98124
0.98537
0.98870
0.99134
0.99343
0.99506
0.99632
0.99728
0.99801
0.99856
0.53586
0.57535
0.61409
0.65173
0.68793
0.72240
0.75490
0.78524
0.81327
0.83891
0.86214
0.88298
0.90147
0.91774
0.93189
0.94408
0.95449
0.96327
0.97062
0.97670
0.98169
0.98574
0.98899
0.99158
0.99361
0.99520
0.99643
0.99736
0.99807
0.99861
Solution to Slides I.4.25-I.4.26
Calculating Statistics
AD.2

r2 = y} = g(y) for r1
The (marginal) probability distributions Pr{
r1 = x} = f (x) and Pr{
and r2 are obtained by summing the rows and columns of the joint probability distribution:
x
10%
30%
g(y)
15%
45%
0.2
0.3
0.3
0.2
0.5
0.5
f (x)
0.5
0.5
We then have
E(
r1 ) = 0.10(0.5) + 0.30(0.5) = 0.10;
E(
r2 ) = 0.15(0.5) + 0.45(0.5) = 0.15;
Var(
r1 ) = (0.10 0.10)2 (0.5) + (0.30 0.10)2 (0.5) = 0.04;
Var(
r2 ) = (0.15 0.15)2 (0.5) + (0.45 0.15)2 (0.5) = 0.09;

1 = Var(
r1 ) = 0.04 = 0.20;

2 = Var(
r2 ) = 0.09 = 0.30.
The covariance between r1 and r2 is calculated as follows:
Cov(
r1 , r2 ) = (0.10 0.10)(0.15 0.15)(0.2) + (0.30 0.10)(0.15 0.15)(0.3)
+(0.10 0.10)(0.45 0.15)(0.3) + (0.30 0.10)(0.45 0.15)(0.2)
= 0.012.
The correlation coecient is
12 =
Cov(
r1 , r2 )
0.012
=
= 0.20.
1 2
(0.20)(0.30)
Using the properties of means shown on page I.4.15, we have

r1 + 0.5
r2 )
E(
rp ) = E(0.5
r2 )
= 0.5E(
r1 ) + 0.5E(
= 0.5(0.10) + 0.5(0.15)
= 0.125,
and
r1 + 0.4
r2 )
E(
rq ) = E(0.6
r2 )
= 0.6E(
r1 ) + 0.4E(
= 0.6(0.10) + 0.4(0.15)
= 0.12.
AD.3
Using the properties of variances shown on page I.4.18, we have

r1 + 0.5
r2 )
Var(
rp ) = Var(0.5
= (0.5)2 Var(
r2 ) + 2(0.5)(0.5)Cov(
r1 , r2 )
r1 ) + (0.5)2 E(
= (0.5)2 (0.04) + (0.5)2 (0.09) + 2(0.5)(0.5)(0.012)
= 0.0265,
and
r1 + 0.4
r2 )
Var(
rq ) = Var(0.6
= (0.6)2 Var(
r2 ) + 2(0.6)(0.4)Cov(
r1 , r2 )
r1 ) + (0.4)2 E(
= (0.6)2 (0.04) + (0.4)2 (0.09) + 2(0.6)(0.4)(0.012)
= 0.02304.
This implies that

Var(
rp ) = 0.0265 = 0.1628, and

q = Var(
rq ) = 0.02304 = 0.0937.
p =
Using the properties of covariances shown on page I.4.18, we have

r1 + 0.4
r2 )
r1 + 0.5
r2 , 0.6
Cov(
rp , rq ) = Cov(0.5
r1 , r2 )
= (0.5)(0.6)Cov(
r1 , r1 ) + (0.5)(0.4)Cov(
r2 , r2 )
+ (0.5)(0.6)Cov(
r2 , r1 ) + (0.5)(0.4)Cov(
r1 , r2 )
= (0.5)(0.6)Var(
r1 ) + (0.5)(0.4)Cov(
r2 )
+ (0.5)(0.6)Cov(
r1 , r2 ) + (0.5)(0.4)Var(
= (0.5)(0.6)(0.04) + (0.5)(0.4)(0.012)
+ (0.5)(0.6)(0.012) + (0.5)(0.4)(0.09)
pq
= 0.024, and
Cov(
rp , rq )
0.024
=
=
= 0.9713.
(0.1628)(0.0937)
p q
The following table shows the dierent values that

r1 + 0.5
r2
rp = 0.5
and
r1 + 0.4
r2
rq = 0.6
can take depending on the values that r1 and r2 take:
r1
r2
rp
rq
prob.
-0.10
-0.10
0.30
0.30
-0.15
0.45
-0.15
0.45
-0.125
0.175
0.075
0.375
-0.12
0.12
0.12
0.36
0.2
0.3
0.3
0.2
The probability distribution for rp is therefore given by:

x
Pr{
rp = x}
-0.125
0.075
0.175
0.375
0.2
0.3
0.3
0.2
We then have
E(
rp ) = 0.125(0.2) + (0.075)(0.3) + (0.175)(0.3) + (0.375)(0.2)
= 0.125,
and
Var(
rp ) = (0.125 0.125)2 (0.2) + (0.075 0.125)2 (0.3)
+ (0.175 0.125)2 (0.3) + (0.375 0.125)2 (0.2)
= 0.0265.
Similarly, the probability distribution for rq is given by:
y
Pr{
rq = y}
-0.12
0.12
0.36
0.2
0.6
0.2
We then have
E(
rq ) = 0.12(0.2) + (0.12)(0.6) + (0.36)(0.2)
= 0.12,
and
Var(
rq ) = (0.12 0.12)2 (0.2) + (0.12 0.12)2 (0.6) + (0.36 0.12)2 (0.2)
= 0.02304.
AD.4
Addition to Slide I.4.32
The Variance of a Portfolio of N Assets
AD.5

In this handout, we will derive the formula for the return variance of a portfolio composed of
N assets:
N
N

wi wj Cov(
ri , rj ),
Var(
rp ) =
i=1 j=1
where
rp = w1 r1 + w2 r2 + + wN rN =
N

wi ri .
i=1
To do this, let us rst recall that the variance of a random variable with itself is simply the variance
of that variable (see page I.4.16 in the lecture notes), so that
Var(
rp ) = Cov(
rp , rp )
= Cov(w1 r1 + w2 r2 + + wN rN , w1 r1 + w2 r2 + + wN rN ).
(36)
Y and Z,
a property of covariances (see page I.4.17 in the
Now, for three random variables X,
lecture notes) is that
+ Y , Z)
= Cov(X,
Y ) + Cov(X,
Z).
Cov(X
(37)
Since we can rewrite (36) as
+ Y , Z),
Cov(w1 r1 + w2 r2 + + wN rN , w1 r1 + w2 r2 + + wN rN ) = Cov(X
where
= w1 r1 ,
X
Y = w2 r2 + + wN rN , and
Z = w1 r1 + w2 r2 + + wN rN ,
we can use (37) to obtain
Var(
rp )
=
=
Cov(w1 r1 + w2 r2 + + wN rN , w1 r1 + w2 r2 + + wN rN )
+ Y , Z)
Cov(X
Y ) + Cov(X,
Z)
Cov(X,
Cov(w1 r1 , w1 r1 + w2 r2 + + wN rN )
+ Cov(w2 r2 + + wN rN , w1 r1 + w2 r2 + + wN rN ).
A similar argument shows that

Cov(w2 r2 + + wN rN , w1 r1 + w2 r2 + + wN rN )
=
Cov(w2 r2 , w1 r1 + w2 r2 + + wN rN )
+ Cov(w3 r3 + + wN rN , w1 r1 + w2 r2 + + wN rN ).
AD.6
In fact, we can repeat this same argument N times to obtain

Var(
rp )
Cov(w1 r1 , w1 r1 + w2 r2 + + wN rN )
+ Cov(w2 r2 , w1 r1 + w2 r2 + + wN rN )
+
+ Cov(wN rN , w1 r1 + w2 r2 + + wN rN )
N

Cov(wi ri , w1 r1 + w2 r2 + + wN rN ).
(38)
i=1
We also know that (37) could be rewritten as

X
+ Y ) = Cov(Z,
X)
+ Cov(Z,
Y ).
Cov(Z,
(39)
Since we can rewrite the term in the summation in (38) as

X
+ Y ),
Cov(wi ri , w1 r1 + w2 r2 + + wN rN ) = Cov(Z,
where
= w1 r1 ,
X
Y = w2 r2 + + wN rN ,
Z = wi ri ,
and
we can use (39) to obtain

Cov(wi ri , w1 r1 + w2 r2 + + wN rN ) = Cov(wi ri , w1 r1 ) + Cov(wi ri , w2 r2 + + wN rN ).
Similarly,
Cov(wi ri , w2 r2 + + wN rN ) = Cov(wi ri , w2 r2 ) + Cov(wi ri , w3 r3 + + wN rN ).
After repeating this same reasoning N times, we get
Cov(wi ri , w1 r1 + w2 r2 + + wN rN )
= Cov(wi ri , w1 r1 ) + Cov(wi ri , w2 r2 ) + + Cov(wi ri , wN rN )
=
N

Cov(wi ri , wj rj ),
j=1
so that we can rewrite (38) as follows:

Var(
rp ) =
N
N

Cov(wi ri , wj rj ).
i=1 j=1
Finally, another property of covariances (see page I.4.17 of the lecture notes) says
cY ) = acCov(X,
Y ),
Cov(aX,
(40)
so that we can rewrite (40) as

Var(
rp ) =
N
N

Cov(wi ri , wj rj )
i=1 j=1
N
N

wi wj Cov(
ri , rj )
i=1 j=1
N
N

i=1 j=1
This completes the derivation.
wi wj ij .
AD.7
Solution to Slide I.4.40
Portfolio Selection
AD.8

Portfolio Selection
We are looking for the portfolio (w1 , 1 w1 ) which minimizes the portfolios variance,
p2 = w12 12 + (1 w1 )2 22 + 2w1 (1 w1 )12 1 2 .
Dierentiating this variance with respect to w1 results in
dp2
= 2w1 12 2(1 w1 )22 + 2(1 2w1 )12 1 2 .
dw1
We can then set this last expression equal to zero, and solve for w1 :11
w1 =
22 12 1 2
(0.2)2 (0.2)(0.3)(0.2)
=
= 0.264.
(0.3)2 + (0.2)2 2(0.2)(0.3)(0.2)
12 + 22 212 1 2
Therefore, the minimum variance portfolio consists in investing 26.4% of the portfolio in
asset 1, and 1 26.4% = 73.6% in asset 2. The standard deviation of this portfolio is equal
to

p = (0.264)2 (0.3)2 + (0.736)2 (0.2)2 + 2(0.264)(0.736)(0.2)(0.3)(0.2) = 0.1806,
and its expected return is given by
rp = 0.264(0.1) + 0.736(0.3) = 0.247.
All the portfolios with a variance
smaller than or equal to 0.25 (i.e. with a standard deviation
smaller than or equal to 0.25 = 0.5) lie on the thick line in the following gure:
rp
0.75
0.5
0.25
0
-0.25
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
sp
Obviously, the portfolio (w1 , 1 w1 ) that we are looking for is the one on top of that thick
line. This portfolio solves
0.25 = w12 12 + (1 w1 )2 22 + 2w1 (1 w1 )12 1 2
0.25 = w12 (0.3)2 + (1 w1 )2 (0.2)2 + 2w1 (1 w1 )(0.2)(0.3)(0.2)
0 = 0.106w12 0.056w1 0.21

0.056 (0.056)2 4(0.106)(0.21)
w1 =
= 1.6962500 or 1.1679481
2(0.106)
11
It is easily veried that the second derivative is positive, so that we indeed found a minimum.
Portfolio Selection
AD.9
The reason we get two dierent solutions is that there are two portfolios with a variance of
0.25; in fact, this is quite obvious from the gure above. Since the portfolio we are interested
in has the higher expected return, we only need to calculate rp with w1 = 1.6962500 and
w1 = 1.1679481. We nd rp = 0.0392500 and rp = 0.5335896 respectively, so that
w1 = 1.1679481 is the portfolio that we want. For every dollar invested in this portfolio,
$(1 w1 ) = $2.1679481 is invested in asset 2, and $1.1679481 comes from (short-)selling
asset 1.
Diversication with Multiple Assets
AD.10

Diversification with Multiple Assets
In this handout, we show that the variance of a portfolio containing a larger and larger number of
assets approaches the average covariance between these assets. Using the notation from page I.4.45,
we have
p2
N
N

wi wj ij
i=1 j=1
N
N

1 1
ij
NN
i=1 j=1
N
N N

1 2 1
+
ij
N2 i
N2
i=1
i=1 j=1
j=i
N
N
N
1 2 1
+
ij
i
N2
N2
i=1
i=1
N
N
N
1 1 2 N 1 1
i +
ij
NN
N2 N 1
i=1
j=1
j=i
1
N
N
1 2
i
N
i=1
i=1
j=1
j=i
N
N
1
N 1
+
ij
N 1
N2
i=1
j=1
j=i
1
N 1
(average covariance of asset i with other assets)
(average variance) +
N
N2
i=1

N
average variance N 1 1
(average covariance of asset i with other assets)
+
N
N
N
N
=
=
i=1
=
N
average variance N 1
+
(average covariance between all assets)
N
N
0 + 1 (average covariance between all assets)
average covariance between all assets.
Why is the Average Covariance Greater Than Zero?
AD.11

Why is the Average Covariance Greater Than Zero?
In this handout, we will try to see why the average covariance between the assets of a portfolio is
positive when the number of assets is large (when N ).
Before we start, however, we should make clear that it is possible for the average covariance between
a nite number of assets to be negative. For example, if we take N 1 uncorrelated assets, and
add an asset that has negative correlation < 0 with one of those assets but is uncorrelated with
all the other assets, the average covariance will be
average covariance between assets =
2
2
=
< 0.
N
N (N 1)
(41)
N2
Our point is really a limit argument: it is imposible to keep adding assets, and keep the average
covariance below some negative number c. In other words, the minimum possible average covariance
in the limit is zero. For example, in (41), the average covariance cannot be prevented from going
to zero as N .
To make this argument more convincing, assume for convenience (but without loss of generality)
that the variance of every asset is 2 . Now, let us form an equally-weighted portfolio (w1 = w2 =
= wN = 1/N ) of N such assets. By the denition of variance, we know that the variance of this
portfolio return is positive, i.e. Var(
rp ) > 0, and this is obviously true for any number N of assets.
By splitting the variance and covariance terms, we can also rewrite the variance of this portfolio
return as follows:
N

Var(
rp ) =
wi2 Var(
ri )
N
N

wi wj Cov(
ri , rj )
i=1 j=1
j=i
i=1
N
N N
1 2
1
+ 2
Cov(
ri , rj )
N2
N
i=1
i=1 j=1
j=i
1
N 2 N 2 N
2
2
2
N
N
N N
N
N

i=1 j=1
j=i
Cov(
ri , rj )
2 N 1
+
(average covariance between assets)
N
N
=
So, since Var(
rp ) > 0, we have
2 N 1
+
(average covariance between assets) > 0,
N
N
for any N . If we let N grow to innity, the rst term in this equation goes to zero and
to one, and we then have
average covariance between assets > 0.
This completes the argument.
N 1
N
goes
Equilibrium A Numerical Example
AD.12

In this handout, we will see how the equilibrium of the Capital Asset Pricing Model (CAPM) comes
about. To illustrate this equilibrium, we consider an economy consisting of three rms (A, B and
C) and two investors (Al and Becky).
Suppose that we know the number shares outstanding for each of the three rms, and that we
have estimated the means, the standard deviations, and the correlations of the future value of their
assets in one year:
Firm
(i)
Outstanding
Shares (Si )
Mean
(mi )
Std. Dev.
(si )
A
B
C
40,000
50,000
40,000
400,000
600,000
800,000
160,000
300,000
200,000
Correlation (ij )
with B
with C
0.7
0.1
0.4
Two individuals, Al and Becky, hold all the wealth in the economy. More precisely, they each hold
$500,000. Also, Al is a conservative (very risk averse) investor, whereas Becky is an aggressive (not
too risk averse) investor.12 Finally, the riskfree rate rf in the economy, which can be obtained from
the yield on one-year government bonds, is currently 10%.
Let us rst suppose that the prices of the three securities are $8.77, $10.17, and $17.39 respectively.
It can be shown that this implies that the means, standard deviations and correlations of the three
assets rates of return are given in the following table:13
Firm
(i)
Share
Price (Pi )
Mean of
Return (ri )
Std. Dev. of
Return (i )
A
B
C
$8.77
$10.17
$17.39
14.00%
18.00%
15.00%
45.60%
59.00%
28.75%
Correlation (ij )
with B
with C
0.7
0.1
0.4
Now, given these numbers we can trace out the portfolio opportunity set (similar to that on
page I.4.57 in the lecture notes), nd the tangency portfolio of risky assets that will be chosen
by both Al and Becky, and nd the combination of that risky portfolio and the riskfree asset. This
is illustrated in Figure 4. In fact, the following table shows the number of shares of each stock that
Al and Becky would like to hold if the share prices were as above:
Firm
(i)
A
B
C
12
# of shares demanded
by Al
by Becky
Total
4,892
3,988
15,086
29,350
23,928
90,514
24,459
19,940
75,429
To be completely exact, they each have a utility function Uj (rp , p ) dened over the mean and standard deviation
of the portfolio that they each hold. Of course, since they prefer higher returns and smaller risk, we must have
Uj (rp ,p )
Uj (rp ,p )
a
> 0 and
< 0. In this particular example, we use Uj (rp , p ) = rp 2j p2 , where for Al aA = 1,
rp
p
and for Becky aB = 0.2.
13
To be exact, the means and standard deviations were obtained as follows:
mi
si
ri =
1, and i =
.
Si Pi
Si Pi
AD.13
rp
0.35
0.30
Becky
0.25
tangency
portfolio
0.20
B
0.15
Al
0.10
0.05
0.25
0.50
0.75
1.00
1.25
1.50
sp
Figure 4: This graph shows the portfolio opportunity set, the tangency portfolio, and the portfolios
chosen by Al and Becky, when the prices of the three securities are $8.77, $10.17, and $17.39
respectively.
Obviously, the total demand for securities A and B (29,350 and 23,928 shares respectively) is below
the number of shares oered by these two rms (40,000 and 50,000 shares). Similarly, the total
demand for security C (90,514 shares) is above the number of shares oered by that rm (40,000).
So, with these share prices ($8,77, $10.17, and $17.39), the market does not clear. In order to
increase (decrease) the demand for securities A and B (security C), they (it) will have to be made
more (less) attractive. This can be done by reducing the price of securities A and B, and increasing
that of security C. Of course, this will in turn aect the means and standard deviations of all three
securities rates of return, and therefore the shape of the portfolio opportunity set. This is shown in
Figure 5, using new security prices of $8.62, $10.01, and $17.67 respectively. In fact, the following
table shows the means and standard deviations of asset returns implied by these security prices:
Firm
(i)
Share
Price (Pi )
Mean of
Return (ri )
Std. Dev. of
Return (i )
A
B
C
$8.62
$10.01
$17.67
16.03%
19.83%
13.17%
46.41%
59.91%
28.29%
Correlation (ij )
with B
with C
0.7
0.1
0.4
Given these new prices, Al and Becky will certainly change their demands, as shown in Figure 6
AD.14
rp
0.25
new minimumvariance frontier
0.20
old minimumvariance frontier

0.15
0.10
0.05
0.2
0.4
0.6
0.8
1.0
sp
Figure 5: This graph shows the change in the portfolio opportunity set, after changing the prices
of the three securities from $8.77, $10.17, and $17.39 to $8.62, $10.01, and $17.67 respectively.
and in the following table:
Firm
(i)
A
B
C
by Al
by Becky
Total
6,667
8,333
6,667
40,000
50,000
40,000
33,333
41,667
33,333
It is now obvious that the market clears in the sense that the number of shares demanded by the
investors is exactly equal to the number of shares supplied/oered by the rms. This is what is
meant by equilibrium. The real-life situation that this equilibrium seeks to capture is the obvious
fact that, at all times, all the securities are held by all the investors. For that to happen, there has
to be correct prices for each security.
It can also be veried that the tangency portfolio in Figure 6 (which is the equivalent of Figure 4
with the new prices) is the market portfolio. Indeed, if we multiply the number of shares of each
security purchased by Al and Becky by their price, we nd the total amount of money that they
have invested in each asset. We can also nd the relative weight that they put on each security (in
AD.15
rp
0.35
0.30
Becky
0.25
tangency
portfolio
0.20
B
0.15
Al
C
0.10
0.05
0.2
0.4
0.6
0.8
1.0
1.2
1.4
sp
Figure 6: This graph shows the portfolio opportunity set, the tangency portfolio, and the portfolios
chosen by Al and Becky, when the prices of the three securities are $8.62, $10.01, and $17.67
respectively.
parentheses), as a total of their total risky portfolio.
Firm
(i)
A
B
C
by Al
by Becky
6,667
8,333
6,667
33,333
41,667
33,333
Share
Prices (Pi )
$8.62
$10.01
$17.67
Total
Wealth
by Al
$57,454
$83,455
$117,818
(22.21%)
(32.26%)
(45.54%)
$258,727
$500,000
$241,273
(lending)
in riskfree asset
Amount invested
by Becky
$287,273
$417,273
$589,091
(22.21%)
(32.26%)
(45.54%)
$1,293,637
$500,000
-$793,637
(borrowing)
Also, at the market-clearing prices, the total value of each rm is given in the following table, along
with the fraction of the total market that they account for.
Firm
(i)
Share
Price (Pi )
Outstanding
Shares (Si )
A
B
C
$8.62
$10.01
$17.67
40,000
50,000
40,000
Total
Total Market
Value (Pi Si )
Fraction of
Total Value
344,727
500,727
706,909
22.21%
32.26%
45.54%
1,552,363
100.00%
Clearly, these fractions correspond to the portfolios of risky securities purchased by Al and Becky.
AD.16
Now, what happens when some quantities are aected? For example, what happens if the riskfree
rate falls to 8%? The fact that the riskfree rate is smaller makes it less (more) attractive for lending
(borrowing). This means that both Al and Becky will change their demands for the risky assets
so that, at current prices, the market will not clear. In fact, it can be shown that security prices
will have to increase to $8.78, $10.20, and $18.00 for the market to clear again. Of course, this
again implies that the means and standard deviations of the asset returns will change, so that the
portfolio opportunity set is also aected. Also, with these new prices, it will be the case that the
total market value of each rm will change. However, as can be seen from the following table, it
will not aect the market portfolio:
Firm
(i)
Share
Price (Pi )
Outstanding
Shares (Si )
A
B
C
$8.78
$10.20
$18.00
40,000
50,000
40,000
Total
Total Market
Value (Pi Si )
Fraction of
Total Value
351,111
510,000
720,000
22.21%
32.26%
45.54%
1,581,111
100.00%
Risk Measure for an Ecient Portfolio
AD.17

Risk Measure for an Efficient Portfolio
In this handout, we show that p is indeed an appropriate measure of risk for any ecient portfolio p.
rm + (1 w)rf for some w 0. Therefore, since the
For any ecient portfolio, we have rp = w
riskfree rate is by denition riskfree,14 we have
rp , rm )
pm = Cov(
= Cov(w
rm + (1 w)rf , rm )
= wCov(
rm , rm )
= wVar(
rm )
2
.
= wm
(42)
We can now use (42) in the equation for p on page I.4.70:

p =
2
pm
wm
=
=w
2
2
m
m
This implies that the correct measure of risk for this ecient portfolio is
p m = wm .
If we can verify that p for an ecient portfolio is equal to this last expression, we will have the
desired result. Using the fact that the riskfree rate is riskfree once again, we have
2
rp ) = w2 Var(
rm ) = w 2 m
,
p2 Var(
and this implies that
p = wm .
14
this is why there is no tilde on rf in the preceding equation.
Olympia
AD.19

Olympia
As described on page I.4.103, Et = Et1 (1 + r k), so that the rate of growth in earnings is
r k = (15%)(0.6) = 9%.
We know that Dt = (1 k)Et and Dt1 = (1 k)Et1 , so that the rate of growth in dividends
is given by
Et
(1 k)Et
Dt
1=
1 = r k = 9%.
1=
Et1
Dt1
(1 k)Et1
From page I.4.101, we have
P0
1 PVGO
= +
,
E
r
E
where (using page I.4.105)

PVGO =
(0.15 0.12)(0.6)E
(r r)kE
=
= 5E.
r(r r k)
(0.12)[0.12 (0.15)(0.6)]
Using this last expression in (44), we obtain

P0
1
5E
1
=
+
=
+ 5 = 13.33.
E
0.12
E
0.12
We know that
P1 P0 D1
P1 P0 + D1
=
+
.
P0
P0
P0
So the portion of the returns coming from dividends is
1
D1
D1 E1
P
=
= (1 0.6)(13.33)1 = 3%.
= (1 k)
P0
E1 P
E
r=
This means that the portion of the returns coming from capital gains is
r 3% = 12% 3% = 9%.
We calculate P0 in three dierent ways:
P0 = E
P
E
= (3)(13.33) = 40 (i.e. $40 per share).

P0 =
=
=
=
=
D2
D1
D3
+
+
+
2
1 + r (1 + r)
(1 + r)3
D1
D1 (1 + r k) D1 (1 + r k)2
+
+
+
1+r
(1 + r)2
(1 + r)3
D1
r r k
(1 k)E1
r r k
(1 0.6)(3)
= 40.
0.12 0.15(0.6)
(44)
Olympia
AD.20

P0 =

Et It
(1 + r)t
t=1
E1 I1
E3 I3
E2 I2
+
+
+
2
1+r
(1 + r)
(1 + r)3
Since both the earnings and the reinvestments in the rm are growing at the same rate
r k = 9%, we have
P0 =
=
=
=
E1 I1 (E1 I1 )(1.09) (E1 I1 )(1.09)2

+
+
+
1+r
(1 + r)2
(1 + r)3
E1 I1
r 0.09
E1 kE1
r 0.09
3 (0.6)(3)
= 40.
0.12 0.09
Amalgamated Products
AD.21

Amalgamated Products
What is the cost of capital for each division?

First we need to calculate the asset betas for United Foods, General Electronics, and Associated Chemicals:
DUF
EUF
UF (debt) +
UF (equity) = 0.3(0) + 0.7(0.8) = 0.56;
DUF + EUF
DUF + EUF
DGE
EGE
GE (debt) +
GE (equity) = 0.2(0) + 0.8(1.6) = 1.28;
GE (assets) =
DGE + EGE
DGE + EGE
EAC
DAC
AC (equity) = 0.4(0) + 0.6(1.2) = 0.72.
AC (debt) +
AC (assets) =
DAC + EAC
DAC + EAC
UF (assets) =
The cost of capital for each division of Amalgamated can then be set equal to the cost of
capital for these rms, using the CAPM:
rA (food) = rUF (assets) = 0.07 + (0.15 0.07)(0.56) = 11.48%;
rA (electronics) = rGE (assets) = 0.07 + (0.15 0.07)(1.28) = 17.24%;
rA (chemicals) = rAC (assets) = 0.07 + (0.15 0.07)(0.72) = 12.76%.
What is Amalgamateds equity beta?
First, let us look at Amalgamateds balance sheet (along with the betas in parentheses):
Assets
50%:
30%:
20%:
Food
(0.56)
Electronics
(1.28)
Chemical
(0.72)
Liabilities
40%:
Debt
(0.20)
60%:
Equity
(A (equity))
Since the asset beta must be the same as the liability (nancing) beta, we must have:
0.5(0.56) + 0.3(1.28) + 0.2(0.72) = 0.4(0.2) + 0.6A (equity)
A (equity) = 1.213.
Constant Discount Rate
AD.22

Constant Discount Rate
The cash ows for this project look as follows:
1
0.6
-500,000
-4,000,000
700,000
700,000
0.4
Since the risk of the project warrants an expected return after year 2, it would be wrong to calculate
the projects net present value as follows:

4,000,000 700,000 1
NPV = 500,000 + 0.6
= 1,188,775.51.
+
(1.4)2
0.4 (1.4)2
Instead we need to discount all the cash ows after year 2 at 12%. At the end of year 2, these will
be worth
700,000
NPV2s = 4,000,000 +
= 1,833,333.33
0.12
if the project is successful. Otherwise they will be worth
NPV2u = 0.
We can therefore redraw the projects cash ows as follows:
0
2
0.6
-500,000
0.4
NPV2s = 1,833,333.33
NPV2u = 0
Now, the NPV is easily calculated using the 40% discount rate applicable to the rst two years:
NPV = 500,000 +
0.6(1,833,333.33)
= 61,224.49.
(1.4)2
Since this is positive, the project should be undertaken.
Dividend Survey
Baker and Powell, 1999
AD.23
Dividend Survey
The following table was extracted from How Corporate Managers View Dividend Policy, by
H. Kent Baker and Gary E. Powell, Quarterly Journal of Business and Economics, Spring 1999,
38, 17-35.
Dividend Survey
AD.24
Dividend Survey
AD.25
Dividend Survey
AD.26
Addition to Slides III.27
The Miles-Ezzell Formula General Derivation
AD.27

In this handout, we derive the adjusted cost of capital formula suggested by Miles and Ezzell (1980).
Suppose that we have a T -year project that requires an initial investment of C0 > 0, and that is
expected to generate cash ows of Ct in years t = 1, 2, . . . , T . The rm adopts a policy that it will
borrow a constant fraction L of the projects residual value during its whole life, i.e. it will keep
the ratio of the additional debt issued as a result the project to the present value of the project
constant.
Let Vt denote the present value of the project at the end of year t (or, equivalently, at the beginning
of year t + 1). At the beginning of year T , the debt will be set to LVT 1 , and the value of the
project is given by
CT
tc rD LVT 1
VT 1 =
.
(45)
+
1+r
1 + rD
Of course, we would like to do this calculation in one step, by adjusting the cost of capital:
VT 1 =
CT
.
1 + r
(46)
Solving for CT in (45) and (46), and setting the two results equal yields

1+r
= VT 1 (1 + r ),
VT 1 (1 + r) tc rD LVT 1
1 + rD
which implies
r = r tc r D L
1+r
1 + rD

.
Let us now move back one year, that is to the beginning of year T 1. The debt is then set to
LVT 2 , and the value of the project is given by
VT 2 =
CT 1
tc rD LVT 2
tc rD LVT 1
CT
+
+
+
,
1+r
(1 + r)2
1 + rD
(1 + r)(1 + rD )
where the expected tax shield coming in year T is discounted at r for one year to reect the fact
that the tax shield will only become known at the beginning of year T (i.e. one year from now).
Using (45), we can rewrite this as
VT 2 =
CT 1 + VT 1 tc rD LVT 2
CT 1 tc rD LVT 2 VT 1
+
=
+
+
.
1+r
1 + rD
1+r
1+r
1 + rD
(47)
Again, we would like to be able to calculate VT 2 in one step using an adjusted cost of capital:
VT 2 =
C2
CT 1
+
.
1 + r (1 + r )2
Given (46), we can rewrite this as

VT 2 =
CT 1 + VT 1
.
1 + r
(48)
We can now solve for CT 1 + VT 1 in (47) and (48), and set the two results equal to obtain

1+r
= VT 2 (1 + r ),
VT 2 (1 + r) tc rD LVT 2
1 + rD
which, once again, implies
r = r tc r D L
1+r
1 + rD
This process can be repeated all the way back to time 0.

.
AD.28

Additional Materials

Hochgeladen von

Dokumentinformationen

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Additional Materials

Hochgeladen von

Copyright:

Verfügbare Formate

Short Sales

This handout is a reproduction of section 2.4.2 (pages 25-30) of

Rules Governing Short Sales

Dividends, annual reports,

Figure 6: Before the short sale.

Notifie d that Mr. Jones

Figure 7: The short sale.

Figure 8: After the short sale.

Market Value of Assets Loan

Mathematics and Statistics

Mathematics and Statistics

(b) Solve y x = z for x (using natural logarithms only).

3. To simplify long equations, we will also make use of summation signs.

Mathematics and Statistics

(b) Take the (natural) logarithm on both sides of y x = z to obtain:

(b) This one can be calculated step by step:

= [1(1) + 2(1) + 3(1)] + [2(2) + 3(2)] + [3(3)]

Mathematics and Statistics

= [0 0.5][1 1.4](0.2) + [0 0.5][2 1.4](0.3)

First Method: Substitution

Now that we have found z, we can use it in (14) to obtain x:

Finally, we can use the values of x and z in (11) to obtain y:

Second Method: Diagonal Reduction

to interchange line 1 and

line 2 above. Then we write10

to subtract 2 times line 1 to line 2 of this last matrix. We then write11

The general goal of this method is to eventually obtain an identity matrix,

1. Reduce the numbers below the diagonal to zeros.

The solution to our original system of equations (8)-(10) is therefore given by x = 2, y = 1, z = 4.

Third Method: Matrix Inversion

be written in matrix form as

The solution is then given by

Figure 9: Matrix manipulations using Microsoft Excel.

Matrix Manipulations Using a Spreadsheet Software

A Few Additional Problems

log(xy) = log x + log y

Net present value

Compounding m times a year

Bond Prices and the Term Structure

Real interest rates

(xi X)2 p(xi )

Compounding and Discounting

Present value formula

Compounding m times a year

Net present value

Capital Budgeting Under Certainty

Internal rate of return

Portfolio Selection and the CAPM

Portfolio rate of return

Expected rate of return

Variance of rate of return

Capital Market Line (CML)

Minimum variance portfolio

Capital Budgeting Under Uncertainty

Cost of capital and leverage

Betas and leverage

Certainty equivalent formulas

DIVt DIVt1 = a (p EPSt DIVt1 )

Lintners dividend model

V = VU + PV (tax shield) PV (bankruptcy costs)

Interaction of Investment and Financing Decisions

The Valuation of Options

Bounds on call price

With dividends: S0 = S0 PV (dividends)