Financial Analytics (FIN 3134)

Spring 2022

Author: Bradley Paye, Pamplin College of Business, Virginia Tech

Date: January 14, 2022
Version: 2.4
 Contents

1 Time Value of Money 1

1.1 The Key Time Value Idea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Interest Rates: The Basics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 The Fundamental Time Value Equation . . . . . . . . . . . . . . . . . . . . . 6
1.4 Multiple Cash Flows: The Brute Force Approach . . . . . . . . . . . . . . . . 11
1.5 Perpetuities and Annuities . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.6 Solving TVM Problems with a Financial Calculator or Excel . . . . . . . . . . 21
1.7 Tricks of the Trade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Chapter 1 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

2 TVM: Advanced Concepts 34

2.1 Growing Cash Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.2 Alternative Compounding Frequencies and the True Cost of Borrowing . . . . 38
2.3 Loans and Amortization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.4 Financial Planning and Retirement Problems . . . . . . . . . . . . . . . . . . . 48
Chapter 2 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

3 Interest Rates and Bonds 58

3.1 Interest rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.2 Bonds: Key Institutional Details . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.3 Bond Prices and Yields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
3.4 Holding Periods Returns and Yields . . . . . . . . . . . . . . . . . . . . . . . 79
3.5 Bonds and Interest Rate Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.6 Special bonds: STRIPS and TIPS . . . . . . . . . . . . . . . . . . . . . . . . 91
3.7 The Yield Curve and Term Structure . . . . . . . . . . . . . . . . . . . . . . . 95
3.8 Corporate Bonds and the Structure of Interest Rates . . . . . . . . . . . . . . . 100
3.9 Appendix: Cyclical Impacts on the Supply of and Demand for Loanable Funds 106
Chapter 3 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110

4 Federal Reserve and Monetary Policy 116

4.1 Overview and Readings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
Chapter 4 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

5 Equity and Risk-Return Tradeoffs 118

5.1 Equity: Basic Institutional Details . . . . . . . . . . . . . . . . . . . . . . . . 118
CONTENTS – ii –

5.2 Stock Returns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121

5.3 Stock Returns: Probability Model and Key Statistics . . . . . . . . . . . . . . . 124
5.4 Risk-free returns, risk premiums, and Sharpe ratios . . . . . . . . . . . . . . . 130
5.5 Measuring Characteristics of Risky Asset Returns . . . . . . . . . . . . . . . . 133
Chapter 5 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138

6 Portfolios, Diversification, and the CAPM 142

6.1 Portfolios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
6.2 One risky asset and a risk-free asset . . . . . . . . . . . . . . . . . . . . . . . 144
6.3 Portfolios of two risky assets . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
6.4 More Than Two Risky Assets . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
6.5 Systematic and Idiosyncratic Return Components . . . . . . . . . . . . . . . . 157
6.6 The CAPM and Beta as a Systematic Risk Measure . . . . . . . . . . . . . . . 164
6.7 Weaknesses of CAPM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
Chapter 6 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172

A Mathematical Details 176

A.1 Summation Operator, Infinite Series, and Proofs of the Perpetuity and Annuity
PV Formulas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
A.2 Review of Logarithms and Exponentials . . . . . . . . . . . . . . . . . . . . . 177
A.3 Probability and Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
 Chapter 1 Time Value of Money

1.1 The Key Time Value Idea

What is $100 worth? Seems like a really simple question, right? A dream exam question!
$100 is worth $100. You would be a fool, or more generously an extremely charitable individual,
if you were to hand someone $150, or $125, or even $101, in exchange for $100. And that person
would be equally foolish to accept $90 or even $99 in exchange for their $100 bill. So, fine –
$100 is worth $100...right now.
Wait...what does the ‘right now’ qualifier mean? It means that the simple and obvious
answer to the question makes a subtle assumption that the exchange takes place at the current
time. Does this matter? It depends. If the exchange of dollars occurs simultaneously, then $100
is still worth $100, irrespective of the particular point in time when the exchange takes place.
For example, if you and I agree to meet in six years and negotiate the value of $100, then in six
year’s time we will agree that $100 is worth $100, just as we did today.
But what if the monetary amounts exchanged occur at different times? Suppose, for example,
that we negotiate an exchange in which I hand you a certain amount today, and you agree to hand
me $100 in exactly one year. This distinction concerning timing matters. In fact, this particular
exchange amounts to a loan agreement. I am effectively agreeing to loan you a certain amount
of money today, and you are agreeing to repay that loan by giving me $100 in one year’s time.
So, what amount should I be willing to lend you today in return for $100 in one year?
The answer suddenly does not seem obvious. In order to make progress, let’s address a slightly
different question: What is the absolute maximum I should be willing to consider lending you
today? The first key insight is that I should rule out amounts equal to $100 or greater, because I
am at least as well off by simply holding onto my cash for the next year. This simple reasoning
captures the key idea of the chapter:
Definition 1.1. Time Value of Money
The time value of money refers to the concept that a certain amount of money received
today is equivalent to a greater amount of money received at some future date. Similarly,
a certain amount of money received at some point in the future is equivalent to a lesser
amount of money received today. Simply put: $1 received in the future is worth less than
$1 today, and $1 today is worth more than $1 in the future.
♣

We have reasoned that I should be unwilling to lend more than $100 today for your promise
to repay me $100 in one year’s time. But exactly what amount should I lend? To arrive at a
quantitative answer, we need to make additional assumptions. One assumption involves how
1.1 The Key Time Value Idea –2–

trustworthy I perceive you to be. Will you actually show up in one year with the promised cash?
If I am concerned about this at all, then I should view your promise of future cash as risky.
Assuming that I am averse to risk, all else equal, the riskier I perceive your promise of $100 to
be, the less I should be willing to pay you today.
Many common ‘financial securities’ such as common stock, corporate bonds, etc. involve
risky cash flows. Eventually, we will analyze how the risk associated with future cash flows affects
their value. In this chapter; however, we will avoid complexities associated with risk by making
the assumption that the promised cash flows are nominally safe. This bit of jargon means that all
dollar-denominated future cash flows will be received and/or paid exactly as per contracted terms.
In our example, we are assuming that I am 100% certain that you will deliver exactly $100 in
exactly one year’s time. Although this assumption might seem unrealistic, several very important
financial securities are ‘essentially’ free of risk in this sense. For example, Treasury Bills issued
by the United States government contractually promise certain cash amounts at specified future
dates and are backed by the government’s ability to print new currency.1
Assuming that the loan repayment of $100 in one year’s time is nominally safe, what is
the maximum I should be willing to lend today? Let’s assume that banks are currently paying
an interest rate of 5% per year on deposits. The next section discusses interest rates in further
detail. For the moment we need only understand that each $1 deposited in a bank account today
for one year will grow to a value $1.05 after one year’s time. We’ll also assume that bank deposits
are nominally safe.2 Under these assumptions, I can guarantee myself $100 in exactly one year
by depositing $95.24 today, because

$95.24 + $95.24 × 5% = $95.24 × (1.05) = $100. (1.1)

So, we have reasoned that I should not be willing to lend more than $95.24 today in return for a
repayment of $100 in one year, because otherwise I would be better off depositing my money in
a bank account.
Now, let’s take your perspective with respect to the loan. Should you be willing to accept
any less than $95.24 today with the obligation to repay $100 in one year? Assuming that you are
able to borrow funds from a bank at the same interest rate of 5%, then the answer is “no.”
Finally, then, we have obtained a quantitative answer: $100 received in one year’s time is
equivalent to $95.24 today. Equivalently, the present value of $100 in one year’s time is $95.24.

1The ‘essentially’ qualifier refers to the fact that, despite the fact that sovereign governments such as the United
States, have the ability to print currency in order to satisfy (own-currency denominated) debt or bond obliga-
tions, they might nevertheless choose not to do so. This is termed strategic default and is costly to the gov-
ernment in the sense that default raises the required interest cost of future debt issuance. A number of coun-
tries have never defaulted in modern times. As a recent example, Lebanon defaulted on approximately $1.2bil-
lion in ‘Eurobonds’ earlier this year (https://www.economist.com/middle-east-and-africa/2020/03/12/
for-the-first-time-lebanon-defaults-on-its-debts).
2This is a reasonable assumption with respect to the United States banking system, because bank deposits are protected
against losses via mandatory insurance banks obtain via the Federal Deposit Insurance Corporation (FDIC).
1.2 Interest Rates: The Basics –3–

Definition 1.2. Present Value

The cash value today of a cash flow or a series of cash flows received at specified dates in
the future.
♣

1.2 Interest Rates: The Basics

In the previous section, we computed the appropriate present value of $100 received one
year in the future by assuming a prevailing interest rate on bank deposits. The interest rate equals
the amount charged above the principal (the amount lent) by a lender to a borrower for the use of
assets. For example, a depositor plays the role of the lender via ‘depositing’ funds with the bank.
The bank pays a stated interest rate for the use of these funds. Alternatively, the bank plays the
role of lender when it issues credit cards to its customers, and charges a stated rate of interest on
card balances.
It is quite likely that you have encountered various forms of interest rates before taking this
course. For example, you might have a checking or savings account (or both) at a bank that pays
a specified interest rate on deposits. You might also have a credit card that you use for purchases.
Virtually all credit cards charge a specified interest rate on unpaid card balances.

Remark Do you know what interest rate is charged on your credit card(s)? If not, you should
find out by ‘checking the fine print’ in your card agreement. You might be in for a surprise.
Credit card interest rates are often much higher than the interest rates banks pay on checking or
savings accounts. For example, the interest rate on one of my credit cards is approximately 18%
per year. This is vastly higher than the roughly 1% per year rate I earn on a savings account I
have at a local bank. Needless to say, credit card issuers typically do not focus marketing efforts
on the (high) interest rates they charge. Instead, they focus on incentives such as ‘triple points
on travel expenses,’ etc. I pay off the balance on my credit card every month in order to avoid
building up significant interest charges. I advise you to do the same, recognizing that certain
unusual circumstances might necessitate borrowing on a line of credit.
In this chapter, we will typically assume that there is a single, specified interest rate per
period that we will denote as r. In the context of a bank account, the interest rate r specifies how
much your initial deposit, or principal, grows in value over a single period. For now, we will
interpret r as the rate of interest per year and consider only cases in which deposits, withdrawals,
loan payments, or other cash flows occur on an annual basis.3 Each $1 deposited today grows in
value to $1 × (1 + r) at the end of the year.
A number of common financial terms closely relate to an interest rate. One is the ‘rate of
return’, or just ‘return’:

3In ‘real life,’ cash flows often occur at other frequencies, such as regular monthly payments on a home loan. Later,
we consider cases with non-annual cash flows.
 1.2 Interest Rates: The Basics –4–

Definition 1.3. Rate of Return

The rate of return, or simply ‘return’ (Ret), is defined as the increment in the value of an
arbitrary asset (stock, bond, real estate, etc.), divided by the initial value, converted to
percentage terms. Mathematically:
New Value − Initial Value
Rate of Return = r = × 100 (1.2)
Initial Value ♣

We can see the connection between the notion of an interest rate and a rate of return by
re-arranging the expression for the growth of a savings account deposit over a single period:

New Account Balance = Initial Account Balance × (1 + r)

New Account Balance
r= −1
Initial Account Balance
New Account Balance − Initial Account Balance
r= .
Initial Account Balance
The final expression is in exactly the same form as the rate of return equation, up to the conversion
to a percentage. The specific term ‘interest rate’ is typically used in conjunction with lending
arrangements. The rate of return is more general, in the sense that it is synonymous with the
interest rate in standard lending arrangements, but applies in general to any situation in which
an investor holds some asset (stocks, bonds, foreign currency, real estate, Bitcoin, and so forth).
Many of these assets involve rates of return that are unknown as of the starting time. For example,
if I acquire $1, 000 of gold today, hold this for one year, and then sell the gold, the market price for
gold in one year is not known with certainty. This is an example of a risky investment involving
uncertain future cash flows. This chapter focuses on scenarios where such risk is not present,
such as the context of depositing funds in a savings account at a bank.4
The following introduces a very simple concrete example of interest rates in action:

Example 1.1 Question: You deposit $100 today at a local bank in an account that earns 5%
interest per year. Assuming you do not withdrawal any funds over the course of the ensuing year,
what is your end-of-year account balance?
Answer: You earn an amount of interest equal to 5% of the principal of $100 over the course of
the year. This is $5. Your end-of-year account balance is therefore F V = $100 + $5 = $105.

The notation F V in the previous example refers to future value:

Definition 1.4. Future Value
The value of a single deposit or a series of deposits or cash flows at some point in the
future, where it is assumed that deposited funds earn a specified rate of interest.
♣

When funds are invested for more than one period, then interest compounds. This refers to
the fact that the interest earned in subsequent periods (years) equals the interest rate times the
account balance at the beginning of the period, which includes interest earned in earlier periods.

4We will use the notation r to represent either the interest rate or rate of return on an investment in this section. Some
other texts use i instead.
1.2 Interest Rates: The Basics –5–

The essence of compound interest is that the depositor earns interest-on-interest, i.e., interest is
earned on re-deposited or re-invested previous interest payments.
Definition 1.5. Compound Interest
Interest earned on both the initial principal and the interest re-deposited or re-invested
from earlier periods.
♣

In order for interest to compound, it is critical that earned interest is re-deposited or re-
invested. Suppose, for example, that an initial $100 deposit earns 5% interest per year. Then, as
we reasoned above, your account will earn $5 of interest after one year. But suppose that you
elect to withdraw this $5 at the beginning of the second year. Now, in the second year, because
your beginning year 2 account balance is only $100 rather than $105, you will earn only $5
interest in the second year, rather than the amount of $105 + $105 × 5% = $5.50 had you not
withdrawn the first year’s interest of $5 from the account. The convention of earning interest
only on the initial principal amount (due to the fact that earlier interest is not re-invested), is
called simple interest:
Definition 1.6. Simple Interest
Interest earned only on the initial principal amount because earned interest is not re-
deposited or re-invested.
♣

Figure 1.1 shows how account balance growth differs under compound interest versus simple
interest. The tabulation tracks account balances over years 1–5 for an initial deposit of $100
assuming that the interest rate is 5% per year. The table shows both the compound interest (that
includes interest on the initial principal as well as all previous accumulated interest) and simple
interest. It also tracks the evolution of the account balance under two scenarios: 1) assuming
no withdrawals so that the account balance reflects compound interest; and 2) assuming that the
account only earns simple interest. (The second case does not reflect reality in most cases but is
presented as a comparison point.)

Figure 1.1: Tabulation: Compound Versus Simple Interest

In the above example, the difference between the simple interest earned in year 2 ($5) and
the compound interest earned by re-investing the year 1 interest ($5.25) is only $0.25. This might
lend the impression that there is only a minor difference between compound and simple interest.
However, this impression is absolutely incorrect, especially as the time horizon of investment
1.3 The Fundamental Time Value Equation –6–

increases. This is because the account balance grows linearly under simpler interest, whereas
under compound interest the account balance grows exponentially. Figure 1.2 illustrates the
account balance growth in our example setting over 100 years. Initially, there is relatively little
difference in the end-of-year balance under compound versus simple interest. However, due to
the exponential growth under compound interest, there is a tremendous difference in the account
balances after 100 years.

Figure 1.2: Account Balance Growth: Compound Versus Simple Interest

There is a close relation between the interest rate r the basic ‘time value of money’ notion.
Recall that the time value notion boils down to the idea that $1 today is worth more than $1 in
the future. Therefore an alternative way to express this idea is that the interest rate is positive
(r > 0). Although we will not delve into the economic determinants of the interest rate until later
in the course, it is worth mentioning that a positive interest rate is consistent with the sensible
economic idea that consumers and households exhibit a time preference in consumption such
that consumption today is preferred to an equal amount of consumption deferred to the future,
all else equal.
Remark The time value of money idea, or equivalently the notion that interest rates should be
positive, hopefully seems natural. Note, however, that the time value of money concept only
suggests a positive interest rate and does not pin down a particular level for the interest rate. In
fact, current interest rates in the United States and many other developed countries are extremely
low – close to zero. In addition, some central banks have pursued so-called “negative interest
rate” policies designed to help spur lending by banks. We will discuss the current interest rate
environment and aspects of monetary policy later in the course.

1.3 The Fundamental Time Value Equation

This section introduces some basic, but crucial, computational formulas. In reality, there
is only one key equation involved, but by rearranging the equation we can answer a variety of
different types of questions. Because the key equation describes how the balance of a savings
1.3 The Fundamental Time Value Equation –7–

account evolves over time, I loosely call this ‘basic bank account math’. Let’s work with a
concrete example to get started:

Example 1.2 You deposit $100 today in a savings account at your local bank. The bank pays an
interest rate on such accounts of 3% per year. Assuming that the $100 was your initial deposit,
that you make no additional deposits in the future, and that there are no periodic bank fees,
compute the answers to the following three questions:
1. What will be the value of your savings account in one year?
2. What will be the value of your savings account in two years?
3. What will be the value of your savings account in N years, where N is an arbitrary number
of years?

The answers to the questions posed in Example 1.2 are as follows:

1. After one year, you will earn 3% interest on your initial deposit, and your account value
will therefore equal:

F V (1) = $100 + 3% × $100 = $100 × (1.03) = $103

2. To determine your account balance after two years, note that after the second year you
will earn an additional 3% interest on your beginning-of-year balance of $103 (from the
previous calculation). Therefore, your end of year two balance equals:

F V (2) = $103 + (3% × $103)

= $103 × (1.03)
= $100 × (1.03) × (1.03)
= $100 × (1.03)2
= $106.09

3. Extending the reasoning from the previous example, after N years your account value will
equal:
F V (N ) = $100 × (1.03)N .

As a concrete example, after 10 years your account balance will equal F V (10) = $100 ×
(1.03)10 = 134.39.
The previous example can be generalized easily to compute the future account value for an
arbitrary starting deposit amount of P V (present value) and an arbitrary rate of interest r paid
per period. The following result gives the resulting ‘future value formula:’
1.3 The Fundamental Time Value Equation –8–

Theorem 1.1. Future Value Equation

The following equation describes the future value of an amount P V deposited today in a
bank account for N periods assuming no withdrawals and that the rate of interest paid
equals r percent per period:

F VN = P V × (1 + r)N (1.3)
♡

The future value equation addresses how an amount deposited today, denoted in the formula
as P V , grows in value over time as interest accumulates. Notice that the formula assumes that
there are no withdrawals between today and the end of the deposit horizon N , so that interest
compounds.

Example 1.3 Exactly 20 years ago, your grandmother made a one-time deposit of $10,000 into
an account in your name in order to help fund your college education. Assuming that the annual
(compound) interest rate on deposits is 4%, what is the account balance today?
Solution: Applying the future value formula:

F VN = P V × (1 + r)N
= $10, 000 × (1.04)20
= $21, 911.23

So, you have just over $21,911 in the account today.

1.3.1 The Present Value Equation

A closely related problem to determining the future value of a deposit is to determine today’s
dollar equivalent to a certain amount of cash received at a specified time in the future. In fact,
this is exactly the type of question we answered in Section 1.1 when considering the amount that
should be loaned today in return for a repayment of $100 in one year’s time. To solve general
problems of this nature, we can simply re-arrange the future value equation so that P V is on the
left-hand side:
Theorem 1.2. Present Value Equation
The following equation describes the present value (or today’s value) of an amount F V
received N periods in the future, assuming that the compound rate of interest paid equals
r percent per period:
F VN
PV = (1.4)
(1 + r)N
♡

The following example illustrates the use of the formula:

Example 1.4 I offer to enter into a financial contract with you obligating me to pay you $5,000
in 3 years, in return for some amount of cash today. What is the maximum amount of cash you
would be willing to pay me today under this contract, assuming the interest rate is 8%?
1.3 The Fundamental Time Value Equation –9–

Solution: Applying the present value formula:

F VN
PV =
(1 + r)N
$5, 000
=
(1.08)3
= $3, 969.16

You would therefore be willing to ‘lend’ around $3,969 for three years under this contract.
It is important to understand that the future value and present value equations ultimately
represent a single relation between current and future values. Calling it the ‘future value equation’
simply means that we are choosing to write the equation with F V on the left-hand side, whereas
calling it the ‘present value equation’ implies that we are choosing to write the equation with
P V on the left-hand side.

1.3.2 Discount Factors and the Discount Rate

Here we pause to introduce a bit of important jargon. To do so, we will re-write the basic
PV equation in the following form:
1
PV = F VN ,
(1 + r)N
| {z }
DF (N )
where F VN is a future amount of cash in N periods. The underbrace emphasizes that the present
value is written as the product of two terms: 1) a ‘discount factor for N periods’ (DF (N ));
and 2) the future value received after N periods. The terminology comes from the fact that the
DF (N ) ‘discounts’ $1 received in N periods to an equivalent value today. This gives rise to
the term discount rate as another synonym for the interest rate r. For example, if you are asked
“What is the value of $1 received 6 years from now if the discount rate is 5%?”, then the answer
equals the corresponding discount factor 1/(1.05)6 = 0.746215. Once you have this discount
factor, then it is easy to solve for the present value of any other sum of money received in 6 years’
time. For example, if you are then asked what is today’s value of receiving $200 in 6 years, the
answer is:
DF (6) × $200 = 0.746215 × $200 = $149.24.

1.3.3 Other Variations on the Basic Equation

There are in fact four variables involved in (either) equation: F V , P V, r, and N . We can
re-arrange the equation so that any of these quantities are on the left-hand side. Suppose, for
example, that given a present value (P V ), future (F V ), and number of periods N associated
with this future value, we want to determine the implied rate of interest (r). To do so, we can
1.3 The Fundamental Time Value Equation – 10 –

rearrange Eq. (1.3) so that r appears (alone) on the left-hand side:

F VN = P V × (1 + r)N
F VN
(1 + r)N =
PV
 
F VN 1/N
1+r =
PV
 
F VN 1/N
r= −1
PV
Consider the following example:

Example 1.5 Question: You invest $5,000 today and will have $7,401.22 after 10 years. What
is the corresponding implied annual interest rate or ‘rate of return’ on this investment?
Answer:
 1/N
F VN
r= −1
PV
 1/10
$7, 401.22
r= −1
$5, 000
r = 4.0%

The final variation of problems of this variety involves solving for N , the investment, loan, or
deposit horizon, given the other three formula inputs. Mathematically, solving for N analytically
requires taking a logarithmic transform:

F VN = P V × (1 + r)N
F VN
(1 + r)N =
PV
N ln(1 + r) = ln(F VN ) − ln(P V )
ln(F VN ) − ln(P V )
N= ,
ln(1 + r)
where ln(·) denote the natural logarithm function and the third line uses the fact that ln(A/B) =
ln(A) − ln(B). Let’s apply this result in the context of a concrete problem:

Example 1.6 If you invest at 5%, how long will it take to double your money?
Answer: The trick is to recognize that doubling your money implies that F V = 2P V . Plug this
into the above equation for N :
ln(F VN ) − ln(P V )
N=
ln(1 + r)
ln(2P V ) − ln(P V )
N=
ln(1.05)
ln(2)
N=
ln(1.05)
N ≈ 14.21 years

At this point, it would be a good idea to work the first 6 problems at the end of the chapter,
1.4 Multiple Cash Flows: The Brute Force Approach – 11 –

all of which involve manipulating the basic time value of money equation.

1.4 Multiple Cash Flows: The Brute Force Approach

The basic time value of money equation (in its various forms) assumes a single future cash
flow. But often we want to compute present or future values associated with multiple future
cash flows. The question is how to do this effectively and efficiently. In certain cases, the
structure of the underlying cash flows is such that relatively simple formulas exist for computing
the valuations of interest. We will consider these cases in the next section. Here, we introduce
a ‘brute force’ approach. The main attribute of this approach is generality. Let us consider an
example to motivate the approach.
Example 1.7 Suppose that you will receive three future cash flows as follows: 1) $1,000 one
year from today; 2) $5,000 three years from today; and 3) $12,000 seven years from today. An
associate offers to give you a certain amount of cash today in exchange for the rights to these
three cash flows. If the prevailing interest rate is 6% per year, what the present value of the three
future cash flows?
Answer: The trick to answering this question is to recognize that we can separately compute a
present value for each of the cash flows involved using the basic time value equation from the
previous section (with PV on the left-hand side). Given these computations, the present value of
all three future cash flows is simply the sum of the present values of the individual cash flows.
For the first cash flow of $1,000 in one year, we have:
F VN
PV =
(1 + r)N
$1, 000
PV =
(1.06)1
P V = $943.40

For the second cash flow of $5,000 in three years’ time, we compute:
F VN
PV =
(1 + r)N
$5, 000
PV =
(1.06)3
P V = $4, 198.20

And for the final cash flow of $12,000 in seven years’ time:
F VN
PV =
(1 + r)N
$12, 000
PV =
(1.06)7
P V = $7, 980.69

Finally, to compute the present value of all three future cash flows, we simply add the three
1.4 Multiple Cash Flows: The Brute Force Approach – 12 –

individual PVs:

P V (total) = P V (CF #1) + P V (CF #2) + P V (CF #3)

P V (total) = $943.40 + $4, 198.20 + $7, 980.69
P V (total) = $13, 122.18

The solution is tedious, as we apply the basic time value formula over and over to value each
individual cash flow. However, it is conceptually very simple (at least I hope it is!). A general
statement of the method is as follows:
Theorem 1.3. PV of Multiple Cash Flows
The present value of an arbitrary set of future cash flows is equal to the sum of the present
values of set of cash flows:

P V (CF #1, CF #2, . . . , CF #K) = P V (CF #1) + · · · + P V (CF #K), (1.5)

where K is a positive integer equal to the total number of (nonzero) cash flows. The PV
of each individual cash flow can be found using the basic time value equation.
♡

There is also a ‘future value’ version of this result. This pertains to cases where an investor
makes a set of deposits into a bank account paying a constant rate of interest (or similar contexts).
The deposits can be for arbitrary amounts and can occur at different times. Assuming no
deposited funds are withdrawn, we can compute the total account balance at some specified
date by summing the future values of each of the deposits at this date. The following result
summarizes the method:
Theorem 1.4. FV of Multiple Cash Flows
Let CF1 . . . CFK denote K (a positive integer) arbitrary cash deposits or investments
earning a (common) fixed rate of interest of r per period. Consider an arbitrary future
reference date and let N1 . . . NK indicate the number of periods over which each cash flow
is invested until this future date, assuming no intermediate withdrawals. Then the total
future of the deposits/investments equals the sum of the future values of each individual
deposit/investment:

F V (CF1 , . . . , CFK ) = F V (CF1 , N1 ) + · · · + F V (CFK , NK ), (1.6)

where the notation emphasizes that the number of periods each deposit is invested generally
differs. The FV of each individual deposit (or investment) can be found using the basic
time value equation.
♡

The following example illustrates an application of the method in a relative simple case with
only two deposits:
Example 1.8 Today you open a savings at your local bank by making an initial deposit of $5,000.
The account pays an interest rate of 3% per year. You anticipate making a second deposit of
$8,000 to the account three years from today. Assuming that you do not make additional deposits
1.5 Perpetuities and Annuities – 13 –

nor withdraw any funds from the account, what will be the account balance in 8 years’ time?
Answer: The account balance in 8 years’ time will equal the sum of the future value of our initial
deposit in 8 years plus the future value 8 years from now of the second deposit we will make
three years from today. It is important to note that, 8 years from today, the initial deposit will
have grown at 3% interest for 8 years, whereas the second deposit will have grown at 3% interest
for only 5 years. The future value of the initial deposit equals:

F VN = P V × (1 + r)N
F VN = $5, 000 × (1.03)8
F VN = $6, 333.85

The future value in 8 years of the second deposit, made three years from now, equals:

F VN = P V × (1 + r)N
F VN = $8, 000 × (1.03)5
F VN = $9, 274.19

Finally, the total account balance will then equal the sum of the above two future values:

F V (total) = $6, 333.85 + $9, 274.19 = $15, 608.04

Cases with additional deposits or investments can be handled in the same fashion as in the
previous example. The computations are tedious, particularly if done by hand as opposed to in
a spreadsheet program such as Excel, but cases with many deposits are conceptually no more
difficult than the previous example with two deposits.

1.5 Perpetuities and Annuities

Many real-life financial contracts involve multiple cash flows with relatively simple struc-
tures. Consider, for example, a standard 30-year mortgage loan on a home. Under this form
of loan contract, the borrower repays the principal of the home loan over a 30-year period via
a series of fixed (in dollar terms) monthly payments. Later, we will study in detail how the
series of fixed payments gradually draws down the remaining principal (this process is known
as ‘amortization’), but for now the main point is simply that the loan repayment involves a large
number (360) of equal cash amounts.
Let us analyze a variation on this type of loan contract that better fits for our convention
thus far of focusing on annual cash flows. Suppose in particular that a 30-year home loan is
paid off by 30 equal annual payments. The first payment occurs one year from the time of the
loan, and the final payment occurs in 30 years’ time. Consider the following problem: what is
the present value or amount that can be borrowed under this loan structure if the interest rate
is 5% per year and the annual loan payments are $10,000 each? We will answer this question
numerically later. For now, I want you to notice that you could answer the question using the
1.5 Perpetuities and Annuities – 14 –

‘brute-force’ approach of the previous section. This would involve separately computing the PV
of each of the 30 payments involved and then adding all of these PVs together. This will give you
the correct answer, but it is very tedious. In fact, if this were an exam question and you attempt
to solve it in this way, you will likely spend an excessive amount of time on the question and fail
to complete much of the rest of the exam.
Fortunately, there is another way! It turns out that there are relatively simple, convenient
formulas for computing present values, future values, and related quantities when the cash flows
involved are constant dollar amounts. This section introduces you to these formulas, which
are critical precisely because many, many real-world financial contracts involve constant dollar
payments. These include not only standard amortizing home loan contracts referenced above,
but also important financial securities including government and corporate bonds, which we will
study later in detail.

1.5.1 Perpetuities

The first special case we will analyze has what might seem to be a highly unusual structure.
A ‘perpetuity’ involves a series of fixed, regular payments that go on forever:
Definition 1.7. Perpetuity
Regular cash flows of a fixed dollar amount that contractually continue forever with no
specified terminal date.
♣

A perpetuity involves an infinite number of cash flows. The first key question is whether
such a security has a finite present value. Initially, it might seem that it will not. After all, a
perpetuity entitles the holder to an infinite amount of cash. Critically, however, it entitles the
holder to only a finite amount of cash (the regular payment amount $C) at any particular time.
The specific timing of cash flows is important, because this effects present value. The perpetuity
present value equation assumes that the first cash flow or payment occurs one period from today.
The timing is illustrated in Figure 1.3. The figure emphasizes that the initial or 1st cash flow
(CF) occurs one period from ‘now’ or time 0 on the timeline. Also, the arrow at the end of the
timeline emphasizes that the sequence of cash flows never stops.

Figure 1.3: Cash flow timing for a standard perpetuity

As long as the discount rate is positive, this turns out to be enough to guarantee that the
1.5 Perpetuities and Annuities – 15 –

N PV(N)
1 $95.24
5 $78.35
10 $61.39
100 $0.76
200 $0.01
Table 1.1: PV of $100 in N periods assuming r = 5%

perpetuity has a well-defined, finite present value. For some intuition, consider a contract offering
an annual perpetuity of $100, with an interest rate of r = 5%. Let’s apply the basic PV formula
to computed the present value of each cash flow in the perpetuity, as a function of the number of
years until we receive it (N ). The basic PV formula says that this is:

P V (N ) = C/(1 + r)N
= $100/(1.05)N ,

where the second line plus in the $100 regular cash flow amount and the assumed interest rate.
Table 1.1 evaluates this formula to determine the PV of specific cash flows promised by the
perpetuity for different waiting periods N . Today’s value of the first $100 received under the
perpetuity contract equals about $95.24. The 10-th $100 payment, in contrast, has a present value
of only $61.39. The 200-th payment of $100 has a PV of only one cent, and of course subsequent
payments will have even lower present value. This illustrates the mathematical reason that a
perpetuity has a finite present value despite offering an infinite number of future cash flows: the
present value of the future cash flows decrease to zero at an ‘exponential’ rate (the exponent
is N in the PV formula). Conceptually, we could therefore compute the present value of the
perpetuity to a very close approximation using the ‘brute’ force approach by simply adding the
PVs of payments up to payment #200. This is very tedious (although less tedious in Excel),
though. Fortunately, it turns out that there is a very simple formula that allows us to easily
compute the present value.
Theorem 1.5. Present Value of a Perpetuity
The present value of a perpetuity paying a regular, fixed dollar cash flow C, with the first
cash flow occurring in one period, is given by the following expression, with r denoting
the per-period interest rate:
C
PV = (1.7)
r ♡

The following examples apply the formula to solve practical problems:

Example 1.9 Preferred stock is a form of equity that pays investors a specified, fixed-dividend
amount. Suppose you own 10 shares of Pacific Gas & Electric preferred stock. Each share pays
$10 to the holder per year. For simplicity, assume that the next dividend payment is exactly one
year from now, and that Pacific Gas & Electric is expected to continue in operation forever and
will always make the $10 preferred stock payment in timely fashion. If the discount rate is 4%
1.5 Perpetuities and Annuities – 16 –

per year, what is a fair market price for a share of Pacific’s preferred stock?
Solution: A fair market price should equal the present value of all of the future dividend payments
that an investor would be entitled to by buying a share today. Under our assumptions, this takes
the form of a $10 perpetuity. So, we can compute the value of a share using the formula for the
PV of a perpetuity:
P V = C/r = $10/(0.04) = $250. (1.8)

So the preferred stock should trade for $250 per share.

Example 1.10 NewTech Inc.’s preferred stock pays a $20 dividend once per year. Suppose that
the most recent dividend has just been paid, and the current price (post-dividend) is $400 per
share. What is the annual rate of return associated with buying one of NewTech’s preferred
shares, assuming that all promised dividends will be paid in timely fashion with certainty?
Solution In this case we are given the price of a preferred share, which should equal it’s present
value (PV), as well as the annual corresponding cash flow amount (C). Therefore, we can re-
arrange the equation for the PV of a perpetuity so that the rate of return r appears on the left-hand
side, and then obtain the solution:

P V = C/r
r = C/P V
r = $20/$400
r = 5%.

1.5.2 Annuities and PV of an Annuity

The cash flow streams in a perpetuity continue without end. A far more common structure
for financial securities involve a specified number of fixed dollar (or other currency) cash flows.
This cash flow structure is called an ‘annuity’:
Definition 1.8. Annuity
Regular cash flows of a fixed dollar amount (or other currency amount) that contractually
continue for a specified number (N ) of periods.
♣

Similar to the case of perpetuities, there is a relatively simple formula for computing the
present value of an annuity, so that we do not need to rely on the ‘brute force’ approach:
Theorem 1.6. Present Value of a Annuity
The present value of an annuity paying a regular, fixed dollar cash flow C for N periods,
with the first cash flow occurring in one period, is given by the following expression, with
r denoting the per-period interest rate:
 
C 1
PV = 1− (1.9)
r (1 + r)N
♡
1.5 Perpetuities and Annuities – 17 –

As with the perpetuity case, the PV formula above assumes the ‘standard structure’ for an
annuity, such that the first cash flow occurs one period in the future from ‘today’ or time 0. Figure
1.4 illustrates the timing for a 5-year annuity with cash flows or payments equal to C.

Figure 1.4: Cash flow timing for a standard annuity of 5 periods with payments equal to C

The formula for the PV of an annuity can be derived by considering an annuity as the
difference in two perpetuity streams. For example, consider a four-year annuity, paying $100
per year, with the first payment occurring one year from now. The time line in Figure 1.5 shows
how we can view this annuity as the difference between a standard perpetuity of $100 per year
and a second perpetuity of $100 per year that is ‘delayed’ by four years, so that the first cash
flow is received five years from today. Therefore, the PV of the annuity must equal the difference
between the PVs of these two perpetuities. This idea can be generalized to derive the annuity
formula. The details are in the mathematical Appendix. (I will not ask you to derive the formula
on quizzes or exams, so going through the proof is optional.)

Figure 1.5: A 3-year Annuity as the Difference between Two Perpetuities

Let’s re-examine the problem mentioned at the beginning of this section and see how the
PV of an annuity formula offers a relatively easy solution:

Example 1.11 A 30-year home loan is paid off by 30 equal annual payments. The first payment
occurs one year from the time of the loan, and the final payment occurs in 30 years’ time. What
is the present value or amount that can be borrowed under this loan structure if the interest rate
is 5% per year and the annual loan payments are $10,000 each?
Solution: You could solve this problem using the brute force approach, but that is extremely
1.5 Perpetuities and Annuities – 18 –

inefficient. Instead, recognize that the loan payment structure takes the form of a 30-year
(N = 30) annuity with cash flow amounts (C) equal to $10,000 and an interest rate of r = 5%.
Then just plug into the formula:
 
C 1
PV = 1−
r (1 + r)N
 
$10, 000 1
= 1−
0.05 (1.05)30N
= $200, 000(0.7686226)
= $153, 724.51

Here’s a second example:

Example 1.12 I offer to pay you $30,000 a year for the next 20 years. If the interest rate is 8%,
what is this offer worth to you today?
Solution: The contract amounts to a 20-year (N = 20) annuity with cash flow amounts (C)
equal to $30,000 and an interest rate of r = 8%. Plug into Eq. (1.9) to get:
 
C 1
PV = 1−
r (1 + r)N
 
$30, 000 1
= 1−
0.08 (1.08)20
= $294, 544.42

The ‘flip side’ of solving for the present value of an annuity is solving for the future value
of an annuity. The typical context for this type of problem is that you (or someone) will make a
series of equal dollar-valued deposits, investments, or payments at regular intervals for a specified
number of periods. It is assumed that these deposits or investments will earn interest at a fixed
interest rate r. We want to know what will be the total account balance or investment value at
the time that the final deposit or investment is made.

1.5.3 FV of an Annuity

As in the present value case, there is a convenient formula for determining the future value
of an annuity. It can be derived using a simple ‘trick’ that is worth illustrating. The present value
formula of Eq. (1.9) tells us today’s dollar valuation of all of the cash flows. In other words,
applying the PV formula gives us an amount of cash today that is equivalent to the stream of
annuity cash flows. If, instead, we want to know the value of this stream at some future date,
then all we need to do is use the basic future value equation to determine how the present value
of the annuity would grow with interest over this period. So, to derive the future value formula,
1.5 Perpetuities and Annuities – 19 –

just take the PV formula and multiply both sides by (1 + r)N :

 
C 1
PV = 1−
r (1 + r)N
 
C 1
P V (1 + r) =N
1− (1 + r)N
r (1 + r)N
C

F VN = (1 + r)N − 1
r
The third line uses the basic future value equation to recognize that multiplying the present value
by (1 + r)N gives the future value after N periods. Let’s recap:
Theorem 1.7. Future Value of a Annuity
The future value of an annuity paying a regular, fixed dollar cash flow C for N periods,
with the first cash flow occurring in one period, is given by the following expression, with
r denoting the per-period interest rate:
C

F VN = (1 + r)N − 1 (1.10)
r ♡

Example 1.13 Consider the following retirement plan: Starting from nothing today, you will
contribute $15,000 to a retirement account for the next 30 years, with the first deposit occurring
one year from now and the final deposit occurring 30 years from now. What will be your
retirement account balance 30 years from now (just after your final deposit) if the interest rate is
8% and there are no withdrawals from the account?
Solution: The retirement plan involves an annuity, so the efficient solution is to apply the formula
of Eq. (1.10) (as opposed to applying the brute-force approach). The plan involves 30 deposits
(N = 30) of C = $15, 000. Plugging into Eq. (1.10) gives:
C

F VN = (1 + r)N − 1
r
15, 000

= (1.08)30 − 1
0.08
= $1, 699, 250,

and so your retirement plan will net almost $1.7 million as of the final deposit.

1.5.4 Other Annuity Problems

The previous examples focus on solving for the PV or FV of an annuity. Just as with the
basic TVM equation; however, it is also possible to re-arrange these equations so that we are
solving for one of the other key ‘inputs’ (C, N , or r). Let us therefore discuss several examples
of this variety. Perhaps the most common case involves solving for the payment associated with
a loan. This type of problem has great practical relevance. If you have not already, many of
you will in the future take out loans to finance the purchase of automobiles, homes, or other
durable goods. Although there are many variations of loan contracts, the most common form
involves fixed, equal loan repayments, i.e., the repayments take the form of an annuity. Here is
an example:
1.5 Perpetuities and Annuities – 20 –

Example 1.14 Suppose that you purchase a home for $400,000. You will put 20% ‘down,’, i.e.,
pay 20% of the purchase price today. You plan to finance the remainder of the purchase price via
a home loan issued by a local bank. For the purposes of this problem, we will assume the term
of the loan is 30 years and that you will repay the loan through 30 equal annual payments, with
the first payment occurring one year from now.5 If the local bank offers a 4% annual interest rate
on the loan, what is the amount of your annual loan payment?
Solution: The first step is to determine the amount that you will actually borrow, i.e., the present
value of the loan. This equals $400, 000×(1−20%), or $320,000. The loan involves repayments
taking the form of a 30-year annuity for some amount C, which is what we are asked to determine
given an interest rate of r = 4%. The trick to solving the problem is to rearrange Eq. (1.9) so
that C is on the left-hand side:
 
C 1
PV = 1−
r (1 + r)N
PV × r
C= 
1 − (1+r)
1
N

$320, 000 × 0.04

C=  
1 − (1.04)
1
30

C = $18, 505.63,

so that your loan payment is just over $18, 500 per year.

It is also possible to solve for the number of payments “N” involved in an annuity, given
information about other aspects of the annuity (the interest rate, payment amount, and present
value).

Example 1.15 You have agreed to lend $20,000 to your sister at an annual interest rate of 3%.
Your sister’s budget allows for repayments of 2, 000 per year. Assuming the first repayment will
occur in one year’s time, after how many payments will the loan be paid off? Note that the last
payment to retire the loan would be less than $2,000 but do not worry about the specific amount
in answering the problem.
Solution: We need to solve for N in the the PV of an annuity formula given the other inputs
P V = $20, 000, r = 3%, and C = $2, 000. The algebra is:

5The 30 year term is quite realistic. However, Most home loans in practice involve monthly rather than annual payments.
1.6 Solving TVM Problems with a Financial Calculator or Excel – 21 –

 
C 1
PV = 1−
r (1 + r)N
 
rP V 1
= 1−
C (1 + r)N
rP V
(1 + r)−N = 1 −
C
rP V
−N ln(1 + r) = ln(1 − )
C
ln(1 − rPCV )
N =−
ln(1 + r)
0.03$20,000
ln(1 − $2,000 )
N =−
ln(1.03)
N ≈ 12.066.

Notice that we need to apply the same ‘log trick’ in the fourth line of the solution that was used
to solve for ‘N’ in the basic TVM formula (see Section 1.3). The solution is not an integer.
This implies that the loan will not quite be paid off following the 12th payment, and therefore
13 payments would be required to retire the loan. One of the problem set questions asks you to
determine the amount of the 13th and final payment that would exactly retire the loan.
As the above solution illustrates, it is possible to solve explicitly for ‘N’ in the PV of an
annuity formula, although the algebra is a bit tedious. (The fifth line of the solution gives the
general formula with ‘N’ on the left-hand side.) An alternative is to use a financial calculator (or
Excel). The next section provides a brief discussion of that approach.
The final type of problem involving the PV and FV of an annuity formulas involves solving
for the interest rate r given the other parameters. This case differs from the others because there
is not a convenient algebraic formula. In this case, you should use your financial calculator or
Excel to obtain the solution. The next section will illustrate with an example.

1.6 Solving TVM Problems with a Financial Calculator or Excel

This section briefly covers how to perform important TVM computations using a standard
financial calculator or using Excel.

1.6.1 Using the Financial Calculator

Standard financial calculators have built-in functions to compute the present value or future
value of a standard annuity, as well as for determining the interest rate or number of periods for
an annuity given the other key inputs. For the purposes of this discussion, I will focus on the
‘Texas Instruments BA II Plus’ calculator, which is the recommended one for the course. Other
financial calculators tend to work similarly, though.
1.6 Solving TVM Problems with a Financial Calculator or Excel – 22 –

If you have not previously used a financial calculator, there are a couple of important steps
that you should take in order to appropriately set up your calculator for work in this course. These
steps involve:
1. Set the number of payments per period or year to one. Many financial calculators by
default set this to twelve (12 payments per year).
2. Ensure that your calculator is in ’END’ mode. This means that your calculator will
assume standard ‘end of period’ timing for annuity cash flows. The alternative is ‘BEGIN’
which corresponds to an alternative structure in which the first payment is received/paid
immediately (see next section on an ‘annuity due’).
3. Most calculators set the default number of decimals to two digits. You should reset this to
four or five decimals for sufficient accuracy on problems involving interest rates.
There are a number of useful online videos that show exactly how these options can be reset.
Here are links to two particular YouTube videos that cover this topic:
1. https://www.youtube.com/watch?v=maD6HuaA-2k
2. https://www.youtube.com/watch?v=nScQsMmohZ0
In addition to standard calculator capabilities, your financial calculator has a set of spe-
cial keys and pre-programmed functions for computing common financial quantities. The five
important keys for our purposes correspond to inputs in standard annuity formulas and are as
follows:
1. N The number of periods in the annuity
2. I/Y The per-period interest rate (interpreted as a percentage)
3. PMT The regular, periodic cash flow in the annuity
4. PV: The present value or a cash flow received/paid today
5. FV: The future value or a cash flow received/paid at the time of the final cash flow in an
annuity
These keys can be used to execute standard annuity-related computations. Think of each
key as a storage unit. In order to store a particular value under one of these keys, you enter
the number on your calculator, and then hit the corresponding special key button. For example,
suppose we want to store the value ‘10’ under the ‘N’ key. To do so, we would enter 10 into the
calculator and then push the ‘N’ key button. The calculator will issue a display that confirms ‘N
= 10,’ letting us know that the value 10 is now stored in the N key. Storing other inputs is similar.
Take special note that the ‘I/Y’ interprets input as percentages. So, for example, if we want to
store an interest rate of 5%, we should key in ‘5’ (and not 0.05) and then push the ‘I/Y’ button.
Once the inputs for a particular annuity calculation are stored under the corresponding keys
in the financial calculator, calculations using these inputs are executed by pushing the ‘CPT’
(‘compute’) button and then the button corresponding to the quantity one wishes to compute.
Suppose, for example, that we want to compute the present value of a 3-year annuity of $100
when the annual interest rate is 7%. The steps are:
1.6 Solving TVM Problems with a Financial Calculator or Excel – 23 –

1. Input the value 3 and push the ‘N’ key

2. Input the value 7 and push the ‘I/Y’ key
3. Input the value −100 and push the ‘PMT’ key (the reason for the minus sign will be
explained below)
4. Push the ‘CPT’ button followed by the ‘PV’ button
Upon following these steps, the financial calculator should display the answer: ’PV = 262.43160’
(The exact output will vary according to the number of decimals your calculator is set to display.)
Going forward, and especially in posted problem set solutions, I will often use short-hand to
indicate the financial calculator steps associated with a particular computation. For example, the
shorthand “N = 3, I/Y = 7, PMT=-100; CPT PV” would designate the ‘calculator solution’ to the
previous example.
Why did we input the PMT amount as a negative number? This is because the financial
calculator has an ‘inflow/outflow’ convention. Essentially, the calculator is thinking about two
parties in a ‘loan’ (or related) transaction. One party is paying cash today in exchange for
receiving cash in the future, and the other is doing the opposite. By inputting the PMT as a
negative value, we are indicating to the calculator that we will be receiving a positive cash flow
today, so the sign of the ‘answer’ is positive. If we had instead input the PMT amount as 100,
then the calculator would have returned −262.43160 as the present value. We can then simply
ignore the minus sign under this alternative approach.
Perhaps the most important reason to be comfortable with using the financial key functions
involves solving for the interest rate associated with an annuity, because in this case there is not
a convenient analytical formula. So let’s consider an example of this type:

Example 1.16 Suppose that you take out a loan for $60,000. You will repay the loan by making
10 annual payments of $8,000 per year, with the first payment occurring one year from now.
What is the annual rate of interest that corresponds to this loan repayment schedule?
Solution: The loan repayment scheme involves a standard annuity for 10 years with a cash flow
of $8,000. The present value of the loan is $60,000. We should therefore input these quantities
under the ‘N’, ‘PMT’, and ‘PV’ keys, respectively. Importantly, either the PMT or PV amount
should be entered as negative, and the other one as positive. So, for example, key in ‘8000’ under
PMT and ‘-60000’ under PV. Then we will push ‘CPT’ followed by ’I/Y’. (More concisely, in
our shorthand, N=10, PMT=8000, PV = -60000, FV=0, CPT I/Y.) The calculator will then tell
us that the implied interest rate is 5.6045% per year. Figure 1.6 shows a visual representation of
the calculator inputs (on top of the keys) and the output (on the bottom).
What would have happened if we had instead input the PMT as a negative amount and the
PV as a positive amount? Nothing different – we would obtain exactly the same answer. This is
because we would simply be reversing the notion of who is making the loan versus paying it off,
and of course this does not change the implied interest rate. BUT, if had we entered both the PMT
and PV as positive (or negative) amounts, then our calculator would have ‘puked’ and produced
1.6 Solving TVM Problems with a Financial Calculator or Excel – 24 –

Figure 1.6: Computing the Interest Rate using a Financial Calculator

an error. Intuitively, entering both amounts as positive is like imagining a loan in which you
borrow $60,000 today and then also receive payments of $8,000 each year. This doesn’t make
sense as a loan and therefore the error occurs.
Notice also that Figure 1.6 enters a zero under the FV key. Zero is the default value for each
key, so this is not necessary as long as the current FV key value is at zero.6 However, you will
often use the calculator repeatedly for different TVM problem calculations. Here a an important
word of caution is in order: a stored value under one of the financial keys will remain in place
unless it is explicitly cleared or the calculator is turned off. So, in the above example, if we
had previously input a value under the FV key and not ‘cleared it’ (see below), then that value
would be retained and our interest rate calculation would be incorrect. By explicitly entering
zero under the FV key, we avoid the possibility of this happening. There is a second approach to
avoid this problem. You can clear all of the current input values in the keys by pushing the ‘2nd’
button, followed by the ‘CLR TVM’ button. (The ‘2nd’ button is like a shift key that accesses
functionality above, rather than below, each button.)
The financial calculator is a useful tool and is relatively easy to use with a bit of practice.
However, remember to:
1. Clear out the TVM key values before starting any new problem
2. Be attentive to sign conventions and put a negative sign in front of ‘outflows’ and positive
sign in front of inflows

1.6.2 Excel TVM Functions

Microsoft Excel provides a number of ‘built-in’ functions that perform TVM computations.
The most important of these for our purposes are the following:
1. PV(rate, nper, pmt, [fv], [type])
2. FV(rate,nper,pmt,[pv],[type])
3. PMT(rate, nper, pv, [fv], [type])
4. NPER(rate,pmt,pv,[fv],[type])
5. RATE(nper, pmt, pv, [fv], [type], [guess])
The functions listed above essentially replicate the key financial calculator functions dis-
cussed in the previous subsection. Each function is listed along with its key ‘arguments’ – the

6Don’t necessarily assume that turning the calculator off and then back on will do this – some financial calculators will
continue to retain keyed-in values until they are explicitly cleared.
1.6 Solving TVM Problems with a Financial Calculator or Excel – 25 –

inputs the function requires. The arguments that appear in brackets are optional. To take a
concrete example, the ‘PV()’ function requires three arguments, the interest rate (‘rate’), the
number of periods (‘nper’) and the regular, level, cash payment in an annuity (‘pmt’). Note that
the last of these might be zero. The [fv] argument allows for a potential lump sum amount at
maturity. This can be used to apply the basic TVM formula for a single future cash flow by
omitting (or setting to zero) the payment argument. Finally, the [type] argument defaults to our
standard annuity timing (first cash flow received at the end of the first period), but can optionally
be changed to an ‘annuity due’ structure.
The Excel TVM functions have a ‘sign convention’ similar to the financial calculator
functions. For example, if you input a positive PMT amount in computing the PV of a standard
annuity, the Excel ‘PV()’ function will output a negative number. Therefore, if you would like
the answer to be positive, input a negative quantity as the PMT argument, or alternatively simply
multiply the output of the Excel function by minus one. In order to illustrate how the function
is used, suppose we want to compute the present value of the following set of cash flows: ten
annual payments of 20, 000 per year with an extra 50, 000 received at year ten, when the interest
rate equals 5% per year. Figure 1.7 shows a screenshot that applies the PV formula in Excel to
make the necessary computation. The text that appears below cell B7 shows that formula that
B7 contains. Note that both the PMT and FV inputs are the opposite of the values contained in
cells B3 and B4, respectively, in order to ensure a positive output for the present value.

Figure 1.7: Using Excel’s ‘PV()’ Function

As a second example, we will use Excel to compute the interest rate implied in the loan
example considered above and solved using a financial calculator. Recall that this example
involved a loan of $60,000, repaid via 10 annaul payments of $8,000 with standard timing (first
payment occurring in one year). Figure 1.8 shows proper application of the Excel ‘RATE()’
function to solve this problem. Naturally, we obtain the same answer (5.6045%) as we did using
the financial calculator. Note carefully the signs of the input PV and PMT arguments. The PMT
is entered as a negative value, and the PV as a positive. The opposite convention also works
equally well. However, as with the financial calculator, you cannot input both PV and PMT as
1.7 Tricks of the Trade – 26 –

positive (or negative) and doing so results in an error.

Figure 1.8: Using Excel’s ‘RATE()’ Function

1.7 Tricks of the Trade

This section introduces some very important “tricks of the trade” for solving time-value of
money problems.

1.7.1 Annuities or Perpetuities ’Due’

Our formulas for the PV of a perpetuity or annuity assume that the first cash flow occurs
at the end of the first period. Instead, sometimes the first cash flow is “due,” meaning paid
immediately. How can we adapt our formulas to this scenario? It turns out that there is a simple
link between the value of a standard annuity (first CF at the end of the first period) and an annuity
due (first CF now).
Theorem 1.8. PV of an Annuity Due
The present value of an annuity due involving level payments of C for N periods with an
interest rate of r equals the present value of a standard annuity with the same features,
multiplied by 1 + r. A similar result holds for future values. Summarizing:

PV Annuity Due(C, r, N ) = (1 + r) × PV Annuity(C, r, N ) (1.11)

FV Annuity Due(C, r, N ) = (1 + r) × FV Annuity(C, r, N ), (1.12)

where (C, r, N ) denote the payment amount, interest rate, and number of periods, which
are assumed to be the same for both annuities.
♡

Figure 1.9 illustrates the intuition underlying the above result. Comparing the annuity due
with the corresponding standard annuity, it should be clear that each cash flow earns one additional
period of interest relative to the standard annuity case. The following example illustrates how to
use the result:

Example 1.17 You are offered the opportunity to purchase a contract guaranteeing a series of
10 annual $5,000 payment, with the first payment occurring today. What is the value of this
1.7 Tricks of the Trade – 27 –

Figure 1.9: Annuity Due: Intuition for ‘multiply by 1 + r result’

contract if the annual interest rate is 7%?

Solution: We will solve the problem two ways and show that they are equivalent. The first
approach is to consider the contract to consist of two components: 1) $5,000 now; and 2) a 9-year
(standard) annuity of $5,000 per year. It should be clear that the PV of the first component is
$5,000. To obtain the PV of the second component, we can apply Eq. (1.9):
 
C 1
P V (Component 2) = 1−
r (1 + r)N
 
$5, 000 1
= 1−
0.07 (1.07)9
= $32, 756.16,

and so the total value equals $5, 000 + $32, 756.16 = $37,756.16. The second approach is
to use the annuity due formula. In this case we compute the PV of a standard annuity with
C = 5000, r = 7%, and N = 10. Then we will adjust this for the fact that it is an annuity due
by multiplying by 1 + r. Here’s the math:
 
C 1
P V (annuity due) = (1 + r) 1−
r (1 + r)N
 
$5, 000 1
= (1.07) 1−
0.07 (1.07)10
= 1.07(35, 117.91)
= $37, 756.16.

So, we get precisely the same answer under either approach.

We can also consider a ‘perpetuity due,’ which differs from the standard perpetuity in that
the initial cash flow is received/paid at time 0 rather than at the end of one period. The same
‘multiply’ by 1 + r trick applies.7 A homework problem gives practice.

7To see this, note that the PV of a perpetuity due with regular cash amounts of C equals C + C
r
= C
r
(1 + r).
1.7 Tricks of the Trade – 28 –

1.7.2 Delayed annuities and perpetuities

The second trick involves how to deal with annuities or perpetuities that are “delayed.” This
means that, instead of the first cash flow occurring one period from now, the first cash flow occurs
K periods in the future, where K is some integer greater than one. The key trick to handle these
situations is as follows. First, use the standard PV for an annuity (or perpetuity) to compute the
present value of the cash flows as of one period before the first cash flow occurs. At this point, we
have expressed the stream of cash flows in the form of an equivalent lump sum of cash received
K − 1 periods from now. Second, use the basic PV equation to discount this cash flow back to
the present, i.e., for K − 1 periods. Here is an example of this concept in practice:

Example 1.18 Kitty Toys, Inc. will be acquired by a competitor today. Ms. Cheshire, the
company’s current CEO, will stay on with the new company for 3 years. At the time of her
departure, she is entitled to receive a ‘golden parachute’ of $500,000 per year for the next four
years (first payment occurs three years from now). What is the present value of this set of
payments if the interest rate equals 4%?

Figure 1.10: Sketch of Timeline for Example 1.18

Solution: The golden parachute payments take the form of a delayed annuity. The first step of
the solution is to apply the PV of an annuity formula. This will give us the value of the payments
as of two years from now – two years because the first payment occurs three years from today,
i.e., one year from two years from now as the formula assumes. So we have:
 
C 1
PV = 1−
r (1 + r)N
 
$500, 000 1
= 1−
0.04 (1.04)4
= $1, 814, 947.61

To be clear, although we are applying the ’PV’ formula, this valuation is as of one period before
the first cash flow, which is two years from today. (See Figure 1.10 for a visual illustration.) The
final step, then, is to treat this amount as a future value as of two years from now and discount it
1.7 Tricks of the Trade – 29 –

for two years at 4% to get the true ‘present’ value:

F VN
PV =
(1 + r)N
$1, 814, 947.61
=
(1.04)2
= $1, 678, 021.09

So the golden parachute is worth about $1.68 million.

1.7.3 Breaking Complicated Problems into Pieces

The final trick is to ‘break things into pieces.’ No, not in your dorm room. What this means
is that some cash flow sequences are too complicated to correspond directly to the key formulas
we have developed. But, if you break them into the right pieces, the pieces can be analyzed
using the standard formulas. Then you can solve the original problem by combining results for
the various pieces. The idea is best illustrated via an example:

Example 1.19 Suppose that you have graduated from VT and you are in need of $30,000 in
start-up money as you move to the (expensive!) west coast and take an entry-level job. Your aunt
agrees to lend you this amount. She offers you a ‘family special’ interest rate of 2% per year.
You will pay her back over 10 years, but, because you are just starting your career, your salary
will be lower in the near term but expected to rise. So you agree that you will pay back a certain
amount X for the first 5 years (first payment in one year), and then double this amount to 2X for
the remaining 5 years. What is the starting payment X that exactly pays off the loan?

Figure 1.11: Sketch of Timeline for Example 1.19

Solution: This problem perhaps seems difficult, because the loan repayment does not involve
fixed dollar amounts, as the annuity formula expects. The trick; however, is to realize that we
can break the repayment plan into two separate annuities. The first involves the level amount X
paid for five years. The second is a ‘delayed’ annuity, as discussed above, and involves the level
amount 2X for five years. The standard annuity formula can be applied to value each of these
components of the repayment plan. Our strategy is to then set up an expression in which the PV
borrowed ($30,000) exactly equals the sum of the present values of the two payment components.
1.7 Tricks of the Trade – 30 –

Figure 1.11 provides a visual illustration. The math is as follows:

P V = P V (Payments part 1) + P V (Payments part 2)

For payment part 1, we have:

 
C 1
P V (Payments part 1) = 1−
r (1 + r)N
 
X 1
= 1− ,
0.02 (1.02)5
recalling that the regular cash payment is X (the variable we will eventually solve for). Similarly,
for payment part 2, using the delayed annuity trick, we have:
 
1 C 1
P V (Payments part 2) = 1−
(1 + r)5 r (1 + r)N
 
1 2X 1
= 1− ,
(1.02)5 0.02 (1.02)5
where the first term on the right-hand side occurs because of the delayed structure of the annuity:
the first payment occurs six years from now, hence the resulting valuation is as of five years from
now. Now we can plug these expressions for the PV of components 1 and 2 into the top equation,
plug-in, and solve. The algebra is a bit tedious, but straightforward:

P V = P V (Payments part 1) + P V (Payments part 2)

   
X 1 1 2X 1
$30, 000 = 1− + 1−
0.02 (1.02)5 (1.02)5 0.02 (1.02)5
   
2 1 1
$30, 000 = X 1 + 1−
1.025 0.02 (1.025
(0.02)$30, 000
X=
 
2
1 + 1.025 1 − (1.02
1
5

X = $2, 263.56.

So the payment for the first 5 years will be $2,263.56, and then the payment will double to
$4,527.72 for the second five years. This is a relatively challenging problem, and is similar to
the ‘savings/retirement’ problems we will cover in the next chapter. Breaking things into pieces
is essential in solving such problems.
1.7 Tricks of the Trade – 31 –

K Chapter 1 Problems k
1. One year ago, Bitcoin was trading for $10,069. Suppose you bought a unit of Bitcoin at
that time and sold today. The current value of Bitcoin is $11,666.
(a) What is your rate of return on this Bitcoin investment?
(b) (Non-quantitative) If you had deposited the same amount in a bank savings account,
you would have earned a rate of return of approximately 1%. Comparing this to
the Bitcoin return, why don’t all investors eschew bank accounts in favor of Bitcoin
investments?
2. Suppose that you invest $12,000 today, and the investment earns 6% a year (i.e., 6% is the
rate of return). What will your investment be worth in 10 years?
3. I offer to pay you $25,000 in 8 years. How much is that worth to you today if the annual
interest rate is 10%?
4. If the annual interest rate equals 5.5%, what is the discount factor for cash received eight
years from now (DF (8))? What is today’s value of $50,000 received in 8 years?
5. Suppose that you borrow $25,000 from me today and pay me $39,671.86 in 6 years. What
is the rate of return on this loan?
6. You receive $100 from your grandmother. Ally Bank will guarantee a 2% annual interest
rate indefinitely. How long will it take to quintuple your money (have 5 times more
money)?
7. Your buddy wants to borrow $2,000 for some great scheme. He promises that he can either
pay you back $6,000 in five years or $15,000 in ten years. Assuming he will pay you back
either way, which is the better option? Hint: Solve for the annual rate implied under each
option.
8. You own a perpetuity that pays $300 a year at the end of each year (starting in year 1).
(a) If the current appropriate annual interest rate is 3%, what is the current value of the
perpetuity?
(b) If interest rate rises to 5%, what is the new value of the perpetuity?
9. You currently have $10,000 in savings. You are going to put an additional $15,000 into
your savings next year and $20,000 in two years’ time. If the annual interest rate is 4%,
how many years until you have $100,000 in savings?
10. Suppose you invest $5,000 per year at 7.25%. Assuming the first investment occurs one
year from now, how much money will you have in 15 years?
11. You wish to borrow $19,721 today and that you will repay me with equal annual level
payments over 10 years. If the interest rate is 6.5% and the first payment is one year from
today, how much will you pay me each year?
12. Rank the following cash flow streams according to present value. In all cases, assume that
the relevant interest rate is 5%. HINT: use TVM logic to avoid unnecessary computations.
(a) $1,000 per year in perpetuity (first cash flow occurring one year from now)
1.7 Tricks of the Trade – 32 –

(b) Ten consecutive annual payments of $2,000, with the first payment received one year
from now
(c) Four consecutive annual payments of $5,000, with the first payment received ten
years from now
(d) $20,000 received ten years from now
(e) $20,000 received fifteen years from now
13. Suppose you charge me a 9% annual interest rate and I promise to pay you $3,000 a year
for four years (first payment is one year from today) plus $9,000 five years from now. How
much is this stream of cash flows worth today?
14. Reconsider Example 1.15. Determine the exact balance remaining on the loan in this
example after 12 years. Then, assuming the borrower will retire the loan with a final
payment in 13 years, determine what this payment must be. Hint: Should it equal the
balance after 12 years? Why or why not?
15. Carmelo ‘Melo’ Anthony has decided to retire from the NBA and will coach the Richmond
Elite ABA team post-retirement. The Richmond Elite have offered Melo 3 payment
schemes to choose between. Each involves a signing bonus payable immediately and three
years of salary payable at the end of each year. The first scheme calls for Melo to receive
a signing bonus of $3 million and a salary of $700,000 million per year for 3 years. The
second calls for him to receive a signing bonus of $2 million and an annual salary of $1.2
million per year. The third scheme calls for Melo to receive a signing bonus of $1 million
and annual salaries of $1.2 million, $1.6 million, and $2 million. Melo seeks your advice
on this choice of payment schemes. Advise him on which scheme to pick. Use an annual
interest rate of 8% to evaluate the alternatives. Bonus challenge: Fixing the signing bonus
of Option 1 at $3 million, what signing bonus amounts for the other two options would
make Melo exactly indifferent among the contract options?
16. This problem extends Example 1.13 in the text involving the future value of annuities. In
that example we showed that a series of 30 annual deposits of $15,000 with an interest rate
of 8% grows to $1,699,250 by the time of the final deposit. Now, consider the following
alternative retirement plan: you contribute $20,000 to a retirement account for the next 20
years and nothing thereafter, although your deposited funds continue to earn interest after
20 years. What is the future value of this plan after 30 years if the interest rate is 8%?
17. Suppose that you promise to pay your sister $10,000 a year for five years with the first
payment occurring one year from today. If the interest rate is 5%, what will this stream be
worth to her 11 years from now?
18. Suppose the interest rate is 5.8%. What will be the value seven years from now of
a perpetuity that pays $1,000 per year forever, assuming that the first payment of the
perpetuity starts 14 years from now.
19. Suppose I promise to pay you 5 payments of $12,000 each with the first payment to come
1.7 Tricks of the Trade – 33 –

7 years from today. If the interest rate is 8.25%, what is this stream of cash flows worth to
you today?
20. A 20-year annuity pays $8,000 per year, and payments are made at the end of year. Suppose
the annual interest rate is 6% for the first 10 years, but 9% for the following ten years.
What is the present value of this annuity?
21. You will receive $12,000 per year for six years. The annual interest rate is 4.8%.
(a) What is the present value of these payments if they are a standard annuity (first
payment in one year)?
(b) What is the present value of these payments if they are an annuity due (first payment
today)?
(c) Suppose you will invest the payments for six years and then withdrawal funds. How
much more money will you be able to withdraw if the structure is an annuity due
versus a standard annuity?
22. ABC, Inc. has issued preferred shares that pay the holder $4/share each year. Assuming
these shares trade actively in a financial market, what should happen to the price per share
just after an annual dividend is paid to current shareholders relative to the price just before
this dividend is paid? How does this relate to the distinction between a perpetuity and a
‘perpetuity due’?
23. You are offered a financial contract that pays $10,000 for the next three years (first payment
in one year) and then $5,000 per year thereafter into perpetuity. What is the present value
of this contract if the discount rate is 4%?
24. Old exam question: You currently have $20,000 in savings. You will make no deposits
until three years from today, at which point you will contribute $5,000 to the account, and
then continue to contribute an additional $5,000 to the account for the following 9 years
(so there are ten total $5,000 deposits). You will not make any subsequent deposits. If your
savings account offers a constant 5% interest rate, and assuming no withdrawals, what will
be the balance in your account 20 years from today?
25. Reconsider the problem of Example 1.19. Suppose that you and your aunt renegotiate
the terms of the loan so that you will make no payments at all for two years. Three years
from now, you will make the first of 5 annual payments of X. Then you will make five
additional annual payments of 2X. How does this change the value of X that pays off the
loan?
 Chapter 2 TVM: Advanced Concepts

This chapter continues our study of the time value of money by introducing several extensions
and advanced concepts that build upon the foundation introduced in the previous chapter.

2.1 Growing Cash Flows

The standard perpetuity and annuity are restrictive in the sense that they require constant
cash flow amounts. Certain important real-world cash flows streams do satisfy this requirement.
A prominent example involves standard 15- or 30-year mortgage loans that involve fixed monthly
payments. However, in many other important contexts, dollar-valued cash flows tend to change
over time and often feature a pattern of growth. Consider, for example, a firm that wants to value
the stream of cash flows generated by a planned project. (Finance majors: you will get first-hand
experience with this exercise in Corporate Finance.) In most cases, the firm will anticipate that
the cash flows generated by the project increase as sales ramp-up, etc. In addition, the general
price level tends to increase over time (termed inflation). Inflation also tends to drive increases
in dollar-valued project cash flows.
It is not appropriate to apply the standard perpetuity and annuity formulas to growing cash
flows. Does this mean that we must revert to the ‘brute force’ approach? No. Fortunately, there
are adjusted, convenient formulas that apply to growing cash flows.
Before presenting the key TVM formulas involving growth, it is important to briefly discuss
the nature of growth that is assumed. In particular, the formulas we will introduce assume a
constant growth rate for cash flows (C), so that cash flows evolve according to the equation:

Ct+1 = (1 + g)Ct (2.1)

where t indexes the time period (year, month, etc.). The parameter g in Equation 2.1 can be
interpreted as an assumed constant percentage growth rate in cash flows. Of course, when
g = 0, Ct+1 = Ct and this reduces to the special case of level cash flows that is assumed by
the standard perpetuity and annuity valuation formulas. When g > 0, e.g., g = 5%, then cash
flows are assumed to grow at this rate each period. It is also conceptually possible that g < 0,
which implies that cash flows shrink at a constant rate over time, although this assumption is less
common in practice. The following example illustrates the properties of cash flows featuring a
constant growth rate:

Example 2.1 XYZ Inc. currently generates cash flows from operations of $10 million per year.
An analyst forecasts that these cash flows will grow at the constant rate of 4% per year over the
next decade. What are forecasts for the level of cash flows for XYZ Inc. in three years and ten
years?
2.1 Growing Cash Flows – 35 –

Solution: Applying Equation (2.1) allows us to write:

Ct+3 = (1 + g)Ct+2
= (1 + g)2 Ct+1
= (1 + g)3 Ct
= (1.04)3 × $10
= $11.25 million

For the 10-year prediction, the same approach gives (omitting the recursive substitution
process):

Ct+10 = (1 + g)10 Ct
= (1 + g)10 Ct
= (1.04)10 × $10
= $14.80 million

Now that the constant growth rate assumption is clear, it is time to introduce TVM formulas
that allow for this type of growth. The following provides a simple generalization of the standard
perpetuity formula that allows for growing cash flows:
Theorem 2.1. PV of Growing Perpetuity
Consider a growing perpetuity, defined as an initial cash flow C occurring at the end of
the first period, and then subsequent periodic cash flows at grow at the rate g forever.
Assuming the interest rate is r and r > g, the present value of the growing perpetuity
equals:
C
PV = . (2.2)
r−g
♡

Notice that when g = 0, the PV formula of Eq. (2.2) reduces to P V = C/r, which
coincides as expected with the standard perpetuity formula. Note the notation and timing for the
cash flows. It is important to remember that the first cash flow in the growing perpetuity is ‘C’.
Figure 2.1 illustrates via a timeline.

Figure 2.1: Timeline: A Growing Perpetuity with Growth Rate g

Here are some examples that apply the formula:

2.1 Growing Cash Flows – 36 –

Example 2.2 Suppose the government of Denmark introduces a new form of ’consol bond’
that promises perpetual payments to holders. The initial payment will total 200 Danish Kroner.
Subsequent payments will grow at 1.5% in perpetuity. If the appropriate discount rate is 4%,
what is the present value of this consol bond?
Solution: Apply the formula for the PV of a growing perpetuity in Eq. (2.2). The growth rate
g = 0.015, r = 0.04, and C = 200:
C
PV =
r−g
200
=
0.04 − 0.015
= 8, 000

So the consol bond should be worth 8,000 Kroner today.

Example 2.3 Continuing the previous example, suppose that the Danish government wishes for
the consol bond to trade for 10,000 Kroner rather than 8,000. Assuming the discount rate is fixed
at 4% and the initial payment will remain 200 Kroner, how should the growth rate be adjusted to
achieve the desired valuation?
Solution: In this case, we need to solve for the growth rate g, given the assumptions r = 0.04,
C = 200, and a price (PV) of the growing perpetuity of 10,000:
C
PV =
r−g
200
10, 000 =
0.04 − g
g = 2%,

where the last line follows after a bit of algebra.

Before proceeding, it is worth discussing the relation between r and g. Suppose, for
example, that r = 4% and g = 5%. Naively plugging these values into the formula for the PV
of a perpetuity produces a negative result. However, this is not sensible, because a set of positive
future cash flows cannot have negative value. The resolution is that the formula is valid only for
cases in which r > g. Note that this does not preclude negative g (declining cash flow amounts).
For example, a case in which r = 4% and cash flows are declining at 5% (g = −5%) has a
well-defined PV given by the formula. When g ≥ r, the formula is invalid and the cash flow
stream would actually have infinite value.
There is an extension to the key annuity formulas to accommodate growth that is similar to
the extension discussed above for perpetuities.
Theorem 2.2. PV of Growing Annuity
Consider a growing annuity, defined as an initial cash flow C occurring at the end of the
first period, and then subsequent periodic cash flows that grow at the rate g for a total of
2.1 Growing Cash Flows – 37 –

N cash flows. The present value of this growing annuity is:

  !
C 1+g N
PV = 1− . (2.3)
r−g 1+r
♡

When g = 0, the PV formula reduces to the standard annuity formula, as it should.

Example 2.4 Continuing in the spirit of the previous two examples, suppose the Danish govern-
ment decides to offer a security that will pay growing cash flows for a total of 30 years, rather
than into perpetuity. Specifically, the security will pay 200 Kroner in one year, and then amounts
that grow at 2% per year thereafter, until the final payment 30 years from now. What is the value
of this security if the discount rate remains 4% as in the previous example?
Solution: Apply the formula for the PV of a growing annuity in Eq. (2.3). The growth rate
g = 0.02, r = 0.04, C = 200, and the total number of payments is N = 30:
  !
C 1+g N
PV = 1−
r−g 1+r
  !
200 1.02 30
= 1−
0.04 − 0.02 1.04

= $4, 415.23.

What would happen in the previous example if the growth rate of cash flows were set to 6%
rather than 2%? Note that in this case g > r. Does this mean the formula is inapplicable and
the PV is infinite? No. In contrast to the growing perpetuity case, there is nothing problematic
about the case g > r for growing annuities. The formula gives the correct present value, which
will be positive. Intuitively, a finite set of future cash flows cannot have an infinite present value.
Finally, just as there is an extension to the PV of annuity formula to permit growth, there is
a similar extension for the future value of an annuity formula:
Theorem 2.3. FV of Growing Annuity
Consider a growing annuity, defined as an initial cash flow C occurring at the end of the
first period, and then subsequent periodic cash flows at grow at the rate g for a total of N
cash flows. The future value of this growing annuity N periods from now equals:
C

F VN = (1 + r)N − (1 + g)N . (2.4)
r−g
♡

This formula for the FV of a growing annuity proves particularly helpful in analyzing savings and
retirement problems, because savings contributions often grow over time, both due to effects of
inflation and real salary growth associated with promotions, etc. The following example provides
an illustration:

Example 2.5 You will make a series of bank deposits to finance the purchase of a riding
lawnmower in 8 years time. You plan to deposit $1,000 next year, and then make subsequent
annual deposits growing at rate of 5% per year, with the final deposit occurring 8 years from
2.2 Alternative Compounding Frequencies and the True Cost of Borrowing – 38 –

now. Assuming no withdrawals from the account, what amount will you have available in the
account to spend on the lawnmower at the end of year 8, assuming the interest rate earned on
deposits equals 9%?
Solution: Apply the formula for the future value of a growing annuity, with inputs as follows:
C = $100, r = 9%, g = 5%, N = 8:
C

F VN = (1 + r)N − (1 + g)N
r−g
$1, 000

= (1.09)8 − (1.05)8
0.09 − 0.05
≈ $12, 878.

So, you will have a bit less than $12,900 to spend.

2.2 Alternative Compounding Frequencies and the True Cost of

Borrowing

Thus far, we have restricted attention to cash flow or payment streams that occur annually.
However, this is restrictive because many prominent real-world settings involve cash flows at other
frequencies. For example, standard home and auto loans feature monthly payments. Government
bonds often pay ‘coupons’ semi-annually. Stock dividends are often paid quarterly, and so on.
This section introduces interest rates that compound at frequencies other than annual, and shows
how to obtain several types of ‘annualized’ rates given the underlying periodic interest rate. The
section also briefly discusses how to assess the ‘true cost of borrowing’ and several strategies
lenders often employ to lower the perceived cost of borrowing.

2.2.1 EPRs and Different Compounding Frequencies

Definition 2.1. Effective Periodic Rate (EPR)

The effective periodic rate or EPR equals the underlying rate of interest and frequency
over which interest compounds.
♣

Definition 2.2. Annual Percentage Rate (APR)

The annual percentage rate of interest equals the product of the effective periodic rate
(EPR) times the number of compounding periods per year. Mathematically:

AP R = EP R × m, (2.5)

where m equals the number of compounding periods per year, which can be any positive
number, but is most commonly an integer such as 4 (quarterly compounding) or 12
(monthly compounding).
♣
2.2 Alternative Compounding Frequencies and the True Cost of Borrowing – 39 –

To illustrate the connection between EPRs and APRs in a context of practical interest,
consider the following example:

Example 2.6 Suppose that you purchase a new automobile from a local dealer. The automobile
costs $21,000. After running a credit check, the dealer offers you financing under the following
terms: the dealer will lend you the $21,000 with a 5-year loan term and monthly fixed payments.
The first payment will be due in one month and the final payment will be due in 60 months. As
you review the paperwork, the dealer pauses to note that she is obligated by law to disclose to
you the APR associated with the loan, which is 4%. What is the EPR associated with the loan
and the corresponding monthly payment?
Solution: Equation (2.5) states that the APR equals the EPR times the number of compounding
periods. We can easily reverse this equation to express the EPR as a function of the APR (which
we are given) and the compounding frequency m:
AP R
EP R =
m
0.04
=
12
= 0.333%,

where the second line inputs the APR of 4% and recognizes that the monthly loan structure
implies that m = 12. Note that the final line is expressed as a percentage: the EPR is 0.3333%,
or 33.33 basis points. Now that we know the underlying EPR, we can apply the standard annuity
formula to determine the monthly loan payment as follows:
 
C 1
PV = 1−
r (1 + r)N
PV × r
C= 
1 − (1+r)
1
N

$21, 000 × 0.00333333

C=  
1 − (1.00333333)
1
60

C = $386.75,

and so the monthly payment will be just under $387. The financial calculator recipe for the
payment computing is: N=60, I/Y=0.0033333, PV = -21,000, FV=0, and then CPT PMT.
Note that the APR does not take into account the compounding of interest within a specific
year. As Eq. (2.5) shows, the APR is calculated by simply multiplying the periodic interest
rate (EPR) by the number of periods per year. Despite this weakness, the APR is an important
quantity, in part because the Federal Truth in Lending Act requires that borrowers disclose the
APR in every consumer loan agreement.1
Since the APR ignores the effects of compounding, it does not fully capture the economic
cost associated with borrowing. The effective annual rate (EAR) is an alternative annual

1Perhaps you have seen television or internet advertisements for auto dealerships that promote ‘low APRs’ or even ‘0%
APR’.
2.2 Alternative Compounding Frequencies and the True Cost of Borrowing – 40 –

borrowing cost measure that does incorporate the effects of compounding.

Definition 2.3. Effective Annual Rate (EAR)
An annualized measure of the cost of borrowing that incorporates the effects of the
compounding of interest when interest compounds at a frequency different from annual.
♣

The following result links the EAR with the EPR (and therefore with the APR as well):
Theorem 2.4. Converting from EPR to EAR and vice versa
Given a stated EPR and associated compounding frequency m, the corresponding EAR
equals:
EAR = (1 + EP R)m − 1 (2.6)

Similarly, given a specified EAR and compounding frequency m, the corresponding EPR
equals:
EP R = (1 + EAR)1/m − 1 (2.7)
♡

Note that the second result of Eq. (2.7) follows from the first result of Eq. (2.6) by solving for
EPR on the left-hand side. There is another, equivalent, version of Eq. (2.6) that substitutes the
result EP R = AP R/m, giving:

EAR = (1 + AP R/m)m − 1. (2.8)

This version allows direct conversion from an APR to the corresponding EAR. Finally, the pre-
vious equation can be inverted to give a formula that converts from an APR to the corresponding
EAR:
AP R = m[(1 + EAR)1/m − 1]. (2.9)

Let’s extend Example 2.6 to illustrate how the EAR will generally differ from the APR when
the compounding frequency differs from annual:

Example 2.7 In Example 2.6 we determined that the EPR associated with the auto loan was
4%/12 ≈ 0.3333%. What is the corresponding EAR, and how does this compare with the APR
of 4%?
Solution: Apply the formula of Equation (2.6):

EAR = (1 + EP R)m − 1
= (1.003333)12 − 1
= 4.0742%.

This calculation reveals that the true economic cost of borrowing is about 7.4 basis points higher
than the APR of 4% suggests.
As the compounding frequency m increases, the EAR also increases. Therefore, if the
compounding frequency for the auto loan discussed in the preceding examples were to be weekly
(m = 52) instead of monthly, then the corresponding EAR would be yet higher than 4.0742%. In
2.2 Alternative Compounding Frequencies and the True Cost of Borrowing – 41 –

fact, you should be able to confirm that it would equal approximately 4.0795%. This illustrates
another point concerning EARs and the effects of compounding: although the EAR increases as
the compounding rate increases, the rate at which it does so decreases with m. In other words,
as the compounding frequency becomes large, the EAR does not explode toward infinity, but
rather tends to approach a ‘limiting’ value. This limiting case based on infinitesimally short
compounding periods is called ‘continuous compounding’:
Definition 2.4. Continuous Compounding
Continuous compounding applies an infinitesimally short compounding period for interest.
♣

Continuous compounding involves the ‘exponential constant’ e ≈ 2.718. The reason for
this is given by the following mathematical result:
 r m
lim 1 + = er , (2.10)
m→∞ m
where e is the exponential constant. A key implication of this result is that the EAR associated
with an APR of r% under continuous compounding is equal to er − 1. Let’s apply this result in
the context of the 4% APR from our previous analysis of monthly and weekly compounding. If
we move from weekly compounding to continuous compounding, the EAR becomes:

EAR = er − 1 = e0.04 − 1 = 4.081%

As expected, this is higher than the EAR of 4.0795% associated with an APR of 4% under weekly
compounding, but it is not much higher.
There are also relatively easy formulas for computing PVs and FVs under continuous
compounding. These are:
Theorem 2.5. Continuous Compounding PV and FV
Under continuous compounding at a rate of interest r,

F Vt = P V ert (2.11)

P V = e−rt F Vt , (2.12)

where t is amount of time, measured in year, i.e., t = 1 equals one year from now and
t = 0.5 equals one-half of one year from now.
♡

The following examples show how to apply these results for continuous compounding:

Example 2.8 Your local savings bank is offering a special 5% continuously compounded interest
rate on deposits. Suppose you make a $10,000 deposit today. Assuming no additional deposits
or withdrawals, what will be your account balance in 6 years?
2.2 Alternative Compounding Frequencies and the True Cost of Borrowing – 42 –

Solution: Apply Eq. (2.11) to determine the future value:

F Vt = P V ert
= $10, 000 e0.05×6
= $10, 000 × 1.349858
= $13, 498.59

Example 2.9 You are the beneficiary of a trust fund that will pay you $50,000 in 10.5 years.
What is the present value of this payment if the continuously compounded rate of interest is 6%?
Solution: Apply Eq. (2.12):

P V = e−rt F Vt
= e−0.06×10.5 $50, 000
= 0.5325918 × $50, 000
= $26, 629.59

2.2.2 The True Cost of Borrowing

Lenders can use several tools to manipulate the cost of lending as perceived by borrowers. A
first relates to the compounding frequency and the distinction between APR and EAR discussed
in the preceding section. By specifying an interest rate that compounds at a very high frequency,
such as daily, the lender can understate the true cost of borrowing to the borrower via the
corresponding APR. As we have seen, the EAR will be higher than the APR when compounding
occurs more frequently than annually. For example, it is very common for credit card issuers to
specify underlying rates with daily compounding. Here is a quick example:

Example 2.10 Suppose Capital Two Bank offers a new credit card with a set of travel perks.
As you read over the ‘fine print,’ you notice that the APR on account balances equals 18%.
Assuming that compounding occurs daily, what is the EAR?
Solution: Apply the formula of Equation (2.8) to covert from the stated APR of 18% to the
corresponding EAR. Since compounding is daily, we will set m = 365:

EAR = (1 + AP R/m)m − 1
= (1 + 0.18/365)365 − 1
= 1.000493151365 − 1
= 19.7164%

The EAR is over 19.7% – fully 1.7% higher than the APR suggests. The problem also illustrates
an important practical point when using financial calculators. It is important to maintain a
significant number of decimals when working with interest rates at high frequencies. In this case
the EPR equals a daily rate of around 0.049%, or 0.00049 in decimal form. A calculator showing
2.2 Alternative Compounding Frequencies and the True Cost of Borrowing – 43 –

only three decimals places showing would report this is 0.000!

A second trick involves obfuscating the value of the amount that is being borrowed by
selling the car for more than the ‘cash price.’ Consider the following simple example of this sort:

Example 2.11 You head to a local car dealer to buy a used automobile. The car you want has a
sticker price of $2,000. The dealer offers a 15% discount for a cash deal. You dont have cash,
but you can borrow $1,700 from your uncle at 10% (annual), with the loan repaid in one year.
Alternatively, the dealer offers you 0% financing. You pay nothing today, and then pay the sticker
price in one year. Compute the EAR for both options. Which is cheaper?
Solution: The loan from your uncle has an EAR of 10%. (The interest rate is already annual,
and is therefore an EAR.) Now we need to compute the EAR associated with the second option.
The critical point to recognize is that the dealer has revealed the cash price of the car to be
$1,700, and not the $2,000 sticker price. Consequently the second option is in fact a loan: the
dealer is lending you the cash price value of the car ($1,700) and the loan will be repaid via a
single payment of $2,000 in one year’s time. We can determine the corresponding rate by using
the basic TVM equation, solved for r:
 
F V 1/N
r= −1
PV
$2, 000
= −1
$1, 700
= 17.65%,

where the second line uses the fact that N = 1 for this loan. The bottom line is that it is (much)
better to take out the loan from your uncle.

Remark Although the previous example is extremely simple, there is a realistic element to it.
Many auto dealers offer occasional ‘financing specials’ including ‘0% APR’ loans. How could
it be profitable for a dealer to offer free financing? Banks surely cannot (and do not) do this.
The resolution is that the dealer is selling you a bundle of goods/services: an automobile plus
a loan. The dealer can price one component of the deal more favorably (e.g., the loan) but the
other component less favorably. Indeed, a few months later the dealer will likely not be offering
0% financing but instead a ‘$2,000 rebate’.
Another tool at lenders’ disposal involves various fees that are charged to borrowers in
conjunction with taking out a loan. Many loans are associated with an ‘origination fee’ that
might be on the order of 1% of the loan value but can vary across different lenders. The
origination fee is alternatively known as a ‘processing fee’, ‘underwriting fee,’ or ‘administrative
fee.’ There are also various ‘property fees’ such as appraisal fees and inspection fees. Another
potential source of fees is the ‘mortgage broker’ fee if you work with a mortgage broker to find a
loan. So where does the manipulation come in? Similar to the auto purchase example discussed
2.3 Loans and Amortization – 44 –

above, a lender might quote a low interest rate, but charge relatively large fees at the same time.2
Here is a very simple example to illustrate the idea:

Example 2.12 You need to borrow $20,000 to fund a boat purchase. Suppose that your local
bank will lend you the $20,000 today and the loan will be structured as a 5-year loan at 4%
(annual payments). The bank is highly efficient at loan processing and charges no associated
loan fees. The boat dealership suggests that you instead finance with them, because they are
running a special 3% interest rate deal on 5-year loans. When you go to sign the paperwork,
you discover that the boat dealership charges a 4% ‘origination fee’ to cover ‘various processing
costs’ that is due up-front. What is the actual ‘effective’ interest rate that the boat dealership is
charging? Is it better to finance through the bank or the dealership?
Solution: The local bank loan involves no fees and annual payments, so the 4% rate equals the
relevant EAR for this loan option. It remains to determine the EAR of the alternative financing
plan offered by the boat dealer. The origination fee equals 4% × $20, 000 = $800.00. Because
this is paid up front as a lump sum in return for a loan (today) of $20,000, the effective amount
borrowed equals $20, 000 − $800 = $19, 200. But the 3% rate is applied on the notional loan
amount of $20,000, giving an annual payment of $4,367.09 (N = 5; I/Y=3; PV = -20000; CPT
PMT). Therefore, you have effectively borrowed $19,200 (net of fee) and repay with five annual
payments of 4, 367.09. Using a financial calculator (N = 5; PV = 19200; PMT = -4367.09; CPT
I/Y) gives an EAR of 4.4466%. So it is best to just borrow from you local bank at 4%. Intuitively,
the excessive origination fee more than offsets the 3% advertised rate.

2.3 Loans and Amortization

2.3.1 Types of Loans

Loans and related debt obligations come in many varieties. A basic classification of loan
types is as follows:
1. Discount loan: The structure of a discount loan is very straightforward. A certain amount
P V is borrowed today, and the loan is repaid via a single payment of F V > P V made at
some specified date in the future.
2. Interest-only loan: As the term suggests, an interest-only structure involves the borrower
making only interest payments for a specified period of time. Then, at maturity, the
entire loan principal (P V ) is due. Many government and corporate bonds are effectively
structured as interest-only loans. The bonds involve a ‘face value’ that is repaid at maturity.
This plays the role of the borrowed principal in a loan structure. A series of periodic
‘coupon payments’ play the role of the interest payments in the loan structure. We will
discuss these types of bonds in detail later in the course. Home equity lines of credit

2This practice was common in the ‘subprime loan boom’ of the late 1990s and early 2000s. In some cases origination
fees could be 4–5% of the loan value!
2.3 Loans and Amortization – 45 –

(HELOCs) are also often structured in the form of an interest-only loan. The home owner
can draw on the line of credit, paying only interest charges on borrowed amounts, and the
borrowed principal is due at maturity, which is often 10 years.
3. Amortizing loan: Amortizing loans typically involve regular, fixed dollar payments that
gradually reduce the amount of outstanding principal, such that the loan principal is exactly
paid off at maturity. The process by which the fixed dollar payments gradually reduce the
outstanding loan balance is termed amortization.
There are many subtle variations of loan structures in practice. One relatively common
variation is the partially amortizing loan. This type of loan functions as an amortizing loan
such that borrower makes regular, fixed dollar payments on the loan that gradually reduce the
outstanding loan balance. However, in contrast to a standard amortizing loan, the regular loan
payments do not fully zero out the outstanding balance at maturity. The borrower therefore faces
a ‘balloon’ or ‘bullet’ payment equal to the remaining loan balance at maturity. For these reasons
such loans are sometimes alternatively termed ‘balloon loans’ or ‘bullet loans.’ The rationale
behind this type of loan structure is that the partial amortization affords the borrower a lower
monthly payment. This is potentially preferred for a borrower with relatively low current income
that is expected to grow in the future or a borrower who does not expect to stay in the home for
a long time. If the loan does approach maturity, the borrower can potentially refinance – take
out a new loan in order to retire the remaining balance on the original partially amortizing loan.
However, there is a risk here, particularly if the value of the home has fallen or lending conditions
deteriorate in conjunction with a financial crisis or recession.

2.3.2 Loan Amortization

An important task in loan servicing involves partitioning each loan payment into two
components: 1) the portion of the payment that represents interest charges; and 2) the portion
of the payment that reduces the outstanding principal remaining on the loan. The amortization
process is important for at least two reasons. First, it provides a means for tracking, at each point
in time, the residual principal due on the loan. This is useful, for example, if a borrower wishes
to retire the loan early by paying off all of the outstanding principal. The second reason is that
interest charges some types of loans have important tax implications. The US tax code permits
household to deduct the interest paid on first and second mortgages (home loans secured by a lien
on the property) up to a limit. Consequently, it is important for tax filing purposes to accurately
track the interest paid on such mortgages.
The key steps to partition each loan payment into interest and principal components are:
1. Determine the principal outstanding at the beginning of the period. If this is the first
period, then the starting principal equals the loan amount. Otherwise, you can obtain it in
two alternative ways: 1) the outstanding balance at the end of the previous period (useful
if you are iteratively working through the amortization); or 2) the PV of the remaining
2.3 Loans and Amortization – 46 –

payments on the loan.

2. Compute the interest due as the beginning-of-period outstanding principal times the interest
rate (EPR).
3. Compute the principal paid as the difference between the total loan payment and the
interest paid (the result of the preceding step).
4. Compute the end-of-period principal outstanding as the beginning-of-period balance less
the principal paid (result of step #3). If you are proceeding to the next period, note that
this becomes the new beginning-of-period principal.
As a first illustration of the computations involved in the amortization process, consider the
following simple loan:

Example 2.13 Suppose that a borrower takes out a three-year $150.000 mortgage with annual
payments at a 5% interest rate. The first payment is due in one year and the final payment in
three years.

Figure 2.2: Complete Amortization Table for Example 2.13

Figure 2.13 shows the complete amortization process for this very simple loan. The
computations proceed following the four steps outlined above. For example, for the line in
the table corresponding to the first period, the beginning of period balance equals the starting
the loan amount, $150,000 (note that the beginning of period one is effectively time zero). The
loan payment of $55,018.28 is easily determined using the annuity formula from Chapter 1
(calculator keystrokes: N = 3; I/Y = 5; PV = -150000; CPT PMT). The interest component of
this payment is computed by taking the beginning of period balance times the interest rate, or
$150, 000 × 5% = $7, 500.00. The principal payment is then computed as the total payment less
the previously computed interest component: $55, 018.28 − $7, 500.00 = $47, 581.28. Finally,
the end-of-period balance is computed as the beginning-of-period balance less the principal paid:
$150, 000 − $47, 581.28 = $102, 418.72. This then becomes the beginning-of-period balance
for period 2. You should be able to continue this process and verify the remaining computations
reported in the tabulation.
Now that the algorithm to make the necessary computations is (hopefully) clear, let us
discuss some of the important qualitative features of the amortization process illustrated in
Figure 2.13. First, note that the loan balance exactly zeros out upon the final payment in period 3.
This is precisely what we should expect for a standard amortizing loan assuming that the payment
has been computed correctly. Second, note that although the periodic payment of $55,018.28
is constant each period, the composition of this payment between interest charges and principal
reduction changes each period. In particular, the interest component of each payment decreases
2.3 Loans and Amortization – 47 –

each period, while the principal paid component increases. The intuition behind this pattern is
that, because the loan balance is gradually paid off, each period the interest charge falls due to
the lower remaining balance on the loan.
The 3-year, annual payment structure of the previous example is not particularly realistic.
Most home and consumer loans in practice involve monthly, rather than annual payments. The
good news is that the amortization process is little affected by shifting to payments at the monthly
frequency (or some other frequency). One important detail is that the interest must be computed
using the EPR. For example, for a standard monthly loan the monthly interest expense equals the
monthly EPR (APR/12) times the beginning-of-month loan balance. The following extends the
previous example to feature realistic monthly payments:

Example 2.14 Revisit example 2.13 and now assume that the loan is paid off via monthly
payments, maintaining an APR of 5%, an initial loan balance of $150,000, and a loan term of 3
years (36 months). Show the first six months of the amortization table for this loan.
Solution: The first thing we must do is compute the EPR associated with the loan. The APR
equals 5% and there the EPR equals AP R/12 = 0.05/12 = 0.4167%. The second step
is to compute the monthly loan payment. This can be conveniently accomplished using the
financial calculator (N = 36; I/Y = 0.4167; PV = -150000; CPT PMT), which gives a payment
of $4,495.93 per month. (Note that N =36 because the loan term is 3 years or 36 months.) With
these prelimary steps, it is now straightforward to track the amortization process using the step-
by-step process outlined above. For example, in the first month of the loan, the interest charge
equals $150, 000 × 0.4167% = $625.00. The principal paid is the computed as the total monthly
payment of $4, 495.93 less the interest charge of $625.00, which gives $3,870.63. Finally the
end-of-period balance equals the beginning-of-period balance of $150,000 less the principal paid
of $3,870.63, which equals $146,129.37. This then becomes the beginning-of-period balance for
month 2, and the process continues. Figure 2.14 shows the results of applying these computations
for the first six months of the loan.

Figure 2.3: Amortization Table for Example 2.14

Notice two things about the amortization computations in Figure 2.14. First, it remains
the case that the interest charge component of each loan payment decreases over time, while
the principal paid component increases. Second, the ending balance after 6 months is not zero.
2.4 Financial Planning and Retirement Problems – 48 –

This makes sense, of course, because the loan term is 36 months and not 6 months. Were we to
continue the amortization computations for the remaining 30 months on the loan, we would be
able to verify that the loan balance becomes exactly zero following the final payment.
As a final example, we briefly consider the amortization process for a partially amortizing
loan. In this case, we will see that loan payments gradually reduce the outstanding balance of
the loan, but do not fully zero out the outstanding balance at maturity. The borrower therefore
faces a ‘balloon’ or ‘bullet’ payment equal to the remaining loan balance at maturity.

Example 2.15 Suppose you buy a home in a high-end neighborhood and you take on a $500,000
partially amortizing loan. You have a fixed interest rate of 8.5% (APR, monthly). The bank
agrees to give you a 7-year maturity with a 30-year amortization schedule. Determine the size
of the loan ‘bullet:’ the outstanding loan balance at maturity in 7 years.
Solution: Because the loan features a 30-year amortization schedule, we compute the monthly
loan payment associated with a term that would pay off the loan in 30 years. The EPR is equal
to AP R/m = 0.085/12 = 0.70833%. The monthly payment is then $3,844.57 (N=360; I/Y
= 0.70833; PV=-500000; CPT PMT). How can we then compute the outstanding balance at
maturity in 7-years? We could do a full amortization table through 7 years and then take the final
end-of-period loan balance. This would be correct, but it is very inefficient. An efficient ‘trick’
is to instead recognize that the loan balance at any time is the PV of the remaining payments.
Therefore, all we need to do is compute the PV of the 23 × 12 = 276 remaining payments.
This is relatively straightforward and can be easily done on the financial calculator (N=276;
I/Y=0.70833; PMT = -3844.57; CPT = PV). This gives a loan bullet amount of $465,395.24
It is informative to note that the loan bullet amount after seven years is only around $35,000
lower than the original loan balance of $500,000. This is directly attributable to the lower
payment associated with the partially amortizing structure: most of the payments over the
30-year amortization schedule go toward interest charges over the shorter 7-year term of the loan.

2.4 Financial Planning and Retirement Problems

This section discusses an important application of TVM concepts: financial planning and,
more specifically, retirement planning. This topic will be extremely relevant to those students
pursuing the ‘CFP’ (certified financial planner) track. Our coverage will necessarily only be
introductory and will focus specifically on the issue of important topic of retirement planning.
We will consider only relatively simple situations and ignore some important realistic issues,
especially those pertaining to assessing the risk associated with investments (because we focus
on nominally safe investments) and tax planning issues.
The essence of retirement planning is the specification of a set of (realistic) retirement
wealth or income stream goals, followed by the specification of a savings and investment strat-
egy that, assuming all maintained assumptions prove valid, will achieve the desired retirement
2.4 Financial Planning and Retirement Problems – 49 –

financial goals. Retirement planning inherently involves TVM concepts. For example, the ex-
tent of monthly or annual income that a certain amount of wealth at retirement will support is
fundamentally a TVM computation. Similarly, projections of wealth at retirement involve future
value computations as periodic deposits earn interest over time.
The following example introduces the key elements of retirement and savings problems:

Example 2.16 You are a financial planner assisting an individual with developing a retirement
plan. Your client is aged 50 today and would like to retire at age 65, exactly 15 years from today.
You and your client decide to target a retirement income stream that delivers a purchasing power
equivalent to $75,000 per year in today’s dollars. Retirement income distributions (withdrawals)
are assumed to be received at the end of each year following retirement. The plan calls for the
ability to fund 25 years of post-retirement distributions. Your client has accumulated savings of
$300,000 today. She will save additional money via annual deposits that begin at the end of the
first year (from now). These deposits will grow at an assumed rate of inflation of 2% per year
indefinitely. For retirement planning purposes, it will be assumed that savings deposits earn 7%
interest per year. You assume that the rate of inflation is 2% indefinitely. Given the retirement
planning assumptions, how much should your client save at the end of the first year in order to
exactly meet her retirement goals?
Relevant to problems and examples we have encountered thus far, retirement problems tend
to be relatively complicated. It is critical to ‘break things into pieces’ in order to solve these
problems. Rather than immediately jumping into calculations, it is very worthwhile to read
the problem carefully and determine a solution strategy. For the problem above, the following
analysis should lead us to a solution:
1. Determine the value of the stream of retirement income as of the retirement date (15 years
from today).
2. Determine the value of current savings of $300,000 as of the retirement date in 15 years.
3. Subtract the answer from step #2 from the answer to step #1. This gives the future value
(in 15 years) that the 15 growing savings deposits must exactly equal.
4. Solve for the first year deposit amount that exactly makes the future value in 15 years of
the deposits equal the goal amount determined in step #4.
Now that a valid solution strategy has been determined, it just remains to execute the
necessary computations for each step. These are as follows:
Step 1: Because your client wishes to make regular withdrawals that afford $75,000 in today
dollars, it is necessary to account for the assumed 2% inflation rate, and therefore the annual
withdrawal amounts must grow at a rate of 2% per year. Given that the annual retirements grow
at a constant 2% rate, and since the first withdrawal occur exactly one year from the retirement
date, we can use the formula for the PV of a growing annuity to compute the value need for step
2.4 Financial Planning and Retirement Problems – 50 –

#1. To remind you, this formula is:

 N !
C 1+g
PV = 1−
r−g 1+r
The retirement horizon is 25 years, so N = 25 in applying the formula, r = 7%, and g = 2%. But
what is C? We must be careful here: C must equal $75,000 in today’s purchasing power. So the
withdrawal amount must adjust for all of the inflation that will occur between today and the time
the withdrawal is made. Note that the withdrawal is made 16 periods from today. (Retirement
occurs 15 years from today and by assumption the first withdrawal occurs one year later.)
Therefore, the initial withdrawal amount must be C = $75, 000 × (1.02)16 = $102, 958.93.
Now we can plug in this value and the other inputs to compute the PV of the retirement income
stream:
  !
C 1+g N
PV = 1−
r−g 1+r
  !
$102, 958.93 1.02 25
= 1−
0.07 − 0.02 1.07

= $1, 436, 729.43

This completes step #1 of the solution process. Figure 2.4 shows an illustration of this compu-
tation of the value of retirement needs at the time of retirement, which occurs in 15 years from
today.

Figure 2.4: Retirement needs at retirement date in 15 years

Step 2: This step is simpler. We simply apply the basic FV TVM formula to compute the
value of the initial account balance in 15 years (the retirement date):

F VN = P V (1 + r)N
F V15 = $300, 000(1.07)15
= $827, 709.46

Step 3: This step is also straightforward. We subtract the value from Step 2 from that in
Step 1 to obtain the gap that must be filled by the series of savings deposits we will calibrate in
2.4 Financial Planning and Retirement Problems – 51 –

Step #4:
$1, 436, 729.43 − $827, 709.46 = $609, 019.93.

Step 4: Finally, we set up an equation involving the shortfall determined in step #3 on

the left-hand side, and the future value of an annuity with growth (the series of deposits), as a
function of the initial deposit “C” on the right-hand side. As a reminder, the formula for the
future value of a growing annuity is:
C

FV = (1 + r)N − (1 + g)N
r−g
Then we solve for C:
C

$609, 019.93 = 1.0715 − 1.0215
0.07 − 0.02
C = $21, 548.11,

where the first line uses the above formula for the FV of a growing annuity and the second line
follows after a bit of algebra.
This solves the problem. It can be verified that by making an initial deposit of 21, 548.11,
and subsequent deposits that grow by 2%, our client generates exactly the necessary funds to
support her desired retirement income stream.
There are many variations of retirement planning problems that introduce additional realism
(and complexity). The problem above considered only annual withdrawals and deposits. It is
arguably more realistic to consider higher frequency deposits and withdrawals, such as monthly.
This introduces issues associated with interest rate (and inflation rate) conversions. A good rule
of them is to remember that all TVM computations should be made using EPRs. Of course,
when the underlying frequency is annual, then EPR=EAR=APR and so the three concepts are
equivalent, but in non-annual cases the EPR should be used. The end of chapter problems
(especially problems 16–20) will give you additional practice with solving a variety of types of
retirement problems.
2.4 Financial Planning and Retirement Problems – 52 –

K Chapter 2 Problems k
1. You receive a solicitation to buy a security for $4,000. This security pays a cash flow of
$100 one year from now, with payments each subsequent year that grow at a constant rate
g forever. Suppose the relevant interest rate is 6%. What is the minimum growth rate g
such that you should buy this security?
2. Suppose I promise to pay you 5 payments as follows: $12,000 to be received 7 years from
today, and 4 additional subsequent annual payments that grow at 4% per year thereafter.
If the interest rate is 8.25%, what is this stream of cash flows worth to you today?
3. This problem asks you to confirm that a growing annuity has a well-defined and positive
PV even when the growth rate of cash flows exceeds the discount rate (g > r). Returning
to the Danish government security example, assume the Danish government decides to
offer a security that will pay growing cash flows for a total of 30 years. The security will
pay 200 Kroner in one year, and then amounts that grow at 6% per year thereafter, until
the final payment 30 years from now. What is the value of this security if the discount rate
equals 4%? Hint: Don’t be perturbed if one term in your computation is negative, because
the second term will also be negative, and the product therefore positive!
4. One of your relatives has been offered the opportunity to buy a new form of ’rollercoaster’
contract. The contract is priced at $650,000. Once purchased, the contract guarantees
your relative a set of payments as follows: $35,000 one year from now and subsequent
annual payments that grow at a rate of 6% per year for ten years (so the tenth payment
incorporates 9 periods of growth), followed by annual payments that then decrease at a
certain constant rate per year (call it X%) forever. What is the rate of decrease that makes
the contract price of $650,000 a fair price, assuming that the interest rate equals 4%?
5. You own some land at the edge of Blacksburg. An individual wants to sign a contract where
she rents the land from you for 10 years to operate a golf driving range. She is willing to
pay $20,000 per year (at the end of each year). She will then buy the land from you at
the end of year 10 for $250,000. Alternatively, she is willing to accept a five-year contract
and pay $20,000 in the first year, and then amounts that grow at 5% for the remaining four
years (at the end of each year). In this case, however, she will not buy the land at the end
of the contract. You suspect that in five years, Blacksburg will have grown enough that a
developer will buy your land to build new homes. How much does the developer need to
pay you at the end of five years to make the shorter contract the better option? Assume a
4% interest rate.
6. Compute the EAR associated with an APR of 10% under the following alternative interest
compounding schemes:
(a) quarterly (m = 4)
(b) monthly (m = 12)
(c) daily (m = 365)
2.4 Financial Planning and Retirement Problems – 53 –

(d) continuous (m = ∞)
7. If the EAR associated with a credit card with daily compounding (m = 365) equals 15%,
what is the corresponding APR?
8. Suppose your local bank offers a CD rate of 8% (APR), compounded quarterly. What
continuously compounded rate offers the same economic return?
9. Today you receive an unexpected bonus check for $50,000, which you decide to deposit
in a new bank savings account with the objective of financing a vacation home purchase
in the future. If your target home value is $400,000, how long will it take you to reach
this target assuming no additional deposits or withdrawals assuming the account earns a
continuously compounded interest rate of 7%.
10. Easterwood Credit Corp. wants to earn an effective annual return on its consumer loans
of 16.4 percent per year. The bank uses daily compounding on its loans. What is the APR
that the bank is required by law to report to potential borrowers? Explain why this rate is
misleading to an uninformed borrower.
11. Big Doms Pawn Shop charges an interest rate of 27 percent per month on loans to its
customers. Like all lenders, Big Dom must report an APR to consumers. What rate should
the shop report? What is the effective annual rate?
12. Suppose your uncle sues a manufacturer for a defective product that causes an injury.
The attorney for the manufacturer offers your uncle the following stream of payments as
a settlement of the suit: 40 quarterly payments (4 per year for 10 years) with the first
payment of $5,000 received one quarter from now and subsequent payments growing at a
rate of 1% per quarter. If an appropriate discount rate is 6% (EAR), what is the present
value of this offer?
13. You are looking at a one-year discount loan of $20,000. The interest rate is quoted as
6.25 percent plus 2 points. A point on a loan is 1 percent of the loan amount. (Quotes
similar to this are common with home mortgages.) The interest rate quote in this example
requires the borrower to pay 2 points to the lender up front and repay the loan later with
6.25 percent interest on the $20,000 borrowed. What (effective) rate (EAR) would you
actually be paying?
14. Consider a standard 30-year mortgage loan with initial principal of $400,000, monthly
payments and an APR of 4.8%, and no additional fees.
(a) What is the EAR associated with this loan?
(b) What is the monthly loan payment?
(c) How much of the first monthly payment goes toward interest versus reducing out-
standing principal?
(d) How much of the 36th monthly payment goes toward interest versus reducing out-
standing principal?
15. Your local bank has offered you a 5-year, $100,000 mortgage. If the interest rate is 10%
2.4 Financial Planning and Retirement Problems – 54 –

compounded monthly:
(a). Calculate the monthly mortgage payment.
(b). What is the EAR associated with the loan?
(c). Create an amortization table for the first six months of the loan that shows the amount
of interest you can report for tax purposes each month, as well as the remaining
principal due.
(d). Compute the total interest paid on the loan.
(e). Now suppose the lender charges a $1,000 origination fee for the loan. Assume you
must borrow the same $100,000 net of the fee, i.e., the fee is rolled into the loan.
Compute the new monthly payment and EAR.
16. Determine the bullet payment associated with a partially amortizing loan of $250,000 with
a 30-year amortization schedule (monthly payments) and a 10-year maturity. The APR for
the loan equals 6%. What is the total amount of interest paid on the loan over the 10-year
term?
17. In the fictional land of Westeros, House Lannister vies with rival Houses for control of
the throne. War breaks out. To finance its war operations, House Lannister borrows
1,200,000 gold pieces from the Iron Bank. The loan is structured as a balloon mortgage
with the following details: a term of 15 years, with amortization based on a 30-year loan
with monthly payments. The Iron Bank offers the Lannisters a relatively low APR of 5%
(monthly compounding), because a Lannister always repays his debt.
(a). Determine the monthly loan payment, the EAR associated with the loan and the
remaining principal that will be due in 15 years.
(b). Complete the first two months of the loan amortization schedule for this loan.
(c). Compute the interest paid in the first month of the 12th year of the loan.
(d). Now suppose that the Iron Bank charges an origination fee equal to 1.5%, which is
rolled into the loan value. This means that the Lannisters formally borrow 1,200,000
plus the fee amount but take home only 1,200,000 because the Iron Bank keeps the
fee. Determine the new monthly loan payment and the EAR associated with the loan
when this fee is incorporated.
18. The current debt of the United States government (as of August 2020) is approximately
$26 trillion. Suppose that the government decides to restructure some of this debt. To
accomplish the restructuring, the government will borrow an amount equal to $X today
from the Swiss government (you will solve for X) that will be used to retire outstanding
bonds. The US government will repay this loan as follows. An amount of $1 trillion will
be paid back each year for the next 5 years, with the first payment one year from today.
After this, the government will continue to make annual payments into perpetuity at a
decreasing rate of 10% per year, so the payment in six years will be 10% less than $1
trillion, and so forth. Assuming the interest rate is 4%, how much can the US government
2.4 Financial Planning and Retirement Problems – 55 –

borrow today?
19. You are planning to retire in forty years. You will be spending $25,000 per year, starting
at the end of year 41. Assume you will live for 25 years after retirement and the annual
interest rate is 5%.
(a) What is the present value of your total spending?
(b) Assume that the inflation rate is expected to be 2% per year from now until your
death. If you still want to be able to spend $25,000 in today’s terms (that is, $25,000
worth of goods), what will be the present value of your total spending? Assume the
annual interest rate is still 5%.
(c) You are going to save $3,200 a year for the next forty years. How much money will
you have before retirement? (That is, at the end of year 40)
(d) Will you have enough savings to retire if you assume no inflation (as in part a)? Will
you have enough savings to retire if you assume 2% inflation (as in part b)?
(e) If you have enough savings, how much money will you leave to your heirs? If you
don’t have enough savings, what will be the total deficit at your death? Assume
you will die after your final spending withdrawal at the end of your 25th year of
retirement. Solve for both the part (a) and part (b) spending amounts.
(f) Challenge question: If you are going to run out of money, how many years would
you be able to live before you exhaust your savings? Solve using whichever spending
assumptions had a savings deficit in part (e).
20. Bruce Wayne, AKA Batman, wants to save money to meet three goals. First, Bruce would
like to be able to retire 20 years from now with retirement income of $30,000/month
for 25 years, with the first payment received 20 years and 1 month from now. Second,
he would like to make some upgrades to the Batcave in 8 years at an estimated cost of
$1,500,000. This amount will be paid via a withdrawal from his savings account at that
time. Third, Bruce would like to leave an inheritance of $3,000,000 to Robin at his death,
which we will assume takes place at the end of the 25 years of withdrawals. Bruce currently
has $1,000,000 in savings. Current income from Wayne Enterprises, net of expenditures
including crime-fighting costs in his moonlighting gig as the Caped Crusader, allows Bruce
to save an additional $6,000/month for the next 10 years. If Bruce earns a constant EAR
of 6 percent on his savings deposits, how much will he have to save each month in Years
11 through 20 in order to meet his retirement goals?
21. Old exam question: Mrs. Paye recently expressed to her husband that she would like
to travel more in the future. After some discussion, the Payes planned trips to Australia,
Brazil and Norway. Suppose each trip costs $8,000. (Although it is somewhat unrealistic,
ignore inflation in this problem.) The Payes will make their first trip eight years from now,
and their second and third trips 10 and 12 years from now, respectively. They also wish
to have $10,000 available following their third trip (in 12 years) to fund additional future
2.4 Financial Planning and Retirement Problems – 56 –

travel. Mr. Paye remarked to his wife that travel funding would be easier if she agreed
to give up her daily Double Gulp Diet Coke purchase from 7-11. Mrs. Paye retorted that
Mr. Paye could also go without his weekly 6-pack of artisanal beer. Assume each follows
through, instead drinking (free) water and saving their foregone beverage funds. Assume
a Double Gulp costs $2.50, while beer costs $14.00/6-pack. Mrs. Paye will forgo her first
Double Gulp tomorrow, and Mr. Paye will forgo his first 6-pack in one week. They will
resume normal Diet Coke and beer consumption in twelve years. The relevant savings
account rate is 8% (APR) with daily compounding. Will the Payes dietary adjustments
finance their desired travel expenditures? Determine the present value of any shortfall
or excess. You can pretend there are exactly 52 weeks and 365 days per year in your
calculations, even though this is not precisely true.
22. old exam question Batman redux: Bruce Wayne, AKA Batman, wants to save money
to meet two goals. First, Bruce would like to be able to retire 30 years from now with
retirement income of $50,000 / month for 20 years, with the first payment received 30 years
and 1 month from now. Second, Bruce would like to leave an inheritance of $5,000,000
to Robin at his death, which we will assume takes place at the end of the 20 years of
withdrawals. Bruce currently has $500,000 in savings. Current income from Wayne
Enterprises, net of expenditures including crime-fighting costs in his moonlighting gig as
the Caped Crusader, allows Bruce to save $4,000/month until retirement. Assume that
Bruce earns a constant EAR of 5 percent on his savings deposits.
(a) Compute the amount of money Bruce will require in his retirement account at his
retirement date in order to meet his retirement goals.
(b) Given Bruces current savings and deposit plans, determine how much savings he
will have at retirement. What is the present value of any excess or shortfall relative
to what is needed (as determined in part a)?
(c) Suppose that at some future point prior to retirement Batman decides to increase his
savings deposits to $12,000 / month. At what time (how many months from now)
will he need to increase his deposits to $12,000 / month such that he can exactly meet
his retirement goals?
23. Old exam problem: You plan to retire in 25 years. Prior to retirement, you will make
regular quarterly deposits as follows: $3,000 next quarter, and subsequent deposits that
grow at the rate of inflation, with the final deposit made 25 years from today. Following
retirement, you will begin withdrawing funds. The first withdraw will occur one year after
retirement and you will make exactly 20 total withdraws. Each withdrawal will be for an
equal amount in terms of todays purchasing power and will cover all living expenses, save
the trip discussed next. You plan to take a big trip during retirement. The trip cost is equal
to $25,000 in todays dollars and will be paid at retirement. The EAR is 8% and the annual
inflation rate is 2%. Determine the amount of the first annual withdrawal that you will
2.4 Financial Planning and Retirement Problems – 57 –

make during retirement (in dollars as of that time).

24. How much would parents need to save on an end-of-month basis if their son was going to
begin college exactly 8 years from today, they want to cover 100% of his expenses, he will
complete school in 4 years, current costs of college total $20,000 per year, college costs
grow at the annual rate of 5%, and the annual rate of return on their investments is 10%.
Assume that they will start saving one month from today and will continue to save each
month up to and including the date on which his first college payment is due. Assume that
their savings will also grow as well at 5%. Assume that both rates (growth and return) are
EARs. Assume that you will pay college expenses at the beginning of each year in full for
the year with the first payment coming 8 years from today.
25. Old exam question: You are planning to retire in ten years. You want to be able to spend
$15,000 per quarter in retirement, adjusted for inflation (so you will spend the equivalent
in each quarter of $15,000 in todays dollars). You will withdraw your retirement spending
at the end of each quarter. So your first withdrawal will be at the end of the first quarter of
year 11. Assume you will live for 20 years after retirement and the effective annual interest
rate (EAR) is 8% for the entire problem.
(a). How much is the first retirement withdrawal at the end of the first quarter of year 11
if the inflation rate is 0.5% per quarter?
(b). What is the present value of all your retirement spending? Continue to assume
inflation is 0.5% per quarter and give the PV as of today (year 0).
(c). Before retirement, you are planning to contribute to your savings once a year for the
next 10 years, at the end of each year. You will increase your savings contribution
by 6% each year, and currently already have $250,000 in retirement savings. How
much do you need contribute this year to afford your retirement plans?
(d). You decide to live a more lavish lifestyle in retirement and will withdraw 20% more
each quarter (so whatever you were initially planning to withdraw for a given quarter,
you withdraw 20% more than that). How many years can you do this before you
exhaust your funds and need to sponge off your adult children?
 Chapter 3 Interest Rates and Bonds

This chapter concerns interest rates offered by financial securities and, in particular, bonds
and related fixed income securities. These are an important class of securities in the so-called
‘capital markets’: exchange marketplaces where financial securities and assets are bought and
sold. The pricing of fixed income securities relates closely to the time value of money concepts
introduced in the previous two chapters.
Section 3.1 provides some background concerning the determination of market interest
rates and the connection between real and nominal interest rates. Section 3.2 describes key
institutional features concerning bonds. Section 3.3 covers bond prices, returns, and yields –
implicit interest rates offered by bonds. Section 3.4 discusses the return earned by investors
who hold bonds and how this relates to the bond yield. Section 3.5 discusses interest rate risk
associated with bonds. Section 3.6 introduces the ‘term structure of interest rates,’ which refers to
the relation between (government) bond yields and bond maturity. Section 3.7 introduces several
special types of US Treasury bonds, including bonds that offer a form of protection against
inflation. Finally, Section 3.8 briefly considers other aspects of bonds that contribute to prices
and yields, including potential default risk, differential tax treatment, and liquidity conditions.
NOTE: Sections 3.4–3.8 will be added in the next version of the PDF.

3.1 Interest rates

There are many forms of lending in a modern economy. For example, households lend
money to banks and related financial institutions when they deposit paychecks and other funds
in forms of checking and savings accounts. Banks and financial institutions, in turn, lend money
to households for purchases of homes and consumer durable goods (e.g., automobiles) and to
firms to finance investment and operations. Financial institutions also borrow and lend money
amongst each other on a short-term basis. National, state and local governments who wish to
spend more than they take in via taxes fund deficits by borrowing from the public via the issuance
of bonds, a form of debt security that, as we shall see, functions similarly to a loan. Likewise,
many corporations issue bonds to finance investment activity.
Definition 3.1. interest rate
An interest rate is the cost of borrowing as asset (often dollars) over a period of time,
expressed as a percentage of the amount borrowed.
♣

Due to large number of ‘lending’ activities in the economy, it is somewhat misleading to

talk about ‘the interest rate.’ However, it is common to broadly refer to ‘interest rates’ and in this
context we usually mean the general average level of interest rates in the economy.
3.1 Interest rates – 59 –

3.1.1 Facts about US Interest Rates

It is useful to begin by discussing the historical behavior of interest rates in the US. Figures
3.1 and 3.2 show two alternative ‘benchmark’ interest rate measures. Figure 3.1 shows the
so-called ‘prime rate’ over the period 1965–2020. This refers to the interest rate that banks
charge for corporate loans with the best category of credit rating, i.e., the strongest likelihood
of repaying loans in a timely fashion. The Wall Street Journal collects this information from
the largest US banks and publishes periodic revisions to the rate. The rate is important because
it serves as the benchmark for many loans with adjustable rates, such as most credit cards, etc.
Figure 3.2 shows the ‘yield’ - a form of implied interest rate – on 10-year US Treasury bonds
over the same period.

Figure 3.1: Historical Prime Rate in the United States

Figure 3.2: Historical 10-year Treasury Yield

Note the following key insights from these two figures:

1. There is a lot of historical variability in US interest rates. Current rates are now historical
lows for the modern economic era. Rates peaked in the early 1980s at levels of around
20% (prime rate) and 15% (10-year Treasury rate).
2. The grey region of the figures show recessions (economic downturns). Interest rates
typically fall in recessions. For example the prime rate fell very recently with the COVID
outbreak and ensuing recession, as well as in the financial crisis period. Rates then tend to
increase somewhat when the economy exists a recession. This implies that interest rates
tend to have a pro-cyclical component.
3. There is a lot of commonality in the long-run movement between different types of interest
rates. The general arc of interest rates is very similar for both the prime rate and 10-year
3.1 Interest rates – 60 –

Treasury rate in Figures 3.1 and 3.2.

4. Although different interest rates tend to move together, the relation is not perfect. First, at
a given point in time different types of interest rates will exhibit different levels. In other
words, at a given point in time the prime rate and 10-year Treasury rate are likely to differ
to some extent. Second, different interest rates can behave differently over the business
cycle. For example, the relative drop in the 10-year Treasury rate in the financial crisis is
considerably smaller than the drop in the prime rate.
In the next few sub-sections, we will gain an understanding concerning why interest rate
data exhibit the above features. We will primarily focus on observations 1–3 above in this section.
Later in the chapter, we will discuss in further depth why differences exist across alternative types
of interest rates.

3.1.2 Interest Rate Determination: Economic Foundations

What determines the interest rate? Why is it 3% and not 30%? A first insight is that the
interest rate is a form of price, specifically, an interest reflects the cost to borrowers of obtain
extra resources today, and simultaneously the benefit, in the form of additional future resources,
to lenders for providing excess resources today. Given that interest rates are forms of prices, it is
natural to think about the determination of these prices just as we think about the determination
of prices that obtain in many economic markets. What determines the price of corn, gasoline,
for example? The classic competitive market model of microeconomics holds that prices are
determined by the forces of supply and demand.
Let us apply the competitive market model in the context of lending. We will first analyze a
natural setting in which the lending takes place in dollars. The price in this dollar-based lending
market is called the nominal interest rate. We will keep things abstract, in the sense that we
will not specify explicitly a lending horizon other other details of the lending arrangement. The
‘good’ in this market is often termed loanable funds: dollars that are lent/borrowed.
Consider first the demand for loanable funds. The following entities contribute to the
demand for loanable dollars:
1. Households: borrow to finance purchases of homes or durable goods, credit card borrow-
ing, etc.
2. Firms: seek to borrow to fund investments or to cover shorter term obligations
3. Governments: seek to borrow to fund expenditures that exceed tax income
Now consider the supply side of the market. What entities are willing to potentially lend dollars?
1. Households: some households spend less than income and are potentially willing to lend
the difference
2. Firms: Occasionally firms might have excess cash that they are willing to lend
3. Governments: Sometimes governments run a surplus, such that tax receipts exceed expen-
ditures and these governments are willing to lend funds
3.1 Interest rates – 61 –

Note that (different) households, firms, and governments operate on both sides of the lending
market.
Figure 3.3 shows a standard ‘microeconomics 101’ supply and demand picture for the dollar
lending market. There is a downward sloping demand curve and an upward sloping supply curve
for loanable funds that intersect at exactly one point. The demand curve reflects the desired
borrowing of households, firms, and governments at different potential levels of the market price,
which is the nominal interest rate. All else equal, a higher nominal interest rate reduces the
demand for borrowing dollars, and this explains the downward slope of the curve. The supply
curve depicts the quantity of dollars households, firms, and the government are willing to supply
at each possible level of the nominal interest rate. The curve slopes upward because these entities
desire to lend more when the nominal interest rate they earn for lending increases. The model
implies that the stable or ‘equilibrium’ nominal interest rate obtains at the point where supply
equals demand: at exactly this nominal interest rate there is neither excess supply nor excess
demand of loanable funds. The graph depicts this equilibrium nominal interest rate (re ) as well
as the corresponding equilibrium quantity of dollars lent (Qe ).

Figure 3.3: The nominal interest rate as an equilibrium price in the loanable funds market.

So far we have reasoned that standard forces of supply and demand determine the level of
the interest rate. But this will not help us explain the variation in interest rates we see in Figures
3.1 and 3.2. It is therefore important to think about the forces that causes changes in interest
rates. Again referring to the traditional competitive market model, changes in equilibrium prices
occur when supply and demand conditions change. In particular, the equilibrium nominal rate
will change when there is a shift to the demand or supply curve for loanable funds. Consider
the following concrete example. Suppose that, all else equal, there is a positive technological
3.1 Interest rates – 62 –

shock that increases firms’ productivity in the economy. Firms would like to obtain additional
resources to take advantage of this productivity increase. Consequently, the demand curve for
loanable funds shifts outward (to the right). This shift delivers an upward movement in the
equilibrium nominal interest rate, as well as an increase in the equilibrium quantity of loanable
dollars. Figure 3.4 below illustrates this shift and the corresponding effects on the equilibrium
interest rate and quantity of lending. In contrast, a reduction in firms’ productivity, all else equal,
would decrease the equilibrium nominal interest rate and associated quantity of loaned funds.
Note that shifts in productivity are one potential source of business cycle fluctuation. Insofar as
technology shocks drive the business cycle, therefore, the previous reasoning helps explain why
interest rates tend to behave pro-cyclically.

Figure 3.4: An outward shift in the demand for loanable funds raises the equilibrium nominal interest
rate.

There are, of course, many other potential sources for shifts in the supply and demand curves
in the lending model. The chapter end problems ask you to think through several other types of
market shocks and the likely impact on equilibrium nominal interest rates.

3.1.3 Real versus Nominal Rates and the Fisher Effect

The nominal interest rate refers to the cost, measured in dollars, of borrowing in dollars.
This will be natural to most readers, as it characterizes most loans in practice. Let us imagine,
however, an alternative form of lending that is dominated not in dollars, but rather in underlying
consumption goods. This is somewhat abstract and it might be helpful to imagine the borrowing
and lending of ‘baskets of goods and services.’ Let us imagine a competitive borrowing and
lending market for these baskets. This essentially amounts to relabeling the y- and x-axes in the
standard market model we developed above. The y-axis now becomes the price associated with
3.1 Interest rates – 63 –

borrowing consumption goods. We call this the real interest rate (rreal ). The x-axis becomes
‘loanable purchasing power.’ Figure 3.5 below illustrates the real interest rate as the equilibrium
price in the market for loanable purchasing power.

Figure 3.5: The real interest rate as the equilibrium price in the market for loanable purchasing power.

What causes changes in the real interest? Similar to the micro-economic model for the
nominal interest rate, shifts in either the supply or demand for loanable purchasing power causes
equilibrium real rate changes. Many of the same factors that influence shifts in the demand or
supply of loanable funds also drive similar shifts in the demand and supply of loanable purchasing
power. For example, a reduction in government taxes increases household resources both in the
form of dollars and in the form of purchasing power. This would be naturally modeled as an
outward shift in the supply loanable funds and loanable purchasing power. Both the equilibrium
real and nominal interest rates decrease, and the equilibrium quantity of lending increases. This
example illustrates that the real and nominal interest rate are far from disconnected objects –
they are related prices driven by many of the same underlying market forces. What then, is the
distinction? The main distinction involves inflation.
Unexpected inflation harms lenders by decreasing the real return to lending. Consider the
following stark example that illustrates the main idea:

Example 3.1 Suppose that in recent history there has been no inflation. A lender and borrower
agree to a one-year, dollar-denominated discount loan with an interest rate of 5%. Over the next
year, prices unexpectedly increase by 10%. What is the real rate of return earned by the lender?

Solution: The loan involves a nominal interest rate of 5%. Therefore, in the benchmark case
in which prices do not change, the lender would be repaid 5% over the principal in interest and
could purchase 5% more in goods and services. However, when prices increase by 10%, the
3.1 Interest rates – 64 –

lender experiences a significant negative real return. We can quantify this using the following
key equation that links real and nominal returns:

Real return = nominal return − inflation rate

= 5% − 10%
= −5%

Although the unexpected inflation is not welcome to the lender, note that it benefits the borrower,
who is able to repay the (nominal) interest and principal with less valuable future dollars. As a
second remark, imagine that instead there was unexpected deflation of 5%: in this case the price
level falls by 5%. This unexpected deflation has the reverse effect, benefiting the lender and
hurting the borrower, who must now repay with more valuable dollars. Unpredictable variation
in the inflation rate is therefore a risk factor associated with nominal debt obligations. Note that
this risk factor would not be present in a debt obligation denoted in ‘real’ terms. Later we will
talk about special types of bonds that seek to eliminate the risk associated with unpredictable
variation in inflation.
Some inflation is predictable. After all, prices tend to rise gradually over time. In addition,
central banks such as the Federal Reserve maintain targets for the inflation rate and they sometimes
take steps to either dampen or accelerate inflation when it deviates from the target rate. Finally,
inflation tends to be persistent, and households are generally aware of recent trends in inflation.
In the early 20th century, an economist by the name of Irving Fisher recognized that
predictable inflation influences nominal interest rates. The logic is relatively straightforward:
lenders are intelligent and they will make reasonable forecasts or expectations of future inflation
rates. They will therefore want to charge a nominal interest rate that still leaves them with
a reasonable real return if the predicted level of inflation occurs. With this basic insight as
motivation, we can arrive at a simple quantitative connection that links nominal and real interest
rates via expected inflation. Start with a simple equation recognizing that the gross real interest
rate implied in a dollar denominated lending transaction equals the ratio of the gross nominal
rate to the gross expected rate of inflation:
1+r
1 + rreal = ,
1 + ∆Pe
where ∆Pe equals expected inflation. (∆ means change, P means price level, and the ‘e’ stands
for ‘expected’, i.e., the expected change in the price level.) Multiplying both sides and doing a
bit of algebra gives:

1 + r = (1 + rreal )(1 + ∆Pe ) (3.1)

= 1 + rreal + ∆Pe + rreal ∆Pe (3.2)

The last term on the right-hand side of the above equation is the product of the real interest rate
and the expected rate of inflation. For typical or ‘reasonable’ values for these rates (at least for
developed countries), the product is quite small. As a useful approximation, we can ignore this
3.1 Interest rates – 65 –

term, which gives us the following important (and very simple) result:
Theorem 3.1. Basic Fisher Equation
r ≈ rreal + ∆Pe
♡

This basic form of the Fisher equation states that the nominal rate is approximately equal
to the sum of the real rate and predicted inflation. Here is a simple example that applies the
equation:

Example 3.2 Suppose that the real interest rate equals 5% per year and market participants expect
a constant inflation rate of 2% for the foreseeable future. What should be the nominal interest
rate accouring to the Fisher equation?

Solution: Plugging 5% and 2% into the Fisher equation for rreal and ∆Pe , respectively gives
a predicted nominal interest rate of 5% + 2% = 7%. This uses the simple form of the Fisher
equation. What if we used the exact form of the Eq. (3.2)? In this case we would get 7% +
(0.05)(0.02) = 0.001, or 7.1%, a slightly higher value.
The equation implies that the level of the nominal interest rate will move directly with
changes in the level of expected inflation. This is an important insight that helps us understand
the historical pattern of interest rates in the US. In particular, the US experienced oil price shocks
in the mid 1970s that caused a significant amount of inflation. US inflation rates during the late
1970s were considerably higher than they had been historically. Households and firms began
to expect continued higher inflation in the future. The so-called Fisher effect dictates that these
inflationary expectations would lead to higher nominal interest rates. This explains why nominal
interest rates were so high (relatively) during this period. Ultimately, the Federal reserve took
certain steps to stamp out the inflation (we will discuss this in the next chapter), and this then
explains why interest rates began to fall again after the early 1980s.
The Fisher effect relates to how predictions of future inflation influence nominal interest
rates. But then the future actually occurs. Unfortunately, actual or realized inflation often deviates
to some degree from predictions. There can be ‘surprise’ inflation that exceeds expectations.
Alternatively, the surprise might be less inflation than expected. These deviations of actual
inflation from expectations influence the real rates of return earned by lenders and (equivalently)
the real rates paid by borrowers. To capture this idea, the ‘realized real rate of return’ (RRRR)
is defined as the nominal interest rate (set in advance) less the actual or realized rate of inflation:

Realized real rate of return = r − ∆Pa (3.3)

where ∆Pa denotes the actual rate of inflation that occurs over the time period corresponding to
the nominal rate. The following example illustrates:

Example 3.3 Suppose that the real rate of interest equals 3% and expected inflation equals 2%.
A borrower and lender agree on a one-year loan at a 5% nominal rate of interest (consistent with
the Fisher equation). Determine the realized real rate of return in the following three scenarios
3.2 Bonds: Key Institutional Details – 66 –

for actual inflation over the coming year: 1) inflation is 2%; 2) inflation is 6%; 3) inflation is
-2%.
Solution: In each case, we can apply the formula for the RRRR:

Case 1: RRRR = r − ∆Pa = 5% − 2% = 3%

Case 2: RRRR = r − ∆Pa = 5% − 6% = −1%
Case 3: RRRR = r − ∆Pa = 5% − (−1%) = 6%

The calculations show that, in the first case when actual inflation equals the expected amount
(there is no surprise inflation), then the RRRR equals the ex ante real rate. In case 2, there is
considerable surprise inflation and the realized real rate of return is actually negative. This is bad
news for lenders but good news for borrowers who end up paying a negative real rate. Finally, in
case 3 there is actually deflation, which benefits lenders but hurts borrowers.
We conclude this section by briefly reconsidering the historical US interest rate data in
Figures 3.1 and 3.2. The Fisher equation and economic model for interest rate determination
give us useful tools to interpret this history. Nominal interest rates fluctuate as economic forces
shift the supply and demand for loanable funds. One component of these fluctuations involves
factors that influence the underlying real interest rate, which fluctuates in a generally pro-cyclical
manner. In addition, significant time-variation in expected inflation over the past six decades
of US economic history have produced low-frequency variation in nominal interest rates, which
peaked in the early 1980s and then began to gradually fall as the relatively high US inflation
of the late 1970s and early 1980s subsided. Different interest rates, e.g., savings account rates
and auto loan rates, generally tend to move together because they share common exposure to the
underlying economic forces driving real rates and inflation in the economy.

3.2 Bonds: Key Institutional Details

When governments or companies wish to borrow money, they often issue forms of financial
securities called bonds:
Definition 3.2. Bonds
Debt securities issued (sold) to the public by governments, companies, and related orga-
nizations. ♣

The government or company that sells the bonds to the public to raise funds is called the
bond issuer or simply the issuer. Those who buy and own the bonds are called bondholders.
Many bonds function as a form of interest-only loan (see Section 2.3 of the previous chapter).
The issuer acts in the role of borrower and the investors who buy the bond at issuance act in
the role of lender. Typically, the bond contract calls for the issuer to make a series of periodic
payments to the bondholders called coupons that act as a form of interest payment. Then, at a
3.2 Bonds: Key Institutional Details – 67 –

specified date called the bond maturity, the issuers pay the bondholder the principal amount of
the bond. This principal amount is often called the face value or par value of the bond.
Although bond contracts can include many features, some of which are rather complicated,
the minimal essential bond contract details include:
1. Face value: The principal amount of the bond. The face value amount varies across
different types of bonds and different issuers, but is often a round value such as $100,
$1,000 or $5,000.
2. Maturity: The date at which the bond matures and the face value is paid to the bondholder.
Bonds differ widely with respect to maturity. For example, the US government issues
bonds with maturities ranging from 30 days to 30 years.
3. Coupon and coupon rate: The periodic interest payments made by the bondholder. This
is typically specified in the bond contract as a coupon rate expressed as an annualized
percentage of the face value. For example, a coupon rate of 5% means that the bondholder
is paid 5% of the face value as a coupon per year.
4. Coupon frequency: The number of times per year that coupons are paid. This again
varies across bond types. Bonds issued by the US government typically pay semi-annual
coupons (twice per year). Some corporate bonds pay quarterly coupons. It is important to
understand that the coupon rate typically acts like an APR (see Section 2.2). If a bond pays
a 5% coupon semi-annually, this means that the bondholder will receive two payments per
year (at six month intervals) and each payment will be for 5%/2 = 2.5% of the face value
of the bond.
It is important to recognize that the aforementioned key bond contract details are fixed over
the life of the bond, at least for the types of bonds we will examine in this course. in other words,
the bond face value, maturity date, coupon rate and coupon frequency are determined at the time
the bond is issued and, once set, do not change over the life of the bond.
Consider the following stylized example of a bond:

Example 3.4 The bond contract specifies a 3-year maturity and annual coupon payments with
a coupon rate of 5%. The bond contract also specifies that the par value of the bond equals
$1,000. Suppose you purchase this bond at issuance and hold the bond until maturity. Draw a
time line that indicates the cash flow this investment will produce, assume that all promised bond
payments are made in a complete and timely fashion.

Solution: The coupon rate equals 5%. Multiplying the $1,000 par value of the bond by this
amount gives an annual coupon payment of $50. Since the coupon payment convention is annual,
the bond will pay 3 total coupons of $50. The first will be paid one year from issuance, the
second two years from issuance, and the final coupon will be paid three years from issues. This
is also the bond maturity date, at which point the investor receives the par value of $1,000 in
addition to the final coupon. The timeline is therefore as depicted in Figure 3.6.
What happens if the coupon frequency is not annual? To illustrate, let us consider the follow-
3.2 Bonds: Key Institutional Details – 68 –

Figure 3.6: Cash flows associated with a 3-year bond with $1,000 par value, 5% coupon rate and annual
payments.

ing adaptation of the previous example that assumes a semiannual (twice per year) compounding
frequency:

Example 3.5 Consider the same bond as in Example 3.4, except that the coupon is paid semian-
nually. Show a timeline of the cash flow received by an investor who buys the bond at issuance
and holds the bond to maturity, again assuming that all bond payments are made in a timely
fashion.

Solution: The annual coupon amount remains equal to 5% × $1, 000 = $50 as in the previous
example. However, this annual payment is now split equally into two payments of 25, once
received halfway through the year and the other at the end of the year. This occurs for each of the
3 years in the life of the bond. The par value of $1,000 is still received at maturity (in 3 years)
just as in the previous example. The new timeline for cash flows appears in Figure 3.7.

Figure 3.7: Cash flows associated with a 3-year bond with $1,000 par value, 5% coupon rate and
semiannual payments.

Some corporate bonds (bonds issued by corporations as opposed to governments) entail

quarterly coupon payments. In this case the annual coupon amount is again computed as the
product of the coupon rate and the par value, and this amount is then split equally into 4 quarterly
coupon payments to bondholders.
Remark Note that the coupon rate functions as an ‘APR’ (annual percentage rate, see Section
2.3). When the coupon frequency is quarterly, for example, the quarterly coupon rate equals the
annual coupon rate divided by 4. This is exactly the computation of a quarterly EPR given an
APR equal to the annual coupon rate.
Bonds are originally issued to investors in what is commonly termed the primary market:
3.3 Bond Prices and Yields – 69 –

Definition 3.3. Primary Markets

Markets in which firms, governments and related entities sell (float) new stocks, bonds, or
other financial securities to the public for the first time.
♣

Suppose, for example, that Boeing decides to raise additional funds for investment purposes
by issuing bonds. It is likely that Boeing would retain the services of an investment bank, such
as Goldman Sachs, to help them navigate the bond issuance process. (Goldman Sachs plays
the role of an ‘intermediary’ in the issuance process and earns compensation the form of fees,
etc.) The investment bank would help Boeing determine the appropriate set of bond contract
details (maturity, face value, and many other contract details). Then, the investment bank would
help Boeing market the new bonds, primarily to large financial institutions such as pension and
mutual funds. Ultimately, the new bonds would be ‘issued’ to some set of investors and Boeing
would receive proceeds based on the final issuance price and number of bonds issued. (This is
just a sketch of the process, but it will suffice for this course.)
Once bonds have been issued, bondholders often wish to sell some or all of their holdings
prior to maturity. This accomplished via the so-called secondary market:
Definition 3.4. Secondary Markets
Markets in which existing (previously issued) financial securities such as stocks and bonds
are traded (bought and sold by investors).
♣

We will not discuss in detail the nature of the secondary markets for government and
corporate bonds in this course. Most, although not all, bonds trade “over-the-counter,” which
means that they are traded among bond broker-dealers acting on their clients’ or their own behalf.
In Section 3.8, we will return to this topic when we discuss how the relative ease of buying and
selling a bond is likely to impact its pricing.
Most data and discussion in the financial press pertaining to bonds relates to the secondary
markets. Regulatory law requires the disclosure of much information concerning corporate bond
transactions in these markets. As an illustration, Figure 3.8 shows data published by FINRA (a
financial regulatory agency) for a set of bonds issued by the pharmacy company CVS Health
Corp. The information includes many details concerning the underlying bond contract (maturity,
coupon rate, etc.) as well as price and yield information based on trades in the secondary market.1

3.3 Bond Prices and Yields

This section addresses how prices for bonds are determined and introduces the concept of
a bond yield, which essentially reflects the implied interest rate being offered by a bond at a
particular point in time.

1Data source: https://finra-markets.morningstar.com/BondCenter/BondDetail.jsp?ticker=

C744753.
3.3 Bond Prices and Yields – 70 –

Figure 3.8: Real-world bond data for bonds issued by CVS Health Corp.

3.3.1 Bond prices as present values

As the previous section indicates, a standard bond can be viewed as a security that promises
a set of specified cash flows are specified times in the future. Consequently, the bond price
should equal the present value of the cash flows that the bond contract promises the bondholder.
As a concrete example, consider the basic 3-year bond with a 5% coupon rate (annual payments)
from Example 3.4. Suppose that we additionally assume that the relevant discount rate for the
cash flows promised by this bond equals 4% (per year). What is the correct or ‘fair’ price for this
bond? To determine this, we can simply apply the time value of money techniques introduced in
Chapters 1 and 2 to compute the present value (PV) of the bond’s cash flows. As a reminder, the
bond pays $50 cash flows one, two, and three years from now and the investor also receives the
$1,000 par or face value at maturity in three years (see Figure 3.6).

$50 $50 $1, 050

Bond price (P) = + 2
+ (3.4)
1.04 1.04 1.043
X $50
3
$1, 000
= n
+
1.04 1.043
n=1
 
$50 1 $1, 000
= 1− 3
+
0.04 1.04 1.043
= $1, 027.75.
3.3 Bond Prices and Yields – 71 –

The first line of Eq. (3.4) states that the bond price equals the present value of the bond cash
flows, where the latter is written out using the ‘brute-force’ approach: we discount each separate
cash flow back to today using the 4% discount rate applied for the number of periods we must wait
until the corresponding cash flow is received. The second line just writes this more efficiently
using summation notation. The third line recognizes an important insight: the coupon payments
associated with the bond take the form of a 3-year standard annuity. Therefore we can apply the
formula for the PV of an annuity to value the coupons (and then we must tack on the discounted
par value). The last line states that all of these approaches equivalently give the bond price as
equal to $1,207.75. Note further that we could easily obtain this bond price using a financial
calculator (N = 3; PMT = -50; FV = -1000; I/Y = 4; CPT PV).
What happens if the bond pays coupons at a frequency other than annual? Consider the
slightly ‘fancier’ version of the bond in Example Y that pays the 5% coupon semi-annually.
Continuing to assume a 4% discount rate, what is the price of this alternative version of the
bond? To determine this, consult the timeline of cash flows for the bond depicted in Figure 3.7.
It should be clear that we can view the bond as providing an annuity consisting of six semi-annual
payments of $25 plus the par value of $1,000 received in three years’ time. We can therefore use
a technique similar to that for the basic (annual) bond to compute the price. But what should
be the discount rate r for these calculations? We cannot use 4%, because this is an annual rate,
whereas the cash flows are received semi-annually. Let us treat the 4% annual rate as an APR
(whether it is an EAR or APR must be clarified). Then the corresponding EPR is computed
simply as EP R = AP R/m = 4%/2 = 2%. With this in hand, we can now compute the bond
price as:

X 6
$25 $1, 000
Bond price (P) = n
+ (3.5)
1.02 1.026
n=1
 
$25 1 $1, 000
= 1− +
0.02 1.026 1.026
= $1, 028.01.

One can also obtain the price using a financial calculator (N = 6; PMT = -25; PV = -1000; I/Y =
2; CPT PV).
Notice that the price (present value) we obtain in this case is just slightly higher than that
obtained for the version of the bond with annual payments. This makes complete sense from a
time value of money perspective: the only difference between the bond cash flow patters is that
the semi-annual bond offers slightly accelerated (in time) coupon payments. It is this slight time
value effect that creates the price difference between the two bonds. The computational approach
for these examples can be generalized to apply to any standard bond. The following gives the
general result:
3.3 Bond Prices and Yields – 72 –

Theorem 3.2. Bond Pricing Formula

Consider a standard bond with N years to maturity, an annual coupon payment C =
c × F V , where c is the (annual) coupon rate and F V is the face or par value of the bond,
paying coupons m times per year. Assuming that the bond has either just been issued or
that the most recent coupon has just been paid, the price of the bond is given by:
 
C/m 1 FV
P = 1− mN
+ (3.6)
r/m (1 + r/m) (1 + r/m)mN
♡

For practice, let us apply this result to a somewhat realistic example:

Example 3.6 Suppose that a US Treasury bond currently has 28.5 years to maturity. The bond
has a coupon rate of 3.5%, paid semi-annually, and a face value of $10,000. If the discount rate
for the bond is 2.5% (APR), what is the price of the bond?

Solution: We can apply Eq. (3.6) with r = 2.5%, m = 2 because the bond is semiannual,
c = 3.5%, F V = $10, 000, and N = 28.5. Noting that the annual coupon payment is
C = c × F V = 3.5% × $10, 000 = $350, we obtain:
 
C/m 1 FV
P = 1− mN
+
r/m (1 + r/m) (1 + r/m)mN
 
350/2 1 $10, 000
= 1− +
0.025/2 (1 + 0.025/2)2×28.5 (1 + 0.025/2)2×28.5
 
175 1 $10, 000
= 1− 57
+
0.0125 (1.0125) (1.0125)57
= $12, 029.79

As with previous examples, the price can be alternatively be computed using a financial calculator.
The corresponding steps are: N = 57; PMT = -175; FV = -10000; I/Y = 1.25; CPT PV. The
problems at the end of the chapter give additional practice using the general bond price formula.

3.3.2 Supply and Demand Perspective and the Yield-to-Maturity

The bond price calculations above assumed a particular value for the discount rate (r in the
general formula of Eq. (3.6)). How is this value obtained? The discount rate r is effectively
determined by the market forces of supply and demand for the bond. Financial institutions such as
mutual funds and pensions and individual investors trade (buy-and-sell) bonds in the secondary
market. We can roughly think of this market as a form of the standard competitive market model
from microeconomics. There is a market for US government bonds with approximately 10-years
to maturity, for example, and a different market for a set of corporate bonds issued by the Coca
Cola company with 8 years remaining to maturity. The phrase ’the bond markets’ generically
refers to all of these markets in which active bond trading takes place.
In any particular bond market, the competitive market model holds that the bond price is
determined as the price at which supply and demand are equal and the market ‘clears’ or reaches
an equilibrium. Economic forces that shift the supply and demand curves in the bond market
3.3 Bond Prices and Yields – 73 –

cause changes in the equilibrium price. Figure 3.9 illustrates the idea of the bond price as the
equilibrium market price in a generic bond market. Later, in section 3.8, we will further explore
the factors that drive demand and supply (and shifts) in particular bond markets. For now, it is
enough to understand that we can consider the bond price at any point in time as determined via
the standard competitive market mechanism.

Figure 3.9: Bond price as the equilibrium in a competitive market.

A key question is how the supply-and-demand market model perspective of bond price
determination squares with the present value perspective we introduced earlier. The answer to
this question is that, as the forces of supply and demand determine an equilibrium bond price,
they are equivalently determining an equilibrium interest rate or discount rate that makes the
present value relation hold for the bond. The market ‘implied’ discount rate that makes the bond
pricing present value relation hold is called the yield-to-maturity:
Definition 3.5. Yield-to-maturity
The yield-to-maturity (YTM), or simply the ‘yield’ equals the discount rate r such that
makes the present value relation of Eq. (3.6) hold exactly, given the current market price
P for the bond and other bond contract characteristics including the maturity, coupon
rate, and face value.
♣

Consider the following quantitative example:

Example 3.7 Suppose that the current market price for a 3-year bond with coupon rate equal to 5%
(annual payments) and face value equal to $1,000 equals $1,027.75. What is the yield-to-maturity
for the bond?

Solution: This bond seems familiar, right? It is actually the same bond contract as in Example
3.3 Bond Prices and Yields – 74 –

3.4. In this case, however, we are given the bond price and not the discount rate (yield). The
price of $1,027.75 is exactly the price we computed under the assumption in Example 3.4 that
the discount rate equals 4%. Therefore, the bond yield must equal 4%. Imagine, however, that
we did not already know that a discount rate of 4% is exactly consistent for the price we seek.
How can we determine the yield directly? It is challenging to solve algebraically for the yield
for a coupon-paying bond. But we can determine the yield easily using a financial calculator
(or Excel). To do so, we input the following: N = 3; PMT = -50; FV = -1000; PV = 1027.75;
CPT I/Y. The calculator responds with 4.00%, which is the yield-to-maturity. Note the sign
convention in the financial calculator inputs. The PMT and FV inputs share the same sign, as
they represent cash flows to the bondholder, whereas the PV input has the opposite sign. Failing
to adopt consistent signs for bond cash flows is a common source of error in computing the yield.
The next example illustrates the computation of the yield-to-maturity for a bond paying
coupons more frequently than annually:

Example 3.8 Consider the bond of Example (3 years to maturity, a 5% coupon rate paid semi-
annually, and a face value of $1,000). If the current bond price equals $1,028.01, what is the
yield-to-maturity?

Solution: We know from previous calculations that the yield must equal 4%. We can determine
this directly as follows. Working in 6-month time increments and noting that the bond pays 6
semi-annual coupons equal to $25 each, we can do the following computation on a financial
calculator: N = 6; PMT = -25; FV = -1000; PV = 1028.01; CPT I/Y. The calculator gives the
output ‘I/Y = 2.00%.’ Wait – we were expecting a yield of 4%, right? We must remember that
the computation is based on half-year intervals. So the rate output should be interpreted as a
6-month rate (EPR). We can obtain an annual yield by doubling the calculator output to obtain
2 × 2.00% = 4.00%, as expected.
Remark The previous example illustrates that the conventional yield-to-maturity functions as
an ‘APR’. For coupon payment schemes that are non-annual, computing the rate that makes the
bond pricing equation hold gives a rate in the form of an EPR. It is conventional to then multiply
this by m, the coupon frequency, to obtain an annualized yield, which takes the form of an APR.
Sometimes an alternative annualized yield measure called the “effective annual yield (EAY)” is
reported that uses the EAR conversion rule (from section 2.3) rather than the APR conversion
rule. We will use the APR conversion rule in this class.

3.3.3 The Price-Yield Relation

This topic will be brief but it is so important that it deserves its own subsection. Here is the
key point:
3.3 Bond Prices and Yields – 75 –

Theorem 3.3. The Price-Yield Relation

There is a negative or inverse relation between bond prices and yields. When bond prices
go up, yields go down, and vice versa.
♡

The intuition behind this result is (hopefully) straightforward. From Eq. (3.6), the yield
to maturity is equivalent to the discount rate that makes the bond present value equation hold
exactly. Now, from our basic time value of money material (Chapter 1), remember that when
the discount rate increases, all else equal, the present value of future cash flows decreases. It
is exactly the same here. When the yield to maturity increases the bond price decreases and
vice versa. From the opposite (price) perspective, if the price rises with all other bond attributes
(maturity, coupon, etc.) held constant, it must be that the yield (discount rate) has fallen.
It is crucial to be very comfortable with the price-yield relation in reading the financial
press. The financial media speak interchangeably in terms of prices or yields in describing bond
market conditions. One day the headline in the markets section of the Wall Street Journal might
state that “Treasury yields rise on weak employment report.” You need to immediately realize
that this means that Treasury bond prices increased in conjunction with the employment data
release (due to for example a flight-to-safety effect from stock and corporate bonds to Treasuries).
The next day the Journal might contain the headline “Corporate bond prices rise in the wake of
Fed move.” Here you need to understand that corporate bond yields (or implied interest rates)
fell.

3.3.4 Zero Coupon Bonds

Not all bonds pay coupons. Some bonds, called zero coupon bonds, or alternatively, discount
bonds, pay no coupons. These are particularly simple types of bonds. They pay the bondholder
the face or par value at maturity, with no intermediate coupon payments. The most well-known
form of discount bonds are Treasury Bills, or T-bills for short, issued by the US government. The
government regularly issues Treasury bills with maturities ranging from 30 days to one year.
Consider first the pricing of a zero coupon bond given an assumed discount rate. Here is a
very simple example:

Example 3.9 Suppose that a zero coupon bond has one year to maturity. The face value of the
bond is $100. If the appropriate discount rate equals 3%, what is the price of the zero coupon
bond?

Solution: The general bond pricing equation becomes very simple for zero coupon bonds.
3.3 Bond Prices and Yields – 76 –

Since C = 0 and m = 1, we are left with:

FV
P =
(1 + r)N
100
=
1.031
$97.09.

Notice that the price is below the face value of $100. Basic time value of money logic dictates
that this will always be the case for a zero coupon bond so long as the discount rate is positive.
It is for this reason that zero coupon bonds are alternatively called discount bonds: these bonds
trade at a ‘discount’ to their face or par value.
Now we turn to the computation of the yield-to-maturity for a zero coupon bond. In this
special case, we can solve algebraically for the underlying discount rate that makes the bond
pricing equation hold.

Example 3.10 A six-month zero coupon bond with face value of $100 trades for $98. What is
the yield-to-maturity?

Solution: In this case we can simply set up the present value equation for the bond and solve
for r, noting that N = 0.5 there are 6-months to maturity:
FV
P =
(1 + r)N
100
98 =
(1 + r)0.5
100
(1 + r)0.5 =
98
 
100 2
r= −1
98
r = 4.1233%

This analytical solution directly gives us an annualized yield, because the time to maturity N
is measured in years. There is a subtlety here though: the yield is in the form of an ‘EAR.’
But earlier we said that yields are typically in the form of APRs. The resolution for this is that
the implicit compounding frequency in the calculation is annual, and not semiannual despite
the 6-month maturity. To more completely appreciate the issue here, suppose we take the
reverse perspective and compute the price of the zero coupon bond using the 4.1233% yield we
determined. If we interpret the yield as an APR with m = 2 (semi-annual), we would compute
the EPR as 4.1233%/2 ≈ 2.0616% and then the PV as:
100
PV = = $97.98
1.020616
But this isn’t exactly right, because we know that the true bond price is exactly $98. What is
going on? Suppose we first convert the EAR of 4.1233% to an APR under the assumption of
3.3 Bond Prices and Yields – 77 –

semiannual compounding. This gives:

AP R = m × [(1 + EAR)1/m − 1]
= 2 × [(1.020616)1/2 − 1]
= 4.08163%

Now let’s re-price the bond using this rate as an APR:

100
PV = = $98.00
1 + 4.08163/2
Now we get the expected result. This illustrates that, if we want to express the yield for a zero
coupon as an ‘APR’ with the compounding frequency equal to the maturity of the zero coupon,
then we have to convert the ‘EAR’ that the analytical solution provides to an APR with the
appropriate compounding frequency.
The basic issue is that coupon-paying bonds have a ‘natural’ compounding frequency that
is equal to the coupon frequency, but zero coupon bonds do not. Consider a 5-year zero coupon
bond, for example. It probably does not make much sense to assume an interest compounding
frequency of 5 years. But this is what we would be doing if we took the EAR and converted
to an APR using the same approach as with the 6-month bond. One approach to this awkward
situation is to use ‘continuously compounded’ yields for zero coupon bonds. Using our time
value of money results for continuous compounding (see section 2.2), we can write:

F V = P V ert (3.7)
ln(F V ) − ln(P V )
r= , (3.8)
t
where t is time measured in years and r should be interpreted as the “continuously compounded
yield” i.e., it is the continuously compounded rate that gives the correct present value for the zero
coupon bond. Applying this to our previous example gives:
ln(F V ) − ln(P V )
r=
t
ln(100) − ln(98)
=
0.5
= 4.04054%

This is the continuously compounded yield. Finally, we can convert this to an EAR:
EAR = er − 1 = e0.04054 − 1 = 4.1233%, which is exactly the yield we originally obtained.
See Problem #18 in the chapter end problem set for more practice with these ideas.

3.3.5 Important bond jargon and bond relations

There is a lot of special terminology or ‘jargon’ related to bonds. Here we discuss some
important jargon and associated bond relations. As noted earlier, a zero coupon bond is sometimes
referred to as a discount bond. More generally, a bond that trades at a price below its face value
is said to be “trading at a discount.” Similarly, a bond that trades at a price greater than its face
value is said to be “trading at a premium” or a “premium bond.” Finally, a bond that trades at a
3.3 Bond Prices and Yields – 78 –

price exactly equal to its face value is said to be “trading at par” or a “par bond.”
What makes a bond trade at par, versus at a discount or premium? As we have noted above,
a zero coupon bond will always trade at a discount. By definition a zero coupon bond has a
coupon rate equal to zero. This suggests a connection between the coupon rate and the pricing
of the bond. In fact, what is crucial is the coupon rate relative to the yield to maturity. A bond
trades at a discount precisely when the yield to maturity exceeds the coupon rate. Naturally,
this occurs when the coupon rate is zero (as long as the yield is positive), and therefore zero
coupon bonds trade at a discount. But coupon paying bonds can also trade at a discount, and do
so precisely when the coupon rate is lower than the current yield-to-maturity. If the coupon rate
is above the current yield-to-maturity for a coupon paying bond, then the bond will trade at a
premium rather than at a discount. Finally, a bond trades at par precisely when the coupon rate
and current yield-to-maturity are identical. Summarizing:
1. A bond trades at par if (and only if) the coupon rate equals the yield-to-maturity
2. A bond trades at a discount if (and only if) the coupon rate is below the yield-to-maturity
3. A bond trades at a premium if (and only if) the coupon rate is above the yield-to-maturity
Another common metric is the current yield:
Definition 3.6. Current Yield
The current yield is equal to the ratio of the annual coupon on the bond C to its current
price P , expressed as a percentage.
♣

Note that important distinction between the coupon rate and the current yield. Both measures
have the annual coupon payment as the numerator. However, the current yield has the current
price as the denominator, whereas the coupon rate has the face value as the denominator. Because
prices tend to change over the life a bond, the current yield moves around over time like the yield
to maturity. In contrast, the coupon rate is fixed for the life of a bond (at least for standard ‘plain
vanilla’ bonds with fixed coupon rates).
The key difference between the coupon rate and current yield illustrate how the two measures
relate to one another. By definition, when a bond trades at a premium, the current bond price
exceeds the face value. This implies that the current yield (with the price in the denominator)
will be lower than the coupon rate (with the face value in the denominator). The reverse will
occur whenever the bond trades at a discount, and the two measures are equal exactly whenever
the bond is priced at par.
The following characterizes the joint relations between the coupon rate, yield to maturity
and current yield for coupon paying bonds:
1. When a bond is priced at par, the coupon rate, current yield, and yield to maturity are all
equal
2. When A bond is priced at a discount, the yield to maturity exceeds the current yield, which
in turn is greater than the coupon rate
3.4 Holding Periods Returns and Yields – 79 –

3. When a bond is priced at a premium, the coupon rate exceeds the current yield, which in
turn exceeds the yield to maturity

Example 3.11 The current yield on a bond is 4% and the coupon rate is 3%. Is the bond priced at
a discount, at par, or at a premium? How much can you say about the current yield to maturity?
Solution: The fact that the current yield exceeds the coupon rate implies that the current bond
price is lower than the face value. Therefore we conclude that the bond currently trades at a
discount. Without additional information, we can only conclude that the current yield to maturity
must be greater than 4%.

3.4 Holding Periods Returns and Yields

The yield to maturity can be loosely interpreted as the rate of return offered by the bond. But
how loose is this interpretation, exactly? This section considers more precisely the connection
between bonds yields and the return an investor experiences in holding the bond. The main
takeaway is that the yield only precisely captures the return an investor will experience upon
buying the bond under special circumstances. In general, the return will differ from the yield,
primarily driven by the fact that bond yields (and hence prices) fluctuate over time.
Recall the general definition of the holding period return on an asset from Chapter 1:
New Value − Initial Value
Ret = × 100.
Initial Value
We will first consider the case in which an investor holds a bond for exactly one coupon period,
i.e., the investor buys the bond immediately after a coupon distribution, holds the bond until
receiving the next coupon, and then immediately (re-)sells the bond. We can specialize the
general return formula to this setting as follows:
Pt+1 + C̃ − Pt
HP Rt,t+1 = × 100, (3.9)
Pt
where Pt is the initial bond price, Pt+1 is the bond price next period, C̃ = C/m is the periodic
coupon payment equal to the annual coupon amount C dividend by the number of times coupons
are paid per year. The t, t + 1 subscript emphasizes that this is the return over a single period,
where a period is defined as the time interval over which coupons are paid, e.g., one half of a
year for a semi-annual bond or one year for an annual bond.
Re-writing the above formula as the sum of two terms helps with intuition:
Pt+1 − Pt C̃
HP Rt,t+1 = × 100 + × 100 . (3.10)
Pt Pt
| {z } | {z }
Capital gain Current yield
The first term on the right-hand side of 3.10 captures the capital gain component of the bond
holding period return that reflects the increase (or decrease) in the bond price over the holding
period. The second term effectively captures the current yield of the bond.2 So the bond return

2Remember that the current yield equals the ratio of the annual coupon amount C to the current price Pt . So when the
3.4 Holding Periods Returns and Yields – 80 –

can be described in words as current yield plus capital gain. A subtle point from Eq. (3.10)
is that the current yield component of the bond return is known at the beginning of the period.
This is because the coupon rate is a fixed bond characteristic and current price Pt is, of course,
known at time t. This means that all uncertainty associated with the bond return comes via the
capital gain component and reflects the uncertainty associated with future bond prices.
The holding period return HP Rt,t+1 is only annualized if the number of coupon payments
per year equals one (m = 1). If the bond is semi-annual, for example, then HP Rt,t+1 is a
6-month return. If we want to annualize this return, then we treat the return as an ‘EPR’ and
convert to an EAR:

Ann.
HP Rt,t+1 = (1 + HP Rt,t+1 )m − 1, (3.11)

where the ‘Ann.’ superscript emphasizes that this is an annualized version of the holding period
return.
Now that we understand the definition of the holding period return on a bond, let us examine
the main issue in this section: will the holding period return (HPR) equal the bond yield?
Consider the following simple example bond:

Example 3.12 An investor purchases a $100 FV bond with exactly 10-years to maturity and a
coupon rate of 6% (annual payments) at par. The investor will hold the bond for one year, receive
the next coupon payment, and immediately sell the bond. Determine the holding period return
under the following three scenarios:
1. The bond yield remains fixed at 6% in one year
2. The bond yield falls from 6% to 5% in one year
3. The bond yield rises from 6% to 7% in one year
Solution: First, note that we can infer several important facts from the observation that the bond
is trading at par when purchased. In particular, the current yield on the bond must equal 6%,
because a bond is priced at par exactly when the coupon rate and yield are equal. Second, the
initial price of the bond equals the face value, or $100. In order to compute the holding period
return, we first compute the current yield component of the return C̃/Pt = 6/100 = 6%.. Note
that this component of the bond return will remain fixed across all three scenarios. Next, we
compute the new bond price after one year and the corresponding capital gain component of the
return in each scenario. Then we add to this the current yield component to obtain the HPR.
What is the new bond price after one year? We can determine this using the bond pricing
method discussed in the previous section (PV of future payments). It is important to note that
the maturity of the bond shortens after a holding period. In other words, one year later this bond
has 9 years to maturity, rather than 10 as when we bought the bond. The coupon payment is

bond pays coupons annually m = 1, the second term literally equals the current yield. When the bond pays coupons
more frequently than annually, e.g., the semi-annual case (m = 2), then the second term in Eq. (3.10) is technically
equal to the current yield dividend by m.
3.4 Holding Periods Returns and Yields – 81 –

6% × $100 or $6. With this in hand, it is straightforward to compute the bond price in one year
in each of the three scenarios using a financial calculator:
Scenario 1: N = 9; PMT = -6; FV = -100; I/Y = 6; CPT PV = $100
Scenario 2: N = 9; PMT = -6; FV = -100; I/Y = 5; CPT PV = $107.11
Scenario 3: N = 9; PMT = -6; FV = -100; I/Y = 7; CPT PV = $93.48
Now, with these new prices in hand, we can compute the holding period return in each
scenario as follows:
Pt+1 − Pt C̃
Scenario 1: HP Rt,t+1 = × 100 + × 100
Pt Pt
100 − 100 6
= × 100 + × 100
100 100
= 0% + 6% = 6%
107.11 − 100 6
Scenario 2: HP Rt,t+1 = × 100 + × 100
100 100
= 7.11% + 6% = 13.11%
93.48 − 100 6
Scenario 3: HP Rt,t+1 = × 100 + × 100
100 100
= −6.52% + 6% = −0.52%

Note that, because the bond pays coupons annually, the holding period returns computed above
are already annualized. The results of our calculations provide important insights regarding the
relation between the YTM and the HPR. In the first scenario, when the YTM stays constant at
6%, then the holding period return equals the yield of 6%. However, the second two scenarios
illustrate that this outcome is a special case. When the yield falls (rises) over the period that
the investor holds the bond, this causes the price to rise (fall), and the resulting HPR deviates
considerably from the YTM of 6% when we bought the bond. We can summarize this insight as
follows:

Theorem 3.4. HPR versus YTM for one holding period

When a bond is held for exactly one period, the (annualized) one-period holding period
return equals the yield to maturity when the bond is purchases if (and only if) the yield
does not change over the period. If the yield rises (falls), then the HPR will be less than
(greater than) the YTM when the bond was purchased.
♡

In Example 3.12, the bond in question is initially priced at par. This turns out to be something
of a special case, in the sense that the capital gain component of the return equals zero when the
yield did not change. In general, this does not hold true. To explore this issue, let us modify the
previous example so that the bond is initially not priced at par.

Example 3.13 Consider the same bond as in Example 3.12, except assume that the initial yield
equals 7% instead of 6%. Determine the one-period HPR if the yield stays the same, decreases
3.4 Holding Periods Returns and Yields – 82 –

to 6%, or increases to 8%.

Solution: The approach is similar to the previous example. However, in this case we must first
compute the initial bond price at purchase. A straightforward computation using the financial
calculator (N = 10; PMT = -6; FV = -100; I/Y = 7) reveals that the initial price equals $92.98.
The current yield equals C̃/Pt = 6/92.98 = 6.453%. Next, we determine the end-of-period
bond price in each scenario:
Scenario 1: N = 9; PMT = -6; FV = -100; I/Y = 7; CPT PV = $93.48
Scenario 2: N = 9; PMT = -6; FV = -100; I/Y = 6; CPT PV = $100.00
Scenario 3: N = 9; PMT = -6; FV = -100; I/Y = 8; CPT PV = $87.51
As before, we now compute the HPR in each scenario as follows:
Pt+1 − Pt C̃
Scenario 1: HP Rt,t+1 = × 100 + × 100
Pt Pt
93.48 − 92.98 6
= × 100 + × 100
92.98 92.98
= 0.547% + 6.453% = 7.00%
100 − 92.98 6
Scenario 2: HP Rt,t+1 = × 100 + × 100
92.98 92.98
= 7.554% + 6.453% = 14.007%
93.48 − 92.98 6
Scenario 3: HP Rt,t+1 = × 100 + × 100
92.98 92.98
= −5.883% + 6.453% = 0.570%

The key insight in this second example concerns the first scenario, when the YTM does not
change over the holding period. Notice that, in contrast to the first example when the bond was
priced at par, the capital gain component of the HPR does not equal zero. Even though the yield
does not change, the bond price rises and there is a positive capital gain component of a bit over
0.5%. The intuition behind this is that the bond price will revert toward the face value as the
maturity shortens when the yield stays constant. Indeed, as the maturity approaches zero, but the
bond value much approach the terminal cash flow by basic TVM logic. Figure 3.10 illustrates
how the price of this particular bond will track gradually toward the face value as the maturity
shortens with the yield fixed and constant at 7%. The price gradually tacks back toward the face
value of $100 as the time to maturity shortens. It is this maturity effect that explains the small
positive capital gain in this case.
A second crucial point concerning Example 3.13 is that, despite the fact that the capital
gain component is now positive, the sum of the capital gain and current yield components equals
7%, so that the total HPR does in fact equal the (unchanged) YTM, consistent with the result in
Theorem 3.4.
Thus far we have only addressed the relation between the yield to maturity and bond holding
period returns for a single holding period. In practice, bond investors often hold bond positions
3.4 Holding Periods Returns and Yields – 83 –

Bond Price as Function of

Time to Maturity
(Fixed 7% Yield to Maturity)

$100.00
$99.07
$98.19
$97.38
$96.61
$95.90
$95.23
$94.61
$94.03
$92.98 $93.48

10 9 8 7 6 5 4 3 2 1 0

Figure 3.10: Bond price as function of time-to-maturity (yield fixed at 7%).

for more than one period, perhaps even to maturity. Consequently, it is important to understand
how multiple holding periods impacts the relation. Before diving into an illustrative example,
we need to discuss how to measure multi-period holding period returns. The complication here
is that the investor receives a series of cash flows in the form of periodic coupon payments, in
addition to the cash received when the bond matures or is sold prior to maturity. Suppose that
we purchase a bond today (at time t) and hold the bond for H periods, where H denotes the
number of coupon periods for which the bond is held. For example, if a bond pays semi-annual
coupons and the position is held for six-months years, then H = 1, whereas in the bond is held
for 3 years, then H = 6. With this notation, the annualized holding period return HP Rt,t+H is
given by
 (m/H)
Ending value
Ann.
HP Rt,t+H = − 1, (3.12)
Starting value
where m is the number of coupon periods per year, “Starting value” and “Ending value” denote
the dollar value invested in the position initially and at the time the position is liquidated or
terminated. The starting value is generally straightforward as this is simply the initial investment
value. The ending value is less straightforward, because it depends not only on the coupon
payment and resale or face value received at the time the position is liquidated, but also on the
intermediate coupon payments and the assumed rates at which these are reinvested.
What should be assumed regarding the reinvestment rates for intermediate coupons? This
is not a straightforward question. One possible assumption is that intermediate coupons are
re-invested such that they earn a return equal to the yield when the bond was purchased. An
alternative assumption is that they are re-invested at the prevailing yield at the time the coupon
3.4 Holding Periods Returns and Yields – 84 –

is received. Of course, if the yield does not change, then these two concepts are the same.
But if yields change over time, they differ. As we will soon see, this distinction is important
in determining whether the holding period return will equal the yield at the time the bond is
purchased.

Example 3.14 Consider the same 6% coupon (annual) bond we have been studying. Suppose that
the yield is constant at 7% and an investor purchases the bond with exactly two years remaining
to maturity and holds the bond until maturity. What is the (annualized) holding period return?
Solution: First, note that the purchase price of the bond equals $98.19 (see Figure 3.10 and
verify yourself). The bond pays a coupon of $6 with one year remaining to maturity. By
assumption, this is reinvested at a rate of 7% for the following year. Finally, at maturity, the
investor receives an additional cash flow of $106 in the form of the final bond coupon and the
face value. The total cash received by the investor (including the reinvested coupon) at maturity
is therefore:
$106 + $6(1.07) = $112.42

As noted above, the bond purchase price was $98.19. Therefore, the annualized holding period
return equals:
 
Ending value m/H
Ann.
HP Rt,t+2 = −1
Starting value
 
112.42 1/2
= −1
98.19
= 7.00%

This illustrates that, if the yield does not change and we assume that intermediate coupons
are re-invested at this yield, then the (annualized) HPR equals the YTM when the bond was
purchased. What if the yield changes? When the yield changes, then typically the HPR will
differ from the YTM when the bond was purchased. Then intuition is similar to the one-period
case: changes in the YTM cause related price changes and these make the HPR differ from the
original YTM. The following extension of the previous example illustrates a more subtle point:
the HPR and YTM will generally differ if the yield changes even if the final yield equals the
original yield if coupons are reinvested at the yield that prevails when they are received.

Example 3.15 Let us re-consider Example 3.14, except that we will now assume that the yield
jumps from 7% to 10% one year from now, only to fall back to 7% at maturity. Assume
intermediate coupons are re-invested at the yield that prevails at the time the coupon is received.
What is the HPR in this scenario and does it still equal to the YTM when the bond was purchased
(7%)?
Solution: The solution is very similar to that for Example 3.14. In fact, we only need to make
one modification: whereas the intermediate coupon of $6 previously earned a return of 7% when
it was re-invested, it now will earn a return of 10% (the yield when this coupon is received). This
3.5 Bonds and Interest Rate Risk – 85 –

means that the ending value for the position becomes:

$106 + $6(1.10) = $112.60

As noted above, the bond purchase price was $98.19. Therefore, the annualized holding period
return equals:
 
Ending value m/H
Ann.
HP Rt,t+2 = −1
Starting value
 
112.60 1/2
= −1
98.19
≈ 7.09%

The punchline is that the HPR now exceeds the YTM. The intuition is that the temporary increase
in yields during the bond holding period increases the rate of return on re-invested coupons. It
is exactly this effect that causes the HPR to deviate positively from the 7% YTM when we
purchased the bond. One of the end-of-chapter problems asks you to to consider yet another
variation in which a YTM change is permanent rather than temporary, both in a setting where
the bond is held to maturity and another setting in which it is not.
The following result summarizes the general insights from our analysis:
Theorem 3.5. Yield to Maturity and Holding Period Return
The holding period return on a bond will equal the yield to maturity of the bond at the
time the bond is purchased if 1) the yield does not change over the holding period of the
bond or the investor holds the bond to maturity; and 2) all coupons received during the
holding period are reinvested at a rate equal to the yield at the time the bond is purchased.
In general, when yields change (and the bond is sold before maturity) or coupons are
reinvested at rates other than the initial yield, the HPR will differ from the yield at the
time the bond is purchased.
♡

3.5 Bonds and Interest Rate Risk

There is a common perception that investments in US Treasury bills, notes, and bonds are
“safe.” After all, the government can, in principle, print the dollars that it requires in order satisfy
debt obligations, and so the investor is virtually assured to received the cash promised by these
bonds in timely fashion. Indeed, this much is true – the cash promised by US government bills,
notes, and bonds is nominally safe (assuming ‘the market’ has minimal concerns about sovereign
default for the US, which is likely true at least at the present time). The point of this section,
however, it to convince you that the nominally safe cash flows associated with government bonds
does not make them free of risk. The main risk associated with bonds is typically dubbed interest
rate risk. In essence, it captures the possibility that yields might rise after bonds are purchased,
causing their value (price) to fall. This erodes investor returns and can lead to negative returns
3.5 Bonds and Interest Rate Risk – 86 –

(loss of value) on the investment position. But bonds are not equivalent in terms of their exposure
to interest rate risk. Our second goal, therefore, is to understand which bonds entail more interest
rate risk and why. Finally, we discuss basic strategies investors might employ to limit the interest
rate risk associated with bond positions.

3.5.1 The components of interest rate risk

Although it is true that, historically, the risk associated with bond investments is lower than
that associated with equity (common stock), bonds are generally not ‘risk-free’ investments.
Indeed, as we saw in the previous section, the holding period return associated with a bond
investment depends on the (unknown) future path of yields. In particular, if yields increase
during the holding period, bond prices will decline leading to a capital loss and potentially
negative holding period returns. If yields increase sharply, the associated losses can be large for
certain types of bond portfolios. The risk to bond investors associated with unpredictable future
changes in yields is termed ‘interest rate risk:’
Definition 3.7. Interest rate risk
Interest rate risk relates to unpredictable variation in interest rates/yields that causes the
total return for a bond to differ from the promised yield or YTM.
♣

We can break down interest rate risk into two specific components, one that impacts investors
via the channel of the price of the bond, and another, more subtle, effect that operates via the
returns earned when intermediate coupons are reinvested:
Definition 3.8. Price risk
Price risk is the component of interest rate risk that arises due to variability in bond prices
caused by their inverse relationship with interest rates / yields.
♣

Definition 3.9. Reinvestment risk

Reinvestment risk refers to the component of interest rate risk that arises due to the impact
of unforeseen changes in future interest rates on the return earned by reinvested coupons
associated with a bond position
♣

It is important to note that price risk and reinvestment risk work against each other. When
interest rates fall, bond prices rise (good for the investor) BUT coupons are reinvested at a lower
return (bad for the investor). In contrast, when interest rates rise, bond prices fall, but coupons
are reinvested at at higher return.
The extent to which a bondholder experiences price risk and reinvestment risk depend both
upon the characteristics of the bond and on the investor’s horizon, especially whether or not the
investor will hold the bond to maturity. Two key insights regarding the nature of interest rate
exposure facing a bondholder are as follows:
1. Zero coupon bonds do not entail reinvestment risk: This idea is hopefully straightfor-
3.5 Bonds and Interest Rate Risk – 87 –

ward. If a bond does not pay coupons, then there is no uncertainty concerning reinvestment
returns on intermediate coupons (there are none!), and therefore the bondholder does not
face reinvestment risk. Just to be clear, this does not mean that a zero coupon bond does
not face interest rate risk – these bonds will generally entail price risk.
2. A bond held to maturity does not face price risk: This idea is a bit more subtle. As
we have seen in plenty of earlier examples, yield changes impact bond prices. However,
remember that as a bond approaches maturity its value approaches the terminal cash flow.
In other words, if a bondholder commits to holding a bond all the way to maturity, the cash
received at maturity is certain. (This assumes nominally safe bonds such as US Treasury
bonds.) This means that, irrespective of the path of yields over the period the investor
holds the bond (yields might skyrocket, then fall, then skyrocket again), the investor will
receive the same fixed cash flow at maturity. Again, this does not imply that a bond held to
maturity is free of all aspects of interest rate risk: coupon bearing bonds held to maturity
face reinvestment risk.
The previous insights reveal a potential strategy for avoiding interest rate risk. If an investor
wants to entirely avoid interest rate risk, she might purchase zero coupon bonds and hold the
bonds to maturity. Because there are no coupons, this strategy is free of reinvestment risk.
Additionally, the strategy is free of price risk because the bond is held to maturity. In fact, the
investor’s certain future cash flow exactly equals the face value of the zero coupon bond. There
are two issues associated with implementing this strategy, though. First, the US government
only issues zero coupon bonds with maturities up to one year. Many real-world bond investors
have much longer horizons. In the next section we will discuss a certain type of security (bond
“STRIPS”) that effectively offer the investor longer maturity zero coupon US Treasury bonds –
so this concern can be addressed. The second concern is imply that the investor might need to
sell the bond prior to maturity for liquidity reasons. If there is any chance that the investor must
sell the bond prior to maturity, then the investor effectively faces a form of price risk.

3.5.2 Interest rate risk and bond characteristics

When yields change, bond prices change. This is true for any conventional bond. But the
relative magnitude of the bond price change – in other words the extent to which the yield change
alters the capital gain component of the bond return – depends upon the characteristics of the
bond.
1. Maturity: All else equal, the effect of an increase in the yield on the bond price is increasing
in the maturity of the bond.
2. Coupon rate: All else equal, the effect of an increase in the yield on the bond price is
decreasing in the maturity of the bond.
The intuition behind these relations stems from basic time value of money logic. To
illustrate, consider the present value of $100 received N periods in the future when the discount
3.5 Bonds and Interest Rate Risk – 88 –

rate equals 5%. Our basic TVM PV formula says the corresponding value equals:

P V = $100/(1.05)N ,

where N is the number of years until the $100 is received. Now, imagine that the discount rate
instantaneously increases from 5% to 6%. Let us compute the percentage change in the value of
the future cash flow when this occurs in two cases: 1) the $100 will be received in one year; or
2) the $100 will be received in 20 years. In the first case, we have:

P V (old) = $100/1.05 = $95.24

P V (new) = $100/1.06 = $94.34
(94.34 − 95.24)
%∆P V = = −0.943%
95.24

So the value of the $100 falls by a bit less than one percent. In the second scenario, we have:

P V (old) = $100/1.0520 = $37.69

P V (new) = $100/1.0620 = $31.18
(31.18 − 37.69)
%∆P V = = −17.269%
37.69

Now the value of the $100 falls by over 17 percent. This explains the logic behind why a yield
increase has a greater proportional impact on the value of a long maturity bond relative to a short
maturity one: yields play the role of discount rates in the bond price equation, and therefore
yield changes have more dramatic impacts on value for long maturity bonds with cash flows that
occur in the distant future.
The reason that interest rate sensitivity is smaller the larger the coupon rate is rather similar.
All else equal, as the coupon rate on a bond increases, the proportion of the value of the bond
that is received earlier increases, i.e., more of the total cash flow promised by the bond will be
received earlier. Because cash flow received earlier are less sensitive to discount rate fluctuations
than cash flows received in the more distant future, an increase in the coupon rate (all else equal)
reduces the sensitivity of the bond to interest rate changes.

3.5.3 Quantifying interest rate risk: the duration measure

Thus far we have mostly focused on qualitative insights regarding bonds and interest rate
risk. Now we turn to an effort to quantify this risk. This will ultimately lead us to introduce a
convenient numerical measure called duration that helps captures the interest rate risk associated
with a particular bond.
Suppose that the yield is currently 4% and consider a 5-year, 10% coupon bond (annual
payments). The current bond price equals $126.71 per $100 of face value. (Be sure you can
verify this.) Imagine that you work at a large pension fund that invests in bonds. Your fund has
3.5 Bonds and Interest Rate Risk – 89 –

taken a relatively large position $500,000 in this particular bond. There is a Federal Reserve
meeting in the next few days, and a risk manager at the fund is concerned about the possibility that
yields could increase and asks you how much interest rate risk is associated with this particular
bond investment.
One way to answer this question would be to imagine alternative possible scenarios for the
bond yield in the very short run (let’s say a week). For example, the yield could fall to 3.7%, or
it could increase to 4.3%, or even to 4.5%, and so forth. In each scenario, you could then work
out the new price for the bond, the corresponding return, and the effect on the value of the fund’s
position in the bond. You could then present these tabulations to the risk manager, who could
then determine whether the risk was acceptable, or needed to be reduced in some way, perhaps by
selling some of the position. Note that, for the purpose of these calculations, we are holding the
maturity fixed at 10% – this is approximately correct if we are considering only a week horizon.
Figures 3.11 and 3.12 show calculations of this sort for the bond in question. The solid blue line
in Figure 3.11 shows the bond price as a function of the yield, while the solid blue line in Figure
3.12 shows the corresponding return. Although both blue lines appear roughly linear, they are in
fact nonlinear and convex in shape (as a reminder, a convex curve ’holds water’).3

Figure 3.11: Bond price as function of yield

Figure 3.12 shows that if yield were to suddenly spike from 4% to 5%, the bond position
would lose a bit over than 4% of its vale (4.11% in fact). Using this, the risk manager could
assess whether this potential loss on the position (a loss in value of $500, 000 × 4.11% or just
under $20,000) is acceptable or not.
Although the preceding calculations do quantify the interest rate risk exposure, they are a

3In bond and fixed income analysis, you will sometimes hear discussion of bond ‘convexity’ – this refers to the degree of
curvature in the bond price function. The greater the curvature, the less accurate is the duration-based approximation
that will be discussed below.
3.5 Bonds and Interest Rate Risk – 90 –

Figure 3.12: Bond return as a function of yield

bit tedious to execute, as the bond must be re-priced at a wide range of alternative yields. In
addition, the fund is likely to hold a large number of bond positions, and each of these positions
would be influenced by a potential yield change to some extent or another. It would be nice to
have a very simple way to determine the magnitude of interest rate risk exposure without having
to go through so much trouble. A quantity called the bond duration is very useful in this regard.
Definition 3.10. Duration
The (Macaulay) duration (D) equals the weighted average time to maturity of a bond.a
a
The ‘Macaulay’ prefix refers to the fact that there are several slightly different, related measures of duration
that are used in fixed income analysis. We will leave discussion of these to more advanced courses.
♣

We will not consider the details of the computation of the duration quantity (you will be
given the value if necessary). It will suffice to note the following key points:
The duration of a zero coupon bond is equal to the time to maturity
All else equal, the longer the maturity, the higher the duration
All else equal, the larger the coupon rate, the shorter the duration
The bond duration provides a very convenient way to assess bond risk and specifically to quantify
how much a bond position would stand to lose if interest rates rise by a specified amount. The
key result is:
Theorem 3.6. Duration and Bond Return
Given a bond or portfolio of bonds with duration D, the approximate percentage change
in price (return) associated with an immediate change of the yield of an amount ∆r from
a starting yield of r is given by:
∆r
%∆P ≈ −100 × D × (3.13)
1+r ♡
3.6 Special bonds: STRIPS and TIPS – 91 –

Returning to Figures 3.11 and 3.12, the orange dashed line shows the duration-based
approximation to the price and return, respectively, associated with various values of the yield.
The duration-based approximations are linear, whereas the true relations are actually nonlinear
as mentioned above. However, the approximation is quite good. There is only a noticeable
difference between the duration-based approximation and the true price or return for relatively
large deviations of the yield from the current level of 4%. The takeaway is that the duration-based
approximation formulas are accurate for small yield changes.
One reason that duration is a convenient bond analysis tool is the fact that the duration of
a portfolio of bonds is equal to the weighted average of the duration values for the bonds in the
portfolio, where the weights are given by the fraction of total portfolio value associated with each
bond.
Example 3.16 Consider a bond with 3 years to maturity and a coupon of 4% when market yield
is 10%. The duration of this bond equals 2.88. Suppose yield increases to 12%. What is the
approximate return on the bond based on the duration? Determine the exact return and assess
how these compare.
Solution: First, using the duration result, we have:

%∆P = −100 × 2.88 × (0.02/1.10) = −5.24%

Using the exact approach (reprice the bond with the new yield of 12%), it is easy to verify that
the precise price drop equals 5.05%. In this case, the difference is somewhat noticeable (almost
20 basis points). This is because the hypothesized yield change is very large (a full 2%). This
would be an extremely large sudden change in yields by US historical standards.

3.6 Special bonds: STRIPS and TIPS

Here we consider two special classes of bonds issued by the US government. The first of
these types of bonds (STRIPS) effectively act as long-horizon zero-coupon bonds. We will focus
on how the prices for these bonds can be interpreted as discount factors, and introduce the idea of
‘arbitrage relations’ among securities. The second type of bond (TIPS) offers investors protection
against inflation. These bonds are interesting because their yields represent a market-based proxy
for real interest rates.

3.6.1 STRIPS

The acronym STRIPS stands for “Separate Trading of Registered Interest and Principal of
Securities.” That sounds rather fancy and complicated, doesn’t it? But in truth STRIPS are
quite simple. STRIPS are simply zero-coupon bonds that are synthesized out of regular, coupon-
bearing US Treasury notes and bonds by ‘stripping’ the notes/bonds into claims to the individual
cash flows (each coupon and the principal) promised by the notes/bonds. STRIPS exist due to
investor demand for longer maturity zero-coupon government bonds. They were introduced in
3.6 Special bonds: STRIPS and TIPS – 92 –

1985 by the Treasury to reduce the cost of financing government debt “by facilitating competitive
private market initiatives.”
Here is an example taken straight from the US ‘Treasury Direct’ website:
Example 3.17 A Treasury note with 10 years remaining to maturity consists of a single principal
payment, due at maturity, and 20 interest payments, one every six months over a 10 year duration.
When this note is converted to STRIPS form, each of the 20 interest payments and the principal
payment becomes a separate (zero-coupon) security.
STRIPS trade differently than regular US government bills, notes, and bonds. In particular,
STRIPS can only be bought and sold through a financial institution or brokerage firm and are
held in a book-entry system. Figure 3.13 shows data for the yield on an index of 20–30 year
STRIPS over the past year.

Figure 3.13: Recent for yields on 20–30 year STRIPS.

It is important to realize that the price of a STRIP security, and more generally the price
of any zero-coupon bond, provides a market-based measure of the discount factor associated
with cash received at the maturity date of the corresponding zero coupon bond. Recall that the
discount factor at a particular future time equals the present value of $1 received at that time.
Suppose, for example, that a STRIP claim that matures in 3 years currently trades for $96.42 (per
$100 of FV). This price implies that the value of $1 received in 3 years time equals $0.9642.
Equivalently, a market-based measure of the discount factor for (nominally safe) cash received
in 3 years time is 0.9642.
There are quite a lot of STRIPS trading at a given point in time. At this moment, for
example, there are STRIPS trading that mature in a few days, in a few weeks, in approximately
4 months, in two and a half years, and so on. This means that, to an approximation, the STRIP
market gives us a set of market-based discount factors for cash received at any point in the future.
If we have such a rich set of market-based discount factors, we can use them to compute the ‘fair’
price for any Treasury bond trading. The following example provides a simple illustration:

Example 3.18 Consider a Treasury bond that pays a 10% coupon (semi-annually) with exactly
one year to maturity. If a STRIP with 6 months to maturity trades at $99 and another STRIP with
12 months to maturity trades at $97.80 (both per $100 of FV), then what is the implied price of
3.6 Special bonds: STRIPS and TIPS – 93 –

the Treasury bond?

Solution: The 6-month STRIP price implies that the present value of $1 received in 6 months is
$0.99 and the 12-month STRIP price implies that the PV of $1 received in a year equals $0.978.
The bond offers the investor $5 in 6 months (the coupon rate of 10%, divided by 2, times the
FV), and $105 in 12 months (the final coupon of $5 plus the FV of $100). Consequently the
appropriate price of the bond is given by:

P = 0.99 × $5 + 0.978 × $105 = $107.64.

So, the STRIP-implied value of the Treasury bond equals $107.64.

What if the Treasury bond in the above example is actually being traded by investors for
a higher value, e.g., $110. This would suggest an arbitrage opportunity available to savvy
investors. To exploit this opportunity, an investor would ‘short sell’ the relatively high-priced
Treasury bonds, and at the same time purchase 6-month STRIPS and 12-month STRIPS in
calibrated proportions to mimic the Treasury bond. We will not go through all the details, but
suffice it to say that the investor could conceptually make a sure profit. However, as the investor
(and potentially other investors who notice the discrepancy) executes the strategy, she will be
putting upward pressure on the STRIP prices and downward pressure on the Treasury bond
prices, which will ultimately tend to push the prices toward equality, eliminating the arbitrage
opportunity.
Does the above reasoning imply that we will always see US Treasury notes and bonds trading
at precisely the prices implied by the STRIP-market implied discount factors? Not exactly, for
two reasons. First, although there are many STRIPS trading, there is not necessarily a STRIP
trading at exactly each date on which the bond in question pays coupons. This means that the
discount factors associated with such cash flows are, to some degree, approximate. Second, and
perhaps more importantly, there are transactions costs associated with executing bond trade, both
for regular Treasury bonds and for STRIPS. If the price differential between the Treasury bond
and the STRIP-implied price is small, it might not be profitable for investors to trade to exploit
the difference. So relatively small differences can and do exist in practice.

3.6.2 TIPS

Treasury Inflation Protected Securities (TIPS) are offered by the U.S.Treasury. These special
bonds help protect investors against inflation. TIPS are issued with (original) maturities of five,
10 and 30 years. In most respects, TIPS bonds are similar to regular US Treasury notes and bonds.
However, they have a special features that shelters investors from inflation risk. The key feature
of a TIPS bond is that the principal (face value) is adjusted semiannually for inflation based
on changes in the Consumer Price Index-Urban Consumers (CPI-U), a widely used measure of
inflation. Coupon (interest) payments are calculated using the inflated principal. To the extent
that inflation occurs throughout the life of the bond, therefore, the coupon payments will increase
3.6 Special bonds: STRIPS and TIPS – 94 –

correspondingly. At maturity, if the adjusted principal is greater than the face or par value, the
investor receives the adjusted value.4
The following example illustrates the adjustment feature associated with TIPS bonds:

Example 3.19 Consider the following TIPS note: Original principal amount of $100,000, 10
years to maturity, and a 3% annual coupon (1.5% semiannually). Suppose that over the next 6
months there is 1% inflation, and in the following 6-months there is 1.4% inflation according to
the CPI-U measure (NOTE: treat these as ‘EPRs’, i.e., 6-month rates). Determine the adjusted
principal for the TIPS and the first two (adjusted) coupon payments.
Solution: After 6-months, the principal adjusts to $100, 000 × (1 + 1%) = $101, 000. The first
coupon payment therefore equals the adjusted principal amount times the semi-annual coupon
rate of 1.5%. This gives:

First coupon = $101, 000 × 1.5% = $1, 515.

Similarly, after the second 6-months, the principal adjusts upward again to $101, 000×(1.014) =
$102, 414. The second coupon payment is then:

Second coupon = $102, 414 × 1.5% = $1, 536.21.

This process would continue for the remaining life of the bond depending on the realized inflation
path. Note that in this way the coupons (and face value) received by the investor maintain a
constant real value (purchasing power).
TIPS bonds provide a market-based measure of the real interest rate at various horizons.
This is very useful from the perspective of the Fisher equation. Remember the approximate form
of the Fisher equation states that r = rreal + ∆Pe , or, in words, the nominal rate equals the
real rate plus expected inflation. The difference between yields on regular US Treasury bonds
and TIPS bonds with the same maturity therefore provides a market-based estimate of inflation
expectations. For example, if current US Treasury notes with one-year to maturity have a yield
of 2%, whereas TIPS securities with one year to maturity have a yield of 1%, then we can infer
that expected inflation over the upcoming year is approximately 2% − 1% = 1%.
Figure 3.14 shows recent data for TIPS bonds from Bloomberg. The most notable feature of
these data (taken as of mid-October 2021) is the fact that the yields for TIPS bonds are negative.
Yields on 5-year maturity TIPS are around -1.7%, whereas long maturity TIPS have yields closer
to zero (but still negative). Why would bond investors accept negative yields? Note that this is
different than accepting a negative nominal rate – in that case the seemingly superior alternative
would be to simply hold cash. But holding cash does not protect purchasing power when there is
inflation. Still, the negative TIPS yields suggest that current real interest rates are negative. These
is atypical, but presumably reflects an uncertain economic and financial environment associated
with the COVID pandemic and the ongoing economic recovery. Investors have great appetite for

4The ‘if’ here subtly implies that if there is actually deflation, then the investor who holds the bond to maturity would
not receive less than the original face value of the bond.
3.7 The Yield Curve and Term Structure – 95 –

a ‘safe haven’ and one interpretation is that they are willing to pay a slight fee (in real terms) to
retain (net of the fee implied by the negative yield) current purchasing power.

Figure 3.14: Recent TIPS data from Bloomberg (10/15/2021).

3.7 The Yield Curve and Term Structure

This section discusses the so-called ‘term structure of interest rates.’ In plain English, that
phrase translates to “how bond yields vary with maturity.”

3.7.1 Term Structure: Definitions and Facts

Figure 3.15 shows bond yield data provided by Bloomberg on October 15, 2021. Current
yields-to-maturity are shown for US Government bills, notes, and bonds with a time to maturity
ranging from 3-months to 30-years. The yield on a 3-month US Treasury bill is approximately
0.04%, or 4 ‘basis points.’ Two-year US Treasury notes, which pay a coupon, have a yield
of 0.39%. Finally, 10-year and 30-year US Treasury bonds have yields of 1.57% and 2.04%,
respectively. It is important to understand that all of these yields are expressed on an annualized
basis. This means that an investor who buys the 2-year note, and holds it to maturity can be
expected to earn a rate of return of around 0.40% per year, whereas an investor who purchases
the 30-year bond and holds it to maturity can be expected to earn a rate of return of around 2%.5
The current difference in yields between 2-year US Treasury notes and 30-year US Treasury
bonds seems quite large from a practical perspective. But, how can the difference be explained?
In particular: Why are investors willing to purchase and hold 2-year Treasury notes, or even
30-day Treasury bills, when 30-year Treasury bonds offer a much higher yield? More generally,
why do yields generally vary as a function of the time to maturity for US Treasuries?
Before diving into explanations, let us introduce some important terminology. Suppose we
consider the set of outstanding US Treasury bills, notes and bonds. These securities all promise
investors cash flows that are back by the full faith and credit of the US government. In other

5As a reminder, the holding period return will only exactly equal the current yield if it is assumed that all intermediate
coupons are reinvested at the current yield. This is unlikely to be realistic, but here we can just interpret the yield as
an approximate or rough measure of the return earned by holding the security to maturity.
3.7 The Yield Curve and Term Structure – 96 –

Figure 3.15: Treasury Yields for Different Maturities from October 15, 2021, via Bloomberg.

words, there is a minimal risk of default. The bonds differ widely, however, with respect to their
term to maturity, or simply the ‘term.’ The term structure of US government bonds refers to how
the yields on US government bills, bonds, and notes depends varies with the term to maturity.
The relation is often depicted graphically in the form of the ‘yield curve:’
Definition 3.11. Yield curve
The yield curve is a graph of the (annualized) yield-to-maturity for US Treasury bonds (or
more broadly for a set of bonds or loans having similar default risk) as a function of the
time to maturity.
♣

Figure 3.16 shows the current yield curve for US Treasuries (as of mid-October 2021),
depicted as a blue line, along with the yield curve that prevailed one and two years ago, i.e., in
October of 2020 and 2019, respectively. The current yield curve is very close to zero until a
maturity of around 1 year, and then it increases with maturity and is therefore ‘upward sloping’.
The yield curve that prevailed one year ago in 2020 was qualitatively similar to the current curve;
however, the 2020 curve is somewhat flatter, especially in the maturity range from 1–5 years.
The 2019 curve is quite different from the more recent two curves. First, short maturity yields are
considerably higher than current analogs (around 1.5%). Second, the 2019 curve is flatter than
the more recent two yield curves. Finally, the 2019 yield curve is partially inverted, meaning that
the curve sloped downward rather than upward. The curve slopes downward until a maturity of
around 5 years, before changing to an upward slope, and therefore the entire curve has something
of a ‘U’-shape.
Definition 3.12. Term Spread
The term spread refers to the difference between the yield on a relatively long-maturity
government bond and that on a short-maturity government bond. The term spread reflects
the general slope of the yield curve.
♣
3.7 The Yield Curve and Term Structure – 97 –

Figure 3.16: Yield curve as of October, 2021 versus preceding two years.

The following summarizes some key facts regarding the historical evolution of the yield
curve and term spread for US Treasury bonds:
1. The level of the yield curve has varied quite a bit historically. The current level of the yield
curve is very low by historical standards. The level of the curve was relatively high in the
late 1970s and early 1980s.
2. The yield curve typically slopes upward, i.e., the term spread is usually positive.
3. Although the yield curve typically slopes upward to some degree, the particular shape of
the yield curve and the magnitude of the term spread vary substantially over time.
4. The yield curve sometimes inverts, or slopes downward, exhibiting a negative term spread.

3.7.2 Understanding the Yield Curve

From the Fisher equation, the nominal yield is (approximately) the sum of the real yield
and a component reflecting expected inflation. This is true at each possible maturity. The level
of the yield curve, then is determined by the average real yield and average expected inflation
across maturities. The level of the yield curve increases when the sum of these two components
increases, and falls when the sum of the components falls. For example, the level of the yield
curve was high in the late 1970s because expected inflation at that time was high both at short and
long horizons. The level of the current yield curve is extremely low because expected inflation
is low at short and long horizons and because the real yield is extremely low (in fact, probably
negative) across the maturity spectrum. (Recall from the previous section that current TIPS
3.7 The Yield Curve and Term Structure – 98 –

yields are negative across the maturity spectrum.)

What about the slope of the yield curve and the associated term spread? What drives
time-variation in the slope of the curve? Let us first consider the role of inflation. Predictions
of inflation can vary across horizon. Figure 3.17 shows some US data for (annualized) inflation
predictions at a 1-year and 10-year horizon. The first important insight is that the inflation rate
predictions at short and long horizons often differ materially. In the mid 2000s, for example,
longer-run inflation expectations were much higher than one-year inflation expectations. In
contrast, the short- and long-run expectations were very similar in the late 1990s and at several
other points in recent macroeconomic history. Finally, in a few instances, the one-year expected
inflation rate is higher than the 10-year expected inflation rate. A prominent example in the figure
occurs in 1999.

Figure 3.17: Expected inflation at 1- and 10-year horizons (Phila. Fed).

To understand how the yield curve slope can depend on the pattern of inflation expectations,
consider first a case in which the longer-run expected inflation rate exceeds the short-run expected
inflation rate. Imagine, for example, that short-run expected inflation (think for the upcoming
year) equals 1% whereas ‘long run’ (beyond a year) expected inflation is 2%. Suppose that the
real interest rate is 2% at all horizons. The Fisher equation implies a corresponding nominal one
year rate (yield) of 2% + 1% = 3%, but longer term rates must be higher, because they must
compensate for the expected future increase in inflation. For example, the ten-year rate should be
close to 4% because expected inflation over a ten-year horizon is close to 2%.6 In other words,
in this environment, the term structure should slope upward due to the increasing pattern of the
inflation premium.
Continuing the previous example, suppose that expectations of the short-term inflation rate
suddenly increase to the level of longer-run inflation rate expectations (2%). Then the same

6The lower expected inflation of 1% next year makes the implied rate under the Fisher equation slightly lower than 2%.
3.8 Corporate Bonds and the Structure of Interest Rates – 99 –

Fisher equation logic suggests that the ‘short end’ of the yield curve will rise to 4%, and the
curve will then become flat at 4%. As yet another example, imagine that short-run expected
inflation soars to 3%, while longer-run inflation expectations stay anchored at 2%. Now the
Fisher equation indicates that short-term yields will equal approximation 5%, whereas long-term
yields will remain close to 4%. This illustrates that ‘inversions’ of the yield curve can be driven
by the fact that inflation rates are expected to decline in the future. A stark example of this
scenario took place in 1980 and 1981, when the Federal Reserve increased short term interest
rates to combat inflation.
The slope of the (nominal) yield curve can also be influenced by the slope of the underlying
real interest rate. Figure 3.14 (in the previous section) show current TIPS yields as measures of
real rates at different maturities. Although the current TIPS yields are negative at all maturities,
they are less negative for long-maturity TIPS. In other words, the TIPS yields imply an upward
slope to the term structure of real interest rates. This pattern can shift as economic conditions
change, however, and the real yield curve can also be flat or even inverted.
Although expected changes in future inflation (or in real rates) can help explain why the
yield curve slopes upward at some points in time (future inflation and/or real rates expected to
rise) and slopes downward at other points in time (future inflation and/or real rates expected to
fall) , it is hard for this channel to explain why the yield curve usually slopes upward. After all,
it is strange to think that we would consistently expected future inflation to increase.
What then explains the typical upward slope? A reasonably compelling explanation focuses
on the interest rate risk ideas we discussed in Section 3.5. Recall that, all else equal, the extent of
interest rate risk is increasing in the maturity of a bond. Consequently, long-maturity Treasury
securities have greater interest rate risk relative to shorter-term Treasury securities. Investors
presumably recognize this risk differential. In order to induce them to hold the riskier long-term
bonds, investors demand extra compensation in the form of a higher yield (interest rate). This
extra yield associated with greater interest rate risk is termed the interest rate risk premium.
Incorporating the interest rate risk premium allows us to explain the key ‘stylized facts’
associated with the yield curve and term spread for US Treasuries. In particular, shifting
expectations of future changes in inflation rates and/or real rates contributes to variation over
time in the term spread. In addition, the interest rate risk premium explains why the curve usually
slopes upward. To clarify this idea, consider a benchmark case in which neither inflation rates
nor real rates are expected to change in the future. Without the interest rate risk premium, it
would seem that the yield curve should be flat. However, the interest rate risk premium concept
suggests that instead the yield curve will slope upward in this benchmark case. In order for the
curve to invert, expected future decreases in inflation rates or real rates must more than offset the
interest rate risk premium. This happens relatively rarely in historical US yield data.
3.8 Corporate Bonds and the Structure of Interest Rates – 100 –

3.8 Corporate Bonds and the Structure of Interest Rates

Thus far we have focused on bonds issued by the US Government. These are an extremely
important set of bonds in the financial markets: as of April, 2020, the value of outstanding US
Treasury bonds (including bills, notes, and bonds) was around $18 trillion. However, there are
other important components of the bond markets. Corporate bonds – bonds issued by companies
in order to finance investments and other activities – also represent an important class of financial
securities. The current US corporate bond market has an aggregate estimated outstanding value
of approximately $10 trillion. The size of the market for municipal bonds – bonds issued by states
and local governments – is roughly $4 trillion.7 This section considers corporate and municipal
bonds with particular emphasis on what factors influence the yields on these bonds and cause
them to differ from the yields on comparable maturity Treasury securities.

3.8.1 Corporate bond yields and default risk

In most respects, corporate bonds are similar to the US Treasury notes and bonds we have
considered in detail to this point. In particular, corporate bonds share the same basic bond contract
elements: a face value and specified maturity date when investors received this value, as well
as a coupon rate, with coupons frequently paid semi-annually or quarterly. There is, however, a
critical difference between corporate bonds and US Treasury securities: the cash flows promised
by Treasury securities are back by the ‘full faith and credit’ of the US government, whereas
corporate bond cash flows are simply backed by the cash flow produced by the issuing company’s
operations. If necessary, the US government can print new dollars to satisfy its bond obligations.
A corporation certainly cannot do this. Therefore, investors in corporate bonds face an added
default risk that is not borne by investors in US Treasuries.
Definition 3.13. Default risk
Default risk refers to the possibility that an investor or lender will not receive full and
timely payment of contractually promised payments associated with bonds or other lending
arrangements.
♣

Formally, default refers to the failure of the bond issuer to satisfy any aspects of the bond
contract concerning cash flows paid to investors. This includes timing aspects. For example,
if Ford delays a coupon payment to investors for one week beyond the contractually specified
coupon date, then this is formally a form of default event. However, assuming that Ford makes
the coupon payment one week later and makes other bond payments in a complete and timely
manner, the associated loss to investors in this case is quite small. In other cases, default can
amount to the complete loss of promised interest and principal associated with a bond. This

7Current bond market size figures are sourced from the International Capital Market Association (ICMA) for Treasuries
and corporate bonds and from Fidelity for municipal bonds.
3.8 Corporate Bonds and the Structure of Interest Rates – 101 –

might occur, e.g., in the event that a company files for bankruptcy. The actual losses incurred
by investors upon default can therefore vary considerably across different corporate bond default
events.
Figure 3.18 shows data for historical default rates on US corporate bonds. The average
default rate is roughly 5%, but there is quite a bit of variation over time. Default activity
consistently rises during economic recessions, indicated by the shaded areas in the figure. This
makes sense: when macroeconomic conditions are poor, firms tend to generate fewer sales and
less cash/profit, and are more likely to be unable to meet bond obligations. Of course, factors
other than macroeconomic conditions influence default likelihood. Industry conditions and ‘firm
specific’ variation in performance also impact the likelihood that a particular corporate bond will
experience default.

Figure 3.18: Historical and projected corporate bond default rates.

The possibility of default is obviously an important consideration for corporate bond in-
vestors. It is critical to understand that the added risk of default relative to US Treasury bonds
is ‘priced in’ to corporate bonds. In other words, all else equal, including for example, maturity,
investors will be willing to pay less for a corporate bond relative to a Treasury bond because
the former is subject to default risk whereas the latter is not. Given the inverse relation between
bond prices and yields, an equivalent statement is that a corporate bond should offer a yield that
is higher than the yield for an otherwise comparable Treasury security. This additional yield is
the ‘premium’ that investors demand for bearing the default risk associated with the corporate
bond, which gives rise to an important piece of bond jargon:
Definition 3.14. Default premium
The default premium, sometimes alternatively termed the ‘default spread’ or ‘credit
spread,’ is the additional yield offered by a security that is subject to default risk rel-
ative to an otherwise similar, default-free bond.
♣
3.8 Corporate Bonds and the Structure of Interest Rates – 102 –

In practice, it is common to measure the default premium or default spread by taking the
difference between the yield for a corporate bond (or the average yield for a collection of such
bonds) and the yield for a comparable maturity US Treasury note or bond.
Because the likelihood of default is a critical corporate bond characteristic that affects the
bond price/yield, investors are very interested in accurately assessing default potential. An
industry has emerged in which specialized financial firms provide and update scores or ‘ratings’
that categorize the relative likelihood of default for corporate bonds and other bonds with credit
risk. The assessment of bond credit strength is a highly concentrated industry with the“Big Three”
credit rating agencies – Fitch Ratings, Moody’s and Standard & Poor’s (S&P) – controlling over
90% of the business. Ratings are typically provided on a letter-based grading system, with ‘AAA’
(or ‘Aaa’ depending on the particular agency) representing the strongest credit rating, i.e., the
lowest likelihood of default. Yields for bonds with AAA ratings are typically only modestly
higher than comparable maturity US Treasury bonds. An important cutoff point in ratings is
‘investment grade,’ which is a rating of BBB or above for S&P, and Baa or above for Moody’s.
Bonds with ratings below this threshold are termed ’speculative grade’, ’high-yield,’ or simply
‘junk bonds’ in light of relatively high likelihoods of default. Investment grade designation is
important because falling below this threshold often means that certain large institutions, such
as pension funds, cannot hold the bond due to fiduciary (risk-management) concerns.
Figure 3.19 below shows a picture of the evolution over time of the aggregate or market-wide
default premium associated with corporate bonds. The measure is constructed as the difference
between the average yield on a basket of ‘Baa’-rated corporate bonds (bonds in the lowest class of
the investment grade tier) less the yield on a 10-year Treasury bond. Over a long span of time, the
average default premium is around 3%. However, note that the default premium tends to increase
during recessions, and increased substantially during the financial crisis period (2008–2011), as
well as recently during the COVID-related economic contraction. The logic behind this effect
is that when investors perceive macroeconomic conditions to be unfavorable, they sell corporate
bonds and buy Treasuries, which are insulated from the higher default rates likely to prevail
during a recession.

Figure 3.19: Historical default spread: Baa yields less 10-year Treasury yields (source: FRED II)
3.8 Corporate Bonds and the Structure of Interest Rates – 103 –

3.8.2 Corporate bond contract features and yields

Corporate bond contracts can be quite complicated. Beyond specifying standard bond
elements including face value, maturity, and coupon rate and frequency, corporate bond contracts
often include the following:
1. Specification of collateral and mortgages that protect bondholders in the event of default
2. Provisions concerning the seniority of the bond relative to other company debt in the event
of default
3. Various options possessed by the issuer or the bondholder, including call options and
convertibility options
4. Protective covenants limiting corporate actions during the term of the bond
The first two contract features listed above are motivated by the desire to help reduce investor
concerns about potential default. One way to reduce these concerns involves pledging valuable
assets against default. The pledged assets can vary in nature. In some cases, for example, the
collateral takes the form of financial securities such as common stock or other securities that are
held by a trustee (typically a financial institution) specified in the bond contract. Upon default,
the investors can recoup losses by selling the collateral securities. Other bonds might pledge real
estate assets – land and/or buildings. Similar to a standard home mortgage lending arrangement,
the bondholders can foreclose on the mortgaged assets in the event of default. Such bonds
are sometimes termed mortgage bonds. In practice, only a minority of corporate bonds pledge
explicit collateral against default. Many bonds are debentures, a fancy term for an unsecured
bond that is not back by particular, specified collateral.
Default and bankruptcy events can impact multiple types of debtholders for a company.
In this case, the debtholders can differ in terms of seniority. Seniority conveys a preference in
position over other creditors or bondholders in the event of default. Effectively, bonds that have
higher seniority offer investors greater protection against default than more junior bonds. The
jargon subordinated is a synonym for a more junior bond or debt instrument.
Corporate bond contracts often contain various form of options that can be exercised either
by the issuer or by bondholders. Perhaps the most common type is a call provision. Such a
provision allows the the issuer to elect to repurchase some or all of the outstanding bond amount
under specified terms. Often the call provision will be delayed or deferred, meaning that the
issuer cannot call the bond during the first part of the bond’s life (e.g., for the first 10 years). The
key point to recognize regarding call provisions is that issuing companies are likely to exercise
call provisions when interest rates are falling. The intuition is similar to that associated with
‘refinancing’ a home loan. When yields fall, companies with outstanding bonds at higher yields
have an incentive to call the outstanding bonds (if the bond contract allows it) and then issue new
bonds at lower yields.
While it is desirable for the issuer to have a call option in place, the bondholder is less
enthusiastic. This is because the bond is likely to be called exactly when yields are lower, and
3.8 Corporate Bonds and the Structure of Interest Rates – 104 –

therefore the bondholder must reinvest the principal at these lower yields. (This is similar to
what occurs with reinvested coupons when yields fall.) For this reason, investors will demand a
slightly higher yield (pay a lower price) for a bond that includes a call provision relative to an
otherwise identical bond that omits this feature.
A convertibility option provides the bondholder the opportunity to convert the bond to
shares of common stock in the company at a pre-specified conversion rate. The conversion rate
is usually set so that it is only advantageous or profitable to convert if the company’s stock price
increases significantly. The convertibility provision gives the bond additional ‘upside’ potential
that it does not have absent this feature. For this reason, investors will pay more for such a bond
or equivalently accept a lower yield, again relative to an otherwise identical bond that omits this
feature.
Finally, protective covenants amount to a set of restrictions that the issuing firm agrees to
abide by in conjunction with issuing the bond. For example, the bond contract might state that
the issuer agrees not to issue and additional bonds that are senior in priority to the bonds in
question for the life of those bonds. Covenants might also limit activities such as investment
spending, under the logic that this will consume cash that could otherwise be used to pay the
cash owed to bondholders.
The bottom line takeaway is that relevant corporate bond features and contract details are
factored into the prices that investors are willing to pay for these bonds, therefore equivalently
into the yields that investors demand in order to hold the bonds.

3.8.3 Munis, Taxes, and Liquidity

We will not spend much time considering municipal bonds in this course. Municipal bonds
are issued by states and local governments. In many cases, the bonds are issued in order to
finance particular, specified public works projects, such as the construction or renovation of a
local school or hospital, or a new road. Some municipal bonds or “munis” are backed by the
taxing power of the local or state government that issues the bonds. Often taxpayers must approve
the issuance of such bonds via a vote. (You might recall seeing voting ballot items pertaining to
a local municipal bond issuance.) Other munis are backed by the revenue stream associated with
a particular project such as, for example, a toll road.
A key feature of municipal bonds is that the interest income they offer (think coupons) is
exempt from federal taxes and often from state or local taxes (if you buy them in the state or
locality in which you live). In contrast, interest income from US Treasuries is taxed at the federal
level, but exempt from state and local taxes. Because federal tax rates are higher than state and
local rates, municipal bonds have the most favorable tax status among the major classes of US
bonds.
How does the tax status of municipal bonds impact their pricing and yields? Like many of the
corporate bond attributes we discussed in the previous section, the tax status of munis is “priced
3.8 Corporate Bonds and the Structure of Interest Rates – 105 –

in.” In other words, investors bid up the price of a municipal bond relative to a comparable
Treasury or corporate bond that lacks the favorable tax status of munis. Consequently, the
favorable tax status tends to lower the yield on munis – holding all else equal of course. Indeed,
this is part of the motivation for the tax exemption in the first place: the favorable tax status helps
states and local governments raise funds at lower cost. The favorable tax status of municipal
bonds makes them a preferred investment option for relative high income investors. This is
because the marginal tax rate for such investors is the highest and therefore they have the most to
gain from the tax exemption.
A final bond characteristic that influences prices/yields is liquidity. Liquidity refers to the
ease and speed with which a financial security can be bought and sold. A security or asset
that is easy to buy and sell is said to be liquid. An illiquid asset is difficult, costly, and/or time-
consuming to sell. Homes are relatively illiquid, for example. In contrast, common stock for IBM
is extremely liquid. There is enormous trading volume in the stock and relatively large quantities
of IBM shares can be bought and sold quickly with minimal transactions costs. In general, the
bond markets tend to be less liquid than equity (stock) markets. Among different types of bonds,
US Treasury bills, notes, and bonds tend to be relatively liquid securities. Corporate bonds
vary in liquidity. Municipal bonds tend to be relatively illiquid securities. Like the other bond
features we have discussed, the liquidity of a bond impacts the bond yield. It is reasonable to
think that, all else equal, investors desire to hold a liquid security. Consequently, a very illiquid
bond might require a significantly higher yield in order to induce investors to hold the bond given
the difficulty and cost associated with selling it.

3.8.4 The structure of interest rates

Sections 3.7 and 3.8 describe a host of features that influence bond yields. These include
maturity (the term premium), default risk considerations (the default premium), tax and liquidity
considerations, and so forth. The complex pattern of variation in yields across different types of
bonds is sometimes termed the “structure of interest rates.” The following equation provides a
summary of the various components that help determine the yield for any generic bond security:

Bond yield ≈ real yield + inflation premium + interest rate risk premium
(3.14)
+ default premium + liquidity premium + tax/other features component
3.9 Appendix: Cyclical Impacts on the Supply of and Demand for Loanable Funds – 106 –

3.9 Appendix: Cyclical Impacts on the Supply of and Demand for

Loanable Funds

When we describe interest rates and their determinants, we often begin by separating the
question of what determines the general level of interest rates across time and what makes the
interest rate on one bond different from the interest rate on another bond at a particular moment
in time (e.g., today). In this appendix, we focus on what factors affect the general level of interest
rates. Earlier sections of chapter 3 describe the impacts of maturity, credit risk, tax status, and
liquidity on the interest rate for a particular bond at a given point in time.
What makes interest rates change across time? Discussion in section 3.1 identified changes
in inflation rates across time as a factor that affects interest rates across time and causes the
general level of rates to vary across time. For example, inflation rates were high in the late
1970’s and early 1980’s. The high inflation rates pushed all interest rates up during that period.
Additionally, inflation rates tend to vary with macroeconomic activity, with inflation rates rising
as the economy reaches its peak performance and with inflation falling as economic performance
weakens (possibly with inflation becoming negative if the economy shrinks enough). Besides
variation in inflation rates across time, what else might cause rates to change? In this appendix,
we will discuss the impact of the business cycle on the general level of interest rates. For this
discussion, we will hold the expected inflation rate constant and focus on other factors for most
of this discussion. We will place this discussion in the context of the supply of and demand for
loanable funds that we employed in section 3.1. Under the assumption that expected inflation
is constant, movement along the vertical axis in the supply and demand graphs represents an
equal change in both the real and nominal interest rate. For simplicity of our discussion, we will
ignore borrowing by all levels of government and focus attention on the roles of businesses and
households.
As noted in 3.1, the demand for loanable funds reflects the desire of businesses and house-
holds to borrow money. For our purposes, we will think of businesses evaluating potential
investment projects using the NPV approach discussed the first week of class. When firms
calculate an NPV on a proposed investment project, they discount the potential future operating
cash flows that the project may produce at the cost of financing the project (or the interest rate
in this context) and subtract the cost of the investment today. If the present value of the future
operating profits outweigh the cost of the investment, the firm will want to undertake the project.
In looking at the aggregate investment across all firms, investment tends to be higher when the
economy is operating at a high level and it declines when the economy is in a recession.
How might households respond to changes in economic activity and how would that affect
the demand for loanable funds? Households seek to finance the purchase of durable goods by
borrowing all or part of the cost of the durable good. Durable goods are goods whose benefit to
consumers is enjoyed over several periods of time. Durable goods typically have prices that are
3.9 Appendix: Cyclical Impacts on the Supply of and Demand for Loanable Funds – 107 –

high relative to household income. The classic examples of this are homes, automobiles, and
large household appliances. When the economy performs at a higher level, consumer spending
on durable goods typically increases (and vice versa when the economy contracts). An expanding
economy is thus accompanied by an increase in consumer spending on durables and by additional
borrowing to finance those purchases.
?? displays a demand curve for loanable funds. The demand curve is downward sloping in
the interest rate (both real and nominal rates since we are holding expected inflation constant).
As interest rates decline, businesses and households wish to borrow more funds.

Figure 3.20: Demand for Loanable Funds

An increase in GDP and other measures of economic activity is likely to be accompanied

by a greater demand for funds by businesses and households at any interest rate or a shift in the
demand curve to the right. The graph below depicts this with the demand increasing from DLF 1
to DLF 2 .

Figure 3.21: An increase in the Demand for Loanable Funds

The graphs above just show demand curves for loanable funds. We will also have a supply
of loanable funds available in financial markets at each moment. We expect the quantity of
loanable funds supplied by households to increase as the real and nominal interest rate increase
or the supply curve to be upward sloping.
In this framework, the increase in the demand for loanable funds combined with a stationary
supply curve causes the interest rate to rise. The upward pressure on the interest rate encourages
households to lend more funds. The graph illustrates that the shift in demand leads to an increase
in the quantity of funds supplied, an increase in the interest rate, and an increase in the amount
of funds borrowed/lent in the financial market.
The above discussion illustrates the likely impact of an increase in economic activity but
assumes the supply curve for funds will remain constant (although changing rates leads to a
change in the quantity supplied). Do we expect the supply curve to remain stationary with an
expanding economy? Perhaps, we should not expect supply to remain stationary. An economic
3.9 Appendix: Cyclical Impacts on the Supply of and Demand for Loanable Funds – 108 –

Figure 3.22: Impact of an Increase in the Demand for Loanable Funds with Supply Constant

expansion should cause household income and wealth to increase.8 This is depicted in the graph
below where an economic expansion leads to an increase in the supply of loanable funds as well
as an increase in the demand for loanable funds. The increase in supply is illustrated as the shift
in the supply curve to the right or from SLF 1 to SLF 2 .

Figure 3.23: Impact of increases in both Supply and Demand for Loanable Funds

This graph clearly indicates that as the economy expands the total amount lent and borrowed
in financial markets will increase since borrowers want to borrow more funds at each interest
rate and lenders want lend more funds at each rate. What about the interest rate – how should
it change? The answer is that if supply and demand curves are equally impacted by economic
growth (i.e., SLF and DLF curves shift right by the same amount) then the interest rate will not
change. However, if demand for loanable funds is impacted by more than supply, then the interest
rate will rise. On the other hand, if supply of loanable funds is impacted by more than demand,
then interest rates will fall.
Are there other factors that might affect the market for loanable funds? There are two factors
that could affect the supply of loanable funds in addition to the wealth or income effect described

8We should say exogenous wealth or the wealth not associated with the impact of interest rate changes on bond prices.
Since decreases (increases) in interest rates are associated with increases (decreases) in bond prices, part of household
wealth depends on interest rates or is endogenous.
3.9 Appendix: Cyclical Impacts on the Supply of and Demand for Loanable Funds – 109 –

above. The first factor is the impact of a change in economic activity on aggregate credit risk.9
An inspection of data on defaults in US financial markets shows that defaults increase in the
middle or latter stages of recessions and that credit or default spreads increase in advance of actual
defaults. Default is a negative factor for lenders. Recessions are expected to increase defaults
and could further reduce the supply of loanable funds beyond the simple income or wealth effect.
Similarly, improvements in the economy are expected to decrease defaults and could further
increase the supply beyond the simple income or wealth effect. Thus incorporating the impact
of recessions and economic expansions on aggregate credit risk exacerbates the fluctuations in
the supply and demand for loanable funds.
The second factor in addition to the wealth or income effect is the impact of a change in
economic activity on aggregate liquidity. Recall that liquidity refers to the ease and speed with
which a financial security can be bought and sold. An examination of the liquidity in financial
markets indicates that markets tend to be more liquid (i.e., securities can be more easily traded)
when the economy is performing at a higher level and that deterioration in economic performance
can be accompanied by decreases in aggregate liquidity. Since investors consider liquidity to be
a desirable attribute, lenders may be willing to supply more funds when the economy is operating
at a high level than than the simple wealth effect implies. Correspondingly, if the economy
deteriorates and liquidity in financial markets decreases, this suggests that lenders may be less
willing to hold securities and the shift in the SLF curve to the left is larger than the wealth effect
suggests.

9In this discussion, aggregate credit risk refers to the credit risk associated with all private issuers (businesses and
housenolds) as opposed to the credit risk that is entailed when lending to a specific borrower.
3.9 Appendix: Cyclical Impacts on the Supply of and Demand for Loanable Funds – 110 –

K Chapter 3 Problems k
1. Consider the competitive market models of the market for loanable funds and the market
for loanable purchasing power. Suppose that both markets are initially in equilibrium.
Now assume that households’ preference for current consumption increases, all else equal.
Show in standard market model diagrams how this change would impact the equilibrium
nominal interest, equilibrium real interest rate, and equilibrium quantity of lending.
2. A commercial bank offers a one-year certificate of deposit (CD) that pays a 5% nominal
interest rate. If the expected rate of inflation equals 2% per year, what is the real interest
rate that the bank CD offers? Compute this using both the ‘approximate’ Fisher equation
and the exact version.
3. Suppose you and I agree today that the expected inflation rate for the next year is 3%. We
then enter into a one-year loan agreement with a nominal interest rate of 6%. If the actual
inflation rate over the next year turns out to be 2%, explain whether the lender or borrower
benefits, and why.
4. The Survey of Professional Forecasters regularly asks a set of forecasters for quantitative
forecasts of key economic variables, including economic growth rates and inflation rates.
In the third quarter of 2020, the median forecast for the one-year inflation rate was 1.77%
and the median forecast for the 10-year inflation rate was 2.03%. As of early October,
2020, the interest rates associated with one-year US government bonds are around 15
basis points (0.15%) and interest rates associated with 10-year US government bonds are
approximately 70 basis points. Use the Fisher equation to compute estimates of the real
interest rate at one-year and 10-year horizons. Briefly discuss what the results imply for
the real return on savings that bond investors are willing to accept.
5. ABC Inc. bonds currently have 10 years to maturity and pay a coupon rate of 4.5% (annual
payments). Assuming the most recent coupon was just paid and the YTM on these bonds
equals 6%, determine the current bond price, assuming a face value of $100.
6. XYZ Co. issued 15-year bonds one year ago at a coupon rate of 5.8%. The bonds make
semi-annual payments and have a face value of $1,000. if the yield-to-maturity on these
bonds is 4.3%, what is the current bond price?
7. Greenbrier Industrial Products bonds have a 7.60 percent coupon and pay interest annually.
The face value is $1,000 and the current market price is $1,062.50 per bond. The bonds
mature in 16 years. What is the yield-to-maturity (YTM)?
8. The bonds issued by Stainless Tubs bear a 6 percent coupon, payable semiannually. The
bonds mature in 11 years and have a $1,000 face value. Currently, the bonds sell for $989.
What is the yield-to-maturity?
9. Collingwood Homes has a bond issue outstanding that pays an 8.5 percent coupon and
matures in 18.5 years. The bonds have a par value of $1,000 and a market price of $964.20.
Interest is paid semiannually. What is the yield-to-maturity?
3.9 Appendix: Cyclical Impacts on the Supply of and Demand for Loanable Funds – 111 –

10. You purchased an annual interest coupon bond one year ago that now has 6 years remaining
to maturity. The coupon rate is 10% and the par value is $1,000. At the time you purchased
the bond, the yield-to-maturity was 8%. Determine the amount you paid for this bond one
year ago.
11. old exam question A US Treasury note with 5-years to maturity currently trades for
$97.865 per $100 of face value. The bond pays coupons semi-annually. Assuming a 6%
market yield, determine the (annualized) coupon rate for this Treasury note.
12. On October 7, 2020 the Wall Street Journal contained a price quote of$153.2040 (per $100
of FV) for a US Treasury bond that matures on November 15, 2040 and that pays a coupon
rate of 4.25% (semiannually). The WSJ lists the corresponding bond yield as 1.230%. Do
your own bond yield computation and compare the result with the WSJ number. Are they
close? Exactly the same? If there is a difference, what do you think causes this?
13. Suppose a 30-year US Treasury bond was issued in the early 1970s with a 30-year maturity
and a 6.2% coupon rate. Initially the bond was priced at par. Describe how the bond price
likely evolved over the next several decades and, specifically, whether the bond would have
traded at a discount or premium. HINT: recall the history of US interest rates discussed in
Section 3.1.
14. A bond paying a 3% coupon rate (annual payments) with 10 years to maturity has a current
yield of 4.2% (this refers to the ‘current yield’ and not the yield to maturity). Does the
bond trade at a premium, discount, or at par? What are the current bond price (per $100
of face value) and yield to maturity?
15. Consider two zero coupon bonds, both with face value of $1,000. One matures in 6 months
and the other in 18 months. Suppose that the discount rate (YTM) on the bond is 6% (treat
this as an ‘APR’ using semiannual compounding). What are the corresponding prices for
the two zero coupon bonds?
16. A zero coupon bond trades for $98 per $100 of face value. Assuming the appropriate
discount rate for the bond equal 5% (EAR), what is the time to maturity for the bond?
17. What is the yield to maturity on a one-year zero coupon bond that trades at $94.78 per
$100 of FV?
18. What is the yield to maturity on a zero coupon bond with a price of $92.57 per $100 of
face value and 6.3 years to maturity? Does this two ways: 1) by solving for the yield using
the basic time value equation; 2) by solving for the continuously compounded yield. Show
that, after converting the continuously compounded yield to an EAR, you obtain the same
value under both approaches.
19. Consider the following extension of Examples 3.12 and 3.13: now assume that the same
bond (10-years to maturity, coupon rate of 6% (annual payments)) currently has a yield
to maturity of 5%. Determine the one-period HPR for this bond in the following three
scenarios: 1) the yield remains at 5% over the next year; 2) the yield falls to 4% over the
3.9 Appendix: Cyclical Impacts on the Supply of and Demand for Loanable Funds – 112 –

next year; 3) the yield increases to 6% over the next year. Explain what happens to the
bond price in the first scenario and what drives this effect.
20. You purchase a 20 year, 6.5% semiannual coupon rate bond with an 8% YTM today.
Assume the bond is issued by a sovereign government (e.g., Treasury) such that there is
no risk of default.
(a). What price did you pay? (Assume the FV is $100 for convenience.)
(b). If you hold the bond to maturity what will be your rate of return, assuming you can
reinvest intermediate coupon payments at 8%? Why?
(c). Suppose you sell the bond six months from today and the yield is unchanged. What
is your (annualized) holding period return?
(d). Suppose you sell the bond six months from today and yields on this type of bond
have fallen to 7%. What is the annualized holding period return in this case? Discuss
how it compares to the 8% YTM at the time you purchased the bond.
21. You have just purchased a 12-year zero-coupon bond with a yield to maturity of 9% and a
par value of $1,000. Suppose you sell the bond one year from now. Determine the holding
period return, assuming that the YTM on the bond is 10% when you sell.
22. In Example 3.15, an investor purchases a bond with a 6% coupon (annual) and $100 face
value with two years to maturity at a yield of 7%. The investor holds the bond to maturity.
The yield increases to 10% in one year, and then falls back to 7% at maturity. It was shown
that the HPR equals around 7.09%.
(a). What, if anything, changes if we now assume that the yield change to 10% is
permanent?
(b). Now assume that the bond is purchased with three years to maturity, rather than two,
at a yield of 7%. Suppose the yield permanently increases to 10% in the following
year and the investor sells the bond after holding for two years, i.e. with one year
to maturity. Now what is the HPR and how does it compare with the initial yield of
7%?
23. Consider the following three bond positions: 1) a zero coupon Treasury bond held to
maturity; 2) a 4.5% (coupon rate) Treasury bond held to maturity; and 3) a 4.5% Treasury
bond with holding period less than maturity. Briefly discuss the extent and nature of
interest rate risk associated with each of these positions.
24. Consider the following three bonds: 1) a bond with 10 years to maturity and a 5% coupon
rate; 2) a bond with 10 years to maturity and a 10% coupon rate; and 3) a bond with 5
years to maturity and a 10% coupon rate. Rank these bonds in order of least interest rate
risk to greatest. Briefly explain your reasoning.
25. old exam question Suppose you purchase US Treasury securities to fund an obligation
that is due in 5 years. Briefly discuss (1 or 2 sentences) the specific form(s) of interest rate
risk that you face for each of the following specific bond investment strategies:
3.9 Appendix: Cyclical Impacts on the Supply of and Demand for Loanable Funds – 113 –

(a). You purchase a 10-year Treasury note with a 4% coupon rate. You will liquidate
your position (sell the bond) after five years.
(b). You purchase a Treasury STRIP security with five years to maturity and you will
hold this security until maturity.
(c). You purchase a Treasury note with a 3% coupon rate and five years to maturity. You
will hold this security until maturity.
26. A bond portfolio manager recently determined that the duration of her portfolio equals 7.
She is concerned about a potential increase in interest rates, and is evaluating the potential
merits of reducing the duration of her portfolio. Conduct a what if analysis that quantifies
the implications on her portfolio of a potential 50 basis point increase in market yields,
assuming that the current yield is 5%. Quantify the benefit that would be associated with
reducing her portfolio duration from 7 to 5 today, should this rate change materialize.
Discuss potential costs associated with reducing her portfolio duration.
27. Figure 3.24 shows the relation between bond price and yield for three $1,000 face value
bonds with coupon rates of 5% and maturities of 5, 10, and 30 years. The current yield to
maturity is 5%.
(a). Which bond has the highest duration? Which bond has the lowest?
(b). Suppose the yield to maturity increases from 5% to 9%. Which bond will lose the
most value? Which bond will lose the least?
(c). Will the duration of the 30-year bond change at the new 9% yield to maturity? Will
it be higher or lower than at the original 5% yield to maturity? (Hint: Think of the
duration as being the straight line which is tangent to the new bond price.)

Figure 3.24: Relation between price and yield for 3 bonds.

3.9 Appendix: Cyclical Impacts on the Supply of and Demand for Loanable Funds – 114 –

28. State one way in which a financial institution can reduce the duration of its bond portfolio.
29. You invest in a 5-year Treasury STRIP (so no coupon payments) with an annual yield
to maturity of 2.09%. Assume you hold the bond until maturity, so your holding period
return is the same as the yield to maturity. After five-years the actual annual inflation rate
turned out to be 1.82%. What is your realized real rate of return (RRRR)?
30. old exam question Briefly explain how a fixed-income investor with a ten-year horizon
can use a STRIP security to avoid exposure to interest rate risk.
31. Consider a US Treasury bond with 3 years to maturity paying a 2 percent coupon semian-
nually. Suppose that the following table provides STRIP prices at the following maturity
points (per 100 of FV). Compute a price for the 3-year Treasury bond based on these data?
Will actual prices for this bond equal the value you determined? Explain.
Maturity (months) STRIP price
6 99.80
12 99.40
18 99.00
24 98.35
30 98.00
36 97.20
32. Consider the following statement: TIPS bonds are completely free of risk, because they
are not only backed by the full faith and credit of the US government, they also protect
investors against any potential inflation. Is this statement true or false? Explain.
33. The table below has recent yields to maturity (YTMs) for TIPS and traditional Treasury
securities from the Wall Street Journals market section (actually from around 2018):
Security 1-yr maturity 5-yr maturity 10-yr maturity 30-yr maturity
Treasury Sec. 1.100% 2.031% 2.521% 3.135%
TIPS -0.738% -0.011% 0.527% 1.009%
Assuming the TIPS YTM is a good proxy for the real interest rate and the Treasury YTM
is the nominal interest rate, calculate the expected annual inflation for each time-period (1
year, 5 years, 10 years, 30 years). Use the exact calculation rather than the approximation.
Is inflation expected to increase or decrease over time?
34. Provide two reasons why the yield curve may not be flat. Why is it relatively rare that the
yield curve slopes downward? Explain briefly.
35. Discuss how the interest rate risk premium contributes to the shape of the yield curve. In
particular, what feature(s) about the shape of the yield curve does the risk premium help
explain?
36. Consider yields on corporate bonds that differ only with respect to the particular feature
mentioned below. For each case, state which bond would be expected to trade at a higher
yield, and briefly explain your reasoning.
3.9 Appendix: Cyclical Impacts on the Supply of and Demand for Loanable Funds – 115 –

(a). Senior debt (first in line to receive asset liquidation value in the case of default) versus
subordinated debt (a secondary claim in the case of default)
(b). A callable bond versus a non-callable bond
(c). A convertible bond versus a non-convertible bond
37. A Treasury bond due in one year has a yield of 4.3%; a Treasury bond due in 5 years
has a yield of 5.06%. A bond issued by Boeing due in 5 years has a yield of 7.63%; a
bond issued by Caterpillar due in one year has a yield of 7.16%. What are the default risk
premiums on the bonds issued by Boeing and Caterpillar?
 Chapter 4 Federal Reserve and Monetary Policy

4.1 Overview and Readings

This is a very short chapter concerning the Federal Reserve and monetary policy. The
‘chapter’ just consists of a set of review and practice questions pertaining to the material we
discussed in class. In lieu of typing up a complete discussion of the chapter topics, we will rely
on a nice set of publicly available introductory articles put together by the Federal Reserve itself.
These readings are posted on Canvas. For reference, the following provides direct links to the
key articles:
1. Federal Reserve Overview: https://www.federalreserve.gov/aboutthefed/files/
pf_1.pdf
2. Key Federal Reserve Entities and Procedures: https://www.federalreserve.gov/
aboutthefed/files/pf_2.pdf
3. Monetary Policy: https://www.federalreserve.gov/aboutthefed/files/pf_3.
pdf
See also the PDFs of class notes on these topics, also posted on Canvas.
4.1 Overview and Readings – 117 –

K Chapter 4 Problems k

1. True / False questions pertaining to the Federal Reserve and monetary policy. Label each
of the following statements as true or false. If false, briefly explain why the statement is
false.
(a). The Federal Reserve decreases the monetary base and money supply whenever it
sells government securities.
(b). Stable employment is one of the objectives of monetary policy.
(c). The Federal Open Market Committee consists of the 12 Presidents of the Fed Regional
Banks.
(d). The Fed exclusively controls the money supply.
(e). The Fed has never specified a long-run goal for the inflation rate.
(f). The Fed’s “QE” purchases have included large quantities of agency. mortgage-backed
securities (MBS), Treasury bonds, and corporate bonds and equities.
(g). Pledging to keep short term interest rates “lower for longer” has been a strategy
adopted by the Federal Reserve at the zero lower bound.
2. Suppose that The Fed decides to buy $10,000 of government securities from Goldman
Sachs. What impact would this have on the Federal Funds rate when the current Federal
Funds rate is above the zero lower bound (ZLB)? What if the current Federal funds rate is
at the ZLB?
3. Briefly discuss key similarities and differences regarding quantitative easing (QE) versus
traditional open market transactions conducted by the Federal Reserve.
4. Describe how the interest rate on reserves paid by the Federal Reserve contributes to
monetary policy. In particular, how can the IOR be used to to manage the Fed Funds rate
target in the current situation in which excessive reserves exist in the banking system?
5. The FOMC met most recently on Nov. 2–3, 2021. As is the case with all FOMC
meetings, the committee released a statement discussing the stance of monetary policy
immediately following the meeting. Read this statement, accessible here: https://www.
federalreserve.gov/newsevents/pressreleases/monetary20211103a.htm.
(a). What is the announced target range for the federal funds rate?
(b). Does this represent a change, or a continuation of the previous target range?
(c). What are the current FOMC goals with respect to employment and inflation rates?
(d). Does the announcement specify what is the maximum employment rate target?
(e). What asset purchases does the committee intend to make in the immediate future?
 Chapter 5 Equity and Risk-Return Tradeoffs

5.1 Equity: Basic Institutional Details

Previously we have discussed corporate bonds, which are financial securities that constitute
an implicit form of lending arrangement in which the firm receives bond sale proceeds in return
for future cash flow obligations to bondholders (coupons and face value payments). Companies
also issue other forms of financial securities to raise funds. In particular, companies typically
also issue stocks or equity claims. In contrast to bonds, stocks convey to the holder a form of
(partial) ownership claim in the issuing firm, often accompanied by certain rights such as voting
rights with respect to the board of directors for the company.
Definition 5.1. Stock
Stock (also known as ’equity’ or ‘capital stock’) of a corporation represents a financial
claim that conveys to the holder a proportional ownership claim on the issuing firm. A
single share of the stock represents fractional ownership of the corporation in proportion
to the total number of shares. Typically, this entitles the stockholder to the corresponding
fraction of earnings paid out in the form of dividends, proceeds from liquidation of assets
(after satisfying debt claims), and voting power.
♣

Figure 5.1 shows an old stock certificate from the late 1970s for 1,000 shares of Motorola,
Inc., which is now Motorola Solutions, Inc., and S&P 500 firm with shares trading on the New
York Stock Exchange.

Figure 5.1: Old Stock Certificate for 1,000 Shares of Motorola, Inc.
5.1 Equity: Basic Institutional Details – 119 –

The value of stock ultimately derives from shareholders’ claims to earnings paid out in the
form of dividends.
Definition 5.2. Dividend
A dividend is the distribution of some of a company’s earnings to a class of its shareholders,
as determined by the company’s board of directors. Shareholders are typically eligible
as long as they own the stock before the ‘ex-dividend date’ – a date specified after the
dividend amount is announced but prior to the date upon which it is paid.
♣

Dividends are often paid quarterly and may be paid in the form of cash or in the form of
additional company stock. In determining dividend policy, the company (and its directors) must
consider the question of how much earnings to retain (‘retained earnings’ on a firm’s balance
sheet) versus how much to pay out to current shareholders as dividends. Retained earnings
provide the firm with an internal source of funds for future investment and growth, as opposed
to the firm having to raise capital via loans or security issuance. Consequently, young firms with
significant growth opportunities often elect to retain all earnings, i.e., not to pay a dividend. The
companies that pay a regular dividend therefore tend to be older, more mature, and more profitable
firms. Over time, the proportion of US stocks that pay a regular dividend to common stock holders
(see below for the definition of common stock) has decreased substantially. Currently fewer than
20% of firms pay a regular dividend. At the same time, the aggregate dollar amount of dividends
paid has increased, and therefore dividend payout is increasingly concentrated among a relatively
small number of large, mature, and profitable companies.
Companies often issue different types of stock, known as stock classes. The different stock
classes vary with respect to the priority structure of receiving dividends, the extent of voting
rights conveyed, and on other dimensions. Broadly, the most important distinction among stock
classes involves common stock versus preferred stock.
Definition 5.3. Common Stock
This type of share gives the stockholder the right to share in the profits of the company,
and to vote on matters of corporate policy and the composition of the members of the
board of directors.
♣

Common stock is the most prominent type of stock in the sense of representing most of
the outstanding value of issued stock. Common stock is a ‘residual claim’ on the firm: holder
are entitled (proportionally) to the entire value of the firm net of debt and preferred stock (see
below) claims. This gives common stock more ‘upside potential’ than bonds or preferred stock.
However, it is also riskier in the sense that, in the event of liquidation, the claims of common
stock holders are behind debt claims as well as the claims of preferred shareholders. In practice,
common stock holders are frequently ‘wiped out’ (receive no value) in such cases. Common stock
holders usually have voting rights that enable them to determine the composition of the board
of directors, vote on CEO pay issues, and other important matters such as potential acquisition
5.1 Equity: Basic Institutional Details – 120 –

offers.
Definition 5.4. Preferred Stock
Preferred stock is a form of stock that provides holders with a higher claim to dividends
or asset distribution relative to common stockholders. ♣

Preferred shareholders have priority over common stockholders with respect to dividends.
The dividends to be paid to preferred stock holders can be fixed or set in terms of a benchmark
interest rate, such as the London Inter-Bank Offer Rate (LIBOR) and are often quoted as a
percentage of par value in the issuing description.
The fact that preferred stock specifies a fixed or variable (with respect to a benchmark interest
rate) dividend payment to shareholders makes preferred stock somewhat similar to corporate
bonds. Moreover, in the event of a liquidation, preferred stockholders’ claim on assets is greater
than common stockholders, although less than bondholders. There is an important distinction
between bonds and preferred stock, though. Failure to pay promised coupons to bondholders
represents contractual default. In contrast, the decision to pay a dividend to preferred shareholders
is at the discretion of a company’s board of directors. Failure to pay the dividend specified in the
preferred share does not amount to default.
Preferred stock features often include:
Cumulative Provision: This form of provision states that, if the company omits dividend
obligations to preferred shareholders, these obligations must be paid in arrears before
common shareholders can receive additional dividends.
Call provision: similar to call provisions in corporate bond contracts, this provision allows
the company to buy-back the shares at par value after a specified date
Convertibility provision: similar to convertibility options for corporate bonds, such a
provision allows for the potential conversion of the preferred stock to common stock at a
specified rate. In some cases this option is possessed by the shareholders, or alternatively
the conversion might occur automatically at a specified date or upon the decision of the
board of directors.
In contrast to common stock, preferred stock often does not convey voting rights to holders,
although in some cases holders receive voting rights under certain conditions, such as if the
company is unable to pay the promised dividends in a timely fashion.
Similar to bonds, there is a primary and secondary market for corporate stock. The primary
market denotes transactions in which firms raise funds by issuing new shares of equity. An
initial public offering (IPO), as the name suggests, represents the first time that a firm issues such
shares to the public. A seasoned offering represents the issuance of additional shares by a firm
that has previously issued public shares. Companies typically work with one or more investment
banks in the issuance process, and the newly issued shares are generally ‘placed’ primarily with
institutional investors such as mutual funds. The secondary market refers to markets in which
existing shares of equity are traded among investors. The New York Stock Exchange (NYSE)
5.2 Stock Returns – 121 –

and Nasdaq are the two most prominent marketplaces for equity in the United States.
Figure 5.2 shows an excerpt from the Yahoo! Finance site for Exxon-Mobil Corporation
common stock (’ticker symbol’ XOM). Exxon-Mobil common stock is ‘listed’ on the NYSE and
is an extremely liquid stock, as evidenced by the very high trading volume indicated in Figure
5.2 as well as the very small ‘spread’ between bid and ask prices for the stock. (We will not
cover the details of market trading mechanism in detail in this class, but the bid ask spread is
one measure of the trading cost associated with a security.) The closing price of the stock was
$33.17 on November 6, 2020. The stock pays a regular quarterly dividend – note the positive
‘dividend yield’ indicated in the figure.

Figure 5.2: Recent stock price data for Exxon-Mobil

5.2 Stock Returns

The realized return to holding stock derives from dividends received and from appreciation
in the stocks market value. Consider first the ‘single holding period’ return for a stock. In this
case, we hold the stock for one ‘period,’ which might be a day, a month, a year, etc. In this case,
the corresponding return equals:

(Pt+1 − Pt ) + Dt+1 (Pt+1 − Pt ) Dt+1

Rt+1 = = + (5.1)
Pt Pt Pt
The formula for the single-period return on a stock is nearly identical to the analogous
formula for a bond, except that instead of the coupon payment received by bondholders, the
shareholder (potentially) receives a dividend.1 The second equality breaks the holding period
return into two components: 1) a capital gain (or loss) determined by change in the share price
over the holding period relative to the initial share price; and 2) the dividend yield (Dt+1 /Pt ),
which is equal to the ratio of the dividend received to the initial share price.

1We will use the simpler notation Rt+1 rather than HP Rt+1 for single-period returns, reserving the ‘HPR’ notation
for multi-period cases discussed below.
5.2 Stock Returns – 122 –

If a stock does not pay a dividend during the holding period, then the dividend yield
component of the return simply equals zero. It is also worth remarking that the return given in
Eq. 5.1 is only an annualized return if the holding period equals a year. In order to compute an
annualized HPR, one can use the following formula:
Ann.
Rt+1 = (1 + Rt+1 )m − 1, (5.2)

where m is the number of periods within one year, e.g., m = 4 if the stock is held for one quarter.

Example 5.1 You purchase shares of ABC Company common stock at the beginning of a quarter
for a price of $20/share. The following table shows the beginning and end of quarter share prices
for the next four quarters, as well as the dividend paid in each quarter. Complete the table by
computing the corresponding quarterly returns for each of the four quarters.

Quarter Starting price Ending price Dividend Returns

1 $20 $23 $0 ?
2 $23 $27.75 $0 ?
3 $27.75 $24 $1 ?
4 $24 $20 $0 ?

Solution: In each case we apply the formula of Eq. 5.1. For the first quarter, this gives:
(Pt+1 − Pt ) Dt+1
R1 = +
Pt Pt
(23 − 20) 0
R1 = +
20 20
R1 = 15%

Proceeding similarly for the second quarter:

(Pt+1 − Pt ) Dt+1
R2 = +
Pt Pt
(27.75 − 23) 0
R2 = +
23 23
R2 = 20.65%

A non-zero dividend is paid in the third quarter, which must be included in the return calculation:
(Pt+1 − Pt ) Dt+1
R3 = +
Pt Pt
(24 − 27.75) 1
R3 = +
27.75 27.75
R3 = −9.91%
5.2 Stock Returns – 123 –

Finally, for the final quarter we have:

(Pt+1 − Pt ) Dt+1
R4 = +
Pt Pt
(20 − 24) 0
R4 = +
24 24
R4 = −16.67%

The completed table is therefore:

Quarter Starting price Ending price Dividend Returns

1 $20 $23 $0 15%
2 $23 $27.75 $0 20.65%
3 $27.75 $24 $1 -9.91%
4 $24 $20 $0 -16.67%

Suppose now that an investor holds a stock for multiple periods. Specifically, suppose the
investor buys the stock at the end of period t and holds the stock for H periods (think of H as
the investor‘s ‘horizon’ in terms of the number of holding periods). Then the multi-period HPR
formula is given by:

HP Rt,t+H = (1 + Rt+1 )(1 + Rt+2 ) . . . (1 + Rt+H ) − 1 (5.3)

The holding period formula of Eq. (5.3) is not annualized. Suppose, for example, we
measure periods in terms of years and an investor holds a stock for five years. Then the HPR
computed using Eq. (5.3) will equal a five-year return. If we want to express the holding period
return on an annualized basis, we can apply the ‘EAR’ conversion formula from the time value
of money segment (see Chapter 2):
Ann.
HP Rt,t+H = [(1 + Rt+1 )(1 + Rt+2 ) . . . (1 + Rt+H )]m/H − 1, (5.4)

where m is the number of periods per year and H is the number of periods for which the stock
is held. In the five-year example discussed above, H = 5 and m = 1.

Example 5.2 Consider the quarterly stock price and return data from the previous example.
Assume you buy the stock at the beginning of quarter 1 and hold it either until the end of quarter
2 or the end of quarter 4. Compute the 2-quarter and 4-quarter HPRs for these cases. Then
convert the 2-quarter HPR to an annualized HPR and compare it with the (already annualized)
4-quarter HPR. What explains any difference?
Solution: Start with the 2-quarter HPR case. Plugging into the HPR formula gives:

HP Rt,t+2 = (1 + Rt+1 )(1 + Rt+2 ) − 1

HP Rt,t+2 = (1 + 0.15)(1 + 0.2065) − 1
= 38.75%
5.3 Stock Returns: Probability Model and Key Statistics – 124 –

This is a two-quarter (6 month) return. In order to annualize the return, we set m = 4 and H = 2
and obtain:
Ann.
HP Rt,t+2 = [(1 + Rt+1 )(1 + Rt+2 )]4/2 − 1 = 1.38752 − 1 = 92.51%

Turning to the 4-quarter case, the HPR equals:

HP Rt,t+4 = (1 + Rt+1 )(1 + Rt+2 )(1 + Rt+3 )(1 + Rt+4 ) − 1

HP Rt,t+4 = (1 + 0.15)(1 + 0.2065)(1 − 0.0991)(1 − 0.1667) − 1
= 4.16%

Because this is a 4-quarter return, it is already annualized. The annualized 2-quarter HPR is
much higher than the 4-quarter HPR because the stock price rose significantly over the first two
quarters, only to fall back to the starting price of $20/share at the end of the holding period. As
an exercise, you should be able to verify that the only reason the 4-quarter HPR is not zero is
due to the $1/share dividend received in the 3rd quarter of the holding period. (To check this,
re-compute the 3rd quarter return assuming no dividend, and then recompute the 4-quarter HPR.)
The previous example reveals something subtle about multi-period HPR calculations. Notice
that rough intuition might have suggested a 4-quarter HPR of 5%: the starting and ending price
are both $20, so that there is no capital gain component, and the investor receives a total dividend
of $1 on a $20 initial investment. But our calculations show that the HPR is actually slightly
less than this. Why? This occurs is because the HPR computation implicitly assumes that all
dividends received during the holding period are re-invested in the stock. So our $1 dividend at
the end of quarter 3 is reinvested for the final quarter, which had a -16.67% return (see the table).
It is exactly this effect that leads to the lower 4.16% HPR.

5.3 Stock Returns: Probability Model and Key Statistics

Future stock prices are subject to uncertainty: an investor who buys stock today cannot
with certainty predict the path of the stock price in the future. Indeed, directional movements in
stock prices are notoriously hard to predict. This uncertainty and lack of predictability actually
makes a great deal of economic sense: if everyone agreed that stock prices tomorrow were
going to increase, then everyone would buy stock today to exploit this insight. But this would
drive up today’s price of the stock upward until there is considerable uncertainty about whether
tomorrow’s price will be higher or lower than today’s price. This property of stock prices is
known as market efficiency:
Definition 5.5. Market efficiency
An efficient market is a market in which the price of a security fully reflects all available
value-relevant (public) information.
♣

Loosely, if a market is efficient, then there is no way to “beat the market”because there are
5.3 Stock Returns: Probability Model and Key Statistics – 125 –

no ‘undervalued’ or ‘overvalued’ stocks (or other securities). The prices of securities completely
reflect all value-relevant information. The extent to which actual financial markets satisfy this
property is an important and classic topic in finance. For example, the entire premise of so-called
‘technical analysis’: methods intended to derive signals of future price movements based on
patterns of past price movements, is that markets are to some degree not fully efficient. The
success of such strategies is a topic of ongoing debate within academic and practical areas of
finance.

5.3.1 A Simple Probability Model for Returns

Given the inherent unpredictable nature of future stock prices, it is useful to model (or
formally think of ) stock returns using the tools of probability. That is, we view stock returns as
random variables subject to uncertain forces that can be modeled using the tools of probability.
Let us start with a very simple ‘blackboard’ probability model for stock returns. The model
is intended to provide intuition behind random variation in returns and introduce some of the key
measures that characterize probability distributions (means, variances, and standard deviations).
The model is highly stylized and simple. In particular, we will assume that all uncertainty
pertains to the realization of a small number of possible economic ‘states.’
Table 5.1 lists three possible economic states, along with their corresponding probabilities
or occurring. The random variables RA and RB equal the returns on two stocks (company A
and company B) in each possible economic state. There is also a risk-free asset Rf that pays a
constant return of 4.5% in each possible state.

State # State Description Probability RA RB Rf

1 Economic Boom 30% 14% 8% 4.5%
2 Normal 50% 8% 4% 4.5%
3 Recession 20% -6% 6% 4.5%
Table 5.1: A simple probability model for stock returns.

In the model, the economy is in a ‘normal’ state 50% of the time. A boom occurs with 30%
probability, and a recession or ‘bust’ with 20% probability. The states are (arbitrarily) number
1 to 3. Note that these probabilities sum to one as they must. (See the Appendix for a more
general review of probability in a finite state model such as this example.) Each realization of
the economic is associated with particular returns for stocks A and B. Stock A is a ‘pro-cyclical’
stock that performs best when the economy booms, and looses value in recessions. Stock B
is a more unusual stock: it’s returns are relatively stable across the three states and it actually
performs worst in the normal state. This might be a firm that has a segment of its business that
is actually counter-cyclical and is particular profitable during sessions, for example.
Suppose you had to choose between investing funds in stock A or B. Which would you
choose? In deciding, it might be useful to summarize the random behavior of returns in several
 5.3 Stock Returns: Probability Model and Key Statistics – 126 –

ways. A first important metric is the average return (also known as the mean return or the
expected value). The average return for stock A equals the weighted sum of the returns in each,
where the weights are equal to the state probabilities:

X
3
E(RA ) = µA = Prob(S = s)Rs,A
s=1

= 30% × 14% + 50% × 8% + 20% × −6%

= 7%

The notation E(RA ) and µA are just alternative common ways of referring to the mean of a
random variable. (The E()˙ is reference to ‘expected value’.) Similarly, for Stock B we can
compute the mean as:

X
3
E(RB ) = µB = Prob(S = s)Rs,B
s=1

= 30% × 8% + 50% × 4% + 20% × 6%

= 5.6%

These calculations reveal that Stock A provides the highest return on average, followed by
Stock B, which also provides a higher average return relative to the risk-free asset.2
Next we turn to the variance and standard deviation of the random stock returns. The
variance is a measure of the dispersion in a random variance and equals the probability-weighted
average of the squared deviations from the mean. For stock A, the computation is:
X
3
2
Var(RA ) = σA = Prob(S = s)(Rs,A − µA )2
s=1

= 30% × (14% − 7%)2 + 50% × (8% − 7%)2 + 20% × (−6% − 7%)2

= 30% × (7%)2 + 50% × (1%)2 + 20% × (−13%)2
= 0.0049

The variance of stock A is 0.0049. Unfortunately, this number is in ‘squared return’ units and
is therefore difficult to interpret. Converting to the standard deviation by taking the square root
gives a quantity that is in the same units as the underlying returns, which is convenient for
interpretation:
p
σ(RA ) = σ 2 (RA )
√
= 0.0049
= 7%

2It is important that the risk-free rate provides a return greater than 4% in this setting. If it did not, then Stock B would
‘dominate’ the risk-free asset in the sense of providing a higher return in all possible states. Risk-averse investors
would not be willing to hold the risk-free asset in this case.
5.3 Stock Returns: Probability Model and Key Statistics – 127 –

Similar calculations deliver the variance and standard deviation of Stock B:

X3
Var(RB ) = σB 2
= Prob(S = s)(Rs,B − µB )2
s=1

= 30% × (8% − 5.6%)2 + 50% × (4% − 5.6%)2 + 20% × (6% − 5.6%)2

= 0.003
p
σ(RB ) = σ 2 (RB )
√
= 0.003
= 1.74%

The computations reveal that, although Stock B has a lower average return than Stock A, Stock
B also has a lower standard deviation. In other words, returns on Stock B are less volatile,
confirming the intuition from the table of outcomes indicating that Stock B outcomes are more
‘stable.’
Finally, it is trivial to verify that the standard deviation of the risk-free asset is zero. The
table below summarizes the key characteristics of the three available assets:

Asset µ σ
Stock A 7% 7%
Stock B 5.6% 1.74%
Risk-free asset 4.5% 0.0%

Note the ‘risk-return’ trade-off that exists among these three assets. An investor who wishes to
earn a higher average return must be willing to take on more risk as measured by the standard
deviation of the return. We will see various forms of this trade-off as we move forward.

5.3.2 A More General Probability Model for Stock Returns

The previous example provides helpful intuition, but it is too simple to be of much practical
use. In reality, there are recessions and expansions of differing intensity, and stock returns
are driven by more than just broad macroeconomic conditions (industry and ‘firm-specific’
circumstances matter as well). We can generalize our model for the return on an asset, such as a
stock or a bond, in the following way:

Ri,t+1 = µi + σi ϵi,t+1 , (5.5)

where µi and σi > 0 are constants and ϵi,t+1 is a ‘random shock’ that equals zero on average
and has a standard deviation of one. It is worth walking through this equation component-by-
component. The component µi is the average return on the asset (equivalently the mean or
5.3 Stock Returns: Probability Model and Key Statistics – 128 –

expected value). The component σi equals the standard deviation of the asset return.3 Recall
from probability and statistics that the standard deviation of a random quantity captures the
extent of its variation around the mean. Finally, the ϵi,t+1 component represents the surprise or
unpredictable component of the asset return. A positive realization of ϵi,t+1 can be interpreted as
‘good news’ next period that increases the stock return, whereas a negative realization of ϵi,t+1
reflects ‘bad news’ that decreases the return. This good or bad news comes from many sources.
For example, there might be good or bad news about the entire macroeconomy that contributes
to ϵi,t+1 . In addition, good of bad news at the industry level, or even ‘firm-specific’ news, such
as the outcome of vaccine efficacy trial for a pharmaceutical company, contributes to ϵi,t+1 . The
shock ϵi,t+1 is normalized to be on a common scale (standard deviation of one) for all securities
– think of it as putting ‘good versus bad news’ on a common scale. The parameter σi then drives
how volatile each asset’s return is by scaling the underlying good/bad news shock each period.

5.3.3 Return Shock Distribution

The model of Eq. (5.5) highlights two key parameters governing risky returns: the average
return µi and the standard deviation of returns σi . Although informative, these two attributes
do not entirely determine the probabilistic behavior of returns. In order to do this we must
specify the distribution of the return shock ϵi,t . A natural benchmark case is to assume that ϵi,t
follows the standard normal distribution. Recall from probability and statistics that the standard
normal distribution has a mean of zero and a standard deviation of one, consistent with the
assumptions we made on ϵi,t . Figure 5.3 plots the density of a standard normal distribution. The
density function for the standard normal displays a famous ‘bell curve’ shape. The figure also
summarizes key probability intervals associated with the standard normal. For example, a value
in the range [−2, 2] will be obtained with approximately 95% probability under the standard
normal. When the shock distribution ϵi,t is standard normal, then the stock return given by Eq.
(5.5) follows a normal distribution with mean µi and standard deviation σi .
Although the standard normal distribution is a natural benchmark assumption for the be-
havior of return shocks, it is important to understand that this is not the only possibility and
not the most realistic assumption. Two specific types of deviations from the standard normal
distribution are worth discussion. The first possibility is that the distribution of return shocks
might be skewed. A positively skewed random variable has a density function that has a ‘long
right tail.’ This implies that there is a substantially greater likelihood of obtain extremely large
(positive) realization relative to an extreme small (or negative) realization.4 A negatively skewed

3Mathematically, this is true because the average return equals µi and the standard deviation of the shock ϵi,t+1 = 1
by assumption. More explicitly, the variance of the return equals:

Var(Ri,t+1 ) = Var(µi + σi ϵi,t+1 ) = σi2 Var(ϵi,t+1 ) = σi2 (5.6)

where the second equality holds because µi and σi are constants. Since the σi2 is the return variance, σi is the return
standard deviation.
4A classic example of a positive distributed random variable is the payoff on a typical lottery ticket. With very high
5.4 Risk-free returns, risk premiums, and Sharpe ratios – 129 –

Figure 5.3: Density plot for the Standard Normal Distribution

random variable has the opposite properties, i.e., the density function has a long ‘left tail’ and
extremely low realizations are more likely than extremely high realizations. Figure 5.4 contrasts
the density plots of so-called ‘skew-normal’ shocks that follow a distribution that generalizes
the normal distribution to allow skewness. Positive (negative) values of the parameter α imply
positive (negative) skew.5
Return shocks can also be fat-tailed. This means that they can deliver extreme realizations
more frequently than the normal distribution would imply. An important distribution with this
feature is the t-distribution (also known as ‘Student’s t-distribution). Figure 5.5 compares the
density plots for several versions of the t-distribution with the standard normal. The key point to
notice is how the t-distributed variables have ‘fatter tails’ – it is more likely to obtain observations
greater or less than 3 standard deviations from the mean relative to the normal.
It turns out that there is strong evidence that most asset returns (stock returns, corporate bond
returns, returns on currencies, etc.) do not appear to be consistent with the normal distribution.
In particular, there is much evidence indicating that asset returns appear to be fat-tailed. This is
important in assessing the nature of risk associated with these investments, as it indicates that the
usual ‘confidence intervals’ associated with the normal distribution are not likely to be reliable.
There is also evidence of skewness for various types of returns. For example, individual stock
returns tend to be positively skewed.

probability, the ticket-holder experiences a small loss, but with a small probability the holder wins the lottery and
receives an extremely large payoff.
5Note that the means of the skew-normal random variables depicted here are not the same and generally not zero. This
can be corrected by adding/subtracting the non-zero mean to obtain a zero mean version.
5.4 Risk-free returns, risk premiums, and Sharpe ratios – 130 –

Figure 5.4: Density plots for ‘Skew-normal’ distributed shocks

Figure 5.5: Density plots for t-distributed shocks

5.4 Risk-free returns, risk premiums, and Sharpe ratios

A special case of Eq. (5.5) holds when σi = 0. In this case the asset return is certain (no
variation) and so this corresponds to a risk-free security. In this case we use the notation Rf for
the risk-free return. By assumption, there is no variation in such a risk-free return, so its average
value also equals Rf . If two or more risk-free assets exist, they should have the same return,
otherwise an arbitrage opportunity – essentially a sure-fire money machine – would exist. To see
this, suppose there are two risk-free assets one offering a sure 5% return and the other offering
a sure 3% return. In such a case, investors would have a massive incentive to ‘short-sell’ the
risk-free asset offering a 3% return and use the proceeds to invest in the alternative risk-free asset
5.4 Risk-free returns, risk premiums, and Sharpe ratios – 131 –

offering 5%. As investors take advantage of this opportunity, they will push down prices (and
hence push up returns) on the risk-free asset offering 3% and push up prices (and hence push
down returns) on the risk-free asset offering 5%. If there are no transactions costs, this will occur
until the two risk-free rates become the same. We will therefore assume that there is at most one
risk-free asset offering the sure return Rf .
A risk-free asset provides an important benchmark and comparison point for alternative
risky assets. Consider the mean or average return µi on a particular risky asset, such as Exxon-
Mobil stock. It might be tempting to think that this expected return should be zero in light of
the ‘unpredictable’ feature of stock prices discussed above. But this is unlikely to be the case,
especially when there is a risk-free asset that offers a positive return Rf > 0. This is because most
investors are risk averse and would prefer to hold the risk-free asset rather than Exxon-Mobil if
Exxon-Mobil stock did not have a higher average return than risk-free option. So it is likely that
the average return µi is positive for Exxon-Mobil stock and greater than that of the risk-free asset.
The difference between the average return on a risky asset and the risk-free return is sufficiently
important that there is particular jargon for it:
Definition 5.6. Risk premium
The risk premium associated with a risky asset equals the difference between the average
return for the risky asset and the risk-free return:

Risk premiumi = µi − Rf (5.7)

The term ‘risk premium’ comes from the presumption that the difference in average return
between a security with uncertain (risky) future payoffs and a risk-free (sure) future payoff
represents the compensation investors demand in the form of a higher average return for
holding the risky security.
♣

It is worth noting that the idea of a risk premium applies to risky assets in general and does
not solely pertain to equity. For example, Chapter 4 discussed bonds and in particular the ideas
the interest rate risk premium for long term Treasury bonds and the default premium for corporate
bonds. These types of bonds are risky securities – returns depend on unknown future outcomes
involving yield changes and default activity. The ‘interest rate’ and ‘default’ descriptors simply
specify particular sources of uncertainty that is viewed as driving unpredictable variation in
returns that investors demand compensation for bearing.
Notice that the risk premium has an i subscript, which implies that the risk premium can vary
across different stocks and other risky securities. Why does it vary? Intuitively, it presumably
varies because different securities involve different degrees of risk, and therefore different degrees
of investor compensation in the form of a risk premium. But what is the right measure of risk?
This is arguably the central question of finance, or at least the area of finance that is often labeled
‘asset pricing.’ We seek answers to the following critical questions:
5.4 Risk-free returns, risk premiums, and Sharpe ratios – 132 –

1. How does the risk premium vary across the wide range of risky assets available (different
stocks, government bonds, corporate bonds, real estate, etc.)?

2. What concept of ‘risk’ best explains the variation in risk premia that we see across assets?

The answers to these critical questions are not obvious and remain debated among scholars and
practitioners. One natural conjecture is that the quantity σi captures the relevant notion of risk
that drives the premium investors require to hold a risky asset. This reasoning would suggest that
µi (and hence the risk premium µi − Rf ) should be increasing with σi . In evaluating the quality
of an investment opportunity, one might compare the risk premium earned per unit of risk as
measured by the standard deviation of the investment return. This called the ‘Sharpe ratio’ (after
economist William Sharpe) or more descriptively the reward-to-risk ratio:
Definition 5.7. Sharpe Ratio
The Sharpe ratio for a risky asset (SRi ) is the ratio of the asset risk premium to its standard
deviation:
µi − R f
SRi = (5.8)
σi
♣

The Sharpe ratio is sometimes called the ‘reward-to-risk’ ratio, with the risk premium
representing the ‘reward’ and the standard deviation of the return representing the ‘risk’ measure.
A higher Sharpe ratio is therefore presumably an attractive feature for an asset.
Although the idea that risk premia should be determined by σi seems initially compelling,
this is NOT the conclusion of the theory of modern finance. We will discuss this point in detail
in the next Chapter, but for now it is sufficient to mention the main problem with this argument:
a risky asset’s return variation is partly driven by idiosyncratic sources that ‘smart’ investors
can (and should) avoid by forming portfolios. Therefore, this idiosyncratic component of σi
should not get a ‘reward’ because there is no need for an investor to bear this type of risk. The
conclusion is then that only sources of risk that cannot be avoided by forming portfolios should
receive compensation. The next chapter further develops these ideas.
There is at least one other possible objection to the idea that the risk premium should be
drive by σi alone. This is that it ignores the role of the shape of the distribution of ϵi,t+1 . In
other words, although the standard deviation of a random variable is a measure of risk, it is not
the only possible measure of risk. To appreciate this, consider two stocks, A and B such that
σA = σB , i.e. their stock returns have identical standard deviations. Suppose, however, that ϵA
is positively skewed whereas ϵB is negatively skewed. In essence, stock A has shocks that are
‘lottery like’ with a relatively high likelihood of a small negative piece of news, but the potential
of a very large positive shock (think of a home-run pharmaceutical development). Stock B, in
contrast, often exhibits small positive shocks but is subject to the possibility of extreme bad
news outcomes. (Think of the potential of the development of a new technology that threatens
Firm B’s entire business model.) It is plausible that investors do not perceive these stocks as
 5.5 Measuring Characteristics of Risky Asset Returns – 133 –

equivalently ‘risky’ just because their returns have the same standard deviation. In particular,
stock B is arguably riskier given its potential for a large capital loss. If this is the case, then stock
B should command the higher risk premium, despite the fact that the two stocks have identical
standard deviations.

5.5 Measuring Characteristics of Risky Asset Returns

The previous section discusses various characteristics of risky asset returns, including the
mean, standard deviation, the risk premium, and Sharpe ratio. But how do we learn about
these characteristics in the ‘real world’? One obvious approach is to examine data consisting of
historical returns on various assets. In this section, we examine a dataset of historical returns on
several benchmark risky asset classes. The data consist of annual returns from 1926–2019 on
the following assets6:
1. A portfolio of Short-term Treasury bills
2. A portfolio of Long-term Treasury bonds
3. A portfolio of investment-grade corporate bonds
4. A market-cap weighted portfolio of the 500 stocks in the ‘S&P 500’ index, which consists
of 500 relatively large-cap stocks listed on US stock exchanges.
The short term Treasury bill returns are a natural proxy for returns on a risk-free asset. We also
include data on annual US inflation rates over the same period for comparison. The underlying
data are available for you to download on Canvas.

5.5.1 Cumulative Returns Across Asset Classes

As a first perspective, let us compare the cumulative return on each of the alternative
portfolios over a long history. Imagine that we invested $1 in each of these portfolios at the
beginning of the year 1926 and the position is held all the way until the end of the year 2019.
(It is implicitly assumed that all income such as bond coupons or dividends are reinvested in the
portfolio.) Figure 5.6 tracks the cumulative value of this investment over time. The figure also
plots cumulative inflation over the same horizon, which is informative concerning the cumulative
real return (over and above inflation) for each investment option. Note that the vertical axis is
presented in a logarithmic scale (base 10). The first key takeaway is that the cumulative return
for the investment in the S&P 500 stock portfolio is much greater than the bond alternatives. A
single dollar invested in the S&P 500 in 1926 grows to around $8,060 in 2019. In contrast, $1
invested in Treasury (corporate) bonds grows to approximately $257 ($158) by 2019. Both the
Treasury and corporate bond cumulative returns are greater in turn than the cumulative risk-free

6Rather than individual bonds or stocks, these are all portfolios, or weighted combinations of positions in multiple
underlying stocks and bonds, which we discuss in detail in the following chapter. It is difficult to analyze the properties
of individual stock or bond returns over a long horizon of time because, for example, relatively few stocks survive
over multiple decades as firms tend to either fail or be acquired.
5.5 Measuring Characteristics of Risky Asset Returns – 134 –

(Treasury bill) return. A dollar invested in Treasury bills grows to around $22 by 2019. This is
greater still than cumulative inflation (about 14). In other words, all of the investment options
produced a positive real return over the long horizon examined.

Figure 5.6: Cumulative value of $1 investment in alternative asset classes (1926-2019)

The extremely large cumulative return for the stock index raises a fundamental question:
why are investors willing to hold Treasury and corporate bonds, or even Treasury bills, when
the long-run cumulative return to stock appears so much higher? A second aspect of Figure
5.6 perhaps sheds light on the answer. The is a marked difference in the ‘smoothness’ of the
cumulative value plots across the alternative portfolios. The curve for Treasury bills is the
smoothest and nearly always slopes upward. Intuitively, this reflects the fact that wealth is nearly
always increasing for the Treasury bill investment, and that the corresponding returns are not
very volatile. The curves for Treasury and investment-grade corporate bonds are ‘choppier’
and exhibit periods in which wealth is decreasing, i.e., years with negative returns. Finally, the
cumulative value plot for the S&P 500 stock index investment is by far the choppiest. Returns
are particularly volatile during the Great Depression period. As an illustration, suppose that an
investor’s horizon (to retirement, perhaps) was, until 1932. This investor would be the worst off
with the S&P 500 strategy, which would have yield roughly a 30% cumulative loss between 1926
and 1932. In contrast, an investment in Treasury or corporate bonds over this period would not
have produced losses.

5.5.2 Return Summary Statistics

The cumulative returns in Figure 5.6 provide some useful information about the properties
of stock, Treasury bond, and corporate bond returns in US markets over a long history. Let’s
dive in further and compute some key summary statistics using the annual return data for these
asset classes.
5.5 Measuring Characteristics of Risky Asset Returns – 135 –

The first set of statistics we consider are measures of the central location of the asset
returns. The first statistics of interest is the mean, otherwise known as the ‘sample mean’, or the
‘arithmetic mean’. The sample mean equals:
X
T
R̄i = µ̂i = (1/T ) Ri,t (5.9)
t=1
The sample mean R̄i provides a data-based estimate of the expected return µi . We therefore
alternatively denote the sample mean return as µ̂i . It is important to understand the distinction
and relation between these two objects. The expected return µi is a population quantity – it
equals the true average value of the asset return. The arithmetic mean, or ‘sample mean’ denotes
the average of a particular realized sample of returns. In general µ̂i ̸= µi ; however, as the number
of observations or the ‘sample size’ becomes large, then the sample mean takes on values closer
and closer to the true expected return µi .
An alternative geometric mean is based on the per-period cumulative return:

R̄iGeom = [(1 + Ri,1 )(1 + Ri,2 ) . . . (1 + Ri,T )]1/T − 1 (5.10)

The geometric mean is closely related to the cumulative return and holding period return formulas
we have encountered previously. In fact, note that the geometric mean is simply equal to the
cumulative return stated on a per-period basis. For annual returns, for example, the geometric
mean is equal to the T -period HPR expressed on an annualized basis. The geometric mean is
less than or equal to the arithmetic mean, where equality holds when the return is constant, i.e.,
the security is risk-free.7 The deviation between the arithmetic and geometric means is larger
the larger is the standard deviation of the asset returns.
Finally, we can obtain an estimate of the risk premium for a risky asset in two equivalent
ways: 1) as the difference between the sample mean for the risky asset and the risk-free return;
and 2) as the sample average of the excess returns for the risky asset:

Risk Premium Estimate = µ̂i − Rf (5.11)

where µ̂i denotes the average return on the asset.

Table 5.2 shows the arithmetic mean, geometric mean, and estimated risk premium for
Treasury bonds, corporate bonds, and the S&P 500 (stock) returns. The computations implicitly
use the average Treasury bill return over the 1926–2019 period (also shown) as the risk-free rate
in the computation of the risk premium estimate.
The average inflation rate is just under 3%. This reflects the influence of the high inflation
period of the late 1970s and 80s around other periods during which inflation has been quite low.
The arithmetic mean for Treasury bill returns is around 3.4% while the geometric mean is only
slightly lower. This reflects the low variability of Treasury bill returns, consistent with the idea

7When the return takes on the constant value R, then the T -period cumulative return equals (1 + R)T and hence the
geometric mean becomes simply 1 + R − 1 = R. Of course, this is also equal to the arithmetic mean when the return
is constant.
5.5 Measuring Characteristics of Risky Asset Returns – 136 –

Table 5.2: Means and Risk Premium Estimates for Alternative Asset Classes: 1926–2019 Annual
Inflation Treasury Bills LT Treas Bonds Corp. Bonds S&P 500
Mean 2.95% 3.40% 5.96% 6.40% 11.95%
Geo Mean 2.87% 3.35% 5.53% 6.08% 10.04%
Risk Premium - - 2.57% 3.00% 8.55%

that they serve as a proxy for a risk-free return. The arithmetic average return for Treasury bonds
is around 6% while the average corporate bond return is slightly higher at 6.4%. The higher
average return for corporate bonds in part reflects compensation for some default risk relative
to Treasury bonds (even though the bonds here are investment grade). Note that the difference
between the arithmetic and geometric means is more pronounced for the long-term bonds, as the
underlying returns exhibit more variation than Treasury bill returns (see also the table below with
standard deviation estimates). The arithmetic average S&P 500 return as nearly 12%, and the
geometric mean is significantly lower at around 10%. Finally, the estimates of the risk premiums
associated with Treasury bonds, corporate bonds, and a broad stock index (the S&P 500) are
around 2.6%, 3.0%, and 8.6%, respectively over a long history for US markets.
We next turn to statistics that are informative regarding the key parameter σi , the standard
deviation of returns, and concerning the nature of the distribution of shocks ϵi,t . The sample
variance of returns is computed as:

1 X
T
σ̂i2 = (Ri,t − R̄i )2 (5.12)
T −1
t=1

the ‘hat’ in the notation σ̂i2 reminds us that this quantity is an estimate of the true, unknown
variance of returns, which is equal to σi2 . Again, in general σ̂i2 ̸= σi2 , but as the number of returns
sampled grows large, then the estimate becomes closer and closer to the true return variance.
The sample variance can be viewed as the average of the squared deviations of the returns from
the sample mean. Note; however, that the sum of squared deviations from the mean is divided
by T − 1 rather than T . This reflects a so-called ‘degree of freedom’ adjustment. Put loosely,
we have used the data once in order to obtain the sample mean. In relatively large samples, i.e.,
for large T , the distinction hardly makes a difference. But it can be important for small sample
sizes.
The sample standard deviation equals the square root of the sample variance:
q
σ̂i = σ̂i2 (5.13)

As with the population analogs, the sample standard deviation is in the same units as the
underlying data, so for annual returns the standard deviation is in annual percentage units. This is
convenient for interpretation and preferable to the sample variance, which is difficult to interpret.
Given an estimate of the risk premium and the sample standard deviation, we can compute
a natural estimate of the Sharpe ratio or ‘reward-to-risk’ ratio by ‘plugging-in’ estimates of the
numerator and denominator of the Sharpe ratio formula:
5.5 Measuring Characteristics of Risky Asset Returns – 137 –

c i = µ̂i − Rf .
SR (5.14)
σ̂i
The ‘hat’ again emphasizes that this is only an estimate based on a sample of data of the
true population Sharpe ratio. As with the previous estimates, the Sharpe ratio estimate becomes
increasingly accurate as the sample size becomes large.
The following table summarizes standard deviation, Sharpe ratio, and a few additional
statistics for the three asset classes:
Table 5.3: Standard Deviation, Sharpe Ratios, and Other Statistics for Alternative Asset Classes
LT Treasury Bonds Corporate Bonds S&P 500
Standard Deviation 9.81% 8.46% 19.78%
Sharpe ratio 0.26 0.36 0.43
Sample Skew 0.95 1.18 -0.45
Minimum -14.90% -8.09% -45.53%
Maximum 40.35% 42.56% 53.25%

Despite the relatively high standard deviation of nearly 20% per year, the S&P 500 achieved a
higher annual Sharpe ratio over the 1926–2019 period than Treasury bonds or corporate bonds.
The sample skewness statistics in Table 5.3 indicate that the annual stock index returns are slightly
negatively skewed, while bond returns skew positively.8 The positive skew in bond returns is
also indicated by the fact that the maximum absolute return for both bond classes (around 40%)
is much greater than the minimum annual return.

8We will not cover the explicit formula for the sample skewness in this class. It is enough to understand that a positive
sample skewness indicates some degree of (estimated) positive skewness, whereas a negative skew statistic indicates
negative skewness. It should be remembered that these are statistics subject to variation over different samples. In
other words, even if a variable has no skew in truth, the estimated skewness statistic is likely to differ from zero.
5.5 Measuring Characteristics of Risky Asset Returns – 138 –

K Chapter 5 Problems k
1. On November 4, 2019, Moody’s Investors Service assigned a Baa2 rating to $300 million
of non-cumulative perpetual preferred stock to be issued by The Allstate Corporation
(Allstate; NYSE: ALL).
(a). Why do ratings agencies such as Moody’s provide ratings for preferred stock?
(b). What does ‘non-cumulative’ mean with respect to Allstate’s preferred stock?
(c). How would investors view the non-cumulative provision relative to an otherwise
comparable alternative preferred stock that is cumulative? How would this be factored
into the price and corresponding expected return?
2. Go to this site and read about the preferred stock issued by Fortive Corporation: https:
//www.preferredstockchannel.com/symbol/ftv.pra/. Answer the following:
(a). What does the phrase ‘liquidation preference shares’ mean?
(b). How often per year does the preferred stock pay a dividend?
(c). What is the annual dividend amount per share in dollars?
(d). Are the preferred shares callable? As of what date?
(e). Are the preferred shares convertible? What is the nature of the convertibility provi-
sion?
3. A dividend initiation event is defined as the first time that a publicly-listed company
pays a dividend to its shareholders. Briefly discuss some of the key considerations for a
firm debating whether to initiate a dividend. What kind of characteristics do you think
companies that initiate dividends typically possess?
4. The following table shows daily closing prices for Tesla (TSLA) over a week in November,
2020 (five tradings days). The stock did not pay a dividend over this time period. Use
these data to construct daily returns for five trading days: Nov, 6, Nov. 9, Nov. 10, Novl.
11, and Nov. 12. (Assume that stock is initially purchased at the close on Nov. 5.) Then
compute the five-day holding period return (HPR). Finally, compute an annualized version
of the 5-day HPR. Then answer the following: is the annualized 5-day HPR indicative
of the return that an investor would earn by holding Tesla stock for a full year? Briefly
explain your response.
Date Close Price Volume
11/12/2020 $ 411.76 19855100
11/11/2020 $ 417.13 17357700
11/10/2020 $ 410.36 30284200
11/9/2020 $ 421.26 34833000
11/6/2020 $ 429.95 21706000
11/5/2020 $ 438.09 28414500
5. The following table gives returns in four possible states of the world for two assets, a bond
index and a stock index:
5.5 Measuring Characteristics of Risky Asset Returns – 139 –

State Probability Stock Index Return Bond Index Return

Depression 0.1 -25% -5%

Recession 0.15 -10% 10%

Normal 0.5 10% 5%

Boom 0.25 40% 4%

(a). Calculate the expected returns for the stock and bond in this setting.
(b). Calculate the variance and standard deviations for the stock and bond returns in this
setting.
(c). Suppose there is a third, risk-free asset in this setting, that pays an average return of
2% in each state. Compute the risk premium for both the stock and bond indices, as
well as the Sharpe ratios.
6. Return to the 3-state model described by Table 5.1 in the chapter.
(a). Compute the risk premium on Stock A and Stock B.
(b). Compute Sharpe-ratios for the two risky stocks.
(c). Show how the model can be written in the form of Eq. (5.5). (Hint: take the
deviations from the mean and scale (divide) them by the standard deviation to produce
standardized shocks.)
7. A stock has had returns of 22%, 14%, 38%, -4%, 16%, and -12% over the past six years.
Compute the arithmetic and geometric means for the stock over this period.
8. Considering the following quarterly prices and dividends for two share classes for Bank of
America Corporation for the period December 2015–March 2017:

Common Shares Preferred Shares

(BAC) (BAC-D)
Qtr. End Date Price Dividend Price Dividend
Mar-17 23.59 0.075 25.22 0.39
Dec-16 23.09 0.075 24.58 0.39
Sep-16 15.65 0.075 25.74 0.39
Jun-16 14.49 0.05 25.81 0.39
Mar-16 13.52 0.05 25.4 0.39
Dec-15 16.83 0.05 25.49 0.39

(a). Calculate the quarterly holding period returns for each share class from January 2016
through March 2017. Hint: Assume the beginning of quarter price is the price from
the last quarter (so the beginning share price for the quarter ending March 2016
is from December 2015). Note: Do these and the remaining calculations in this
problem explicitly, i.e., not by using Excel formulas.
(b). Calculate the arithmetic average return for both share classes.
5.5 Measuring Characteristics of Risky Asset Returns – 140 –

(c). Calculate the average quarterly holding period return (geometric average) for both
share classes.
(d). Calculate the total (cumulative) 5-quarter holding period return for both share classes.
Which share class offered the better return?
(e). Calculate the variance and standard deviation of the quarterly holding period return
for each share class. If the quarterly risk-free return over this period was 0.05% (that
is five basis points, not 5%!), what is the Sharpe Ratio (reward-to-risk ratio) for each
share class. Which one has the higher Sharpe Ratio?
9. Suppose that Durango Manufacturing Inc.’s monthly stock return follows the equation:

Rt = 1% + 0.012ϵt ,

where ϵt follows the standard normal distribution.

(a). Determine the monthly and annualized average return for Durango stock.
(b). What is the monthly standard deviation of Durango stock? What does this equal
on an annualized basis? HINT: Variances scale linearly with the time interval, so
standard deviations scale with the square root of the time interval.
(c). Provide an interval within which Durango’s monthly stock returns will fall with
approximately 95% probability.
(d). Assume the monthly risk-free rate equals 25 basis points. What is the monthly risk
premium for Durango stock?
(e). Continuing to assume a 25 basis point risk-free rate, suppose that the realization of
ϵt this month equals 0.5. Compute the corresponding return and excess return this
month for Durango.
10. If company ABC’s stock return has an annual mean of of 13% and a standard deviation
of 15%. Provide a 99% confidence interval for the stock return assuming that the return
is normally distributed. If, instead, the stock return distribution is fat-tailed, what can be
said about the likelihood of obtaining values outside of this range?
11. Based on a sample of 5 years of data, the arithmetic average monthly return for Genetic
Technologies Inc. stock equals 1.2% and the sample standard deviation of monthly returns
equals 3.2%. Similarly, the arithmetic average monthly return for Big Construction Co.
stock equals 0.9% and the standard deviation equals 1.2%. Assume the monthly risk-free
rate over this period was 0.20%.
(a). Suppose we were to compute the geometric mean of the monthly returns for both
stocks. How would these compare to the arithmetic means? For which stock would
the difference likely be greater?
(b). Compute estimates of Sharpe ratios for Genetic Technologies and Big Construction
using the provided information.
(c). Do these Sharpe ratios clearly indicate which is the better investment between Genetic
and Big Construction? Why or why not?
5.5 Measuring Characteristics of Risky Asset Returns – 141 –

12. Excel problem Download the data of annual returns for Treasury bills, Treasury bonds,
corporate bonds, and the S&P 500 index over the period 1926–2019. (The data are located
in the ‘Spreadsheets’ folder on Canvas.) Build a spreadsheet that replicates as closely as
possible Figure 5.6 showing the cumulative value of a $1 investment in each asset class
starting at the beginning of 1926, as well as Tables 5.2 and 5.3.
13. Excel problem Download monthly data on risk-free returns that posted on Canvas. The
data are located in the ‘Spreadsheets’ folder on Canvas. Then go to Yahoo! Finance and
look up Nike stock (ticker symbol NKE). See the following address: https://finance.
yahoo.com/quote/NKE. Go under the Historical Data tab and download monthly data
for NKE from December 2010 through October 2020. Use the ‘Adjusted close’ (which
incorporates dividends already and adjusts for any stock splits) to compute monthly returns
from January 2011 – October 2020. Merge these data with the monthly risk-free rate data.
Note: I will post a short Canvas video that illustrates how to get the NKE stock data.
Create a new column that computes excess returns for NKE in each month from January
2011 – October 2020. Then do the following:
(a). Provide a figure that shows the cumulative return for $1 invested in NKE at the
beginning of January 2011 through the end of October 2020. (You will need to
create a column tracking the cumulative return before creating the figure.)
(b). Compute and display in a prominent and nicely formatted way the following statistics
for NKE monthly stock returns:
I. Arithmetic mean monthly return
II. Geometric mean monthly return
III. The standard deviation of monthly returns
IV. The skewness of monthly returns
V. The minimum monthly return
VI. The maximum monthly return
VII. Estimated risk premium
VIII. Estimated Sharpe ratio
 Chapter 6 Portfolios, Diversification, and the CAPM

The previous chapter considers the properties of risky asset returns, including the average
return, standard deviation, risk premium, and Sharpe ratio. Should an investor then search across
stocks or other securities for the particular security with the greatest Sharpe ratio, and invest
their wealth in this security? Although this might seem a reasonable strategy, one of the main
takeaways of the current chapter will be that this is not a wise investment strategy, because it
ignores the benefits of diversification across a range of risky assets. We will develop a set of
results that illustrates the wisdom of the classic adage: don’t put all of your eggs in one basket!.
The second key result developed in this chapter is among the most important in finance. This
result is an implication of a classic model known as the capital asset pricing model, or ‘CAPM,’
developed in the 1960s. The result states that expected returns for risky assets should be a linear
function not of the standard deviation of returns, but rather of their beta, which captures how
the returns ‘co-vary’ with aggregate or market returns. Consequently, the theory holds that this
beta quantity, and not the standard deviation of returns, is the correct measure of risk that is
‘priced’ in financial markets, meaning that investors receive a compensation in the form of a risk
premium for bearing this risk.

6.1 Portfolios

We begin by describing investment portfolios. A portfolio is simply an allocation of wealth

across a set of available assets or securities:
Definition 6.1. Portfolio
A portfolio is a set of positions in financial securities. The securities involved can include
stocks, government bonds, corporate bonds, real estate, and other assets.
♣

A portfolio can be efficiently described via the weights on various assets. Mathematically,
given an assumed set of N possible securities (in principle N can be very large), a portfolio can
be described as a set of N numbers, or weights, wn , n = 1, . . . , N , such that:
X
N
wn = 1 (6.1)
n=1
Eq. (6.1) simply says that the weights in the various assets must sum to one, in other words, the
invested money must all go somewhere. Note that Eq. (6.1) does not preclude the possibility
of negative weights, such as wi = −025 (or −25%) for some asset, nor of weights greater than
one, such as wj = 1.25 (or 125%) for some other asset. In fact, a two-asset portfolio with
w1 = −0.25 and w2 = 1.25 is a perfectly valid portfolio according to Eq. (6.1).
What does it mean to have a negative portfolio weight for an asset? Such a position is
6.1 Portfolios – 143 –

consistent with the practice of short selling the asset. Short selling involves borrowing shares
of a stock or other asset today. The investor then sells these borrowed shares to buyers at the
current market price. At a specified future date, the shares must be returned to the lender, which
is often a broker or related financial institution. The ‘short position’ in the asset (negative weight)
will be profitable (earn a positive return) if the price of the asset falls during the holding period.
Consequently, one motivation for short-selling securities is to speculate that the stock or asset is
overvalued and that the price will therefore fall in the future. In practice, there are limitations to
the extent of short positions that brokers will permit traders to take. We will not discuss further
these details in this course.
An asset wn > 1 is said to be a leveraged position, because it is often obtained by short-
selling a risk-free asset, which is tantamount to borrowing funds at the risk-free rate, and using
the borrowed funds to invest more than one’s starting wealth in a particular risky asset. Two
other notable types of portfolios are equal-weighted portfolios and value-weighted portfolios:
Definition 6.2. Equal-weighted portfolio
An equal-weighted portfolio is a portfolio of N assets with weights equal to 1/N for
wn = 1, . . . , N .
♣

Definition 6.3. Value-weighted portfolio

A value-weighted portfolio is a portfolio that applies weights to each security in proportion
to the securities market capitalization value. For stocks, the market capitalization value
equals the product of the number of shares outstanding times the current share price.
Letting M Cn denote the market capitalization in dollars for stock n, the weights in a
value-weighted portfolio equal ∑NM Cn for n = 1, . . . , N .
n=1 M Cn
♣

Given a set of portfolio weights, the return on the corresponding portfolio equals the
weighted combination of the returns on the individual securities held in the portfolio, where the
weights are equal to the portfolio weights. Mathematically, we have:
X
N
Rp = wn Rn , (6.2)
n=1
where Rp denotes the periodic return on a portfolio with weights w1 , . . . , wN , and Rn denotes
the periodic return on the n-th security for n = 1, . . . , N .

Example 6.1 The following table shows total returns for Best Buy (BBY) and Disney (DIS) stock
for the month of October, 2020. Suppose that you formed a portfolio at the beginning of October,
2020 consisting of these two stocks with a weight of w on BBY and a weight of (1 − w) on DIS
(since the portfolio weights must sum to one). Compute the return for this portfolio under the
following alternative scenarios concerning the portfolio weights:
1. An equal-weighted portfolio
2. A value-weighted portfolio
6.2 One risky asset and a risk-free asset – 144 –

3. A portfolio with -20% allocated to BBY

The market capitalization data in the table are in billions of USD and as of the beginning of
October, 2020.

Stock Ticker Return (Oct. 2020) Market Cap ($B, 9/30)

Best Buy BBY 6.80% $ 28.87

Disney DIS 16.35% $ 219.46

Solution: For the equal-weighted case, we have w = (1 − w) = 0.5. Applying Eq. (6.2) to
compute the portfolio return gives:
X
N
Rp = wn Rn
n=1

Rp = wR1 + (1 − w)R2
= 0.5(6.80%) + 0.5(16.35%)
= 11.58%

For the value-weighted case, we first need to obtain the appropriate weights. To do so, we
sum the market capitalization across the two firms. This equals 28.87 + 219.46 = 248.33 billion
USD. Next, we compute the value-weighted allocation to BBY, which is the ratio of BBY’s
market cap to the total: w = 28.87/248.33 ≈ 11.626%. Finally, the weight on DIS is 1 minus
this, or 1 − 0.11626 = 0.88374. Then, we compute the portfolio return as:

Rp = wR1 + (1 − w)R2
= 0.11626(6.80%) + 0.88374(16.35%)
= 15.24%

Finally, for the case involving a 25% short position in BBY, we have:

Rp = wR1 + (1 − w)R2
= −0.2(6.80%) + 1.2(16.35%)
= 28.25%

Since DIS had the greater return during this month, the final portfolio has the highest return
among the three.

6.2 One risky asset and a risk-free asset

The first portfolio problem we will consider is quite limited, in the sense that we will only
consider two investment opportunities: a single risky asset, with return R, and a risk-free asset
that pays a certain return Rf . The weight in the risky asset will be given by w, with the weight in
the risk-free asset then equal to 1 − w (because portfolio weights must sum to one). The return
to this simple portfolio equals:
6.2 One risky asset and a risk-free asset – 145 –

Rp = wR + (1 − w)Rf .

Next, we want to understand the key characteristics of the portfolio return: its mean and
standard deviation, which in turn allow us to determine the risk premium and Sharpe ratio. The
mean portfolio return is simply the weighted average of the mean return for the risky asset and
the risk-free return:
E(Rp ) = wE(R) + (1 − w)Rf . (6.3)

The variance of the portfolio return is given by the following simple expression:

σ 2 (Rp ) = w2 σ 2 (R),

and consequently the standard deviation of the portfolio equals:

σ(Rp ) = w σ(R).

Intuitively, the risk-free asset has a variance of zero by definition. Therefore, the standard
deviation of the portfolio return is determined by the weight invested in the risky asset. Finally,
the risk premium and Sharpe ratio associated with the portfolio are given by:

Risk premium(Rp ) = E(Rp ) − Rf = w(E(R) − Rf ) (6.4)

and
Risk premium(Rp ) E(R) − Rf
Sharpe ratio(Rp ) = = (6.5)
σ(Rp ) σ(R)
The top equation says that the risk premium of the portfolio equals the weight in the risky
asset times the risky asset’s risk premium. The bottom equation says that the Sharpe ratio of
the portfolio combining the risky asset with the risk-free asset is exactly equal to the Sharpe
ratio of the risky asset. Note in particular how the proportion invested in the risky asset w
does not appear in the Sharpe ratio expression, due to the fact that w cancels in the numerator
and denominator when plugging in the previously derived expressions for the risk premium and
standard deviation. This implies that whether an investor invests 10%, 50%, 100%, or 150% in
the risky asset, she still achieves the exact same Sharpe ratio. In other words, it is not possible
to increase (or decrease) the Sharpe ratio metric by changing the weight allocated to the risky
asset. This is because when the weight is increased, the risk premium and standard deviation of
the portfolio return increase proportionally.
It is useful to visually the above results in the form of a graph that plots the expected return
of a combination of the risky asset with the risk-free asset (as a function of the weight w in the
risky asset), against the corresponding standard deviation. Figure 6.1 shows such a plot. An
allocation of 100% to the risk-free asset corresponds to the y-intercept in the figure, with an
expected return equal to the risk-free rate and a standard deviation of zero. Starting from this
point, as the allocation to the risky asset increases, both the expected return of the portfolio and
the standard deviation increase proportionally, moving in a northeasterly direction along the red
6.2 One risky asset and a risk-free asset – 146 –

line in the figure. The slope of the red line capturing these possible expected return-standard
deviation combinations is equal to the Sharpe ratio of the risky asset. (The last point can be
easily verified using the ‘rise-over-run’ formula for the slope of a line.) The line depicted in
Figure 6.1 is often called the capital allocation line (picking up the idea of how risk and return
change as you allocate more capital to the risky alternative).

Figure 6.1: Portfolios of one risky asset and a risk-free asset

Example 6.2 Suppose Pfizer common stock returns are 15% on average per year with a standard
deviation equal to 22%. A risk-free asset offers a 7% annual return. Compute the risk premium
and Sharpe ratio associated with Pfizer stock. Then complete the table below regarding the
attributes of portfolios consisting of Pfizer and the risk-free asset.

Alloc. to Pfizer (w) E(Rp ) Risk premium σ(Rp ) Sharpe ratio

0% 7% 0% 0% NA
50%
75%
100% 15% 8% 22% 0.364
150%

Solution: The risk premium for Pfizer stock is equal to the expected return of Pfizer stock less
the risk-free rate, or 15% − 7% = 8%. The Sharpe ratio for Pfizer stock is:
Risk premium
SR =
σ
8%
= = 0.364
22%

Now consider a portfolio that is 50% Pfizer and 50% risk-free asset. The expected return for
the portfolio equals 50%×15%+50%×7% = 11%. The risk premium for the portfolio is then the
6.3 Portfolios of two risky assets – 147 –

previously computed expected return for the portfolio minus the risk-free rate: 11% − 7% = 4%.
The standard deviation for the portfolio equals the weight in the risky asset times the risky asset’s
standard deviation, or 50% × 22% = 11%. Finally, the Sharpe ratio equals the ratio of the risk
premium for the portfolio to the portfolio’s standard deviation: 4%/11% = 0.364. The other
portfolio weights are handled similarly, leading to the following completed table:

Alloc. to Pfizer (w) E(Rp ) Risk premium σ(Rp ) Sharpe ratio

0% 7% 0% 0% NA
50% 11% 4% 11% 0.364
75% 13% 6% 16.5% 0.364
100% 15% 8% 22% 0.364
150% 19% 12% 33% 0.364

6.3 Portfolios of two risky assets

We next move on to consider portfolios that allocate wealth to two risky assets. Label the
returns to these two risky assets as R1 and R2 respectively. The expected returns for the two
risky assets are µ1 and µ2 , respectively, and we will assume that µ1 > µ2 (otherwise we can just
switch the assets ordering). Let σ1 and σ2 denote the standard deviations of the returns for the
first and second risky assets. Finally, let w denote the portfolio weight allocated to the first risky
asset, and therefore 1 − w is the weight on the second asset. The portfolio return is then:

Rp = wR1 + (1 − w)R2 .

The expected return for the portfolio of the two risky assets is simply a weighted combination
of the expected returns on the two underlying assets:

E(Rp ) = wµ1 + (1 − w)µ2 .

The variance of the portfolio return, however, is more complicated. This is because the
returns on the two risky assets are likely to be correlated. Recall from probability and statistics
that the covariance between two random variables measures the extent to which the two random
variables move or ‘co-vary’ together. A positive covariance between two random variables means
that when one random variable experiences a positive shock, the other is also likely to experience
a positive shock, and vice versa. In contrast, a negative covariance means that the two variables
tend to experience oppositely-signed shocks, such that if the first experiences a positive shock
the second is more likely to experience a negative shock, and vice versa. Although the sign
of the covariance is informative regarding the type of relation between two random variables,
its magnitude is hard to interpret. The correlation is defined as the covariance dividend by the
products of the standard deviation of the two random variables. Taking our case of two risky
returns as a concrete example, we have:
6.3 Portfolios of two risky assets – 148 –

σ1,2
ρ1,2 = ,
σ1 σ2
where ρ1,2 denotes the correlation between the two risky asset returns, and σ1,2 denotes the
covariance.
It might seem convenient to ignore the possibility of return correlations, i.e., to simply
assume that returns on different stocks are uncorrelated. This makes the computation of portfolio
variances relatively straightforward. In statistics, we often make an assumption that observations
are ‘independently distributed,’ which implies that the observations are uncorrelated. Unfortu-
nately, it is simply not realistic to assume that returns on different stocks will be uncorrelated.
This is because stock prices respond to news about macroeconomic conditions as well as about
firm-specific developments. When important macroeconomic news hits the market, the prices of
many stocks tend to move up or down together, implying that the returns are correlated. A good
example occurred in late February and early March of 2020, when it became gradually clear that
the novel coronavirus would transmit globally and a pandemic was likely to occur. The stock
market dropped substantially in value and the vast majority of stocks lost value at the same time.
The formula for the variance of the two-risky asset portfolio is as follows:1

σ 2 (Rp ) = w2 σ12 + (1 − w)2 σ22 + 2w(1 − w)σ1 σ2 ρ1,2 . (6.6)

The first two terms on the right-hand side of Eq. (6.6) equal the variance of the portfolio return
in the special case in which the returns are uncorrelated (ρ1,2 = 0). The final term captures
the effect of return correlation. When the correlation is positive (ρ1,2 > 0) then the variance of
the portfolio return exceeds the variance under the benchmark uncorrelated case. The intuition
is that, when the returns are positively correlated, the stocks experience positive and negative
‘swings’ together, thereby magnifying the variation of the portfolio value. In contrast, when the
correlation is negative, the returns tend to offset in sign, and this causes the portfolio variance to
be less than the variance under the benchmark uncorrelated case.

Example 6.3 Suppose that Stock A has an average annual return of 12% and standard deviation
of 15%. Stock B has an average annual return of 8% and standard deviation of 7%. Compute
the mean portfolio return as well as the variance and standard deviation of returns of an equal-
weighted portfolio of these stocks assuming:
1. The correlation between the stock returns is 0.5
2. The correlation between the stock returns is 0.2

1An alternative, equivalent way to write this formula is:

σ 2 (Rp ) = w2 σ12 + (1 − w)2 σ22 + 2w(1 − w) σ1,2 ,

because σ1,2 = σ1 σ2 ρ1,2 .

6.3 Portfolios of two risky assets – 149 –

Solution: The mean portfolio return does not depend on the correlation and equals:

E(Rp ) = wµ1 + (1 − w)µ2

= 0.5(12%) + 0.5(8%)
= 10%

To determine the variance of the portfolio, we apply Eq. (6.6) as follows (when ρ1,2 = 0.5):

σ 2 (Rp ) = w2 σ12 + (1 − w)2 σ22 + 2w(1 − w)σ1 σ2 ρ1,2

= 0.52 0.152 + 0.52 0.072 + 2(0.5)(0.5)(0.15)(0.07)(0.5)
= 0.009475
√
The standard deviation is σ(Rp ) = 0.009475 = 9.734%.
The only thing that changes for the case of ρ1,2 = 0.2 is that particular input to the variance
equation, so we now have:

σ 2 (Rp ) = 0.52 0.152 + 0.52 0.072 + 2(0.5)(0.5)(0.15)(0.07)(0.2) = 0.0079

√
Finally, the standard deviation equals σ(Rp ) = 0.0079 = 8.888%.
The previous computations show how to obtain the mean and standard deviation of the
portfolio return given particular portfolio weights on the two risky assets. Imagine now that we
did these computations over and over again, each time for a different allocation to the risky asset,
e.g.. 1%, 2%, 3,%, and so on. What would the characteristics of the portfolios we generate
in this way look like? To illustrate this, consider stock returns for Disney (DIS) and Best Buy
(BBY), with the following assumed attributes:

Attribute Disney (DIS) Best Buy (BBY)

Mean (Annual %) 12.127% 30.277%

St. Dev. (Annual %) 24.891% 51.669%
DIS/BBY Correlation 0.38 -

The means and standard deviations for DIS and BBY given in the table are based on estimates
from actual historical monthly returns for these two stocks over the period 1996–2019. Similarly,
the sample correlation of the monthly stock returns over this period equals 0.38. Of course, in
reality these are only estimates of the true, unknown means, etc. However, for the purposes of
this example, we will pretend as though these are the true characteristics.
Figure 6.2 shows a plot of the expected return (y-axis) and standard deviation (x-axis) for
portfolios formed using the stocks DIS and BBY. Each point corresponds to a particular allocation
to DIS (and therefore BBY as well). These allocations range from -30% DIS and 130% BBY
(top-right-most point) to 130% DIS and -30% BBY. Note, therefore that in some cases we are
allowing a short position in one asset and a leveraged position in the other.
Altering the portfolio weights allocated to DIS and BBY traces out a curve in the mean -
standard deviation space. Note that, because these are two (imperfectly correlated) risky assets,
6.3 Portfolios of two risky assets – 150 –

Figure 6.2: Portfolios of two risky assets (DIS and BBY)

all of the portfolios entail risk. (The points are associated with positive portfolio return standard
deviations in the graph.) The left-most point on this curve represents the minimum variance
portfolio. As the name suggests, this particular portfolio represents the lowest possible variance
and standard deviation that can be achieved by combining these two assets in a portfolio. The
minimum variance portfolio in this case is close to, but not exactly, 100% DIS and 0% BBY.
The points below and to the right of the minimum variance portfolio are colored red in
Figure 6.2. This highlights the fact that all portfolios in this region of the curve are dominated
by points on the green portion of the curve directly above the red-colored points. Specifically,
starting from a point on the red portion of the curve and going straight upward, one can identify
an alternative portfolio that has the same standard deviation (risk) and a higher mean. For
this reason, the red-colored portion is called the inefficient portion of the frontier. Note that,
on the green portion of the curve, there is not an obviously ‘dominating’ portfolio. Starting at
an arbitrary point on the green portion, it is possible to increase the mean return by moving
upward/right-ward, but doing so also increases the standard deviation associated with the return.
This portion of the curve is called the efficient frontier. Investors forming portfolios of the two
risky assets should choose portfolios on this portion of the curve.
The inward-bowed shape of the curve in Fig. 6.2 captures the benefits of diversification
across multiple risky assets. The fact the the absolute value of the correlation between stocks is
less than or equal to one implies:

wσ1 + (1 − w)σ2 ≥ σ(Rp ), (6.7)

where σ(Rp ) denotes the standard deviation of a portfolio of stocks 1 and 2, which depends
implicitly on the weight w. In the space of Fig. 6.2, wσ1 + (1 − w)σ2 describes points on a
line that connects the means and standard deviations associated with assets 1 and 2, i.e., a line
connecting the points corresponding to a 100% allocation to DIS and a 100% allocation to BBY.
6.3 Portfolios of two risky assets – 151 –

As long as returns to DIS and BBY are not perfectly correlated; however, the standard deviation
of a portfolio of the two with 0 < w < 1 is less than that implied by the line – visually this
refers to the fact that the plot curves inward toward the y-axis. The imperfect correlation implies
that the same mean return can be achieved with a lower standard deviation relative to the ‘no
diversification’ benchmark case when the asset returns correlated perfectly.
Figure 6.3 helps clarify the preceding points by showing what happens in the hypothetical
scenario in which DIS and BBY stocks returns are perfectly correlated. (The assumed means
and standard deviations of the returns remain the same.) It is important to understand that this
example is purely hypothetical. In practice, different risky assets will not be perfectly positively
or negatively correlated. The cases ρ = 1 and ρ = −1 are therefore unrealistic, but show what
happens in the limit as the correlation between returns becomes more and more extreme. When
ρ = 1, the ‘curve’ becomes a line connecting the two risky asset points, as discussed above.
Portfolios formed with 0 < w < 1 represent points on this line and there is no diversification
benefit. Intuitively, this is because there exists a perfect linear relation between shocks to the two
risky stocks, so in effect their returns are driven by the same underlying shock. When the returns
are perfectly negatively correlated (ρ = −1), then there exists a particular combination, with
weight denoted w∗ so that the resulting portfolio is risk-free (has a standard deviation of zero).
Since a risk-free asset exists, we can think of the portfolio possibilities as similar to those that
obtain when one combines a risky asset with a risk-free asset: we move along a line connecting
the risk-free return to the risky asset characteristics (see previous section).

Figure 6.3: Two risky assets: perfect positive or negative correlation

Figure 6.4 shows how the frontier changes under various other alternative assumptions
for the correlation between DIS and BBY, again with all other features (means and standard
deviations) held fixed. The original case with ρ = 0.38 is plotted as the green- and red-dotted
curve and is exactly the same as in the previous figure. Additional curves are plotted for
alternative assumptions about the correlation between the stock returns. Figure 6.4 shows that,
6.3 Portfolios of two risky assets – 152 –

as the correlation falls in value, the curve ‘bows inward’ more. This indicates a greater potential
benefit associated with diversification, because an investor can now achieve the same mean return
(on the y-axis) with a lower standard deviation (on the x-axis). This is particularly pronounced
for the case ρ = −0.5. We emphasize that the correlation values illustrated here are hypothetical
– in reality it is very unlikely that DIS and BBY returns would have a strong negative correlation,
as both businesses are exposed to common macroeconomic shocks.

Figure 6.4: Two risky assets: impact of correlation on curve

To continue our analysis, now suppose that, in addition to the two risky stocks, there is also a
risk-free asset paying a return of Rf . What portfolio properties can be obtained by investors given
the addition of the new risk-free opportunity? Recall from the previous section that any risky
asset, when combined with the risk-free asset, produces a line in the mean-standard deviation
plot space with a y-intercept at the risk-free return and intersecting the point that corresponds to
the mean and standard deviation of the risky asset return.
Figure 6.5 illustrates the points that may be obtained by combining the new risk-free asset
with DIS (black line) or BBY (blue line). As noted in the previous section, the slope of the line
corresponds to the Sharpe ratio associated with the combinations. From this perspective, if an
investor had to choose between combining BBY or DIS with the risk-free asset, combining BBY
is preferable, as this leads to a higher Sharpe ratio (greater slope of the line).
Importantly, however, the investor is not restricted from only combining DIS or BBY
individually with the risk-free asset. The investor can combine any portfolio with the risk-free
asset that corresponds to a point on the risky asset frontier. Among these, which is best? If we
define ‘best’ as generating the highest possible Sharpe ratio, then the optimal portfolio of the
two risky assets is characterized as the point of tangency between a line drawn from the risk-free
return that is just tangent to the risky asset frontier. Figure 6.6 illustrates, with the purple line
just tangent to the DIS and BBY frontier, and the purple square indicating the tangency portfolio.
For this example, the tangency portfolio is approximately (but not exactly) 50% DIS and 50%
6.4 More Than Two Risky Assets – 153 –

Figure 6.5: Combining DIS and BBY with a risk-free asset

BBY.
There are several additional key insights:
1. In this scenario, all investors optimally hold portfolios consisting of the risk-free asset and
the tangency portfolio. This result is called two fund separation, referring to the fact
that all investors hold two ‘funds’: 1) a risky asset fund corresponding to the tangency
portfolio; and 2) a ‘money market’ fund that corresponds to the risk-free asset
2. Different individual investors hold more or less of the risk-free asset the particular mix
depends on investor risk aversion
3. Every investor holds the risky assets (in this case DIS and BBY) in the same relative
proportion driven by tangency portfolio composition
4. To illustrate the previous point in a concrete example, suppose that the tangency portfolio
composition was exactly 50% DIS and 50% BBY. Assume an investor decides to allocate
20% of her wealth to the risk-free asset. Then the remaining 80% will be allocated to the
tangency portfolio, such that she would hold 40% DIS and 40% BBY.

6.4 More Than Two Risky Assets

In reality, there are a wide range of risky assets available to investors. Focusing only on
US stocks, for example, there are thousands of stocks listed on NYSE and Nasdaq. How do the
concepts developed in the previous sections extend to a setting with many assets? Fortunately, it
turns out that the key ideas extend almost directly. This section develops these results.
Let us now consider a setting with N risky assets, where N can be any positive integer. A
portfolio of these risky assets can be described as the set of weights wn , n = 1, . . . , N that sum
6.4 More Than Two Risky Assets – 154 –

Figure 6.6: The tangency portfolio for two risky assets

to one. The portfolio return is given by:

X
N
Rp = wn Rn . (6.8)
n=1
The average portfolio return, or expected portfolio return, is equal to a similar weighted combi-
nation of the individual assets expected returns:
X
N X
N
E(Rp ) = wn E(Rn) = wn µ n . (6.9)
n=1 n=1
The formula for the variance of the portfolio returns quickly becomes complicated as the
number of assets grows. The two asset case involves only two asset standard deviations and a
single pairwise correlation. A three asset case, however, involves three individual asset standard
deviations and three pairwise correlations: the correlation between stocks 1 and 2, the correlation
between stocks 2 and 3, and the correlation between stocks 1 and 3. The variance for the N asset
case involves N standard deviations and N (N − 1)/2 correlations. As the following example
illustrates, this quickly becomes large:

Example 6.4 How many different pairwise correlations exist for a set of 10 risky asset returns?
Solution: Applying the previous formula, there are 45 different pairwise correlations:
N (N − 1) 10(9)
= = 45. (6.10)
2 2
The formula for the portfolio variance can be written as:
X
N X
N
σ 2 (Rp ) = wi wj σi,j (6.11)
i=1 j=1
XX
= wi wj σi σj ρi,j (6.12)
i=1 j=1
Notice that for the N terms in the sum for which i = j in the above formula, we just get the
squared weight times the asset variance. The standard deviation of the portfolio return is, of
6.4 More Than Two Risky Assets – 155 –

course, the squared root of the variance:

q
σ(Rp ) = σ 2 (Rp ) (6.13)

The following gives an example of the explicit computation of a portfolio variance and standard
deviation for the three asset case.

Example 6.5 Suppose you allocate 20% to Disney, 30% to Best Buy, and 50% to Exxon-Mobil.
Suppose the following tables gives the means and standard deviations for these assets. In addition,
assume that the correlation between DIS and BBY returns equals 0.38, the correlation between
DIS and XOM equals 0.27, and the correlation between BBY and XOM equals 0.13. Compute
the expected return, variance, and standard deviation of the portfolio return.

Attribute Disney (DIS) Best Buy (BBY) Exxon-Mobil (XOM)

Mean (Annual %) 12.127% 30.277% 10.629%

St. Dev. (Annual %) 24.891% 51.669% 17.932%

Solution: The portfolio expected return is computed as the weighted average of the individual
stock returns:

E(Rp ) = w1 E(R1 ) + w2 E(R2 ) + w3 E(R3 )

= 0.2(12.127%) + 0.3(30.277%) + 0.5(10.629%)
= 16.82%

The portfolio variance is somewhat complicated because it involves three different correla-
tions:2

σ 2 (Rp ) = w12 σ 2 (R1 ) + w22 σ 2 (R1 ) + w32 σ 3 (R1 )

+ 2w1 w2 σ(R1 )σ(R2 )ρ1,2 + 2w1 w3 σ(R1 )σ(R3 )ρ1,3 + 2w2 w3 σ(R2 )σ(R3 )ρ2,3
= 0.22 (24.891%)2 + 0.32 (51.669%)2 + 0.52 (17.932%)2
+ 2(0.2)(0.3)(24.891%)(51.669%)(0.38)
+ 2(0.2)(0.5)(24.891%)(17.932%)(0.27)
+ 2(0.3)(0.5)(51.669%)(17.932%)(0.13)
= 0.046433
√
The standard deviation of the portfolio return is then σ(Rp ) = 0.046433 = 21.548%.
What happens to the minimum variance frontier when we consider more than two risky
assets? Figure 6.7 shows a plot of the frontier generated by four stocks: DIS, BBY, XOM, and
Boston Beer (SAM). Three of these four stocks have appeared in previous examples. The fourth
is Boston Beer, maker of Samuel Adams – hence the SAM ticker symbol – and owner of several

2It is much easier to represent portfolio variances using matrix multiplication. For those familiar with linear algebra
and matrix operations, the portfolio variance equals σ 2 (Rp ) = w′ Σw, where Σ denotes the N × N ‘covariance
matrix’ for the individual security returns.
6.4 More Than Two Risky Assets – 156 –

other beverage brands. The means, standard deviations, and pairwise correlations among for the
stock returns of the four companies are calibrated using historical data.

Figure 6.7: The minimum variance frontier for four risky assets

Figure 6.7 looks similar to the two asset case in many ways. The curve remains ‘bullet-
shaped’ and, as in the case with two risky assets, there is both a global minimum variance
portfolio (indicated in the figure) and an inefficient portion of the frontier which lies below and
to the right of the global minimum variance portfolio. The interpretation is the same: investors
should shun portfolios on the inefficient portion of the frontier in favor of portfolios on the
efficient portion, which lies above the global minimum variance portfolio.
There is one important difference between the case with two assets and cases with more
than two assets, such as the four asset case in Figure 6.7. In the two-asset case, a 100% allocation
to either stock represented points on the frontier. In contrast, when there are more than two risky
assets, 100% positions in each of the stocks will in general be points strictly inside the frontier.
Notice that each of the four 100% positions associated with the four stocks is a point inside
the frontier. This is an important result, as it clearly indicates the benefits of diversification.
No investor should allocate 100% to any of the four individual stocks. A better portfolio can
be obtained by moving upward to the frontier, where a portfolio exists with the same standard
deviation and a higher return.
When a risk-free asset is added to the set of available investments, the situation is very
similar to the case with two risky assets. Once again, any risky asset portfolio can be combined
with the risk-free asset to obtain points along the line that connects the risk-free return with the
point corresponding to the risky asset portfolio in the mean-standard deviation plot. The slope of
the line reflects the Sharpe ratio associated with all combinations of that risky asset portfolio and
the risk-free asset. Just as in the two risky asset case, the highest Sharpe ratio is obtained when
the risk-free asset is combined with the tangency portfolio, defined as the risky asset portfolio
that corresponds to the line from the risk-free return that is just tangent to the minimum variance
6.5 Systematic and Idiosyncratic Return Components – 157 –

Figure 6.8: The tangency portfolio for four risky assets

frontier. Figure 6.8 shows the tangency portfolio and the corresponding ‘capital allocation line’
(CAL) for the example with DIS, BBY, SAM, and XOM.
The same two fund separation result holds here as did for the two risky asset case. All
investors will hold a position on the CAL that is a combination (portfolio) of the tangency
portfolio of risky assets and the risk-free asset. The exact position on the CAL for each investor
will depend on his or her level of risk aversion. However, all investors will hold risky assets in
the same proportion as in the tangency portfolio. Finally, although the example developed above
consists of four assets, the general N asset case is essentially the same. In particular, the N
risky assets will generate a minimum variance frontier and, when a risk-free asset is available,
investors will choose to hold a combination of the tangency portfolio and the risk-free asset.

6.5 Systematic and Idiosyncratic Return Components

Recall our model of the return on a stock or other risky asset:

Ri,t = µi + σi ϵi,t . (6.14)

To briefly review, recall that the term µi is simply the expected return for the asset, i.e., µi =
E(Ri,t ). The term σi captures the standard deviation of the returns and ϵi,t is a standardized
‘news’ component (in the sense of having being mean zero with a standard deviation of one).
One limitation of the model is that all forms of ‘news’ that influence returns are lumped together
in the composite shock ϵi,t . But there are different kinds of news that influence stock returns.
In particular, some news pertains to the macroeconomy. Critically, this type of news tends to
influence many different risky stock returns simultaneously, although not necessarily in precisely
the same way. Another type of news might be described as ‘industry news’ – this news would
have a significant impact on a collection of firms operating in the same industry, but would
have minimal impact on the returns of firms outside of this industry. Finally, we can imagine
6.5 Systematic and Idiosyncratic Return Components – 158 –

‘firm-specific’ news: news that only materially impacts the returns of a single firm. A concrete
example of such news might be the outcome of a set of drug trials for a new drug intended to
treat a rare disease. The outcome of the trial will have a significant impact on the value of the
firm developing the potential treatment, but may not have a material impact on any other firms’
valuations, especially if no other firms are competing to develop a treatment for this particular
rare disease.
In this section, we adapt the basic stock return model to recognize at least two distinct forms
of news: systematic news and idiosyncratic or firm-specific news. First, we need to settle on a
way to measure systematic news. It might seem sensible to obtain a measure of macroeconomic
output, such as gross domestic product (GDP). Instead, however, we will use a financial measure
of the systematic component of stock returns: the return on the stock market, which we will
denote Rm . We can think of Rm as the return to a market-capitalization weighted index of a
broad collection of stocks. For example, the S&P 500 index or the Russell 1000 index might
serve as a reasonable proxy for the US stock market return. Applying our basic model of returns
to the market gives:
Rm,t = E(Rm,t ) + ϵ̃m,t , (6.15)

where ϵ̃m,t = σm ϵm,t . Note that we are not bothering to pull the standard deviation of the returns
outside of the shock here, so ϵ̃m,t has standard deviation of σm rather than one. The component
ϵ̃m,t of Rm,t represents a measure of systematic risk:

Definition 6.4. Systematic risk

Systematic risk (or alternatively ’market risk’ or ‘undiversifiable risk’) is risk that affects
the overall market and not just a particular stock or industry. As a result, it is not possible
to avoid this risk by holding a diversified portfolio of many stocks.
♣

How does the return on a particular individual stock relate to the return on the market? It
is natural to think that there will in general be a relation, as most firms’ prospects are tied to
the state of the macroeconomy. Since both the individual stock return and the market return are
random variables (subject to uncertain variation), it is natural to frame the relation in a linear
regression model:
Ri,t = ai + βi Rm,t + νi,t , (6.16)

where ai is a constant or regression ‘intercept’, βi captures the nature of the relation between
market returns and the returns on the i-th stock, and νi,t is an idiosyncratic shock to firm i
returns. Formally, this means that the shock νi,t is uncorrelated with the market return Rm,t and
uncorrelated with the idiosyncratic shock to any other company, i.e., ρ(νi,t , νj,t ) = 0 whenever
i ̸= j. Intuitively, νi,t reflects firm-specific news, such as the pharmaceutical trial outcome
example described above, that is unrelated to aggregate stock market news. This model is
sometimes called a single index model because there is a single source of common variation
6.5 Systematic and Idiosyncratic Return Components – 159 –

among firm stock returns: the market return. (The model can be generalized to permit other
sources of common variation, but we will not explore this extension in this course.)

6.5.1 Diversified portfolios and systematic risk

To further understand the effects of portfolio diversification, let us think about forming an
equally-weighted portfolio of N stocks, where each of the N stock returns follows Eq. (6.16).
The equal-weighted portfolio return equals:
X
N
Rp = (1/N ) Ri
i=1
XN
= (1/N ) (ai + βi Rm,t + νi,t )
i=1
X
= ā + β̄Rm,t + (1/N ) νi,t ,
P P
where ā = (1/N ) αi (the average of the ai ’s across the N stocks), and β̄ = (1/N ) βi (the
average of the βi values). The second line just plugs in the expression for Ri from Eq. (6.16) and
the third line just does some simplifying and re-arranging. The equation says that the portfolio
return has two random components: 1) the piece β̄Rm,t which is ‘systematic’ in the sense that it
P
involves the market return; 2) the piece (1/N ) νi,t , which is an equal-weighted combination
of the idiosyncratic component of the returns on the N stocks.
P
The key insight is that, as N becomes large, the piece (1/N ) νi,t will become small and
in particular its variance approaches zero. The loose intuition for this result is that the νi,t terms
are each uncorrelated, zero-mean return shocks. As we add in more and more such return shocks,
i.e., as N becomes large, it becomes less and less likely that the average of these shocks will
deviate significantly from zero. Applying this result gives, as N → ∞:

Var(Rp ) ≈ µ2β Var(Rm ), (6.17)

where µβ is the average βi value in the cross-section of firms. In words, the equation says that
the portfolio return variance will be proportional to the variance the market return. In fact, there
are good reasons to think that the quantity µβ will be close to one (it would be exactly one if
the portfolio was market-cap weighted rather than equal-weighted), and in that case the portfolio
variance is simply the market return variance. The rough intuition is that, when an investor holds
a lot of stocks with small weights on each stock, then the portfolio increasingly behaves like the
market portfolio.
Eq. 6.17 shows how a ‘well-diversified’ portfolio is subject only to systematic risk, i.e.,
uncertain about future market returns, and not by idiosyncratic risk, which as been ‘diversified
away.’ Figure 6.9 shows an illustration of how the addition of more and more stocks into
a portfolio gradually drives down the standard deviation of portfolio returns by reducing the
exposure to idiosyncratic risk. (Note that ‘unique risk’ is a synonym for idiosyncratic risk.)
At the same time, there is a bound on the ability of diversification to mitigate portfolio return
6.5 Systematic and Idiosyncratic Return Components – 160 –

variation. This is because the systematic component of return variation remains irrespective
of how diversified the portfolio becomes. In other words, an investor cannot avoid market risk
simply by investing in a large number of stocks.

Figure 6.9: The impact of diversification on portfolio return variation.

6.5.2 Interpretation of Beta

The parameter βi turns out to be a key quantity in financial theory, as we will see in the next
section. It is therefore important to clearly understand what βi represents. Using results from
linear regression theory, we can write βi as:
Cov(Ri,t , Rm,t ) σi,m σi ρi,m
βi = = 2 = , (6.18)
Var(Rm,t ) σm σm
where Cov(·) denotes the co-variance and Var(·) the variance. The third equality follows by
plugging in the result σx,y = σx σy ρx,y for two random variables x and y (equal to the stock i
return and the market return in our context) and simplifying.
1. The sign of βi is equal to the sign of ρi,m , which is the correlation between firm i returns
and market returns. In other words, if stock i’s returns are positively (negatively) correlated
with market returns, then βi is positive (negative). A value of zero for βi corresponds to
the case in which firm i’s stock returns are uncorrelated with market returns.
2. Since most firms’ prospects are better when the aggregate economy performs well, it is
reasonable to think that most stocks will exhibit a positive βi . A negative beta stock would
imply that the particular stock’s performance negatively relates to the performance of the
overall market, which is likely to occur only rarely.
3. The component σm on the far right-hand side of Eq. (6.18) is common for all stocks.
Consequently, the relative magnitude of βi across stocks is entirely determined by ρi,m σi .
This shows that 1) all else equal, a stock has a higher beta if it is a higher correlation with
market returns; and 2) all else equal, a stock has a higher beta if its return has a higher
standard deviation.
6.5 Systematic and Idiosyncratic Return Components – 161 –

The last point above implies that there is an important distinction between βi and the standard
deviation σi . In particular, a stock or other risky asset might have a very large σi ; however, if
that stock’s return is uncorrelated with market returns, then its beta is zero.
A final result concerning beta that proves useful is that the beta of a portfolio equals the
weigthed average of the betas of the stocks or other securities in the portfolio, where the weights
equal the portfolio weights:
X
N
βp = wn β n , (6.19)
n=1

where βp equals the beta on a portfolio of N assets, for n = 1, . . . , N , βn is the beta for the n-th
asset, and wn is the portfolio weight for the n-th asset.

6.5.3 Excess return form and α

Now that we understand more clearly what the key parameter βi represents, let us push a bit
further with the single index model of Eq. 6.16. Suppose there is a risk-free asset with return
Rf . Subtracting this from both sides of Eq. 6.16 and doing a bit of algebra gives:

Ri,t − Rf = ai + βi Rm,t − Rf + νi,t

Ri,t − Rf = ai + βi (Rm,t − Rf ) + βi Rf + νi,t
Ri,t − Rf = αi + βi (Rm,t − Rf ) + νi,t , (6.20)

where we make the definition αi ≡ ai + Rf (βi − 1) in moving from the second line to the final
line.
Eq. 6.20 is sometimes called the ‘excess return form’ of the single index model. It says that
the excess return for stock i equals a constant term αi , plus a ‘systematic’ component that equals
the excess return on the market, plus the same idiosyncratic or firm-specific shock as in the base
model. In a sense, all we have done here is re-written the basic model in terms of excess returns
rather than standard returns. This becomes convenient, though, when we want to think about the
risk premiums on different stocks. Recall that the risk premium equals the average excess return
for the asset. Taking the average of Eq. 6.20 gives:

E(Ri,t ) − Rf ) = αi + βi E(Rm,t − Rf ) (6.21)

Eq.( 6.21) directly relates the risk premium on a stock to the market risk premium, where the
connection is determined by the parameter βi . Note that the market risk premium is a market
concept and therefore does not vary across stocks. (In other words, it is simply a number, such
as 6% annually.) Eq.( 6.21) implies that, holding αi constant, risk premium for a stock is higher
the larger is its βi . This is almost the kind of ‘big-picture’ result we have been seeking. We
want to understand why risk premiums vary across stocks and what notion of ‘risk’ explains this
variation. Eq.( 6.21) suggests that βi , which is a measure of systematic or market risk exposure
helps determine the risk premium.
Although the above model is helpful, it is ultimately not a complete theory because of the αi
6.5 Systematic and Idiosyncratic Return Components – 162 –

term. The αi can be interpreted as an ‘extra’ component of the risk premium that is not explained
our measure of systematic risk exposure βi . A stock with a positive αi , therefore, earns an extra
risk premium relative to what is driven by systematic risk exposure. A stock with a negative αi
would earn a insufficient risk premium given its systematic risk exposure as measured by βi . In
other words, positive alpha stocks represent a ‘good deal’ to investors because they offer a higher
average return than their exposure to market risk suggests is warranted. Negative alpha stocks
are ‘bad deal’ stocks because they do not provide a sufficiently high average return to compensate
for exposure to market risk. Zero alpha stocks are ‘fair deal’ stocks – the risk premium is exactly
what is warranted by exposure to market risk. Because we have no theoretical basis for saying
what αi should look like, at this stage we do not have a complete theory of the determination of
the risk premiums on stocks. It would be nice if we could assert that αi = 0 for all stocks. This
would imply that differences in risk premium across different stocks are entirely determined by
differences in market betas, and we would then have a complete theory. The problem is that,
at this point, we do not have a basis for making this claim. The next section introduces a very
important model that will allow us to do so.

6.5.4 Measuring alphas and betas: linear regressions

The key parameters αi and βi in Eqs. (6.20) and (6.21) are population parameters that are
not directly observable. We can obtain estimates of these parameters by collecting a sample of
data consisting of excess returns on the stock of interest and excess returns on a proxy for the
market (such as S&P 500 index returns). We can then run a linear regression with excess returns
on the stock of interest as the ‘y-variable’ and market excess returns as the ‘x-variable’. The
regression should include a constant or ‘intercept’ term as well, which will give us an estimate
of α. Recall from statistics that linear regression picks values of the unknown parameters (αi
and βi in this case) that correspond to the line that best fits the data in the particular sense of
minimizing the sum of squared errors.
Let us consider a concrete example to further illustrate how linear regression provides
estimates of alpha and beta for a particular stock. Figure 6.10 plots monthly excess returns for
Disney stock (DIS) against monthly excess returns for the S&P 500 over the period 2011–2019.
The ‘cloud’ of points in the figure reveals a clear positive relation between the two variables:
excess returns on DIS tend to be high when market excess returns are high, and vice versa. Yet
it is also clear that the correlation between the two pairs of excess returns is well below one – no
line will perfectly fit these points.
In addition to the scatter plot of excess returns, Figure 6.10 shows the line fit by running a
linear regression of DIS excess returns on S&P 500 excess returns (including a constant). The
fitted regression line slopes upward as expected. In the northeast quadrant of the figure, the
fitted regression line equation is shown. This equation indicates that the regression estimate of
βi is β̂i = 1.132. Note that this value is positive as expected given the positive relation between
6.5 Systematic and Idiosyncratic Return Components – 163 –

Figure 6.10: DIS excess returns versus market excess returns, 2011-2019 monthly.

DIS excess returns and market excess returns. The regression estimate of αi is α̂i = 0.0044.
Converting to a percentage, the alpha estimate for DIS equals around 44 basis points a month,
or about 12 × 0.44 = 5.28% per year. An interpretation of this quantity is that DIS stock
seems to have earned an annualized risk premium that is around 5% larger than what would be
expected given its exposure to market risk. The figure also shows the R2 -statistic associated
with the regression. This is a measure of the quality of fit of the regression. An R2 value of
one (or 100%) would correspond to a situation in which all the points were exactly on the fitted
regression line, i.e., a case in which DIS returns were perfectly correlated with market returns.
In fact the R2 -value is around 41.9%, which can be interpreted as indicating that about 42% of
the variation in DIS excess returns can be ‘explained’ by variation in the market excess return.
What drives the other 58% of variation? This represents idiosyncratic return variation for DIS.
In general, a relatively low (high) R2 for a given stock indicates that most of the variation in that
stock’s return is idiosyncratic (systematic).
Have we learned that the true beta for DIS is 1.132? Unfortunately, no. The value 1.132
is simply an estimate of the true (unknown) beta value. Similarly, the value of 44 basis points
a month is just an estimate of the true unknown alpha value for DIS. Typical linear regression
output from Excel or other computer programs that execute linear regressions includes standard
errors for the alpha and beta estimates. These are measures of the sampling variation in the
estimates. A useful rule of thumb is that a 95% ‘confidence interval’ for the unknown parameter
can be obtained by taking the estimate and subtracting / adding two times the standard error of
the parameter estimate.
Figure 6.11 shows more detailed regression output for the same regression of DIS excess
returns on market excess returns. (These results are produced using Excel’s Data Analysis
 6.6 The CAPM and Beta as a Systematic Risk Measure – 164 –

toolpack, and I have edited the raw output slightly to simplify and clarify it.) The ‘coefficients’
column shows the regression estimates of alpha (‘intercept’) and beta (‘S&P 500 Excess return’).
The next column shows the standard error for each estimate. For example, the standard error
associated with the beta estimate is around 0.13. We can therefore obtain a 95% confidence
interval for DIS’s beta estimate by taking 1.13 and subtracting and adding 0.26 (2 × 0.13). This
gives (0.87, 1.39). A more exact (without my heavy rounding) version of this confidence interval
is given by the last two columns of the output.3 The ‘P -Value’ column shows the likelihood of
obtaining the parameter estimate by random chance under the assumption that the true parameter
value is in fact zero. This is basically zero for the beta estimate. In other words, we can be very
confident that the true DIS beta value is greater than zero. Note that the 95% confidence interval
for our α estimate includes zero. The P -value is about 32.23%. Together, these facts illustrate
that our alpha estimate is quite imprecise and we cannot rule out the possibility that the true
alpha might be zero.

Figure 6.11: Regression output for regression of DIS excess returns on market excess returns.

6.6 The CAPM and Beta as a Systematic Risk Measure

The Capital Asset Pricing Model (CAPM) was jointly developed by several financial
economists in the 1960s. Although the model has weaknesses, as we will eventually dis-
cuss, the CAPM was a critically important development in financial economics and remains the
‘benchmark model’ in financial theory. A particular, notable feature of the model is that it links
the risk premium that a security should earn not to the standard deviation of the security’s return,
but rather to a firm’s market beta (or β) – the same β that we introduced in the previous section
that captures to what extent the security’s returns co-move with the market.

6.6.1 CAPM assumptions and key implications

To introduce the CAPM, let us return to the mean-variance frontier with many assets
discussed in Section 6.4. Recall that a set of N risky assets will generate a bullet-shaped frontier

3The ‘t Stat’ column shows the so-called t-statistic for the regression, which is computed by dividing the estimate
(1.13 for beta) by the standard error. A t-statistic with absolute value of 2 or greater indicates strong evidence against
the null hypothesis that the true value of the corresponding parameter equals zero.
6.6 The CAPM and Beta as a Systematic Risk Measure – 165 –

and, when a risk-free asset is also available, the best portfolio of risky assets is given by the
tangency portfolio (determined as the portfolio on the frontier that is on a line emanating from
the risk-free return and just tangent to the curve). But what is the identity of the tangency
portfolio? So far we have not said this. Under the CAPM, however, the tangency portfolio is
very specifically characterized.
The CAPM makes the following key assumptions:
1. Investors choose portfolios based on trading off the mean and variance of portfolio returns.
2. All investors perceive the same set of investment opportunities – this means that all
investors agree about the position of the curve we draw in mean-standard deviation space.
3. Investors consider a single period and common investment horizon
4. The market is in equilibrium: supply equals demand for all securities.
5. There are no taxes, limits on positions, or other sources of frictions in asset markets.
6. The market is ‘competitive’: there are many investors and investors act as ‘price takers’.
Because all investors choose portfolios based on mean-variance considerations, and they
all agree about the positioning of the mean-variance frontier, two-fund separation holds and all
investors hold a combination of the tangency portfolio and the risk-free asset. The key addition
is the implication of the assumption that the market is in equilibrium, i.e., demand equals supply
for all assets. This implies that the market portfolio – the value-weighted portfolio of all
available risky assets – must be the tangency portfolio.
Given that the market portfolio is the tangency portfolio under CAPM, all investors hold
a combination of the risk-free asset and the market, with risk aversion levels dictating the
proportions of each. We then refer to the general ’capital allocation line’ depicting combinations
of the tangency portfolio with the risk-free asset as the ’capital-market line’ (CML). Figure
6.12 shows the situation under the CAPM. There is a curve in mean-standard deviation space
that represents the minimum variance frontier for all available risky assets. There is the usual
tangency portfolio, except that we now identify this portfolio as synonymous with the market
portfolio. The CML is the line that connects the risk-free return with the tangency portfolio.
The second key implication of the CAPM is that the risk premium for an asset is proportion
to its market beta. This result is sufficiently important that we will state it as a theorem:
Theorem 6.1. Expected returns under the CAPM
Assuming the CAPM holds, the risk premium on any asset is given by the following
equation:
E(Ri ) − Rf = βi [E(Rm ) − Rf ], (6.22)

where Ri is the return on an arbitrary stock or other asset, Rm is the market return, Rf
is the risk-free return, and βi is the market beta for the asset.
♡

Notice the similarity between Eq. (6.22) and Eq. (6.21) for the single index model discussed
in the previous section. The equations are the same except for the fact that, under the CAPM, the
6.6 The CAPM and Beta as a Systematic Risk Measure – 166 –

Figure 6.12: The capital-market line (CML) under the CAPM.

‘alpha’ term is equal to zero for all stocks. This is worth stating as another theorem:
Theorem 6.2. CAPM and alpha
If the CAPM holds, then the parameter αi associated with the single index model equals
zero for all securities and portfolios.
♡

The CAPM delivers a number of key predictions:

Risky stocks - those commanding a high risk premium – are high beta stocks. This is
because beta measures the sensitivity of an asset to market risk. Because investors hold
diversified portfolios on the efficient frontier, they do not care about idiosyncratic risk,
which is diversified away, and therefore only market risk as measured by beta receives a
‘premium’ in the form of a higher expected return.
The key Eq.(6.22) implies that, under CAPM, the only factor that determines differences
in averages returns across assets is differences in market betas.
Holding the market portfolio (along with the risk-free asset) is efficient. investors can skip
trouble of individual portfolio analysis, etc. The CAPM allows little scope for entities such
as actively managed mutual funds or hedge funds – why pay a professional investor to tell
you to invest in the market?
The CAPM’s key prediction is that α = 0 for all securities and (therefore) for all portfolios
of securities, including those of mutual and hedge fund managers. This provides a key
“testable implication” for the model.

6.6.2 The SML and using the CAPM equation

The security-market line (SML) offers a visual depiction of the key implication of the
CAPM. The SML is simply the line that corresponds to Eq. (6.22), rewritten below in ‘slope-
6.6 The CAPM and Beta as a Systematic Risk Measure – 167 –

intercept’ form:
E(Ri ) = (E(Rm ) − Rf ) βi + Rf . (6.23)

To map this to the usual y = mx + b, note that E(Ri ) plays the role of y, βi plays the role of
x, the market risk premium (E(Rm ) − Rf ) plays the role of the slope m, and Rf plays the role
of b. The equation should be viewed as describing how expected returns vary across different
stocks, securities, or portfolios, and states that expected returns are linear in the market beta of
these stocks, securities, or portfolios. Figure 6.13 shows an illustration of the SML. The security
β is on the x-axis. According to the CAPM, the expected return on any zero-beta security equals
the risk-free rate Rf . As the value of beta increases, the expected return increases proportionally.
Any security with a β = 1 should have an expected return equal to the market expected return
according to the CAPM. The slope of the SML is therefore equal to the market risk premium
(apply the ‘rise over run’ formula).
It is important to emphasize that, if the CAPM holds, the expected return and beta combi-
nations for every stock, security, or portfolio will fall exactly on the SML. Deviations of security
expected return-beta combinations from the SML correspond to nonzero alphas. For example,
a security with a positive alpha (if one exists) would appear in Figure 6.13 as a point above the
SML. Similarly, a negative alpha security would appear as a point below the SML.

Figure 6.13: The security-market line (SML).

Example 6.6 TechFirm stock has a market beta of 2.1. FastFoodBrands Inc. stock as a market
beta of 0.75. The market risk premium is 8% annually and the (annual) risk-free rate is 2%.
Assuming the CAPM holds, compute the expected return for both stocks. Suppose that the
actual expected return of TechFirm is 16%. Where does a point corresponding to TechFirm’s
characteristics fall relative to the SML?
6.6 The CAPM and Beta as a Systematic Risk Measure – 168 –

Solution: First, apply the CAPM equation for the expected return to determine the CAPM-
implied expected return for both stocks. For Techfirm, we have:

E(Ri ) = (E(Rm ) − Rf ) βi + Rf
= (8%)2.1 + 2%
= 18.8%

For FastFoodBrands, we have:

E(Ri ) = (E(Rm ) − Rf ) βi + Rf
= (8%) 0.75 + 2%
= 8%

Finally, if the actual expected return for TechFirm equals 16%, then TechFirm has an α value of
16% − 18.8% = −2.8% per year. The point corresponding to its expected return and market
beta characteristics would fall below the SML.
The CAPM applies not only to individual stocks but also to portfolios. The following
example illustrates this idea.

Example 6.7 Fidelity runs a large number of mutual funds. One of these is the Fidelity
Contrafund, which is described as a ‘large cap growth’ style fund. Suppose that the beta
for Fidelity’s Contrafund is 0.99. Suppose the market risk premium is 7%. What does the
CAPM imply should be the risk premium for Contrafund? Suppose that Contrafund has earned
an annual return of 10% over the past 5 years and the risk-free rate is 2%. What is an estimate
of the alpha associated with Contrafund? Interpret your answer in terms of whether this fund is
an attractive investment opportunity.
Solution: Contrafund’s beta is very close to the market beta, which is 1 by definition. Therefore,
the CAPM implies that Contrafund’s risk premium should be nearly identical to the market risk
premium. In particular, the predicted risk premium under CAPM equals βi (E(Rm ) − Rf ) =
0.99(7%) = 6.93%. The corresponding expected return would be the previously computed risk
premium plus the risk-free rate, or 6.93% + 2% = 8.93%. The actual historical average return
for Contrafund of 10% therefore implies an estimated alpha equal to 10% − 8.93% = 1.07%.
The positive alpha estimate suggests that the fund is an attractive investment. However, it is
important to bear in mind that this is only an estimate of the true unknown alpha. Without
computing a standard error (or being given one), we cannot determine whether there is strong
statistical evidence against the null hypothesis that Contrafund’s true alpha is zero. This fact
might temper our enthusiasm about pouring money into this fund.
As a final application of the CAPM, let us consider a simple stock valuation exercise. This
involves estimating the fundamental value of a stock by computing the discounted value of the
future dividends that shareholders can expect to receive. Consider the following relatively simple
example of this variety:
6.7 Weaknesses of CAPM – 169 –

Example 6.8 ABC stock currently pays a dividend of $2/share. Suppose that an analyst forecasts
that ABC’s dividend will grow at a steady rate of 4% annually into perpetuity (forever). In
addition, suppose the market risk premium is 6%, the risk-free rate is 3%, and the market beta
of ABC stock is 1.2. Determine a valuation for ABC stock (a price for the stock) as the present
value of the future predicted dividends for ABC using the CAPM-implied expected return as the
discount rate. Assume that the most recent dividend of $2/share was just paid.
Solution: The first step is to obtain the CAPM-implied expected return that we will use as a
discount rate. To do so, apply the CAPM equation as in previous examples:

E(RABC) = (E(Rm ) − Rf ) βABC + Rf

= (6%) 1.2 + 3%
= 10.2%

The above computations give us a 10.2% annual discount rate to use in the present value
calculation. Note that we can interpret this discount rate as the sum of the 3% rate that would be
used to discount truly risk-free cash flows (the risk-free rate) and a 7.2% adjustment for the risk
associated with the company’s dividends.
The next step is to apply the formula for the present value of a perpetuity with growth from
the time value of money material we covered in Chapter 1. As a reminder, this formula is:
C
PV = (6.24)
r−g
We will set r = 10.2% and g = 4% in this formula. To obtain C, recall that the C should be the
initial cash flow received in one period. Consequently we should set C = $2 × (1.04) = $2.08.
Finally, plugging in we obtain:
C 2.08
PV = = ≈ $33.55
r−g 0.102 − 0.04
Consequently, our valuation indicates a fundamental value of ABC stock of $33.55.
As an epilogue to the previous example, suppose that we see that shares of ABC are
currently trading for $28/share. Then our valuation computation suggests that ABC shares are
undervalued. However, we should be cautious before taking a significant long position (buying)
ABC. In particular, our valuation computation is only as good as the underlying assumptions we
made. These include our assumptions about ABC’s market beta and, critically, the growth rate of
future dividends. One of the chapter end problems extends this example by considering a more
involved set of assumptions about dividend growth for ABC.

6.7 Weaknesses of CAPM

In this final section, we briefly discuss the question of whether real world financial data
support the key predictions of the CAPM. This is an important question both for theoretical and
practical reasons. From a theoretical perspective, if the CAPM is not supported by the data, then
this suggests that one or more of the underlying assumptions fail in important ways, and that
6.7 Weaknesses of CAPM – 170 –

news theories should be devised. From a practical perspective, recall that the CAPM leaves little
scope for, e.g., actively managed mutual funds, hedge funds, and the like.
It turns out that a significant body of research points to weaknesses in the CAPM. In other
words, the data do not appear to support some of the key predictions of the model. Among the
most important set of testable predictions is that alphas should be zero across all securities and
portfolios. Many researchers have formed portfolios of stocks by sorting on particular observed
characteristics, such as size (market capitalization), financial ratios, accounting measures of
profitability, measures of past performance, and so on. If the CAPM holds, none of these
characteristics should be associated with differences in average returns once we control for
potential differences in market betas. Yet, in practice, this does not appear to be the case.
One of the more well-known examples of an apparent failure in the CAPM involves sorts of
firms on the book-to-market (BTM) ratios. The BTM is the ratio of the (accounting) book value
of equity to the market value of equity. Firms with a low BTM are firms whose market value
is much higher than the book value of equity. Such firms are often growth firms or ‘glamour’
stocks whose prices have been bid up, presumably in anticipated of high future growth. Low
BTM firms are those firms with market values that are closer to the accounting value of assets in
place (as measured by the book value of equity). They are often termed ‘value stocks.’ The key
prediction of the CAPM is that the BTM ratio should not relate differences in average returns
once we control for differences in market betas. Figure 6.14 shows a summary of statistical
evidence to the contrary. Here stocks are sorted each year into 10 portfolios ranked by BTM
(high BTM stocks or value stocks are decile 10 and low BTM or growth stocks are decile 1).
There is a clear increasing pattern in expected returns as the BTM increases. Most importantly,
this increase is not fully explained by differences in market beta across the portfolios. High BTM
firms earn positive alphas and low BTM firms earn negative alphas. A ‘long-short’ portfolio that
buys high BTM stocks and shorts low BTM stocks earns a statistically significant positive alpha.

Figure 6.14: Non-zero alphas for portfolios sorted on book-to-market.

Other characteristics, such as recent momentum in stock prices, also seem to generate
nonzero alphas in the data. Collectively, these results indicate that the CAPM is an imperfect
model. So-called ‘multi-factor’ extensions of the CAPM add additional risk factors, derived from
6.7 Weaknesses of CAPM – 171 –

alternative perspectives, that seek to address the shortcomings of the CAPM. The conjecture here
is that, once we account for exposure to these additional risk factors, alphas will then disappear.
But the precise nature of such factors remains a hotly debated topic in finance. Another potential
source of alphas are behavioral factors – the notion that stock prices (and thus expected returns)
can deviate from the ‘rational’ levels due to a variety of flaws in human reasoning skills. The
extent to which apparent alphas represent behavioral effects or some form of omitted risk factor
remains a point of contention in the field of finance.
6.7 Weaknesses of CAPM – 172 –

K Chapter 6 Problems k
1. At the beginning of this month, Stock A had a price of $50/share with 100 million shares
outstanding. Stock B had a price of $20/share with 700 million shares outstanding. Finally,
Stock C had a price of $150/share with 20 million shares outstanding. Consider forming an
equal-weighted versus value-weighted portfolio of these three stocks. Compute the return
for both portfolios this month assuming that Stock A has a monthly total return (including
any dividends) of 5%, Stock B has a monthly total return of -3%, and Stock C has a total
return of 11%.
2. Consider the following scenario with two risky securities:

State of Economy Probability Stock A Return Stock B Return

Recession 20% -6% -30%
Normal Growth 60% 8% 5%
Boom 20% 12% 55%
A. Complete the following table based on the above information:
Stock A Stock B

Expected return

Variance

Standard Deviation

Covariance (A, B) NA

Correlation (A, B) NA
B. If you hold a portfolio that is invested 45% in Stock A and 55% in Stock B, what is
the portfolio expected return and standard deviation?
C. Referring to the previous portfolio (45% A / 55% B), if the current risk-free rate is 3
percent, what is the risk premium on the portfolio?
3. Your sister holds a portfolio consisting of a 50% allocation to a (risk-free) money market
account that earns 2% per year and a 50% allocation to an ETF that tracks a broad stock
index. Suppose you are informed that the risk premium for her portfolio is 6% per year
and the standard deviation of the portfolio return equals 10% per year. She would like to
alter her allocation to target a high expected return of 20% per year, but she is concerned
about the level of risk she would be taking. Assist her by computing the new weights that
will deliver a 20% expected return and determine the corresponding standard deviation of
the portfolio return.
4. Returns on stock ABC have a standard deviation of 10%. Returns on stock XYZ have a
standard deviation of 24%.
A. Determine the maximum possible standard deviation of an equally-weighted portfolio
of these two stocks.
6.7 Weaknesses of CAPM – 173 –

B. Compute the standard deviation of an equally-weighted portfolio of the two stocks

assuming in addition that their returns are uncorrelated.
C. Is it possible for the standard deviation of an equally-weighted portfolio of the two
stocks to be lower than that obtained in part B.? Explain briefly.
5. Ford stock has an expected return of 9% per year and a standard deviation of 15%. Yum!
Brands stock has an expected return of 13.5% per year and a standard deviation of 30%
per year. Suppose the risk-free rate equals 5%.
A. If you can form a portfolio consisting of either Ford or Yum! Brands and the risk-free
asset, which is better to combine with the risk-free asset? Explain your reasoning.
B. At what alternative standard deviation for Yum! Brands returns would you be just
indifferent?
C. (Optional) What does the previous case imply about the nature of the tangency
portfolio?
6. There are two risky assets (A and B). Asset A has an expected return of 5% and standard
deviation of 10%. Asset B has an expected return of 9% and a standard deviation of
20%. Returns for the two assets are perfectly negatively correlated. Consider a portfolio
with 2/3 invested in Asset A and 1/3 invested in Asset B. Determine the variance of the
corresponding portfolio return. Suppose that a risk-free asset is also available. What is
the most you can say about the risk-free return in this situation?
7. Consider an investment universe with three risky assets (Stocks 1, 2, and 3) and a risk-free
asset. Suppose that the composition of the tangency portfolio is 20% Stock 1, 15% Stock
2, and 65% Stock 3. If an investor chooses to allocate 40% of wealth to the risk-free asset,
and the rest to the tangency portfolio, what are the portfolio weights on Stocks 1, 2, and 3?
8. You are going to form a portfolio out of two assets: a risk-free asset (the current 3-month
Treasury bill return) and a risky asset (the Vanguard Total Stock Market ETF, Ticker:
VTI). The main details are given below:

Average Return Standard Deviation

Total Stock Market ETF 7.69% 15.8%
3 Month Treasury Bill 0.67% 0%

A. Graph in terms of expected returns and standard deviation the possible set of portfolios
if you restrict yourself to these two assets.
B. What would be the expected return and standard deviation of the portfolio if you hold
80% in the risky asset and 20% in the risk-free asset?
C. How much of your portfolio would you need to invest in the risky asset to have a
portfolio with a standard deviation of 5%? (Hint: solve for the weight in the risky
asset, ω, using the fact that the weight in the risk-free asset is (1 − ω).)
D. How much of your portfolio would you need to invest in the risky asset to have a
portfolio with an expected return of 10%?
6.7 Weaknesses of CAPM – 174 –

9. How many correlations are relevant in computing the standard deviation of the return on
a portfolio consisting of six risky securities?
10. Four stocks have identical standard deviations of 20%. In addition, all pairwise correlations
among the stocks are equal (call this value ρ. What is the variance and standard deviation
of an equally-weighted portfolio of the stocks?
11. A researcher runs a regression of monthly excess returns on Z-Tech company on excess
returns for the Russell 1000 (as a market return proxy). The regression output indicates
that the intercept estimate equals -0.0026, with a standard error of 0.0018, and the slope
estimate equals 0.875 with a standard error of 0.231.
(a). What is regression-based estimate of the market beta for Z-Tech?
(b). What is the regression-based estimate of the alpha for Z-Tech?
(c). Can the null hypothesis that the true market beta for Z-Tech equals zero be rejected
at the 95% confidence level?
(d). Can the null hypothesis that the true alpha for Z-Tech equals zero be rejected at the
95% confidence level?
12. Consider the following stocks:

Stock Beta Share Price Market Capitalization

Agilent Technologies (A) 1.55 52.57 17.17B
AT&T Corp (T) 0.44 39.93 245.53B
Citigroup (C) 1.55 57.72 158.92B
Dominion Resources (D) 0.24 77.54 48.71B

A. Calculate the beta of an equally-weighted portfolio of the four stocks.

B. Calculate the beta of a market-capitalization (value) weighted portfolio of the four
stocks.
C. Calculate the expected return according to the CAPM of the two portfolios, assuming
the expected market risk premium, (E[Rm ] − Rf ), is 7% and the risk-free rate, Rf ,
is 2%.
13. PepsiCo.’s (PEP) estimated market beta on Yahoo! Finance is 0.58, whereas Coca-Cola’s
(KO) is 0.57. Assuming these beta values are accurate, discuss what the CAPM implies
about the expected return of a so-called ‘long-short’ portfolio that is long $0.5 million in
PEP and short $0.5 million in KO. Discuss briefly what might be the motivation for such
a portfolio.
14. The Sweet Spot hedge fund has a market beta of 1.2. Suppose that the average annual
return of the fund is 19%. If the risk free rate is 3.5% and the market risk premium is
8%, what is the fund’s alpha? How would the fund appear relative to the SML in the SML
graph?
15. A portfolio that combines the risk-free asset and the market portfolio has an expected
return of 7.5% and standard deviation of 17%. The risk-free rate is 3.8%, and the expected
6.7 Weaknesses of CAPM – 175 –

return on the market portfolio is 10%. If the CAPM holds, what expected return would
an asset exhibit if its return had a correlation equal to 0.44 with the market return and a
standard deviation of 58%?
16. Revisit Example 6.8 dealing with the fundamental value of ABC stock. Continue to
maintain all of the assumptions in that example, except now assume that dividends will
grow at the higher rate of 8% per year for the next four years, and then at a rate of 4%
per year thereafter into perpetuity. How much does this alter the valuation of ABC stock
relative to the computations in Example 6.8? HINT: consider the stream of dividends as
an annuity with growth followed by a delayed perpetuity.
17. Excel problem Download the spreadsheet "Disney.xlsx" from the Canvas page for the
course. Use the provided data to replicate the regression analysis discussed in Section 6.5.
18. Excel problem Copy the risk-free rates and S&P 500 prices and returns given in the
“Disney.xlsx” spreadsheet to a new sheet. Go to Yahoo! Finance and download monthly
data for Best Buy stock (BBY) over the same time period. Compute returns and excess
return BBY. Then use Excel to perform a linear regression of BBY excess returns on market
excess returns. Clearly indicate the corresponding alpha and beta estimates for BBY. Obtain
95% confidence intervals for these estimates and briefly interpret your results.
 Appendix Mathematical Details

This appendix covers some of the basic mathematics used in the course.

A.1 Summation Operator, Infinite Series, and Proofs of the

Perpetuity and Annuity PV Formulas

The summation operator is an abbreviation used to express the summation of numbers. If

{xi : i = 1, 2, . . . , n} is a sequence of n numbers, the summation of the n numbers is:

X
n
xi ≡ x1 + x2 + · · · + xn (A.1)
i=1

Mathematically, the present value of a perpetuity involves the sum of an infinite series.
Not all such series converge to well-defined sums. As a simple example, consider the trivial
infinite series comprised of ones: 1, 1, 1, . . . . It is hopefully obvious that the sum of this infinite
series is infinitely large.1
The particular infinite series of interest for this course takes the form: α, α2 , α3 , . . . for
0 < α < 1. In this case, the sum of the corresponding series is well-defined (for intuition, note
that the N -th term αN is tending rapidly to zero as N increases). The sum equals:
X∞
α
αn = . (A.2)
1−α
n=1
The above result allows us to easily prove the PV formula for a standard perpetuity.
Proposition A.1. PV of Perpetuity Formula
C
The PV of a perpetuity is equal to r ♠

Proof Write out the PV as an infinite sum:

C C C
PV = + 2
+ + ...
(1 + r) (1 + r) (1 + r)3


= C α + α2 + α3 + ... ,

where 0 < α = 1/(1 + r) < 1 (so long as r > 0 as is assumed). Finally, applying the result of
Eq. (A.2) gives:
α
PV = C
1−α
C
, =
r
where the second line plugs back in the definition of α and simplifies.

1Pick any integer N > 0, then the sum of an infinite series of ones is greater than the sum of a series N + 1 ones,
which in turn is greater than N . Since N is arbitrary, the sum is greater than any N , and hence infinite.
A.2 Review of Logarithms and Exponentials – 177 –

Given the result for the PV of a perpetuity, the formula for the PV of an annuity can be
proved by considering the annuity as a difference between two perpetuities.
Proposition A.2. PV of Annuity Formula
 
The PV of a standard annuity is equal to Cr 1 − 1
(1+r)N
♠

Proof The annuity is equivalent to the difference of a standard perpetuity of C per period and
a second ‘delayed’ perpetuity with first cash flow occurring N + 1 periods in the future. (See
Figure 1.5 for an illustration.) Using the formula for the PV of a perpetuity, and discounting the
value of the second delayed perpetuity back to the present we have:
C 1 C
P V (Annuity) = − N
r
|{z} (1 + r) r
| {z }
PV of Perpetuity PV of Delayed Perpetuity
 
C 1
= 1− .
r (1 + r)N

A.2 Review of Logarithms and Exponentials

This section reviews key properties of logarithms and exponentials. We will focus on the
natural logarithm, or log to base e, where e is the exponential number. Key properties include:

ln(1) = 0; ln(exp) = 1 (A.3)

ln(a × b) = ln(a) + ln(b) (A.4)

a
ln = ln(a) − ln(b) (A.5)
b

ln(ab ) = b ln(a) (A.6)

Key properties of the exponential function are:

exp(a) exp(b) = exp(a + b) (A.7)

exp(a)b = exp(ab) (A.8)

A.3 Probability and Statistics

This section reviews some key results from probability and statistics. First we focus on a
discrete outcome setting, i.e., a setting in which only a finite number of outcomes are possible.
Let the s = 1, 2, . . . , S denote the random state that occurs, where S equals the total number of
A.3 Probability and Statistics – 178 –

possible outcome. The set of possible states or outcomes of the underlying random environment
is known as the sample space in probability.
In a coin flip setting, for example, there are two possible outcomes, heads or tails. The
sample space is therefore heads, tails, which we can label as states one and two, respectively,
where s = 1 corresponds to the outcome ‘heads’ while s = 2 corresponds to ‘tails’. In an
economic environment, we can imagine a random business cycle scenario with 3 states: s = 1
denotes ‘expansion/boom’, s = 2 denotes ‘normal’, and s = 3 denotes ‘recession/bust’.
Each state s occurs with a specified probability P (s). In the coin toss setting, for example,
P (1 (’heads’)) = P (2 (’tails’)) = 0.5. Probabilities must be non-negative and sum to one across
the states:

P (s) ≥ 0, s = 1, ..., S (A.9)

X
S
P (s) = 1 (A.10)
s=1
A random variable is formally a mapping (function) from the sample space to the set of
real numbers (or the set n-dimensional vectors in a multivariate setting). As a simple example,
consider the random experiment of flipping a coin 3 times and recording the number of heads.
The sample space for this experiment can be written as

S = HHH, HHT, HT H, HT T, T T T, T T H, T HT, T HH

. We can define a random variable X as the number of heads that occur in the three tosses. This
random variable takes four possible values: 0, 1, 2, or 3. Note also that the probability that X
takes on each of these values can be readily deduced from the underlying probability of the coin
tosses. For example, the probability X = 0 is 1/8, or 12.5%, as only one of the eight possible
coin toss outcomes (which are equally likely) involves zero heads. Similarly, the probability that
X <= 2 equals 7/8, or 87.5%.
Given a random variable and its associated probability distribution, it is often of interest
to examine certain key attributes of the random variable. The first among these is the mean or
expected value of the random variable, denoted E(X), µ(X), or simply µ if the random variable
involved is clear. The definition of the expected value for a discrete random variable is:

X
S
E(X) = µ(X) = P (s)X(s) (A.11)
s=1

Note that Eq. (A.11) takes the form of a weighted average of the outcomes in each state
X(s), where the weights equal the probabilities of the corresponding states P (s). In the special
P
case in which the states are equally likely, then we can write E(X) = (1/S) Ss=1 , but it must
be remembered that this is a special case and only valid under equal state probabilities.
A.3 Probability and Statistics – 179 –

A second key attribute of random variable is the variance

X
S
Var(X) = σ 2 (X) = P (s)(X(s) − µ(X))2 (A.12)
s=1
Like the equation for the expected value, Eq. A.12 also takes the form of a weighted
average, where the the weights are the probabilities of each state. In this case, however, the
weight average is computed based on the squared deviations of the variable from its mean.
Intuitively, the variance provides a measure of the extent of dispersion around the mean. A
random variable that takes a constant value in every state (such as a ‘risk-free return’ in finance)
has a variance of zero. The standard deviationof a random variable, denoted σ(X) or simply
σ , equals the square root of the variance;
p
σ(X) = Var(X) (A.13)

A useful property of the standard deviation is that it is measured in the same units as the
underlying random variable. This is not the case for the variance, which ,makes it somewhat
difficult to interpret. In the context of stock returns, for example, the standard deviation of the
annual return on a particular is in percent per year.