CAS Exam 8 Notes - Parts A&B

Portfolio Theory and Equilibrium in Capital Markets

Fixed Income Securities

Part I
Table of Contents
A Portfolio Theory and Equilibrium in Capital Markets

BKM - Ch. 6: Risk aversion and capital allocation to risky assets . . . . . . . . . . . . . . . . . . .

BKM - Ch. 7: Optimal risky portfolios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

BKM - Ch. 8: Index models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

17

BKM - Ch. 9: The Capital Asset Pricing Model (CAPM) . . . . . . . . . . . . . . . . . . . . . . .

27

BKM - Ch. 10: Arbitrage pricing theory and multifactor models of risk and return . . . . . . . . .

37

BKM - Ch. 11: The efficient market hypothesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

43

BKM - Ch. 12: Behavioral finance and technical analysis . . . . . . . . . . . . . . . . . . . . . . .

51

BKM - Ch. 13: Empirical evidence on security returns . . . . . . . . . . . . . . . . . . . . . . . . .

57

B Fixed Income Securities

BKM - Ch. 14: Bond prices and yields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

69
71

Hull - Ch. 4: Interest rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

81

BKM - Ch. 15: The term structure of interest rates . . . . . . . . . . . . . . . . . . . . . . . . . .

89

Hull - Ch. 6.1: Day count and quotation conventions . . . . . . . . . . . . . . . . . . . . . . . . . .

95

Hull - Ch. 22 - Part 1: Credit risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

97

Altman: Measuring corporate bond mortality and performance . . . . . . . . . . . . . . . . . . . . 101

Cummins: CAT Bonds and other risk-linked securities . . . . . . . . . . . . . . . . . . . . . . . . . 105
Additional Notes

113

Portfolio Theory and Equilibrium in Capital Markets

BKM - Ch. 6: Risk aversion and capital allocation to risky assets

Introduction
The process of constructing an investor portfolio can be viewed as a sequence of two steps:
1. Selecting the composition of ones portfolio of risky assets such as stocks and long-term bonds
2. Deciding how much to invest in that risky portfolio vs. safe assets such as short-term T-bills
Fundamental part of asset allocation problem: Characterize risky portfolio risk-return trade-off
Then, the fundamental decision is capital allocation between the risk-free and the risky portfolio
Two themes in portfolio theory:
1. Investors will avoid risk unless they can anticipate a reward for engaging in risky investments
2. Utility model allows to quantify investors trade-offs between portfolio risk/expected return
Risk and risk aversion
Risk, speculation, and gambling
Speculation: The assumption of considerable investment risk to obtain commensurate gain
Considerable risk: The risk is sufficient to affect the decision
Commensurate gain: Positive risk premium, i.e., expected profit > risk-free alternative
Gamble: To bet or wager on an uncertain outcome
The central difference with speculation is the lack of commensurate gain
To turn a gamble into a speculative prospect requires an adequate risk premium to compensate
risk-averse investors for the risks they bear
A risky investment with a risk premium of zero, aka fair game, amounts to a gamble
A risk-averse investor will reject it
In some cases a gamble may appear to the participants as speculation:
Economists call this case of differing beliefs heterogeneous expectations
Risk aversion and utility values
Risky assets command a risk premium in the marketplace Most investors are risk averse
Investors who are risk averse reject investment portfolios that are fair games or worse
A risk-averse investor penalizes the expected rate of return of a risky portfolio by a certain
% (or penalizes the expected profit by a dollar amount) to account for the risk involved
The greater the risk, the larger the penalty
We will assume that each investor can assign a welfare, or utility, score to competing investment
portfolios based on the expected return and risk of those portfolios
Higher utility values are assigned to portfolios with more attractive risk-return profiles
Portfolios have higher utility scores for higher expected returns/lower scores for higher volatility
E.g.,utility score for portfolio with expected return E(r) and variance of returns 2 : utility score:
U = E(r) 12 A 2

(1)

Where U is the utility value and A is an index of the investors risk aversion
How variance of risky portfolios lowers utility depends on A, the investors degree of risk aversion
More risk-averse investors (larger values of A) penalize risky investments more severely
Investors select the investment portfolio providing the highest utility level
Risk-free portfolios utility score = their (known) rate of return (no penalty for risk)
We can interpret the utility score of risky portfolios as a certainty equivalent rate of return
Certainty equivalent rate: The rate that risk-free investments would need to offer to provide
the same utility score as the risky portfolio
Natural way to compare the utility values of competing portfolios
A portfolio is desirable only if its certainty equivalent return > risk-free alternative
Risk-neutral investors (A = 0) judge risky prospects solely by their expected rates of return
The level of risk is irrelevant to the risk-neutral investor: There is no penalty for risk
For this investor, a portfolios certainty equivalent rate is simply its expected rate of return
3

E(r)

Indifference curve
Northwest
(preferred direction)
Q

II

E(rp)

III

IV

Figure 1: The trade-off between risk and return of a potential investment portfolio P
A risk lover (A < 0) is willing to engage in fair games and gambles: This investor adjusts the
expected return upward to take into account the fun of confronting the prospects risk
Portfolio P (expected return E(rp ), standard deviation p ) is preferred by risk-averse investors to
any portfolio in quadrant IV because it has an expected return any portfolio in that quadrant
and a standard deviation any portfolio in that quadrant
Conversely, any portfolio in quadrant I is preferable to portfolio P
Mean-variance (M-V) criterion: Portfolio A dominates B if:
E(rA ) E(rB )

and

A B

In the E- plane in Fig. 1, the preferred direction is northwest, because we simultaneously increase
the expected return/decrease the variance of the rate of return
Indifference curve: Equally preferred portfolios will lie in the mean-standard deviation plane on
a curve called the indifference curve that connects all portfolio points with the same utility value
Estimating risk aversion
One way is to observe individuals decisions when confronted with risk
Consider an investor with risk aversion A whose entire wealth is in a piece of real estate
Suppose that in any given year there is a probability p of a disaster that will wipe out the
investors entire wealth. Such an event would amount to a rate of return of 100%
With probability 1 p, real estate remains intact, and rate of return is zero
The expected rate of return of this prospect is:
E(r) = p (1) + (1 p) 0 = p
The variance of the rate of return equals the expectation of the squared deviation:
2 (r) = p (p 1)2 + (1 p) p2 = p(1 p)
Utility score:
U = E(r) 12 A 2 (r) = p 12 Ap(1 p)

(2)

We can relate the risk-aversion parameter to the amount that an individual would be willing
to pay for insurance against the potential loss. Suppose an insurance company offers to cover
any loss over the year for a fee of dollars per dollar of insured property
Such a policy amounts to a sure negative rate of return of , with a utility score: U =
Maximum value of the investor is willing to pay? Equate the utility score of the uninsured
property to that of the insured property, and solve for
= p[1 + 21 A(1 p)]

(3)

Square brackets in Eq. 3 = multiple of expected loss p the investor is willing to pay
Economists estimate that investors exhibit degrees of risk aversion in the range of 2 to 4
More support for the hypothesis that A is somewhere in the range of 2 to 4 can be obtained from
estimates of the expected rate of return and risk on a broad stock-index portfolio
4

Capital allocation across risky and risk-free portfolios

The most straightforward way to control the risk of the portfolio is through the fraction of the portfolio
invested in Treasury bills and other safe money market securities versus risky assets
This capital allocation decision is an example of an asset allocation choice - a choice among broad
investment classes, rather than among the specific securities within each asset class
Most investment professionals: Asset allocation = most important part of portfolio construction
Take composition of risky portfolio as given and focus on allocation between it/risk-free securities
When we shift wealth from the risky portfolio to the risk-free asset, we do not change the relative
proportions of the various risky assets within the risky portfolio
Rather, we reduce the relative weight of the risky portfolio as a whole in favor of risk-free assets
As long as we do not alter the weights of each security within the risky portfolio, the probability
distribution of the rate of return on the risky portfolio remains unchanged by the asset reallocation
What will change is the probability distribution of the rate of return on the complete portfolio that
consists of the risky asset and the risk-free asset
The risk-free asset
There are no true risk-free assets
Only the government can issue default-free bonds
Even the default-free guarantee by itself is not sufficient to make the bonds risk-free in real terms
The only risk-free asset in real terms would be a perfectly price-indexed bond
Moreover, a default-free perfectly indexed bond offers a guaranteed real rate to an investor only if
the maturity of the bond is identical to the investors desired holding period
Even indexed bonds are subject to interest rate risk: Real interest rates change unpredictably
Nevertheless, it is common practice to view Treasury bills as the risk-free asset
Their short term nature makes their values insensitive to interest rate fluctuations
An investor can lock in a short-term nominal return by buying a bill and holding it to maturity
Inflation uncertainty over a few weeks/months  uncertainty of stock market returns
In practice, most investors use a broader range of money market instruments as a risk-free asset
All the money market instruments are virtually free of interest rate risk because of their short
maturities and are fairly safe in terms of default or credit risk
Most money market funds hold three types of securities: (i) T-bills, (ii) Bank certificates of deposit
(CDs), and (iii) Commercial paper (CP), differing slightly in their default risk
Portfolios of one risky asset and a risk-free asset
The concern is with the proportion of the investment budget y to be allocated to the risky portfolio P .
The remaining proportion 1 y is to be invested in the risk-free asset F
Denote the risky rate of return of P by rp , its expected rate of return by E(rp ) and its standard
deviation by p . The rate of return on the risk-free asset is denoted as rf
The risk premium on the risky asset is: E(rp ) rf
The rate of return on the complete portfolio C is rc = yrp + (1 y)rf
E(rc ) = rf + y[E(rp ) rf ]
Base rate of return = risk-free rate. The portfolio is also expected to earn a risk premium that
depends on risk premium of risky portfolio E(rp ) rf and investors position y in P
When combining risky and risk-free assets, standard deviation c of complete portfolio = standard
deviation p of risky asset multiplied by weight y of risky asset:
c = yp

(4)

Investment opportunity set with risky/risk-free asset in the E- plane

Equation for the straight line between F and P :
E(rc ) = rf + y[E(rp ) rf ] = rf +

c
[E(rp ) rf ]
p

(5)

E(r)

Capital Allocation Line (CAL)

E(rp)

S(y > 1)

rf

E(rp){rf

S(y < 1)

rB
f
F
Slope S =

E(rp) { rf
p

Capital Allocation Line (CAL)

with borrowing rate rB
f
p

Figure 2: The investment opportunity set in the expected return-standard deviation plane
Investment opportunity set
The set of feasible expected return and standard deviation pairs of portfolios resulting from
different values of y
The Capital Allocation Line (CAL) and the Sharpe ratio
The CAL depicts all the risk-return combinations available to investors
The slope S of the CAL equals the increase in the expected return of the complete portfolio
per unit of additional standard deviation, i.e. incremental return per incremental risk
The slope is called the reward-to-volatility ratio or the Sharpe ratio
S=

E(rp ) rf
p

(6)

If investors can borrow at rf , they can construct portfolios to the right of P on the CAL
However, non-government investors cannot borrow at the risk-free rate
Then in the borrowing range, the reward-to-volatility ratio (i.e. the slope of the CAL) will be lower
The CAL will therefore be kinked at point P
Risk tolerance and asset allocation
Investor confronting the CAL must choose one optimal portfolio C from set of feasible choices
This choice entails a trade-off between risk and return
Differences in risk aversion Different investors choose different positions in risky asset
Investors attempt to maximize utility by choosing the best allocation to the risky asset y
As allocation to risky asset increases (y %), expected return increases, but so does volatility
Solving the utility maximization problem:
max U = E(rc ) 21 Ac2 = rf + y[E(rp ) rf ] 21 Ay 2 p2
y

Setting the derivative of this expression to zero and solving for y yields the optimal position:
y? =

E(rp ) rf
Ap2

(7)

The optimal position in the risky asset is inversely proportional to the level of risk aversion and
the level of risk (variance) and directly proportional to the risk premium offered by the risky asset
Indifference curve analysis
First calculate the utility value of a risk-free portfolio yielding rf
Then, find the expected return the investor would require to maintain the same level of utility
when holding a risky portfolio for a given
This yields all combinations of expected return/volatility with a given constant utility level
Any investor prefers a portfolio on higher indifference curve (higher certainty equivalent)
Portfolios on higher indifference curves offer a higher return for any given level of risk
Higher indifference curves correspond to higher levels of utility
6

E(r)

Higher U

E(rc)

E(rp)

Complete portfolio maximizing U

U = rf

CAL
P

rf

Figure 3: Indifference curve analysis

More risk-averse investors have steeper indifference curves than less risk-averse investors
Steeper curves: Investors require greater increase in expected return for increase in risk
Investors thus attempt to find the complete portfolio on the highest possible indifference curve
Superimpose indifference curves on the investment opportunity set represented by the CAL as
in Fig. 3, and identify highest possible indifference curve that still touches CAL
That indifference curve is tangent to the CAL, and the tangency point corresponds to the
standard deviation and expected return of the optimal complete portfolio
The choice for y ? the fraction of overall investment funds to place in the risky portfolio versus the safer
but lower expected-return risk-free asset, is in large part a matter of risk aversion
Passive strategies: The capital market line
Passive strategy
Describes a portfolio decision that avoids any direct or indirect security analysis
Natural candidate for passively held risky asset is a well-diversified portfolio of common stocks
Because a passive strategy requires that we devote no resources to acquiring information on any
individual stock or group of stocks, we must follow a neutral diversification strategy
Select diversified stock portfolio that mirrors the value of the US corporate sector
This results in a portfolio in which, e.g., the proportion invested in Microsoft stock will be the ratio
of Microsofts total market value to the market value of all listed stocks
The Capital Market Line (CML)
Defined as the CAL provided by l-month T-bills and a broad index of common stocks
A passive strategy generates an investment opportunity set that is represented by the CML
How reasonable is it for an investor to pursue a passive strategy?
1. The alternative active strategy is not free, Whether you choose to invest the time and cost to
acquire the information needed to generate an optimal active portfolio of risky assets, or whether
you delegate the task to a professional who will charge a fee
2. Free-rider benefit: Another reason to pursue a passive strategy
If there are many active investors who quickly bid up prices of undervalued assets and force
down prices of overvalued assets, then at any time most assets will be fairly priced
Therefore, a well-diversified portfolio of common stock will be a reasonably fair buy, and the
passive strategy may not be inferior to that of the average active investor
To summarize, a passive strategy involves investment in two passive portfolios: (i) Virtually risk-free
short-term T-bills (or a money market fund) and (ii) A fund of common stocks that mimics a broad
market index. The capital allocation line representing such a strategy is called the capital market line
Criticisms of index funds dont hold up
They are undiversified: The same complaint could be leveled at actively managed funds
They are top-heavy: True, but S&P 500 not so narrow focused (77.2% of US stock-market value)
They are chasing performance: This is what all investors do
You can do better: As a group, investors in cant outperform market (collectively, they = market)

BKM - Ch. 7: Optimal risky portfolios

Introduction
The investment decision can be viewed as a top-down process:
1. Capital allocation between the risky portfolio and risk-free assets
2. Asset allocation across broad classes (US stocks, international stocks, long-term bonds)
3. Security selection of individual assets within each asset class
The optimal capital allocation is determined by risk aversion as well as expectations for the risk-return
trade-off of the optimal risky portfolio
In principle, asset allocation and security selection are technically identical:
Both aim at identifying that optimal risky portfolio, namely, the combination of risky assets that
provides the best risk-return trade-off
In practice, however, asset allocation and security selection are typically separated into two steps:
1. The broad outlines of the portfolio are established first (asset allocation)
2. Details concerning specific securities are filled in later (security selection)
Diversification and portfolio risk
When all risk is firm-specific, diversification can reduce risk to arbitrarily low levels
With all risk sources independent, exposure to any specific source of risk reduced to negligible level
The insurance principle: Reduction of risk to very low levels for independent risk sources
Risk eliminated by diversification: Unique/firm-specific/nonsystematic/diversifiable risk
When common sources of risk affect all firms, even extensive diversification cannot eliminate risk
The risk that remains even after extensive diversification is called market risk, risk that is attributable to marketwide risk sources (aka systematic risk, or nondiversifiable risk)
Portfolio of two risky assets
Efficient diversification: Constructing risky portfolios to provide the lowest possible risk for any given
level of expected return
Consider two mutual funds: A bond portfolio D, and a stock fund E
wD is invested in the bond fund, and the remainder 1 wD = wE is invested in the stock fund
Denoting rD , rE the rate of return on debt/equity funds, the rate of return rp on portfolio is:
rp = wD rD + wE rE

(1)

The expected return on the portfolio is a weighted average of expected returns on the component
securities with portfolio proportions as weights:
E(rp ) = wD E(rD ) + wE E(rE )

(2)

The variance of the two-asset portfolio is:

2 2
2 2
p2 = wD
D + wE
E + 2wD wE Cov(rD , rE )

(3)

Variance is reduced if the covariance term is negative

Even if covariance term 0, the portfolio standard deviation is less than weighted average of
individual security s, unless the two securities are perfectly positively correlated
The covariance can be computed from the correlation coefficient DE :
2 2
2 2
Cov(rD , rE ) = DE D E p2 = wD
D + wE
E + 2wD wE D E DE

A hedge asset has negative correlation with the other assets in the portfolio
Such assets will be particularly effective in reducing total risk
Expected return is unaffected by correlation between returns
Always prefer to add to portfolio assets with low or negative correlation with existing position

(4)

 How low can portfolio standard deviation be?

With perfect negative correlation, = 1:
p2 = (wD D wE E )2

(5)

When = 1, a perfectly hedged position can be obtained by choosing the portfolio proportions
to solve wD D wE E = 0. Then:
wD =

E
D + E

and

wE =

D
= 1 wD
D + E

(6)

Portfolio Standard Deviation

What happens when wD > 1 and wE < 0?

Strategy: Sell the equity fund short and invest proceeds of short sale in debt fund
This will decrease the expected return of the portfolio
The reverse happens when wD < 0 and wE > 1
Strategy: Sell bond fund short and use proceeds to buy more of equity fund
= {1
= 0
= :3
= 1
E

10

-.5

.5

Weight in Stock Fund

Figure 1: Portfolio standard deviation as a function of investment proportions

As the portfolio weight in the equity fund increases from zero to 1, portfolio standard deviation
first falls with the initial diversification from bonds into stocks, but then rises again as the portfolio
becomes heavily concentrated in stocks, and again is undiversified
Pattern generally holds as long as the correlation coefficient between funds is not too high
If < D /E , volatility initially falls when we start with all bonds and move into stocks
For a pair of assets with a large positive correlation of returns, the portfolio standard deviation
will increase monotonically from the low-risk asset to the high-risk asset. Even in this case,
however, there is a positive (if small) value from diversification
What is the minimum level to which portfolio standard deviation can be held?
The portfolio weights that solve this minimization problem turn out to be:
wmin (D) =

2
D

2 Cov(r , r )
E
D E
2 2Cov(r , r )
+ E
D E

The minimum-variance portfolio has a standard deviation smaller than that of either of the individual component assets Effect of diversification
Portfolio opportunity set
Pair of investment weights (wD , wE ) Resulting pair of expected return/standard deviation
These constitute the portfolio opportunity set that can be constructed from the two available assets
Fig. 2 shows the portfolio opportunity set for other values of the correlation coefficient
The solid black line connecting the two funds shows that there is no benefit from diversification
when the correlation between the two is perfectly positive ( = 1)
The dashed colored line demonstrates the greater benefit from diversification when the correlation coefficient is lower than .30
10

Expected Return (%)

= {1
= 1

10

= :3

= 0
9
8
D
10

12

14

16

Standard Deviation (%)

Figure 2: Portfolio expected return as a function of standard deviation

For = 1, the portfolio opportunity set is linear, but now it offers a perfect hedging opportunity and the maximum advantage from diversification
The lower the correlation, the greater the potential benefit from diversification
Suppose now an investor wishes to select the optimal portfolio from the opportunity set
The best portfolio will depend on risk aversion
Portfolios to the northeast in Fig. 2 provide higher rates of return but impose greater risk
The best trade-off among these choices is a matter or personal preference
Investors with greater risk aversion prefer southwest portfolios (lower expected return, lower risk)
Given a level of risk aversion, determine the portfolio that provides highest level of utility
Using U = E(rp ) 12 Ap2 and the portfolio mean/variance determined by the portfolio weights
in the two funds wE and wD , the optimal investment proportions in the two funds are:
wD =

2
E(rD ) E(rE ) + A(E
D E DE )
2
2
A(D + E 2D E DE )

and

wE = 1 wD

Expected Return (%)

Asset allocation with stocks, bonds, and bills

The optimal risky portfolio with two risky assets and a risk-free asset
Graphical solution
Ratchet the CAL upward until it reaches point of tangency with investment opportunity set
This must yield the CAL with the highest feasible reward-to-volatility ratio
Thus, the tangency portfolio (P in Fig. 3) is the optimal risky portfolio to mix with T-bills
We can read the expected return and standard deviation of portfolio P from the graph
CAL(E)
CAL(P)

Indifference Curve

Optimal
14 Complete
Portfolio
12
10
8
6
r = 5%
4 f
2
0

E
P

Efficient Frontier

Opportunity Set of Risky Assets

Minimum-Variance Frontier

Global M-V
Portfolio
10

15

20

Standard Deviation (%)

Figure 3: The opportunity set of the debt and equity funds with the CAL
Portfolio construction with only two risky assets and a risk-free asset
The objective is to find the weights wD and wE that result in the highest slope of the CAL
(i.e., the weights that result in the risky portfolio with the highest reward-to-volatility ratio)
11

 Thus our objective function is the slope (equivalently, the Sharpe ratio) Sp :
Sp =

E(rp ) rf
p

For portfolio with two risky assets, expected return and standard deviation of portfolio P are:
E(rp ) = wD E(rD ) + wE E(rE )
2 2
2 2
p = [wD
D + wE
E + 2wD wE Cov(rD , rE )]1/2

Therefore, we solve an optimization problem formally written as (subject to

max Sp =
wi

wi = 1):

E(rp ) rf
p

In the case of two risky assets, the solution for the weights of the optimal risky portfolio P ,
using excess rates of return R rather than total returns r, is:
wD =

2 E(R )Cov(R , R )
E(RD )E
E
D
E
and wE = 1 wD
2
2
E(RD )E + E(RE )D [E(RD ) + E(RE )]Cov(RD , RE )

(7)

The steps to arrive at the complete portfolio are:

1. Specify the return characteristics of all securities (expected returns, variances, covariances)
2. Establish the risky portfolio:
(a) Calculate the optimal risky portfolio P [Eq. (7)]
(b) Calculate properties of P using weights determined in step (a) and Eqs. (2) and (3)
3. Allocate funds between the risky portfolio and the risk-free asset:
(a) Calculate the fraction y of the complete portfolio allocated to portfolio P (the risky portfolio) and to T -bills (the risk-free asset)
y=

E(rp ) rf
Ap2

(8)

(b) Calculate the share of the complete portfolio invested in each asset and in T-bills
Our two risky assets, the bond and stock mutual funds, are already diversified portfolios. The
diversification within each of these portfolios must be credited for a good deal of the risk reduction
compared to undiversified single securities
Optimizing the asset allocation between bonds and stocks contributed incrementally to the improvement in the reward-to-volatility ratio of the complete portfolio
The CAL with stocks, bonds, and bills shows that the standard deviation of the complete portfolio
can be further reduced while maintaining the same expected return as the stock portfolio
The Markowitz portfolio selection model
Generalizing the portfolio construction problem to the case of many risky securities and a risk-free asset
1. Identify the risk-return combinations available from the set of risky assets
2. Identify optimal portfolio of risky assets by finding portfolio weights resulting in steepest CAL
3. Choose appropriate complete portfolio by mixing risk-free asset with optimal risky portfolio
Security selection
In the risk-return analysis, the portfolio manager needs as inputs a set of estimates for the expected
returns of each security and a set of estimates for the covariance matrix
Hence, we have n estimates of E(ri ) and the n n estimates of the covariance matrix in which
the n diagonal elements are estimates of the variances i2 and the n2 n = n(n 1) off-diagonal
elements are the estimates of the covariances between each pair of asset returns
12

 The expected return/variance of any risky portfolio with weights in each security wi is:
E(rp ) =
p2 =

n
X

wi E(ri )
i=1
n X
n
X

(9)

wi wj Cov(ri , rj )

(10)

i=1 j=1

Markowitz model is precisely step one of portfolio management: The identification of the efficient
set of portfolios, or the efficient frontier of risky assets
M-V frontier: Graph of the lowest possible variance for a given portfolio expected return
All individual assets lie to the right inside the frontier, at least when short sales are allowed.
When short sales are prohibited, single securities may lie on the frontier
The part of the frontier above the global M-V portfolio is the efficient frontier
The principal idea behind the frontier set of risky portfolios is that, for any risk level, we are
interested only in that portfolio with the highest expected return
The frontier is the set of portfolios that minimizes variance for any target expected return
Some clients may be subject to additional constraints. E.g., prohibited from taking short positions
For these clients the portfolio manager will add to the optimization program constraints that
rule out negative (short) positions in the search for efficient portfolios
In this special case, single assets may be, in and of themselves, efficient risky portfolios
E.g., asset with highest expected return is a frontier portfolio because, without short sales, the
only way to obtain that rate of return is to hold the asset as ones entire risky portfolio
Some may want to ensure minimal level of expected dividend yield from optimal portfolio
In this case the input list will be expanded to include a set of expected dividend yields d1 , , dn
and the optimization program will include an additional constraint that ensures that the expected dividend yield of the portfolio will equal or exceed the desired level d
Any constraint carries a price tag in the sense that an efficient frontier constructed subject to extra
constraints will offer a reward-to-volatility ratio inferior to that of a less constrained one
Another type of constraint is aimed at ruling out investments in industries or countries considered
ethically or politically undesirable. This is referred to as socially responsible investing
Capital allocation and the separation property
We ratchet up the CAL by selecting different portfolios until we reach portfolio P which is the
tangency point of a line from F to the efficient frontier
Portfolio P maximizes the reward-to-volatility ratio and is the optimal risky portfolio
The most striking conclusion is that a portfolio manager will offer the same risky portfolio P to all
clients regardless of their degree of risk aversion
The degree of risk aversion comes into play only in the selection of desired point along CAL
More risk-averse clients invest more in risk-free asset and less in optimal risky portfolio than
less risk-averse clients. However, both use portfolio P as their optimal risky investment vehicle
Separation property: Portfolio choice problem may be separated into 2 independent tasks:
1. Determination of the optimal risky portfolio (purely technical)
2. Allocation of the complete portfolio to T-bills vs. the risky portfolio (personal preference)
In practice, different managers estimate different input lists, thus deriving different efficient frontiers, and offer different optimal portfolios Source of disparity lies in security analysis
This analysis suggests that a limited number of portfolios may be sufficient to serve the demands
of a wide range of investors Theoretical basis of the mutual fund industry
The power of diversification
Consider the naive diversification strategy in which an equally weighted portfolio is constructed
wi = 1/n for each security. In this case Eq. (10) becomes:
p2

n
n
n
X
X
1X1 2
1
=
i +
Cov(ri , rj )
n
n
n2
i=1

j=1,j6=i i=1

13

(11)

 Define the average variance and average covariance of the securities as:
n

1X 2

=
i
n
2

and

i=1

n
n
X
X
1
Cov =
Cov(ri , rj )
n(n 1)

(12)

j=1,j6=i i=1

Then, the portfolio variance is:

p2 =

1 2 n1
Cov

+
n
n

(13)

Effect of diversification
When the average covariance among security returns is zero, as it is when all risk is firm-specific,
portfolio variance can be driven to zero
Hence when security returns are uncorrelated, the power of diversification is unlimited
However, usually, economy-wide risk factors impart positive correlation among stocks
The irreducible risk of a diversified portfolio depends on covariance of the returns of component securities, which is a function of the importance of systematic economic factors
Suppose that all securities have a common and all security pairs have a common
The covariance between all pairs of securities is 2 and Eq. (13) becomes:
p2 =

1 2 n1 2
+

n
n

(14)

When = 0, we again obtain the insurance principle: p 0 as n

For > 0, however, portfolio variance remains positive
For = 1, portfolio variance equals 2 regardless of n Diversification is of no benefit
In the case of perfect correlation, all risk is systematic
More generally, as n becomes greater, Eq. (14) shows that systematic risk becomes 2
For diversified portfolios, the contribution to portfolio risk of a particular security depends on the
covariance of that securitys return with other securities, and not on the securitys variance
Asset allocation and security selection
Why distinguish between asset allocation and security selection? 3 reasons:
1. As a result of greater need and ability to save (for college, recreation, longer life, health care
needs, etc.), the demand for sophisticated investment management has increased enormously
2. The widening spectrum of financial markets and financial instruments has put sophisticated
investment beyond the capacity of many amateur investors
3. There are strong economies of scale in investment analysis
The end result is that the size of a competitive investment company has grown with the industry,
and efficiency in organization has become an important issue
The practice is therefore to optimize the security selection of each asset-class portfolio independently
At the same time, top management continually updates the asset allocation of the organization,
adjusting the investment budget allotted to each asset-class portfolio
Risk pooling, risk sharing, and risk in the long run
Risk pooling and the insurance principle
The insurance principle
Suppose an insurer sells 10,000 uncorrelated policies, each with a standard deviation
Because the covariance between any two insurance policies is zero and is the same for each
policy, the standard deviation of the rate of return on the 10,000-policy portfolio is:
p2 =

1 2

p =
n
n

(15)

p could be further decreased by selling even more policies This is the insurance principle

14

 It seems that as the film sells more policies, its risk continues to fall. Flaw in this argument:
Probability of loss = inadequate measure of risk: Does not account for the magnitude of loss
If 10,000 policies are sold, maximum possible loss is 10,000 bigger and comparison with a
one-policy portfolio cannot be made based on means/standard deviations of rates of return
Similar flaw as the argument that investing in stocks for the long run reduces risk
Increasing the size of the bundle of policies does not make for diversification! Diversifying a portfolio
means dividing a fixed investment budget across more assets
When we combine n uncorrelated insurance policies, each with an expected profit $, both expected
total profit and standard deviation (SD) grow in direct proportion to n:
E(n) = nE()
V ar(n) = n2 V ar() = n2 2
SD(n) = n
The ratio of mean to standard deviation does not change when n increases
The risk-return trade-off therefore does not improve with the assumption of additional policies
Risk sharing
If risk pooling (sale of additional independent policies) does not explain insurance, what does?
The answer is risk sharing, the distribution of a fixed amount of risk among many investors
An underwriter will contact other underwriters who each will take a piece of the action: Each will
choose to insure a fraction of the project risk
Each underwriter has a fixed amount of equity capital. Underwriters engage in risk sharing. They
limit their exposure to any single source of risk by sharing that risk with other underwriters
Underwriter diversifies its risk by allocating its budget across many projects that are not perfectly
correlated One underwriter will decline to u/w too large a fraction of any single project
This is the proper use of risk pooling: Pooling many sources of risk in a portfolio of given size
The bottom line is that portfolio risk management is about the allocation of a fixed investment budget
to assets that are not perfectly correlated
In this environment, rate of return statistics (expected returns, variances, and covariances) are
sufficient to optimize the investment portfolio
Choices among alternative investments of a different magnitude require that we abandon rates of
return in favor of dollar profits
This applies as well to investments for the long run

15

16

BKM - Ch. 8: Index models

Introduction
The Markowitz procedure suffers from two drawbacks:
1. The model requires a huge number of estimates to fill the covariance matrix
2. The model does not provide any guideline to the forecasting of the security risk premiums
Index models simplify estimation of covariance matrix and enhance analysis of risk premiums
By allowing to explicitly decompose risk into systematic and firm-specific components, these models
also shed considerable light on both the power and limits of diversification
Further, they allow to measure these components of risk for particular securities and portfolios
Despite simplification, index models remain true to concepts of efficient frontier/portfolio optimization
Empirically, index models are as valid as the assumption of normality of rates of return on securities
A single-factor security market
The input list of the Markowitz model
The success of a portfolio selection rule depends on the quality of the input list
The Markowitz model necessitates n expected returns, n variances, and n(n 1)/2 covariances
Markowitz model: Errors in estimation of correlation coefficients can lead to nonsensical results
This can happen because some sets of correlation coefficients are mutually inconsistent
Introducing a model that simplifies the way we describe the sources of security risk allows us to
use a smaller, consistent set of estimates of risk parameters and risk premiums
The simplification emerges because positive covariances among security returns arise from
common economic forces that affect the fortunes of most firms
Examples of common economic factors: Business cycles, interest rates, cost of natural resources
The unexpected changes in these variables cause, simultaneously, unexpected changes in the
rates of return on the entire stock market
By decomposing uncertainty into these system-wide versus firm-specific sources, we vastly simplify
the problem of estimating covariance and correlation
Normality of returns and systematic risk
Decompose rate of return on security i into its expected plus unanticipated components:
ri = E(ri ) + ei

(1)

Where the unexpected return ei has a mean of zero and a standard deviation of i
When security returns can be well approximated by normal distributions that are correlated across
securities, we say that they are joint normally distributed
At any time, security returns are driven by one or more common variables
Suppose the common factor m that drives innovations in security returns is some macroeconomic
variable that affects all firms Decompose sources of uncertainty into uncertainty about economy
as a whole (captured by m) and uncertainty about firm in particular (captured by ei )
ri = E(ri ) + m + ei

(2)

The macroeconomic factor m measures unanticipated macro surprises

m has a mean of zero (over time, surprises average out) with standard deviation of m
m and ei are uncorrelated, because ei is firm-specific Independent of shocks to the common
factor that affect the entire economy
The variance of ri thus arises from two uncorrelated sources, systematic and firm specific
2
i2 = m
+ 2 (ei )

(3)

The common factor m generates correlation across securities

Because all securities will respond to the same macroeconomic news
17

 Since m is uncorrelated with any firm-specific surprises, the covariance between any i and j is:
2
Cov(ri , rj ) = Cov(m + ei , m + ej ) = m

(4)

We recognize that some securities are more sensitive than others to macroeconomic shocks:
Capture this refinement by assigning each firm a sensitivity coefficient to macro conditions
This leads to the single-factor model:
ri = E(ri ) + i m + ei

(5)

2 , and its total risk is:

The systematic risk of security i is i2 m
2
i2 = i2 m
+ 2 (ei )

(6)

The covariance between any pair of securities also is determined by their betas:
2
Cov(ri , rj ) = Cov(i m + ei , j m + ej ) = i j m

(7)

Normality of security returns alone guarantees that portfolio returns are also normal and that there
is a linear relationship between security returns and the common factor
Seek a variable that can proxy for common factor. To be useful, variable must be observable, so
we can estimate its volatility/sensitivity of individual securities returns to variation in its value
The single-index model
A reasonable approach is to assert that the rate of return on a broad index of securities such as the S&P
500 is a valid proxy for the common macroeconomic factor
Single-index model single-factor model because market index = proxy for common factor
The regression equation of the single-index model
Denote the market index by M , with excess return of RM = rM rf and standard deviation of M
We regress the excess return of a security Ri = ri rf on the excess return of the index RM
We collect a historical sample of paired observations Ri (t) and RM (t) where t denotes the date of
each pair of observations. The regression equation is:
Ri (t) = i + i RM (t) + ei (t)

(8)

The intercept i is the securitys expected excess return when the market excess return is zero
The slope coefficient i is the securitys sensitivity to the index
ei is the zero-mean, firm-specific surprise in the security return in time t, aka the residual
The expected return-beta relationship
Taking expected values, we obtain the expected return-beta relationship of the single-index model:
E(Ri ) = i + i E(RM )

(9)

Part of a securitys risk premium is due to the risk premium of the index
The market risk premium is multiplied by the relative sensitivity of the individual security
This is the systematic risk premium because it derives from the risk premium that characterizes
the entire market, which proxies for the condition of the full economy or economic system
is the non-market premium
Risk and covariance in the single-index model
Both variances/covariances are determined by security betas/properties of market index:
Total risk = Systematic risk + Firm-specific risk
2
i2 = i2 M
+ 2 (ei )

(10)

Covariance = Product of betas Market index risk

2
Cov(ri , rj ) = i j M

(11)

Correlation = Product of correlations with the market index

Corr(ri , rj ) =

2
i j M
= Corr(ri , rM ) Corr(rj , rM )
i j

18

(12)

 The set of parameter estimates needed for the single-index model consists of only , , and (e)
for the individual securities, plus the risk premium and variance of the market index
The set of estimates needed for the single-index model
1. i : Stocks expected return if the market is neutral, i.e. rM rf = 0
2. i (rM rf ): The component of return due to movements in the overall market
3. ei : The unexpected component of return due to unexpected firm specific events
2 : The variance attributable to the uncertainty of the common macroeconomic factor
4. i2 M
2
5. (ei ): The variance attributable to firm-specific uncertainty
Advantages of the model
The index model is a very useful abstraction because for large universes of securities, the
number of estimates is only a small fraction of what otherwise would be needed
The index model abstraction is crucial for specialization of effort in security analysis: If a
covariance term had to be calculated for each security pair, then no specialization by industry
Disadvantages
Cost of the model: Restrictions it places on structure of asset return uncertainty
Classification of uncertainty into dichotomy - macro vs. micro risk - oversimplifies sources of
real-world uncertainty and misses important sources of dependence in stock returns
E.g., dichotomy rules out industry events (affect an industry but not the broad macroeconomy)
The optimal portfolio derived from the single-index model therefore can be significantly inferior
to that of the full-covariance (Markowitz) model when stocks with correlated residuals have large
alpha values and account for a large fraction of the Portfolio
If many pairs of the covered stocks exhibit residual correlation, it is possible that a multiindex model, which includes additional factors to capture those extra sources of cross-security
correlation, would be better suited for portfolio analysis and construction
The index model and diversification
Suppose that we choose an equally weighted portfolio of n securities. The excess rate of return on
each security is given by: Ri = i + i RM + ei
Similarly, we can write the excess return on the portfolio of stocks as: Rp = p + p RM + ep
As the number of stocks included in this portfolio increases, the part of the portfolio risk attributable
to nonmarket factors becomes ever smaller: This part of the risk is diversified away
In contrast, market risk remains, regardless of the number of firms combined into the portfolio
!
n
n
n
n
n
X
1X
1X
1X
1X
Rp =
wi Ri =
(i + i RM + ei ) =
i +
i RM +
ei
(13)
n
n
n
n
i=1

i=1

i=1

i=1

i=1

Portfolios sensitivity to market, nonmarket return and avg. of firm-specific components:

n

p =

1X
i
n

p =

i=1

1X
i
n

ep =

i=1

1X
ei
n

(14)

i=1

Hence the portfolios variance is:

2
p2 = p2 M
+ 2 (ep )

(15)

The systematic risk component of the portfolio variance, which depends on marketwide movements,
2 and depends on the sensitivity coefficients of the individual securities. This part of the
is p2 M
risk will persist regardless of the extent of portfolio diversification
The nonsystematic component of the variance is 2 (ep ) and is attributable to firm-specific ei
Because ei s are independent and have zero expected value, as more stocks are added to the
portfolio, firm-specific components cancel out, resulting in ever-smaller nonmarket risk
Such risk is thus termed diversifiable: Because the ei s are uncorrelated,
n  2
X
1
1 2
2
(ep ) =
2 (ei ) =
(e)
(16)
n
n
i=1

19

Estimating the single-index model

The security characteristic line for stock i
Regression: Line with intercept i , slope i = security characteristic line (SCL) for stock i:
Ri (t) = i + i RS&P 500 (t) + ei (t)
The explanatory power of the SCL for stock i
The R-square tells us the percentage of the variation in the stock i series that is explained by the
variation in the S&P 500 excess returns
The adjusted R2 (slightly smaller) corrects for upward bias in R2 that arises because we use fitted
values of parameters (slope and intercept ) rather than true, but unobservable values
2 ) is derived from the unadjusted by:
In general, the adjusted R-square (RA
(n1)
2 = 1 (1 R2 )
RA
where k is the number of independent variables
(nk1)
An additional degree of freedom is lost to the estimate of the intercept
Analysis of variance
The sum of squares (SS) is the portion of the variance of the dependent variable (stock is return)
2
that is explained by the independent variable (S&P 500 return); it is equal to i2 S&P
500
The MS column for the residual shows the variance of the unexplained portion of stock is return,
i.e. the portion of return that is independent of the market index. The square root of this value is
the standard error (SE) of the regression
Dividing the total SS of the regression by the number of dof (59 for 60 observations), we obtain
the estimate of the variance of the dependent variable (stock i), per month
R2 (ratio of explained/total variance) = [explained (regression) SS]/[total SS]
R2 =

2
i2 S&P
2 (ei )
500
=
1

2
2
2
2
i2 S&P
i2 S&P
500 + (ei )
500 + (ei )

The estimate of alpha

The intercept is the estimate of stock is alpha for the sample period (per month)
The standard error of the estimate is a measure of the imprecision of the estimate. If the standard
error is large, the range of likely estimation error is correspondingly large
We can relate the standard error of to the standard error of the residuals as:
s
(Avg. S&P 500)2
1
SE(i ) = (ei )
+
n V ar(S&P 500) (n 1)
The t-statistic is the ratio of the regression parameter to its standard error
This statistic equals the number of standard errors by which our estimate exceeds zero, and
therefore can be used to assess the likelihood that the true but unobserved value might actually
equal zero rather than the estimate derived from the data
For , we are interested in avg. value of stock is return net of market movements
Define the nonmarket component of return as actual return minus the return attributable to
market movements. Call this stock is firm-specific return Rf s = Ri i RS&P 500
If Rf s were normally distributed with a mean of zero, the ratio of its estimate to its standard
error would have a t-distribution
From a table of the t-distribution, we can find the probability that the true is actually zero
This is called the level of significance or the probability or p-value. Conventional cut-off for
statistical significance is a probability of less than 5%, which requires t-statistic 2.0
Even if was both economically and statistically significant within the sample, we still would not
use that as a forecast for a future period
Overwhelming empirical evidence shows that 5-year alpha values do not persist over time
Virtually no correlation between estimates from one sample period to the next

20

 The estimate of beta

When the value of beta and its SE produce a large t-statistic and a p-value of practically zero, we
can confidently reject the hypothesis that stock is true beta is zero
SE(i ) =

(e )
i
i n 1

A more interesting t-statistic might test a null hypothesis that i is greater than the market-wide
average beta of 1. This t-statistic would measure how many standard errors separate the estimated
beta from a hypothesized value of 1:
t-statistics =

Estimated value Hypothesized value

Standard error

Firm-specific risk

The annual standard deviation of stock is residual is 12 times its monthly standard deviation
The standard deviation of systematic risk is (S&P 500)
It is common for individual stocks to have a firm-specific risk as large as its systematic risk
Correlation and covariance matrix
2 + 2 (e )
The variance estimates for the individual stocks equal i2 M
i
2
The off-diagonal terms are covariance values and equal i j M
Portfolio construction and the single-index model
Alpha and security analysis
Most important advantage of single-index model: Framework it provides for macroeconomic/security
analysis in preparing input list critical to efficiency of the optimal portfolio
The single-index model separates macroeconomic and individual-firm sources of return variation
and makes it easier to ensure consistency across analysts
Hierarchy of the preparation of the input list using the single-index model:
1. Macroeconomic analysis is used to estimate the risk premium and risk of the market index
2. Statistical analysis used to estimate s of all securities and their residual variances 2 (ei )
3. The portfolio manager uses the estimates for the market-index risk premium and the beta
coefficient of a security to establish the expected return of that security absent any contribution
from security analysis. This market-driven expected return can be used as a benchmark
4. Security-specific expected return forecasts (s) are derived from security-valuation models
The alpha value distills the incremental risk premium attributable to private information
developed from security analysis
is more than just one of the components of expected return: It is the key variable that tells us
whether a security is a good or a bad buy
The index portfolio as an investment asset
To avoid inadequate diversification, include the S&P 500 portfolio as one of the assets
S&P 500 passive portfolio that the manager would select in the absence of security analysis
The single-index model input list
If the portfolio manager plans to compile a portfolio from a list of n actively researched firms and
a passive market index portfolio, the input list will include the following estimates:
1. Risk premium on the S&P 500 portfolio
2. Estimate of the standard deviation of the S&P 500 portfolio
3. n sets of estimates of (i) coefficients, (ii) Stock residual variances, and (iii) values
The optimal risky portfolio of the single-index model
With the estimates of and coefficients, plus the risk premium of the index portfolio, we can
generate the n + 1 expected returns using Eq. (9)
With the estimates of coefficients and residual variances, together with the variance of the index
portfolio, we can construct the covariance matrix using Eq. (10)
Given risk premiums and the covariance matrix, we can conduct the optimization program
21

 The , , and residual variance of a weighted portfolio are the simple averages of those parameters
across component securities (where the index n + 1 corresponds to the market index: n+1 = M =
0, n+1 = M = 1, and (en+1 ) = (eM ) = 0):
p =

n+1
X

wi i

p =

i=1

n+1
X

wi i

(ep ) =

i=1

n+1
X

wi2 2 (ei )

(17)

i=1

The objective is to maximize the Sharpe ratio of the portfolio by using portfolio weights w1 , , wn+1 .
With this set of weights, the portfolio expected return, standard deviation, and Sharpe ratio are:
E(Rp ) = p + E(RM )p =

n+1
X

wi i + E(RM )

n+1
X

i=1

2
2
p = [p2 M
+ 2 (ep )]1/2 = M

n+1
X
i=1

Sp =

wi i

(18)

i=1

!2
wi i

n+1
X

1/2
wi2 2 (ei )

(19)

i=1

E(Rp )
p

(20)

The optimal risky portfolio trades off search for against departure from efficient diversification
The optimal risky portfolio turns out to be a combination of two component portfolios:
1. An active portfolio A comprised of the n analyzed securities
2. The market-index passive portfolio is the (n + l)-th asset included to aid in diversification
Assume first that the active portfolio has a beta of 1:
In that case, the optimal weight in active portfolio is proportional to ratio A / 2 (eA )
This ratio balances the contribution of the active portfolio (its alpha) against its contribution
to the portfolio variance (residual variance)
2 and hence the initial position in the
The analogous ratio for the index portfolio is E(RM )/M
active portfolio (i.e., if its beta were 1) is:
0
wA
=

A / 2 (eA )
2
E(RM )/M

(21)

Next, we amend this position to account for the actual beta of the active portfolio
2 , the correlation between the active and passive portfolios is greater when
For any level of A
the beta of the active portfolio is higher
This implies less diversification benefit from the passive portfolio and a lower position in it.
Correspondingly, the position in the active portfolio increases:

wA
=

0
wA
0
1 + (1 A )wA

(22)

The information ratio

Sharpe ratio of optimally constructed risky portfolio > index portfolio (passive strategy):


A 2
2
2
Sp = SM +
(eA )

(23)

The information ratio

) to the Sharpe
The contribution of the active portfolio (when held in its optimal weight wA
ratio of the overall risky portfolio is determined by the ratio of its alpha to its residual standard
deviation, aka the information ratio
Ratio measures extra return we can obtain from security analysis compared to firm-specific
risk we incur when we over-/underweight securities relative to passive market index
22

 To maximize the overall Sharpe ratio, maximize the information ratio of active portfolio
Information ratio maximized if we invest in each security its ratio of i / 2 (ei )
, weight in security i is:
Rescaling so that total position in active portfolio = wA
, n
X i
i

wi = wA 2
(24)
(ei )
2 (ei )
i=1

With this set of weights, the contribution of each security to the information ratio of the active
portfolio depends on its own information ratio:



n 
A 2 X i 2
=
(25)
(eA )
(ei )
i=1

In contrast to alpha, the market (systematic) component of the risk premium i E(RM ) is offset by
2 and both are driven by the same beta
the securitys nondiversifiable (market) risk i2 M
The beta of a security is neither vice nor virtue:
It is a property that simultaneously affects the risk and risk premium of a security
Only the aggregate of the active portfolio, rather than each individual securitys matters
The index portfolio is an efficient portfolio only if all alpha values are zero
Unless a security has 6= 0, including it in active portfolio makes portfolio less attractive
In addition to the securitys systematic risk, which is compensated for by the market risk
premium (through beta), the security would add its firm-specific risk to portfolio variance
If all securities have zero alphas, the optimal weight in the active portfolio will be zero, and
the weight in the index portfolio will be 1
However, when security analysis uncovers securities with nonmarket risk premiums (nonzero
alphas), the index portfolio is no longer efficient
Summary of the optimization procedure
1. Compute the initial position ofPeach security in the active portfolio as wi0 = i / 2 P
(ei )
0/
2. Scale initial positions to force
weights = 1 by dividing
by
their
sum:
w
=
w
wi0
i
i
P
3. Compute the alpha of the active portfolio: A =
wi i
P
4. Compute the residual variance of the active portfolio: 2 (eA ) = wi2 2 (ei )
0
2
2
5. Compute the initial position in the active portfolio:
P wA = [A / (eA )]/[E(RM )/M ]
6. Compute the beta of the active portfolio: A = wi i
= w 0 /[1 + (1 )w 0 ]
7. Adjust the initial position in the active portfolio: wA
A
A
A
= 1 w and w = w w
8. Note: The optimal risky portfolio now has weights: wM
i
A
A i
9. Calculate the risk premium of the optimal risky portfolio from the risk premium of the index
+ w )E(R ) + w
portfolio and the alpha of the active portfolio: E(Rp ) = (wM
M
A A
A A
10. Compute the variance of the optimal risky portfolio from the variance of the index portfolio and
+ w )2 2 + [w (e )]2
the residual variance of the active portfolio: p2 = (wM
A
A A
M
A
Practical aspects of portfolio management with the index model
Is the index model inferior to the full-covariance model?
A parsimonious model that is stingy about inclusion of independent variables is often superior:
Predicting the value of the dependent variable depends on two factors: (i) The precision of
the coefficient estimates and (ii) The precision of the forecasts of the independent variables
When we add variables, we introduce errors on both counts
Markowitz model more flexible in for asset covariance structure compared to single-index model
Advantage illusory if we cant estimate covariances with any degree of confidence
Using full-covariance matrix invokes estimation of thousands of terms Possible that cumulative effect of so many estimation errors results in an inferior than the single-index model
Advantages of the single-index framework:
Clear practical advantage
Decentralizes macro and security analysis
23

 The industry version of the index model

Merrill Lynch publishes a monthly Security Risk Evaluation book (aka beta book) which uses
the S&P 500 as the proxy for the market portfolio
It relies on the 60 most recent monthly observations to calculate regression parameters
Uses total returns, rather than excess returns, in the regressions Variant of index-model:
r = a + brM + e

instead of

r rf = + (rM rf ) + e

(26)

If rf is constant, both equations have same independent variable rm and residual e

The slope coefficient will be the same in the two regressions
However, the intercept that Merrill Lynch calls alpha is really an estimate of + rf (1 )
Merrill Lynch departs from the index model is in its use of percentage changes in price instead
of total rates of return Ignores the dividend component of stock returns
For most firms, R2  0.5, indicating that stocks have far more firm-specific than systematic risk
Highlights the practical importance of diversification
Adjusted beta
Motivation for adjusting estimates: On average, s of stocks move toward 1 over time
As it grows, a firm diversifies, expanding to similar products and later to more diverse operations. As firm becomes more conventional, it resembles rest of economy even more
Its beta coefficient will tend to change in the direction of 1
Another explanation for this phenomenon is statistical:
When we estimate this beta coefficient over a particular sample period, we sustain some
unknown sampling error of the estimated beta
The greater the difference between our estimate and 1, the greater the chance that we
had a large estimation error and that in a later sample period will be closer to 1
Merrill Lynch adjusts beta estimates in a simple way:
Adjusted beta = 23 Sample beta + 13
estimates are ex post (after the fact) measures. They do not mean that anyone could have
forecast these alpha values ex ante (before the fact)
Given the R2 of the regression and the residual standard deviation of the stock (ei ), we can solve
for the total standard deviation i of stock i:
r
2 (ei )
i =
1 R2

i = stock is monthly std. dev. for sample period Annualized std. dev. is 12i
Predicting betas
Simple approach: Collect data on in different periods and then estimate a regression:
Current beta = a + b(Past beta)

(27)

Given estimates of a and b, we would then forecast future betas using:

Forecast beta = a + b(Current beta)

(28)

If belief that firm size/debt ratios are two determinants of , an expanded version of Eq. (27) is:
Current beta = a + b1 (Past beta) + b2 (Firm size) + b3 (Debt ratio)
Variables to help predict betas:
1. Variance of earnings
2. Variance of cash flow

3. Growth in earnings per share

4. Market capitalization (firm size)

5. Dividend yield
6. Debt-to-asset ratio

Even after controlling for a firms financial characteristics, industry group helps to predict
24

 Index models and tracking portfolios

Suppose a portfolio manager believes she has identified an underpriced portfolio. The index model
equation for this portfolio is:
Rp = .04 + 1.4RS&P 500 + ep

(29)

Manager confident in quality of security analysis but wary about near term performance of broad
market. Wants a position that takes advantage of teams analysis but is independent of overall
market performance To this end, a tracking portfolio (T ) can be constructed
Tracking portfolio
A tracking portfolio for P is designed to match the systematic component of P s return
The idea is for the portfolio to track the market-sensitive component of P s return
This means the tracking portfolio must have the same beta on the index portfolio as P and as
little nonsystematic risk as possible. This procedure is also called beta capture
A tracking portfolio for P will have a levered position in the S&P 500 to achieve a beta of 1.4
Therefore, T includes positions of 1.4 in the S&P 500 and -0.4 in T-bills
Because T is constructed from the index and bills, it has an alpha value of zero
Now buy P but offset systematic risk by assuming a short position in tracking portfolio
The short position in T cancels out the systematic exposure of the long position in P
The overall combined position is thus market neutral
Therefore, even if the market does poorly, the combined position should not be affected
But the alpha on portfolio P will remain intact
The combined portfolio C provides an excess return per dollar of:
RC = Rp RT = (.04 + 1.4RS&P 500 + ep ) 1.4RS&P 500 = .04 + ep

(30)

While this portfolio is still risky (residual risk ep ), the systematic risk has been eliminated, and if
P is reasonably well-diversified, the remaining nonsystematic risk will be small
Manager can take advantage of the 4% without inadvertently taking on market exposure
Process of separating the search for from the choice of market exposure is called transport
This long-short strategy is characteristic of the activity of many hedge funds

25

26

BKM - Ch. 9: The Capital Asset Pricing Model (CAPM)

Introduction
CAPM gives precise prediction of relationship between risk of an asset and its expected return
CAPM serves two vital functions:
1. It provides a benchmark rate of return for evaluating possible investments
2. Helps make a guess as to the expected return on assets not yet traded in the marketplace
Although the CAPM does not fully withstand empirical tests, it is widely used because of the insight it
offers and because its accuracy is deemed acceptable for important applications
The Capital Asset Pricing Model
Simplifying assumptions that lead to the basic version of the CAPM:
1. There are many investors, each with a wealth that is small compared to the total endowment of
all investors. Investors act as though security prices are unaffected by their own trades
2. All investors plan for one identical holding period. This behavior is myopic: It ignores everything
that might happen after the end of single-period horizon. Myopic behavior is suboptimal
3. Investments limited to publicly traded financial assets and to risk-free borrowing/lending
Rules out investment in nontraded assets such as human capital, private enterprises, and governmentally funded assets. Assumes that investors may borrow/lend any amount at risk-free rate
4. Investors pay no taxes on returns and no transaction costs on trades in securities
5. All investors are rational M-V optimizers, they all use the Markowitz model
6. All investors analyze securities in the same way and share the same economic view of the world.
I.e., for any set of security prices, they all derive the same input list to feed into the Markowitz
model. This assumption is often referred to as homogeneous expectations
Summary of the equilibrium that will prevail in this hypothetical world of securities and investors:
1. All investors will choose to hold a portfolio of risky assets in proportions that duplicate representation of the assets in the market portfolio M , which includes all traded assets. The proportion of
each stock in the market portfolio equals the market value of the stock (price per share number
of shares outstanding) divided by the total market value of all stocks
2. Not only will the market portfolio be on the efficient frontier, but it also will be the tangency
portfolio to the optimal capital allocation line (CAL) derived by each and every investor
The capital market line (CML), the line from the risk-free rate through the market portfolio
M is also the best attainable capital allocation line. All investors hold M as their optimal risky
portfolio, differing only in the amount invested in it versus in the risk-free asset
3. The risk premium on the market portfolio will be proportional to its risk and the degree of risk
aversion of the representative investor:
2
M
E(rM ) rf = A
2 = variance of market portfolio and A
= avg. degree of risk aversion across investors
Where M
4. The risk premium on individual assets will be proportional to the risk premium on the market
portfolio M and the beta coefficient of the security relative to the market portfolio, where:

i =

Cov(ri , rM )
2
M

And the risk premium on individual securities is:

E(ri ) rf = i [E(rM ) rf ]
Why do all investors hold the market portfolio?
When we aggregate the portfolios of all individual investors, lending/borrowing cancel out, and the
value of the aggregate risky portfolio will equal the entire wealth of the economy
This is the market portfolio M
27

 The CAPM implies that as individuals attempt to optimize their personal portfolios, they each
arrive at the same portfolio, with weights on each asset equal to those of the market portfolio
If all investors use identical Markowitz analysis (Assumption 5) applied to the same universe of
securities (Assumption 3) for the same time horizon (Assumption 2) and use the same input list
(Assumption 6), they all must arrive at the same composition of the optimal risky portfolio, the
portfolio on the efficient frontier identified by the tangency line from T-bills to that frontier
As a result, the optimal risky portfolio of all investors is simply a share of the market portfolio
All assets have to be included in the market portfolio
When all investors avoid a stock i, the demand is zero, and is price takes a free fall
Ultimately, stock i reaches a price attractive enough to be included in the optimal portfolio
Such price adjustment process guarantees that all stocks are included in optimal portfolio
The passive strategy is efficient
In the simple world of the CAPM, M is the optimal tangency portfolio on the efficient frontier
Thus the passive strategy of investing in a market index portfolio is efficient
We sometimes call this result a mutual fund theorem
Another incarnation of separation property: Separate portfolio selection into 2 components:
1. A technical problem: Creation of mutual funds by professional managers
2. A personal problem that depends on an investors risk aversion: Allocation of the complete
portfolio between the mutual fund and risk-free assets
The risk premium of the market portfolio
Each individual investor chooses a proportion y allocated to the optimal portfolio M such that:
y=

E(rM ) rf
2
AM

(1)

Net borrowing and lending across all investors must be zero Substituting the representative
investors risk aversion A for A, the average position in the risky portfolio is 100%, or y = 1
Risk premium on market portfolio related to its variance by avg. degree of risk aversion:
2
E(rM ) rf = A
M

(2)

Expected returns on individual securities

The contribution of one stock i to portfolio variance can be expressed as the sum of all the covariance
terms corresponding to the stock Contribution of stock i to variance of market portfolio:
wi [w1 Cov(r1 , ri ) + + wi Cov(ri , ri ) + + wn Cov(rn , ri )]

(3)

When there are many stocks, there will be many more covariance terms than variance terms
Covariance of a stock with all others dominates that stocks contribution to total portfolio risk
Measure stock is contribution to market portfolio risk by its covariance with that portfolio:
Stock i contribution to variance = wi Cov(ri , rM )
Since rM =

wk rk , the covariance of return on stock i with market portfolio is:

!
n
n
X
X
Cov(ri , rM ) = Cov ri ,
wk rk =
wk Cov(ri , rk )
k=1

(4)

k=1

The contribution of our holding of stock i to the risk premium of the market portfolio is wi [E(ri )rf ]
Therefore, the reward-to-risk ratio for investments in i is:
wi [E(ri ) rf ]
E(ri ) rf
Stock i contribution to risk premium
=
=
Stock i contribution to variance
wi Cov(ri , rM )
Cov(ri , rM )
28

 The market portfolio is the tangency (efficient M-V) portfolio with reward-to-risk ratio:
E(rM ) rf
Market risk premium
=
2
Market variance
M

(5)

Market price of risk: Ratio in Eq. (5), quantifies extra return demanded to bear portfolio risk
A basic principle of equilibrium is that all investments should offer the same reward-to-risk ratio
The reward-to-risk ratios of i and the market portfolio should be equal:
E(ri ) rf
E(rM ) rf
=
2
Cov(ri , rM )
M

(6)

Hence, the fair risk premium of stock i is:

E(ri ) rf =

Cov(ri , rM )
[E(rM ) rf ]
2
M

(7)

Expected return-beta relationship

2 measures contribution of stock i to variance of market portfolio as
Ratio Cov(ri , rM )/M
fraction of total variance of market portfolio. This ratio is called and:
E(ri ) = rf + [E(rM ) rf ]

(8)

Assumptions that made individuals act similarly are very useful:

If everyone holds an identical risky portfolio, then everyone will find that the beta of each asset
with the market portfolio equals the assets beta with his or her own risky portfolio
Hence everyone will agree on the appropriate risk premium for each asset
Even if one does not hold precise market portfolio, a well-diversified portfolio is so highly
correlated with market that a stocks relative to market is a useful risk measure
Modified CAPMs hold true even if differences among individuals lead them to different portfolios
If CAPM relationship holds for any individual asset, it must hold for any combination of assets:
w1 E(r1 )
+w2 E(r2 )
+w2 E(r2 )

+wn E(rn )
E(rp )

=
=
=
=
=
=

w1 rf + w1 1 [E(rM ) rf ]
w2 rf + w2 2 [E(rM ) rf ]
w2 rf + w2 2 [E(rM ) rf ]

wn rf + wn n [E(rM ) rf ]
P
P
rf + p [E(rM ) rf ]
where E(rp ) = wk E(rk ) and p =
wk k

This result has to be true for market portfolio itself: E(rM ) = rf + M [E(rM ) rf ] M = 1
This also establishes 1 as the weighted-average value of beta across all assets
s greater than 1 are considered aggressive, s below 1 can be described as defensive
The security market line
CAPM relationship as a reward-risk equation: of a security = appropriate measure of its risk
because is proportional to the risk that security contributes to optimal risky portfolio
CAPM states that the securitys risk premium is directly proportional to both: (i) The beta and,
(ii) The risk premium of the market portfolio
The security market line (SML) graphs individual asset risk premiums as a function of asset risk
Relevant measure of risk for individual assets part of diversified portfolios is not assets std.
dev./variance. It is the asset contribution of the to portfolio variance (measured by assets )
The SML is valid for both efficient portfolios and individual assets
Fairly priced assets plot exactly on the SML
If stock perceived to be a good buy (underpriced) it will provide an expected return in excess
of fair return stipulated by SML Underpriced stocks plot above SML: Given their s, their
expected returns are greater than implied by CAPM. Overpriced stocks plot below the SML
Difference between fair/actually expected rates of return on a stock is called the stocks
29

E(r)

Security Market Line (SML)

Stock

E(rM)

Market
E(rM) { rf = Slope of SML
rf

M = 1:0

1.5

Figure 1: The security market line

The security market line provides a benchmark for the evaluation of investment performance: Given
the risk of an investment, as measured by its beta, the SML provides the required rate of return
necessary to compensate investors for both risk as well as the time value of money
In contrast, the capital market line (CML) graphs the risk premiums of efficient portfolios (i.e.
portfolios composed of the market and the risk-free asset) as a function of portfolio
Analysis suggests: Starting point of portfolio management = passive market-index portfolio
The portfolio manager will then increase the weights of securities with positive alphas and
decrease the weights of securities with negative alphas
The CAPM is also useful in capital budgeting decisions
For a firm considering a new project, the CAPM can provide the required rate of return that
the project needs to yield, based on its beta, to be acceptable to investors
Yet another use of the CAPM is in utility ratemaking cases
In this case the issue is the rate of return that a regulated utility should be allowed to earn
on its investment in plant and equipment. The firm would be allowed to set prices at a level
expected to generate rate of returns indicated by the CAPM
The CAPM and the index model
Actual returns versus expected returns
One central prediction of CAPM: The market portfolio is a M-V efficient portfolio
However, testing its efficiency has not been feasible
CAPM involves expected returns. We can only observe actual/realized HPRs
The second central set of CAPM predictions is the expected return-beta relationship
Problem of measuring expectations as well
We must make additional assumptions to make CAPM implementable and testable
The index model and realized returns
To go from expected to realized returns, use the index model in excess return form:
Ri = i + i RM + ei

(9)

Covariance between the returns on stock i and the market index

Firm-specific component independent of market wide component Cov(RM , ei ) = 0
Hence, the covariance of the excess rate of return on security i with the market index is:
2
Cov(Ri , RM ) = Cov(i RM + ei , RM ) = i M

Sensitivity coefficient i in Eq. (9), which is the slope of the index model regression, equals:
i =

Cov(Ri , RM )
2
M

The index model = in CAPM relationship, except that we replace the CAPM (theoretical)
market portfolio with the well-specified/observable market index
30

 The index model and the expected return-beta relationship

If the index M in Eq. (9) represents the true market portfolio, we can take the expectation of each
side of the equation to show that the index model specification is: E(ri ) rf = i [E(rm ) rf ]
A comparison of the index model relationship to the CAPM expected return-beta relationship Eq.
(8) shows that the CAPM predicts that i should be zero for all assets
i is the expected return in excess of (or below) fair expected return predicted by CAPM
If the stock is fairly priced, its alpha must be zero
Therefore, if we estimate the index model for several firms, using Eq. (9) as a regression, we should
find that ex post/realized s (regression intercepts) center around zero
CAPM states that E[] = 0 for all securities, whereas index model representation of CAPM holds
that realized value of should average out to zero for sample of historical observed returns
Indirect evidence on the efficiency of the market portfolio can be found in a study that estimates
values for a large sample of equity mutual funds: The distribution of s is roughly bell shaped,
with a mean that is slightly negative but statistically indistinguishable from zero
The market model
Other applicable variation on the intuition of the index model, which divides returns into firmspecific and systematic components somewhat differently from the index model
The market model states that the return surprise of any security is proportional to the return
surprise of the market, plus a firm-specific surprise:
ri E(ri ) = i [rM E(rM )] + ei
If CAPM is valid, substituting for E(ri ) from Eq. (8), the market model equation is identical to
the index model Index model and market model are used interchangeably
Is the CAPM practical?
If CAPM was valid, single-index model in which index includes all traded securities also valid
All alpha values in security risk premiums would be identically zero
All 0 s = 0 feasible in principle, but cannot be expected to emerge in real markets
Such an equilibrium may be one that the real economy can approach, but not necessarily reach
Actions of security analysts drive security prices to proper levels at which = 0
But if all 0 s 0, there would be no incentive to engage in such security analysis
A more reasonable standard, that the CAPM is the best available model to explain rates of return on
risky assets, means that in the absence of security analysis, one should take security alphas as zero
Is the CAPM testable?
A model consists of (i) A set of assumptions, (ii) Logical/mathematical development of the model,
and (iii) a set of predictions. We can test a model in two ways: Normative and positive
Normative tests examine the assumptions of the model
Positive tests examine the predictions
Model robust wrt. assumption if predictions not highly sensitive to assumption violation
If we use only assumptions to which the model is robust, the models predictions will be
reasonably accurate despite its shortcomings
Tests of models are almost always positive: Judge model on success of empirical predictions
Because the nonrealism of CAPM assumptions precludes a normative test, the positive test is really
a test of the robustness of the model to its assumptions
The CAPM implications are embedded in two predictions:
1. The market portfolio is efficient
2. Security market line (CAPM relationship) accurately describes risk-return trade-off (0 s = 0)
Central problem in testing predictions: The hypothesized market portfolio is unobservable
It is difficult to test the efficiency of an observable portfolio, let alone an unobservable one
These problems alone make adequate testing of the model infeasible

31

 The CAPM fails empirical tests

Tests use proxies such as the S&P 500 index to stand in for the true market portfolio. Assumption
that the market proxy is sufficiently close to the true, unobservable market portfolio
CAPM fails these tests: Data reject hypothesis that 0 s 0 at acceptable levels of significance
E.g., on average, low-beta securities have > 0 and high-beta securities have < 0
The economy and the validity of the CAPM
Some industries are regulated, with rate commissions either setting or approving prices
The normative framework of the typical rate hearing is that shareholders, who have made an
investment in the firm, are entitled to earn a fair rate of return on their equity investment
The CAPM provides a clear criterion: If the rates under current regulation are too low, then
the rate of return to equity investors would be less than commensurate with risk
Similar applications arise in many legal settings
Contracts with payoffs that are contingent on a fair rate of return
Disputes involving damages (to get the proper discount rate)
CAPM is an accepted norm in the US, despite its empirical shortcomings. Two reasons:
1. The logic of the decomposition to systematic and firm-specific risk is compelling
2. There is impressive, albeit less-formal, evidence that the central conclusion of the CAPM (the
efficiency of the market portfolio) may not be all that far from being valid
The investments industry and the validity of the CAPM
More than other practitioners, investment firms must take a stand on the validity of the CAPM
E.g., CAPM provides discount rates that help analysts assess a firms intrinsic value
If they judge CAPM invalid, they need a substitute framework to construct optimal portfolios
The curriculum of the CFA Institute also suggests a widespread acceptance of the CAPM, at least
as a starting point for thinking about the risk-return relationship
Explicitly or implicitly, practitioners do use a CAPM
Econometrics and the expected return-beta relationship
When assessing empirical success of CAPM, we must also consider our econometric technique:
If our tests are poorly designed, we may mistakenly reject the model
Similarly, some empirical tests implicitly introduce additional assumptions not part of CAPM
Estimates are not independent
If the beta estimate is inefficient and/or biased, so will be the estimate of the intercept
Unfortunately, statistical bias is easily introduced
There are several potential problems with the estimation of beta coefficients:
1. When residuals correlated (firms in same industry), standard estimates are not efficient
Simple approach to this problem: Use statistical techniques designed for these complications
E.g., replace OLS (ordinary least squares) regressions with GLS (generalized least squares)
regressions, which account for correlation across residuals
2. Both coefficients, alpha and beta, as well as residual variance, are likely time varying
There is nothing in the CAPM that precludes such time variation, but standard regression
techniques rule it out and thus may lead to false rejection of the model
There are now well-known techniques to account for time-varying parameters
3. Betas may vary not purely randomly over time, but in response to changing economic conditions
A conditional CAPM allows risk and return to change with a set of conditioning variables
4. The beta of a security can be decomposed into two components:
One of which measures sensitivity to changes in corporate profitability
Another which measures sensitivity to changes in the markets discount rates
Improved econometric techniques may help resolve part of the empirical failure of the simple CAPM
Extensions of the CAPM
The Zero-beta model
Efficient frontier portfolios have a number of interesting characteristics:
1. Any portfolio that is a combination of two frontier portfolios is itself on the efficient frontier
32

2. The expected return of any asset can be expressed as an exact linear function of the expected
return on any two efficient-frontier portfolios P and Q:
E(ri ) E(rQ ) = [E(rP ) E(rQ )]

Cov(ri , rP ) Cov(rP , rQ )
P2 Cov(rP , rQ )

(10)

3. Every portfolio on the efficient frontier, except for the global M-V portfolio, has a companion
portfolio on bottom (inefficient) half of the frontier with which it is uncorrelated
Uncorrelated Companion portfolio called zero-beta portfolio of efficient portfolio
If we choose the market portfolio M and its zero-beta companion portfolio Z, then Eq.
(10) simplifies to the CAPM-like equation:
E(ri ) E(rZ ) = [E(rM ) E(rZ )]

Cov(ri , rM )
= i [E(rM ) E(rZ )]
2
M

(11)

Eq. (11) resembles the SML of the CAPM, except that the risk-free rate is replaced with the
expected return on the zero-beta companion of the market index portfolio
Eq. (11): CAPM equation for investors restricted when borrowing/investing in risk-free asset
Because average returns on the zero-beta portfolio are greater than observed T-bill rates, the zerobeta model can explain why average estimates of alpha values are positive for low-beta securities
and negative for high-beta securities, contrary to the prediction of the CAPM
Labor income and nontraded assets
An important departure from realism is the CAPM assumption that all risky assets are traded.
Two important asset classes that are not traded are:
1. Human capital
2. Privately held businesses
Privately held business may be the lesser of the two sources of departures from the CAPM
Nontraded firms can be incorporated or sold at will, save for liquidity considerations
Owners of private business also can borrow against their value
Suppose that privately held business have similar risk characteristics as those of traded assets
Individuals can partially offset the diversification problems posed by their nontraded entrepreneurial assets by reducing their portfolio demand for securities of similar, traded assets
CAPM equation may not be greatly disrupted by presence of entrepreneurial income
To the extent that risk characteristics of private enterprises differ from those of traded securities,
a portfolio of traded assets that best hedges the risk of typical private business would enjoy excess
demand from the population of private business owners
The price of assets in this portfolio will be bid up relative to the CAPM considerations, and
the expected returns on these securities will be lower in relation to their systematic risk
Adding proprietary income to a standard asset-pricing model improves its predictive performance
The size of labor income and its special nature is of greater concern for validity of CAPM
Despite individuals borrowing against labor income (via a home mortgage) and reducing some
uncertainty about future labor income via life insurance, human capital is less portable across
time and more difficult to hedge using traded securities than nontraded business
This induces pressure on security prices and results in departures from CAPM equation
Equilibrium expected return-beta equation for an economy in which individuals are endowed with
labor income of varying size relative to their nonlabor capital: The resultant SML equation is:
E(Ri ) = E(RM )
Where

PH
PM
PH

=
=
=

Cov(Ri , RM ) +
2 +
M

PH
PM Cov(Ri , RH )

PH
PM Cov(RM , RH )

Value of aggregate human capital

Market value of traded assets (market portfolio)
Excess rate of return on aggregate human capital

33

(12)

 The CAPM measure of systematic risk is replaced in the extended model by an adjusted beta
that also accounts for covariance with the portfolio of aggregate human capital
Because we expect Cov(Ri , RH ) to be positive for the average security, the risk premium in
this model will be greater, on average, than predicted by the CAPM for securities with beta
less than 1, and smaller for securities with beta greater than 1
Model predicts a security market line less steep than that of standard CAPM
This may help explain the average negative alpha of high-beta securities and positive alpha of
low-beta securities that lead to the statistical failure of the CAPM
A multiperiod model and hedge portfolios
Merton relaxes the single-period myopic assumptions about investors
He envisions individuals who optimize a lifetime consumption/investment plan, and who continually adapt consumption/investment decisions to current wealth/planned retirement age
When uncertainty about portfolio returns is the only source of risk and investment opportunities remain unchanged through time (no change in probability distribution of return on market
portfolio or individual securities) Mertons intertemporal capital asset pricing model
(ICAPM) predicts same expected return-beta relationship as single-period model
But the situation changes when we include additional sources of risk, of two general kinds:
1. One concerns changes in the parameters describing investment opportunities, such as future
risk-free rates, expected returns, or the risk of the market portfolio
Investors will sacrifice some expected return if they can find assets whose returns will be
higher when other parameters (e.g., the risk-free rate) change adversely
2. The other additional source of risk concerns the prices of the consumption goods that can be
purchased with any amount of wealth (e.g. inflation risk)
Investors may be willing to sacrifice some expected return to purchase securities whose
returns will be higher when the cost of living changes adversely
Suppose we identify K sources of extramarket risk and find K associated hedge portfolios
Mertons ICAPM equation generalizes the SML to a multi-index version:
E(Ri ) = iM E(RM ) +

K
X

iK E(RK )

where iK = beta on k-th hedge portfolio

(13)

k=1

A consumption based CAPM

The logic of the CAPM together with the hedging demands noted in the previous subsection
suggests that it might be useful to center the model directly on consumption
In a lifetime consumption plan, investor must in each period balance allocation of current
wealth between todays consumption and savings/investment supporting future consumption
When optimized, utility value from additional dollar of consumption today = utility value of
expected future consumption financed by that additional dollar of wealth
Future wealth will grow from: (i) Labor income, and (ii) Returns on that dollar when invested
in the optimal complete portfolio
As a general rule, investors will value additional income more highly during difficult economic times
An asset will therefore be viewed as riskier in terms of consumption if it has positive covariance
with consumption growth Equilibrium risk premiums will be greater for assets that exhibit
higher covariance with consumption growth
We can write the risk premium on an asset as a function of its consumption risk as:
E(Ri ) = iC RPC

(14)

Portfolio C may be interpreted as a consumption-tracking portfolio,i.e. the portfolio with

the highest correlation with consumption growth
iC = slope in regression of asset is excess returns on consumption-tracking portfolio
RPC = risk premium associated with consumption uncertainty: RPC = E(RC ) = E(rC ) rf
34

 Consumption-tracking portfolio in CCAPM plays role of market portfolio in CAPM

Focus on risk of consumption opportunities rather than risk/return of portfolios dollar value
Excess return on consumption-tracking portfolio Excess return on market portfolio
In contrast to CAPM, M on the market factor of the CCAPM is not necessarily 1
Advantage: Compactly incorporates consumption hedging/changes in investment opportunities
However, consumption growth figures are published infrequently (monthly at the most) compared with financial assets, and are measured with significant error
Liquidity and the CAPM
Standard models of asset pricing assume frictionless markets (securities can be traded costlessly)
In reality, trading (and trading costs) will be of great importance to investors
Liquidity
The liquidity of an asset is the ease and speed with which it can be sold at fair market value
Part of liquidity is the cost of engaging in a transaction, particularly the bid-ask spread
Another part is price impact: Adverse movement in price encountered when attempting large trade
Another part is immediacy: Ability to sell quickly without reverting to fire-sale prices
Conversely, illiquidity can be measured in part by the discount from fair market value seller must
accept if asset is to be sold quickly. A perfectly liquid asset would entail no illiquidity discount
Stock investors pay high price for liquidity
Overall, least-liquid stocks average 8.5%/year higher return than most-liquid stocks
On average, a 1% point increase in spread is associated with a 2.5% higher annual return (NYSE)
The relationship held after results were adjusted for size and other risk factors
Cost of spread incurred for each trade Illiquid stocks can quickly become prohibitively expensive
Liquidity (or lack of it) is an important characteristic that affects asset values
Legal cases: Courts apply steep discounts to values of businesses not publicly traded
Liquidity is increasingly viewed as an important determinant of prices and expected returns
Early models of liquidity focused on the inventory management problem faced by security dealers
Dealers in OTC markets post prices at which they buy (the bid price) or sell (the ask price)
The fee they earn for supplying this liquidity is the bid-ask spread
Part of the bid-ask spread is compensation for bearing price risk involved in holding an inventory
of securities and allowing inventory levels to absorb fluctuations in overall security demand
Assuming the fair price of the stock is the average of the bid and ask prices, an investor pays half
the spread upon purchase and another half upon sale of the stock
The spread is one important component of liquidity: It is the cost of transacting in a security
In electronic markets, the limit-order book contains the inside spread: Difference between highest
price at which some investor will purchase and lowest price at which another investor will sell
The effective bid-ask spread will also depend on the size of the desired transaction
Greater emphasis today on the component of the spread that is due to asymmetric information
Traders who post prices need to be worried about being picked off by better-informed traders who
hit their limit prices only when they are out of line with the intrinsic firm value
Investors trade securities for two reasons:
1. Some trades are driven by noninformational motives
E.g., selling assets to raise cash for a big purchase, or even just for portfolio rebalancing
These trades, not motivated by private info, are called noise trades
Security dealers will earn a profit from the bid-ask spread when transacting with noise traders
(also called liquidity traders because their trades may derive from needs for liquidity)
2. Other transactions are motivated by private information known only to the seller or buyer
These transactions are generated when traders believe they have come across information that
a security is mispriced, and try to profit from that analysis
Information traders impose a cost on both dealers and other investors who post limit orders
The greater the relative importance of information traders, the greater the required spread to
compensate for the potential losses from trading with them
35

 Discount in a security price from illiquidity can be very large, far larger than bid-ask spread
The PV of all three future trading costs (spreads) are discounted into the current price
For any given spread, the price discount increases almost in proportion to frequency of trading
For 1% spread, if security is traded once a year forever, current illiquidity cost = immediate cost
plus PV of a .5% perpetuity. At 5% discount rate, .005 + .005/.05 = 10.5%
The reduction in the rate of return due to trading costs is lower the longer the security is held
In equilibrium, investors with long holding periods, on average, hold more of the illiquid securities
Short-horizon investors will more strongly prefer liquid securities
This clientele effect mitigates the effect of the bid-ask spread for illiquid securities
The end result is that liquidity premium should increase with bid-ask spread at a decreasing rate
What about unanticipated changes in liquidity?
Investors may also demand compensation for liquidity risk
The bid-ask spread of a security is not constant through time, nor is the ability to sell a security
at a fair price with little notice. Both depend on overall conditions in security markets
There may be a systematic component to liquidity risk that affects the equilibrium rate of return
and hence the expected return-beta relationship
Acharya and Pedersen consider the impacts of both the level and the risk of liquidity on security pricing
They include three components to liquidity risk: Each captures the extent to which liquidity varies
systematically with other market conditions
Therefore, expected return depends on expected liquidity, as well as the conventional CAPM
beta and three additional liquidity-related betas:
E(Ri ) = kE(Ci ) + ( + L1 L2 L3 )
where

E(Ci )
k

Li

=
=
=
=
=

(15)

Expected cost of illiquidity

Adjustment for average holding period over all securities
Market risk premium net of average market illiquidity cost, E[RM CM ]
Measure of systematic market risk
Liquidity betas

Compared to the conventional CAPM, the expected return-beta equation now has a predicted
firm-specific component that accounts for the effect of security liquidity
Such an effect would appear to be an alpha in the conventional index model
The market risk premium itself is measured net of the average cost of illiquidity = E[RM CM ],
where CM is the market-average cost of illiquidity
Overall risk of each security accounts for three elements of liquidity risk:
1. L1 measures sensitivity of the securitys illiquidity to market illiquidity. Investors want extra
compensation for holding security that becomes illiquid when general liquidity is low
L1 =

Cov[Ci , CM ]
V ar[RM CM ]

2. L2 measures the sensitivity of the stocks return to market illiquidity. This coefficient appears
with a negative sign in Eq. (15) because investors are willing to accept a lower average return
on stocks that will provide higher returns when market illiquidity is greater
L2 =

Cov[Ri , CM ]
V ar[RM CM ]

3. L3 measures sensitivity of security illiquidity to the market rate of return. This sensitivity
also appears with a negative sign, because investors accept a lower average return on securities
that can be sold more easily (have low illiquidity costs) when the market declines
L1 =

Cov[Ci , RM ]
V ar[RM CM ]

All liquidity variants improve on the explanatory power of the CAPM

36

BKM - Ch. 10: Arbitrage pricing theory and multifactor models of risk and return
Introduction
Arbitrage: The exploitation of security mispricing so that risk-free profits can be earned
Involves simultaneous purchase/sale of equivalent securities to profit from price discrepancies
The most basic principle of capital market theory is that equilibrium market prices are rational in
that they rule out arbitrage opportunities
If actual security prices allow for arbitrage, the result will be strong pressure to restore equilibrium
Security markets ought to satisfy a no-arbitrage condition
Multifactor models of security returns can be used to measure and manage exposure to each of many
economy-wide factors such as business-cycle risk, interest or inflation rate risk, energy price risk
Factor models combined with a no-arbitrage condition lead to simple relationship between expected
return and risk This approach to risk-return trade-off is called Arbitrage Pricing Theory (APT)
Multifactor models: An overview
Sometimes, rather than using market proxy, more useful to focus directly on ultimate sources of risk
Useful in risk assessment when measuring ones exposures to particular sources of uncertainty
Factor models allow to quantify the different factors affecting the rate of return on a security
Factor models of security returns
Uncertainty in asset returns has two sources:
1. A common or macroeconomic factor
2. Firm-specific events
The common factor is constructed to have zero expected value, because we use it to measure new
information concerning the macro-economy which, by definition, has zero expected value
If F is the deviation of the common factor from its EV, i the sensitivity of firm i to that factor, and
ei the firm-specific disturbance, factor models state that the actual return on firm i equals its initially expected return plus a (zero EV) random amount attributable to unanticipated economywide
events, plus another (zero EV) random amount attributable to firm-specific events:
ri = E(ri ) + i F + ei

(1)

Example: Factor models

Suppose that macro factor F is taken to be news about state of business cycle, measured by
unexpected % change in GDP, and that consensus is that GDP will increase by 4% this year
Suppose stocks value is 1.2. If GDP increases by only 3%, then F = 1%
This disappointment would translate into a return on the stock that is 1.2% lower than expected
When we estimate a single-index regression, we implicitly impose an (incorrect) assumption that each
stock has the same relative sensitivity to each risk factor
A more explicit representation of systematic risk, allowing for different stocks exhibiting different
sensitivities to its various components, is a useful refinement of the single-factor model
multifactor models can provide better descriptions of security returns
Apart from their use in building models of equilibrium security pricing, multifactor models are
useful in risk management applications: These models give us a simple way to measure our exposure
to various macroeconomic risks, and construct portfolios to hedge those risks
Two-factor model
Macroeconomic sources of risk: Unanticipated growth in GDP and changes in interest rates:
ri = E(ri ) + iGDP GDP + iIR IR + ei

(2)

Both macro factors have zero EV: They represent unanticipated changes in these variables
The coefficients of each factor in Eq. (2) measure the sensitivity of share returns to that factor
Coefficients are called factor sensitivities, factor loadings, or factor betas
Increase in interest rates is bad news for most firms Interest rate betas generally negative
37

 Factor betas can provide a framework for a hedging strategy:

The idea for investors who wish to hedge a source of risk is to establish an opposite factor
exposure to offset that source of risk
A multifactor Security Market Line (SML)
The multifactor model is no more than a description of the factors that affect security returns.
Where does E(r) comes from, i.e., what determines a securitys expected rate of return
A multifactor index model leads to a multifactor SML in which the risk premium is determined by
exposure to each systematic risk factor, and by risk premium associated with those factors
E.g., in a two-factor economy in which risk exposures can be measured by Eq. (2), we would
conclude that the expected rate of return on a security i would be the sum of:
1. The risk-free rate of return
2. [Sensitivity to GDP risk (iGDP )] [Risk premium for bearing GDP risk]
3. [Sensitivity to interest rate risk (iIR )] [Risk premium for bearing interest rate risk]
Hence, the two-factor security market line:
E(r) = rf + GDP RPGDP + IR RPIR

(3)

One difference between single/multiple factor economy: A factor risk premium can be negative
E.g., a security with a positive interest rate beta performs better when rates increase, and thus
would hedge the value of a portfolio against interest rate risk
Investors accept a lower rate of return (negative risk premium) as cost of hedging attribute
How to estimate the risk premium for each factor?
Like CAPM, the risk premium for each factor is the risk premium of a portfolio that has
= 1.0 on that particular factor and = 0 on all other factors
I.e., it is the risk premium one expects to earn by taking a pure play on that factor
Arbitrage pricing theory
Like CAPM, APT predicts a SML linking expected returns to risk. APT relies on 3 key propositions:
1. Security returns can be described by a factor model
2. There are sufficient securities to diversify away idiosyncratic risk
3. Well-functioning security markets do not allow for the persistence of arbitrage opportunities
Arbitrage, risk arbitrage, and equilibrium
Arbitrage opportunity: When investor can earn riskless profits without making a net investment
E.g., if shares of a stock sold for different prices on two different exchanges
The Law of One Price
If two assets are equivalent in all economically relevant respects, then same market price
The Law of One Price is enforced by arbitrageurs: They will bid up the price where it is low
and force it down where it is high until the arbitrage opportunity is eliminated
Market prices move to rule out arbitrage opportunities: Fundamental concept in capital markets
The critical property of a risk-free arbitrage portfolio is that any investor, regardless of risk
aversion or wealth, will want to take an infinite position in it
Because those large positions will quickly force prices up or down until the opportunity vanishes,
security prices should satisfy a no-arbitrage condition
Difference between arbitrage/risk-return dominance arguments in support of equilibrium prices:
A dominance argument holds that when an equilibrium price relationship is violated, many
investors will make limited portfolio changes, depending on their degree of risk aversion
Aggregation of these limited portfolio changes is required to create a large volume of buying
and selling, which in turn restores equilibrium prices
By contrast, with arbitrage, each investor wants to take as large a position as possible
Few investors needed to bring about price pressures necessary to restore equilibrium
Therefore, implications for prices derived from no-arbitrage arguments are stronger than implications derived from a risk-return dominance argument
CAPM: Example of dominance argument, implying that all investors hold M-V efficient portfolios
38

 Arbitrageur often refers to a professional searching for mispriced securities in specific areas such
as merger-target stocks, rather than to one who seeks strict (risk-free) arbitrage opportunities
Such activity is sometimes called risk arbitrage to distinguish it from pure arbitrage
Well-diversified portfolios
If we construct an n-stock portfolio with weights wi , then the rate of return on this portfolio is:
X
X
rp = E(rp ) + p F + ep where p =
wi i and E(rp ) =
wi E(ri )
(4)
The portfolio nonsystematic component (which is uncorrelated with F ) is ep =
The portfolio variance is:

wi e i

p2 = p2 F2 + 2 (ep )
Where F2 is the variance of the factor F and 2 (ep ) is the nonsystematic risk of the portfolio:
hX
i X
2 (ep ) = V ar
wi ei =
wi2 2 (ei )
If the portfolio was equally weighted wi = 1/n, then the nonsystematic variance would be:
2

(ep ) = V ar

hX

wi ei =

X  1 2
n

2 (ei ) =

1 2
1 X 2 (ei )
=
(ei )
n
n
n

Where the last term is the average value across securities of nonsystematic variance
When n is large, nonsystematic variance approaches zero Effect of diversification
Well-diversified portfolio
A portfolio that is diversified over a large enough number of securities, with each weight wi
small enough that for practical purposes the nonsystematic variance 2 (ep ) is negligible
Because the expected value of ep for any well-diversified portfolio is zero, and its variance also
is effectively zero, we can conclude that any realized value of ep will be virtually zero
Hence, for a well-diversified portfolio, for all practical purposes:
rp = E(rp ) + p F
Betas and expected returns
Because nonfactor risk can be diversified away, only factor risk should command a risk premium
Portfolios with same beta
Well-diversified portfolio A with A = 1 for systematic factor F : E(rA ) + A F = 10% + 1.0 F
Consider another well-diversified portfolio B with an expected return of 8% and B = 1
Could portfolios A and B coexist with the return pattern depicted? Clearly not
If one sells short $1 million of B and buy $1 million of A, a zero net investment strategy, the
riskless payoff would be $20,000:
(.10 + 1.0 F ) $1M
(.08 + 1.0 F ) $1M
.02 $1M = $20, 000

from long position in A

from short position in B
net proceeds

The profit is risk-free because the factor risk cancels out across the long and short positions
Moreover, the strategy requires zero net investment
One should pursue it on an infinitely large scale until the return discrepancy disappears
Well-diversified portfolios with equal s must have equal expected returns in equilibrium
Portfolios with different betas
Risk-free rate = 4%. Well-diversified portfolio C ( = .5) has expected return of 6%
Consider new portfolio D composed of half of portfolio A and half of the risk-free asset
Portfolio Ds = .5 0 + .5 1.0 = .5, and expected return = 5 4 + .5 10 = 7%
D has same beta but greater expected return than C Arbitrage opportunity
39

E(r)

10

Risk Premium

7
6

rf = 4
F
with respect
to macro factor
0

0.5

1.0

Figure 1: An arbitrage opportunity

We conclude that, to preclude arbitrage opportunities, the expected return on all well-diversified
portfolios must lie on the straight line from the risk-free asset in Fig. 1
The equation of this line will dictate the expected return on all well-diversified portfolios
Risk premiums are indeed proportional to portfolio betas
Risk premium shown by vertical arrow (distance between risk-free rate/expected return)
The risk premium is zero for = 0 and rises in direct proportion to
The one-factor security market line
Consider the market index portfolio M as a well-diversified portfolio, and let us measure the
systematic factor as the unexpected return on that portfolio
Because index portfolio must be on the line in Fig. 1 and M = 1, the equation of the line is:
E(rp ) = rf + [E(rM ) rf ]p

(5)

Arbitrage and the security market line

Suppose market index is well-diversified with expected return 10% and deviations of its return
from expectation (rM 10%) serve as systematic factor. T-bill rate is 4%
Then the SML, Eq. (5) implies that the expected rate of return on well-diversified portfolio E
with a beta of 2/3 should be 4% + 2/3(10 4) = 8%
If its expected return actually is 9%, then there will be an arbitrage opportunity
Buy $1 worth of E and sell $1 of portfolio invested 1/3 in T-bills and 2/3 in market
Risk-free profit per dollar invested = Deviation of expected return on E from SML
Unlike CAPM, APT does not require benchmark SML portfolio = true market portfolio
Any well-diversified portfolio lying on the SML may serve as the benchmark portfolio
APT more flexible than CAPM (unobservable market portfolio not a concern)
APT provides further justification for index model in practical SML implementation
Individual assets and the APT
If no arbitrage, each well-diversified portfolios expected excess return must be proportional to its
If satisfied by all well-diversified portfolios, it must be satisfied by almost all individual securities
Suppose that the expected return-beta relationship is violated for all single assets. Now create a
pair of well-diversified portfolios from these assets
What are the chances that in spite of the fact that for any pair of assets the relationship does not
hold, the relationship will hold for both well-diversified portfolios? The chances are small
Now construct yet a third well-diversified portfolio. What are the chances that the violations of the
relationships for single securities are such that the third portfolio also will fulfill the no-arbitrage
expected return-beta relationship? Obviously, the chances are smaller still
Continue with a fourth well-diversified portfolio, and so on. If the no-arbitrage expected returnbeta relationship has to hold for infinitely many different, well-diversified portfolios, it must be
virtually certain that the relationship holds for all but a small number of individual securities
40

 Imposing no-arbitrage condition implies maintenance of the expected return-beta relationship for all
well-diversified portfolios and for all but possibly a small number of individual securities
The APT and the CAPM
The APT serves many of the same functions as the CAPM
Rate of return benchmark for capital budgeting/security valuation/performance evaluation
APT highlights the crucial distinction between nondiversifiable risk (factor risk) that requires
a reward in the form of a risk premium and diversifiable risk that does not
Advantages of the APT
Depends on assumption that rational equilibrium in markets precludes arbitrage opportunities
Uses diversified portfolio that can be constructed practically from large number of securities
In contrast, the CAPM is derived assuming an inherently unobservable market portfolio
The CAPM argument rests on mean-variance efficiency
Disadvantages of the APT
The APT does not fully dominate the CAPM
CAPM: Unequivocal statement on expected return-beta relationship for all securities
APT: Implies that this relationship holds for all but perhaps a small number of securities
APT cannot rule out a violation of expected return-beta relationship for any particular asset
A multifactor APT
Multifactor version of the APT to accommodate multiple sources of risk. E.g., two-factor model:
ri = E(ri ) + i1 F1 + i2 F2 + ei

(6)

Factor portfolio
Well-diversified portfolio with = 1 on one of the factors and = 0 on all other factor
Think of factor portfolio as a tracking portfolio: Returns on such portfolio track evolution of
particular sources of macroeconomic risk, but are uncorrelated with other sources of risk
It is possible to form such factor portfolios because we have a large number of securities to choose
from, and a relatively small number of factors
Factor portfolios will serve as the benchmark portfolios for a multifactor security market line
The multifactor APT states that the overall risk premium on a portfolio must equal the sum of the risk
premiums required as compensation for each source of systematic risk
The factor exposures of any portfolio P are given by its betas P 1 and P 2
A competing portfolio Q can be formed by investing in factor portfolios with the following weights:
P 1 in the first factor portfolio
P 2 in the second factor portfolio
1 P 1 P 2 in T-bills
By construction, portfolio Q will have betas equal to those of portfolio P and expected return of:
E(rQ ) = P 1 E(r1 ) + P 2 E(r2 ) + (1 P 1 P 2 )rf = rf + P 1 [E(r1 ) rf ] + P 2 [E(r2 ) rf ] (7)
Any well-diversified portfolio with P 1 /P 2 must have Eq. (7) return if no arbitrage opportunities
Extension of multifactor SML to individual assets: Same as for one-factor APT
Eq. (7) cannot be satisfied by all well-diversified portfolio unless satisfied by virtually every security
individually Eq. (7) represents multifactor SML for economy with multiple sources of risk
Multifactor APT can be used to provide fair rates of return for regulated utilities, like CAPM
Where should we look for factors?
Multifactor APT gives no guidance to determine relevant risk factors/their risk premiums
Two principles guide us when we specify a reasonable list of factors:
1. Limit ourselves to systematic factors with considerable ability to explain security returns
2. Choose factors likely to be important risk factors, i.e. factors that concern investors sufficiently
that they will demand meaningful risk premiums to bear exposure to those sources of risk
41

 One of many possible sets that might be considered:

IP
EI
UI
CG
GB

=
=
=
=
=

% change in industrial production

% change in expected inflation
% change in unanticipated inflation
Excess return of long-term corporate bonds over long-term government bonds
Excess return of long-term government bonds over T-bills

Thus, five-factor model of security returns during holding period t:

rit = i + iIP IPt + iEI EIt + iU I U It + iGC GCt + iGB GBt + eit

(8)

The Fama-French (FF) three-factor model

An alternative approach to specifying macroeconomic factors as candidates for relevant sources of
systematic risk uses firm characteristics that seem to proxy for exposure to systematic risk
Factors are variables that seem to predict avg. returns well and thus may capture risk premiums
Fama and French three-factor model:
rit = i + iM RM t + iSM B SM Bt + iHM L HM Lt + eit

(9)

SMB: Small Minus Big = Return of small stocks portfolio in xs of large stocks portfolio
HML: High Minus Low = Return of a portfolio of stocks with a high book-to-market ratio in
excess of the return on a portfolio of stocks with a low book-to-market ratio
Market index expected to capture systematic risk originating from macroeconomic factors
These two firm-characteristic variables are chosen because of observations that corporate capitalization (firm size) and book-to-market ratio predict deviations of avg. stock returns from CAPM
While SMB and HML are not themselves obvious candidates for relevant risk factors, the hope
is that these variables proxy for yet-unknown more fundamental variables
Fama and French: Firms with high ratios of book-to-market value are more likely to be in
financial distress and small stocks are more sensitive to changes in business conditions
These variables may capture sensitivity to risk factors in the macroeconomy
The problem with empirical approaches which use proxies for extramarket sources of risk, is that
factors in the models cannot be clearly identified as hedging significant source of uncertainty
The firm-characteristic basis of Fama-French factors raises the question of whether they reflect
APT model or approximation to a multi-index ICAPM based on extra-market hedging demands
Important distinction for the debate over the proper interpretation of the model
The validity of FF-style models may constitute either: (i) A deviation from rational equilibrium
(as there is no rational reason to prefer one or another of these firm characteristics per se),
or (ii) That firm characteristics identified as empirically associated with average returns are
correlated with other (yet unknown) risk factors

The multifactor CAPM and the APT

Multi-index CAPM inherits risk factors from risks that investors deem important enough to hedge
If hedging demands are common to many investors, prices of securities with desirable hedging
characteristics will be bid up and their expected return reduced. This process requires a multifactor
model to explain expected returns, where each factor arises from particular hedging motive
Risk sources that are priced in market equilibrium presumably will be systematic sources of
uncertainty that affect investors broadly
In contrast, the APT is largely silent on where to look for priced sources of risk
Lack of guidance problematic, but accommodates less structured search for risk factors

42

BKM - Ch. 11: The efficient market hypothesis

Random walks and the efficient market hypothesis
Attempts to find recurrent patterns in stock price movements are likely to fail
A forecast about favorable future performance leads instead to favorable current performance, as
market participants all try to get in on the action before the price jump
Any info that could be used to predict stock performance should already be reflected in stock prices
New information and random walk
If prices are bid immediately to fair levels, they increase/decrease only in response to new info
New info, by definition, must be unpredictable: Stock prices that change in response to new
(unpredictable) information also must move unpredictably
This is the essence of the argument that stock prices should follow a random walk
Far from market irrationality, randomly evolving stock prices are the necessary consequence of
intelligent investors competing to discover relevant information on which to buy/sell
If stock price movements were predictable, then evidence of stock market inefficiency, because
ability to predict prices indicates that all available info is not already reflected in prices
Efficient market hypothesis (EMH)
EMH: The notion that stocks prices already reflect all available information
Dramatic evidence of rapid response to new information found in intraday prices: Most of price
response to corporate dividend/earnings announcements occurs within 10 minutes
Competition as the source of efficiency
Why should we expect stock prices to reflect all available information?
If willing to spend time/money on gathering info, it might seem reasonable that you could turn
up something that has been overlooked by the rest of the investment community
When information is costly to uncover and analyze, one would expect investment analysis
calling for such expenditures to result in an increased expected return
The degree of efficiency differs across various markets
E.g., emerging markets that are less intensively analyzed than US markets and in which accounting disclosure requirements are less rigorous may be less efficient than US markets
Although it may not literally be true that all relevant information will be uncovered, competition
among many well-backed, highly paid, aggressive analysts ensures that, as a general rule, stock
prices ought to reflect available information regarding their proper levels
Versions of the efficient market hypothesis
Three versions of the EMH: (i) The weak, (ii) The semistrong, and (iii) The strong form
Versions differ by their notions of what is meant by the term all available information
The weak form hypothesis
Asserts that stock prices already reflect all information that can be derived by examining
market trading data such as the history of past prices, trading volume, or short interest
Implies that trend analysis is fruitless since past stock price data are publicly available/costless
The weak-form hypothesis holds that if such data ever conveyed reliable signals about future
performance, all investors already would have learned to exploit the signals
The semistrong form hypothesis
States that all publicly available info must be reflected already in the stock price
Such info includes, in addition to past prices, fundamental data on firms product line, quality
of management, balance sheet, patents held, earning forecasts, accounting practices
The strong form hypothesis
States that stock prices reflect all info, even info available only to company insiders
This version of the hypothesis is quite extreme
Few would argue with the proposition that corporate officers have access to pertinent info long
enough before public release to enable them to profit from trading on that info

43

Implications of the EMH

Technical analysis
Technical analysis is essentially the search for recurrent and predictable patterns in stock prices
Although technicians recognize the value of info regarding future economic prospects of the
firm, they believe that such info is not necessary for a successful trading strategy
If stock price responds slowly enough, can identify/exploit trend during adjustment period
Key to technical analysis: Slow response of prices to fundamental supply-and-demand factors
This prerequisite, of course, is diametrically opposed to the notion of an efficient market
Example of technical analysis - The relative strength approach
Chartist compares stock performance over recent period to market/stocks in same industry
Strength of a stock presumably may continue for long enough to offer profit opportunities
Resistance levels or support levels
Price levels above which it is difficult for stock prices to rise, or below which it is unlikely for
them to fall, and they are believed to be levels determined by market psychology
The efficient market hypothesis implies that technical analysis is without merit
Will a technical rule that seems to work continue to work in the future once widely recognized
Once discovered, a useful tech rule ought to fail invalidated when mass of traders exploit it
In this sense, price patterns ought to be self-destructing
Thus market dynamic: Continual search for profitable trading rules, followed by destruction
by overuse of successful rules, followed by more search for yet-undiscovered rules
Fundamental analysis
Uses earnings/dividend prospects, expectations of future interest rates, and risk evaluation of the
firm to determine proper stock prices. Represents an attempt to determine the PV of all the
payments a stockholder will receive from each share of stock
If value exceeds stock price, fundamental analyst recommends purchasing the stock
Fundamental analysts usually start with a study of past earnings/company balance sheets
Supplement analysis with detailed economic analysis, including evaluation of the firms management quality, firms standing within industry, and prospects for industry
The EMH predicts that most fundamental analysis also is doomed to failure
Only analysts with a unique insight will be rewarded
Fundamental analysis much more difficult than identifying well-run firms with good prospects
Trick is not to identify good firms, but firms better than everyone elses estimate
Similarly, poorly run firms can be great bargains if not as bad as stock prices suggest
Active versus passive portfolio management
Limitations of active portfolio management
Casual efforts to pick stocks are not likely to pay off
Only serious/uncommon analyses generate differential insight needed to yield profits
Moreover, these techniques are economically feasible only for managers of large portfolios
The small investor probably is better off investing in mutual funds. By pooling resources in this
way, small investors can gain from economies of scale
EMH proponents believe active management is wasted effort/unlikely to justify expenses
Therefore, they advocate a passive investment strategy (no attempt to outsmart market)
Passive strategy aims only at building well-diversified portfolio without attempting to find
under-/overvalued stocks, usually characterized by a buy-and-hold strategy
One common strategy for passive management is to create an index fund, which is a fund
designed to replicate the performance of a broad-based index of stocks
Exchange-traded funds (ETFs) are a (lower-expense) alternative to indexed mutual funds
These are shares in diversified portfolios that can be bought/sold like shares of individual stock
ETFs matching several broad stock market indexes are available to investors who want to hold
a diversified sector of a market without attempting active security selection
Hybrid strategy common: Fund maintains passive core, plus actively managed portfolio(s)
44

 The role of portfolio management in an efficient market

There is a role for rational portfolio management, even in perfectly efficient markets
Even if all stocks are priced fairly, firm-specific still need to be diversified away
Therefore, rational security selection, even in an efficient market, calls for the selection of a
well-diversified portfolio providing the systematic risk level that the investor wants
Rational investment policy also requires that tax considerations be reflected in security choice
Another argument for rational portfolio management relates to investors particular risk profile
Investors of varying ages need different portfolio policies with regard to risk bearing
Investors optimal positions vary with age, tax bracket, risk aversion, and employment
Role of portfolio manager is to tailor portfolio to these needs, rather than to beat the market
Resource allocation
Deviations from efficiency allow better-informed traders to profit from less-informed traders
However, deviations from informational efficiency would also result in a large cost that will be
borne by all citizens, namely, inefficient resource allocation
Capital market prices guide resource allocation because security mispricing could entail severe
social costs by fostering inappropriate investments on the real side of the economy
Corporations with overpriced stocks would obtain capital too cheaply, and corporations with
undervalued stocks might forgo investment opportunities because raising capital too expensive
Inefficient capital markets would diminish one of the most potent benefits of a market economy
Event studies
The notion of informationally efficient markets leads to a powerful research methodology
If security prices reflect all currently available info, then price changes must reflect new info
Measure the importance of an event by examining price changes when the event occurs
Event study
Technique of empirical financial research to assess impact of an event on firms stock price
Isolating the part of a stock price movement that is attributable to a specific event is not trivial
Starts with a proxy for what the stocks return would have been in absence of the event
Abnormal return due to event estimated as [stocks actual return] [this benchmark]
Many researchers use a market model to estimate abnormal returns (based on index models):
The stock return rt during a given period t would be expressed as:
rt = a + brM t + et

et = rt (a + brM t )

(1)

Firm-specific/abnormal return et interpreted as unexpected return resulting from event

Market model is highly flexible, it can be generalized to richer models of benchmark returns
E.g., by including industry as well as broad market returns on the right-hand side of Eq. (1),
or returns on indexes constructed to match characteristic such as firm size
However, intercept a/slope b must be estimated using data sufficiently separated in time from
the event so that they are not affected by event-period abnormal stock performance
We measure the impact of an event by estimating the abnormal return on a stock (or group of stocks)
at the moment the information about the event becomes known to the market
One concern that complicates event studies arises from leakage of information
Leakage occurs when info is released to a small group of investors before official public release
Any abnormal return on announcement date is then
P poor indicator of total impact of info release
Better indicator: Cumulative abnormal return ( abnormal returns over time period)
Event study has become a widely accepted tool to measure the economic impact of many events
E.g., SEC regularly uses event studies to measure illicit gains from insider trading
Event studies are also used in fraud cases, where the courts must assess damages
Are markets efficient?
The issues
EMH implies that most of portfolio managers activity is at best wasted effort, and probably
harmful to clients because it costs money and leads to imperfectly diversified portfolios
45

 The magnitude issue

As observers, we cannot statistically measure portfolio managers contribution
A 0.1% contribution would be swamped by the yearly volatility of the market
The selection bias issue
Only investors who find that an investment scheme cannot generate abnormal returns will be
willing to report their findings to the whole world
Selection bias: Outcomes we can observe are preselected in favor of failed attempts
Cannot fairly evaluate true ability of portfolio managers to generate winning strategies
The lucky event issue
Under the hypothesis that any stock is fairly priced given all available information, any bet on
a stock is simply a coin toss: There is equal likelihood of winning or losing the bet
However, if many investors using a variety of schemes make fair bets, statistically speaking,
some of those investors will be lucky and win a great majority of the bets
For every big winner, there may be many big losers, but we never hear of these managers
The proper test would be to see whether the successful investors can repeat their performance
in another period, yet this approach is rarely taken
Weak form tests: Patterns in stock returns
Returns over short horizons
To discern trends in stock prices, measure the serial correlation of stock returns
Serial correlation refers to the tendency for stock returns to be related to past returns
Positive serial correlation: Positive returns tend to follow positive returns (momentum)
Negative serial correlation: Positive returns tend to be followed by negative returns (a
reversal or correction property)
Studies demonstrate weak price trends over short periods, but no obvious trading opportunities
While broad market indexes demonstrate only weak serial correlation, there is stronger momentum in performance across market sectors exhibiting best/worst recent returns
While individual stocks performance is unpredictable, portfolios of best-performing stocks in
recent past outperform other stocks with enough reliability to offer profit opportunities
Returns over long horizons
Tests of long-horizon returns (multiyear periods) have found suggestions of pronounced negative
long-term serial correlation in the performance of the aggregate market
Fads hypothesis: The stock market might overreact to relevant news
Such overreaction leads to positive serial correlation (momentum) over short time horizons
Subsequent correction of the overreaction leads to negative serial correlation
Alternative interpretation: The market risk premium varies over time
Other studies: Over long horizons, extreme performance in particular securities reverse itself
Reversal effect: Losers rebound and winners fade back, suggests that the stock market
overreacts to relevant news. After the overreaction is recognized, performance is reversed
Predictors of broad market returns
Several studies have documented the ability of easily observed variables to predict market returns:
E.g., Fama and French showed that the return on the aggregate stock market tends to be
higher when the dividend/price ratio, the dividend yield, is high
Campbell and Shiller found that the earnings yield can predict market returns
Keim and Stambaugh showed that bond market data such as the spread between yields on
high- and low-grade corporate bonds also help predict broad market returns
On the one hand, they may imply that stock returns can be predicted, in violation of the EMH
More probably, however, these variables are proxying for variation in the market risk premium
Semistrong tests: Market anomalies
Market anomalies
Surprisingly, several easily accessible statistics, e.g. a stocks price-earnings ratio or its market
capitalization, seem to predict abnormal risk-adjusted returns

46

 Such findings difficult to reconcile with EMH Referred to as efficient market anomalies
A difficulty in interpreting these tests is that we usually need to adjust for portfolio risk before
evaluating the success of an investment strategy Tests of risk-adjusted returns are joint tests
of the efficient market hypothesis and the risk adjustment procedure
Example: Portfolios of low P/E stocks provide higher returns than high P/E portfolios
The P/E effect holds up even if returns are adjusted for portfolio beta
An interpretation of these results is that returns are not properly adjusted for risk
If two firms have same expected earnings, riskier stock will sell at lower price/lower P/E
Because of its higher risk, the low P/E stock also will have higher expected returns
Unless CAPM fully adjusts for risk, P/E acts as useful additional descriptor of risk, and
is associated with abnormal returns if CAPM is used to establish benchmark performance
The small-firm-in-January effect
Small firm effect: Average annual returns are consistently higher on small-firm portfolios
Of course, the smaller-firm portfolios tend to be riskier. But even when returns are adjusted
for risk using CAPM, there is still a consistent premium for the smaller-sized portfolios
Later studies showed that small-firm effect occurs virtually entirely in January (first 2 weeks)
The neglected-firm effect and liquidity effects
Interpretation of the small-firm-in-January effect:
Small firms neglected by large traders Info about smaller firms less available
This info deficiency makes smaller firms riskier investments that command higher returns
Thus, the neglected firm premium is not a market inefficiency, but a type of risk premium
The effect of liquidity on stock returns are related to both small-/neglected-firm effects
Investors demand a rate-of-return premium for less-liquid stocks (higher trading costs)
These stocks exhibit abnormally high risk-adjusted rates of return
Because small and less-analyzed stocks as a rule are less liquid, the liquidity effect might
be a partial explanation of their abnormal returns
Book-to-market ratios
A powerful predictor of returns across securities is the firms book/market ratio
The dramatic dependence of returns on book-to-market ratio is independent of beta, suggesting
either that high book-to-market ratio firms are relatively underpriced, or that the book-tomarket ratio is serving as a proxy for a risk factor that affects equilibrium expected returns
Fama and French found that after controlling for the size and book-to-market effects, beta
seemed to have no power to explain average security returns
This finding is an important challenge to the notion of rational markets, because implies that
a factor that should affect returns - systematic risk - seems not to matter, while a factor that
should not matter - book-to-market ratio - seems capable of predicting future returns
Post-earnings-announcement price drift
Fundamental principle of efficient markets: Any new info is reflected in prices very rapidly
Puzzle: Sluggish response of prices to firms earnings announcements
The News content of announcement is evaluated by comparing announcement of actual
earnings to value previously expected by market participants. Difference = earnings surprise
The correlation between ranking by earnings surprise and abnormal returns is as predicted:
Large abnormal return (jump in cumulative abnormal return) on announcement day (t = 0)
Abnormal return > 0 for positive-surprise firms and < 0 for negative-surprise firms
More interesting result concerns stock price movement after announcement date:
Cumulative abnormal returns of positive-surprise stocks continue to rise (momentum) even
after info public, and negative-surprise firms keep suffering negative abnormal returns
Market adjusts to earnings info gradually Sustained period of abnormal returns
Strong form tests: Inside information
It would not be surprising if insiders were able to make superior profits trading in their firms stock
We do not expect markets to be strong-form efficient

47

 Tendency for prices to rise after insiders intensively bought shares/fall after intensive insider sales
Although there is some tendency for stock prices to increase even after the Official Summary reports
insider buying, abnormal returns are not of sufficient magnitude to overcome transaction costs
Interpreting the evidence
Risk premiums or inefficiencies?
The price-earnings, small-firm, market-to-book, momentum, and long-term reversal effects are
currently among the most puzzling phenomena in empirical finance
The feature that small firms, low-market-to-book firms, and recent losers seem to have
in common is a stock price that has fallen considerably recently This group may contain
a high proportion of distressed firms that have suffered recent difficulties
Fama and French argue that these effects can be explained as manifestations of risk premiums:
Using their three-factor model, they show that stocks with higher betas on size or marketto-book factors have higher average returns
They interpret these returns as evidence of a risk premium associated with the factor
While size or book-to-market ratios per se are obviously not risk factors, they perhaps might
act as proxies for more fundamental determinants of risk
Fama and French argue that these patterns of returns may be consistent with an efficient
market in which expected returns are consistent with risk
The opposite interpretation argues that these phenomena are evidence of inefficient markets,
more specifically, of systematic errors in the forecasts of stock analysts
Analysts extrapolate past performance too far into the future, and therefore overprice firms
with recent good performance and underprice firms with recent poor performance
Ultimately, when market participants recognize their errors, prices reverse
This explanation is consistent with the reversal effect and also, to a degree, is consistent
with the small-firm and book-to-market effects because firms with sharp price drops may
tend to be small or have high book-to-market ratios
Analysts seem overly pessimistic about firms with low growth prospects and overly optimistic about firms with high growth prospects
Anomalies or data mining?
If one reruns the database of past returns over and over and examines stock returns along
enough dimensions, simple chance will cause some criteria to appear to predict returns
Some anomalies have not shown much staying power after being reported in the literature
Real puzzle: Value stocks-defined by low P/E ratio, high book-to-market ratio, or depressed prices
vs. historic levels have provided higher average returns than glamour/growth stocks
The noisy market hypothesis and fundamental indexing
EMH argues for capitalization-weighted indexed portfolios (broad diversification, low trading costs)
Some argue that cap-weighted indexing is inferior to fundamental indexing
The rational for their argument goes by the name noisy market hypothesis
The hypothesis begins with the observation that market prices may well contain pricing errors or
noise relative to the intrinsic or true value of a firm
Because indexed portfolios invest in proportion to market capitalization, portfolio weights will track
these pricing errors, with greater amounts invested in overpriced stocks (which have poor expected
returns) and lesser amounts invested in underpriced stocks (which have high expected returns)
Conclusion: Cap-weighted strategy overweights precisely the firms with worst return prospects
In contrast, a fundamental index that invests intrinsic value avoids detrimental association
between portfolio weights and markets pricing errors Outperform cap-weighted index
How can we find the intrinsic values necessary to form a fundamental index?
Necessary inputs: True stock values, market pricing errors (holy grail of active managers)
Advocates of fundamental indexing propose that portfolio weights determined by indicators of
intrinsic value such as dividends/earnings
Despite its name, it is not indexing, but rather a form of active investing

48

Mutual fund and analyst performance

Issue of market efficiency reduces to whether skilled investors can make consistent abnormal profits
Stock market analysts
Analysts have tended to be overwhelmingly positive in their assessment of the prospects of firms
We cannot take positive recommendations at face value
Focuses on changes in analysts recommendations:
Positive changes 5% increased stock prices, negative changes 11% price decreases
One might wonder whether these price changes reflect markets recognition of analysts superior
info or, simply result from new buy/sell pressure brought on by recommendations
Price impact seems permanent Consistent with hypothesis that analysts do reveal new info
Portfolio strategies based on analyst consensus recommendations would result in extremely heavy
trading activity with associated costs that would wipe out potential profits from the strategy
Mutual fund managers
Evidence does not support claim that managed portfolios can consistently beat the market
Revisit the question of mutual fund performance, paying more attention to the benchmark against
which performance ought to be evaluated
Style choice, i.e. the exposure to small stocks (asset allocation decision) dominates the
evaluation of performance but has little to do with managers stock-picking ability
After controlling for the relative performance of three asset classes - large stocks, small stocks,
and bonds - mutual fund managers as a group do not demonstrate an ability to beat passive
index strategies that would simply mix index funds from among these asset classes
Conventional performance benchmark today is a 4-factor model, using the three Fama-French
factors (return on market index, and returns to portfolios based on size and book-to-market ratio)
augmented by a momentum factor (portfolio based on prior-year stock return)
Results show that distribution of s is roughly bell shaped, with slightly negative mean
On average, these funds do not outperform their style-adjusted benchmarks
Issue of consistency in mutual fund performance: The hot hands phenomenon
After controlling for the 4 factors, small persistence in relative performance across managers
Seems due to expenses/transactions costs rather than gross investment returns
However, in a fully efficient market, there can be consistently inferior performers
Repeated weak performance not due to ability to pick bad stocks consistently but results from
consistently high expense ratio/portfolio turnover, or higher-than-avg. transaction costs
There have been few studies of the performance of bond fund managers
On average, bond funds underperform passive fixed-income indexes by an amount roughly
equal to expenses. No evidence that past performance can predict future performance
Performance of professional managers broadly consistent with market efficiency. Amounts by which
managers as a group beat/are beaten by market is within margin of statistical uncertainty
Survivor bias in mutual fund studies
Studies of mutual fund performance can be affected by survivorship bias, the tendency for less
successful funds to go out of business over time, thus leaving the sample
Can give rise to the appearance of persistence in performance, even if there is none in reality
Data to assess performance of managers must be free of survivorship bias. Many are not
Are markets efficient?
An overly doctrinaire belief in efficient markets can paralyze the investor and make it appear that
no research effort can be justified. This extreme view is probably unwarranted. Enough anomalies
in the empirical evidence to justify the search for underpriced securities that clearly goes on
The bulk of the evidence, however, suggests that any supposedly superior investment strategy
should be taken with many grains of salt. The market is competitive enough that only differentially
superior information or insight will earn money
Conclusion: Markets very efficient, but rewards to the very diligent/creative may be waiting

49

50

BKM - Ch. 12: Behavioral finance and technical analysis

Introduction
Behavioral finance argues that the literature on trading strategies has missed a larger/more important
point by overlooking the first implication of efficient markets: The correctness of security prices
Behavioral school argues that even if security prices are wrong, it still is difficult to exploit them
Failure to uncover obviously successful trading rules/traders cannot be proof of market efficiency
Conventional theories presume that investors are rational, behavioral finance assumes they may not be
The behavioral critique
Introduction
The premise of behavioral finance is that conventional financial theory ignores how real people
make decisions and that people make a difference. Irrationalities fall into two broad categories:
1. Investors do not always process information correctly and therefore infer incorrect probability
distributions about future rates of return
2. Even given distribution of returns, they often make inconsistent/suboptimal decisions
Existence of irrational investors not by itself sufficient to render capital markets inefficient
If such irrationalities did affect prices, then sharp-eyed arbitrageurs taking advantage of profit
opportunities might be expected to push prices back to their proper values
Thus, the second leg of the behavioral critique is that in practice the actions of such arbitrageurs
are limited and therefore insufficient to force prices to match intrinsic value
If behaviorists are correct about limits to arbitrage activity, then the absence of profit opportunities
does not necessarily imply that markets are efficient
Information processing
Errors in information processing can lead investors to misestimate the true probabilities of possible
events or associated rates of return Four of the more important biases:
Forecasting errors
Too much weight to recent experience vs. prior beliefs when forecasting (memory bias)
People make forecasts too extreme given the uncertainty inherent in their information
E.g., the P/E effect can be explained by earnings expectations that are too extreme
Overconfidence
People overestimate precision of their beliefs/forecasts, and they overestimate their abilities
Such overconfidence may be responsible for prevalence of active versus passive investment
E.g., compare trading activity/average returns in brokerage accounts of men/women:
Men (in particular single men) trade far more actively than women, consistent with the
greater overconfidence among men well documented in the psychology literature
Trading activity highly predictive of poor performance: Top 20% of accounts by turnover
had avg. returns 7% pts lower than the 20% with lowest turnover
Conservatism
A conservatism bias means that investors are too slow (too conservative) in updating their
beliefs in response to new evidence Momentum in stock market returns
Sample size neglect and representativeness
Notion of representativeness: People commonly do not take into account sample size, apparently reasoning that a small sample is just as representative of a population as a large one
Such a pattern would be consistent with overreaction and correction anomalies
Behavioral biases
Even if information processing were perfect, many studies conclude that individuals would tend to
make less-than-fully-rational decisions using that information
These behavioral biases largely affect how investors frame questions of risk versus return
Framing
Decisions seem to be affected by how choices are framed
Individuals may act risk averse in terms of gains but risk seeking in terms of losses
51

 Mental accounting
Mental accounting is a specific form of framing in which people segregate certain decisions
E.g., an investor takes lots of risk with one account, but very conservative with another
Rationally, it might be better to view both accounts as part of the investors overall portfolio
with the risk-return profiles of each integrated into a unified framework
Mental accounting is consistent with some investors irrational preference for stocks with high
cash dividends and with a tendency to ride losing stock positions for too long
Mental accounting effects also can help explain momentum in stock prices
The house money effect: Gamblers more willing to accept bets if currently ahead
Analogously, after a stock market run-up, individuals may view investments as largely
funded out of a capital gains account
Regret avoidance
Psychologists have found that individuals who make decisions that turn out badly have more
regret (blame themselves more) when that decision was more unconventional
E.g., buying a blue-chip that turns down not as painful as same losses on start-up firm
Regret avoidance is consistent with both the size and book-to-market effect
Prospect theory
Modifies analytic description of rational risk-averse investors of standard financial theory
Prospect theory proposes competing description of preferences characterized by loss aversion.
Utility depends not on level of wealth, but on changes in wealth from current levels
Utility function in prospect theory always recenters on current wealth, ruling out such decreases
in risk aversion and helping to explain high avg. historical equity risk premiums
Moreover, the convex curvature to the left of the origin will induce investors to be risk seeking
rather than risk averse when it comes to losses (see Fig. 2)
Utility function

Utility function

Utility

3
2
1
Utility

2.0

-20

-10

10

20

-1
1.0
-2
-3
0.0
0

10

Wealth

20

Change in wealth

Figure 1: Conventional utility function

Figure 2: Utility function under prospect theory

Limits to arbitrage
Behavioral biases would not matter if arbitrageurs could exploit mistakes of behavioral investors
Behavioral advocates argue that in practice, several factors limit ability to profit from mispricing
Fundamental risk
Suppose that a share of IBM is underpriced. Buying it may present a profit opportunity, but
it is hardly risk-free, because the presumed market underpricing can get worse
Prices converge to intrinsic value, but may not happen until after traders horizon
Fundamental risk incurred in exploiting profit opportunity Limit traders activity
Implementation costs
Exploiting overpricing can be particularly difficult: Short selling a security entails costs
Short-sellers may have to return the borrowed security on little notice Uncertain horizon
Many pension/mutual fund managers face strict limits on their discretion to short securities
This can limit the ability of arbitrage activity to force prices to fair value
52

 Model risk
One always has to worry that an apparent profit opportunity is more apparent than real
Mispricing may make a position a good bet, but it is still a risky one
Limits to arbitrage and the law of one price
The Law of One Price
Effectively identical assets should have identical prices
Should be satisfied in rational markets
Yet there are several instances where the law seems to have been violated
Siamese Twin companies
In 1907, Royal Dutch Petroleum and Shell Transport merged their operations into one firm.
The two companies, continuing to trade separately, agreed to split all profits on a 60/40 basis
Expect Royal Dutch to sell for exactly 60/40 = 1.5 [price of Shell]. This was not the case
This arbitrage opportunity posed fundamental risk
Equity carve-outs
Several equity carve-outs also have violated the Law of One Price
In 1999, 3Com decided to spin off Palm. It first sold 5% of its stake in Palm in an IPO,
announcing that it would distribute the remaining 95% of its Palm shares to 3Com shareholders
6 months later. Each 3Com shareholder would receive 1.5 shares of Palm
Once Palm began trading, prior to spinoff, 3Com should have been at least 1.5 Palm. Each
share of 3Com entitled to 1.5 shares of Palm plus ownership stake in profitable company
Instead, Palm shares at the IPO actually sold for more than the 3Com shares
Again, an arbitrage strategy seems obvious. Why not buy 3Com and sell Palm?
The limit to arbitrage in this case was the inability of investors to sell Palm short. Virtually
all available shares in Palm were already borrowed and sold short
Closed-end funds
Closed-end funds often sell for substantial discounts or premiums from net asset value
This is nearly a violation of the Law of One Price, because one would expect the value of
the fund to equal the value of the shares it holds
Nearly because, in practice, there are a few wedges between the value of the closed-end fund
and its underlying assets: One is expenses
Others argue that patterns of discounts/premiums on closed-end funds driven by changes in
investor sentiment. One might consider buying funds selling at discount/selling those trading
at premium, but discounts/premiums can widen, subjecting strategy to fundamental risk
Closed-end fund discounts: Example of anomalies that may have rational explanations
Can reconcile with rational investors even if expenses/fund abnormal returns are modest
If fund has dividend yield , risk-adjusted abnormal return , and expense ratio , then
using constant-growth DDM, fund premium over its net asset value is:
Price - NAV

=
N AV
+
If fund performance exceeds expenses (i.e., if > ), fund sells at premium
Bubbles and behavioral economics
Irrational exuberance
In 6-year period beginning in 1995, NASDAQ index increased by a factor of more than 6
Greenspan famously characterized dot-com boom as example of irrational exuberance
Assessment correct: 10/02, index fell to < 1/4 of peak value reached 2.5 years earlier
Behavioral school: Example of market moved by irrational investors
Moreover, in accord with behavioral patterns, as dot-com boom developed, it fed on itself, with
investors increasingly confident of their investment prowess (overconfidence bias) and willing
to extrapolate short-term patterns into distant future (representativeness bias)
On the other hand, bubbles are a lot easier to identify as such once they are over
53

 At the time, boom was justified as consistent with glowing forecasts for new economy
Hard to tie down the fair value of stock investments
Using constant-growth DDM, the estimate is highly sensitive to the input values: Even a small
reassessment of the expected dividend growth leads to very different stock valuation
Evaluating the behavioral critique
Behavioral literature is largely silent on whether there is money to be made from mispricing
Many believe that behavioral approach is too unstructured, allowing any anomaly to be explained
by combination of irrationalities chosen from laundry list of behavioral biases
Critics would like a consistent/unified behavioral theory that explains range of anomalies
Anomalies are inconsistent in terms of their support for one type of irrationality versus another
Moreover, the statistical significance of many of these results is less than meets the eye
Small errors in choosing benchmark cumulate to large apparent abnormalities in long-term
Technical analysis and behavioral finance
Introduction
Technical analysis attempts to exploit recurring and predictable patterns in stock prices
Technicians value fundamental info, but believe prices only gradually close in on intrinsic value
As fundamentals shift, astute traders can exploit the adjustment to a new equilibrium
Disposition effect
The tendency of investors to hold on to losing investments
Behavioral investors seem reluctant to realize losses
Disposition effect leads to momentum in prices even if fundamental values follow random walk
Behavioral biases may also be consistent with technical analysts use of volume data
Important behavioral trait: Overconfidence, systematic tendency to overestimate ones abilities
As traders become overconfident, they trade more Relation trading volume/market returns
Technical analysis thus uses volume data as well as price history to direct trading strategy
Technicians also believe that markets can be perturbed by behavioral factors (sentiment variables)
Trends and corrections
Much of technical analysis seeks to uncover trends in market prices
This is in effect a search for momentum
Momentum can be absolute (search for upward price trends), or relative (look to invest in one
sector over another, or take on a long-short position in the two sectors)
Relative strength statistics are designed to uncover these potential opportunities
Dow theory
Grandfather of trend analysis: Posits three forces simultaneously affecting stock prices:
1. Primary trend = Long-term movement of prices, lasting for several months/years
2. Secondary/intermediate trends: Caused by short-term deviations of prices from underlying
trend line. Such deviations are eliminated when prices revert back to trend values
3. Tertiary/minor trends: Daily fluctuations of little importance
Dow theory based on notion of predictably recurring price patterns. Yet EMH holds that if a
pattern is exploitable, many investors attempt to profit from such predictability, which would
ultimately move stock prices and cause the strategy to selfdestruct
Can we see trend only after the fact? Recognizing patterns as they emerge is very difficult
Recent variations on Dow theory: Elliott wave theory and theory of Kondratieff waves
Like Dow theory, idea behind Elliott waves: Stock prices described by set of wave patterns
Long-term/short-term wave cycles are superimposed and result in complicated pattern of
price movements, but by interpreting the cycles, one can predict broad movements
Similarly, Kondratieff waves come from Russian economist who asserted that the macroeconomy (hence the stock market) moves in broad waves lasting between 48/60 years
Moving averages
Moving average of a stock index: Average level of the index over a given interval of time
After a period in which prices have been falling, the moving average is above current price
54

 When prices have been rising, the moving average will be below the current price
When the market price breaks through the moving average line from below, it is taken as a
bullish signal because it signifies a shift from a falling trend to a rising trend
Conversely, when prices fall below the moving average, its considered time to sell
Two popular measures are 200-day and 53-week moving averages
Breadth
The breadth of the market is a measure of the extent to which movement in a market index is
reflected widely in the price movements of all the stocks in the market
Most common measure of breadth: Spread between # of stocks that advance/decline in price
If advances outnumber declines by wide margin, market viewed as stronger (widespread rally)
Analysts might use a moving average of cumulative breadth to gauge broad trends
Sentiment indicators
Trin statistic
Market volume is sometimes used to measure the strength of a market rise or fall
Level of investor participation in market advance/retreat measures the significance of the move
The trin statistic is defined as:
Trin =

Volume declining / Number declining

Volume advancing / Number advancing

Therefore, trin = ratio of avg. volume in declining issues to avg. volume in advancing issues
Ratios above 1.0 are considered bearish because the falling stocks would then have higher
average volume than the advancing stocks, indicating net selling pressure
Confidence index
Barrons computes a confidence index using data from the bond market. The presumption is
that actions of bond traders reveal trends that will emerge soon in the stock market
The confidence index is the ratio of the average yield on 10 top-rated corporate bonds divided
by the average yield on 10 intermediate-grade corporate bonds
Ratio always below 100% (higher rated bonds offer lower promised YTM)
When bond traders optimistic, they require smaller default premiums on lower rated debt
Hence, the yield spread will narrow, and the confidence index will approach 100%
Higher values of the confidence index are bullish signals
Put/call ratio
The ratio of outstanding put options to outstanding call options is called the put/call ratio
Typical put/call ratio around 65%. Because put options do well in falling markets while call
options do well in rising markets, deviations of the ratio from historical norms are considered
to be a signal of market sentiment Predictive of market movements
Technicians see increase in ratio as bearish (interest in puts hedge against market declines)
A rising ratio is taken as a sign of broad investor pessimism and a coming market decline
Contrarians: Good time to buy when rest of market bearish (stock prices unduly depressed)
They take an increase in the put/call ratio as a signal of a buy opportunity
A warning
Search for patterns is irresistible, and human eye ability to discern patterns is remarkable
Unfortunately, it is possible to perceive patterns that really dont exist
Possible to simulate price path exhibiting patterns, constructed from patternless returns
A problem related to the tendency to perceive patterns where they dont exist is data mining
After the fact, always find patterns/rules that would have generated huge profits
If you test enough rules, some will have worked in the past
Unfortunately, picking a theory that would have worked carries no guarantee of future success

55

56

BKM - Ch. 13: Empirical evidence on security returns

Introduction
Many implications of CAPM/APT have already been accepted in a wide range of applications:
1. Professional portfolio managers use the expected return-beta relationship of security returns. Furthermore, many firms rate the performance of portfolio managers according to reward-to-volatility
ratios they maintain and avg. rates of return they realize relative to the CML/SML
2. Regulatory commissions use the expected return-beta relationship along with forecasts of the market index return as one factor in determining the cost of capital for regulated firms
3. Court rulings on torts cases sometimes use the expected return-beta relationship to determine
discount rates to evaluate claims of lost future income
4. Many firms use the SML to obtain a benchmark hurdle rate for capital budgeting decisions
The financial community has passed a favorable judgment on CAPM/APT, if only implicitly
CAPM - Theoretical construct:
Difficult to test CAPM because expected returns/covariances/market portfolio are unobservable
Most tests conducted with highly diversified portfolios rather than on individual securities
Tests directed at CAPM actually have been more suitable to establish validity of APT
It is more important to distinguish the empirical work on the basis of the factor structure that is assumed
or estimated than to distinguish between tests of the CAPM and the APT
The index model and the single-factor APT
The expected return-beta relationship
If the expected return-beta relationship holds with respect to an observable ex ante efficient index
M , the expected rate of return on any security i is:
E(ri ) = rf + i [E(rM ) rf ]

2
where i = Cov(ri , rM )/M

(1)

Early simple tests followed three basic steps: (i) Establishing sample data, (ii) Estimating the SCL
(security characteristic line), and (iii) Estimating the SML (security market line)
Setting up the sample data
Determine a sample period of, e.g., 60 monthly holding periods
Collect returns on 100 stocks, market portfolio proxy (S&P 500), and 1-month T-bills
The data constitutes a table of 102 60 = 6, 120 rates of return
rit
rM t
rf t

=
=
=

Returns on 100 stocks over 60-month period; i = 1-100, and t = 1-60

Returns on the S&P 500 index over the sample period
Risk-free rate each month

Estimating the SCL

View Eq. (1) as a security characteristic line (SCL)
For each stock i, we estimate the beta coefficient as the slope of a first-pass regression equation:
rit rf t = ai + bi (rM t rf t ) + eit
ri rf
bi
rM rf
2 (ei )

=
=
=
=

Sample avg. (60 observations) of excess return on each stock

Sample estimates of the beta coefficients of each of the 100 stocks
Sample average of the excess return of the market index
Estimates of the variance of the residuals for each of the 100 stocks

Estimating the SML

Now view Eq. (1) as security market line (SML) with 100 observations for stocks in sample
Estimate 0 /1 in 2nd-pass regression with estimates bi from 1st-pass as independent variable:
ri rf = 0 + 1 bi

i = 1, , 100

If the CAPM is valid, then 0 = 0 and 1 = rM rf

57

(2)

 The key property of the expected return-beta relationship described by SML is that expected
excess return on securities is determined only by systematic risk (measured by ) and should
be independent of nonsystematic risk, measured by the variance of residuals 2 (ei )
These estimates can be added as a variable in Eq. (2) of an expanded SML:
ri rf = 0 + 1 bi + 2 2 (ei )

(3)

This 2nd-pass regression is estimated with hypotheses: 0 = 0, 1 = rM rf , and 2 = 0

The hypothesis that 2 = 0 is consistent with the notion that nonsystematic risk should not
be priced, i.e. there is no risk premium earned for bearing nonsystematic risk
More generally, according to the CAPM, the risk premium depends only on beta
Any additional RHS variable in Eq. (3) beyond beta should have a coefficient that is
insignificantly different from zero in the second-pass regression
Tests of the CAPM
Early tests of CAPM by Lintner, and replicated by Miller/Scholes used annual data on 631 NYSE
stocks for 1954-63. Results inconsistent with CAPM:
The estimated SML is too flat: The 1 coefficient is too small
SML intercept 0 (should be zero) is 20 its standard error
Two-stage procedure is straightforward, and rejection of CAPM using this approach is disappointing
However, there are several difficulties with this approach:
1. Stock returns are extremely volatile, which lessens the precision of any tests of average return
2. The market index used in the tests is surely not the market portfolio of the CAPM
3. In light of asset volatility, security s from 1st-stage regressions are estimated with substantial
sampling error and therefore cannot readily be used as inputs to 2nd-stage regression
4. Investors cannot borrow at the risk-free rate, as assumed by the simple version of CAPM
The market index
In what has become known as Rolls critique, Richard Rolls pointed out that:
1. Single testable CAPM hypothesis: Market portfolio is mean-variance efficient
2. All other implications follow from market portfolios efficiency Not independently testable
3. In any sample of individual returns, there are an infinite number of ex post M-V efficient
portfolios using sample-period returns and covariances. Sample s calculated between each
such portfolio/individual assets are exactly linearly related to sample avg. returns
If s are calculated against such portfolios, they will satisfy the SML relation exactly
whether or not the true market portfolio is mean-variance efficient in an ex ante sense
4. CAPM not testable unless we know exact composition of true market portfolio
The theory is not testable unless all individual assets are included in the sample
5. Using a proxy such as the S&P 500 for the market portfolio is subject to two difficulties:
(a) Proxy itself might be M-V efficient even when true market portfolio is not. Conversely,
proxy may be inefficient, but true market portfolio could still be efficient
(b) Most reasonable market proxies will be very highly correlated with each other and with
the true market portfolio whether or not they are mean-variance efficient
This problem is referred to as benchmark error, because it refers to the use of an incorrect
benchmark (market proxy) portfolio in the tests of the theory
Roll and Ross expanded Rolls critique, arguing that tests that reject positive relationship between
avg. return/ point to inefficiency of market proxy, rather than refuting CAPM relationship
Plausible that even highly diversified portfolios, such as value-/equally weighted portfolios of
all stocks in the sample, fail to produce significant avg. return- relationship
Kandel and Stambaugh considered the properties of the usual two-pass test of the CAPM in an
environment in which borrowing is restricted but the zero-beta version of the CAPM holds
The expected return-beta relationship describes the expected returns on a stock i, a portfolio
E on the efficient frontier, and that portfolios zero-beta companion Z:
E(ri ) E(rZ ) = i [E(rE ) E(rZ )]
58

(4)

 Kandel and Stambaugh asked what would happen if we followed the common procedure of
using market proxy portfolio M in place of E, and used the more efficient generalized least
squares regression in estimating 2nd-pass regression for zero-beta version of CAPM:
ri rZ = 0 + 1 (Estimated i )
Results: Estimates of 0 /1 biased by term relative efficiency of market proxy
2nd-pass regression provides a poor test of CAPM if market proxy is not efficient
Unfortunately, it is difficult to tell how efficient our market index is relative to the theoretical
true market portfolio, so we cannot tell how good our tests are
Measurement error in beta
Suppose that we could get past Rolls problem by obtaining data on the returns of the true market
portfolio. We still would have to deal with the statistical problems caused by measurement error
in the estimates of beta from the first-stage regressions
Well known in statistics that if RHS variable of a regression is measured with error (here ), then
the slope coefficient of the regression is biased downward and the intercept biased upward
Consistent with 0 higher than predicted by CAPM and 1 lower than predicted
Miller and Scholes confirmed these arguments in a well-controlled simulation:
They constructed stock returns obeying SML/agreeing with CAPM using a random generator
They used those simulated rates of returns in a test of CAPM as if they were from real data
They obtained the same test results as with real data, i.e. a rejection of CAPM
Next wave of tests designed to overcome the measurement error problem
Innovation (Black, Jensen, and Scholes): Use portfolios rather than individual securities
Combining securities into portfolios diversifies away most of the firm-specific part of returns, thereby
enhancing precision of estimates and expected rate of return of the portfolio
This mitigates the statistical problems that arise from measurement error in the beta estimates
However, combining stocks into portfolios reduces the number of observations left for 2nd-pass
To get best of trade-off: Construct portfolios with largest possible dispersion of coefficients
Rather than allocate 20 stocks to each portfolio randomly, we can rank portfolios by betas
Widely spaced betas will yield reasonably powerful tests of the SML
Fama and MacBeth used this to verify that the observed relationship between avg. excess returns and
is indeed linear and that nonsystematic risk does not explain avg. excess returns
Using 20 portfolios constructed according to the Black, Jensen, and Scholes methodology, Fama
and MacBeth expanded the estimation of the SML equation to include:
The square of coefficient (test for linearity of relationship between returns and s)
The estimated std. dev. of the residual (test for explanatory power of nonsystematic risk)
For a sequence of many subperiods, they estimated for each subperiod the equation:
ri = 0 + 1 i + 2 i2 + 3 (ei )

(5)

According to CAPM, both 2 /3 should have zero coefficients in 2nd-pass regression (confirmed)
With respect to the expected return-beta relationship, however, the picture is mixed:
The estimated SML is too flat, consistent with previous studies
1 is, on average, less than rM rf
On the + side, difference does not appear to be significant CAPM not clearly rejected
In conclusion, these tests of the CAPM provide mixed evidence on the validity of the theory:
1. The insights that are supported by the single-factor CAPM and APT are:
(a) Expected rates of return are linear and increase with beta, the measure of systematic risk
(b) Expected rates of return are not affected by nonsystematic risk
2. The single-variable expected return-beta relationship predicted by either the risk-free rate or the
zero-beta version of the CAPM is not fully consistent with empirical observation
CAPM qualitatively correct ( matters, (ei ) does not), but tests invalidate quantitative predictions
59

 The EMH and the CAPM

Roll: CAPM and the expected return-beta relationship follow directly from efficiency of the market
portfolio If we establish that market portfolio is efficient, we prove CAPM equation
Proxies for market portfolio (S&P 500, NYSE index) have proven hard to beat by investors
Strongest evidence for the empirical content of the CAPM and APT
Accounting for human capital and cyclical variations in asset betas
Two important deficiencies of the tests of the single-index models:
1. Not all assets traded in capital markets. Most important nontraded asset is human capital
2. There is ample evidence that asset betas are cyclical
One of the CAPM assumptions is that all assets are traded and accessible to all investors
Mayer proposed a version of the CAPM that accounts for a violation of this assumption
This requires an additional term in the expected return-beta relationship
An important nontraded asset is human capital
The value of future wages and compensation for expert services is a significant component of
the wealth of investors who expect years of productive careers prior to retirement
Moreover, it is reasonable to expect that changes in human capital are far less than perfectly
correlated with asset returns, and hence they diversify the risk of investor portfolios
Jagannathan and Wang used a proxy for changes in the value of human capital based on the rate
of change in aggregate labor income
In addition to the standard security betas estimated using the value-weighted stock market
index vw , they also estimated s of assets with respect to labor income growth labor
They considered the possibility that business cycles affect asset betas
They used the difference between the yields on low-/high-grade corporate bonds as proxy for
state of business cycle and estimate asset s relative to business cycle variable prem
With estimates of these three s for several stock portfolios, they estimated a 2nd-pass regression which includes firm size (market value of equity, denoted M E):
E(Ri ) = c0 + csize log(M E) + cvw vw + cprem prem + clabor labor

(6)

Jagannathan/Wang chose indirect way to add cyclical component to expected return-beta

relationship, as default premium can be very different from expected return on market index
Eq. (6) combines the labor factor that would be a natural ICAPM factor with two other factors
(size and the default premium) that might be justified on APT grounds
Jagannathan and Wang test their model with 100 portfolios that are designed to spread securities on the basis of size and beta. Stocks are sorted to 10 size portfolios, and the stocks
within each size decile are further sorted by beta into 10 subportfolios
Size explains variations in average returns quite well while beta does not
Substituting the default premium and labor income for size results in a similar increase in
explanatory power, but the CAPM expected return-beta relationship is not redeemed
Default premium is significant, while labor income is borderline significant. When we add size
as well, we find it is no longer significant and only marginally increases explanatory power
Conclusions from the Jagannathan and Wang study
1. Conventional 1st-pass estimates of security betas are greatly deficient. They do not fully
capture cyclicality of returns and thus do not accurately measure systematic risk of stock
2. Human capital important in any version of CAPM that better explains systematic risk
Accounting for nontraded business
Heaton and Lucas estimate the importance of proprietary business in portfolio choice
We expect that private-business owners will reduce demand for traded securities that are
positively correlated with their specific entrepreneurial income
If this effect is sufficiently important, aggregate demand for traded securities will be determined
in part by the covariance with aggregate noncorporate business income
Higher risk premium on securities with high covariance with noncorporate business income
60

 Consistent with theory, Heaton and Lucas find that households with higher investments in private
business do in fact reduce the fraction of total wealth invested in equity
In their regression analysis, allocation of overall portfolio to stocks is the dependent variable
The share of private business in total wealth (aka relative business) receives negative and
statistically significant coefficients
Risk-attitude (self-reported risk aversion) also receives a negative/significant coefficient
Heaton/Lucas extend Jagannathan/Wangs equation to include rate of change in proprietarybusiness wealth This variable also is significant and improves explanatory power of regression
Tests of multifactor CAPM and APT
Multifactor CAPM/APT are elegant theories, but provide little guidance concerning which factors
(sources of risk) ought to result in risk premiums. A test would require three stages:
1. Specification of risk factors
2. Identification of portfolios that hedge these fundamental risk factors
3. Test of the explanatory power and risk premiums of the hedge portfolios
A macro factor model
Chen, Roll, and Ross identify several possible variables that might proxy for systematic factors:
IP
EI
UI
CG
GB

=
=
=
=
=

Growth rate in industrial production

Changes in expected inflation measured by changes in short-term (T-bill) rates
Unexpected inflation = Actual expected inflation
Unexpected changes in RP = Returns on corporate Baa bonds long-term gov. bonds
Unexpected changes in the term premium = Returns on long- short-term gov. bonds

By using these factors, they implicitly assumed that factor portfolios exist that proxy for factors
A critical part of the methodology is the grouping of stocks into portfolios
In single-factor tests, portfolios constructed to span wide range of s to enhance power of test
In a multifactor framework the efficient criterion for grouping is less obvious
Chen, Roll, and Ross chose to group sample stocks into 20 portfolios by size (market value)
They first used 5 years of monthly data to estimate the factor betas of the 20 portfolios in a
first-pass regression: For each portfolio, estimate:
r = a + M rM + IP IP + EI EI + U I U I + CG CG + GB GB + e

(7)

Used as market index both value-weighted NYSE (VWNY) and equal weight NYSE (EWNY)
Using the 20 sets of first-pass estimates of factor betas as the independent variables, they now
estimated the second-pass regression (with 20 observations, one for each portfolio):
r = 0 + M M + IP IP + EI EI + U I U I + CG CG + GB GB + e

(8)

Where the gammas become estimates of the risk premiums on the factors
Chen, Roll, and Ross ran this second-pass regression for every month of their sample period,
reestimating the first-pass factor betas once every 12 months. The estimated risk premiums (the
values for the parameters ) were averaged over all the second-pass regressions
Industrial production (IP ), the risk premium on corporate bonds (CG), and unanticipated inflation
(U I) are the factors that appear to have significant explanatory power
The Fama-French three-factor model
Systematic factors in Fama-French model: Firm size, book-to-market ratio and market index
These additional factors are empirically motivated by the observations that historical average returns on stocks of small firms and on stocks with high ratios of book equity to market equity (B/M)
are higher than predicted by the security market line of the CAPM
This suggests that size or book-to-market ratio may be proxies for exposures to sources of systematic
risk not captured by CAPM and results in return premiums associated with these factors
They measure the size factor in each period as differential return on small vs. large firms
61

 Similarly, the other extra-market factor is typically measured as the return on firms with high book-tomarket ratios minus that on firms with low ratios (HML)
E(ri ) rf = ai + bi [E(rM ) rf ] + si E[SM B] + hi E[HM L]

(9)

bi , si and hi are the betas of the stock on each of the three factors
If these factors fully explain asset returns, the intercept ai of the equation should be zero
Tracking portfolios
To create portfolios that track the size and book-to-market factors, Davis, Fama, and French sort
industrial firms by size (market cap) and by book-to-market (B/M) ratio
Size premium SMB constructed as difference in returns between smallest/largest third of firms
HML in each period is the difference in returns between high and low book-to-market firms
They use a broad market index, the value-weighted return on all stocks traded on US national
exchanges (NYSE, AMEX, and NASDAQ) to compute the excess return on the market portfolio
relative to the risk-free rate, taken to be the return on 1-month T-bills
Test: Davis, Fama, and French form nine portfolios with a range of sensitivities to each factor
They construct the portfolios by sorting firms into three size groups (small S, medium M , and big
B) and three book-to-market groups (high H, medium M , and low L)
For each of these nine portfolios, they estimate Eq. (9) as a first-pass regression over the 816
months between 1929 and 1997 by using the regression model:
ri rf = ai + bi (rM rf ) + si SM B + hi HM L + ei

(10)

Intercepts (ai ) are small and statistically insignificant, with t-statistics 2

Large R2 statistics Returns well explained by three-factor portfolios, and large t-statistics on
size/value loadings show that these factors contribute significantly to explanatory power
How should we interpret these tests?
Size and relative value (B/M ratio) proxy for risks not fully captured by CAPM beta
Consistent with the APT in that it implies that size and value are priced risk factors
Another explanation attributes these premiums to investor irrationality/behavioral biases
Risk-based interpretations
Liew and Vassalou show that returns on style portfolios (HML or SMB) seem to predict GDP
growth, and thus may in fact capture some aspects of business cycle risk
Returns on HML/SMB portfolios are positively related to future growth in macroeconomy
Petkova and Zhang also try to tie avg. return premium on value portfolios to risk premiums
Conditional CAPM, allowing both market RP and firm s to vary over time, and to co-vary
What might lead to such an association between beta and the market risk premium?
Irreversible investments: Value firms (high B/M ratios) have more tangible capital
Investment irreversibility puts such firms more at risk for economic downturns because
they will suffer from excess capacity from assets already in place
In contrast, growth firms better able to deal with downturn by deferring investment plans
Moreover, evidence suggests that market risk premium also is higher in down markets
These factors imparts a positive correlation between of high B/M firms and market RP
Petkova/Zhang fit both beta/market risk premium to a set of state variables (variable that
summarizes the state of the economy):
DIV
DEF LT
T ERM
TB

=
=
=
=

Market dividend yield

Default spread on corporate bonds (Baa - Aaa rates)
Term structure spread (10-yr - 1-yr Treasury rates)
1-month T-bill rate

They estimate a first-pass regression, but first substitute these state variables for beta:
rHM L = + rM t + ei = + [b0 + b1 DIVt + b2 DEF LTt + b3 T ERMt + b4 T B t ]rM t + ei
Where t = b0 + b1 DIVt + b2 DEF LTt + b3 T ERMt + b4 T B t is a time-varying beta
62

 Similarly, the determinants of a time-varying market risk premium, using same state variables:
rM kt,t rf t = c0 + c1 DIVt + c2 DEF LTt + c3 T ERMt + c4 T Bt + et
Petkova and Zhang examine the relationship between beta and the market risk premium:
They define the state of economy by the size of the premium
A peak is defined as the periods with the 10% lowest risk premiums
A trough has the 10% highest risk premiums
The results support the notion of a counter-cyclical value beta: The beta of the HML
portfolio is negative in good economies, meaning that the beta of value stocks (high bookto-market) is less than that of growth stocks (low B/M). Reverse true in recessions
Behavioral explanations
Several authors make the case that the value premium is a manifestation of market irrationality
The essence of the argument is that analysts tend to extrapolate recent performance too far out
into the future, and thus tend to overestimate the value of firms with good recent performance
Chan, Karceski, and Lakonishok makes the case for overreaction:
Firms are sorted into deciles based on income growth in the past 5 years
Book-to-market ratio for each decile at end of 5-year period tracks recent growth very well
B/M falls steadily with growth over past 5 years
This is evidence that past growth is extrapolated and then impounded in price
But B/M at beginning of 5-year period shows little or even a positive association with subsequent growth Market capitalization today is inversely related to growth prospects
Implication: Market ignores evidence that past growth cannot be extrapolated into future
B/M reflects past growth better than future growth, consistent with extrapolation error
La Porta et al. examine stock performance when actual earnings are released to public
Firms are classified as growth versus value stocks
Growth stocks underperform value stocks surrounding these announcements
When news of earnings is released, market is disappointed in stocks priced as growth firms
Liquidity and asset pricing
Measuring liquidity is hard. Effect of liquidity on expected return is composed of two factors:
1. Transaction costs that are dominated by the bid-ask spread
2. Liquidity risk resulting from covariance between changes in asset liquidity cost with both changes
in market-index liquidity cost and with market-index rates of return
Both factors are unobservable and their effect on equilibrium rates of return is hard to estimate
Observed (inside) bid-ask spreads apply only to small trades and therefore may be highly unreliable
indicators of trading costs for larger transactions
Most studies of liquidity instead use a proxy variable that can distinguish liquidity costs across firms
and then calibrate the distribution of such costs to average observed spreads
One widely used measure of illiquidity cost was proposed by Amihud:


Absolute value(Stock return)
ILLIQ = Monthly average of daily
Dollar volume
This measure of illiquidity is based on the price impact per dollar of transactions in the stock and can
be used to estimate both liquidity cost and liquidity risk
Acharya and Pedersen calculate this statistic Cit for stock i in month t, using daily data for 1963-99
The market measure of illiquidity CM t is the average value of Cit over all stocks in month t
Using Cit , CM t , the excess returns Rit (net of Cit ) on each stock, as well as the market excess
return RM t (net of CM t ), they calculate the market beta M and three illiquidity s:
1. The sensitivity of asset illiquidity to market illiquidity: L1
2. The sensitivity of stock returns to market illiquidity: L2
3. The sensitivity of stock illiquidity to the market return: L3
63

 To reduce errors in measurement, they form 25 portfolios sorted from low to high illiquidity
Results:
Liquidity is correlated with all three Fama-French factors, suggesting that some of the predictive power of the FF factors for average returns may in fact be liquidity related
Average excess return increases with portfolio illiquidity, although part of that increase is
attributable to the higher systematic risk associated with higher illiquidity
Liquidity cost (effective bid-ask spread) increases a lot with illiquidity, up to 8.83%/month
After adjustment for the more typical holding periods, as reflected by monthly turnover, this
cost is economically significant for illiquid portfolios, on the order of 2.5% per year
Liquidity s are small relative to market and highly collinear, but still significantly improve
explanatory power of a CAPM augmented with liquidity considerations
Acharya and Pedersens work establishes important point that liquidity is a priced factor
Time-varying volatility
We may associate the variance of the rate of return on the stock with the rate of arrival of new information
because new information may lead investors to revise their assessment of intrinsic value
The rate of arrival of new information is time varying
Expect variances of rates of return on stocks (and covariances among them) to be time varying
When we consider a time-varying return distribution, we must refer to the conditional mean, variance,
and covariance, that is, the mean, variance, or covariance conditional on currently available info
Conditions that vary over time are values of variables that determine level of these parameters
In contrast, the usual estimate of return variance, the average of squared deviations over the sample
period, provides an unconditional estimate (treats variance as constant over time)
Engle ARCH model - Autoregressive conditional heteroskedasticity
Based on the idea that a natural way to update a variance forecast is to average it with the most
recent squared surprise (i.e., the squared deviation of the rate of return from its mean)
Engle GARCH model - Generalized autoregressive conditional heteroskedasticity
Most widely used model to estimate conditional variance of stocks/stock-index returns
Allows greater flexibility in the specification of how volatility evolves over time
Uses rate-of-return history as the information set used to form our estimates of variance
The model posits that the forecast of market volatility evolves relatively smoothly each period in
response to new observations on market returns
The updated estimate of market-return variance in each period depends on both the previous
estimate and the most recent squared residual return on the market
This technique essentially mixes in a statistically efficient manner the previous volatility estimate
with an unbiased estimate based on new observation of market return. Updating formula:
2
t2 = a0 + a1 2t1 + a2 t1

(11)

2
Updated forecast = function of most recent variance forecast t1
and most recent squared predic2
tion error in market return t1 . Parameters a0 , a1 , and a2 estimated from past data
ARCH-type models clearly capture much of the variation in stock market volatility

Consumption-based asset pricing and the equity premium puzzle

Equity premium puzzle
Historical excess returns on risky assets in the US are too large to be consistent with economic
theory and reasonable levels of risk aversion
This observation has come to be known as the equity premium puzzle
Suggests that forecasts of the market risk premium should be lower than historical averages
Consumption growth and market rates of return
ICAPM derived from lifetime consumption/investment plan of representative investor
Each individuals plan is set to maximize a utility function of lifetime consumption, and consumption/investment in each period is based on age and current wealth, as well as the risk-free
rate and the market portfolios risk and risk premium
64

 What matters to investors is not their wealth per se, but their lifetime flow of consumption
There can be slippage between wealth and consumption due to variation in factors such as the
risk-free rate, the market portfolio risk premium, or prices of major consumption items
Better measure of consumer well-being than wealth is consumption flow that wealth can support
Given this framework, the generalization of the basic CAPM is that instead of measuring security
risk based on the covariance of returns with the market return (focuses only on wealth), we are
better off using the covariance of returns with aggregate consumption
Hence, we expect the risk premium of the market index to be related to that covariance as:
E(rM ) rf = A Cov(rM , rC )

(12)

Where A depends on avg. coeff of risk aversion and rC is the rate of return on a consumptiontracking portfolio (highest possible correlation with growth in aggregate consumption)
The first wave of attempts to estimate consumption-based asset pricing models used consumption
data directly rather than returns on consumption-tracking portfolios
By and large, these tests found the CCAPM no better than the conventional CAPM
Difficulty: Consumption data are collected far less frequently and with substantial error
Recent research improves the quality of estimation in several ways:
1. Rather than using consumption growth directly, it uses consumption-tracking portfolios:
Data on aggregate consumption used only to construct consumption-tracking portfolio
Frequent/accurate data on the return on these portfolios then used to test model
2. Investors adjust consumption levels most substantially in 4-th quarter of calendar year
Data from other quarters obscure the reaction of annual consumption to annual returns
3. To improve the models power to explain returns, some newer studies allow for several classes
of investors with differences in wealth and consumption behavior
They also may separate expenditures on consumer nondurables versus durable goods
Standard CCAPM focuses on representative consumer/investor, thereby ignoring info about
heterogeneous investors with different levels of wealth/consumption habits
Jagannathan and Wang study
Focus on year-over-year 4-th quarter consumption, employs consumption-mimicking portfolio
Find that annual consumption growth measured by comparing 4-th quarter data in successive
years is substantially better than other intervals in explaining portfolio returns
Show that FF factors are associated with consumption s as well as excess returns
High book-to-market ratio is associated with higher consumption beta
Larger firm size is associated with lower consumption beta
The suggestion is that the explanatory power of the Fama-French factors for average returns
may in fact reflect the differing consumption risk of those portfolios
Other tests reported by Jagannathan/Wang show that CCAPM explains returns even better
than FF three-factor model, which in turn is superior to single-factor CAPM
The equity premium puzzle refers to the fact that using reasonable estimates of A, the covariance
of consumption growth with the market-index return Cov(rM , rC ) is far too low to justify observed
historical average excess returns on the market-index portfolio
The equity premium puzzle can be interpreted in several ways:
Focus on observed historical returns: Does Eq. (12) fail empirical tests because those historical
returns were not representative of investors expectations at the time?
Conflicting interpretation: Puzzle is real, and is yet another nail in CAPM coffin
Third approach: Extensions of CAPM may resolve the puzzle
4-th interpretation from behavioral finance: Pins the puzzle on irrational behavior
Expected versus realized returns
Fama and French offer another interpretation of the equity premium puzzle:
Using stock index returns from 1872-1999, they report avg. risk-free rate, avg. ROE (represented by S&P 500), and resultant risk premium for overall period and subperiods
65

 The big increase in the average excess return on equity after 1949 suggests that the equity
premium puzzle is largely a creature of modern times
FF suspect that estimating the risk premium from avg. realized returns may be the problem
Using constant-growth DDM to estimate expected returns, they find that for 1872-1949, DDM
yields similar estimates of expected risk premium as avg. realized excess return
But for 1950-99, DDM yields much smaller risk premium High avg. excess return in this
period exceeded returns investors expected to earn at the time
In constant-growth DDM, expected capital gains rate = growth rate of dividends
The expected total return on the firms stock will be the sum of dividend yield (dividend/price) plus the expected dividend growth rate g:
E(r) =

D1
+g
P0

where D1 = end-of-year dividends and P0 = current stock price (13)

For any sample period t = 1, , T , FF estimate expected return from arithmetic avg. of
dividend yield (Dt /Pt1 ) plus dividend growth rate (gt = Dt /Dt1 1)
In contrast, realized return = dividend yield plus rate of capital gains (Pt /Pt1 1)
Because dividend yield is common to both estimates, the difference between expected/realized
return equals the difference between the dividend growth and capital gains rates
Capital gains significantly exceeded the dividend growth rate in modern times
FF conclusion: Equity premium puzzle due to unanticipated capital gains in latter period
FF argue that dividend growth rates produce more reliable estimates of capital gains investors
actually expected to earn than the avg. of their realized capital gains. Three reasons:
1. Avg. realized returns over 1950-99 exceeded IRR on corp. investments. If those avg. returns
were representative of expectations, then firms were engaging in negative-NPV investments
2. Statistical precision of DDM estimates far higher than using avg. historical returns
3. The reward-to-volatility (Sharpe) ratio derived from DDM far more stable than that from
realized returns. If risk aversion constant, then Sharpe ratio should be stable
Fama and Frenchs study provides a simple explanation for the equity premium puzzle:
Observed rates of return in the recent half-century were unexpectedly high
Implies that forecasts of future excess returns will be lower than past averages
Goetzmann and Ibbotson lends support to Fama and Frenchs argument
They extend data on rates of return on stocks/long-term corporate bonds back to 1792
These statistics suggest a risk premium that is much lower than the historical average for
1926-2005, which is the period that produces the equity premium puzzle
Survivorship bias
The equity premium puzzle emerged from long-term averages of US stock returns. There are
reasons to suspect that these estimates of the risk premium are subject to survivorship bias
Jurion and Goetzmann: Database of capital appreciation indexes for 39 stock markets over 1921-96
US equities had highest real return of all countries (4.3% annually vs. 0.8% median)
Unlike US, many countries had equity markets that closed (permanently/extended time)
Using avg. US data leads to survivorship bias for estimate of expected returns
Estimating risk premiums from experience of most successful country and ignoring evidence from
stock markets that did not survive for full period impart an upward bias in expected returns
The high realized equity premium obtained for US may not be indicative of required returns
Extensions to the CAPM may resolve the equity premium puzzle
Constantinides argues that the standard CAPM can be extended to account for observed excess
returns by relaxing some of its assumptions, in particular, by recognizing that consumers face
uninsurable and idiosyncratic income shocks, e.g., the loss of employment
In addition, life-cycle considerations are important and often overlooked
Borrowing constraints become important when placed in the context of the life cycle
The imaginary representative consumer does not face borrowing constraints
Young consumers, however, do face meaningful borrowing constraints
66

 Constantinides traces their impact on the equity premium, the demand for bonds, and on the
limited participation of many consumers in the capital markets
Adding habit formation to conventional utility function helps explain higher risk premiums than
those obtained by covariance of stock returns with aggregate consumption growth
Constantinides argues for integrating notions of habit formation, incomplete markets, life cycle,
borrowing constraints, and sources of limited stock market participation
Behavioral explanations of the equity premium puzzle
Barberis and Huang explain the puzzle as an outcome of irrational investor behavior
Key elements: Loss aversion and narrow framing
Narrow framing is the idea that investors evaluate every risk they face in isolation
Investors ignore low correlation of the risk of a stock portfolio with other components of
wealth, and therefore require a higher risk premium than rational models would predict
Combined with loss aversion, investor behavior will generate large risk premiums despite the
fact that traditionally measured risk aversion is plausibly low
Incorporating these effects generates large equilibrium equity risk premium and low/stable risk-free
rate, even when consumption growth is smooth/only weakly correlated with stock market
Analysis for the equity premium also has implications for the stock market participation puzzle
When accounting for heterogeneity of preferences, behavioral approach explains why segment
of population that should participate in stock market despite frictions/other rational explanations, still avoids it

67

68

Fixed Income Securities

BKM - Ch. 14: Bond prices and yields

Introduction
Debt securities
A debt security is a claim on a specified periodic stream of income
Often called fixed-income securities because they promise either a fixed stream of income or a
stream of income that is determined according to a specified formula
Risk considerations are minimal as long as the issuer of the security is sufficiently creditworthy
The bond is the basic debt security
Includes Treasury, corporate, and international bonds
Bond prices are set in accordance with market interest rates
Measures of bond returns: (i) Yield to maturity (YTM), (ii) Yield to call, (iii) Holding-period
return (HPR), or (iv) Realized compound rate of return
Important to consider the impact of default or credit risk on bond pricing and look at the determinants of credit risk and the default premium built into bond yields
Bond characteristics
A bond is a security that is issued in connection with a borrowing arrangement
Borrower issues (sells) bond to lender for some cash. Bond is the IOU of the borrower
The arrangement obligates issuer to make specified payments to bondholder on specified dates
Typical coupon bond obligates issuer to make semiannual payments of interest for life of bond
When bond matures, issuer repays debt by paying bondholder the bond s par value (face value)
The coupon rate of the bond serves to determine the interest payment
The annual payment is the coupon rate times the bonds par value
The coupon rate, maturity date, and par value of the bond are part of the bond indenture, which
is the contract between the issuer and the bondholder
Sometimes, zero-coupon bonds are issued that make no coupon payments
In this case, investors receive par value at maturity date but no interest payments until then
The bond has a coupon rate of zero
These bonds are issued at prices considerably below par value, and the investors return comes
solely from the difference between issue price and the payment of par value at maturity
Treasury bonds and notes
Generalities
Treasury note maturities up to 10 years, while Treasury bonds have 10-30 years maturities
Bonds/notes issued in denominations of $1,000 or more (minimum reduced to $100 in 2008)
Both make semiannual coupon payments
Bonds prices are quoted as a percentage of par value
The yield to maturity (YTM) is a measure of the average rate of return to an investor who
purchases the bond for the asked price and holds it until its maturity date
Accrued interest and quoted bond prices
Bond prices quoted in financial pages are not actually prices that investors pay for the bond
because quoted price does not include interest that accrues between coupon payment dates
If a bond is purchased between coupon payments, the buyer must pay the seller for accrued
interest, the prorated share of the upcoming semiannual coupon
The sale, or invoice, price of the bond would equal the stated price plus the accrued interest:
Accrued interest =

Annual coupon payment

Days since last coupon payment

2
Days separating coupon payments

Corporate bonds
Although some bonds trade on a formal exchange operated by NYSE, most bonds are traded
over-the-counter in a network of bond dealers linked by a computer quotation system
Safer bonds with higher ratings promise lower YTM than other bonds with similar maturities
71

 Call provisions on corporate bonds

Although Treasury no longer issues callable bonds, some corporate bonds are issued with call
provisions allowing issuer to repurchase bond at specified call price before maturity
E.g., if a company issues a bond with a high coupon rate when market interest rates are high,
and interest rates later fall, the firm might like to retire the high-coupon debt and issue new
bonds at a lower coupon rate to reduce interest payments (aka refunding)
Callable bonds typically come with a period of call protection, an initial time during which
the bonds are not callable (aka deferred callable bonds)
To compensate investors, callable bonds issued with higher coupons/promised YTM
Convertible bonds
Convertible bonds give bondholders an option to exchange each bond for a specified number
of shares of common stock of the firm
The conversion ratio is the number of shares for which each bond may be exchanged
Market conversion value = current value of shares for which bonds may be exchanged
The conversion premium is the excess of the bond value over its conversion value
Convertible bondholders benefit from price appreciation of the companys stock
Convertible bonds offer lower coupon rates/promised YTM
Puttable bonds
Extendable/put bond gives bondholder the option to extend/retire the bond
If bonds coupon rate exceeds current market yields, bondholder will extend bonds life
Floating-rate bonds
Floating-rate bonds make interest payments tied to some measure of current market rates
The major risk involved in floaters has to do with changes in the firms financial strength
The yield spread is fixed over the life of the security. If financial health of firm deteriorates,
then investors will demand a greater yield premium than offered by the security
Coupon rate adjusts to changes in level of market rates, but not to firm financial condition
Preferred stock
Although preferred stock is (strictly) equity, it is often included in fixed-income universe
Like bonds, preferred stock promises to pay a specified stream of dividends
However, unlike bonds, the failure to pay the promised dividend does not result in corporate
bankruptcy. Instead, the dividends owed simply cumulate
In the event of bankruptcy, preferred stockholders claims to the firms assets have lower priority
than those of bondholders, but higher priority than those of common stockholders
Preferred stock commonly pays a fixed dividend Perpetuity, providing level CF indefinitely
Floating-rate preferred stock is much like floating-rate bonds: The dividend rate is linked to a
measure of current market interest rates and is adjusted at regular intervals
Unlike interest on bonds, dividends on preferred stock are not tax-deductible expenses to the firm
This reduces their attractiveness as a source of capital to issuing firms
On the other hand, there is an offsetting tax advantage to preferred stock
When firm buys preferred stock of another firm, tax on only 30% of dividends received
Preferred stock rarely gives its holders full voting privileges in the firm. However, if the preferred
dividend is skipped, the preferred stockholders may then be provided some voting power
Other issuers
State/local gov. issue municipal bonds. Outstanding feature: Interest payments are tax-free
Gov. agencies such as Federal Home Loan Bank Board, Farm Credit agencies, and mortgage
pass-through agencies Ginnie Mae/Fannie Mae/Freddie Mac also issue bonds
International bonds
International bonds are commonly divided into: (i) Foreign bonds and (ii) Eurobonds
Foreign bonds are issued by borrower from country other than the one in which bond is sold
The bond is denominated in the currency of the country in which it is marketed
E.g., German firm sells dollar-denominated bond in the US Foreign bond

72

 Eurobonds are bonds issued in the currency of one country but sold in other national markets
E.g., the Eurodollar market refers to dollar-denominated bonds sold outside the US
Because Eurodollar market falls outside US jurisdiction, such bonds not regulated by US Feds
Similarly, Euroyen bonds are yen-denominated bonds selling outside Japan
Innovations in the bond market
Inverse floaters
The coupon rate on these bonds falls when the general level of interest rates rises
Investors in these bonds suffer doubly when rates rise: Not only does the PV of each dollar of
CF from the bond fall as the discount rate rises, but the level of those CFs falls as well
Of course, investors in these bonds benefit doubly when rates fall
Asset-backed bonds
The income from a specified group of assets is used to service the debt
E.g., mortgage-backed securities or securities backed by auto or credit card loans
Walt Disney issued bonds with coupon rates tied to financial performance of its films
Catastrophe bonds
These bonds are a way to transfer catastrophe risk from the firm to the capital markets
Investors in these bonds receive compensation in the form of higher coupon rates
But in the event of a catastrophe, the bondholders will give up all/part of their investments
Indexed bonds
Indexed bonds make payments tied to a general price index/price of a particular commodity
E.g., Treasury Inflation Protected Securities (TIPS): By tying par value to general level of
prices, coupon payments/final repayment of par value on these bonds increase CPI
Bond with 3yr maturity, par value of $1,000, and coupon rate of 4%. Assume annual
coupon payments and inflation turns out to be 2%, 3%, and 1% in next 3 years
The nominal rate of return on the bond in the first year is:
Nominal return =

Interest + Price Appreciation

40.80 + 20
=
= 6.08%
Initial Price
1000

The real rate of return is precisely the 4% real yield on the bond:
Real return =

1 + Nominal Return
1.0608
1=
1 = 4%
1 + Inflation
1.02

Bond pricing
Because bonds coupon/principal repayments occur in the future, the price an investor pays for a claim
to those payments depends on value of dollars to be received in future vs. dollars in hand today
This present value calculation depends in turn on market interest rates
In addition, because most bonds are not riskless, discount rate embodies an additional premium
that reflects bond-specific characteristics such as default risk/liquidity/tax attributes/call risk
CFs from a bond consist of coupon payments until maturity date plus final payment of par value
Bond value = Present value of coupons + Present value of par value
If we call the maturity date T and call the interest rate r, the bond value is:
Bond value =

T
X
Coupon
t=1



Par value
1
1
1
+
= Coupon
1
+ Par value
t
T
T
(1 + r)
(1 + r)
r
(1 + r)
(1 + r)T

= Coupon Annuity factor(r, T ) + Par value PV factor(r, T )

(1)
(2)

At higher interest rate, PV of payments to be received is lower Bond price falls as interest rates rise
Convex shape of the bond price curve
Progressive increases in interest rate result in progressively smaller reductions in bond price
An increase in the interest rate results in a price decline that is smaller than the price gain resulting
from a decrease of equal magnitude in the interest rate
73

 Corporate bonds typically are issued at par value

Underwriters of bond issue must choose a coupon rate that closely approximates market yields
If the coupon rate is inadequate, investors will not pay par value for the bonds
After bonds are issued, bondholders may buy/sell bonds in secondary markets (NYSE or OTC)
The inverse relationship between price and yield is a central feature of fixed-income securities
Interest rate fluctuations represent the main source of risk in the fixed-income market
One key factor that determines that sensitivity is the maturity of the bond
The longer the maturity, the greater the sensitivity to fluctuations in interest rates
Bond pricing between coupon dates
In principle, the fact that bond is between coupon dates does not affect pricing problem
The procedure is always the same: Compute the PV of each remaining payment and sum up
Bond prices are typically quoted net of accrued interest
These prices, which appear in the financial press, are called flat prices
Invoice price = Flat price + Accrued interest

Prices ($)

Bond yields
The current yield of a bond measures only the cash income provided by the bond as a percentage of
bond price and ignores any prospective capital gains or losses
The yield to maturity is the standard measure of the total rate of return which accounts for both current
income and the price increase or decrease over the bonds life
Yield to maturity (YTM)
Defined as the interest rate that makes the PV of a bonds payments equal to its price
Measure of avg. rate of return earned if bond is bought now and held until maturity
To calculate the YTM, solve the bond price equation for the interest rate given the bonds price
The financial press reports yields on an annualized basis, and annualizes the bonds semiannual
yield using simple interest techniques, resulting in an annual percentage rate, or APR
Yields annualized using simple interest are also called bond equivalent yields
The effective annual yield of the bond, however, accounts for compound interest
The bonds yield to maturity is the IRR on an investment in the bond
The yield to maturity can be interpreted as the compound rate of return over the life of the bond
under the assumption that all bond coupons can be reinvested at that yield
Yield to maturity is widely accepted as a proxy for average return
YTM differs from current yield = [bonds annual coupon payment] [bond price]
For premium bonds (bonds selling above par value), coupon rate is greater than current yield,
which in turn is greater than YTM because the YTM accounts for the built-in capital loss
For discount bonds (bonds selling below par value), these relationships are reversed
Yield to call
When interest rates fall, the PV of the bonds scheduled payments rises, but the call provision
allows the issuer to repurchase the bond at the call price

Straight Bond

1500
1200
900

Callable Bond

600

Call Price

300

10

Interest Rate (%)

Figure 1: Bond prices: Callable and straight debt

74

 If call price < PV of scheduled payments, the issuer may call the bond back
At high interest rates, risk of call negligible because PV of scheduled payments < call price
The values of the straight and callable bonds converge
At lower rates, however, the values of the bonds begin to diverge, with the difference reflecting
the value of the firms option to reclaim the callable bond at the call price
At very low rates, PV of scheduled payments exceeds the call price, so the bond is called
Its value at this point is simply the call price
The yield to call is calculated just like the yield to maturity except that the time until call replaces
time until maturity, and the call price replaces the par value
Premium bonds selling near their call prices are especially apt to be called if rates fall further
Investors in premium bonds often more interested in bonds yield to call rather than YTM
Realized compound return versus yield to maturity
The yield to maturity will equal the rate of return realized over the life of the bond if all coupons
are reinvested at an interest rate equal to the bonds yield to maturity
Denoting Vf the final value of the investment (par value plus coupons payment bearing interest),
the realized compound return r is determined by: Vi (1 + r)n = Vf
Problem with conventional yield to maturity when reinvestment rates can change over time: Conventional yield to maturity will not equal realized compound return
Disadvantages of the realized compound return:
It cannot be computed in advance without a forecast of future reinvestment rates
Reduces much of the attraction of the realized return measure
Horizon analysis
Forecasting the realized compound yield over various holding periods/investment horizons
The forecast of total return depends on forecasts of both the price of the bond when sold at
the end of your horizon and the rate at which coupon income is reinvested
The sales price depends in turn on the yield to maturity at the horizon date
Longer investment horizon Reinvested coupons are larger component of final proceeds
As interest rates change, bond investors are actually subject to two sources of offsetting risk:
On the one hand, when rates rise, bond prices fall, which reduces the value of the portfolio
On the other hand, reinvested coupon income compounds faster at higher rates
This reinvestment rate risk will offset the impact of price risk
Bond prices over time
A bond will sell at par value when its coupon rate equals the market interest rate
The investor receives fair compensation for TVM with the recurring coupon payments
No further capital gain is necessary to provide fair compensation
When the coupon rate is lower than the market interest rate, the coupon payments alone will not provide
investors as high a return as they could earn elsewhere in the market
To receive a fair return, investors also need to earn price appreciation on their bonds
The bonds have to sell below par value to provide built-in capital gain on the investment
When bond prices are set according to PV formula, any discount from par value provides an anticipated
capital gain that augments a below-market coupon rate to provide a fair total rate of return
Conversely, if coupon rate exceeds market interest rate, the interest income by itself is greater than
elsewhere in market Investors bid up price of these bonds above their par values
As bonds approach maturity, they fall in value because fewer above-market coupon payments
remain Resulting capital losses offset large coupon payments Fair rate of return
Although capital gain vs. income components differ, bond prices are set to provide competitive rates
Yield to maturity versus holding-period return
When the YTM is unchanged over the period, the rate of return on the bond will equal that yield
However, unanticipated changes in market rates result in unanticipated changes in bond returns
Increase in bonds yield acts to reduce its price HPR will be less than initial yield
Conversely, a decline in yield will result in a holding-period return greater than the initial yield
75

 Yield to maturity vs. holding-period return

YTM depends only on bonds coupon, current price, and par value at maturity
All of these values are observable today, so yield to maturity can be easily calculated
YTM = Measure of avg. rate of return if bond held until maturity
In contrast, HPR is the rate of return over a particular investment period and depends on the
market price of the bond at the end of that holding period
Bond prices respond to unanticipated changes in interest rates HPR can at most be forecast
Zero-coupon bonds and Treasury strips
Original-issue discount bonds
Less common than coupon bonds issued at par
Bonds issued intentionally with low coupon rates that cause bond to sell at a discount
An extreme example of this type of bond is the zero-coupon bond
Treasury bills are examples of short-term zero-coupon instruments
If face value = $10,000, Treasury issues it for less, agreeing to repay $10,000 at maturity
All of the investors return comes in the form of price appreciation
Longer-term zero-coupon bonds created from coupon-bearing notes/bonds with help of US Treasury
Dealer who purchases Treasury coupon bond may ask Treasury to break down CFs to be paid
into independent securities, where each security is one of the original bond payments
E.g., a 10-year coupon bond would be stripped of its 20 semiannual coupons, and each
coupon payment would be treated as a stand-alone zero-coupon bond
The maturities of these bonds would thus range from 6 months to 10 years
The final payment of principal would be treated as another stand-alone zero-coupon security
Treasury program which perform coupon stripping: STRIPS (Separate Trading of Registered
Interest and Principal of Securities). Such zero-coupon securities Treasury strips
Prices of zeros as time passes
At maturity, zeros must sell for par value As time passes, price should approach par value
If the interest rate is constant, a zeros price will increase at exactly the rate of interest
Consider a zero with 30 years until maturity, and suppose the market interest rate is 10% per
year. The price of the bond today will be 1, 000/(1.10)30 = 57.31
Next year, the price is 1, 000/(1.10)29 = 63.04, a 10% increase over previous-year value
The bond prices rise exponentially, not linearly, until its maturity
After-tax returns
Tax authorities recognize that built-in price appreciation on original-issue discount (OID) bonds
such as zero-coupon bonds represents an implicit interest payment to security holder
Therefore, IRS calculates price appreciation schedule to impute taxable interest income for built-in
appreciation during tax year, even if asset is not sold/does not mature until future year
Interest income based on constant yield method (ignores changes in market interest rates)
Any additional gains or losses that arise from changes in market interest rates are treated as capital
gains or losses if the OID bond is sold during the tax year
Default risk and bond pricing
Introduction
US government bonds may be treated as free of default risk. Not true of corporate bonds
Bond default risk, aka credit risk, is measured by Moodys, S&P, and Fitch, which provide financial
info on firms and quality ratings of large corporate/municipal bond issues
The top rating is AAA or Aaa, a designation awarded to only about a dozen firms
Moodys modifies each rating class with a 1, 2, or 3 suffix (e.g., Aaal, Aaa2, Aaa3) to provide
a finer gradation of ratings. The other agencies use a + or modification
Bonds rated BBB or above (S&P, Fitch) or Baa and above (Moodys) are considered investmentgrade bonds, whereas lower-rated bonds are classified as speculative-grade or junk bonds
Defaults on low-grade issues are not uncommon
E.g., almost half of bonds rated CCC by S&Ps at issue have defaulted within 10 years
76

 Junk bonds
Junk bonds, aka high-yield bonds, are speculative-grade (low-rated/unrated) bonds
Before 1977, almost all junk bonds were fallen angels, i.e. bonds issued by firms that
originally had investment-grade ratings but that had since been downgraded
1977: Original-issue junk started: Lower-cost financing alternative than bank borrowing
High-yield bonds notorious in 80s as financing in leveraged buyouts/hostile takeover attempts
Shortly thereafter, the market suffered. Legal difficulties of Drexel/Mike Milken with Wall
Streets insider trading scandals of late 80s tainted junk bond market which nearly dried up
Since then, the market has rebounded dramatically
Determinants of bond safety
Bond rating agencies base their quality ratings largely on an analysis of the level and trend of some
of the issuers financial ratios. The key ratios used to evaluate safety are:
1. Coverage ratios - Ratios of company earnings to fixed costs
Times-interest-earned ratio: Ratio of EBIT to interest obligations
Fixed-charge coverage ratio: Includes lease payments and sinking fund payments with
interest obligations to arrive at the ratio of earnings to all fixed cash obligations
Low or falling coverage ratios signal possible cash flow difficulties
2. Leverage ratio / Debt-to-equity ratio
A too-high leverage ratio indicates excessive indebtedness, signaling the possibility the firm
will be unable to earn enough to satisfy the obligations on its bonds
3. Liquidity ratios
The two most common liquidity ratios are the current ratio (current assets/current liabilities)
and the quick ratio (current assets excluding inventories/current liabilities)
These ratios measure the firms ability to pay bills coming due with its most liquid assets
4. Profitability ratios
Measures of rates of return on assets or equity. Indicators of a firms overall financial health
The return on assets (earnings before interest and taxes divided by total assets) or return on
equity (net income/equity) are the most popular of these measures
Firms with higher returns on assets or equity should be better able to raise money in security
markets because they offer prospects for better returns on the firms investments
5. Cash flow-to-debt ratio
This is the ratio of total cash flow to outstanding debt
S&Ps periodically computes median values of selected ratios for firms in several rating classes
Ratios evaluated in context of industry standards. Analysts differ in ratios weights
Default rates vary dramatically with bond rating
Historically, 1% of bonds originally rated AA/better at issuance had defaulted after 15 years
That ratio is around 7.5% for BBB-rated bonds, and 40% for B-rated bonds
Credit risk clearly varies dramatically across rating classes
Studies have tested whether financial ratios can in fact be used to predict default risk
Altman used discriminant analysis to predict bankruptcy
With this technique a firm is assigned a score based on its financial characteristics
If its score exceeds a cut-off value, the firm is deemed creditworthy. A score below the cut-off
value indicates significant bankruptcy risk in the near future
Altman found the following equation to best separate failing and nonfailing firms:
Z = 3.3

EBIT
Sales
Market value of equity
Retained earnings
Working capital
+ 99.9
+ .6
+ 1.4
+ 1.2
Total assets
Assets
Book value of debt
Total assets
Total assets

Bond indentures
A bond is issued with an indenture, which is the contract between issuer/bondholder
Part of the indenture is a set of restrictions that protect rights of the bondholders
Include provisions relating to collateral/sinking funds/dividend policy/further borrowing
Issuer agrees to protective covenants to market its bonds to investors concerned about safety
77

 Sinking funds
Payment of par value at end of bonds life constitutes large cash commitment for issuer
To help ensure the commitment does not create a CF crisis, the firm agrees to establish a
sinking fund to spread the payment burden over several years
The fund may operate in one of two ways:
1. Firm may repurchase fraction of outstanding bonds in the open market each year
2. Firm may purchase outstanding bonds at special call price from SF provision
Firm has option to purchase bonds at either market price/sinking fund price, whichever
is lower (bonds chosen for the call are selected at random based on serial number)
Although less common, sinking fund provision may call for periodic payments to trustee, with
payments invested so that accumulated sum is used for retirement of entire issue at maturity
The sinking fund call differs from a conventional bond call in two important ways:
1. Firm can repurchase only a limited fraction of bond issue at sinking fund call price
2. Usually: Callable bonds call prices > par value, sinking fund call price = par value
Although sinking funds protect bondholders (principal repayment more likely), they can hurt
investors: Firm will buy back discount bonds (selling below par) at market price, and buy back
premium bonds (selling above par) at par
One bond issue that does not require a sinking fund is a serial bond issue
In a serial bond issue, the firm sells bonds with staggered maturity dates
As bonds mature sequentially, principal repayment is spread over time ( sinking fund)
One advantage of serial bonds over sinking fund issues is that there is no uncertainty
introduced by the possibility that a particular bond will be called for the sinking fund
The disadvantage of serial bonds, however, is that bonds of different maturity dates are
not interchangeable, which reduces the liquidity of the issue
Subordination of further debt
One of the factors determining bond safety is total outstanding debt of the issuer
Subordination clauses restrict the amount of additional borrowing: Additional debt might be
required to be subordinated in priority to existing debt
Subordination is sometimes called a me-first rule, meaning the senior (earlier) bondholders
are to be paid first in the event of bankruptcy
Dividend restrictions
Covenants also limit the dividends firms may pay. These limitations protect bondholders
because they force the firm to retain assets rather than paying them out to stockholders
A typical restriction disallows payments of dividends if cumulative dividends paid since the
firms inception exceed cumulative retained earnings plus proceeds from sales of stock
Collateral
Some bonds are issued with specific collateral behind them
Collateral = particular asset of firm that bondholders receive if firm defaults on bond
If the collateral is property, the bond is called a mortgage bond
If collateral is another security held by firm Collateral trust bond
In the case of equipment, the bond is known as an equipment obligation bond
Collateralized bonds generally are considered the safest variety of corporate bonds
General debenture bonds by contrast do not provide collateral: They are unsecured bonds
Because they are safer, collateralized bonds generally offer lower yields than general debentures
Yield to maturity and default risk
Bonds subject to default risk Distinguish between bonds promised YTM and its expected yield
The stated yield is the maximum possible yield to maturity of the bond
The expected yield to maturity must take into account the possibility of a default
When bond becomes more subject to default risk, its price falls Its promised YTM rises
Similarly, default premium (spread between stated YTM/comparable Treasury bonds) rises
However, its expected YTM (ultimately tied to systematic risk of bond) will be far less affected

78

 Default premium
To compensate for the possibility of default, corporate bonds must offer a default premium
The default premium is the difference between the promised yield on a corporate bond and
the yield of an otherwise-identical government bond that is riskless in terms of default
Risk structure of interest rates Pattern of default premiums offered on risky bonds
The greater the default risk, the higher the default premium
One particular manner in which yield spreads vary over time is related to business cycle
Yield spreads tend to be wider when the economy is in a recession
Investors perceive a higher probability of bankruptcy when the economy is faltering, even
holding bond rating constant: They require a commensurately higher default premium
Aka flight to quality, meaning that investors move their funds into safer bonds unless they
can obtain larger premiums on lower-rated securities
Credit risk and collateralized debt obligations (CDOs)
Collateralized debt obligations: Major mechanism to reallocate credit risk in fixed-income markets
To create a CDO, a financial institution, commonly a bank, first establishes a legally distinct entity
to buy and later resell a portfolio of bonds or other loans
A common vehicle for this purpose is the Structured Investment Vehicle (SIV)
Legal separation of bank/SIV allows ownership of the loans to be conducted off the banks
balance sheet, and thus avoids capital requirements the bank would otherwise encounter
The SIV raises funds, often by issuing short-term commercial paper, and uses the proceeds to
buy corporate bonds or other forms of debt such as mortgage loans or credit card debt
These loans are first pooled together and then split into a series of classes known as tranches
Each tranche is given a different level of seniority in terms of its claims on the underlying loan
pool, and each can be sold as a stand-alone security
As the loans in the underlying pool make their interest payments, the proceeds are distributed
to pay interest to each tranche in order of seniority
This priority structure implies that each tranche has a different exposure to credit risk
The bottom tranche - aka the equity, first-loss, or residual tranche - has last call on payments
from the pool of loans (head of the line in terms of absorbing default/delinquency risk)
Using junior tranches to insulate senior tranches from credit risk, one can create Aaa-rated
bonds even from a junk-bond portfolio
While Aaa-rated bonds are rare, Aaa-rated CDO tranches are common
Investors in tranches with the greatest exposure to credit risk demand the highest coupon rates
Investors with greater expertise in credit risk are natural investors in these securities
Often, the originating bank holds the residual tranche: Provides incentives to originator to
perform careful credit analysis of bonds included in structure
Mortgage-backed CDOs were an investment disaster in 2007
Highly rated tranches suffered extreme losses as default rates far higher than anticipated
Rating agencies that certified these tranches as investment-grade came under fire
Questions were raised concerning conflicts of interest: Because the rating agencies are paid by
bond issuers, the agencies were accused of responding to pressure to ease their standards

79

80

Hull - Ch. 4: Interest rates

Types of rates
The higher the credit risk, the higher the interest rate that is promised by the borrower
Treasury rates
Treasury rates are the rates an investor earns on Treasury bills and Treasury bonds
These are the instruments used by a government to borrow in its own currency
Usually assumed that no chance a government will default on an obligation in its own currency
Traders argue that Treasury rates are too low to be used as risk-free rates because:
1. T-bills/Treasuries must be purchased by institutions to fulfill many regulatory requirements
This increases demand for these Treasury instruments driving price up and yield down
2. Amount of capital a bank is required to hold for investment in T-bills/bonds is substantially smaller
than capital required to support similar investment in other instruments with very low risk
3. In the US, Treasuries are given a favorable tax treatment (no tax at the state level)
LIBOR
LIBOR is short for London Interbank Offered Rate
A LIBOR quote by a particular bank is the rate of interest at which the bank is prepared to
make a large wholesale deposit with other banks
Large banks/financials quote LIBOR in all major currencies for maturities up to 12 months
AA-rated financial institutions regard LIBOR as their short-term opportunity cost of capital: They
can borrow short-term funds at the LIBOR quotes of other financial institutions
LIBOR rates are not totally free of credit risk
However, they are close to risk-free
Traders regard LIBOR rates as a better indication of the true risk-free rate than Treasury
rates, because many tax/regulatory issues cause Treasury rates to be artificially low
LIBID rates (London Interbank Bid Rate)
Large banks also quote LIBID rates
This is the rate at which they will accept deposits from other banks
At any time, usually small spread between quoted LIBID/LIBOR rates (LIBOR > LIBID)
The rates themselves are determined by active trading between banks
LIBOR and LIBID trade in what is known as the Eurocurrency market
This market is outside the control of any one government
Repo rates
Sometimes trading activities are funded with a repo or repurchase agreement
This is a contract where an investment dealer who owns securities agrees to sell them to another
company now and buy them back later at a slightly higher price
The other company is providing a loan to the investment dealer
Difference between prices at which securities are sold/repurchased is the interest it earns
The interest rate is referred to as the repo rate
If structured carefully, the loan involves very little credit risk
Measuring interest rates
Suppose that an amount A is invested for n years at an interest rate of R per annum
If the rate is compounded once per annum, the terminal value of the
investment is: A(1 + R)n
R mn
If rate compounded m times per annum, terminal value = A 1 + m
When m = 1, the rate is sometimes referred to as the equivalent annual interest rate
Continuous compounding
With continuous compounding (m ), A invested for n years at rate R grows to: AeRn
Discounting at a continuously compounded rate R for n years involves multiplying by eRn

81

 Suppose that Rc is a rate of interest with continuous compounding and Rm is the equivalent rate
with compounding m times per annum:


Rm
Rc = m ln 1 +

Rm = m(eRc /m 1)
(1)
m
Zero rates
n-year zero-coupon interest rate = rate of interest on investment starting today and lasting for n years
All the interest and principal is realized at the end of n years, there are no intermediate payments
n-year zero-coupon interest rate: Aka n-year spot rate, n-year zero rate, or n-year zero
Bond pricing
Price of a bond
Most bonds pay coupons to the holder periodically
The bonds principal (aka par value or face value) is paid at the end of its life
Theoretical price of bond calculated as PV of all CFs that will be received by bondholder
Sometimes bond traders use the same discount rate for all the CFs underlying a bond, but a more
accurate approach is to use a different zero rate for each CF
E.g., the theoretical price of a 2-year Treasury bond with a principal of $100 provides coupons at
the rate of 6% per annum semiannually is:
3e0.050.5 + 3e0.0581.0 + 3e0.0641.5 + 103e0.0682.0 = 98.39
Bond yield
Bonds yield = discount rate that, when applied to all CFs, gives bond price = market price
If y is the yield on the bond, expressed with continuous compounding, then:
3ey0.5 + 3ey1.0 + 3ey1.5 + 103ey2.0 = 98.39 y = 6.76%
Par yield
Par yield for a bond maturity = coupon rate that causes bond price = par value
Suppose that coupon on 2-year bond is c per annum (c/2 per 6m). Bond price = par value when:
c 0.050.5 c 0.0581.0 c 0.0641.5 
c  0.0682.0
e
+ e
+ e
+ 100 +
e
= 100 c = 6.87
2
2
2
2
The 2-year par yield is therefore 6.87% per annum with semiannual compounding
If d = PV of $1 received at maturity, A = value of annuity that pays $1 on each coupon date, and
m = # of coupon payments per year, then the par yield c is:
A

c
+ 100d = 100
m

c=

(100 100d)m
A

Determining treasury zero rates

One way of determining Treasury zero rates is to observe the yields on strips
Bootstrap method
Another way to determine Treasury zero rates is from Treasury bills and coupon-bearing bonds
Start with short-term instruments and move progressively to longer-term instruments, making sure
that zero rates calculated at each stage are consistent with prices of instruments
Bond principal ($)
100
100
100
100
100

Time to maturity (years)

0.25
0.50
1.00
1.50
2.00

82

Annual coupon ($)

0
0
0
8
12

Bond price ($)

97.5
94.9
90.0
96.0
101.6

 1st three bonds pay no coupons Zero rates corresponding to their maturities easily calculated
3-month bond provides return of 2.5 in 3 months on an initial investment of 97.5. With
quarterly compounding, 3-month zero rate is (4 2.5)/97.5 = 10.256% per annum. With
continuous compounding, rate becomes: 4 ln(1 + 0.10256/4) = 10.127% per annum
The 6-month bond provides a return of 5.1 in 6 months on an initial investment of 94.9. With
semiannual compounding the 6-month rate is (2 5.1)/94.9 = 10.748% per annum. With
continuous compounding, it becomes: 2 ln(1 + 0.10748/2) = 10.469% per annum
Similarly, the 1-year rate with continuous compounding is: ln(1+10/90) = 10.536% per annum
4th bond lasts 1.5 years. Payments: $4 at 6 months, $4 at one year, and $104 at 1.5 years
From earlier calculations, the discount rate for the payment at the end of 6 months is 10.469%
and that the discount rate for the payment at the end of 1 year is 10.536%
The bonds price $96 must equal the PV of all the payments received by bondholder
Denoting the 1.5-year zero rate by R, we have:
4e0.104690.5 + 4e0.105361.0 + 104eR1.5 = 96 R = 10.681%
This is the only zero rate consistent with 6-month rate, 1-year rate, and data in table
2-year zero rate calculated similarly from 6m, 1yr, and 1.5yr zero rates, and info on last bond
If R is the 2-year zero rate, then:
6e0.104690.5 + 6e0.105361.0 + 6e0.106811.5 + 106eR2.0 = 101.6 R = 10.808%
A chart showing the zero rate as a function of maturity is known as the zero curve
Common assumption: Zero curve is linear between points determined using bootstrap method
The 1.25-year zero rate is 0.5 10.536 + 0.5 10.681 = 10.608%
Usually assumed that zero curve is horizontal prior to first point/beyond last point
Using longer maturity bonds, zero curve would be more accurately determined beyond 2 years
In practice, we do not usually have bonds with maturities equal to exactly 1.5 years, 2 years, etc.
Approach: Interpolate between bond price data before it is used to calculate zero curve
E.g., if a 2.3-year bond with a coupon of 6% sells for 98 and a 2.7-year bond with a coupon of 6.5%
sells for 99, assume that a 2.5-year bond with a coupon of 6.25% sells for 98.5
Forward rates
Forward interest rates are rates of interest implied by current zero rates for future periods of time
Year
1
2
3
4
5

Zero rate for n-yr investment

3.0
4.0
4.6
5.0
5.3

Forward rate for n-th year

5.0
5.8
6.2
6.5

The forward interest rate in the above table for year 2 is 5% per annum
Rate of interest implied by zero rates for period between end of 1st year/end of 2nd year
Calculated from 1-year zero rate of 3% per annum and 2-year zero rate of 4% per annum
Equals rate for year 2 that, when combined with 3% for year 1, gives 4% overall for the 2 years
When interest rates are continuously compounded and rates in successive time periods are combined,
the overall equivalent rate is simply the average rate during the whole period
If R1 /R2 are zero rates for maturities T1 /T2 , and RF = forward rate for period [T1 , T2 ], then:
RF =

R2 T2 R1 T1
T2 T1

RF = R2 + (R2 R1 )

T1
T2 T1

(2)

If the zero curve is upward sloping between T1 and T2 , so that R2 > R1 , then RF > R2 (i.e., the
forward rate for a period of time ending at T2 is greater than the T2 zero rate)
If zero curve is downward sloping (R2 < R1 ), then RF < R2 (forward rate < T2 zero rate)
83

 Instantaneous forward rate for a maturity of T

Taking limits as T2 approaches T1 :
RF = R + T

R
T

where R is the zero rate for a maturity of T

Value of RF obtained in this way is known as instantaneous forward rate for a maturity of T
This is the forward rate applicable to a very short future time period that begins at time T
Defining P (0, T ) as the price of a zero-coupon bond maturing at time T , since P (0, T ) = eRT ,
the instantaneous forward rate can also be written as:
RF =

ln P (0, T )
T

Assuming that the zero rates for borrowing and investing are the same (which is close to the truth for
a large financial institution), an investor can lock in the forward rate for a future time period
Lending at the locked forward rate
E.g., investor borrows $100 at 3% for 1 year and invests the money at 4% for 2 years
Result: Cash outflow of 100e0.031 = 103.05 at EoY1 , inflow of 100e0.042 = 108.33 at EoY2
108.33 = 103.05e0.05 Return equal to forward rate (5%) earned on $103.05 during 2nd year
Borrowing at the locked forward rate
E.g., investor borrows $100 for four years at 5% and invests it for three years at 4.6%
Result: Cash inflow of 100e0.0463 = 114.80 at EoY3 , outflow of 100e0.054 = 122.14 at EoY4
Since 122.14 = 114.80e0.062 money is being borrowed for the 4th year at forward rate of 6.2%
If an investor thinks that rates in the future will be different from todays forward rates there are many
trading strategies that the investor will find attractive
Forward rate agreements
A forward rate agreement (FRA) is an over-the-counter agreement that a certain interest rate will apply
to either borrowing or lending a certain principal during a specified future period of time
Assumption underlying the contract: Borrowing/lending would normally be done at LIBOR
Consider FRA where X agrees to lend money to Y for period [Tl , T2 ]. Define:
RK :
RF :
RM :
L:

The
The
The
The

rate of interest agreed to in the FRA

forward LIBOR interest rate for the period between T1 and T2 , calculated today
actual LIBOR interest rate observed in the market at time T1 for period [T1 , T2 ]
principal underlying the contract

Assume that the rates RK , RF , and RM are all measured with a compounding frequency reflecting
the length of the period to which they apply
Normally X would earn RM from the LIBOR loan. FRA means that it will earn RK
Extra interest rate (may be negative) earned as result of entering into FRA is RK RM
The interest rate is set at time T1 and paid at time T2
The extra interest rate therefore leads to a cash flow to X at time T2 of:
L(RK RM )(T2 T1 )

(3)

Similarly there is a cash flow to Y at time T2 of:

L(RM RK )(T2 T1 )

(4)

Interpretation of FRA
It is an agreement where X will receive interest on the principal between T1 and T2 at the fixed
rate of RK and pay interest at the realized market rate of RM
Y will pay interest on principal between T1 and T2 at fixed rate of RK and receive interest at RM
Usually FRAs are settled at time T1 rather than T2
The payoff must then be discounted from time T2 to T1
84

 The payoff at time T1 for each company is then:

X:

L(RK RM )(T2 T1 )
1 + RM (T2 T1 )

and

Y:

L(RM RK )(T2 T1 )
1 + RM (T2 T1 )

Valuation of FRAs
An FRA is always worth zero when RK = RF . This is because, a large financial institution can at
no cost lock in the forward rate for a future time period
E.g., it can ensure that it earns the forward rate for the time period between years 2 and 3 by
borrowing a certain amount of money for 2 years and investing it for 3 years
Similarly, it can ensure that it pays the forward rate for the time period between years 2 and
3 by borrowing for a certain amount of money for 3 years and investing it for 2 years
Compare two FRAs:
1st promises that LIBOR forward rate RF will be earned on principal of L between T1 and T2
2nd promises that RK will be earned on the same principal between same two dates
The two contracts are the same except for the interest payments received at time T2
Excess value of 2nd contract over 1st = PV of difference between interest payments:
L(RK RF )(T2 T1 )eR2 T2

where R2 = riskless zero rate for maturity T2

Hence, the value of the FRA where RK is earned is:

VF RA = L(RK RF )(T2 T1 )eR2 T2

(5)

Similarly, for company receiving interest at floating rate/paying interest at RK , value of FRA is:
VF RA = L(RF RK )(T2 T1 )eR2 T2

(6)

An FRA can be valued if we:

1. Calculate the payoff on the assumption that forward rates are realized (i.e., RM = RF )
2. Discount this payoff at the risk-free rate
Duration
Duration: Measure of how long on avg. the bondholder has to wait before receiving cash payments
A zero-coupon bond that lasts n years has a duration of n years
However, a coupon-bearing bond lasting n years has a duration of less than n years
Suppose that a bond provides the holder with cash flows ci at time ti (1 i n). The bond price B
and bond yield y (continuously compounded) are related by:
X
B=
ci eyti
(7)
The duration D of the bond is defined as:
P
 yti 
n
X
ti ci eyti
ci e
D=
=
ti
B
B

(8)

i=1

The term in square brackets is the ratio of the PV of the CF at time ti to the bond price
The bond price is the PV of all payments
For purposes of duration, all discounting is done at the bond yield rate of interest y
The duration is a weighted average of the times when payments are made, with the weight applied
to time ti being equal to the proportion of the bonds total PV provided by the CF at time ti
When a small change y in the yield is considered, it is approximately true that:
n

B =

X
dB
y = y
ci ti eyti
dy

(9)

i=1

85

 Key duration relationship

B = BDy

B
= Dy
B

(10)

Modified duration
If y is expressed with annual compounding, the approximate relationship in Eq. (10) becomes:
B =

BDy
1+y

More generally, if y is expressed with a compounding frequency of m times per year, then:
B =

BDy
1 + y/m

Modified duration D
D =

D
1 + y/m

When y is expressed with a compounding frequency of m times per year, the bonds modified
duration D allows the duration relationship to be simplified to:
B = BD y

(11)

Dollar duration
This is the product of modified duration and bond price
B = D y, where D = B D is the dollar duration
Bond portfolios
The duration D of a bond portfolio can be defined as a weighted average of the durations of the
individual bonds in the portfolio, with the weights being proportional to the bond prices
Eqs. (10) and (11) then apply, with B being defined as the value of the bond portfolio
They estimate change in value of bond portfolio for small change y in yields of all the bonds
Implicit assumption that the yields of all bonds will change by approximately the same amount
When bonds have differing maturities, this means a parallel shift in zero-coupon yield curve
Convexity
The duration relationship applies only to small changes in yields
For large yield changes, convexity measures the curvature of the yield curve [improves Eq. (10)]
A measure of convexity C is:
P 2 yti
1 d2 B
ci ti e
C=
=
2
B dy
B
From Taylor expansion, we obtain a more accurate expression than Eq. (10):
B =

dB
1 d2 B
y +
y 2
dy
2 dy 2

B
1
= Dy + C(y)2
B
2

Convexity of bond portfolio greatest when portfolio provides payments evenly over long period of time
It is least when the payments are concentrated around one particular point in time
By choosing a portfolio of assets and liabilities with a net duration of zero and a net convexity of zero,
a financial institution can make itself immune to relatively large parallel shifts in the zero curve
However, it is still exposed to nonparallel shifts
Theories of the term structure of interest rates
What determines the shape of the zero curve?
86

 Expectations theory
Conjectures that long-term interest rates should reflect expected future short-term interest rates
More precisely, it argues that a forward interest rate corresponding to a certain future period is
equal to the expected future zero interest rate for that period
Market segmentation theory
Conjectures that no relationship between short-, medium-, and long-term interest rates
Investors choose bonds of certain maturity and do not readily switch from one maturity to another
The short-term interest rate is determined by supply/demand in short-term bond market
Medium-term interest rate is determined by supply/demand in medium-term bond market . . .
Liquidity preference theory
Most appealing theory: Argues that forward rates always higher than expected future zero rates
Basic assumption: Investors prefer to preserve liquidity and invest funds for short periods of time
Borrowers, on the other hand, usually prefer to borrow at fixed rates for long periods of time
Liquidity preference theory leads to forward rates > expected future zero rates
Consistent with empirical result that yield curves tend to be upward sloping
The management of net interest income
Banks net interest income: Excess of interest received over interest paid. Needs to be managed
Suppose you have money to deposit and agree with the prevailing view that interest rate increases
are just as likely as interest rate decreases
Would you deposit your money for 1 year at 3% per annum or for 5 years at 3% per annum?
Choose one year because more financial flexibility: Funds tied up for shorter period of time
Now suppose that you want a mortgage
Choose 5-year mortgage because it fixes borrowing rate Less refinancing risk
When banks have same rates for various maturities, customers opt for 1-yr deposits/5-yr mortgages
This creates an asset/liability mismatch for the bank and subjects it to risks
If rates rise, deposits financing 6% loans will cost more and net interest income declines
Asset/liability management group must ensure that the maturities of the assets on which interest
is earned and the maturities of the liabilities on which interest is paid are matched
One way it can do this is by increasing the five-year rate on both deposits and mortgages
The net result of all banks behaving in such way is liquidity preference theory:
Long-term rates tend to be higher than predicted by expected future short-term rates
The yield curve is upward sloping most of the time
It is downward sloping only when market expects a really steep decline in short-term rates
Sometimes derivatives such as interest rate swaps are also used to manage exposure
The result of all this is that net interest income is very stable

87

88

BKM - Ch. 15: The term structure of interest rates

Introduction
Relationship between time to maturity and YTM can vary dramatically from one period to another
Term structure of interest rates: Structure of interest rates to discount CFs of different maturities
Reveals market-consensus forecasts of future interest rates. Interest rate risk may affect those inferences
Traders use term structure to compute forward rates on forward or deferred loans
The yield curve
Yield curve
Summarizes the relationship between yield and maturity
Plot of yield to maturity as a function of time to maturity
Key concerns of fixed-income investors and central to bond valuation
Allows investors to gauge their expectations for future interest rates against those of the market
Bond pricing
If yields on different-maturity bonds are not all equal, how to value coupon bonds that make
payments at many different times? Consider each bond cash flow - either coupon or principal
payment - as at least potentially sold off separately as a stand-alone zero-coupon bond
Treasury stripping suggests exactly how to value a coupon bond
If each cash flow can be sold off as a separate security, then the value of the whole bond should
be the same as the value of its cash flows bought piece by piece in the STRIPS market
If it werent, then easy profits to be made: If bond sold for less than the amount at which the
sum of its parts could be sold, buy the bond, strip it into stand-alone zero-coupon securities,
sell off the stripped CFs, and profit by the price difference
Both bond stripping and bond reconstitution offer opportunities for arbitrage
Because each coupon payment matures at a different time, discount by using yield appropriate to
its particular maturity, i.e. yield on a Treasury strip maturing at the time of that CF
Example - Valuing coupon bonds
Value a 10% coupon bond with a maturity of 3 years when the yield to maturity for zero
coupon bonds of maturities of 1, 2 and 3 years are 5%, 6% and 7% respectively
The first cash flow, the $100 coupon paid at the end of the first year, is discounted at 5%. The
second cash flow, the $100 coupon at the end of the second year, is discounted at 6%
Final CF (final coupon plus par value = $1,100) discounted at 7% Bond value:
100
100
1100
+
+
= 1, 082.17
1.05 1.062 1.073
Its YTM is 6.88%: While its maturity matches that of the 3-year zero, its yield is a bit lower
This reflects the fact that the 3-year coupon bond may usefully be thought of as a portfolio of
three implicit zero-coupon bonds, one corresponding to each cash flow.
Yield on coupon bond = amalgam of yields on each of the three portfolio components
The pure yield curve refers to the curve for stripped, or zero-coupon, Treasuries
In contrast, the on-the-run yield curve refers to the plot of yield as a function of maturity for
recently issued coupon bonds selling at or near par value
On-the-run Treasuries have greatest liquidity Traders interested in their yield curve
The yield curve and future interest rates
The yield curve under certainty
If interest rates are certain, what should we make of the fact that the yield on the 2-year zero
coupon bond is greater than that on the 1-year zero?
It cant be that one bond is expected to provide a higher rate of return than the other
Upward-sloping yield curve: Evidence that short-term rates are going to be higher next year
Terminology to distinguish yields on long-term bonds from short-term rates available in future:
They call the yield to maturity on zero-coupon bonds the spot rate y, meaning the rate that
prevails today for a time period corresponding to the zeros maturity
89

 In contrast, the short rate r for a given time interval (e.g., 1 year) refers to the interest rate
for that interval available at different points in time
E.g., the 2-year spot rate y2 is an average of todays short rate r1 and next years short rate r2 .
But because of compounding, that average is a geometric one: (1 + y2 )2 = (1 + r1 ) (1 + r2 )
At least in part, the yield curve reflects the markets assessments of coming interest rates
When next years short rate r2 is greater than this years short rate r1 , the average of the two
rates is higher than todays rate, so y2 > r1 , and the yield curve slopes upward
If next years short rate were less than r1 , the yield curve would slope downward
Example - Finding a future short rate
Compare 3yr strategies: (i) Buy a 3yr zero, with YTM = 7%, and hold to maturity, (ii) Buy
a 2yr zero yielding 6%, and roll proceeds in year 3, at short rate r3
We must have: Buy/hold 3-year zero = Buy 2-year zero/roll into 1-year bond
(1 + y3 )3 = (1 + y2 )2 (1 + r3 )

r3 = (1 + y3 )3 /(1 + y2 )2 1

Yield on 3-year bond reflects geometric average of discount factors for next 3 years:
1 + y3 = [(1 + r1 ) (1 + r2 ) (1 + r3 )]1/3
Yield/spot rate on long-term bond reflects path of short rates anticipated by market
Holding-period returns
Multiyear cumulative returns on all of our competing bonds ought to be equal
This conclusion holds for holding-period returns over shorter periods such as a year
In a world of certainty, all bonds must offer identical returns, or investors will flock to the higherreturn securities, bidding up their prices, and reducing their returns
Example - Holding-period returns on zero-coupon bonds
1-year bond can be bought today for 1, 000/1.05 = 952.38 and matures to par value in 1 year
Rate of return is (1, 000 952.38)/952, 38 = .05
The 2-year bond can be bought for 1, 000/1.062 = 890, 00. Next year, the bond will have a
remaining maturity of 1 year and the 1-year interest rate will be 7.01%
Therefore, its price next year will be 1, 000/1.0701 = 934.49, and its 1-year holding-period rate
of return will be (934.49 890.00)/890.00 = .05, for an identical 5% rate of return
Forward rates
Generalization to inferring a future short rate from the yield curve of zero-coupon bonds:
Equate total return on two n-year strategies: (i) Buying/holding n-year zero-coupon bond vs.
(ii) Buying (n 1)-year zero and rolling over proceeds into 1-year bond
(1 + yn )n = (1 + yn1 )n1 (1 + rn )

(1)

Given the observed yield curve, we can solve Eq. (1) for the short rate in the last period:
(1 + rn ) =

(1 + yn )n
(1 + yn1 )n1

(2)

Numerator on RHS: Total growth factor of investment in n-year zero held to maturity
Similarly, the denominator is the growth factor of an investment in an (n 1)-year zero
The difference in these growth factors is the rate of return available in year n when the (n 1)year zero can be rolled over into a 1-year investment
Forward interest rate
Future interest rates are uncertain Interest rate infered is the forward interest rate rather
than the future short rate (need not be interest rate that actually prevails at future date)
If the forward rate for period n is denoted fn , we then define fn by:
(1 + fn ) =

(1 + yn )n
(1 + yn )n = (1 + yn1 )n1 (1 + fn )
(1 + yn1 )n1

Forward rates equal future short rates in the special case of interest rate certainty
90

(3)

Interest rate uncertainty and forward rates

Short-term investors
When we account for risk, it is clear that short-term investors will shy away from the long-term
bond unless it offers an expected return greater than that of the 1-year bond
Investors require a risk premium to hold longer-term bond: Risk-averse investors will hold longterm bond only if expected value of short rate < break-even value f2 (the lower the expectation of
r2 , the greater the anticipated return on the long-term bond)
If most individuals are short-term investors, bonds must have prices that make f2 > E(r2 )
The forward rate will embody a premium compared with the expected future short-interest rate
This liquidity premium compensates short-term investors for the uncertainty about the price at
which they will be able to sell their long-term bonds at the end of the year
Long-term investors
Whenever that risk is important, the long-term investor will not be willing to engage in the rollover
strategy unless its expected return exceeds that of the 2-year bond
LT investor requires that expected value of next years short rate > forward rate: E(r2 ) > f2
Therefore, if all investors were long-term investors, no one would be willing to hold short-term
bonds unless those bonds offered a reward for bearing interest rate risk
Bond prices would be at levels such that rolling over short bonds resulted in greater expected return
than holding long bonds Forward rate < expected future spot rate
The liquidity premium f2 E(r2 ) therefore is negative
Whether forward rates equal expected future short rates depends on investors readiness to bear interest
rate risk and willingness to hold bonds that do not correspond to their investment horizons
Theories of the term structure
The expectations hypothesis
Simplest theory of the term structure: States that the forward rate equals the market consensus
expectation of the future short interest rate: f2 = E(r2 ) and liquidity premiums are zero
If f2 = E(r2 ), we may relate yields on long-term bonds to expectations of future interest rates
Can use forward rates derived from yield curve to infer market expectations of future short rates
Liquidity preference
Short-term investors unwilling to hold long-term bonds unless f2 > E(r2 ), whereas long-term
investors unwilling to hold short-term bonds unless E(r2 ) > f2
Both groups require a premium to hold bonds with maturities 6= investment horizons
Liquidity preference theory
Short-term investors dominate the market so that forward rate > expected short rate
The excess of f2 over E(r2 ), the liquidity premium, is predicted to be positive

Interest Rate (%)

Interest Rate (%)

Forward rate
Forward rate
Yield curve is upward-sloping
Constant liquidity premium
Expected short rate is constant

Yield

curve

Expected short rate is falling

Liquidity premium
Increases with
maturity
) Rising yield curve
despite falling expected
interest rates
Year

Interest Rate (%)

Interest Rate (%)

te

Constant liquidity premium

Expected short rate is falling

Year

d ra
war
For

Yield curve is humped

Figure 1: Yield curves

91

d
war
For

rate

Year
Liquidity premium
Increases with
maturity
) Yield curve
rises sharply

Expected short rate is increasing

Year

Interpreting the term structure

Yield curve reflects expectations of future short rates Powerful tool for fixed-income investors
If we can use the term structure to infer future interest rate expectations of other investors in the
economy, we can use those expectations as benchmarks for our own analysis
Unfortunately, the yield curve also reflects other factors such as liquidity premiums
Moreover, forecasts of interest rate changes may have different investment implications depending
on whether those changes are driven by changes in the expected inflation rate or the real rate
Certainty vs. uncertainty
Under certainty, 1 plus the yield to maturity on a zero-coupon bond is simply the geometric average
of 1 plus the future short rates that will prevail over the life of the bond
1 + yn = [(1 + r1 )(1 + r2 ) (1 + rn )]1/n
When future rates are uncertain, we replace future short rates with forward rates:
1 + yn = [(1 + r1 )(1 + f2 ) (1 + fn )]1/n

(4)

Direct relationship between yields on various maturity bonds and forward interest rates
What factors can account for a rising yield curve?
The yield curve is upward-sloping at any maturity date n for which the forward rate for the coming
period is greater than the yield at that maturity
What can account for that higher forward rate?
The forward rate can be related to the expected future short rate according to:
fn = E(rn ) + Liquidity premium

(5)

Two explanations for a higher forward rate:

1. Either investors expect rising interest rates, meaning that E(rn ) is high
2. Or they require a large premium for holding longer-term bonds
Although expectations of increases in future interest rates can result in rising yield curve,
converse not true: Rising yield curve does not necessarily imply expectations of higher future rates
Liquidity premiums effects confound simple attempt to extract expectations from term structure
Rough approach to deriving expected future spot rates: Assume that liquidity premiums are constant
Markets expected interest rate = [Forward rate] - [Estimate of liquidity premium]
This approach has little to recommend it for two reasons:
1. It is next to impossible to obtain precise estimates of a liquidity premium:
The general approach to doing so would be to compare forward rates and eventually realized
future short rates and to calculate the average difference between the two
However, the deviations between the two values can be quite large and unpredictable
because of unanticipated economic events that affect the realized short rate
2. There is no reason to believe that the liquidity premium should be constant
Still, steep yield curves are interpreted by professionals as warning signs of impending rate increases:
In fact, the yield curve is a good predictor of the business cycle as a whole, because long-term rates
tend to rise in anticipation of an expansion in economic activity
When steep curve, far lower probability of recession next year than when curve inverted/falling
The yield curve is a component of the index of leading economic indicators
Valid to interpret downward-sloping yield curve as evidence that interest rates are expected to decline?
If term premiums (spread between yields on long- and short-term bonds) generally are positive,
then downward-sloping curve signals anticipated declines in rates (possible impending recession)
Why might interest rates fall? Two factors: (i) The real rate, and (ii) The inflation premium
Nominal interest rate composed of real rate plus factor to compensate for inflation:
1 + Nominal rate = (1 + Real rate)(1 + Inflation rate)
92

 Important to distinguish between changes in expected real rate/expected inflation rate because
economic environments associated with them may vary substantially
High real rates indicate rapidly expanding economy, high gov. deficits, tight monetary policy
High inflation rates can arise out of rapidly expanding economy, but also caused by rapid
expansion of money supply/supply-side shocks to the economy (oil supplies trouble)
Forward rates as forward contracts
In general, forward rates 6= eventually realized short rate, or even todays expected short rate
But there is still an important sense in which the forward rate is a market interest rate
Suppose that you wanted to arrange now to make a loan at some future date:
Need to agree today on the interest rate that will be charged
Interest rate on such forward loan would be the forward rate of interest for the loan period
To construct synthetic forward loan, sell (1 + f2 ) 2-year zeros for every 1-year zero that you buy
Denote B0 (T ) todays price of a zero maturing at time T
Pay B0 (1) for a zero maturing in 1 year, receive B0 (2) for each zero you sell maturing in 2 years
Initial CF = 0 because prices of the 1-/2-year zeros differ by factor (1 + f2 )
B0 (1) =

$1, 000
(1 + y1 )

while

B0 (2) =

$1, 000
$1, 000
=
2
(1 + y2 )
(1 + y1 )(1 + f2 )

Therefore, selling (1 + f2 ) 2-year zeros generates just enough cash to buy one 1-year zero
Both zeros mature to face value of $1,000 Difference between cash inflow at t = 1 and cash
outflow at t = 2 is 1 + f2 (see Fig. 2) f2 is the rate on the forward loan
General forward rate: The short rates in the two periods are r1 (which is observable
today) and r2 (which is not).
Amount borrowed one year from now: $1,000

Present
0

r1

r2

Amount repaid one year after borrowing: -$1,000(1 + f2)

The rate that can be locked in for a one-period-ahead loan is f2

Figure 2: Engineering a synthetic forward loan

93

94

Hull - Ch. 6.1: Day count and quotation conventions

Day counts
The day count defines the way in which interest accrues over time
The day count convention is usually expressed as X/Y
X = Number of days between the two dates, as measured by convention
Y = Total number of days in the reference period, as measured by convention
The interest earned between the two dates is:
Number of days between dates
Interest earned in reference period
Number of days in reference period
Three day count conventions commonly used in the US:
1. Actual/actual (in period)
2. 30/360
3. Actual/360
Actual/actual day count
Used for Treasury bonds in the US
The interest earned between two dates is based on the ratio of the actual days elapsed to the actual
number of days in the period between coupon payments
Suppose that the coupon payment dates are March 1 and September 1, the coupon rate is 8%, and
we wish to calculate the interest earned between March 1 and July 3
The reference period is from March 1 to September 1
There are 184 (actual) days in this period, and interest of $4 is earned during the period
There are 124 (actual) days between March 1 and July 3
The interest earned between March 1 and July 3 is therefore
124
4 = 2.6957
184
30/360 day count
Used for corporate and municipal bonds in the US
Assumes 30 days per month and 360 days per year when carrying out calculations
With the 30/360 day count, the total number of days between March 1 and September 1 is 180
The total number of days between March 1 and July 3 is (4 30) + 2 = 122
In a corporate bond with the same terms as the Treasury bond above, the interest earned between
March 1 and July 3 would therefore be:
122
4 = 2.7111
180
Actual/360 day count
Used for money market instruments in the US
The reference period is 360 days
The interest earned during part of a year is calculated by dividing the actual number of elapsed
days by 360 and multiplying by the rate
The interest earned in a whole year of 365 days is 365/360 times the quoted rate
Conventions vary from country to country and from instrument to instrument
LIBOR is quoted on an actual/360 for all currencies except sterling (actual/365 basis)
Euro-denominated and sterling bonds are usually quoted on an actual/actual basis
Price quotations
The prices of money market instruments are sometimes quoted using a discount rate
This is the interest earned as a % of the final face value rather than as a % of the initial price paid

95

 Example: Treasury bills in the US

If price of 91-day T-bill is quoted as 8, the annualized rate of interest earned is 8% of face value
Interest of $2.0222 (= 100 0.08 91/360) is earned over the 91-day life
This corresponds to true rate of interest of 2.0222/(100 2.0222) = 2.064% for the 91-day period
In general, the relationship between the cash price and quoted price of a Treasury bill in the US is:
P =

360
(100 Y )
n

Where P = quoted price, Y = cash price, and n = remaining life of T-bill measured in calendar days
US Treasury bonds
Treasury bond prices in the United States are quoted in dollars and thirty-seconds of a dollar
The quoted price is for a bond with a face value of $100 A quote of 90-05 indicates that the
quoted price for a bond with a face value of $100,000 is $90,156.25 = 1, 000 (90 + 5/32)
Definitions
Clean price: The quoted price
Dirty price: The cash price paid by the purchaser of the bond
Cash price = Quoted price + Accrued interest since last coupon date
Example
Suppose on March 5, 2010, the bond under consideration is an 11% coupon bond maturing on July
10, 2018, with a quoted price of 95-16 or $95.50
Because coupons are paid semiannually on government bonds, the most recent coupon date is
January 10, 2010, and the next coupon date is July 10, 2010
The number of days between January 10, 2010, and March 5, 2010, is 54, whereas the number of
days between January 10, 2010, and July 10, 2010, is 181
On a bond with $100 face value, the coupon payment is $5.50 on January 10 and July 10
Accrued interest on March 5, 2010 = share of July 10 coupon accruing on March 5, 2010
Because actual/actual in period is used for Treasury bonds in the US, this is:
54
$5.5 = $1.64
181
The cash price per $100 face value for the bond is therefore: 95.50 + 1.64 = 97.14
The cash price of a $100,000 bond is $97,140

96

Hull - Ch. 22 - Part 1: Credit risk

Credit ratings
Credit risk arises from possibility that borrowers/counterparties in derivatives transactions may default
Rating agencies (Moodys, S&P, Fitch) provide ratings describing creditworthiness of corporate bonds
Moodys ratings
Moodys best rating: Aaa. Aaa bonds considered to have almost no chance of default
The next best rating is Aa. Following that comes A, Baa, Ba, B, Caa, Ca, and C
Only bonds with ratings of Baa or above are considered to be investment grade
S&P ratings
The S&P ratings corresponding to Moodys Aaa, Aa, A, Baa, Ba, B, Caa, Ca, and C are AAA,
AA, A, BBB, BB, B, CCC, CC, and C, respectively
Fitchs rating categories are similar to those of S&P
To create finer rating measures, Moodys divides its Aa rating category into Aal, Aa2, and Aa3, . . .
Similarly, S&P divides its AA rating category into AA+, AA, and AA-, . . .
Moodys Aaa/S&Ps AAA category not subdivided, nor are the two lowest rating categories
Historical default probabilities
The data produced by rating agencies shows the default experience during a 20-year period of bonds
that had a particular rating at the beginning of the period
For investment grade bonds, probability of default in a year is an increasing function of time
This is because the bond issuer is initially considered to be creditworthy, and the more time
that elapses, the greater the possibility that its financial health will decline
For bonds with a poor credit rating, probability of default is often a decreasing function of time
Reason: For a bond with a poor credit rating, the next year or two may be critical
The longer the issuer survives, the greater the chance that its financial health improves
Default intensities
Unconditional default probability
The probability of default during the n-th year year as seen at time 0
Conditional default probabilities - Default intensities / Hazard rates
Probability that bond defaults during the n-th year conditional on no earlier default
The default intensity (t) at time t is then defined so that (t)t is the probability of default
between time t and t + t conditional on no earlier default
Denoting V (t) the cumulative probability of the company surviving to time t (i.e., no default by
time t), the conditional probability of default between time t and t + t is [V (t) V (t + t)]/V (t)
Since this is equal to (t)t, it follows that:
Rt
dV (t)
= (t)V (t) V (t) = e 0 ( )d
dt
Defining Q(t) as the probability of default by time t, so that Q(t) = 1 V (t), gives:
Rt

Q(t) = 1 e 0 ( )d = 1 e(t)t

Where (t)
is the average default intensity (hazard rate) between time 0 and time t

(1)

Recovery rates
Recovery rate for a bond: Bonds market value immediately after default, as % of its face value
Senior secured debt holders had an average recovery rate of 54.44 cents per dollar of face value
Junior subordinated debt holders had avg. recovery rate of 24.47 cents per dollar of face value
Recovery rates are significantly negatively correlated with default rates
Moodys looked at avg. recovery rates/avg. default rates each year between 1982-06
Recovery = 59.1 - 8.356 Default rate
The recovery rate is the average recovery rate on senior unsecured bonds in a year measured as a
% and the default rate is the corporate default rate in the year measured as a %
Bad year for default rate is doubly bad because accompanied by low recovery rate
97

Estimating default probabilities from bond prices

The probability of default for a company can be estimated from the prices of bonds it has issued
Assumption: Corporate bond sells for less than similar risk-free bond due to possibility of default
In practice, bond prices are affected by liquidity: The lower the liquidity, the lower the price
Quick calculation of the average default intensity
Suppose that a bond yields s basis points more than a similar risk-free bond and that the expected
recovery rate in the event of a default is R%
The holder of the bond must be expecting to lose s basis points per year from defaults
Given recovery rate of R%, probability of default per year conditional on no earlier default is:
= s
= avg. default intensity (hazard rate) per year

where
(2)
1R
A more exact calculation
Suppose that the corporate bond we have been considering lasts for 5 years, provides a coupon 6%
per annum (paid semiannually) and that the yield on the corporate bond is 7% per annum (with
continuous compounding). The yield on a similar risk-free bond is 5%
The yields imply that corporate bond price is 95.34 and risk-free bond price is 104.09
The expected loss from default over the 5-year life of the bond is 104.09 95.34 = 8.75
Denoting the probability of default per year Q, the table below calculates the expected loss from
default in terms of Q on the assumption that defaults can happen at times 0.5, 1.5, 2.5, 3.5, and
4.5 years (immediately before coupon payment dates)
Time
(years)
0.5
1.5
2.5
3.5
4.5
Total

Default
probability
Q
Q
Q
Q
Q

Recovery
amount ($)
40
40
40
40
40

Risk-free
value ($)
106.73
105.97
105.17
104.34
103.46

Loss given
default ($)
66.73
65.97
65.17
64.34
63.46

Discount
factor
0.9753
0.9277
0.8825
0.8395
0.7985

PV of expected
loss ($)
65.08Q
61.20Q
57.52Q
54.01Q
50.67Q
288.48Q

Calculation example
Expected value of corp. bond at time 3.5 years (using forward interest rates, no default) is:
3 + 3e0.050.5 + 3e0.051.5 + 103e0.051.5 = 104.34
Amount recovered if default = 40 loss given default is 104.34 40 = 64.34
The present value of this loss is 54.01 The expected loss is 54.01Q
Total expected loss = 288.48Q = 8.75 Value for Q of 8.75/288.48 = 3.03%
With several bonds, several parameters describe the term structure of default probabilities
Suppose we have bonds maturing in 3, 5, 7, and 10 years
We could use the first bond to estimate a default probability per year for the first 3 years, the
second bond to estimate default probability per year for years 4 and 5, . . .
Approach analogous to bootstrap procedure for calculating a zero-coupon yield curve
The risk-free rate
Key issue when bond prices used to estimate default probabilities: risk-free rate/risk-free bond
In Eq. (2), spread s = excess of corporate bond yield over yield on similar risk-free bond
In table above, risk-free value of the bond must be calculated using risk-free discount rate
Benchmark risk-free rate for corporate bond yields = yield on similar Treasury bonds
Traders usually use LIBOR/swap rates as proxies for risk-free rates when valuing derivatives
Traders also use LIBOR/swap rates as risk-free rates when calculating default probabilities
When they determine default probabilities from bond prices, the spread s in Eq. (2) is the
spread of the bond yield over the LIBOR/swap rate
Risk-free discount rates used in calculations in table above are LIBOR/swap zero rates
Credit default swaps can be used to imply the risk-free rate assumed by traders
Implied rate LIBOR/swap rate minus 10 basis points on average
98

 Asset swaps
In practice, traders use asset swap spreads to extract default probabilities from bond prices
Asset swap spreads provide direct estimate of bond yields spread over LIBOR/swap curve
E.g., asset swap spread for a bond is quoted as 150 basis points. Three possible situations:
1. The bond sells for its par value of 100
The swap then involves one side (company A) paying the coupon on the bond and the
other side (company B) paying LIBOR plus 150 basis points
Only coupons are exchanged; Exchanges take place regardless of whether bond defaults
2. The bond sells below its par value, say, for 95
In addition to coupons, company A pays $5 per $100 of notional principal at outset
Company B pays LIBOR plus 150 basis points
3. The underlying bond sells above par, say, for 108
In addition to LIBOR plus 150 bp, company B pays $8 per $100 of principal at outset
Company A pays the coupons
The effect of all this is that the PV of the asset swap spread is the amount by which the price of
the corporate bond is exceeded by the price of a similar risk-free bond where the risk-free rate is
assumed to be given by the LIBOR/swap curve
Suppose that instead of knowing bonds price we know that asset swap spread is 150 basis points
This means that the amount by which the value of the risk-free bond exceeds the value of the
corporate bond is the present value of 150 basis points per year for 5 years
Assuming semiannual payments, this is $6.55 per $100 of principal
The total loss in the table above would in this case be set equal to $6.55
The default probability Q per year would be 6.55/288.48 = 2.27%
Comparison of default probability estimates
Default probabilities estimated from historical data  those derived from bond prices
Default intensities from historical data [Eq. (1)]: Default intensities from bond prices [Eq. (2)]:
= 1 ln[1 Q(t)]
(t)
t

=
(t)

s
1R

Ratio of default probability backed out from bond prices to default probability from historical data
is very high for investment grade companies and declines as a companys credit rating declines
Difference between the two default probabilities tends to increase as the credit rating declines
Excess return
The expected excess return on bond (in basis points) is the bond yield spread over Treasuries
minus the spread of risk-free rate over Treasuries minus the spread for historical default
A large percentage difference between default probability estimates translates into a small (but
significant) excess return on the bond
The excess return tends to increase as credit quality declines
Real-world vs. risk-neutral probabilities
The default probabilities implied from bond yields are risk-neutral probabilities of default
The calculations assume that expected default losses can be discounted at the risk-free rate
The risk-neutral valuation principle shows that this is a valid procedure providing the expected
losses are calculated in a risk-neutral world
The default probability Q in the table above is a risk-neutral probability
By contrast, default probabilities from historical data are real-world default probabilities
Expected excess return arises from difference real-world/risk-neutral default probabilities
If there were no expected excess return, then real-world = risk-neutral default probabilities
Why do we see such big differences between real-world/risk-neutral default probabilities?
One reason for the results is that corporate bonds are relatively illiquid and the returns on bonds
are higher than they would otherwise be to compensate for this
But this is small part of what is going on: Explains 25 basis points of excess return
99

 Other reason: The subjective default probabilities of bond traders are much higher
Bond traders may be allowing for depression scenarios much worse than anything seen earlier
However, difficult to see how this explains a large part of the observed excess return
By far, most important reason for results: Bonds do not default independently of each other
There are periods of time when default rates are very low and periods when very high
The year-to-year variation in default rates gives rise to systematic risk (i.e., risk that cannot
be diversified away) and bond traders earn an excess expected return for bearing the risk
Variation in default rates from year to year stems from overall economic conditions and because
a default by one company has ripple effect Defaults by other companies (credit contagion)
In addition to systematic risk, nonsystematic (idiosyncratic) risk associated with each bond
Bond returns are highly skewed with limited upside. This type of risk is difficult to diversify
It would require tens of thousands of different bonds
In practice, many bond portfolios are far from fully diversified
Bond traders earn extra return for bearing nonsystematic risk and systematic risk
When to use real-world or risk-neutral default probabilities in the analysis of credit risk?
When valuing credit derivatives/estimating impact of default risk, risk-neutral default probabilities should be used because the analysis calculates the PV of expected future CFs and
almost invariably (implicitly or explicitly) involves using risk-neutral valuation
For scenario analyses to calculate potential losses from defaults, use real-world probabilities
Using equity prices to estimate default probabilities
Credit ratings are revised relatively infrequently Some analysts argue that equity prices can provide
more up-to-date information for estimating default probabilities
Merton model
In Mertons model, a companys equity is an option on the assets of the company
Suppose firm has one zero-coupon bond outstanding and that bond matures at T . Define:
V0 :
VT :
E0 :
ET :

Value
Value
Value
Value

of
of
of
of

companys
companys
companys
companys

assets today
assets at time T
equity today
equity at time T

D:
V :
E :

Debt repayment due at time T

Volatility of assets (assumed constant)
Instantaneous volatility of equity

If VT < D, rational for company to default on debt at time T Value of equity = zero
If VT > D, company should make debt repayment at time T Value of equity at T = VT D
Mertons model gives the value of the firms equity at time T as: ET = max(VT D, 0)
Equity = call option on assets value with strike price equal to repayment required on debt
The Black-Scholes formula gives the value of the equity today as:

2 /2)T
ln V0 /D+(r+V

E0 = V0 N (d1 ) DerT N (d2 ) with d1 =

,
d
=
d

T
(3)
2
1
V

T
V

The value of the debt today is V0 E0

The risk-neutral probability that the company will default on the debt is N (d2 )
However, neither V0 nor V is directly observable
If the company is publicly traded, we can observe E0
Eq. (3) provides one condition that must be satisfied by V0 and V
We can also estimate E from historical data or options
From It
os lemma,
E E0 =

E
V V0
V

E E0 = N (d1 )V V0

(4)

Eqs. (3) and (4) provide a pair of simultaneous equations that can be solved for V0 and V
How well do default probabilities from Mertons model correspond to actual default experience?
They produce a good ranking of default probabilities (risk-neutral or real-world)
Monotonic transformation can be used to convert probability of default output from Mertons model
into good estimate of either real-world/risk-neutral default probability
100

Altman: Measuring corporate bond mortality and performance

Introduction
In addition to default risk, investors also consider the effects of the two other major risk dimensions of
investing in fixed-interest instruments: (i) Interest rate risk and, (ii) Liquidity risk
The appropriate measure of default risk and the accuracy of its measurement are critical in:
The pricing of debt instruments
The measurement of their performance
The assessment of market efficiency
Analysts have concentrated their efforts on measuring the default rate for finite periods of time (e.g.,
one year) and then averaging the annual rates for longer periods
The rate of default has been measured simply as the value of defaulting issues for some specific
population of debt compared with the value of bonds outstanding that could have defaulted
Annual default rates are then usually compared with observed promised yield spreads in order to
assess the attractiveness of particular bonds or classes of bonds
A corollary approach is to compare default rates with ex-post returns to assess whether investors
were compensated for the risks they bear
Although it is informative to measure default rates and losses based on the average annual rate method,
that traditional technique has at least two deficiencies:
1. It fails to consider that there are other ways in which a bond dies, namely redemptions from calls,
sinking funds, and maturation Fails to consider the surviving population of bonds
2. It does not answer the question of the probability of default for various time periods in the future
on the basis of an issues specific attributes at issuance, summarized into its bond rating
This study does explicitly consider the surviving population as the relevant basis or denominator in the
default calculation and addresses the initial default assessment by the following questions
Given an issues initial bond rating:
1. What is the probability of default/loss from default over a specific time horizon of N years?
2. What are the estimates of the cumulative annual mortality rates and losses for various time frames
as well as the marginal rates for specific one-year periods?
3. Given estimates of cumulative mortality losses suffered by investors and expected return spreads
earned on the surviving population of bonds, what were the net return spreads earned or lost in
comparison with returns on risk-free securities?
Prior studies
Previous works in the area of default were of three general kinds:
1. Hickman-style (1958) reports
Presents statistics on annual default rates/actual returns over various time frames
2. Studies which emphasize default risk potential of individual-company debt by examining determinants of risk premiums over risk-free securities, or by constructing uni-/multivariate classification
models based on combination of micro-finance measures/statistical classification techniques
Variants on those models were based on the gamblers ruin concept, recursive partitioning
techniques, and market indicators of survival
3. Fons study
Attempts to combine observed pricing and the inherent default risk premium with estimates
of corporate bond default experience
Incorporates default experience with risk-neutral investment strategy: The only factor that
matters is the return distribution of debt with no relevance for volatility or liquidity factors
Traditional measures of default rates and losses
Main distinction: Between so-called investment-grade and non-investment-grade categories
More precise ratings: 4 classes of investment-grade debt, 3 classes of lower quality junk bonds
Default rates are calculated on average annual basis, with individual rates for each year combined with
rates for other years, over longer time horizon to form estimate for the average annual rate
101

 The rate for each year is based on the dollar amount of defaulting issues in that year divided by
the total population outstanding as of some point during that year
Average annual default rate, measured in such way, for period 1978-87 was 1.86% per year
Default losses
More relevant default statistic for investors: Not the rate of default but amount lost from defaults
E.g., measure amount lost by tracking price for the defaulting issue just after default and
assuming that investor had purchased issue at par value and sold just after default
The investor also is assumed to lose one coupon payment
Average annual default loss over the sample period has been approximately 1.2% per year
That lower percentage of loss compared with default rates stems from the fact that defaulting
debt, on average, sells for approximately 40% of par at the end of the defaulting month
The

mortality rate concept

Begin with specific cohort (rating category) and track groups performance for multiple time periods
Consider mortalities in relation to survival population and input defaults to calculate mortality rates
Bonds can exit from the original population by means of at least four different events:
(i) Defaults, (ii) Calls, (iii) Sinking funds, and (iv) Maturities
The individual mortality rate for each year (marginal mortality rate = M M R) is calculated by:
M M R(t) =

Total value of defaulting debt in the year (t)

Total value of the population of bonds at the start of the year (t)

We then measure the cumulative mortality rate (CM R) over a specific time period (1, 2, , T years)
by subtracting the product of the surviving populations of each of the previous years from unity:
CM T(T ) = 1

T
Y
t=1

Where

CM T(T )
SR(t)

=
=

SR(t)

T
Y
= 1 (1 M M R(t) )
t=1

Cumulative mortality rate in (T )

Survival rate in (t) = 1 M M R(t)

The individual year marginal mortality rates for each bond rating are based on a compilation of that
years mortality measured from issuance
E.g., all one-year mortalities combined for 17-year sample period to get the one-year rate
All of the second-year mortalities are combined to get the two-year rate, etc . . .
Mortality rate = value-weighted rate for particular year after issuance, not simple average
Simple average Results susceptible to significant specific-year bias
Weighted-average correctly biases results toward larger-issue years, especially more recent years
Empirical results
Mortality rates
The relative results across cohort groups are pretty much in line with expectations, with the
mortality rates very low for the higher-rated bonds and increasing for lower rated issues
AAA debt: Zero mortality rate for first 5 yrs after issuance, 0.13% from 6-10 yrs
The mortality rates for BBB and lower bonds begin to increase almost immediately after
issuance, with BBB (the lowest investment-grade debt level) showing a cumulative rate of
0.91% after five years and 2.12% after ten years
The marginal mortality rates are fairly constant after year three
The longer term mortality results should be analyzed with considerable caution with respect to
expectations about future mortality rates and return spreads because of thin volume of data
In addition, later years biased since portion of original population redeemed by then
The traditional default rates are calculated on the basis of the population on June 30, while our
mortality rates use survival population data from the start of each year
The old way probably understates default rates somewhat
102

 Since we adjust population for all redemptions, mortality rates higher than if data unadjusted
Both the adjusted and unadjusted methods of calculating the results are meaningful
Mortality figures over time should adjust for changing population size, while unadjusted data
helpful for probability of default of specific rating from a given years issuance
Strictly speaking, however, the unadjusted figures are not rates
Losses
The loss to investors from defaults is of paramount importance
In analysis of net return spreads, we use the actual recovery amount for which investors were able
to sell the defaulting issue and also assume that one coupon payment was lost
The average recovery rate was slightly below 40% of par
We did look at the relation between individual bond ratings at issuance and the subsequent average
price that could be realized upon default and found essentially that no relationship existed
Virtually no correlation between initial bond rating and average price after default
No correlation between price after default and number of years bond in existence before default
While the marginal default rate is relatively low in the first three years after issuance, the
recovery rate is unaffected by the age of the issue
Net return performance
Analysis tracks performance of bonds from issuance, across ratings and over relevant time horizons
Compare performance of various risky bond categories with default risk-free US Treasuries
Factor into the analysis actual losses from defaults and yield spreads over Treasuries
We calculate actual return-spread performance
The spreads, expressed in terms of basis points compounded over a ten-year investment horizon,
are based on actual yield spreads for the 18-year period
Results: AAA bonds can be expected to earn 45 basis points (0.45%) more than Treasuries
over one year (two semiannual coupon payments) and 1245 basis points after ten years
BB bonds earn 326 basis points more than Treasuries after one year and 7637 basis points after
10 years. $100 would return $76.37 more than Treasuries over 10 years
Assumptions
These results use actual long-term Treasury coupon rates, yield spreads at birth for the different rating categories, the sale of defaulted debt, the loss of one coupon payment, and the
reinvestment of cash flows at the then prevailing interest rates for that bond-rating group
Cash flows are reinvested from coupon payments on the surviving population as well as the
reinvestment of sinking funds, calls, and the recovery from defaulted debt
The results assume no capital gains or losses over the measurement period, and the investor
follows a buy-and-hold strategy for the various horizons
For all holding periods, all bond types do well and have positive spreads over Treasuries
Average historical yield spreads ranged from 0.47% (AAA) to 3.05% (BB) to 7.07% (CCC)
Historical average 4.09% yield spread for B-rated debt provides an ample cushion to compensate
for losses, but performance relative to BB category is inferior in later years
This changes, however, if we adjust our initial yield spread assumptions to reflect different
market conditions, assuming the same default experience
Implications
Despite higher than expected cumulative mortality rates over long holding periods, return spreads on
all corporate bonds are positive, with impressive results for high-yield, low-grade categories
Investors have been more than satisfactorily compensated for investing in high-risk securities
Indeed, if expected default losses are fully discounted in the prices (and yields) of securities, our
return spread results should be insignificantly different from zero
Possible explanation: Fixed-income market has been mispricing corporate debt issues and discrepancy
has persisted, perhaps because of lack of appropriate info Market inefficiency
If default losses are consistently lower than yield spreads and this comparison is the only relevant
determinant of future yield spreads, inefficiency is a reasonable conclusion
103

If all other things are not equal, however, for determining yield spreads on corporate bonds, then
the market inefficiency conclusion is difficult to reach
Liquidity risk is often mentioned as important to price determination
If liquidity risk increases with lower bond ratings, then the excess returns noted earlier may in part
be the returns necessary to bear this risk
Indeed, during post-October 19, 1987 period, poor liquidity was cited as one cause of precipitous
drop in common stock prices and the rise in yields of certain high-yield debt issues
The other risk element that is not isolated in our study is interest rate or reinvestment risk
Actual returns on bonds are obviously affected by interest rate changes
Our results include actual reinvestment rates over time, and we have not factored in any capital
gains or losses, assuming a buy-and-hold strategy for investors
However, lower grade bonds have lower volatility from interest rate changes than Treasuries
Another explanation of the persistent positive return spreads attributed to lower rated bonds is the
variability of retention values after default
Our 40% recovery rate of par value just after default is an expected value
Investors might require positive spreads based on possibility that retention values < 40%
In addition, the 40% retention is relevant only for a portfolio of defaulting bonds
Investor not well-diversified is vulnerable to higher than avg. mortality losses on specific issues
If the market prices low-quality issues as individual investments and not as portfolios, required
spreads are likely to be higher than is perhaps necessary
On the other hand, if defaults are correlated with market returns, risks may not be as diversifiable
as we assume to be the case for equities
Investors might also be restricted in relation to the risk class of possible investments, thereby creating
an artificial barrier to supply-demand equilibrium
For instance, certain institutions are prohibited from investing in low-grade bonds or are limited
in the amount that they can invest in such securities
That reduces demand and inflates yield and possibly return spreads

104

Cummins: CAT Bonds and other risk-linked securities

Introduction
Risk-linked securities are financing devices that enable insurance risk to be sold in capital markets,
raising funds that insurers/reinsurers can use to pay claims arising from CAT/other loss events
The most prominent type of risk-linked security is the catastrophic risk (CAT) bond, which is a fully
collateralized instrument that pays off on the occurrence of a defined catastrophic event
CAT bonds and other risk-linked securities are potentially quite important because they have the ability
to access the capital markets to provide capacity for insurance and reinsurance markets
CAT bond market has expanded significantly in recent years and seems to have reached critical mass
Although the CAT bond market is small in comparison with the overall nonlife reinsurance market, it
is of significant size in comparison with the property-catastrophe reinsurance market
The structure of risk-linked securities
Early developments
Following Hurricane Andrew in 1992, efforts began to access securities markets directly as a mechanism for financing future catastrophic events
First contracts launched by the Chicago Board of Trade (CBOT), which introduced catastrophe
futures in 1992 and later introduced catastrophe put and call options
Options based on aggregate CAT loss indices compiled by Property Claims Services (PCS)
The contracts were later withdrawn due to lack of trading volume
In 1997, Bermuda Commodities Exchange (BCE) also tried to develop a market in CAT options, but contracts were withdrawn within 2 years (lack of trading)
Insurers had little interest in the CBOT and BCE contracts for various reasons:
Thinness of the market
Possible counterparty risk on the occurrence of a major catastrophe
Potential for disrupting long-term relationships with reinsurers
Possibility of excessive basis risk, i.e., the risk that payoffs under the contracts would be
insufficiently correlated with insurer losses
In 2007, the Chicago Mercantile Exchange (CME) and the New York Mercantile Exchange (NYMEX)
introduced futures and options contracts on US hurricane risk
Their introduction was motivated by the 2005 US hurricane season, which revealed the limitations on the capacity of insurance and reinsurance markets
CME contracts settle on Carvill (reinsurance intermediary) Hurricane Indices. NYMEX contracts settle on CAT loss indices by Gallagher Re based on PCS data
Given that both the CME and NYMEX contracts are based on broadly defined geographical
areas, they will be subject to significant basis risk
Given the existence of a secondary market as well as dedicated CAT bond mutual funds, it is
possible that the CME or NYMEX contracts could be used for hedging purposes
Another early attempt at securitization involved contingent notes aka Act of God bonds
In 1995, Nationwide issued $400 million in contingent notes through a special trust
Proceeds from the sale of the bonds were invested in 10-year Treasury securities, and investors
were provided with a coupon payment equal to 220 basis points over Treasuries
Embedded in these contingent notes was a substitutability option for Nationwide: Given a
pre-specified event that depleted Nationwides equity capital, Nationwide could substitute up
to $400 million of surplus notes for Treasuries in the Trust at any time during a 10-year period
for any business reason, with surplus notes carrying a coupon of 9.22%
Structure did not achieve significant segregation of Nationwides liabilities, leaving investors
exposed to general business risk and to risk that Nationwide might default on the notes
In addition, unlike CAT bonds, the withdrawal of funds from the trust would create the
obligation for Nationwide eventually to repay the Trust
Contingent notes have not emerged as a major solution to the risk-financing problem
105

 CAT bonds
Introduction to CAT Bonds
CAT bonds modeled on asset-backed-security transactions executed for a wide variety of financial assets including mortgage loans, automobile loans, aircraft leases, and student loans
CAT bonds are part of event-linked bonds, which pay off on occurrence of a specified event
Most event-linked bonds issued to date have been linked to catastrophes such as hurricanes/earthquakes,
although bonds also have been issued that respond to mortality events
First successful CAT bond was $85 million issue by Hannover Re in 1994 (Swiss Re, 2001)
The first CAT bond issued by a nonfinancial firm, occurring in 1999, covered earthquake losses
in the Tokyo region for Oriental Land Company, the owner of Tokyo Disneyland
More recently CAT bonds more standardized Driven by need for bonds to respond to
requirements of principal stakeholders (sponsors, investors, rating agencies, and regulators)
Properties of CAT Bonds
CAT bonds often issued to cover the high layers of reinsurance (return period of 100+ years)
Higher layers of protection often go unreinsured by ceding companies for two reasons:
1. For huge events, ceding insurers more concerned about credit risk of reinsurer
2. High layers have highest reinsurance margins/pricing spreads above expected loss
CAT bonds are fully collateralized Eliminate concerns about credit risk
CATs have low correlations with investment returns CAT bonds may provide lower spreads
than high-layer reinsurance because attractive for diversification
CAT bonds also can lock in multi-year protection, unlike traditional reinsurance, and shelter
the sponsor from cyclical price fluctuations in the reinsurance market
The multi-year terms (or tenors) of most CAT bonds also allow sponsors to spread the fixed
costs of issuing the bonds over a multi-year period, reducing costs on an annualized basis
Typical CAT bond structure
Transaction begins with formation of single purpose reinsurer (SPR): SPR issues bonds and
invests proceeds in safe, short-term securities (gov. bonds, AAA corp.), held in trust
Embedded in the bonds is a call option triggered by a defined CAT event: On occurrence of
event, proceeds are released from SPR to pay claims arising from the event
In most CAT bonds, principal fully at risk. In return, insurer pays premium to investors
Fixed returns on securities held in trust are swapped for floating returns based on LIBOR
Immunize insurer/investors from interest rate (mark-to-market) risk and default risk
The investors receive LIBOR plus the risk premium in return for providing capital to the trust.
If no contingent event occurs, principal returned to investors upon expiration
Some CAT bonds include principal protected tranches, with guaranteed return of principal
In this tranche, the triggering event would affect the interest and spread payments and the
timing of the repayment of principal
E.g., a 2-year CAT bond subject to the payment of interest and a spread premium might
convert into a 10-year zero-coupon bond that would return only the principal
Principal-protected tranches have become relatively rare, primarily because they do not provide
as much risk capital to the sponsor as a principal-at-risk bond
Insurers and investores prefer to use a SPR
Insurers: To capture tax/accounting benefits associated with traditional reinsurance
Investors: To isolate the risk of their investment from the general business and insolvency risks
of the insurer, thus creating an investment that is a pure play in catastrophic risk
Bonds are fully collateralized, with collateral in trust, insulating investors from credit risk
The issuer of the securitization can realize lower financing costs through segregation
The transaction also is more transparent than a debt issue by the insurer, because the funds
are held in trust and are released according to carefully defined criteria
The bonds are attractive to investors because catastrophic events have low correlations with returns
from securities markets and hence are valuable for diversification purposes

106

 Types of triggers
CAT bonds/other ILS structured to pay off on three types of triggering variables:
1. Indemnity triggers: Payouts based on size of the sponsoring insurers actual losses
2. Index triggers: Payouts based on an index not directly tied to sponsors losses
3. Hybrid triggers, which blend more than one trigger in a single bond
There are three broad types indices that can be used as CAT bond triggers:
(i) Industry loss indices, (ii) Modeled loss indices, and (iii) Parametric indices
With industry loss indices, the payoff on the bond is triggered when estimated industrywide losses from an event exceed a specified threshold
A modeled-loss index is calculated using a model provided by one of the major catastrophemodeling firms (actual events physical parameters are used in running simulations)
With a parametric trigger, the bond payoff is triggered by specified physical measures of
the catastrophic event such as the wind speed and location of a hurricane
Number of factors to consider in the choice of a trigger when designing a CAT bond: Trade-off
between moral hazard (transparency to investors) and basis risk
Indemnity triggers
Often favored by insurers/reinsurers because they minimize basis risk (risk that loss payout of
bond will be greater or less than sponsors actual losses)
However, indemnity triggers require investors to obtain info on risk exposure of sponsors
underwriting portfolio (difficult, especially for complex commercial risks)
Disadvantage to sponsor: Require disclosure of confidential info on sponsors portfolio
Contracts based on indemnity triggers may require more time than nonindemnity triggers to
reach final settlement because of the length of the loss adjustment process
Index triggers
Index triggers tend to be favored by investors because they minimize the problem of moral
hazard: I.e., they maximize the transparency of the transaction
Moral hazard occurs if issuer fails to settle CAT losses carefully/appropriately
Insurer might also excessively expand its premium writings in geographical areas covered
Although CAT bonds almost always contain copayment provisions to control moral hazard,
moral hazard remains a residual concern for some investors
Indices measurable more quickly after event Sponsor receives quicker payment
Disadvantage of index triggers: Expose sponsor to higher degree of basis risk
Degree of basis risk varies depending upon several factors. Parametric triggers tend to have
the lowest exposure to moral hazard but may have the highest exposure to basis risk
However, even with parametric trigger, basis risk can be reduced substantially by appropriately
defining location where event severity is measured. Similarly, industry loss indices based on
narrowly defined geographical areas have less basis risk than those based on wider areas
Modeled-loss indices may become the favored mechanism for obtaining benefits of index trigger
without significant basis risk. However, modeled-loss indices subject to model risk, diminishing over time as modeling firms refine their models
Sidecars
Sidecars are special purpose vehicles formed by insurance and reinsurance companies to provide
additional capacity to write reinsurance, usually for property catastrophes and marine risks, and
typically serve to accept retrocessions exclusively from a single reinsurer
Sidecars are typically off-balance sheet, formed to write specific types of reinsurance such as
property-catastrophe quota share or excess of loss, and generally have limited lifetimes
Reinsurers receive override commissions for premiums ceded to sidecars
Sidecars capitalized by private investors (hedge funds), but insurers/reinsurers also participate
Sidecars receive premiums for reinsurance underwritten and pay claims under contracts terms
Sidecars also enable sponsoring reinsurer to move some risks off balance sheet, improving leverage
Sidecars can also be formed quickly and with minimal documentation/administrative costs

107

 Catastrophic Equity Puts (Cat-E-Puts)

Unlike CAT bonds, Cat-E-Puts are not asset-backed securities but options
In return for a premium paid to the writer of the option, the insurer obtains the option to issue
preferred stock at a preagreed price on the occurrence of a contingent event
Enables insurer to raise equity capital at good price after a CAT, when stock price depressed
Cat-E-Puts have lower transactions costs than CAT bonds because no need to set up an SPR
However, they are not collateralized Expose insurer to counterparty performance risk
In addition, issuing the preferred stock can dilute the value of the firms existing shares
Thus, although Cat-E-Puts have been issued, they are not nearly as important as CAT bonds
Catastrophe risk swaps
Catastrophe risk swaps are not prefunded but rely on agreement between two counterparties
Catastrophe swaps executed between two firms with exposure to different types of CAT risk
In some instances, a reinsurer may serve as an intermediary between the swap partners, but in
most instances CAT swaps are done directly between two (re)insurers
Swaps are facilitated by the Catastrophic Risk Exchange (CATEX), a web-based exchange where
insurers and reinsurers can arrange reinsurance contracts and swap transactions
Swap agreement
Event(s) that trigger payment under swap are carefully defined in the agreement
The swap can be designed such that the two sides of the risk achieve parity, i.e., such that the
expected losses under the two sides of the swap are equivalent
Swap defines amount to be paid if event occurs. Some contracts have sliding scale payoffs
Swaps can be annual or can span several years
Swaps also can be executed that fund multiple risks simultaneously
Swaps may be attractive substitutes for reinsurance/CAT bonds/other risk financing devices:
Advantage that the reinsurer simultaneously lays of some of its core risk and obtains a new
source of diversification by exchanging uncorrelated risks with counterparty
Swaps may enable reinsurers to operate with less equity capital
Swaps also are characterized by low transactions costs and reduce current expenses because
no money changes hands until the occurrence of a triggering event
The potential disadvantages of swaps are that modeling the risks to achieve parity can be
challenging and is not necessarily completely accurate
Swaps also may create more exposure to basis risk than some other types of contracts and also
create exposure to counter-party nonperformance risk
Industry Loss Warranties (ILW)
Possible impediment to growth of CAT securitization market has to do with whether securities are
treated as reinsurance by regulators, and given favorable regulatory accounting treatment:
Properly structured indemnity CAT securities (those that pay off based on the losses of the
issuing insurer) will be treated as reinsurance
Sponsors and their bankers have found various ways to finesse potential regulatory problems:
E.g., even if the SPV is an offshore vehicle, the trust holding the assets can be onshore,
mitigating regulatory concerns regarding credit risk of offshore entities
Dual-trigger contracts, aka ILW, also overcome regulatory objections to nonindemnity bonds:
ILWs: Dual-trigger reinsurance contracts with retention trigger based on incurred losses of
insurer buying the contract and also a warranty trigger based on an industry-wide loss index
Both triggers have to be hit in order for the buyer of the contract to receive a payoff
ILWs cover events from specified catastrophe perils in a defined geographical region
The term of the contact is typically 1 year
ILWs may have binary triggers, where full amount pays off once the two triggers are satisfied
or pro rata triggers where payoff depends upon how much loss exceeds warranty
The principal advantages of ILWs are that they are treated as reinsurance for regulatory purposes,
and that they can be used to plug gaps in reinsurance programs

108

 The ILW market is roughly of the same order of magnitude as the CAT bond market
Capital market participants provide majority of risk capital in ILW/CAT bonds markets
ILWs can be packaged and securitized, broadening the investor base
The risk-linked securities market
The CAT bond market: Size and bond characteristics
Although the CAT bond market seemed to get off to a slow start in late 1990s, the market has
matured and now has become a steady source of capacity for both primary insurers/reinsurers:
Market growing steadily. New records for market issuance volume in 2005, 2006, and 2007
CAT bonds make sound economic sense as a mechanism for funding mega-catastrophes
$100B < 0.5% of US securities markets value Easily absorbed with securitized transactions
Securities markets more efficient in reducing info asymmetries/facilitating price discovery
Recently, event-linked bonds have also been issued to cover third-party commercial liability, automobile quota share, and indemnity-based trade credit reinsurance
The amount of risk capital outstanding in CAT bond markets has also grown steadily
Risk-capital outstanding represents the face value of all bonds still in effect in each year
Characteristics of CAT bonds continue to evolve. Overall trend is toward higher standardization
Between 2000 and 2006, index or hybrid bonds accounted for 80% of total issue volume
Leading type of index by volume is the parametric index (34% of total issuance)
Market has converged on shorter-term issues, with mostly 3-yr bonds
Maturities greater than 1 year favored because they provide a steady source of risk capital that
is insulated from year-to-year swings in reinsurance prices and because they permit issuers to
amortize costs of issuance over a longer period, reducing per period transactions costs
Bonds > 5 years not favored: Market participants want to reprice risk periodically to reflect
new info on frequency/severity of CATs and to recognize changes in u/w risk profile of sponsor
2000-06: Insurers account for 47.9% of bonds by issue volume, reinsurers for 47.5%
In 2006, the first government issued disaster-relief bond placement was executed to provide funds
to the government of Mexico to defray costs of disaster recovery
Such bonds illustrate how securitization can be used by governments to prefund disaster relief
programs, rather than waiting for disaster relief from donor countries ex post
Obtaining a financial rating is a critical step in issuing a CAT bond
Buyers use ratings to compare yields on CAT bonds with other corporate securities
Almost all bonds are issued with financial ratings
The vast majority of CAT bonds issued in 2005 and 2006 have been below investment grade
(ratings below BBB). In 2007, there has been a resurgence in investment-grade bonds, although
the majority of CAT bonds are below investment grade in 2007 as well
Although lower than investment grade ratings are generally bad news for insurers/reinsurers
and other corporate bond issuers, they are not necessarily adverse in CAT bond market
Because CAT bonds are fully collateralized, CAT bond ratings tend to be determined by
the probability that the bond principal will be hit by a triggering event
Bond ratings indicate layer of CAT-risk coverage being provided by the bonds
The modeling firms analysis drives the price more than the actual rating
There is broad market interest in CAT bonds among institutional investors
1999: Insurers/reinsurers among leading investors in the bonds, 55% of the market
By 2007, insurers/reinsurers = 7% of demand External capital attracted to market
Dedicated CAT funds = 55% of market in 2007, money managers/hedge funds = 36%
The declining spreads and increasingly broad market interest suggest that bonds are attractive
to investors and are playing an increasingly important role vs. conventional reinsurance
In addition to CAT bonds, significant amount of new capital raised through sidecars in 05 and 06
Indication that sidecars were competing with CAT bonds for risk capital of interested investors
in 2005, leading to rising prices and tightening capacity in the CAT bond market

109

 The first publicly acknowledged total loss of principal for a CAT bond took place in 2005
Short-term impact of wipeout: Increase investor wariness of indemnity-based transactions
Indemnity transactions rebounded in 07 due to surge of primary insurer CAT bond issues
Longer-term impact of the KAMP Re wipeout on CAT bond market is favorable
Smooth settlement of KAMP Re bond established important precedent, showing that CAT
bonds function well, with minimal confusion and controversy between sponsor/investors
CAT Bond prices
CAT bonds are priced at spreads over LIBOR, meaning that investors receive floating interest plus
a spread or premium over the floating rate
In the past, CAT bonds have been somewhat notorious for having high spreads
However, there are now significant indications that the spreads are not as high as they might
seem relative to the cost of reinsurance, such that CAT bonds are more competitive
Because CAT bonds are not publicly traded, difficult to obtain data on CAT bond yields
However, there is an active (nonpublic) secondary market that provides guidance on yields
Prior to Katrina, there was a more or less steady decline in yields and a slight increase in the
expected loss, implying a general decline in the cost of financing through CAT bonds
The ratio of the premium to expected loss was about six in early 2001
However, ratio of premium to expected loss steadyly declined to 2.1 for 2001-05
Not surprisingly, yields and spreads increased following Katrina as market tightened and investors had opportunities to place capital in other CAT-risk vehicles such as sidecars
The CAT bond market was able to withstand the post-Katrina competition for capital without
returning to the high relative spreads of earlier periods
This is the expected result in a market where there is growing investor interest and expertise
as well as growing volume, which adds to market liquidity
Comparison of CAT bond and catastrophe reinsurance pricing is difficult because of the general
lack of systematic data on reinsurance prices
The rate on line (ROL) is defined as the reinsurance premium divided by the policy limit
The loss on line (LOL) is the expected loss on the contract divided by the policy limit
Ratio ROL-to-LOL is analogous to ratio of yield/expected loss on CAT bonds
Like the CAT bond yield to ELR, the ratios of ROL to expected LOL are significantly higher
in 2006 than in 2005, reflecting the effects of Hurricanes Katrina, Rita, and Wilma
ROL-to-LOL ratios are significantly larger for national insurers than for regional insurers
Finally, the ratios are lower for contracts with higher expected LOL, because policies with low
expected LOL are covering the more risky upper tails of the loss distribution
CAT bonds on average tend to have expected losses of between 1-3% of principal, and thus are
most comparable to catastrophe reinsurance contracts with relatively low LOLs
CAT bonds are in the ballpark in terms of pricing for national companies
CAT bonds do not appear to be expensive relative to catastrophe reinsurance
Moreover, investment banks have reduced transactions costs/time to market as they have
gained experience with ILS Bonds more attractive to insurers/reinsurers
CAT bond prices look less attractive relative to reinsurance for regional companies
However, because regional firms have not been active in the CAT bond market, it is not
clear what the bond premia would be for these firms
Another comparative indication of trends in CAT bond spreads is provided by a comparison of the
Mexican CAT bonds with previously issued earthquake bonds
The spreads on the Mexican bonds are very low in comparison to the prior bonds
Two phenomena (not precisely separated) influence spreads:
1. The Mexican bonds are more recent and CAT bond spreads have been declining
2. The Mexican bonds are valuable to CAT bond investors for diversification purposes because
they cover a previously unsecuritized area of the world and permit investors to diversify
their current large proportionate exposure to US hurricane risk

110

 Relevant to compare CAT bond yields relative to yields on comparably rated corporate bonds
BB CAT bond yields comparable to yields on BB corporate bond for 2001-05 (until Katrina)
At the peak, yields on CAT bonds were 2-3% higher than the yields on BB corporates
Nevertheless, considering reinsurance prices in 06 and uncertainty created by Katrina/recent
CATs, the CAT bond market has weathered the storms very well
Regulatory, Accounting, Tax (RAT), and rating issues
Regulatory and accounting issues do not pose a material impediment to the growth of the market at
the present time, and the statistics on market size and growth clearly seem to bear this out
Regulatory issues
Some commentators have argued that CAT bonds have mostly been issued off-shore for regulatory
reasons and that lack of onshore issuance represents a barrier to market developments
Encouraging onshore issuance might reduce transactions costs/facilitate market growth
However, offshore jurisdictions provide low issuance costs/high levels of expertise in issuing ILS
Transactions costs for onshore CAT bonds generally higher than for offshore issues
Off-shore jurisdictions perform very effectively/efficiently in issuing/settling ILS
Nonindemnity CAT bonds currently face uncertain prospects with respect to regulatory treatment
Regulators concerned about basis risk and use of securitized risk instruments to speculate
Denied reinsurance accounting for nonindemnity CAT bonds impedes market development
However, other industry experts indicate that regulatory treatment does not presently pose a
significant obstacle to market development
Market participants have found a variety of structuring mechanisms to blunt regulatory concerns about alternative risk financing with respect to nonindemnity CAT bonds
E.g., contracts can be structured to pay off on narrowly defined geographical indices or combinations of indices that are highly correlated with the insurers losses
Concerns about speculative investing can be addressed through dual-trigger contracts that pay
off on an index but where the insurer cannot collect more than its ultimate net loss
In the area of risk-linked securities, it would be helpful if regulators were to codify the rules
and regulations relating to the SAP treatment of various types of risk-linked securities and avoid
imposing any unnecessary regulatory impediments in the future
Tax issues
Offshore CAT bonds do not create taxation problems for sponsors
No income/corporate/withholding taxes in offshore jurisdictions apply to CAT bonds
The bonds SPRs are also not taxable for US federal income tax purposes
Main tax issue for US investors: Treatment of bond premia under US tax law
Tax Code/IRS do not address tax treatment of income received from CAT bonds
Bonds are presently being treated as passive foreign investment companies (PFICs) Income
from CAT bonds included in taxable income as dividends rather than interest
US sponsors have been deducting premium payments on offshore bonds for income tax purposes, i.e., bond interest is currently treated similarly to reinsurance premiums
Dissemination of information on bonds
Although the ultimate objective should be the development of a public market for CAT bonds,
privately placed bonds are likely to continue to play an important role
Market development impeded to the extent that info on existing bonds is not available
Under current securities regulations, bond prospectuses for privately placed bonds can be
distributed only to investors falling under the definition of accredited investors
These rules have the unintended consequence of inhibiting research on CAT bonds
The SEC rules should be changed to allow sponsors to distribute bond prospectuses to researchers
who are not necessarily accredited investors
E.g., post prospectuses on repository maintained by appropriate governmental entity

111

 Issues to be explored
Insurance regulators in key jurisdictions such as the US, the EU, and Japan could mandate catastrophe loss reporting for events above a given industry threshold such as $1 billion
This would solve an important current problem, i.e., the lack of a PCS-equivalent index for
the EU and Japan, and would enhance the market by providing more info on US losses
Until loss turns into recoverable, credit quality of counterparty is ignored by regulators
In RBC, the charges for reinsurance are not graded by reinsurer credit quality
Explicitly incorporating reinsurance credit quality into regulatory capital calculations and related regulatory credit evaluations has the potential to provide an important boost to the ILS
market as well as improving insurance solvency regulation in more general terms
In several key CAT-prone states, regulators are most reluctant to allow price increases at precisely
the times when insurer loss expectations/reinsurance prices are increasing most rapidly
The best solution to this problem would be to deregulate prices at the state level so that
primary insurers would not be caught in this price-cost bind
Short of deregulating prices, regulators could help ease the problem by giving primary insurers
credit for locking in multi-year pricing and capacity by issuing ILS
The application of the Employee Retirement Income Security Act (ERISA) to CAT bond collateral
trusts is complex and would benefit from some thorough research
Conclusions
The CAT bond market is thriving and seems to have reached critical mass
The market achieved record bond issuance in 2005, 2006, and 2007
Bond premia have declined significantly since 2001
The bonds now are priced competitively with catastrophe reinsurance
The bonds now account for a significant share of the property-catastrophe reinsurance market
CAT bonds have an important role to play for high coverage layers and in retrocession market
Regulatory and accounting issues such as the regulatory accounting treatment of non-indemnity CAT
bonds and the issuance of most bonds offshore, which have been cited as impediments to the development
of the market, do not presently seem to pose serious problems
However, there are a number of issues/reforms that should be explored:
Fostering better reporting of CAT losses to facilitate development of better index products
Solvency regulation should be adapted to recognize the credit quality of reinsurance receivables
and give recognition to the full collateralization provided by CAT bonds
Primary insurance prices should be deregulated, and primary insurers should receive credit for
entering into contracts providing multi-year pricing/capacity through ILS/conventional reinsurance
Applicability of ERISA to CAT bond collateral trusts and the US GAAP and SAP treatment of
triggers employed in ILW and similar contracts
Issuers of CAT bonds should be required to make available bond prospectuses to researchers who
could provide valuable analysis of CAT risk financing
The future looks bright for the ILS market
CAT bonds, swaps, sidecars, ILW, and other innovative products will play an increasingly important
role in providing risk financing for CAT events
Event-linked bonds are also being used increasingly by primary insurers for lower layers of coverage
and non-CAT coverages such as automobile and commercial liability insurance
It remains to be seen whether CAT futures and options will play an important role in catastrophe
risk management in the years to come
Basis risk and counterparty credit risk, as well as the need to educate insurance industry participants, are the primary impediments to the success of these contracts

112

Additional Notes

113

Additional Notes

114

Additional Notes

115

CAS Exam 8 Notes - Parts C, D, & E

Futures, Forwards and Swaps
Options
Asset-Liability Management

Part II
Table of Contents
C Futures, Forwards and Swaps

Hull - Ch. 2: Mechanics of futures markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Hull - Ch. 3: Hedging strategies using futures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11

Hull - Ch. 5: Determination of forward and futures prices . . . . . . . . . . . . . . . . . . . . . . .

17

Hull - Ch. 7: Swaps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

25

D Options

35

Hull - Ch. 8: Mechanics of options markets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

37

Hull - Ch. 9: Properties of stock options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Hull - Ch. 10: Trading strategies involving options . . . . . . . . . . . . . . . . . . . . . . . . . . .

45
51

Hull - Ch. 11: Binomial trees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

57

Fabozzi: Valuation of bonds with embedded options . . . . . . . . . . . . . . . . . . . . . . . . . .

63

Hull - Ch. 12: Wiener processes and Itos lemma . . . . . . . . . . . . . . . . . . . . . . . . . . . .

69

Hull - Ch. 13: The Black-Scholes-Merton model . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

73

Black: Using the holes in Black-Scholes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

81

Hull - Ch. 15: Options on stock indices and currencies . . . . . . . . . . . . . . . . . . . . . . . . .

87

Hull - Ch. 16: Futures options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

93

E Asset-Liability Management

99

BKM - Ch. 16: Managing bond portfolios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

Feldblum: Asset liability matching for P&C insurers . . . . . . . . . . . . . . . . . . . . . . . . . . 111
Noris: Asset/Liability management strategies for P&C companies . . . . . . . . . . . . . . . . . . . 119
Panning: Managing interest rate risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Additional Notes

129

Futures, Forwards and Swaps

Hull - Ch. 2: Mechanics of futures markets

Background
Both futures/forward contracts are agreements to buy/sell an asset at a future time for a certain price
Futures are traded on an organized exchange, contract terms are standardized by the exchange
By contrast, forward contracts are private agreements between financial institutions/clients
Chicago Board of Trade (CBOT), Chicago Mercantile Exchange (CME): Largest US futures exchanges
Futures trade
On March 5, a trader in New York might call a broker with instructions to buy 5,000 bushels of
corn for delivery in July of the same year. The broker would immediately issue instructions to a
trader to buy (i.e., take a long position in) one July corn contract
Another trader in Kansas might instruct a broker to sell 5,000 bushels of corn for July delivery.
This broker would then issue instructions to sell (i.e., take a short position in) one corn contract
A price would be determined and the deal would be done
Under traditional open outcry system, floor traders representing each party would physically meet
to determine price. With electronic trading, a computer matches trades/monitor prices
Definitions
The trader in New York who agreed to buy has a long futures position in one contract
The trader in Kansas who agreed to sell has a short futures position in one contract
The price agreed to is the current futures price for July corn
This price, like any other price, is determined by the laws of supply and demand
Closing out positions
The vast majority of futures contracts do not lead to delivery:
Most traders choose to close out their positions prior to delivery period specified in contract
Closing out a position means entering into the opposite trade to the original one
Investors gain/loss determined by change in futures price from 3/5 to when contract closes out
Nevertheless, it is the possibility of final delivery that ties the futures price to the spot price
Specification of a futures contract
As stipulated by the exchange, the agreement between the two parties must specify the asset, the
contract size, where delivery will be made, and when delivery will be made
Sometimes alternatives are specified for grade of asset delivered/delivery locations
As a general rule, it is the party with the short position (the party that has agreed to sell the asset)
that chooses what will happen when alternatives are specified by the exchange
When party with short position ready, it files a notice of intention to deliver with the exchange
Notice indicates selections made wrt. grade of asset delivered and delivery location
The asset
When asset is a commodity, the exchange stipulates acceptable grade(s) of the commodity
For some commodities, range of grades can be delivered, but price received depends it
Often, substitutions are allowed with price being adjusted in a way established by the exchange
The financial assets in futures contracts are generally well defined and unambiguous
The contract size
This is an important decision for the exchange
If the contract size is too large, many investors will be unable to use the exchange
If the contract size is too small, trading expensive (cost associated with each contract traded)
Delivery arrangements
The place where delivery will be made must be specified by the exchange
This is particularly important for commodities that involve significant transportation costs
When alternative delivery locations are specified, the price received by the party with the short
position is sometimes adjusted according to the location chosen by that party
Price higher for delivery locations relatively far from main sources of commodity

 Delivery months
A futures contract is referred to by its delivery month
The exchange must specify the precise period during the month when delivery can be made
At any time, contracts trade for closest delivery month/few subsequent delivery months
The exchange specifies when trading in a particular months contract will begin
The exchange also specifies the last day on which trading can take place for a given contract
Trading generally ceases a few days before the last day on which delivery can be made
Price quotes
The exchange defines how prices will be quoted
E.g., crude oil prices on NYMEX are quoted in dollars and cents
Treasury bond/Treasury note futures on CBOT quoted in dollars and thirty-seconds
Price limits and position limits
For most contracts, daily price movement limits are specified by the exchange
A limit move is a move in either direction equal to the daily price limit
Normally, trading ceases for the day once the contract is limit up or limit down
Purpose: Prevent large price movements from speculative excesses
Position limits are the maximum number of contracts that a speculator may hold
Purpose: Prevent speculators from exercising undue influence on the market
Convergence of futures price to spot price
As delivery period approaches, the futures price converges to spot price of underlying asset
When delivery period is reached, futures price ' spot price
Suppose that futures price > spot price during delivery period Arbitrage opportunity:
(i) Sell (i.e., short) a futures contract, (ii) Buy the asset, and (iii) Make delivery
These steps lead to a profit equal to the amount by which the futures price > spot price
As traders exploit this arbitrage opportunity, the futures price will fall
Suppose next that the futures price is below the spot price during the delivery period:
Companies interested in acquiring the asset find attractive to enter into long futures contract and
wait for delivery As they do, futures prices tend to rise
The result is that the futures price is very close to the spot price during the delivery period
Daily settlements and margins
Key role of the exchange: Organize trading so that contract defaults are avoided Margins
The operation of margins
Consider two December gold futures contracts (size = 100 ounces, current futures price is $600/ounce)
The broker will require the investor to deposit funds in a margin account
Initial margin: Amount that must be deposited at time contract is entered into
At the end of each trading day, the margin account is adjusted to reflect investors gain/loss
This practice is referred to as marking to market the account
A trade is first marked to market at the close of the day on which it takes place
It is then marked to market at the close of trading on each subsequent day
Marking to market is not merely an arrangement between broker and client
When there is a decrease in the futures price so that the margin account of an investor with a
long position is reduced by $600, the investors broker has to pay the exchange $600 and the
exchange passes the money on to the broker of an investor with a short position
Investor entitled to withdraw any balance in margin account in excess of initial margin
Maintenance margin (< initial margin): Ensures that margin account never < 0
If balance in margin account falls below maintenance margin, investor receives a margin call
and is expected to top up margin account to initial margin level the next day
The extra funds deposited are known as a variation margin
Further details
Many brokers allow an investor to earn interest on the balance in a margin account
Balance in the account 6= true cost, provided that interest rate is competitive
4

 To satisfy the initial margin requirements (but not subsequent margin calls), an investor can
sometimes deposit securities with the broker
Treasury bills are usually accepted in lieu of cash at about 90% of their face value
Shares are also sometimes accepted in lieu of cash, at 50% of market value
Effect of marking to market: Futures contract settled daily rather than at end of its life
At the end of each day, the investors gain (loss) is added to (subtracted from) the margin
account, bringing the value of the contract back to zero
A futures contract is in effect closed out and rewritten at a new price each day
Minimum levels for initial and maintenance margins are set by the exchange
Individual brokers may require greater margins from clients than specified by exchange
Margin levels are determined by the variability of the price of the underlying asset
The higher this variability, the higher the margin levels
The maintenance margin is usually about 75% of the initial margin
Day trades/spread transactions give rise to lower margin requirements than hedge transactions
Day trade: Trader tells broker an intent to close out position the same day
Spread transaction: The trader simultaneously buys a contract on an asset for one maturity
month and sells a contract on the same asset for another maturity month
Margin requirements are the same on short futures positions as they are on long futures positions
The spot market does not have this symmetry
Long position in spot market involves buying asset for immediate delivery No problems
Short position involves selling an asset not owned More complex (may not be possible)
The clearinghouse and clearing margins
Clearinghouse
Acts as an intermediary in futures transactions
Guarantees the performance of the parties to each transaction
The clearinghouse has a number of members, who must post funds with the exchange
Brokers who are not members themselves must channel their business through a member
The main task of the clearinghouse is to keep track of all the transactions that take place
during a day, so that it can calculate the net position of each of its members
The broker is required to maintain a margin account with a clearinghouse member and the clearinghouse member is required to maintain a margin account with the clearinghouse
The latter is known as a clearing margin
The margin accounts for clearinghouse members are adjusted for gains and losses at the end
of each trading day in the same way as are the margin accounts of investors
However, for clearinghouse members, there is an original margin, but no maintenance margin
Brokers not clearinghouse members maintain a margin account with a clearinghouse member
In determining clearing margins, the exchange clearinghouse calculates the number of contracts
outstanding on either a gross or a net basis
Gross basis: The number of contracts equals the sum of the long and short positions
Net basis: These are offset against each other
Most exchanges currently use net margining
Credit risk
Purpose of margining system: Eliminate risk that a trader who makes a profit is not paid
Overall the system has been very successful
Collateralization in OTC markets
Credit risk has traditionally been a feature of the over-the-counter markets
To reduce credit risk, OTC markets now imitates the margining system with collateralization
Participants in OTC market can enter into collateralization agreement, valuing contract each day
Each day, if value to A increases, B required to pay A (cash amount = increase), and conversely
Interest is paid on outstanding cash balances
Collateralization significantly reduces the credit risk in OTC contracts

Newspaper quotes
Many newspapers carry futures prices
The prices refer to the trading that took place on the previous day
For most commodities, contracts trade with much longer maturities than those shown
However, trading volume tends to decrease as contract maturity increases
The asset underlying the futures contract, the exchange that the contract is traded on, the contract
size, how the price is quoted, and the maturity month of the contract are usually shown
Quotes usually show opening price, highest/lowest prices achieved in trading during the day
Settlement price
The settlement price is the price used for calculating daily gains/losses and margin requirements
Price at which contract traded just before the bell signaling end of trading for the day
Quotes also usually give the change in the settlement price from the previous day
Open interest
This is the total number of contracts outstanding
The open interest is the number of long positions the number of short positions
When volume of trading in a day > open interest at end of day, it indicates many day trades
Patterns of futures prices
Normal market: Settlement prices increase with the maturity of the contract
Inverted market: Futures price is a decreasing function of maturity
Futures prices can also show a mixture of normal and inverted markets
Delivery
Period during which delivery can be made is defined by the exchange and varies by contract
The decision on when to deliver is made by the party with the short position (investor A)
When A ready to deliver, As broker issues a notice of intention to deliver to exchange clearinghouse
This notice states how many contracts will be delivered and, in the case of commodities, also
specifies where delivery will be made and what grade will be delivered
The exchange then chooses a party with a long position to accept delivery
Exchange usually passes notice of intention to deliver to party with oldest outstanding long position
Parties with long positions must accept delivery notices
However, if notices are transferable, long investors have a short period of time (half an hour) to
find another party with a long position that is prepared to accept the notice from them
For commodities, delivery means accepting a warehouse receipt in return for immediate payment
The party taking delivery is then responsible for all warehousing costs
In the case of financial futures, delivery is usually made by wire transfer
For all contracts, the price paid is usually the most recent settlement price
If specified by the exchange, this price is adjusted for grade, location of delivery, and so on
The whole delivery procedure from notice to delivery takes 2-3 days
There are three critical days for a contract:
First notice day: 1st day on which a notice of intention to make delivery can be submitted
Last notice day: The last such day
Last trading day: Generally a few days before the last notice day
To avoid risk of having to take delivery, long positions should be closed out prior to 1st notice day
Cash settlement
Some financial futures (on stock indices) are settled in cash (impossible to deliver underlying asset)
When contract settled in cash, all outstanding contracts are declared closed on a predefined day
Final settlement price = spot price of underlying asset at opening/close of that trading day
Types of traders and types of orders
There are two main types of traders executing trades: Commission brokers and locals
Commission brokers are following instructions of their clients and charge a commission
Locals are trading on their own account

 Individuals taking positions can be categorized as hedgers, speculators, or arbitrageurs

Speculators can be classified as scalpers, day traders, or position traders
Scalpers are watching for very short-term trends and attempt to profit from small changes in the
contract price. They usually hold their positions for only a few minutes
Day traders hold positions < one trading day, unwilling to take risk of adverse overnight news
Position traders hold positions for much longer periods (profit from major market movements)
Orders
Market order: Request that trade be carried out immediately at best price available in market
Limit order: Specifies a particular price: Order executed only at this price or better
Stop order or stop-loss order: Also specifies a particular price: Order executed at the best
available price once a bid or offer is made at that particular price or a less-favorable price
Purpose of stop order: To close out a position if unfavorable price movements take place
It limits the loss that can be incurred
Stop-limit order: A combination of a stop order and a limit order
Suppose a stop-limit order to buy is issued with stop price of $40 and limit price of $41
As soon as there is a bid or offer at $40, the stop-limit becomes a limit order at $41
If stop price = limit price, the order is called a stop-and-limit order
Market-if-touched (MIT) order: Executed at the best available price after a trade occurs at a
specified price or at a price more favorable than the specified price
In effect, an MIT becomes a market order once the specified price has been hit
An MIT is also known as a board order
Consider investor who has a long position and is issuing instructions to close out contract:
Stop order: Designed to limit the loss that can occur if unfavorable price movements
MIT order: Designed to ensure that profits are taken if favorable price movements
Discretionary order or market-not-held order: Traded as a market order except that execution may be delayed at the brokers discretion in an attempt to get a better price
Some orders specify time conditions:
Unless otherwise stated, an order is a day order and expires at the end of trading day
Time-of-day order: Specifies a particular period of time during the day
Open order or a good-till-canceled order: In effect until executed/until end of trading
Fill-or-kill order: Must be executed immediately on receipt or not at all
Regulation
US futures markets regulated federally by the Commodity Futures Trading Commission (CFTC)
This body is responsible for licensing futures exchanges and approving contracts
All new contracts and changes to existing contracts must be approved by the CFTC
To be approved, contract must have some useful economic purpose (serve hedgers and speculators)
The CFTC looks after the public interest:
Responsible for ensuring that prices are communicated to the public and that futures traders report
their outstanding positions if they are above certain levels
Licenses all individuals who offer their services to the public in futures trading
Deals with complaints brought by public and ensures that disciplinary action is taken
Has authority to force exchanges to discipline members in violation of exchange rules
With formation of National Futures Association (NFA) in 1982, some responsibilities of CFTC shifted:
The NFA is an organization of individuals who participate in the futures industry
Its objective is to prevent fraud/ensure that market operates in best interests of general public
Authorized to monitor trading and take disciplinary action when appropriate
NFA set up an efficient system for arbitrating disputes between individuals/members
From time to time, other bodies, such as the SEC, the Federal Reserve Board, and the US Treasury
Department, have claimed jurisdictional rights over some aspects of futures trading
Concerned with effects of futures trading on spot markets for securities (stocks, T-bills, T-bonds)
SEC has effective veto over approval of new stock/bond index futures contracts
7

 Trading irregularities
One type of trading irregularity occurs when an investor group tries to corner the market
The investor group takes a huge long futures position and also tries to exercise some control
over the supply of the underlying commodity
Regulators usually deal with this type of abuse of the market by increasing margin requirements
or imposing stricter position limits or prohibiting trades that increase a speculators open
position or requiring market participants to close out their positions
Other types of trading irregularity can involve the traders on the floor of the exchange:
Overcharging customers, not paying customers the full proceeds of sales, and traders using
their knowledge of customer orders to trade first for themselves (front running)
Accounting and tax
Accounting
Accounting standards require changes in the market value of a futures contract to be recognized
when they occur unless the contract qualifies as a hedge
If the contract does qualify as a hedge, gains or losses are generally recognized in the same period
in which the gains or losses from the item being hedged are recognized (hedge accounting)
Consider a company with a December year end. In September 2007 it buys a March 2008 corn
futures contract and closes out the position at the end of February 2008
Suppose: Futures prices are 250c per bushel when contract entered into, 270c at end of 2007,
and 280c when contract closed out. Contract is for 5,000 bushels
If the contract does not qualify as a hedge, the gains for accounting purposes are:
5, 000 (2.70 2.50) = 1, 000 in 2007 and 5, 000 (2.80 2.70) = 500 in 2008
If contract qualifies for hedge accounting, entire gain ($1,500) realized in 08 for accounting
FAS 133
Issued in June 1998, applies to all types of derivatives (futures, forwards, swaps, options)
FAS 133 requires all derivatives to be included on the balance sheet at fair market value
It increases disclosure requirements
It also gives companies far less latitude than previously in using hedge accounting:
For hedge accounting to be used, the hedging instrument must be highly effective in offsetting exposures and an assessment of this effectiveness is required every three months
A similar standard IAS 39 has been issued by the IASB
Tax
Key issues: (i) Nature of a taxable gain/loss and, (ii) Timing of recognition of the gain/loss
Gains/losses are either classified as capital gains/losses or as part of ordinary income
Corporate taxpayer
Capital gains taxed at same rate as ordinary income, and restricted ability to deduct losses
Capital losses are deductible only to the extent of capital gains
Corporation may carry back capital loss for 3 years/forward for up to 5 years
Non-corporate taxpayer
Short-term capital gains taxed at same rate as ordinary income, but long-term (capital asset
held for 1+ year) capital gains are subject to a maximum capital gains tax rate of 15%
Capital losses are deductible to the extent of capital gains plus ordinary income up to $3,000
and can be carried forward indefinitely
Generally, positions in futures are treated as if closed out on last day of the tax year
For noncorporate taxpayer, this gives rise to capital gains/losses treated as if 60% long term
and 40% short term without regard to holding period (60/40 rule)
A noncorporate taxpayer may elect to carry back for three years any net losses from the 60/40
rule to offset any gains recognized under the rule in the previous three years
Hedging transactions are exempt from this rule. The definition of a hedge transaction for tax
purposes is different from that for accounting purposes

 The tax regulations define a hedging transaction as a transaction entered into in the normal course
of business primarily for one of the following reasons:
1. Reduce the risk of price changes or currency fluctuations with respect to property that is held
or to be held by the taxpayer for the purposes of producing ordinary income
2. Reduce risk of price/interest rate/currency fluctuations wrt. taxpayer borrowings
Hedging transaction must be identified before end of the day on which taxpayer enters into it
The asset being hedged must be identified within 35 days
Gains or losses from hedging transactions are treated as ordinary income
The timing of recognition of gains/losses from hedging transactions generally matches the
timing of recognition of income/expense associated with transaction being hedged
Forward vs. futures contracts
Comparison of forward and futures contracts
Forward
Private contract between two parties
Not standardized
Usually one specified delivery date
Settled at end of contract
Delivery/final cash settlement usually takes place
Some credit risk

Futures
Traded on an exchange
Standardized contract
Range of delivery dates
Settled daily
Contract usually closed out prior to maturity
Virtually no credit risk

Profits from forward and futures contracts

Under forward contract, whole gain/loss is realized at the end of the life of the contract
Under futures contract, gain/loss is realized day by day because of the daily settlement
Foreign exchange quotes
Both forward and futures contracts trade actively on foreign currencies
However, differences in the way exchange rates are quoted in the two markets
Futures prices where one currency is USD are quoted as USD per unit of foreign currency
Forward prices are always quoted in the same way as spot prices:
For British pound, euro, Australian dollar, and New Zealand dollar, forward quotes show USD
per unit of foreign currency (directly comparable with futures quotes)
For other currencies, forward quotes show number of units of foreign currency per USD

10

Hull - Ch. 3: Hedging strategies using futures

Basic principles
Short hedges
A short hedge is a hedge that involves a short position in futures contracts
Appropriate when hedger already owns asset and expects to sell it in the future
A short hedge can also be used when an asset is not owned right now but will be owned
Example
Assume: On May 15, an oil producer negotiates a contract to sell 1 million barrels of crude oil
The price that will apply in the contract is the market price on August 15
The oil producer will gain $10,000 for each 1 cent increase in the price of oil over the next 3
months and lose $10,000 for each 1 cent decrease in the price during this period
On May 15, spot price = $60 per barrel and futures price for August delivery = $59
Because each futures contract on NYMEX is for the delivery of 1,000 barrels, the company
can hedge its exposure by shorting 1,000 futures contracts
If producer closes out position on August 15, he will lock in a price $59 per barrel
Suppose that the spot price on August 15 proves to be $55 per barrel
The company realizes $55 million for the oil under its sales contract
The futures price on August 15 should be very close to the spot price of $55 on that date
Company gains $59 $55 = $4 per barrel from the short futures position
Total from both futures position/sales contract $59 per barrel, or $59M total
Suppose that the price of oil on August 15 proves to be $65 per barrel
Company realizes $65 for oil and loses approximately $65 $59 = $6 per barrel on the futures
Again, the total amount realized is approximately $59 million
Long hedges
Hedges that involve taking a long position in a futures contract
Appropriate when company will purchase an asset in the future and wants to lock in a price now
Long hedges can be used to manage an existing short position:
Consider an investor who has shorted a certain stock
Part of the risk faced by the investor is related to the performance of the whole stock market
The investor can neutralize this risk with a long position in index futures contracts
Making or taking delivery can be costly and inconvenient
Delivery is not usually made even when hedger keeps futures contract until delivery month
Hedgers with long positions usually avoid any possibility of having to take delivery by closing out
their positions before the delivery period
In practice, daily settlement does have a small effect on the performance of a hedge
If the exposure is such that the company gains when the price of the asset increases and loses when the
price of the asset decreases, a short hedge is appropriate
If the exposure is the other way round (i.e., the company gains when the price of the asset decreases
and loses when the price of the asset increases), a long hedge is appropriate
Arguments for and against hedging
Arguments for hedging
Most companies have no particular skills/expertise in predicting interest rates, exchange rates, and
commodity prices Makes sense to hedge such risks to avoid unpleasant surprises
The companies can then focus on their main activities and expertise
Hedging and shareholders
Argument against hedging: Shareholders can do the hedging themselves
Argument assumes that shareholders have as much info about companys risks as management
The argument also ignores commissions and other transactions costs Hedging less expensive
when carried out by company rather than individual shareholders
One thing that shareholders can do far more easily than a corporation is diversify risk
11

 If companies are acting in the best interests of well-diversified shareholders, it can be argued
that hedging is unnecessary in many situations (at least in theory)
Hedging and competitors
If hedging not the norm in an industry, it may not make sense for one company to be different
Competitive pressures within industry may be such that prices of goods/services produced fluctuate
to reflect raw material costs, interest rates, exchange rates, and so on
A company that does not hedge can expect its profit margins to be roughly constant
However, a company that does hedge can expect its profit margins to fluctuate!
This emphasizes the importance of looking at the big picture when hedging
All the implications of price changes on a companys profitability should be taken into account
in the design of a hedging strategy to protect against the price changes
Hedging can lead to a worse outcome
A hedge can result in decrease/increase in a companys profits relative to position without hedging
Treasurer may fear criticism if company gains from underlying asset/loses on the hedge
Ideally, hedging strategies are set by a companys board of directors and are clearly communicated
to both the companys management and the shareholders No misunderstandings
Basis risk
In practice, hedging is often not quite straightforward. Reasons:
1. Asset hedged may not be exactly same as asset underlying futures contract
2. The hedger may be uncertain as to the exact date when the asset will be bought or sold
3. The hedge may require the futures contract to be closed out before its delivery month
These problems give rise to what is termed basis risk
The basis
The basis (hedging situation): Difference between spot price of an asset and its futures price:
Basis = Spot price of asset to be hedged Futures price of contract used
If the asset to be hedged and the asset underlying the futures contract are the same, the basis
should be zero at the expiration of the futures contract
As time passes, spot price/futures price do not change by same amount Basis changes
An increase in the basis is referred to as a strengthening of the basis
A decrease in the basis is referred to as a weakening of the basis
Notations
S1 :
S2 :

Spot price at time t1

Spot price at time t2

F1 :
F2 :

Futures price at time t1

Futures price at time t2

b1 :
b2 :

Basis at time t1
Basis at time t2

Assume that a hedge is put in place at time t1 and closed out at time t2 . E.g., consider the case
where the spot and futures prices at the time the hedge is initiated are $2.50 and $2.20, respectively,
and that at the time the hedge is closed out they are $2.00 and $1.90, respectively
This means that S1 = 2.50, F1 = 2.20, S2 = 2.00, and F2 = 1.90
From the definition of the basis, we have: b1 = S1 F1 = 0.30 and b2 = S2 F2 = 0.10
Consider hedger who knows that asset will be sold at t2 and takes short futures position at t1
The price realized for the asset is S2 and the profit on the futures position is F1 F2
The effective price that is obtained for the asset with hedging is therefore: S2 +F1 F2 = F1 +b2
If b2 were also known at this time, a perfect hedge would result. The hedging risk is the
uncertainty associated with b2 and is known as basis risk
Consider a short hedge (for a long hedge, the reverse holds):
If the basis strengthens (i.e., increases) unexpectedly, the hedgers position improves
If the basis weakens (i.e., decreases) unexpectedly, the hedgers position worsens
When asset of hedgers exposure is different from asset underlying the futures contract:
This increases the basis risk
S2 = price of asset underlying futures contract at t2 , S2 = price of asset being hedged at t2
By hedging, company ensures that price paid (or received) for the asset is:
S2 + F1 F2 F1 + (S2 F2 ) + (S2 S2 )
12

 The terms (S2 F2 ) and (S2 S2 ) represent the two components of the basis:
(S2 F2 ): Basis if asset being hedged = asset underlying futures contract
(S2 S2 ): Basis arising from the difference between the two assets
Choice of contract
Key factor affecting basis risk: Choice of futures contract used for hedging. Two components:
1. The choice of the asset underlying the futures contract
2. The choice of the delivery month
Necessary to carry out a careful analysis to determine which of the available futures contracts has
futures prices that are most closely correlated with the price of the asset being hedged
The choice of the delivery month is likely to be influenced by several factors:
We assumed: If hedge expiration = delivery month, contract with that delivery month chosen
In fact, a contract with a later delivery month is usually chosen in these circumstances
Reason: Futures prices can be quite erratic during the delivery month
Moreover, a long hedger runs the risk of having to take delivery of the physical asset if the
contract is held during the delivery month (can be expensive and inconvenient)
Long hedgers prefer to close out futures contract and buy asset from their usual suppliers
Basis risk increases as time difference between hedge expiration/delivery month increases
Choose a delivery month as close as possible to, but later than, the expiration of the hedge
Cross hedging
Cross hedging occurs when asset underlying futures contract 6= asset whose price is being hedged
Hedge ratio: Ratio of [size of the position taken in futures contracts] [size of the exposure]
When asset underlying futures = asset being hedged, natural to use hedge ratio = 1.0
When cross hedging is used, setting the hedge ratio equal to 1.0 is not always optimal
Choose a hedge ratio value that minimizes the variance of the value of the hedged position
Calculating the minimum variance hedge ratio
Notations
S:
F :
S :
F :
:
h :

Change in spot price S during a period of time equal to the life of the hedge
Change in futures price F during a period of time equal to the life of the hedge
Standard deviation of S
Standard deviation of F
Coefficient of correlation between S and F
Hedge ratio that minimizes the variance of the hedgers position

Minimum variance hedge ratio h

h =

S
F

(1)

If = 1 and F = S , hedge ratio h = 1.0: Futures price mirrors spot price perfectly
If = 1 and F = 2S , hedge ratio h = 0.5: Futures price always changes by 2 spot price
The optimal hedge ratio h is the slope of the best-fit line when S is regressed against F
Intuitively, we require h to correspond to the ratio of changes in S to changes in F
Hedge effectiveness: Defined as proportion of variance eliminated by hedging
This is the R2 from the regression of S against F and equals 2 , or:
2 = h2

F2
S2

Parameters , F , and S in Eq. (1) usually estimated from historical data on S and F
The implicit assumption is that the future will in some sense be like the past
Equal nonoverlapping time intervals are chosen to observe values of S and F
Ideally, length of each interval = length of time interval for which hedge is in effect

13

 Optimal number of contracts

Definitions
QA : Size of position being hedged (units)
QF :
N :

Size of one futures contract (units)

Optimal number of futures contracts for hedging

The futures contracts should be on h QA units of the asset

The number of futures contracts required is therefore given by:
N =

h QA
QF

(2)

Denoting the i-th observations on F and S by xi and

sP
sP
P
x2i
yi2
( xi )2
F =

and S =

n 1 n(n 1)
n1

yi , respectively, we have:
P
( yi )2
n(n 1)

The estimate of is:

P P
xi yi xi yi
= q P
P
P
P
[n x2i ( xi )2 ][n yi2 ( yi )2 ]
n

Tailing the hedge

When futures are used for hedging, a small adjustment, known as tailing the hedge, can be made
to allow for the impact of daily settlement. In practice this means that Eq. (2) becomes:
N =

h V A
VF

(3)

VA = dollar value of hedged position, VF = dollar value of 1 futures contract (futures price QF )
Effect of tailing the hedge: Multiply hedge ratio in Eq. (2) by ratio of spot price to futures price
Ideally, futures position should be adjusted as VA /VF change (not feasible in practice)
Stock index futures
Stock index futures are used to hedge or manage exposures to equity prices
A stock index tracks changes in the value of a hypothetical portfolio of stocks
Weight of a stock in the portfolio equals proportion of the portfolio invested in the stock
% increase in stock index over t = % increase in value of hypothetical portfolio
Dividends usually not included Index tracks only capital gain/loss from investing in portfolio
If hypothetical portfolio fixed, weights assigned to individual stocks do not remain fixed
Some indices are constructed from portfolio consisting of one of each of many stocks
Weights assigned to stocks are then their market prices (with adjustments for stock splits)
Other indices constructed so that weights market cap (stock price shares outstanding)
Underlying portfolio then adjusted to reflect stock splits, dividends, new issues
Stock indices
Dow Jones Industrial Average
Based on a portfolio consisting of 30 blue-chip stocks in the US
The weights given to the stocks are proportional to their prices
CBOT: 2 futures contracts on DJIA: One on $10 index, other on $5 index
Standard & Poors 500 (S&P 500)
500 stocks: 400 industrials, 40 utilities, 20 transportation, and 40 financials
Weights of stocks at any given time market capitalizations
CME: 2 futures on S&P 500: $250 index, $50 index
Nasdaq 100
100 stocks using National Association of Securities Dealers Automatic Quotations Service
CME: 2 contracts, $100 index, and $20 index
14

 Russell 1000
Index of the prices of the 1000 largest capitalization stocks in the US
US Dollar Index
Trade-weighted index of the values of six foreign currencies (the euro, yen, pound, Canadian
dollar, Swedish krona, and Swiss franc)
Futures contracts on stock indices are settled in cash, not by delivery of the underlying asset
Hedging an equity portfolio
Stock index futures can be used to hedge a well-diversified equity portfolio. Define:
P:
F:

Current value of the portfolio

Current value of one futures contract (futures price contract size)

If the portfolio mirrors the index, the optimal hedge ratio h equals 1.0 and Eq. (3) shows that
the number of futures contracts that should be shorted is:
N = P/F

(4)

When portfolio 6= index mirror, use from CAPM to determine the hedge ratio
Beta is the slope of the best-fit line obtained when excess return on the portfolio over the
risk-free rate is regressed against the excess return of the market over the risk-free rate
In general, h = , so that Eq. (3) gives:
N =

P
F

(5)

Formula assumes that maturity of futures contract is close to maturity of the hedge
Example
Suppose that a futures contract with 4 months to maturity is used to hedge the value of a
portfolio over the next 3 months in the following situation:
Value of S&P 500 index = 1,000
S&P 500 futures price = 1,010
Value of portfolio = $5,050,000

Risk-free interest rate = 4% per annum

Dividend yield on index = 1% per annum
Beta of portfolio = 1.5

One futures contract is for delivery of $250 times the index

F = 250 1, 010 = 252, 500 and from Eq. (5), the number of futures contracts that should
be shorted to hedge the portfolio is: 1.5 5, 050, 000/252, 500 = 30
Suppose the index turns out to be 900 in 3 months and the futures price is 902
The gain from the short futures position is: 30 (1010 902) 250 = $810, 000
The loss on the index is 10%. The index pays a dividend of 1% per annum, or 0.25% per 3
months. When dividends are taken into account, an investor in the index would therefore earn
-9.75% in the 3-month period. The risk-free interest rate is approximately 1% per 3 months
Because the portfolio has a of 1.5, the CAPM gives that the expected return (%) on the
portfolio during the 3 months is: 1.0 + [1.5 (9.75 1.0)] = 15.125
The expected value of the portfolio (inclusive of dividends) at the end of the 3 months is
therefore $5, 050, 000 (1 0.15125) = $4, 286, 187 The expected value of the hedgers
position, including gain on the hedge, is: $4, 286, 187 + $810, 000 = $5, 096, 187
Reasons for hedging an equity portfolio
The hedge results in the investors position growing at the risk-free rate
It is natural to ask why the hedger should go to the trouble of using futures contracts:
Hedger could simply sell portfolio/invest proceeds in risk-free instruments (T-bills)
Hedging can be justified if hedger feels that stocks in portfolio have been chosen well
A hedge using index futures removes the risk arising from market moves and leaves the hedger
exposed only to the performance of the portfolio relative to the market
Another reason for hedging may be that the hedger is planning to hold a portfolio for a long period
of time and requires short-term protection in an uncertain market situation
Alternative strategy of selling portfolio/buying back later has very high transaction costs
15

 Changing the beta of a portfolio

Sometimes futures contracts are used to change the beta of a portfolio
Continuing with our earlier example, to reduce the beta of the portfolio from 1.5 to 0.75, the
number of contracts shorted should be 15 rather than 30
To increase the beta of the portfolio to 2.0, a long position in 10 contracts should be taken
To change the portfolio from > , a short position in ( )P/F contracts is required
When < , a long position in ( )P/F contracts is required
Exposure to the price of an individual stock
Hedging an exposure to the price of an individual stock using index futures contracts is similar to
hedging a well-diversified stock portfolio
Number of index futures contracts that hedger should short is P/F , where is the stocks
, P = total value of shares owned, and F = current value of one index futures contract
However, the performance of the hedge is considerably worse
Hedge provides protection only against the risk arising from market movements, and this risk is a
relatively small proportion of the total risk in price movements of individual stocks
Hedge appropriate if stock will outperform the market but the market performance is unsure
It can also be used by an investment bank that has underwritten a new issue of the stock and
wants protection against moves in the market as a whole
It cannot be emphasized often enough that the purpose of a hedge is to reduce risk:
Hedge makes unfavorable outcomes less unfavorable but makes favorable outcomes less favorable
Rolling the hedge forward
Sometimes the expiration date of the hedge is later than the delivery dates of all the futures contracts
that can be used The hedger must roll the hedge forward by closing out one futures contract and
taking the same position in a futures contract with a later delivery date
Result: Creation of a long-dated futures contract by trading a series of short-dated contracts
A company can use the following strategy:
Time
Time

Time
Time

t1 :
t2 :
tn :
T:

Short futures contract 1

Close out futures contract 1, Short futures contract 2

Close out futures contract n 1, Short futures contract n

Close out futures contract n

Example
Date
Oct. 2007 futures price
Mar. 2008 futures price
July 2008 futures price
Spot price

Apr. 2007
68.20

Sept. 2007
67.40
67.00

69.00

Feb. 2008
66.50
66.30

June 2008

65.90
66.00

Results
Oct. 2007 contract shorted at $68.20 and closed out at $67.40 for profit of $0.80 per barrel
Mar. 2008 contract shorted at $67.00 and closed out at $66.50 for profit of $0.50 per barrel
July 2008 contract shorted at $66.30 and closed out at $65.90 for profit of $0.40 per barrel
The final spot price is $66. The dollar gain per barrel of oil from the short futures contracts is:
(68.20 67.40) + (67.00 66.50) + (66.30 65.90) = 1.70
The daily settlement of futures contracts can cause a mismatch between the timing of the cash flows on
hedge and the timing of the cash flows from the position being hedged
When the hedge is rolled forward so that it lasts a long time, this can lead to serious problems

16

Hull - Ch. 5: Determination of forward and futures prices

Investment assets vs. consumption assets
Investment asset
Asset held for investment purposes by many investors (stocks, bonds, gold, silver)
Investment assets do not have to be held exclusively for investment
Consumption asset
Asset held primarily for consumption, not usually held for investment (commodities, copper, oil)
We can use arbitrage arguments to determine the forward and futures prices of an investment asset from
its spot price and other observable market variables
We cannot do this for consumption assets
Short selling
Shorting involves selling asset not owned. Possible for some, but not all, investment assets
An investor with a short position must pay to the broker any income, such as dividends or interest, that
would normally be received on the securities that have been shorted
Broker transfers this income to client account from whom securities have been borrowed
The investor is required to maintain a margin account with the broker
Margin account consists of cash/marketable securities deposited by investor with broker to guarantee that investor will not walk away from short position if share price increases
An initial margin is required and if there are adverse movements (i.e., increases) in the price of the
asset that is being shorted, additional margin may be required
Margin account does not represent a cost to investor: Interest is paid on balance in margin accounts
and marketable securities (T-bills) can be used to meet margin requirements
Purchase of shares
April: Purchase 500 shares for $120
May: Receive dividend
July: Sell 500 shares for $100 per share
Short sale of shares
April: Borrow 500 shares and sell them for $120
May: Pay dividend
July: Buy 500 shares for $100 per share
Replace borrowed shares to close short position

-$60,000
+$500
+$50,000
Net profit = -$9,500
+$60,000
-$500
-$50,000
Net profit = +$9,500

The SEC abolished the uptick rule in the US on July 6, 2007: This rule required the most recent
movement in the price of a stock to be an increase for the shorting of a stock to be permitted
Assumptions and notations
Assume that the following are all true for some market participants:
1. Market participants are subject to no transaction costs when they trade
2. Market participants are subject to the same tax rate on all net trading profits
3. Market participants can borrow money at same risk-free rate as they can lend
4. Market participants take advantage of arbitrage opportunities as they occur
Notations
T:
S0 :
F0 :
r:

Time until delivery date in a forward or futures contract (in years)

Price of the asset underlying the forward or futures contract today
Forward or futures price today
Zero-coupon risk-free rate (continuous compounding), for maturity of T years

The risk-free rate r is the rate at which money is borrowed/lent when there is no credit risk
LIBOR rates rather than Treasury rates are the relevant risk-free rates
Forward price for an investment asset
The easiest forward contract to value is one written on an investment asset that provides the holder
with no income (e.g., non-dividend-paying stocks and zero-coupon bonds)
17

 Arbitrage opportunities when forward price is out of line with spot price for asset providing no income
(Asset price = $40, interest rate = 5%, maturity of forward contract = 3 months)
Forward Price = $43
Action now :
Borrow $40 at 5% for 3 months
Buy one unit of asset
Enter forward to sell asset in 3m for $43
Action in 3 months:
Sell asset for $43
Use $40.50 to repay loan with interest
Profit realized = $2.50

Forward Price = $39

Action now :
Short 1 unit of asset to realize $40
Invest $40 at 5% for 3 months
Enter forward to buy asset in 3m for $39
Action in 3 months:
Buy asset for $39
Close short position/receive $40.50 from inv.
Profit realized = $1.50

We deduce that for there to be no arbitrage, the forward price must be exactly $40.50
A generalization
Consider a forward contract on an investment asset with price S0 that provides no income
F0 = S0 erT

(1)

If F0 > S0 erT , arbitrageurs can buy the asset and short forward contracts on the asset
If F0 < S0 erT they can short the asset and enter into long forward contracts on it
A long forward contract and a spot purchase both lead to the asset being owned at time T
The forward-price is higher than the spot price because of the cost of financing the spot
purchase of the asset during the life of the forward contract
What if short sales are not possible?
Short sales are not possible for all investment assets
Requirement for Eq. (1) to hold: There are many people who hold asset purely for investment
If the forward price is too low, they will find it attractive to: (i) Sell the asset, (ii) Invest the
proceeds at interest rate r for time T and, (iii) Take a long position in a forward contract
If the forward price is too high, investors will find it attractive to: (i) Borrow S0 dollars at an
interest rate r for T years, (ii) Buy the asset and, (iii) Take a short position in a forward contract
Known income
Consider a forward contract on an investment asset that will provide a perfectly predictable cash income
to the holder (e.g., stocks paying known dividends and coupon-bearing bonds)
Arbitrage opportunities when 9-month forward price is out of line with spot price (Asset price $900,
income of $40 occurs at 4 months, 4-month and 9-month rates are 3% and 4% per annum):
Forward price = $910
Action now :
Borrow $900: $39.60 for 4m, $860.40 for 9m
Buy 1 unit of asset
Enter forward to sell asset in 9m for $910
Action in 4 months:
Receive $40 of income on asset
Use $40 to repay 1st loan with interest
Action in 9 months:
Sell asset for $910
Use $886.60 to repay 2nd loan with interest
Profit realized = $23.40

Forward price = $870

Action now :
Short 1 unit of asset to realize $900
Invest $39.60 for 4m, $860.40 for 9m
Enter forward to buy asset in 9m for $870
Action in 4 months:
Receive $40 from 4m investment
Pay income of $40 on asset
Action in 9 months:
Receive $886.60 from 9m investment
Buy asset for $870, close out short position
Profit realized = $16.60

If there are no arbitrage opportunities then the forward price must be $886.60
A generalization
When investment asset provides income with PV of I during life of forward contract:
F0 = (S0 I)erT

(2)

Eq. (2) applies to any investment asset that provides a known cash income
F0 > (S0 I)erT : Arbitrageur locks in profit by buying asset/shorting forward
18

 F0 < (S0 I)erT : Arbitrageur locks in profit by shorting asset/taking long position in forward
If short sales not possible, investors who own asset will sell asset and enter into long forwards
Other derivation of Eq. (2)
Consider: Buy one unit of asset and enter into short forward to sell it for F0 at T
This costs S0 and is certain to lead to a cash inflow of F0 at T and an income with PV of I
The initial outflow is S0 . The present value of the inflows is I + F0 erT
Hence, S0 = I + F0 erT , or equivalently F0 = (S0 I)erT
Known yield
Asset underlying a forward contract provides a known yield rather than a known cash income
The income is known when expressed as a % of the assets price at the time the income is paid
Defining q as the average yield per annum on an asset during the life of a forward contract:
F0 = S0 e(rq)T

(3)

Example
Consider 6m forward on asset providing income = 2% of asset price once during 6m period
Risk-free rate of interest (continuous compounding) is 10% per annum. Asset price = $25
Yield = 4% per annum (semiannual compounding) 3.96% per annum (continuous compounding)
q = 0.0396 and F0 = 25e(0.100.0396)0.5 = $25.77
Valuing forward contracts
The value of a forward contract at the time it is first entered into is zero
At a later stage, it may prove to have a positive or negative value
Important for banks/other financial institutions to value contract each day (marking to market)
Notations
K is the delivery price for a contract that was negotiated some time ago
The delivery date is T years from today, and r is the T -year risk-free interest rate
F0 is the forward price that would be applicable if we negotiated the contract today
f is the value of the forward contract today
At the beginning of the life of the forward contract, the delivery price K is set equal to the forward
price F0 , and the value of the contract f is 0
As time passes, K stays the same (by definition of the contract), but the forward price changes
and the value of the contract becomes either positive or negative
General result, applicable to all long forward contracts (both on investment/consumption assets), is:
f = (F0 K)erT

(4)

Similarly, the value of a short forward contract with delivery price K is:
f = (K F0 )erT
Eq. (4) shows that we can value a long forward contract on an asset by making the assumption that
the price of the asset at the maturity of the forward contract equals the forward price F0
Similarly, value short forward by assuming that current forward price of asset is realized
Using Eqs. (4) and (1), the value of a forward on an investment asset providing no income is:
f = S0 KerT

(5)

Using Eqs. (4) and (2), the value of long forward on investment asset providing income with PV I is:
f = S0 I KerT

(6)

Using Eqs. (4) and (3), the value of long forward on investment asset providing yield at rate q:
f = S0 eqT KerT

(7)
19

 When a futures price changes, the gain or loss on a futures contract is calculated as the change in the
futures price multiplied by the size of the position
This gain is realized almost immediately because of the way futures contracts are settled daily
Eq. (4) shows that, when a forward price changes, the gain or loss is the PV of the change in the
forward price multiplied by the size of the position
Are forward prices and futures prices equal?
An arbitrage argument shows that when risk-free interest rate constant and same for all maturities, the
forward price is the same as the futures price for a contract with same delivery date
Argument can be extended to cover situations where interest rate is a known function of time
When interest rates vary unpredictably, forward and futures prices are in theory no longer the same
Consider situation where price S of underlying asset is positively correlated with interest rates
When S %, investor with long futures makes immediate gain because of daily settlement
The positive correlation indicates that it is likely that interest rates have also increased
The gain will therefore tend to be invested at a higher than average rate of interest
Similarly, when S decreases, the investor will incur an immediate loss
This loss will tend to be financed at a lower than average rate of interest
Forward contracts are not affected in this way by interest rate movements
Long futures contract slightly more attractive than similar long forward contract and when
S strongly positively correlated with interest rates, futures prices & forward prices
When S is strongly negatively correlated with interest rates, a similar argument shows that forward
prices will tend to be slightly higher than futures prices
The theoretical differences between forward and futures prices for contracts that last only a few months
are in most circumstances sufficiently small to be ignored
In practice, many factors not reflected in models may cause forward/futures prices to be different.
These include: (i) Taxes, (ii) Transactions costs, and (iii) Treatment of margins
However, for most purposes it is reasonable to assume that forward/futures prices are the same
Futures prices of stock indices
A stock index can usually be regarded as the price of an investment asset that pays dividends
The investment asset is the portfolio of stocks underlying the index, and the dividends paid by the
investment asset are the dividends that would be received by the holder of this portfolio
Usually assumed that dividends provide a known yield rather than a known cash income
If q is the dividend yield rate, Eq. (3) gives the futures price F0 as:
F0 = S0 e(rq)T

(8)

In practice, dividend yield on portfolio underlying an index varies week by week through the year
q should represent the average annualized dividend yield during the life of the contract
Dividends used for estimating q: Ex-dividend date during the life of the futures contract
Index arbitrage
If F0 > S0 e(rq)T , profits can be made by buying the stocks underlying the index at the spot price
(i.e., for immediate delivery) and shorting futures contracts
If F0 < S0 e(rq)T , profits can be made by doing the reverse that is, shorting or selling the stocks
underlying the index and taking a long position in futures contracts
These strategies are known as index arbitrage
F0 < S0 e(rq)T : Index arbitrage done by pension fund that owns an indexed stock portfolio
F0 > S0 e(rq)T : Done by corporation holding short-term money market investments
For indices involving many stocks, index arbitrage is sometimes accomplished by trading a relatively
small representative sample of stocks whose movements closely mirror those of the index
Often index arbitrage is implemented through program trading (using a computer)
Most of the time, the activities of arbitrageurs ensure that Eq. (8) holds, but occasionally arbitrage
is impossible and the futures price does get out of line with the spot price
20

Forward and futures contracts on currencies

Consider forward and futures foreign currency contracts from the perspective of a US investor
The underlying asset is one unit of the foreign currency
S0 is the current spot price in dollars of one unit of the foreign currency and F0 is the forward
or futures price in dollars of one unit of the foreign currency
This does not necessarily correspond to the way spot and forward exchange rates are quoted
Holder of foreign currency can earn interest at risk-free rate prevailing in foreign country
Define rf as the value of the foreign risk-free interest rate when money is invested for time T
The variable r is the US dollar risk-free rate when money is invested for this period of time
The relationship between F0 and S0 is:
F0 = S0 e(rrf )T ,

well-known interest rate parity relationship

(9)

1,000 units of foreign currency. Two ways it can be converted to dollars at time T :
1. Invest it for T years at rf and enter into forward to sell proceeds for dollars at time T
This generates 1, 000erf T F0 dollars
2. Exchange foreign currency for dollars in spot market and invest proceeds for T years at rate r
This generates 1, 000S0 erT dollars
In the absence of arbitrage opportunities, the two strategies must give the same result Eq. (9)
Example
Suppose that the 2-year interest rates in Australia and the US are 5% and 7%, and the spot
exchange rate between the AUD and the USD is 0.6200 USD per AUD
From Eq. (9), the 2-year forward exchange rate should be 0.62e(0.070.05)2 = 0.6453
Suppose first that 2-year forward exchange rate is less than this, 0.6300. An arbitrageur can:
1. Borrow 1,000 AUD at 5% per annum for 2 years, convert to 620 USD and invest USD at 7%
2. Enter into a forward contract to buy 1,105.17 AUD for 1, 105.17 0.63 = 696.26 USD
To repay principal/interest on the 1,000 AUD borrowed, we need 1, 000e0.052 = 1, 105.17 AUD
The 620 USD that are invested at 7% grow to 620e0.072 = 713.17 USD in 2 years
Of this, 696.26 USD are used to purchase 1,105.17 AUD under terms of forward contract
The strategy therefore gives rise to a riskless profit of 713.17 696.26 = 16.91 USD
Suppose next that 2-year forward rate is 0.6600 (> 0.6453 given by Eq. (9)). Arbitrageur can:
1. Borrow 1,000 USD at 7% for 2 yrs, convert to 1, 000/0.62 = 1, 612.90 AUD, invest AUD at 5%
2. Enter into a forward contract to sell 1,782.53 AUD for 1, 782.53 0.66 = 1, 176.47 USD
The 1,612.90 AUD that are invested at 5% grow to 1, 612.90e0.052 = 1, 782.53 AUD in 2 years
The forward contract has the effect of converting this to 1,176.47 USD
The amount needed to payoff the USD borrowings is 1, 000e0.072 = 1, 150.27 USD
The strategy therefore gives rise to a riskless profit of 1, 176.47 1, 150.27 = 26.20 USD
For currencies such that r > rf , futures prices for these currencies increase with maturity
For currencies such that r < rf , futures prices of these currencies decrease with maturity
A foreign currency as an asset providing a known yield
Eq. (9) is identical to Eq. (3) with q replaced by rf
A foreign currency can be regarded as an investment asset paying a known yield, where the
yield is the risk-free rate of interest in the foreign currency
The value of interest paid in a foreign currency depends on the value of the foreign currency
Suppose that the interest rate on British pounds is 5% per annum
To US investor, provides an income equal to 5% of the value of the per annum
In other words it is an asset that provides a yield of 5% per annum

Futures on commodities
Income and storage costs
Background
Gold owners such as central banks charge interest in the form of what is known as the gold
lease rate when they lend gold. The same is true of silver
21

 Gold and silver can therefore provide income to the holder

Like other commodities they also have storage costs
Storage costs can be treated as negative income:
If U = PV of all storage costs, net of income, during life of forward, from Eq. (2):
F0 = (S0 + U )erT

(10)

If storage costs net of income incurred at any time are price of commodity negative yield
Then, from Eq. (3), and denoting u the storage costs per annum as a proportion of the spot
price net of any yield earned on the asset:
F0 = S0 e(r+u)T

(11)

Consumption commodities
Commodities that are consumption assets rather than investment assets usually provide no income,
but can be subject to significant storage costs
Suppose that, instead of Eq. (10), we have:
F0 > (S0 + U )erT

(12)

To take advantage of this opportunity, an arbitrageur can:

1. Borrow S0 + U at rf and use it to buy one unit of commodity and pay storage costs
2. Short a forward contract on one unit of the commodity
This strategy leads to a profit of F0 (S0 + U )erT at time T
There is no problem in implementing the strategy for any commodity
However, as arbitrageurs do so, S0 will increase and F0 decrease until Eq. (12) no longer true
Eq. (12) cannot hold for any significant length of time
Suppose next that:
F0 < (S0 + U )erT

(13)

When commodity is an investment asset and Eq. (13) holds, it is profitable to:
1. Sell commodity, save storage costs, and invest proceeds at risk-free interest rate
2. Take a long position in a forward contract
Result: Riskless profit at maturity of (S0 + U )erT F0 Eq. (13) cannot hold for long
Because neither Eqs. (12) nor (13) can hold for long, we must have F0 = (S0 + U )erT
Argument cannot be used for commodity that is consumption asset rather than investment asset
Individuals/companies who own consumption commodity usually plan to use it
Reluctant to sell commodity in spot market and buy forward/futures (not consumable)
There is nothing to stop Eq. (13) from holding
All we can assert for a consumption commodity is:
F0 (S0 + U )erT

(14)

If storage costs are expressed as a proportion u of the spot price, the equivalent result is:
F0 S0 e(r+u)T

(15)

Convenience yields
Not necessarily equality in Eqs. (14)/(15) because users of consumption commodity may feel that
ownership of physical commodity provides benefits not obtained by holders of futures
Benefits from holding physical asset are sometimes referred to as the convenience yield
If dollar amount of storage costs has a PV of U , then the convenience yield y is:
F0 eyT = (S0 + U )erT
22

 If storage costs u per unit are a constant proportion of spot price, then y is defined so that:
F0 eyT = S0 e(r+u)T

F0 = S0 e(r+uy)T

(16)

For investment assets, the convenience yield must be zero: Otherwise, arbitrage opportunities
The convenience yield reflects markets expectations for future availability of the commodity
The greater the possibility that shortages will occur, the higher the convenience yield
If users of the commodity have high inventories, there is very little chance of shortages in the
near future and the convenience yield tends to be low
The cost of carry
Relationship between futures/spot prices can be summarized in terms of the cost of carry
Cost of carry: Storage cost + interest paid to finance the asset income earned on the asset
For non-dividend-paying stock, cost of carry = r (no storage costs and no income earned)
For a stock index, it is: r q, because income is earned at rate q on the asset
For a currency, it is: r rf
For commodity that provides income at rate q/requires storage costs at rate u, it is: r q + u
Denote the cost of carry as c:
For an investment asset, the futures price is:
F0 = S0 ecT

(17)

For a consumption asset, denoting y is the convenience yield, it is:

F0 = S0 e(cy)T

(18)

Delivery options
Futures often allows party with short position to choose to deliver at any time during a certain period
The choice introduces a complication into the determination of futures prices
Should maturity of futures be assumed to be beginning, middle, or end of delivery period?
If the futures price is an increasing function of the time to maturity, Eq. (18) c > y
Benefits from holding asset (including convenience yield, net of storage costs) are less than rf
Usually optimal in such a case for party with short position to deliver as early as possible, because
interest earned on the cash received outweighs benefits of holding asset
As a rule, futures prices in these circumstances should be calculated on the basis that delivery will
take place at the beginning of the delivery period
If futures prices are decreasing as time to maturity increases (c < y), the reverse is true
It is then usually optimal for the party with the short position to deliver as late as possible
Futures prices should, as a rule, be calculated on this assumption
Futures prices and expected future spot prices
Expected spot price
Markets avg. opinion about what the spot price of an asset will be at a certain future time
The futures price converges to the spot price at maturity
If the expected spot price is less than todays September futures price, the market must be expecting
the September futures price to decline Short traders gain, long traders lose
If expected spot price is greater than todays September futures price, reverse must be true
Traders with long positions gain while traders with short positions lose
Keynes and Hicks
They argued that, if hedgers tend to hold short positions and speculators tend to hold long positions,
the futures price of an asset will be below the expected spot price
Speculators require compensation for their risks and trade only if they expect to make money
Hedgers lose money on avg. but accept this because futures reduce their risks
If hedgers tend to hold long positions while speculators hold short positions, Keynes and Hicks
argued that the futures price will be above the expected spot price for a similar reason
23

 Risk and return

The relationship between futures prices and expected spot prices is based on the relationship
between risk and expected return in the economy
An investor generally requires a higher expected return than the risk-free interest rate for bearing
positive amounts of systematic risk
Also, an investor is prepared to accept a lower expected return than the risk-free interest rate when
the systematic risk in an investment is negative
The risk in a futures position
Consider speculator who takes long position in futures contract that lasts for T years hoping that
the asset spot price will be above futures price at end of life of futures contract
Suppose that the speculator puts the PV of the futures price into a risk-free investment while
simultaneously taking a long futures position
The proceeds of the risk-free investment are used to buy the asset on the delivery date. The asset
is then immediately sold for its market price
The cash flows to the speculator are as follows:
Today: F0 erT
End of futures contract: +ST
How do we value this investment?
Discount rate we should use for the expected CF at T equals an investors required return k
on the investment. The PV of this investment is: F0 erT + E[ST ]ekT
Assume that all investments in securities markets are priced so that they have zero NPV
F0 = E[ST ]e(rk)T

(19)

The returns investors require on an investment depend on its systematic risk. The investment
we have been considering is in essence an investment in the asset underlying the futures contract
If the returns from this asset are uncorrelated with the stock market, the correct discount rate
to use is the risk-free rate r, so we should set k = r. Eq. (19) then gives: F0 = E[ST ]
The futures price is an unbiased estimate of the expected future spot price when the return
from the underlying asset is uncorrelated with the stock market
If return from asset is positively correlated with stock market, k > r and F0 < E[ST ]
When the asset underlying the futures contract has positive systematic risk, we should expect
the futures price to understate the expected future spot price (e.g., a stock index)
If return from asset is negatively correlated with stock market, k < r and F0 > E[ST ]
When the asset underlying the futures contract has negative systematic risk, we should expect
the futures price to overstate the expected future spot price
Normal backwardation and contango
Normal backwardation: When the futures price is below the expected future spot price
Contango: When the futures price is above the expected future spot price
Summary for contract with maturity T , on investment asset with price S0 when risk-free rate = r
Asset
Provides no income:
Provides known w/ PV I:
Provides known yield q:

Futures price

Value of long forward (delivery price K)

S0 KerT
S0 I KerT
S0 eqT KerT

rT

S0 e
(S0 I)erT
S0 e(rq)T

24

Hull - Ch. 7: Swaps

Introduction
A swap is an agreement between two companies to exchange cash flows in the future
The agreement defines dates when CFs are to be paid and how they are to be calculated
The first swap contracts were negotiated in the early 1980s
Swaps now occupy a position of central importance in the OTC derivatives market
Whereas a forward contract is equivalent to the exchange of cash flows on just one future date, swaps
typically lead to cash flow exchanges on several future dates
Mechanics of interest rate swaps
The most common type of swap is a plain vanilla interest rate swap
Company agrees to pay CFs = interest at fixed rate on notional principal for number of years
In return, it receives interest at floating rate on same notional principal for same period of time
LIBOR
The floating rate in most interest rate swap agreements is the LIBOR
For each period, the rate of interest is set based on LIBOR rate at beginning of the period
Interest is paid at the end of the period
Illustration
Consider a hypothetical 3-year swap initiated on March 5, 2007, between Microsoft and Intel
Microsoft agrees to pay Intel an interest rate of 5% per annum on principal of $100 million,
and in return Intel agrees to pay Microsoft the 6-month LIBOR rate on same principal
Microsoft is the fixed-rate payer, Intel is the floating-rate payer
Assume payments exchanged every 6 months and 5% interest rate (semiannual compounding)
No uncertainty about 1st exchange (determined by LIBOR at time contract entered into)
In total, there are six exchanges of payment on the swap
The fixed payments are always $2.5 million
Floating-rate payments calculated with 6m LIBOR prevailing 6 months before payment date
Generally, one side remits the difference between the two payments to the other side
The principal itself is not exchanged
For this reason, it is termed the notional principal, or just the notional
If principal was exchanged at end of swap life, nature of deal would not be changed
With final exchange of principal, swap exchange fixed-rate/floating-rate bonds
Using the swap to transform a liability
For Microsoft, the swap could be used to transform a floating-rate loan into a fixed-rate loan
Suppose Microsoft has arranged to borrow $100 million at LIBOR plus 10 basis points
After Microsoft has entered into the swap, it has the following three sets of cash flows:
1. It pays LIBOR plus 0.1% to its outside lenders
2. It receives LIBOR under the terms of the swap
3. It pays 5% under the terms of the swap
These three sets of cash flows net out to a fixed interest rate payment of 5.1%
For Intel, the swap could transform a fixed-rate loan into a floating-rate loan
Suppose that Intel has a 3-year $100 million loan outstanding on which it pays 5.2%
After it has entered into the swap, it has the following three sets of cash flows:
1. It pays 5.2% to its outside lenders
2. It pays LIBOR under the terms of the swap
3. It receives 5% under the terms of the swap
These three sets of CFs net out to a floating interest rate payment of LIBOR plus 0.2%
Using the swap to transform an asset
Swaps can also be used to transform the nature of an asset
E.g., transform asset earning fixed rate of interest into asset earning a floating rate
Suppose Microsoft owns $100M in bonds providing 4.7% per annum over next 3 years
25

 After Microsoft has entered into the swap, it has the following three sets of cash flows:
1. It receives 4.7% on the bonds
2. It receives LIBOR under the terms of the swap
3. It pays 5% under the terms of the swap
These three CFs net out to floating interest rate inflow of LIBOR minus 30 bp
Swap can transform asset earning floating rate into asset earning fixed rate
Suppose Intel has an investment of $100M that yields LIBOR minus 20 bp
After it has entered into the swap, it has the following three sets of cash flows:
1. It receives LIBOR minus 20 basis points on its investment
2. It pays LIBOR under the terms of the swap
3. It receives 5% under the terms of the swap
These three sets of cash flows net out to a fixed interest rate inflow of 4.8%
Role of financial intermediary
Usually, two nonfinancial companies do not get in touch directly to arrange a swap
They each deal with a financial intermediary such as a bank or other financial institution
Plain vanilla fixed-for-floating swaps on US interest rates usually structured so that financial
institution earns 3-4 bp (0.03% or 0.04%) on a pair of offsetting transactions
(a)

(b)
4.985%

5%
5.2%

Intel

Microsoft

5.2%
LIBOR+0.1%

Intel

LIBOR

5.015%
Financial
Institution

LIBOR

Microsoft and Intel use the swap to transform a liability

Microsoft

LIBOR+0.1%

LIBOR

Interest rate swap when financial institution is involved

Figure 1: Interest rate swap with and without financial institution

The financial institution enters into two offsetting swap transactions with Microsoft and Intel
Financial institution has 2 separate contracts: One with Intel, the other with Microsoft
If one company defaults, financial institution still has to honor its agreement with the other
The 3-bp spread earned by the financial institution is partly to compensate it for the risk that
one of the two companies will default on the swap payments
Market makers
Many large financial institutions act as market makers for swaps
Ready to enter swap without having an offsetting swap with another counterparty
Market makers must carefully quantify and hedge the risks they are taking
Instruments to hedge swaps: Bonds, forward rate agreements, and interest rate futures
Swap rate: The average of the bid and offer fixed rates
Consider new swap where fixed rate = current swap rate: The value of this swap is zero
Swap characterized as difference between fixed-rate/floating-rate bonds. Define:
Bf ix :
Bf l :

Value of fixed-rate bond underlying the swap we are considering

Value of floating-rate bond underlying the swap we are considering

Since the swap is worth zero, it follows that: Bf ix = Bf l

Day count issues
In general, a LIBOR-based floating-rate CF on a swap payment date is calculated as LRn/360, where
L = principal, R = relevant LIBOR rate, and n = # days since last payment date
Fixed rate paid in swap transaction is also quoted with particular day count basis specified:
As a result, the fixed payments may not be exactly equal on each payment date
The fixed rate is usually quoted as actual/365 or 30/360
It is not therefore directly comparable with LIBOR because it applies to a full year
To compare: [6-month LIBOR 365/360] vs. [fixed rate 360/365]

26

Confirmations
Confirmation: Legal agreement underlying a swap, signed by representatives of the two parties
Drafting of confirmations facilitated by the International Swaps and Derivatives Association (ISDA)
This organization produced Master Agreements that consist of clauses defining in some detail the
terminology used in swap agreements, what happens in the event of default by either side, . . .
E.g., the confirmation specifies that the following business day convention is to be used and that the
US calendar determines which days are business days and which days are holidays
If payment date falls on a weekend/US holiday, payment is made on the next business day
The comparative-advantage argument
Explanation put forward to explain popularity of swaps concerns comparative advantages:
Some companies, it is argued, have a comparative advantage when borrowing in fixed-rate markets,
whereas other companies have a comparative advantage in floating-rate markets
To obtain a loan, company should go to the market where it has a comparative advantage
Company may borrow fixed when wants floating, or borrow floating when wants fixed
The swap is used to transform a fixed-rate loan into a floating-rate loan, and vice versa
AAACorp and BBBCorp
Key feature of rates offered to AAACorp/BBBCorp: Difference between the two fixed rates >
difference between the two floating rates (AAACorp has lower rates in both cases)
BBBCorp appears to have a comparative advantage in floating-rate markets, whereas AAACorp
appears to have a comparative advantage in fixed-rate markets: The extra amount that BBBCorp
pays over the amount paid by AAACorp is less in the floating-rate market
The total apparent gain from this type of interest rate swap arrangement is always a b, where:
a is the difference between the interest rates facing the two companies in fixed-rate markets
b is the difference between the interest rates facing the two companies in floating-rate markets
Criticism of the argument
Why spreads between rates offered to AAACorp/BBBCorp different in fixed/floating markets?
With swap market, we may expect these differences to have been arbitraged away
Spreads exist due to nature of contracts available to companies in fixed/floating markets
LIBOR rates in floating-rate markets are 6m rates: Lender can review rates every 6 months
If creditworthiness of Corp has declined, lender can increase spread over LIBOR
Providers of fixed-rate financing do not have option to change terms of loan
Probability of default by Corp with low credit rating (BBBCorp) is liable to increase faster
than probability of default by Corp with high credit rating (AAACorp)
Spread between 5-year rates is greater than spread between 6-month rates
Entering into swap, BBBCorp appears to obtain fixed-rate loan at 4.97%. Not really:
In practice, rate = 4.97% only if BBBCorp can continue to borrow at 0.6% over LIBOR
Swap locks in LIBOR 0.33% for AAACorp for the next 5 years, not just for next 6 months
This appears to be a good deal for AAACorp
The downside is that it is bearing the risk of a default by the financial institution
If it borrowed floating-rate funds in the usual way, it would not be bearing this risk
The nature of swap rates
A swap rate is the average of:
(a) Fixed rate that market maker ready to pay for receiving LIBOR (its bid rate)
(b) Fixed rate that it is prepared to receive in return for paying LIBOR (its offer rate)
Like LIBOR rates, swap rates are not risk-free lending rates. However, they are close to risk-free
A financial institution can earn the 5-year swap rate on a certain principal by:
1. Lend principal 6m to AA borrower, then relend it for successive 6m periods to other AA borrowers
2. Enter into a swap to exchange the LIBOR income for the 5-year swap rate
Hence, the 5-year swap rate is an interest rate with a credit risk corresponding to the situation where
10 consecutive 6-month LIBOR loans to AA companies are made
Note that 5-year swap rates are less than 5-year AA borrowing rates
27

 Much more attractive to lend money for successive 6-month periods to borrowers who are always
AA at the beginning of the periods than to lend it to one borrower for the whole 5 years
Determining LIBOR / Swap zero rates
Problem with LIBOR rates: Direct observations are possible only for maturities out to 12 months
One way of extending the LIBOR zero curve beyond 12 months is to use Eurodollar futures
Eurodollar futures produce LIBOR zero curve out to 2 yrs, up to 5 yrs
Traders then use swap rates to extend the LIBOR zero curve further
Resulting zero curve referred to as LIBOR zero curve, or swap zero curve
The value of a newly issued floating-rate bond that pays 6-month LIBOR is always equal to its principal
value (or par value) when the LIBOR/swap zero curve is used for discounting
Reason: The bond provides a rate of interest of LIBOR, and LIBOR is the discount rate
Interest on bond exactly matches discount rate The bond is fairly priced at par
For a newly issued swap where the fixed rate equals the swap rate, Bf ix = Bf l
Since Bf l = notional principal Bf ix = swaps notional principal
Swap rates define set of par yield bonds Use bootstrap method to extend LIBOR/swap zero curve
Example
Suppose 6m, 12m, and 18m LIBOR/swap zero rates are 4%, 4.5%, and 4.8% (continuous compounding) and 2-year swap rate is 5% (semiannual payments)
5% swap rate Bond with $100 principal and 5% per annum semiannual coupon sells for par
It follows that, if R is the 2-year zero rate, then:
2.5e0.040.5 + 2.5e0.0451.0 + 2.5e0.0481.5 + 102.5e2R = 100 R = 4.953%
Valuation of interest rate swaps
An interest rate swap is worth zero, or close to zero, when it is first initiated
After it has been in existence for some time, its value may become positive or negative
There are two valuation approaches:
1. The first regards the swap as the difference between two bonds
2. The second regards it as a portfolio of FRAs
Valuation in terms of bond prices
Principal payments are not exchanged in an interest rate swap
Assume principal payments received/paid at end of swap Same swap value
From the point of view of the floating-rate payer, a swap can be regarded as a long position in
a fixed-rate bond and a short position in a floating-rate bond, so that:
Vswap = Bf ix Bf l
Similarly, from the point of view of the fixed-rate payer, a swap is a long position in a floatingrate bond and a short position in a fixed-rate bond, so that the value of the swap is:
Vswap = Bf l Bf ix
To value floating-rate bond: Bond worth notional principal immediately after interest payment
At this time, bond = fair deal, borrower pays LIBOR for each subsequent accrual period
Suppose that notional principal = L, next exchange of payments is at t , and floating payment
made at t = k (determined at last payment date):
Immediately after payment, Bf l = L Immediately before payment Bf l = L + k
Floating-rate bond Instrument providing a single CF of L + k at t
Discounting at r = LIBOR/swap zero rate for maturity t , value of floating-rate bond today:
(L + k )er

Example
Suppose that a financial institution has agreed to pay 6-month LIBOR and receive 8% per
annum (with semiannual compounding) on a notional principal of $100 million
28

 The swap has a remaining life of 1.25 years. The LIBOR rates with continuous compounding
for 3-month, 9-month, and 15-month maturities are 10%, 10.5%, and 11%. The 6-month
LIBOR rate at the last payment date was 10.2% (with semiannual compounding)
The calculations for valuing the swap in terms of bonds are summarized in the table below
Time
0.25
0.75
1.25
Total:

Bf ix cash flow
4.0
4.0
104.0

Bf l cash flow
105.100

Discount factor
0.9753
0.9243
0.8715

PV Bf ix CF
3.901
3.697
90.640
98.238

PV Bf l CF
102.505

102.505

The fixed-rate bond has CFs of 4, 4, and 104 on the three payment dates. The discount factors
for these CFs are: e0.10.25 , e0.1050.75 , and e0.111.25 [Col (4)]
The table shows that the value of the fixed-rate bond (in millions) is 98.238
L = $100 million, k = 0.5 0.102 100 = $5.1 million, and t = 0.25, so that the floating-rate
bond can be valued as if it produces a CF of $105.1 million in 3 months
The current value of the floating bond (in millions) is 105.100 0.9753 = 102.505
Value of the swap: Difference between the two bond prices = Vswap = 98.238102.505 = 4.267
Valuation in terms of FRAs
A swap can be characterized as a portfolio of forward rate agreements
The first exchange of payments is known at the time the swap is negotiated
The other exchanges can be regarded as FRAs
An interest rate swap can also be valued by assuming that forward interest rates are realized:
1. Use LIBOR/swap zero curve to obtain forward rates for each LIBOR rates used for swap CFs
2. Calculate swap CFs assuming that LIBOR rates will equal forward rates
3. Discount these swap CFs (using LIBOR/swap zero curve) to obtain the swap value
Example
Valuing swap with FRAs: Floating CFs assume realized forward rates
Time
0.25
0.75
1.25
Total:

Fixed CF
4.0
4.0
4.0

Floating CF
-5.100
-5.522
-6.051

Net CF
-1.100
-1.522
-2.051

Discount Factor
0.9753
0.9243
0.8715

PV Net CF
-1.073
-1.407
-1.787
-4.267

The first row of the table shows the cash flows that will be exchanged in 3 months
The fixed rate of 8% leads to a cash inflow of 100 0.08 0.5 = $4 million
Floating rate of 10.2% (set 3 months ago) leads to outflow of 100 0.102 0.5 = $5.1M
Second row: Cash flows exchanged in 9 months assuming that forward rates are realized:
The cash inflow is $4.0 million as before
Cash outflow: First calculate the forward rate for period 3-9 months
From Eq. (4.2), this is: 0.1050.750.100.25
= 0.1075 (continuous compounding), or 11.044%
0.5
(semiannual compounding)
The cash outflow is therefore 100 0.11044 0.5 = $5.522 million
Third row: Cash flows that exchanged in 15 months assuming that forward rates are realized
Discount factors for the payment dates are: e0.10.25 , e0.1050.75 , and e0.111.25
The fixed rate in an interest rate swap is chosen so that the swap is worth zero initially
At the outset of a swap the sum of the values of the FRAs underlying the swap is zero
It does not mean that the value of each individual FRA is zero
In general, some FRAs will have positive values whereas others have negative values
Suppose that term structure of interest rates is upward-sloping when swap is negotiated
The forward interest rates increase as the maturity of the FRA increases
Since the sum of the values of the FRAs is zero, the forward interest rate must be less than
5.0% for the early payment dates and greater than 5.0% for the later payment date
If term structure of interest rates downward-sloping when swap is negotiated, reverse is true

29

Currency swaps
Exchange of principal/interest payments in one currency for principal/interest payments in another
A currency swap agreement requires the principal to be specified in each of the two currencies
Principal amounts in each currency are exchanged at the beginning/end of the swap life
Principal amounts chosen to be equivalent using exchange rate at swaps initiation
When they are exchanged at the end of the life of the swap, their values may be quite different
Illustration
Consider 5-yr currency swap between IBM and BP entered into on February 1, 2007
Suppose IBM pays a 5% fixed rate in and receives a 6% fixed rate in USD from BP
Interest rate payments are made once a year, principal amounts are $18M USD and 10M
When interest rates in both currencies are fixed, we have a fixed-for-fixed currency swap
At the outset of the swap, IBM pays $18 million and receives 10 million
Each year during swap, IBM receives $1.08M (= 6% of $18M) and pays 0.50M (= 5% of 10M)
At the end of the swap, it pays a principal of 10M and receives a principal of $18M USD
Use of a currency swap to transform liabilities and assets
A currency swap can be used to transform borrowings in one currency to borrowings in another
Suppose that IBM can issue $18 million of USD-denominated bonds at 6% interest
Swap transforms this transaction into one where IBM borrowed 10M at 5%
A currency swap can also be used to transform the nature of assets
Suppose IBM invests 10M in the UK to yield 5% per annum for the next 5 years, but feels
that USD will strengthen against and prefers a USD-denominated investment
Swap transforms UK investment into $18M investment in USD yielding 6%
Comparative advantage
Currency swaps can be motivated by comparative advantage
With plain vanilla interest rate swap, comparative advantages are largely illusory
When comparing rates offered in two different currencies, comparative advantages are genuine
One possible source of comparative advantage is tax
Borrowing rates providing basis for currency swap
Quoted rates have been adjusted to reflect the differential impact of taxes
General Electric
Qantas Airways

USD
5.0%
7.0%

AUD
7.6%
8.0%

Difference between USD rates is 2%, whereas difference between AUD rates is 0.4%
We expect the total gain to all parties to be 2.0 0.4 = 1.6% per annum
Fig. 2(a) shows one way swaps might be entered into with a financial institution
(a) Currency swap motivated by comparative advantage

(b) Quantas bears some foreign exchange risk

GE exchange a loan at 5% USD interest for a loan at 6.9% AUD interest rate

5.0%
USD

5.0%
USD
GE

6.9%
AUD

6.3%
USD
Financial
Institution 8.0%
AUD

Quantas

8.0%
AUD

5.0%
USD

Quantas exchanges a AUD loan at 8% for a USD loan at 6.3%

5.0%
USD
GE

6.9%
AUD

5.2%
USD
Financial
Institution 6.9%
AUD

Quantas

8.0%
AUD

The financial institution makes a 0.2% spread in USD

Figure 2: A currency swap motivated by comparative advantage

GE borrows USD and Qantas Airways borrows AUD
Swap transforms USD rate of 5% to AUD rate of 6.9% for GE. As a result, GE is 0.7% better
off than if it went directly to AUD markets
Similarly, Qantas exchanges an AUD loan at 8% per annum for a USD loan at 6.3% per annum
and ends up 0.7% per annum better off than it would be if it went directly to USD markets
The financial institution gains 1.3% per annum on its USD cash flows and loses 1.1% per
annum on its AUD flows, for a net gain of 0.2% per annum
30

 Possible to redesign swap so that financial institution makes a 0.2% spread in USD [Fig. 2(b)]
Unlikely in practice because they do not lead to GE/Qantas being free of FX risk
Qantas bears FX risk because it pays 1.1% in AUD and pays 5.2% in USD
Valuation of currency swaps
Like interest rate swaps, fixed-for-fixed currency swaps can be decomposed into either:
(i) The difference between two bonds or (ii) A portfolio of forward contracts
Valuation in terms of bond prices
Vswap = value in USD of swap where USD are received and foreign currency is paid:
Vswap = BD S0 BF
Where

BF = value, measured in foreign currency, of the bond defined by the foreign CFs on the swap
BD = value of the bond defined by domestic CFs on the swap
S0 = spot exchange rate (number of USD per unit of foreign currency)

The value of a swap can therefore be determined from LIBOR rates in the two currencies, the term
structure of interest rates in the domestic currency, and the spot exchange rate
Similarly, the value of a swap where the foreign currency is received and dollars are paid is:
Vswap = S0 BF BD
Example
Suppose that term structure of LIBOR/swap rates is flat in both Japan/US. Japanese rate is
4% per annum and the US rate is 9% per annum (continuous compounding)
Currency swap: Receives 5% per annum in yen and pays 8% per annum in dollars once a year
The principals in the two currencies are $10 million and 1,200 million yen
The swap will last for another 3 years, and the current exchange rate is 110 yen = $1
Time
1
2
3
3
Total:

CF dollar bond ($)

0.8
0.8
0.8
10.0

PV ($)
0.7311
0.6682
0.6107
7.6338
9.6439

CF yen bond (yen)

60
60
60
1,200

PV (yen)
57.65
55.39
53.22
1,064.30
1,230.55

9.6439 = 1.5430 million

The value of the swap in dollars is therefore: 1,230.55
110
Valuation as portfolio of forward contracts
Each exchange of payments in a fixed-for-fixed currency swap is a forward FX contract
Forward FX contracts valued by assuming that forward exchange rates are realized
The forward exchange rates themselves can be calculated from Eq. (5.9)
Example
Consider previous example: Valuation of currency swap as portfolio of forward contracts
Time
1
2
3
3
Total:

Dollar CF
-0.8
-0.8
-0.8
-10.0

Yen CF
60
60
60
1200

Forward
FX Rate
0.009557
0.010047
0.010562
0.010562

$ Value
of yen CF
0.5734
0.6028
0.6337
12.6746

Net CF ($)
-0.2266
-0.1972
-0.1663
+2.6746

PV
-0.2071
-0.1647
-0.1269
2.0417
1.5430

Financial institution pays 0.08 10 = $0.8M USD/receives 1, 200 0.05 = 60M yen each year
Principal of $10M USD is paid and yen principal of 1,200 is received at end of year 3
Current spot rate is 0.009091 dollar per yen. From Eq. (5.9), the 1-year forward rate is:
0.009091e(0.090.04)1 = 0.009557
If the 1-year forward rate is realized, the yen cash flow in year 1 is worth 600.009557 = 0.5734
million dollars and the net cash flow at the end of year 1 is 0.8 0.5734 = 0.2266 million
dollars. This has a present value of 0.2266e0.091 = 0.2071 million dollars
The value of the other forward contracts are calculated similarly
31

 The value of a currency swap is normally zero when it is first negotiated

However, as in the case of interest rate swaps, this does not mean that each of the individual
forward contracts underlying the swap has zero value
When rates in two currencies are different, payer of currency with high rate is in position where
forward contracts corresponding to early exchanges of CFs have negative values, and forward
contract corresponding to final exchange of principals has positive value
The payer of the currency with the low interest rate is in the opposite position
These results are important when the credit risk in the swap is being evaluated
Credit risk
If neither party defaults, the financial institution remains fully hedged
Decline in value of one contract always offset by increase in value of other contract
However, there is a chance that one party will get into financial difficulties and default
The financial institution then still has to honor the contract it has with the other party
Credit-risk exposure from swap only when swap value to the financial institution is positive
What happens when this value is negative and the counterparty gets into financial difficulties?
Financial institution could realize windfall gain (default would lead to getting rid of liability)
In practice, it is likely that the counterparty would choose to sell the contract to a third party or
rearrange its affairs in some way so that its positive value in the contract is not lost
Potential losses from defaults on swap  potential losses from defaults on loan with same principal
This is because the value of the swap is usually only a small fraction of the value of the loan
Potential losses from defaults on a currency swap are greater than on an interest rate swap
Reason: Principal amounts in two different currencies are exchanged at end of life of currency swap
Currency swap liable to have greater value at time of default than interest rate swap
Market risk arises from the possibility that market variables such as interest rates and exchange rates
will move in such a way that the value of a contract to the financial institution becomes negative
Market risks can be hedged relatively easily by entering into offsetting contracts
Credit risks are less easy to hedge
Banks trading swaps also sometimes bear legal risk
Other types of swaps
Variations on the standard interest rate swap
Tenor (payment frequency) on floating side does not have to match tenor on fixed side
LIBOR: Most common floating rate. Others, e.g. commercial paper rate, are occasionally used
Sometimes floating-for-floating interest rates swaps are negotiated
Allow to hedge exposure when assets/liabilities are subject to different floating rates
Principal in swap can be varied throughout term of swap to meet needs of a counterparty
Amortizing swap: Principal reduces in predetermined way ( loan amortization schedule)
Step-up swap: Principal increases in predetermined way ( loan drawdowns agreement)
Deferred swaps/forward swaps: Exchange interest payments do not begin until future date
Sometimes swaps are negotiated where the principal to which the fixed payments are applied
is different from the principal to which the floating payments are applied
Constant maturity swap (CMS): Agreement to exchange LIBOR for swap rate
Constant maturity Treasury swap (CMT swap): Exchange LIBOR/Treasury rate
Compounding swap: Interest on one/both sides is compounded forward to end of swap life
according to preagreed rules. Only one payment date at end of swap life
LIBOR-in arrears swap: LIBOR on payment date used to calculate payment on that date
Accrual swap: Interest on one side accrues only when floating reference rate in certain range
Other currency swaps
Cross-currency interest rate swap:
A fixed-for-floating currency swap, whereby a floating rate (usually LIBOR) in one currency
is exchanged for a fixed rate in another currency
Combination of a fixed-for-floating interest rate swap and a fixed-for-fixed currency swap
32

 Floating-for-floating currency swap: Floating rate in one currency/floating rate in another

Diff swap or quanto: Rate in one currency applied to principal in another currency
Equity swaps
An equity swap is an agreement to exchange the total return (dividends and capital gains) realized
on an equity index for either a fixed or a floating rate of interest
Equity swaps can be used by portfolio managers to convert returns from a fixed or floating investment to the returns from investing in an equity index, and vice versa
Options
Sometimes there are options embedded in a swap agreement
Extendable swap: One party has option to extend swap life beyond specified period
Puttable swap: One party has the option to terminate the swap early
Options on swaps or swaptions: Provide one party with the right at a future time to enter into
a swap where a predetermined fixed rate is exchanged for floating
Commodity swaps, volatility swaps, and other exotic instruments
Commodity swaps: A series of forward contracts on a commodity with different maturity dates
and the same delivery prices
Volatility swap: A series of time periods where at the end of each period, one side pays a
preagreed volatility, while the other side pays the historical volatility realized during the period.
Both volatilities are multiplied by the same notional principal in calculating payments

33

34

Options

Hull - Ch. 8: Mechanics of options markets

Background
Options are fundamentally different from forward and futures contracts
An option gives the right to do something, but the holder does not have to exercise this right
It costs a trader nothing (except for the margin requirements) to enter into a forward or futures
contract, whereas the purchase of an option requires an up-front payment

Profit ($)

Profit ($)

Types of options
There are two basic types of options:
Call option: Gives the right to buy an asset by a certain date for a certain price
Put option: Gives the holder the right to sell an asset by a certain date for a certain price
Expiration date or maturity date: The date specified in the contract
Exercise price or strike price: The price specified in the contract
Options can be either American or European:
American options can be exercised at any time up to the expiration date
European options can be exercised only on the expiration date itself
Call options
Consider a European call option with strike price of $100 to purchase 100 shares of a stock
Suppose that the current stock price is $98, the expiration date of the option is in 4 months, and
the price of an option to purchase one share is $5
Fig. 1 shows investors net profit/loss on a call option vs. final stock price
Profit from buying a European call option on one share of stock

20

BUYING CALL OPTION

Profit from buying a European put option on one share of stock

20

BUYING PUT OPTION

Option price = $5
Strike price = $100

Option price = $7
Strike price = $70

10

10

80

-5

90

100

110

0
Terminal stock price ($)

50

60

70

80

Terminal stock price ($)

-7

Figure 1: Profit from buying a European call option

Figure 2: Profit from buying a European put option

Put

options
The purchaser of a put option is hoping that the stock price will decrease
Consider a European put option with a strike price of $70 to sell 100 shares of a certain stock
Suppose that the current stock price is $65, the expiration date of the option is in 3 months, and
the price of an option to sell one share is $7
Fig. 2 shows investors profit/loss on a put option vs. terminal stock price
Early exercise
Exchange-traded stock options are generally American rather than European
The investor does not have to wait until the expiration date before exercising the option
Options positions
There are two sides to every option contract:
On one side is the investor who has taken the long position (i.e., has bought the option)
On the other side is the investor who has taken a short position (i.e., has sold or written the option)
The writer of an option receives cash up front, but has potential liabilities later
The writers profit or loss is the reverse of that for the purchaser of the option
37

Profit ($)

5
0

110
80

90

100

120

130

40

50

Terminal stock price ($)

60

70

Terminal stock price ($)

-10

-20

Profit ($)

80

90

-10
WRITING CALL OPTION
Option price = $5
Strike price = $100

WRITING PUT OPTION

Option price = $7
Strike price = $70

-20

-30

-30

Profit from writing a European call option on one share of stock

Profit from writing a European put option on one share of stock

Figure 3: Profit from writing European call option

Figure 4: Profit from writing European put option

There are four types of option positions:

(i) Long position in call, (ii) Long position in put, (iii) Short position in call, (iv) Short position in put

Payoff (a) Long Call

Payoff (b) Short Call

ST

ST

Strike price = K, Price of asset at maturity = ST

Payoff (c) Long Put

Payoff (d) Short Put

ST

ST

Figure 5: Payoffs from positions in European options

It is often useful to characterize a European option in terms of its payoff to the purchaser of the option
The initial cost of the option is then not included in the calculation
If K is the strike price and ST is the final price of the underlying asset, then:
Payoff from a long position in a European call option: max(ST K, 0)
Payoff from a short position in a European call option: max(ST K, 0) = min(K ST , 0)
Payoff from a long position in a European put option: max(K ST , 0)
Payoff from a short position in a European put option: max(K ST , 0) = min(ST K, 0)
Underlying assets
Stock options
Most trading in stock options is on exchanges, options trade on 1,000+ different stocks
One contract gives the holder the right to buy/sell 100 shares at specified strike price
Foreign currency options
Most currency options trading is in OTC market. Some exchange trading (Philadelphia)
The size of one contract depends on the currency
Index options
Many index options trade throughout the world (OTC market, exchange-traded market)
The most popular exchange-traded contracts in the US are (most contracts are European):
S&P 500 Index (SPX), S&P 100 Index (OEX), Nasdaq 100 Index (NDX), and Dow Jones
Industrial Index (DJX) on the Chicago Board Options Exchange (CBOE)
38

 One contract is usually to buy or sell 100 times the index at the specified strike price
Settlement is always in cash, rather than by delivering the portfolio underlying the index
Futures options
When an exchange trades a particular futures contract it often also trades options on that contract
A futures option normally matures just before the delivery period in the futures contract
When a call option is exercised, the holder acquires from the writer a long position in the underlying
futures contract plus a cash amount equal to the excess of the futures price over the strike price
When a put option is exercised, the holder acquires a short position in the underlying futures
contract plus a cash amount equal to the excess of the strike price over the futures price
Specification of stock options
Details of the contract - the expiration date, the strike price, what happens when dividends are declared,
how large a position investors can hold, and so on - are specified by the exchange
Expiration dates
One item used to describe a stock option is the month in which expiration date occurs
Precise expiration date: Saturday following third Friday of expiration month
The last day on which options trade is the third Friday of the expiration month
Stock options are on a January, February, or March cycle:
The January cycle consists of the months of January, April, July, and October
If expiration date for current month has not yet been reached, options trade with expiration
dates in: Current month, following month, and next two months in cycle
If the expiration date of the current month has passed, options trade with expiration dates in:
Next month, next-but-one month, and next two months of the expiration cycle
When one option reaches expiration, trading in another is started
Longer-term options (LEAPS: Long-term equity anticipation securities) trade on 500 US stocks
These have expiration dates up to 39 months into the future
The expiration dates for LEAPS on stocks are always in January
Strike prices
Exchange chooses strike prices at which options are written (spaced $2.50, $5, or $10 apart)
Typically the spacing is $2.50 when the stock price is between $5 and $25, $5 when the stock
price is between $25 and $200, and $10 for stock prices above $200
Stock splits and stock dividends can lead to nonstandard strike prices
When new expiration date introduced, 2-3 strike prices closest to current price are selected
If stock price moves outside range between highest/lowest strike prices, trading is usually
introduced in an option with a new strike price
Terminology
Option class: All options of the same type (calls or puts)
Option series: All options of a given class with same expiration date and strike price
Options are referred to as in the money, at the money, or out of the money:
If S is the stock price and K is the strike price, a call option is in the money when S > K, at
the money when S = K, and out of the money when S < K
Put option in the money (S < K), at the money (S = K), and out of the money (S > K)
In the absence of transactions costs, an in-the-money option will always be exercised on the
expiration date if it has not been exercised previously
Intrinsic value: Max[0, option value if exercised immediately]
For a call option, the intrinsic value is therefore: max(S K, 0)
For a put option, it is: max(K S, 0)
An in-the-money American option must be worth at least as much as its intrinsic value because
the holder can realize the intrinsic value by exercising immediately
Often it is optimal for the holder of an in-the-money American option to wait rather than
exercise immediately The option is said to have time value
Total value of an option = sum of its intrinsic value and its time value
39

 Flex options
Options where the traders on the floor of the exchange agree to nonstandard terms
Can involve strike price/expiration date different from usually offered by exchange
It can also involve the option being European rather than American
FLEX options are an attempt by option exchanges to regain business from the OTC markets
Dividends and stock splits
The early OTC options were dividend protected:
If a company declared a cash dividend, the strike price for options on the companys stock was
reduced on the ex-dividend day by the amount of the dividend
Exchange-traded options are not usually adjusted for cash dividends:
When a cash dividend occurs, there are no adjustments to the terms of the option contract
An exception is sometimes made for large cash dividends
Exchange-traded options are adjusted for stock splits:
A stock split occurs when the existing shares are split into more shares
E.g., in a 3-for-1 stock split, three new shares are issued to replace each existing share
In general, n-for-m stock split should cause stock price to go down to m/n of its previous value
Option contracts are adjusted to reflect expected changes in stock price arising from splits
After an n-for-m stock split, the strike price is reduced to m/n of its previous value, and the
number of shares covered by one contract is increased to n/m of its previous value
Stock options are adjusted for stock dividends:
A stock dividend involves a company issuing more shares to its existing shareholders
E.g., 20% stock dividend: Investors receive one new share for each five already owned
The 20% stock dividend referred to is essentially the same as a 6-for-5 stock split
All else being equal, it should cause the stock price to decline to 5/6 of its previous value
The terms of an option are adjusted to reflect the expected price decline arising from a stock
dividend in the same way as they are for that arising from a stock split
Adjustments are also made for rights issues: The basic procedure is to calculate the theoretical
price of the rights and then to reduce the strike price by this amount
Position limits and exercise limits
Position limit: CBOE often specifies a position limit for options
Maximum number of option contracts that an investor can hold on one side of the market
For this purpose, long calls/short puts are considered to be on same side of market
Also considered to be on the same side are short calls and long puts
Exercise limit: Usually equals the position limit
Maximum number of contracts that can be exercised in five consecutive business days
Options on largest/most frequently traded stocks have positions limits of 250,000 contracts
Position limits and exercise limits are designed to prevent the market from being unduly influenced
by the activities of an individual investor or group of investors
However, whether the limits are really necessary is a controversial issue
Trading
Traditionally, exchanges have had to provide a large open area for individuals to meet and trade options.
This is changing: Many derivatives exchanges are fully electronic
Market makers
Most options exchanges use market makers to facilitate trading
Market maker: Individual who will quote both a bid and an offer price on the option
The bid is the price at which the market maker is prepared to buy
The offer or asked is the price at which the market maker is prepared to sell
Market maker does not know if trader who asked for quotes wants to buy/sell the option
Offer always higher than bid, difference = bid-offer spread
The exchange sets upper limits for the bid-offer spread

40

 Existence of market maker ensures that buy/sell orders can always be executed at some price
without any delays Market makers add liquidity to the market
The market makers themselves make their profits from the bid-offer spread
Offsetting orders
Investor who purchased option can close out position by issuing offsetting order to sell same option
Investor who wrote option can close out position by issuing offsetting order to buy same option
Commissions
Orders that can be placed with a broker for options trading are similar to those for futures trading
For a retail investor, commissions vary significantly from broker to broker
Actual amount charged is often calculated as fixed cost + proportion of dollar amount of trade
If option position is closed out by entering offsetting trade, commission must be paid again
If option exercised, commission is the same as if investor placed order to buy/sell underlying stock
The commission system pushes retail investors to sell options rather than exercise them
A hidden cost in option trading (and in stock trading) is the market makers bid-offer spread
Margins
When shares are purchased in the US, investor can borrow up to 50% of the price from broker
This is known as buying on margin
If the share price declines so that the loan is substantially more than 50% of the stocks current
value, there is a margin call, where the broker requests that cash be deposited by the investor
If the margin call is not met, the broker sells the stock
When call/put options with maturities < 9 months are purchased, option price must be paid in full
Investors are not allowed to buy these options on margin because options already contain substantial
leverage and buying on margin would raise this leverage to an unacceptable level
For options with maturities > 9m, investors can buy on margin, borrowing up to 25% of option value
A trader who writes options is required to maintain funds in a margin account
Writing naked options
Naked option: Option not combined with an offsetting position in underlying stock
Initial margin required by the CBOE for a written naked call option is the greater of:
1. A total of 100% of the proceeds of the sale plus 20% of the underlying share price less the
amount, if any, by which the option is out of the money
2. A total of 100% of the option proceeds plus 10% of the underlying share price
For a written naked put option, it is the greater of:
1. A total of 100% of the proceeds of the sale plus 20% of the underlying share price less the
amount, if any, by which the option is out of the money
2. A total of 100% of the option proceeds plus 10% of the exercise price
The 20% in the preceding calculations is replaced by 15% for options on a broadly based stock
index because a stock index is usually less volatile than the price of an individual stock
Example
An investor writes four naked call option contracts on a stock
The option price is $5, the strike price is $40, and the stock price is $38
Option is $2 out of the money First calculation gives: 400 (5 + 0.2 38 2) = $4, 240
The second calculation gives: 400 (5 + 0.1 38) = $3, 520
The initial margin requirement is therefore $4,240
If put option, $2 in the money and margin requirement = 400 (5 + 0.2 38) = $5, 040
In both cases the $2,000 proceeds of the sale can be used to form part of margin account
A calculation similar to the initial margin calculation (but with the current market price replacing
the proceeds of sale) is repeated every day
Funds can be withdrawn from the margin account when the calculation indicates that the
margin required is less than the current balance in the margin account
When calculation indicates that a greater margin is required, a margin call is made
CBOE has special rules for margin requirements when various trading strategies are used
41

The options clearing corporation (OCC)

The Options Clearing Corporation (OCC) is similar to the clearinghouse for futures markets
It guarantees that options writers will fulfill their obligations under the terms of options contracts
and keeps a record of all long and short positions
The OCC has a number of members, and all option trades must be cleared through a member
Members are required to have a certain minimum amount of capital and to contribute to a special
fund that can be used if any member defaults on an option obligation
Exercising an option
When investor notifies broker to exercise, broker notifies OCC member that clears its trades
This member then places an exercise order with the OCC
The OCC randomly selects a member with an outstanding short position in the same option
The member, using pre-established procedure, selects an investor who has written the option
If the option is a call, this investor is required to sell stock at the strike price
If it is a put, the investor is required to buy stock at the strike price
The investor is said to be assigned
When an option is exercised, the open interest goes down by one
At the expiration of the option, all in-the-money options should be exercised unless the transactions
costs are so high as to wipe out the payoff from the option
Regulation
Options markets are regulated in a number of different ways:
Both the exchange and its OCC have rules governing the behavior of traders
In addition, there are both federal and state regulatory authorities
SEC regulates options markets in stocks, stock indices, currencies, and bonds at fed. level
The Commodity Futures Trading Commission regulates markets for options on futures
Illinois and New York actively enforce their own laws on unacceptable trading practices
Taxation
In the US, the general rule is that (unless the taxpayer is a professional trader) gains and losses from
the trading of stock options are taxed as capital gains or losses
For both the holder and the writer of a stock option, a gain or loss is recognized when:
(a) The option expires unexercised
(b) The option position is closed out
If the option is exercised, the gain or loss from the option is rolled into the position taken in the stock
and recognized when the stock position is closed out:
E.g., when a call option is exercised, the party with a long position is deemed to have purchased
the stock at the strike price plus the call price. This is then used as a basis for calculating this
partys gain or loss when the stock is eventually sold
Party with short call position is deemed to have sold the stock at strike price plus call price
When a put option is exercised, the seller of the option is deemed to have bought the stock for the
strike price less the original put price and the purchaser of the option is deemed to have sold the
stock for the strike price less the original put price
Wash sale rule
Tax authorities have ruled that when the repurchase is within 30 days of the sale (i.e., between 30
days before the sale and 30 days after the sale), any loss on the sale is not deductible
The disallowance also applies where, within the 61-day period, the taxpayer enters into an option
or similar contract to acquire the stock Selling a stock at a loss and buying a call option within
a 30-day period will lead to the loss being disallowed
The wash sale rule does not apply if the taxpayer is a dealer in stocks or securities and the loss is
sustained in the ordinary course of business
Constructive sales
Prior to 1997, if a US taxpayer shorted a security while holding a long position in a substantially
identical security, no gain or loss was recognized until the short position was closed out
42

 Short positions could be used to defer recognition of a gain for tax purposes
The situation was changed by the Tax Relief Act
An appreciated property is now treated as constructively sold when owner does one of:
1. Enters into a short sale of the same or substantially identical property
2. Enters into futures/forward contract to deliver the same/substantially identical property
3. Enters into one/more positions that eliminate substantially all of the loss/opportunity for gain
Transactions reducing only risk of loss/only opportunity for gain should not result in constructive
sales An investor holding a long position in a stock can buy in-the-money put options on the
stock without triggering a constructive sale
Many tax authorities have legislation designed to combat the use of derivatives for tax purposes
Warrants, employee stock options, and convertibles
Warrants
Options issued by a financial institution or nonfinancial corporation
To exercise the warrant, the holder would contact the financial institution
A common use of warrants by a nonfinancial corporation is at the time of a bond issue:
The corporation issues call warrants on its own stock and then attaches them to the bond issue
to make it more attractive to investors
Employee stock options
Call options issued to executives to motivate them to act in the best interests of shareholders
They are usually at the money at the time of issue
They are now a cost on the income statement of the company in most countries, making them a
less attractive form of compensation than they used to be
Convertibles bonds
Bonds that can be converted into equity at certain times using a predetermined exchange ratio
Bonds with an embedded call option on the companys stock
Warrants, employee stock options, and convertibles: A predetermined number of options are issued
By contrast, the number of exchange-traded options outstanding is not predetermined
When these instruments are exercised, the company issues more shares of its own stock and sells them
to the option holder for the strike price Increase in number of shares outstanding
By contrast, when exchange-traded option exercised, underlying company stock not involved
Over-the-counter (OTC) options markets
OTC market for options increasingly important since 80s, now larger than exchange-traded market
OTC options on foreign exchange and interest rates are particularly popular
The chief potential disadvantage of the OTC market is that option writer may default
This means that the purchaser is subject to some credit risk
Market participants are increasingly requiring counterparties to post collateral
Options traded in OTC market often structured by financial institutions to meet precise needs of clients:
E.g., exercise dates, strike prices, and contract sizes different from those traded by exchange
In other cases, option structure is different from standard calls/puts, aka exotic option

43

44

Hull - Ch. 9: Properties of stock options

Factors affecting option prices
There are six factors affecting the price of a stock option:
1. The current stock price S0
2. The strike price K
3. The time to expiration T

4. The volatility of the stock price

5. The risk-free interest rate r
6. The dividends expected during option life

Summary of effect on price of stock option of increasing one variable while keeping all others fixed
+ indicates that an increase in the variable causes the option price to increase
indicates that an increase in the variable causes the option price to decrease
? indicates that the relationship is uncertain
Variable
European Call European Put American Call American Put
Current stock price
+

Strike price

+
Time to expiration
?
?
+
+
Volatility
+
+
+
+
Risk-free rate
+

Amount of future dividends

+
Call option price c

Stock price S0
Put option price p

Stock price S0

Call option price c

Strike price K
Put option price p

Strike price K

Call option price c

Time to expiration T
Put option price p

Time to expiration T

Call option price c

Volatility (%)
Put option price p

Volatility (%)

Call option price c

Risk-free rate r (%)

Put option price p

Risk-free rate r (%)

Figure 1: Effect of changes in stock price, strike price, expiration date, volatility and rf
Stock price and strike price
Call options: More valuable as stock price increases and less valuable as strike price increases
Put options behave in the opposite way from call options: They become less valuable as the stock
price increases and more valuable as the strike price increases
Time to expiration
Put/call American options: More valuable (no decrease in value) as time to expiration increases:
Consider two American options that differ only as far as the expiration date is concerned
Long-life option has all exercise opportunities open to short-life option, and more
The long-life option must always be worth at least as much as the short-life option
European put/call options usually more valuable as time to expiration increases, but not always:
Suppose that a very large dividend is expected in between two expiration dates
Dividend Stock price to decline Short-life option may be > long-life option
Volatility
As volatility increases, the chance that the stock will do very well or very poorly increases
The owner of a call benefits from price increases but has limited downside risk in the event of price
decreases because the most the owner can lose is the price of the option
Similarly, puts benefit from price decreases, but limited downside risk with price increases
The values of both calls and puts increase as volatility increases
45

 Risk-free interest rate

As interest rates increase, expected return required by investors from stock tends to increase
In addition, the PV of any future cash flow received by the holder of the option decreases
Combined impact: Increase value of call options and decrease value of put options
Important: We are assuming that interest rates change while all other variables stay the same
In practice, when interest rates rise (fall), stock prices tend to fall (rise)
Net effect can be to decrease value of a call option and increase value of a put option
Amount of future dividends
Dividends have the effect of reducing the stock price on the ex-dividend date
The value of a call option is therefore negatively related to the size of an anticipated future dividend,
and the value of a put option is positively related to the size of an anticipated future dividend
Assumptions and notations
Assume that for some market participants, such as large investment banks:
1. There are no transactions costs
2. All trading profits (net of trading losses) are subject to the same tax rate
3. Borrowing and lending are possible at the risk-free interest rate
Assume that there are no arbitrage opportunities
Notations:
S0 :
K:
T:
ST :
r:

Current stock price

Strike price of option
Time to expiration of option
Stock price on the expiration date
Risk-free nominal rate for maturity T

C:
P:
c:
p:

Value
Value
Value
Value

of
of
of
of

American
American
European
European

call
put
call
put

option
option
option
option

to
to
to
to

buy one share

sell one share
buy one share
sell one share

Upper and lower bounds for option prices

Upper bounds
American C or European c call option
Gives the holder the right to buy one share of a stock for a certain price
No matter what happens, the option can never be worth more than the stock
Hence, stock price is an upper bound to option price: c S0 and C S0
If not true, arbitrageur could make riskless profit by buying stock/selling call option
American or European put option
Gives the holder the right to sell one share of a stock for K
Even for low stock price, option never worth more than K p K and P K
European put option
For European options, at maturity the option cannot be worth more than K
It follows that it cannot be worth more than the present value of K today: p < KerT
If not true, arbitrageur could profit by writing the option/investing proceeds at r
Lower bound for calls on non-dividend-paying stocks
Lower bound for price of European call option on a non-dividend-paying stock:
max(S0 KerT , 0) c

(1)

Consider the following two portfolios:

Portfolio A: One European call option plus amount of cash = KerT
Portfolio B: One share
In portfolio A, the cash, if it is invested at the risk-free interest rate, will grow to K in time T
If ST > K, the call option is exercised at maturity and portfolio A is worth ST
If ST < K, the call option expires worthless and the portfolio is worth K
Hence, at time T , portfolio A is worth: max(ST , K)
Portfolio B is worth ST at time T
Portfolio A always worth as much as portfolio B (if not more) at maturity
46

 It follows that in the absence of arbitrage opportunities this must also be true today
c + KerT S0

c S0 KerT

Worst for a call option: It expires worthless c 0 and: c max(S0 KerT , 0)

Lower bound for European puts on non-dividend-paying stocks
For a European put option on a non-dividend-paying stock, a lower bound for the price is:
max(KerT S0 , 0) p

(2)

Consider the following two portfolios:

Portfolio C: One European put option plus one share
Portfolio D: An amount of cash equal to KerT
If ST < K, then option in portfolio C is exercised at maturity, and portfolio is worth K. If ST > K,
then put option expires worthless, and portfolio is worth ST at this time
At time T , Portfolio C is worth max(ST , K)
Assuming cash is invested at the risk-free interest rate, portfolio D is worth K at T
Hence, portfolio C is always worth as much as portfolio D (if not more) at T
If no arbitrage, portfolio C must be worth at least as much as portfolio D today:
p + S0 KerT p KerT S0
Worst case: Put option expires worthless p 0 and: p max(KerT S0 , 0)
Put-call parity
The put-call parity is the most important relationship between p and c
Consider the following two portfolios:
Portfolio A: One European call option plus amount of cash = KerT
Portfolio C: One European put option plus one share
Both are worth max(ST , K) at expiration. Because the options are European, they cannot be
exercised prior to the expiration date Portfolios must have identical values today:
c + KerT = p + S0

(3)

The value of a European call with a certain strike price and exercise date can be deduced from the value
of a European put with the same strike price and exercise date, and vice versa
If the put-call parity Eq. (3) does not hold, there are arbitrage opportunities:
Stock price = $31, interest rate = 10%, call price = $3
Both put and call have a strike price of $30 and 3 months to maturity
Three-month put price = $2.25 Three-month put price = $1
Action now:
Buy call for $3
Borrow $29 for 3 months
Short put to realize $2.25
Short call to realize $3
Short the stock to realize $31
Buy put for $1
Invest $30.25 for 3 months
Buy the stock for $31
Action in 3 months if ST > 30:
Receive $31.02 from investment
Call exercised: Sell stock for $30
Exercise call to buy stock for $30
Use $29.73 to repay loan
Net profit = $1.02
Net profit = $0.27
Action in 3 months if ST < 30:
Receive $31.02 from investment
Exercise put to sell stock for $30
Put exercised: buy stock for $30
Use $29.73 to repay loan
Net profit = $1.02
Net profit = $0.27

Put-call parity and capital structure

Consider a company that has assets that are financed with zero-coupon bonds and equity
Suppose that bonds mature in 5 years at which time a principal payment of K is required
If assets are worth more than K in 5 years, equityholders repay bondholders
47

If assets < K, equityholders declare bankruptcy and bondholders own the company
Value of equity in 5 years = max(AT K, 0), where AT = value of companys assets at T
Equityholders have 5-yr European call option on company assets with strike price of K
Bondholders get min(AT , K) = K max(K AT , 0) in 5 yrs Today, bonds are worth PV of [K
minus value of 5-yr European put option on assets with strike price K]
If c and p are the value of the call and put options on the companys assets at time T , then:
Value of equity = c
Value of debt = PV(K) p
A0 = value of company assets today = total value of instruments used to finance assets
A0 must equal the sum of the value of the equity and the value of the debt:
A0 = c + [P V (K) p]

c + P V (K) = p + A0

This is the put-call parity result in Eq. (3) for call/put options on assets of the company
American options
Put-call parity holds only for European options
For American option prices, when there are no dividends:
S0 K C P S0 KerT

(4)

Early exercise: Calls on a non-dividend-paying stock

Never optimal to exercise American call on non-dividend-paying stock early. Two reasons:
1. One relates to the insurance that it provides:
Call option, when held instead of stock, insures against stock price falling below strike price
Once option is exercised and strike price is exchanged for stock price, insurance vanishes
2. The other reason concerns the time value of money:
From perspective of option holder, the later the strike price is paid out, the better
For a more formal argument, we can use Eq. (1): c S0 KerT
American call has all exercise opportunities of the corresponding European call We must have:
Cc

Fig.

C S0 KerT

Given r > 0, it follows that C > S0 K

If optimal to exercise early, C would equal S0 K Never optimal to exercise early
2 shows the general way in which the call price varies with S0
It indicates that the call price is always above its intrinsic value of max(S0 K, 0)
As r or T or the volatility increases, the line relating the call price to the stock price moves in the
direction indicated by the arrows (i.e., farther away from the intrinsic value)

Call option price

Put option price

American
European

Increase in r, T, or

Ke{rT

A
Increase T or , decrease in r
B

Intrinsic value = max(S0-K,0)

Intrinsic value = K - S0

Stock price S0

Figure 2: Price of a call option vs. stock price S0

Stock price S0

Figure 3: Price of a put option vs. stock price S0

48

Early exercise: Puts on a non-dividend-paying stock

It can be optimal to exercise an American put option on a non-dividend-paying stock early:
A put option should always be exercised early if it is sufficiently deep in the money
Like a call option, a put option can be viewed as providing insurance:
Put option, when held with stock, insures against stock price falling below certain level
However, a put option is different from a call option in that it may be optimal for an investor to
forgo this insurance and exercise early in order to realize the strike price immediately
Early exercise of put option more attractive as S0 decreases, as r increases, and as volatility decreases
Eq. (2): p KerT S0 . American put P : Stronger condition P > K S0 holds (immediate exercise)
Fig. 3 shows the general way in which the price of an American put varies with S0
When early exercise is optimal, the value of the option is K S0
Value of put merges into puts intrinsic value, K S0 , for small value of S0 (point A)
The line relating the put price to the stock price moves in the direction indicated by the arrows
when r decreases, when the volatility increases, and when T increases
Because there are some circumstances when it is desirable to exercise an American put option early, it
follows that American put option always worth more than corresponding European put option
Point B (price of option = intrinsic value) represents higher value of stock price than A
Point E is where S0 = 0 and the European put price is KerT
Effect of dividends
In the US most exchange-traded stock options have a life of less than 1 year and dividends payable
during the life of the option can usually be predicted with reasonable accuracy
Let D to denote the PV of the dividends during the life of the option
In the calculation of D, a dividend is assumed to occur at the time of its ex-dividend date
Lower bound for calls and puts
Redefine portfolios A and B as follows:
Portfolio A: One European call option plus amount of cash = D + KerT
Portfolio B: One share
A similar argument to the one used to derive Eq. (1) shows that:
c S0 D KerT

(5)

Also redefine portfolios C and D as follows:

Portfolio C: One European put option plus one share
Portfolio D: An amount of cash = D + KerT
A similar argument to the one used to derive Eq. (2) shows that:
p D + KerT S0

(6)

Early exercise
When dividends expected, no longer true that American call option not exercised early
Sometimes, optimal to exercise American call immediately prior to an ex-dividend date
It is never optimal to exercise a call at other times
Put-call parity
With dividends, the put call parity result in Eq. (3) becomes:
c + D + KerT = p + S0

(7)

Dividends cause Eq. (4) to be modified to:

S0 D K C P S0 KerT

(8)

49

50

Hull - Ch. 10: Trading strategies involving options

b)

Sh

a) Covered call

C
al
l

Strategies involving a single option and a stock

Lo
ng

k
oc

St

Profit

t
or

ST

ST

Lo

ng

l
al

St

tC

oc

or

Sh
d)

tS
or
Sh

c) Protective put

ck

to

Short Put

ST

ST

Long Put

Lo

ng

St

oc

Figure 1: Profit patterns involving a single option and a stock

In all figures: Dashed line = relationship between profit/stock price for individual securities. Solid line
= relationship between profit/stock price for the whole portfolio
Writing a covered call
Fig. 1(a): Portfolio consists of long position in stock plus short position in call option
The long stock position covers the investor from the payoff on the short call that becomes
necessary if there is a sharp rise in the stock price
Reverse of writing a covered call
In Fig. 1(b), a short position in a stock is combined with a long position in a call option
Protective put
In Fig. 1(c), the investment strategy involves buying a put option on a stock and the stock itself
Reverse of a protective put
In Fig. 1(d), a short position in a put option is combined with a short position in the stock
The profit patterns in Fig. 1 have the same general shape as the profit patterns discussed in Ch. 8 for
short put, long put, long call, and short call, respectively
Put-call parity provides a way of understanding why this is so:
p + S0 = c + KerT + D

(1)

Eq. (1) shows that a long position in a put combined with a long position in the stock is equivalent
to a long call position plus a certain amount (= KerT + D) of cash
The profit pattern in Fig. 1(c) is similar to the profit pattern from a long call position
The position in Fig. 1(d) is the reverse of that in Fig. 1(c)
The profit pattern similar to that from a short call position
Eq. (1) can be rearranged to become: S0 c = KerT + D p
This shows that a long position in a stock combined with a short position in a call is equivalent to
a short put position plus a certain amount (= KerT + D) of cash
The profit pattern in Fig. 1(a) is similar to the profit pattern from a short put position
The position in Fig. 1(b) is the reverse of that in Fig. 1(a)
The profit pattern similar to that from a long put position
51

Bull spread from 2 call options

Profit

Profit

Spreads
Spread strategy: Taking a position in two/more options of same type (e.g., two or more calls)
Bull spreads
Created by buying a call option on a stock with a certain strike price and selling a call option on
the same stock with a higher strike price (both options have same expiration date)
Bull spread from 2 put options
Short put option (K2)

Short call option (K2)

K1

K1
K2

K2

ST

ST
Long put option (K1)

Long call option (K1)

Figure 2: Profit from bull spread using calls

Figure 3: Profit from bull spread using puts

Call price always decreases as strike price increases Value of option sold < value of option bought
A bull spread, when created from calls, requires an initial investment
Bull spreads also created by buying put (low strike price) and selling put (high strike price)
Unlike bull spread created from calls, bull spreads from puts involve positive up-front CF to
investor (ignoring margin requirements) and payoff 0
Payoff from a bull spread created using calls
Stock price

Long call payoff

Short call payoff

Total payoff

ST K1
K1 < ST < K2
ST K2

0
ST K1
ST K1

0
0
(ST K2 )

0
ST K1
K2 K1

A bull spread strategy limits the investors upside as well as downside risk:
Strategy: The investor has a call option with a strike price equal to K1 and has chosen to give
up some upside potential by selling a call option with strike price K2 (K2 > K1 )
In return for giving up upside potential, investor gets the price of option with strike price K2
Three types of bull spreads can be distinguished:
1. Both calls are initially out of the money (most aggresive)
2. One call is initially in the money, the other call is initially out of the money
3. Both calls are initially in the money
Type 1: Cost very little to set up/small probability of a high payoff (= K2 K1 )
As we move from type 1 2 and from 2 3, spreads become more conservative
Bear spreads
Bear spreads created by buying put with one strike price/selling put with another strike price
The strike price of the option purchased is greater than the strike price of the option sold
Payoff from a bear spread created with put options
Stock price

Long put payoff

Short put payoff

Total payoff

ST K1
K1 < ST < K2
ST K2

K2 ST
K2 ST
0

(K1 ST )
0
0

K2 K1
K2 ST
0

Bear spread from puts involves initial cash outflow (price of put sold < price of put purchased)
In essence, the investor has bought a put with a certain strike price and chosen to give up some
of the profit potential by selling a put with a lower strike price
In return for the profit given up, the investor gets the price of the option sold
52

Profit

Profit

Bear spread from 2 put options

Bear spread from 2 call options

Short call option (K1)

Short put option (K1)

K2

K2

K1

K1

ST

ST

Long call option (K2)

Long put option (K2)

Figure 4: Profit from bear spread using puts

Figure 5: Profit from bear spread using calls

Bear spreads can be created using calls instead of puts:

Investor buys call with high strike price/sells call with low strike price (Fig. 5)
Bear spreads from calls involve an initial cash inflow (ignoring margin requirements)
Comparison with bull spread
Like bull spreads, bear spreads limit both upside profit potential/downside risk
An investor who enters into a bull spread is hoping that the stock price will increase
By contrast, an investor who enters into a bear spread hopes that stock price will decline
Box spreads
A box spread is a combination of a bull call spread with strike prices K1 and K2 and a bear put
spread with the same two strike prices. The payoff from a box spread is always K2 K1
Stock price

Bull call spread payoff

Bear put spread payoff

Total payoff

ST K1
K1 < ST < K2
ST K2

0
ST K1
K2 K1

K2 K1
K2 ST
0

K2 K1
K2 K1
K2 K1

Profit

Butterfly spread using put options

2 short put options (K2)

op

ll
o

tio

pt

2 short call options (K2)

(K

io

(K

Butterfly spread using call options

Lo

ng

Lo

ca

ng

ll

ca

Profit

The value of a box spread is therefore always the PV of this payoff or (K2 K1 )erT
If it has a different value there is an arbitrage opportunity:
If the market price of the box spread is too high, it is profitable to sell the box
This involves buying a call with strike price K2 , buying a put with strike price K1 , selling a
call with strike price K1 , and selling a put with strike price K2
A box-spread arbitrage only works with European options
Butterfly spreads
A butterfly spread involves positions in options with three different strike prices:
Buy call option with low strike price K1 , buy call option with high strike price K3 , and sell
two call options with strike price K2 halfway between K1 and K3

K1

K3
K2

K1

K3
K2

ST

ST
Long put option (K1)

Long put option (K3)

Figure 6: Profit from butterfly spread using calls

Figure 7: Profit from butterfly spread using puts

53

 Profit if stock price stays close to K2 , but small loss if significant price move in either direction
Appropriate strategy for an investor who feels that large stock price moves are unlikely
The strategy requires a small investment initially
Payoff from a butterfly spread (assuming K2 = (K1 + K3 )/2)
Stock price

1st long call payoff

2nd long call payoff

Short calls payoff

Total payoff

ST K1
K1 < ST < K2
K2 < ST < K3
ST K3

0
ST K1
ST K1
ST K1

0
0
0
ST K3

0
0
2(ST K2 )
2(ST K2 )

0
ST K1
K3 ST
0

Calendar spread using two call options

Profit

Profit

Butterfly spreads can be created using put options:

The investor buys one put with a low strike price, another with a high strike price, and sells
two puts with an intermediate strike price, see Fig. 7
If all options are European, use of puts results in same spread as with calls
Put-call parity can be used to show that the initial investment is the same in both cases
A butterfly spread can be sold or shorted by following the reverse strategy:
Options sold with strike prices K1 /K3 , and two options purchased with middle strike price K2
This strategy produces a modest profit if there is a significant movement in stock price
Calendar spreads
In calendar spreads, the options have the same strike price and different expiration dates
A calendar spread can be created by selling a call option with a certain strike price and buying a
longer-maturity call option with the same strike price
The longer the maturity, the more expensive it is Calendar spread requires initial investment
Calendar spread using two put options

Short call option (K, short maturity)

Short put option (K, short maturity)

ST

ST

Long put option (K, longer maturity)

Long call option (K, longer maturity)

Figure 8: Profit from calendar spread using calls

Figure 9: Profit from calendar spread using puts

Profit diagrams for calendar spreads are usually produced showing profit when short-maturity
option expires on the assumption that long-maturity option is sold at that time (Fig. 8)
The pattern is similar to the profit from the butterfly spread in Fig. 6
The investor makes a profit if the stock price at the expiration of the short-maturity option
is close to the strike price of the short-maturity option. However, a loss is incurred when the
stock price is significantly above or significantly below this strike price
Qualitative explanation of the profit pattern
If the stock price is very low when the short-maturity option expires, the short-maturity option
is worthless and the value of the long-maturity option is close to zero
The investor incurs a loss that is close to the cost of setting up the spread initially
If the stock price ST is very high when the short-maturity option expires, the short-maturity
option costs the investor ST K, and the long-maturity option is worth close to ST K
Again, the investor makes a net loss that is close to the cost of setting up the spread initially
If ST is close to K, the short-maturity option costs the investor either a small amount or
nothing at all. However, the long-maturity option is still quite valuable
In this case a significant net profit is made
54

 Calendar spreads can be created with put options as well as call options:
The investor buys a long-maturity put option and sells a short-maturity put option
As shown in Fig. 9, the profit pattern is similar to that obtained from using calls
Types of calendar spreads
Neutral calendar spread: A strike price close to the current stock price is chosen
Bullish calendar spread: Involves a higher strike price
Bearish calendar spread: Involves a lower strike price
Reverse calendar spread: Opposite to that in Figs. 8 and 9:
The investor buys a short-maturity option and sells a long-maturity option
A small profit arises if the stock price at the expiration of the short-maturity option is well
above or well below the strike price of the short-maturity option
However, a significant loss results if it is close to the strike price
Diagonal spreads
Both expiration date/strike price of calls are different Increases range of possible profit patterns
Combinations
Combination: Strategy that involves taking a position in both calls and puts on the same stock
Straddle
A straddle involves buying a call and put with the same strike price and expiration date
If stock price close to strike price at expiration of the options, straddle leads to a loss
However, if there is a sufficiently large move in either direction, significant profit results
Profit

Bottom straddle (straddle purchase)

Long call option (K, maturity T)
Long put option (K, maturity T)

K
ST

Figure 10: Profit from a straddle

Payoff from a straddle
Stock price

Call payoff

Put payoff

Total payoff

ST K
ST > K

0
ST K

K ST
0

K ST
ST K

Straddle appropriate when investor expects large move in stock price but direction unknown
Carefully consider whether the anticipated jump is already reflected in option prices
The straddle in Fig. 10 is sometimes referred to as a bottom straddle or straddle purchase
A top straddle or straddle write is the reverse position:
It is created by selling a call and a put with the same exercise price and expiration date
It is a highly risky strategy:
If stock price at expiration is close to strike price, significant profits
However, loss arising from a large move is unlimited
Strips and straps
Strip consists of long position in one call/two puts with same strike price/expiration date
Strap consists of long position in two calls/one put with same strike price/expiration date
In a strip the investor is betting that there will be a big stock price move and considers a decrease
in the stock price to be more likely than an increase
55

Profit

Strap

Profit

Strip
Long call (K, T)

2 long calls (K, T)

K
ST

ST

2 long puts (K, T)

Long put (K, T)

Figure 11: Profit from a strip and a strap

ca

ll

(K

Strangle (bottom vertical combination)

ng

Lo

Lo
ng

Profit

In a strap the investor is also betting that there will be a big stock price move. However, in this
case, an increase in the stock price is considered to be more likely than a decrease
Strangles
In a strangle, sometimes called a bottom vertical combination, an investor buys a put and a call
with the same expiration date and different strike prices

)
K1

t(

pu
K1

K2
ST

Figure 12: Profit from a strangle

The call strike price K2 is higher than the put strike price K1
Profit pattern from strangle depends on how close the strike prices are: The farther they are
apart, the less the downside risk/the farther the stock price has to move for a profit
Payoff from a strangle
Stock price

Call payoff

Put payoff

Total payoff

ST K1
K1 < ST < K2
ST > K2

0
0
ST K2

K1 ST
0
0

K1 ST
0
ST K2

A strangle is a similar strategy to a straddle:

Investor betting on large price move, but uncertain direction
Stock price has to move farther in strangle than in straddle for investor to profit
However, downside risk if stock price ends up at a central value is less with a strangle
The sale of a strangle is sometimes referred to as a top vertical combination:
Appropriate for an investor who feels that large stock price moves are unlikely
However, as with sale of straddle, it is a risky strategy involving unlimited potential loss
Other payoffs
If European options expiring at time T were available with every single possible strike price, any payoff
function at time T could in theory be obtained
Through the judicious combination of a large number of very small spikes (payoff from a butterfly
spread), any payoff function can be approximated
56

Hull - Ch. 11: Binomial trees

Binomial tree
Diagram representing different possible paths of a stock price over the life of an option
Useful and very popular technique for pricing an option
The underlying assumption is that the stock price follows a random walk
In each time step, it has certain probabilities of moving up/down by a certain % amount
As time step 0, model leads to lognormal assumption for stock prices (Black-Scholes model)
A one-step binomial model and a no-arbitrage argument
A very simple situation
Stock price currently $20, and at the end of 3 months it will be either $22 or $18
We are interested in valuing a European call option to buy the stock for $21 in 3 months
The only assumption needed is that arbitrage opportunities do not exist
We set up a portfolio of the stock and the option in such a way that there is no uncertainty about
the value of the portfolio at the end of the 3 months
We then argue that, because portfolio has no risk, the return it earns must equal rf
This enables us to work out cost of setting up the portfolio and thus the options price
Because there are two securities (the stock and the stock option) and only two possible outcomes,
it is always possible to set up the riskless portfolio
Consider portfolio: Long position in shares of stock and short position in one call option
We calculate the value of that makes the portfolio riskless
If the stock price moves up from $20 to $22, the value of the shares is 22 and the value of
the option is 1, so that the total value of the portfolio is 22 1
If the stock price moves down from $20 to $18, the value of the shares is 18 and the value of
the option is zero, so that the total value of the portfolio is 18
Portfolio riskless if: 22 1 = 18 = 0.25, i.e. Long: 0.25 shares, Short: 1 option
Riskless portfolios must, in the absence of arbitrage opportunities, earn the risk-free rate
Suppose that in this case the risk-free rate is 12% per annum
Portfolio today must be worth PV of 4.5 (= 22 0.25 1 = 18 0.25) = 4.5e0.123/12 = 4.367
The value of the stock price today is known to be $20. Suppose the option price is denoted by f
The value of the portfolio today is: 20 0.25 f = 5 f = 4.367 f = 0.633
In the absence of arbitrage opportunities, the current value of the option must be 0.633
Generalization
Suppose that the option lasts for time T and that during the life of the option the stock price can
either move up from S0 to a new level S0 u (u > 1), or down from S0 to a new level S0 d (d < 1)
% increase in stock price when up movement: u 1. % decrease when down movement: 1 d
If the stock price moves up to S0 u, we suppose that the payoff from the option is fu . If the stock
price moves down to S0 d, we suppose the payoff from the option is fd
Consider portfolio consisting of long position in shares and short position in one option. We
calculate the value of that makes the portfolio riskless:
S0 u fu = S0 d fd

fu fd
S0 u S0 d

(1)

Eq. (1): = ratio [change in option price] [change in stock price] moving between nodes at T
PV of portfolio = (S0 u fu )erT , and cost of setting up portfolio = S0 f . Hence:
S0 f = (S0 u fu )erT

f = S0 (1 uerT ) + fu erT

Eqs. (2)/(3) enable option to be priced when stock moves are given by a one-step binomial tree:
f = erT [pfu + (1 p)fd ]
p=

(2)

erT d
ud

(3)

57

 Irrelevance of the stocks expected return

Option pricing formula in Eq. (2) does not involve probabilities of stock moving up/down:
This is surprising and seems counterintuitive
The key reason is that we are not valuing the option in absolute terms:
We are calculating its value in terms of the price of the underlying stock
Probabilities of future up/down moves already incorporated into stock price: No need to take
them into account again when valuing option in terms of stock price
Risk-neutral valuation
Natural to interpret variable p in Eq. (2) as the probability of an up movement in stock price
1 p is then probability of down move, and pfu + (1 p)fd = expected payoff from option
Eq. (2) then states that the option value today is its expected payoff discounted risk-free
Expected return from the stock when the probability of an up movement is p:
The expected stock price E[ST ] at time T is:
E[ST ] = pS0 u + (1 p)S0 d = pS0 (u d) + S0 d = S0 erT

(4)

Probability of up move p Assuming that return on stock = r

Risk neutral world
All individuals are indifferent to risk. In such a world, investors require no compensation for risk,
and the expected return on all securities is the risk-free interest rate
Eq. (4): We assume risk-neutral world when we set probability of an up move to p
Eq. (2): Option value = its expected payoff in a risk-neutral world discounted risk-free
Principle of risk-neutral valuation:
States that we can with complete impunity assume the world is risk neutral when pricing options
The resulting prices are correct not just in a risk-neutral world, but in other worlds as well
The one-step binomial example revisited
Define p as the probability of an upward movement in the stock price in a risk-neutral world:
We can calculate p from Eq. (3)
Alternatively, argue that expected return on stock in risk-neutral world = risk-free rate of 12%
p must satisfy: 22p + 18(1 p) = 20e0.123/12 p = 0.6523
At the end of the 3 months, the call option has a 0.6523 probability of being worth 1 and a 0.3477
probability of being worth zero. Its expected value is therefore: 0.6523 1 + 0.3477 0 = 0.6523
In a risk-neutral world, the value of the option today is therefore: 0.6523e0.123/12 = 0.633
No-arbitrage arguments and risk-neutral valuation give the same answer
Real world vs. risk-neutral world
p = probability of up move in risk-neutral world 6= probability of up move in real world (in general)
Suppose, in real world, expected return is 16% and p = probability of up move:
22p + 18(1 p ) = 20e0.163/12

p = 0.7041

The expected payoff from the option in the real world is then: p 1 + (1 p ) 0 = 0.7041
Unfortunately, correct discount rate to apply to expected payoff in real world is unknown
A position in a call option is riskier than a position in the stock
The discount rate to be applied to the payoff from a call option is greater than 16%
Using risk-neutral valuation is convenient because in a risk-neutral world the expected return on
all assets (and thus the discount rate to use for all expected payoffs) is the risk-free rate
Two-step binomial trees
We can extend the analysis to a two-step binomial tree such as that shown in Fig. 1
Stock price initially S0 . During each time step, it moves up/down to [u/d] [its initial value]
Suppose that the risk-free interest rate is r and the length of the time step is t years
Option prices at final nodes of tree are easily calculated: They are the option payoffs
58

fu = e{rt[pfuu + (1{p)fud]

S0u2

rt

{|d
p = e|||
u{d

fd = e{rt[pfud + (1{p)fdd]

fuu

{rt

f=e

[pfu + (1{p)fd]
p

S0
f

S0u
fu
1{p

p
1{p

S0ud
fud

S0d
fd
1{p

{2rt

)f=e

[p fuu + 2p(1{p)fud + (1{p) fdd]

S0d2
fdd

Figure 1: Stock and option prices in general two-step tree

Eqs. (2) and (3) become:
f = ert [pfu + (1 p)fd ]
p=

(5)

ert

d
ud

(6)

Repeated application of Eq. (5) gives:

fu = ert [pfuu + (1 p)fud ]
fd = e

rt

(7)

[pfud + (1 p)fdd ]

(8)

f = ert [pfu + (1 p)fd ]

(9)

Substituting Eqs. (7) and (8) into Eq. (9), we get:

f = e2rt [p2 fuu + 2p(1 p)fud + (1 p)2 fdd ]

(10)

This is consistent with the principle of risk-neutral valuation

p2 , 2p(1 p), (1 p)2 are probabilities that upper/middle/lower final nodes reached
Option price = its expected payoff in risk-neutral world discounted risk-free
As we add more steps to the binomial tree, the risk-neutral valuation principle continues to hold
Option price always = its expected payoff in risk-neutral world discounted risk-free
Put options
The procedures described above can be used to price puts as well as calls
American options
The procedure is to work back through the tree from the end to the beginning, testing at each
node to see whether early exercise is optimal
The value of the option at the final nodes is the same as for the European option
At earlier nodes the value of the option is the greater of:
1. The value given by Eq. (5)
2. The payoff from early exercise
Delta
is an important parameter in the pricing and hedging of options
= ratio of [change in option price] [change in underlying stock price]
= # of units of stock we should hold for each option shorted to create riskless portfolio
The construction of a riskless hedge is sometimes referred to as delta hedging
The delta of a call option is positive, whereas the delta of a put option is negative
changes over time To maintain a riskless hedge, we need to adjust holdings in stock periodically
59

Matching volatility with u and d

In practice, when constructing a binomial tree, we choose u and d to match volatility of stock price
Suppose that the expected return on a stock (in the real world) is and its volatility is
The step is of length t. Stock price starts at S0 and moves either up to S0 u or down to S0 d
Probability of up move in real world is p and in risk-neutral world it is p
The expected stock price at the end of the first time step in the real world is S0 et . On the tree,
the expected stock price at this time is: p S0 u + (1 p )S0 d
To match the expected return on the stock with the trees parameters, we must have:
p S0 u + (1 p )S0 d = S0 et

p =

et d
ud

(11)

The volatility of a stock price is defined so that t is the std. dev. of the return on the stock price
in a short period of time of length t Variance of return = 2 t
On the tree, variance of stock price return = p u2 + (1 p )d2 [p u + (1 p )d]2
To match the stock price volatility with the trees parameters, we must have:
p u2 + (1 p )d2 [p u + (1 p )d]2 = 2 t

(12)

Substituting Eq. (11) into Eq. (12) gives: et (u + d) ud e2t = 2 t and when terms in t2
and higher powers of t are ignored, one solution to this equation is:
u = e

and

d = e

(13)

These are values of u/d proposed by Cox, Ross, Rubinstein (1979) for matching volatility
Risk-neutral analysis
The variable p is given by Eq. (6) as:
p=

ad
ud

where

a = ert

(14)

The expected stock price at the end of the time step is S0 ert
The variance of the stock price return in the risk-neutral world is:
pu2 + (1 p)d2 [pu + (1 p)d]2 = [ert (u + d) ud e2rt ]
Substituting u/d from Eq. (13), this equals 2 t when terms in t2 are ignored
When we move from the real world to the risk-neutral world the expected return on the stock changes,
but its volatility remains the same (at least in the limit as t tends to zero)
Girsanovs theorem
When we move from a world with one set of risk preferences to a world with another set of risk
preferences, the expected growth rates in variables change, but their volatilities remain the same
Moving from one set of risk preferences to another is sometimes referred to as changing the measure
Real-world measure: aka, P -measure. Risk-neutral world measure: aka, Q-measure
Increasing the number of steps
When binomial trees are used in practice, option life is typically divided into 30+ time steps
In each time step there is a binomial stock price movement
With 30 time steps, 31 terminal stock prices and 230 109 stock price paths implicitly considered
Options on other assets
We can construct and use binomial trees for options on indices, currencies, and futures contracts in
exactly the same way as for options on stocks except that the equations for p change
Options on stocks paying a continuous dividend yield
Consider a stock paying a known dividend yield at rate q
The total return from dividends and capital gains in a risk-neutral world is r
The dividends provide a return of q Capital gains must provide a return of r q
60

 If stock starts at S0 , its expected value after one time step t must be S0 e(rq)t . Hence,
pS0 u + (1 p)S0 d = S0 e(rq)t

p=

e(rq)t d
ud

We match volatility by setting u = e t and d = 1/u

We can use Eqs. (13) to (14), except that we set a = e(rq)t instead of a = ert
Options on stock indices
We assume that the stocks underlying the index provided a dividend yield at rate q
Valuation of option on stock index Option on stock paying known dividend yield
Options on currencies
Foreign currency asset providing a yield at foreign risk-free rate of interest rf
Analogy with stock index: For options on a currency, use Eqs. (13)/(14) and set a = e(rrf )t
Options on futures
It costs nothing to take a long or a short position in a futures contract
In a risk-neutral world a futures price should have an expected growth rate of zero
If F0 = initial futures price, expected futures price at end of time step t is also F0 . Thus,
pF0 u + (1 p)F0 d = F0

p=

1d
ud

Therefore, we can use Eqs. (13)/(14) with a = 1

61

62

Fabozzi: Valuation of bonds with embedded options

Introduction
The complication in building a model to value bonds with embedded options and option-type derivatives
is that cash flows will depend on interest rates in the future
The first step is to move from the yield curve to a valuation lattice
Lattice holds all the info required to value certain option-like interest-rate products
First, the lattice is used to generate the CFs across the life of the security
Next, the interest rates on the lattice are used to compute the PV of those CFs
Binomial model: Lattice model where only two rates are possible in next period
Regardless of underlying assumptions, models share common restriction: Interest-rate tree generated
must produce a value for an on-the-run optionless issue consistent with current par yield curve
Output from the model must be equal to observed market price for optionless instrument
Under these conditions, the model is said to be arbitrage free
A lattice that produces an arbitrage-free valuation is said to be fair
The lattice is used for valuation only when it has been calibrated to be fair
The interest rate lattice
Fig. 1 provides an example of a binomial interest-rate tree: Consists of number of nodes/legs
Each leg represents a one-year interval over time
The methodology is the same for smaller time periods
In practice, the selection of the length of the time period is critical
The distribution of future interest rates is represented by the nodes at each point in time
Each node is labeled as N and has a subscript, a combination of Ls and Hs
Subscripts indicate whether node is lower/higher on the tree, relative to other nodes
The root of the tree is N , the only point in time at which we know the interest rate with certainty:
The one-year rate today (at N ) is the current one-year spot rate r0

r2;HH
||
|
NHH

r1;H
||
NH

r
|0
N

r3;HHL
|||
NHHL

r2;HL
||
NHL

r1;L
|
|
NL

Today

r3;HHH
|||
NHHH

Year 1

r
|0
N

r3;HLL
|||
NHLL

r2;LL
||
NLL

r1e2
||
NH
r1
|
|
NL

r3;LLL
|||
NLLL

Year 2

Year 3

Today

Figure 1: 3-yr binomial interest-rate tree

Year 1

r e6
|3||
NHHH

r2e4
||
NHH

r e4
|3||
NHHL

r2e
||
NHL

r e2
|3|
|
NHLL

r2
||
NLL

Year 2

r3
||
|
NLLL
Year 3

Figure 2: 3-yr binomial tree with 1-yr rates

We must make an assumption concerning the probability of reaching one rate at a point in time:
Here, rates at any point in time have the same probability of occurring
The probability is 50% on each leg
The interest-rate model used to construct the binomial tree assumes that the one-year rate evolves over
time based on a log-normal random walk with a known (stationary) volatility
The relationship between any two adjacent rates at a point in time is calculated via:
r1,H = r1,L e2

where = assumed volatility of 1-yr rate and t = time in years

In the second year, there are three possible values for the one-year rate:
r2,HH = r2,LL e4

and

r2,HL = r2,LL e2

and

r2,LL

This relationship between rates holds for each point in time, as shown in Fig. 2
63

 Determining the value at a node

To get a securitys value at a node, we follow the fundamental rule for valuation:
The value is the PV of the expected cash flows
Discount rate for CFs one year forward = one-year rate at node where we are valuing
Fig. 3: Illustration with 1-yr rate r at node where valuation is sought and letting:
VH
VL
C

=
=
=

The bonds value for the higher one-year rate state

The bonds value for the lower one-year rate state
Coupon payment
Bond's value in higher-rate
state one-year forward

Value of bond:
+C V +C
1 [V
V=|
|H|| + |L||]
2 1+r
1+r

One-year rate at
node where bond's
value is sought

VH + C

Cash flow in
higher-rate state

VL + C

Cash flow in
lower-rate state

V
|

Bond's value in lower-rate

state one-year forward

Figure 3: Calculating a value at a node

The future CFs include:
1. The coupon payment one year from now
2. The bonds value one year from now, both of which may be uncertain
Starting process from the last year in tree and working backwards resolves the uncertainty:
At maturity, the instruments value is known with certainty: Par value
Final coupon payment determined from coupon rate/prevailing rates to which it is indexed
This process of working backward is often referred to as recursive valuation
Using our notation, the cash flow at a node is either:
VH + C for the higher one-year rate
VL + C for the lower one-year rate
The PV of these two CFs using the one-year rate at the node r is:
VH + C
= PV for the higher one-year rate,
(1 + r )

VL + C
= PV for the lower one-year rate
(1 + r )

Then the value of the bond at the node is found as:



1 VH + C
VL + C
Value at a node =
+
2 (1 + r ) (1 + r )
Calibrating the lattice
To ensure that no-arbitrage condition holds, lattice must be calibrated to current par yield curve
Ultimately, the lattice must price optionless par bonds at par
Assume the on-the-run par yield curve for a hypothetical issuer as it appears below:
Maturity
1 year
2 years
3 years
4 years

Par Rate
3.50%
4.20%
4.70%
5.20%

Market Price
100
100
100
100

We assume that volatility is 10% and construct a two-year tree using the two-year bond with a
coupon rate of 4.2%, the par rate for a two-year security
The root rate for the tree r0 is simply the current one-year rate, 3.5%
At beginning of year 2, two possible one-year rates: The higher rate and the lower rate
64

 We already know the relationship between the two

A rate of 4.75% at NL has been chosen arbitrarily as a starting point
An iterative process determines the proper rate
The steps are described below and illustrated in Fig. 4: The goal is a rate that, when applied in
the tree, provides a value of par for the two-year 4.2% bond
98.486 = (100 + 4.2) / (1.0582)

NH

NH

NHL 100
4.2

NL

Year 1

NLL 100
4.2

NHL 100
4.2

NL

99.766
4.2
4.4448%

99.466 = (100 + 4.2) / (1.0445)

Year 2

Today

Figure 4: 1-yr rates for Y1 : 1st trial

NHH 100
4.2

98.834
4.2
5.4289%

100.00
3.5000%

100.00 = 1/2 [
(99.766 + 4.2) / (1.035)
+ (98.834 + 4.2) / (1.035)]

99.475
4.2
4.7500%

99.475 = (100 + 4.2) / (1.0475)

Today

98.834 = (100 + 4.2) / (1.0543)

98.486
4.2
5.8017%

99.691
3.5000%

99.691 = 1/2 [
(98.486 + 4.2) / (1.035)
+ (99.475 + 4.2) / (1.035)]

NHH 100
4.2

Year 1

NLL 100
4.2
Year 2

Figure 5: 1-yr rates for Y1 using 2-yr 4.2% issue

Step 1
Select a value for r1 , the lower one-year rate. In this first trial, r1 = 4.75%
Step 2
Determine the corresponding value for the higher one-year rate
This rate is related to the lower one-year rate as: r1 e2
The higher one-year rate is 5.8017% (= 4.75%e20.10 ), reported in Fig. 4 at node NH
Step 3: Compute the bond values one year from now
1. Determine bonds value 2 years from now: Since we are using a two-year bond, the bonds
value is its maturity value ($100) plus its final coupon payment ($4.2) It is $104.2
2. Calculate VH : Cash flows are known, the appropriate discount rate is the higher one-year rate
The present value is $98.486 (= $104.2/1.058017)
3. Calculate VL : CFs known, and discount rate assumed for lower one-year rate is 4.75%
The present value is $99.475 (= $104.2/1.0475)
Step 4: Calculate V
1. Add the coupon to both VH and VL to get the cash flow at NH and NL
We have $102.686 for the higher rate and $103.675 for the lower rate
2. Calculate V : The one-year rate is 3.50%
V = $99.691 = 1/2($102.686/1.035 + $103.675/1.035)
Step 5
Compare the value in step 4 to the bonds market value
If the two values are the same, then the r1 used in this trial is the one we seek
If the value found in step 4 is not equal to the market value of the bond, this means that the
value r1 is not the one-year rate that is consistent with the current yield curve
In this case, the five steps are repeated with a different value for r1
The correct rate for r1 in this example is 4.4448%
The corresponding binomial tree is shown in Fig. 5
The value at the root is equal to the market value of the two-year issue (par)
We can grow this tree for one more year by determining r2
Now we will use the three-year on-the-run issue, the 4.7% coupon bond, to get r2
Same 5 steps in iterative process to find the 1-yr rates in the tree two years from now
Our objective is now to find the value of r2 that produces a bond value of $100
Note that the two rates one year from now of 4.4448% (the lower rate) and 5.4289% (the higher
rate) do not change: These are the fair rates for the tree one-year forward
65

Using the lattice for valuation

To use lattice for valuation, consider a 6.5% option-free bond with 4 years remaining to maturity
Since this bond is option-free, it is not necessary to use the lattice model to value it
All that is necessary to obtain an arbitrage-free value for this bond is to discount the cash flows
using the spot rates obtained from bootstrapping the yield curve shown in the table above
The spot rates are: 1-year 3.5000%, 2-year 4.2147%, 3-year 4.7345%, 4-year 5.2707%
Discounting the 6.5% 4-yr option-free bond (par = $100) at spot rates above Bond value = $104.64
Computed value
Coupon
Short-term rate

100
NHHHH 6.5

NHH
N
104.64
3.5000%

Today

100.23
6.5
NH 5.4289%
103.38
6.5
NL
4.4448%

Year 1

NHL

NLL

NHHH

97.529
6.5
9.1987%

NHHL

99.041
6.5
7.5312%

NHLL

100.31
6.5
6.1660%

NLLL

101.38
6.5
5.0483%

97.925
6.5
7.0053%
100.42
6.5
5.7354%
102.53
6.5
4.6958%

Year 2

Year 3

100
NHHHL 6.5
100
NHHLL 6.5
100
NHLLL 6.5

NLLLL

100
6.5

Year 4

Figure 6: Valuing an option-free bond with four years to maturity and 6.5% coupon
Fig. 6 shows the various values in the discounting process
Root of the tree shows the bond value of $104.64 = value found by discounting at spot rates
The lattice is consistent with the valuation of an option-free bond when using spot rates
Fixed-coupon bonds with embedded options
Embedded option requires adjustment to CFs on the tree depending on option structure:
Decision whether to call/put must be made at nodes where option is eligible for exercise
Valuing a callable bond
In the case of a call option, the call will be made when the PV of the future cash flows is greater
than the call price at the node where the decision to exercise is being made:
Vt = min[Call price, P V (future CF)]
Vt represents the PV of future cash flows at the node
This operation is performed at each node where the bond is eligible for call
Fig. 7 shows that two values are now present at each node of the binomial tree:
Discounting explained earlier is used to calculate 1st of the two values at each node
The second value is the value based on whether the issue will be called
Again, the issuer calls the issue if the PV of future CFs exceeds the call price
This second value is incorporated into the subsequent calculations
Value of call option: [Value of an optionless bond] [value of a callable bond]
Here, the value of the option-free bond is $104.64. The value of the callable bond is $102.899
Hence the value of the call option is $1.744 (= 104.634 102.899)
Valuing a putable bond
Putable bond: Bondholder has the right to force issuer to pay off the bond prior to maturity
The analysis of the putable bond follows closely that of the callable bond
If the PV of the future CFs is less than the put price (i.e. par), then the bond will be put:
Vt = max[Put price, P V (future CFs)]
66

Computed value
Call price / Computed value
Coupon
Short-term rate

N
102.90
3.5000%

100.03
100.00
NH
6.5
5.4289%
101.97
100.00
NL
6.5
4.4448%

NHH

NHL

Year 1

100.27
100.00
6.5
5.7354%

NHHH

NHHL

99.041
99.041
6.5
7.5312%

NHLL
NLL

Today

97.925
97.925
6.5
7.0053%

97.529
97.529
6.5
9.1987%

101.72
100.00
6.5
4.6958%

Year 2

NLLL

100.31
100.00
6.5
6.1660%
101.38
100.00
6.5
5.0483%

Year 3

100
NHHHH 6.5
100
NHHHL 6.5
100
NHHLL 6.5
100
NHLLL 6.5

NLLLL

100
6.5

Year 4

Figure 7: Valuing a callable bond

The value of the putable bond is greater than the value of the corresponding option-free bond
Value of put option: [Value of the putable bond] [value of corresponding option-free bond]
Suppose that a bond is both putable and callable
Adjust the value at each node to reflect whether the issue would be put or called
Specifically, at each node, two decisions about exercising option must be made:
If it is called, the value at the node is replaced by the call price
The valuation procedure then continues using the call price at that node
If call not exercised, determine whether the put option will be exercised
If put exercised, then put price is substituted at that node and used subsequently
Valuation of more exotic structures
Valuing a step-up callable note
Callable instruments whose coupon rate is increased (stepped up) at designated times
Single step-up callable note: When coupon rate increased only once over securitys life
Multiple step-up callable note: Coupon increased more than one time
Valuation using lattice similar to that for valuing a callable bond except that the CFs are altered
at each node to reflect the coupon characteristics of a step-up note
Value of embedded call option: [Optionless step-up note value] [step-up callable note value]
Valuing a range note
A range note is a security that pays the reference rate only if the rate falls within a band:
If reference rate falls outside the band, whether lower/upper boundary, no coupon paid
Typically, the band increases over time
The tree is modified to reflect the fact that either the one-year reference rate is paid, or nothing,
depending on the interest rate at a given node for the calibrated tree
Using recursive method, we work back through the tree to current value of range note
Extensions
Option-adjusted spread
Market transactions determine actual price of fixed income instrument, not lattice calculations
If markets provide meaningful price (function of liquidity), this price can be translated into an
alternative measure of value: The option-adjusted spread (OAS)
The OAS for a security is the fixed spread (measured in basis points) over benchmark rates that
equates the output from valuation process with actual market price of the security
For an optionless security, the calculation of OAS is a relatively simple, iterative process
Process much more analytically challenging with the added complexity of optionality
And just as the value of the option is volatility-dependent, the OAS for a fixed income security
with embedded options or an option-like interest-rate product is volatility-dependent
67

 Suppose info from market that price of callable bond from Fig. 7 is actually $102.218
We need the OAS that equates the value from the lattice with the market price
Since the market price is lower than the valuation, the OAS is a positive spread to the rates
in the exhibit, rates that we assume to be benchmark rates
The solution here is 35 bp: Observed market value ($102.218) = value of callable bond calculated in Fig. 7 after adding 35 bp to each rate, at each node of original calibrated lattice
Effective duration and effective convexity
Duration/convexity provide measure of interest-rate risk inherent in fixed income security
We rely on the lattice model to calculate the effective duration and effective convexity of a bond
with an embedded option and other option-like securities
The formulas for these two risk measures are given by:
Effective duration =

V V+
2V0 r

and

Effective convexity =

V+ + V 2V0
2V0 (r)2

Where V /V+ are values derived following parallel shift in yield curve down/up, by fixed spread
The model adjusts for the changes in the value of the embedded call option that result from the
shift in the curve in the calculation of V and V+
Note that the calculations must account for the OAS of the security
Steps for the proper calculation of V+ (calculation for V is analogous):
Step 1: Given the market price of the issue, calculate its OAS
Step 2: Shift the on-the-run yield curve up by a small number of basis points r
Step 3: Construct a binomial interest-rate tree based on the new yield curve from step 2
Step 4: Shift the binomial interest-rate tree by the OAS to obtain an adjusted tree
The calculation of the effective duration and convexity assumes a constant OAS
Step 5: Use the adjusted tree in step 4 to determine the value of the bond V+
To determine the value of V , the same five steps are followed except that in step 2, the on-the-run
yield curve is shifted down by a small number of basis points r

68

Hull - Ch. 12: Wiener processes and It

os lemma
Stochastic processes
Variable whose value changes over time in uncertain way is said to follow a stochastic process
Stochastic processes can be classified as discrete time or continuous time:
Discrete-time stochastic process: Variable can change only at certain fixed points in time
Continuous-time stochastic process: Changes can take place at any time
Stochastic processes can also be classified as continuous variable or discrete variable:
In a continuous-variable process, the underlying variable can take any value within a range
In a discrete-variable process, only certain discrete values are possible
The Markov property
Markov process: Stochastic process where only the present value of a variable is relevant for the future:
Past history of variable and the way that present has emerged from past are irrelevant
Stock prices are usually assumed to follow a Markov process
The Markov property implies that the probability distribution of the price at any particular future time
is not dependent on the particular path followed by the price in the past
The Markov property of stock prices is consistent with the weak form of market efficiency:
This states that stock present price impounds all info contained in past prices
Competition in the marketplace tends to ensure that weak-form market efficiency holds
Due to the many investors watching closely the stock market, trying to make a profit from it leads
to a situation where a stock price, at any given time, reflects the information in past prices
Continuous-time stochastic processes
Consider a variable that follows a Markov stochastic process:
Suppose that its current value is 10 and that the change in its value during 1 year is (0, 1), where
(m, v) denotes a Normal probability distribution with mean m and variance v
Change in 2 years = sum of two normal distributions, each of which has m = 0, v = 1.0:
Because the variable is Markov, the two probability distributions are independent
When we add two independent normal distributions, the result is a normal distribution where
the mean is the sum of the means and the variance is the sum of the variances
The change in the variable over 2 years has the distribution (0, 2)
More generally, change during any time period of length T distributed according to (0, T )
Wiener process
Type of Markov process with mean change = 0 and variance rate = 1.0/year (aka Brownian motion)
Formally, a variable z follows a Wiener process if it has the following two properties:
Property 1: The change z during a small period of time t is:

z =  t
where  has a standardized Normal distribution (0, 1)
(1)
Property 2: The values of z for any two different short intervals of time t are independent
It follows from the first property that z itself has a normal distribution with:
mean of z
standard deviation of z

=
=

The second property implies that z follows a Markov process

Consider the change in the value of z during a relatively long period of time T :
z(T ) z(0) is the sum of changes in z in N small time intervals t = T /N :
z(T ) z(0) =

N
X

i t

where the i (i = 1, , N ) are distributed (0, 1)

i=1

From the second property of Wiener processes, the i are independent of each other
From Eq. (2), z(T ) z(0) is normally distributed, with:
mean of [z(T ) z(0)]
standard deviation of [z(T ) z(0)]

69

=
=

(2)

 Our uncertainty about the value of the variable at a certain time in the future, as measured by its
standard deviation, increases as the square root of how far
we are looking ahead
Two intriguing properties of Wiener processes related to t  t for small t:
1. The expected length of the path followed by z in any time interval is infinite
2. The expected number of times z equals any particular value in any time interval is infinite
Generalized Wiener process
Drift rate: The mean change per unit time for a stochastic process
Variance rate: The variance per unit time
A generalized Wiener process for a variable x can be defined in terms of dz as:
dx = a dt + b dz

with a and b constants

(3)

The a dt term implies that x has an expected drift rate of a per unit of time:
Without the b dz term, the equation is dx = a dt, which implies that: x = x0 + at
In a period of time of length T , the variable x increases by an amount aT
RHS of Eq. (3): b dz term adding noise/variability to path followed by x
Amount of noise/variability is b [Wiener process] and has a std. dev.of b
In a small time interval t, the change x in the value of x is: x = at + b t
Thus, x has a normal distribution with:
mean of x
standard deviation of x
variance of x

=
=
=

at

b t
b2 t

The change in the value of x in any time interval T is normally distributed with:
mean of change in x
standard deviation of change in x
variance of change in x

=
=
=

aT

b T
b2 T

Thus, the generalized Wiener process in Eq. (3) has expected drift rate a and variance rate b2
It
o process
Generalized Wiener process where a and b are functions of x and t
dx = a(x, t)dt + b(x, t)dz

(4)

Both expected drift rate/variance rate of an Ito process are liable to change over time:

Between [t, t + t], variable changes from x to x + x, with: x = a(x, t)t + b(x, t) t
This relationship assumes that the drift and variance rate of x remain constant, equal to a(x, t)
and b(x, t)2 , during the small time interval between t and t + t
The process for a stock price
Key aspect of stock prices: Expected % return required by investors is independent of stocks price
The assumption of constant expected drift rate is inappropriate and needs to be replaced by the
assumption that the expected return (i.e., expected drift divided by the stock price) is constant
If S = stock price at t, then expected drift rate in S should be S for some constant
In a short interval of time t, the expected increase in S is St
The parameter is the expected rate of return on the stock
If the volatility of the stock price is always zero, then this model implies that:
S = St

ST = S0 eT

(5)

Eq. (5): When variance rate = 0, stock grows at continuously compounded rate of per unit of time
In practice, a stock price does exhibit volatility
Assumption: Variability of % return during t is the same regardless of stock price
Std. dev. of change during t should be stock price and the model becomes:
dS = Sdt + Sdz

dS
= dt + dz
S
70

(6)

 Eq. (6) is the most widely used model of stock price behavior
The variable is the volatility of the stock price
The variable is its expected rate of return
Model Limiting case of random walk (binomial trees in Ch. 11) as t 0
Discrete-time model
This model of stock price behavior is a geometric Brownian motion. Discrete-time version:

S
= t +  t
S

(7)

The LHS of Eq. (7) is the return provided by the stock in a short period of time t
The term t
is the expected value of this return
The term  t is the stochastic component of the return
The variance ofthe stochastic component (and of the whole return) is 2 t
is such that t is the standard deviation of the return in a short time period
t
Eq. (7) shows that S/S is normally distributed with mean t and std. dev. t:
S
(t, 2 t)
S

(8)

Monte Carlo simulation

Procedure for sampling random outcomes for a stochastic process
Path of stock price simulated by sampling repeatedly for  from (0, 1)/substituting into Eq. (7)
Because simulated process is Markov, samples for  should be independent of each other
Repeating simulations, complete distribution of stock price at end of time period is obtained
The parameters
The parameter is the expected return (annualized) earned by an investor in a short period of time
should depend on the risk of the return from the stock (higher risk higher )
More precisely, depends on that part of risk that cannot be diversified away by investor
also depends on level of interest rates (higher interest rates higher )
Fortunately, we do not have to concern ourselves with the determinants of in any detail because
the value of a derivative dependent on a stock is, in general, independent of
(stock price volatility) is by contrast critically important to determine many derivatives values
Typical values of for a stock are in the range 0.15 to 0.60

Std. dev. of proportional change in stock price in a small interval t is t

Std. dev. of proportional change in stock price over long period T is T (rough approx.)
Volatility can be approximated as the std. dev. of the change in stock price in 1 year
It
os lemma
Suppose that the value of a variable x follows the Ito process:
dx = a(x, t)dt + b(x, t)dz

where dz Wiener process, a and b functions of x and t

(9)

The variable x has a drift rate of a and a variance rate of b2

Itos lemma shows that a function G of x and t follows the process:

dG =


G
G 1 2 G 2
G
a+
+
b dt +
bdz
2
x
t
2 x
x

(10)

dz is the same Wiener process as in Eq. (9) G also follows an Ito process, with drift rate of:
G
G 1 2 G 2
a+
+
b
x
t
2 2x


and a variance rate of:

71

G
x

2

b2

 Earlier, we argued that dS = Sdt + Sdz with and constant, is a reasonable model of stock price
movements. From It
os lemma, it follows that the process followed by a function G of S and t is:


G
G
G 1 2 G 2 2
dG =
S +
+
S dt +
Sdz
(11)
2
S
t
2 S
S
Where both S and G are affected by the same underlying source of uncertainty dz
Application to forward contracts
Consider a forward contract on a non-dividend-paying stock
F0 = S0 erT where F0 is the forward price at time zero, S0 is the spot price at time zero, T is
the time to maturity, and r is the risk-free rate of interest (constant for all maturities)
Define F as the forward price at a general time t, and S as the stock price at time t, with
t < T . The relationship between F and S is given by:
F = Ser(T t)

(12)

Assume process for S is given by Eq. (6). Itos lemma determines process for F . From Eq. (12):
F
= er(T t) ,
S

2F
= 0,
2S

F
= rSer(T t)
t

From Eq. (11), the process for F is given by:

h
i
dF = er(T t) S rSer(T t) dt + er(T t) Sdz
Substituting F for Ser(T t) gives:
dF = ( r)F dt + F dz

(13)

Like S, the forward price F follows geometric Brownian motion

It has an expected growth rate of r rather than
The growth rate in F is the excess return of S over the risk-free rate
The lognormal property
Use Itos lemma to derive process followed by ln S when S follows Eq. (6). We define G = ln S. Since:
G
1
= ,
S
S

2G
1
= 2,
2
S
S

G
= 0,
t

It follows from Eq. (11) that the process followed by G is:



2
dG =
dt + dz
2

(14)

Since and are constant, G = ln S follows a generalized Wiener process:

It has constant drift rate 2 /2 and constant variance rate 2
Change in ln S from 0 to T normally distributed, with mean ( 2 /2)T and variance 2 T . Hence,







2
2
2
2
ln ST ln S0
T, T
ln ST ln S0 +
T, T
(15)
2
2
Eq. (15) shows that ln ST is normally distributed
A stocks price at time T , given itsprice today, is lognormally distributed
Std. dev. of log of stock price is T ( square root of how far ahead we look)

72

Hull - Ch. 13: The Black-Scholes-Merton model

Lognormal property of stock prices
The model of stock price behavior used by Black, Scholes, and Merton assumes that percentage changes
in the stock price in a short period of time are normally distributed. Define:
:
:

Expected return on stock per year

Volatility of the stock price per year

The mean of the return in time t is t and the std. dev. of the return is t so that:
S
(t, 2 t)
S

(1)

The model implies that ln ST is normally distributed, so that ST has a lognormal distribution:







2
2
ST
2
2

T, T
ln ST ln S0 +
T, T
ln
S0
2
2

The mean of ln ST is ln S0 + ( 2 /2)T and the standard deviation is T

Properties of the (lognormal) stock price distribution
A variable that has a lognormal distribution can take any value between zero and infinity
95% chance that normally distributed variable is within 1.96 std. dev. of its mean
The expected value E(ST ) of ST is given by:
E(ST ) = S0 eT

(2)

(3)

The variance V ar(ST ) of the stock price ST is given by:

V ar(ST ) = S02 e2T (e

2T

1)

(4)

The distribution of the rate of return

Define the continuously compounded rate of return per annum realized between 0 and T as x. Then:


1 ST
2 2
xT
ST = S0 e
x = ln
(5)
From Eq. (2), it follows:
x ,
T
S0
2 T

Thus, the continuously compounded rate of return is Normal, mean = 2 /2 and std. dev. = / T
As T increases, the standard deviation of x declines
More certain about avg. return over 20 years than about return in any one year
The expected return
Expected return required by investors depends on riskiness of stock and on level of interest rates:
The higher the risk, the higher the expected return
The higher the level of interest rates, the higher the expected return required
Fortunately, we do not have to concern ourselves with the determinants of in any detail:
Value of stock option, when expressed in terms of underlying stock, does not depend on
Reason why expected continuously compounded return x 6= is subtle, but important:
Under our assumptions for stock price behavior, the average of the returns on the stock in each
interval of length t is close to t is close to arithmetic mean of the Si /Si
However, the expected return over the whole period covered by the data E(x), expressed with a
compounding interval of t is close to 2 /2
Reason: Geometric mean of set of numbers (not all the same) < arithmetic mean
Volatility
From Eq. (5), the volatility of a stock price can be defined as the standard deviation of the return
provided by the stock in 1 year when the return is expressed using continuous compounding
73


t std. dev. of the % change in stock price during t
Uncertainty about a future stock price, as measured by its standard deviation, increases - at least
approximately - with the square root of how far ahead we are looking
Estimating volatility from historical data
Define:
n + 1: Number of observations
Si :
:
ui :

Stock price at end of i-th interval, with i = 0, 1, , n

Length of time interval in years
ln(Si /Si1 ) for i = 1, , n

The usual estimate s of the standard deviation of the ui is given by:

v
v
!2
u
u
n
n
n
u 1 X
X
u 1 X
1
t
s=t
(ui u
)2 =
u2i
ui
n1
n1
n(n 1)
i=1

i=1

where u
= mean of ui s

i=1

From Eq. (2), the standard deviation of the ui is

The variable s is an estimate of

itself estimated as
, where:
= s/ , with std. error of estimate being
/ 2n
Choosing an appropriate value for n is not easy:
More data generally lead to more accuracy, but does change over time and data that are too
old may not be relevant for predicting the future volatility
Compromise: Use closing prices from daily data over most recent 90-180 days
Alternatively, n set equal to # of days to which volatility is to be applied
The foregoing analysis can be adapted to accommodate dividend-paying stocks:
Return ui during a time interval that includes an ex-dividend day:
ui = ln

Si + D
Si1

where D is the amount of dividend

The return in other time intervals is still: ui = ln Si /Si1

However, as tax factors play a part in determining returns around an ex-dividend date, it is
probably best to discard altogether data for intervals that include an ex-dividend date
Trading days vs. calendar days
Important issue: Should time be measured in calendar/trading days when volatility is estimated/used?
Volatility is much higher when the exchange is open for trading than when it is closed
As a result, practitioners tend to ignore days when the exchange is closed when estimating
volatility from historical data and when calculating the life of an option
The volatility per annum is calculated from the volatility per trading day using the formula:
p
Volatility per annum = Volatility per trading day # trading days per annum
The life of an option is also usually measured using trading days rather than calendar days:
The number of trading days in a year is usually assumed to be 252 for stocks
The life of an option is calculated as T years, where:
T =

Number of trading days until option maturity

252

It is natural to assume that the volatility of a stock is caused by new info reaching the market:
This is not true, volatility is to a large extent caused by trading itself
The idea underlying the Black-Scholes-Merton differential equation
Equation that must be satisfied by price of any derivative dependent on a non-dividend-paying stock
Derivation: Set up a riskless portfolio consisting of positions in derivative/stock
If no arbitrage, the return from the portfolio must be the risk-free interest rate r
74

 The reason a riskless portfolio can be set up is that the stock price and the derivative price are both
affected by the same underlying source of uncertainty: Stock price movements
In any short period, derivative price perfectly correlated with price of underlying stock
When an appropriate portfolio of the stock and the derivative is established, the gain or loss from
the stock position always offsets the gain or loss from the derivative position
The portfolio overall value at end of the short period of time is known with certainty
Example
Suppose that at a particular time, relationship between small change S in stock price and
resultant small change c in price of a European call option is: c = 0.4S
The riskless portfolio would consist of:
1. A long position in 0.4 shares
2. A short position in one call option
Suppose that stock price increases by 10 cents: Option price increases by 4c and the 400.10 =
$4 gain on the shares = the 100 0.04 = $4 loss on short option position
Important difference between the Black-Scholes-Merton analysis and binomial model in Ch. 11
In Black-Scholes-Merton, position in stock/derivative is riskless for only a very short period
To remain riskless, it must be adjusted, or rebalanced, frequently
Assumptions to derive the Black-Scholes-Merton differential equation
1. The stock price follows the process developed in Ch. 12 with and constant
2. The short selling of securities with full use of proceeds is permitted
3. There are no transactions costs or taxes. All securities are perfectly divisible
4. There are no dividends during the life of the derivative
5. There are no riskless arbitrage opportunities
6. Security trading is continuous
7. The risk-free rate of interest r is constant and the same for all maturities
Derivation of the Black-Scholes-Merton (BSM) differential equation
The stock price process we are assuming is:
dS = Sdt + Sdz

(6)

Suppose f = price of derivative on S. f must be a function of S and t. Hence, from Eq. (12.11):


f
f
1 2f 2 2
f
df =
S +
+
S dt +
Sdz
2
S
t
2 S
S

(7)

The discrete versions of Eqs. (6) and (7) are:

S = St + Sz


f
f
1 2f 2 2
f
f =
S +
+
S t +
Sz
2
S
t
2 S
S

(8)
(9)

From Itos lemma, the Wiener processes underlying f and S are the same
A portfolio of stock/derivative can be constructed so that the Wiener process is eliminated
Such portfolio is:
1: derivative
+f /S:

shares

Holder is short one derivative and long f /S shares. Define = portfolio value:
= f +

f
S
S

(10)

The change in the value of the portfolio in the time interval t is:


f
f
1 2f 2 2
= f +
S =

S t
S
t
2 S 2
75

(11)

 Because z not involved, the portfolio must be riskless during time t. Hence:
= rt

(12)

Substituting Eqs. (10) and (11) into (12), we obtain:





f
f
1 2f 2 2
f
f
1 2 2 2f
t
=
r
f

S
S
t

+
rS
+
S
= rf
t
2 S 2
S
t
S 2
S 2

(13)

Black-Scholes-Merton differential equation

Eq. (13) is the Black-Scholes-Merton differential equation
Many solutions, corresponding to different derivatives defined with S as underlying variable
Particular derivative obtained when equation solved depends on the boundary conditions
European call option: The key boundary condition is: f = max(S K, 0) when t = T
European put option: The key boundary condition is: f = max(K S, 0) when t = T
The portfolio used in the derivation of Eq. (13) is not permanently riskless:
As S and t change, f /S also changes
To keep portfolio riskless, frequently rebalance relative proportions of derivative/stock
Example
Forward contract on non-dividend-paying stock = derivative on stock Must satisfy Eq. (13)
The value of the forward contract f at a general time t is in terms of the stock price S at this time:
f = S Ker(T t) , where K is the delivery price. Hence,
f
= rKer(T t) ,
t

f
= 1,
S

2f
=0
S 2

When substituted into LHS of Eq. (13), we get: rKer(T t) + rS = rf Eq. (13) satisfied
The price of tradeable derivatives
Any function f (S, t) solution of Eq. (13) is the theoretical price of a derivative
If a derivative with that price existed, it would not create any arbitrage opportunities
Conversely, if a function f (S, t) does not satisfy the differential Eq. (13), it cannot be the price of
a derivative without creating arbitrage opportunities for traders
Risk-neutral valuation
BSM Eq. (13) does not involve any variables affected by the risk preferences of investors
Variables: Stock price, time, stock volatility, and risk-free rate, all independent of risk preferences
The Black-Scholes-Merton differential equation would not be independent of risk preferences if it
involved the expected return on the stock because does depend on risk preferences
If risk preferences do not enter the equation, they cannot affect its solution
Any set of risk preferences can be used when evaluating f (e.g., risk neutral world)
Consider derivative with payoff at one particular time. Can be valued using risk-neutral valuation by:
1. Assume the expected return from underlying asset is the risk-free interest rate r (i.e = r)
2. Calculate the expected payoff from the derivative
3. Discount the expected payoff at the risk-free interest rate
Risk-neutral valuation is a device for obtaining solutions to the BSM equation
Solutions obtained are valid in all worlds, not just when investors are risk neutral
When we move from a risk-neutral world to a risk-averse world, two things happen:
1. The expected growth rate in the stock price changes
2. The discount rate that must be used for any payoffs from the derivative changes
It happens that these two changes always offset each other exactly
Application to forward contracts on a stock
Consider a long forward contract that matures at time T with delivery price K. The value of the
contract at maturity is ST K where ST is the stock price at time T
From the risk-neutral valuation argument, the value f of the forward contract at time 0 is its
expected value at time T in a risk-neutral world discounted at the risk-free rate of interest
76

 T K) where E
denotes the expected value in a risk-neutral world
f = erT E(S
Since K is a constant, this equation becomes:
T ) KerT
f = erT E(S

(14)

T ) = S0 erT and f = S0 KerT

Expected return = r in risk-neutral world From Eq. (3), E(S
Black-Scholes pricing formulas
The Black-Scholes formulas for the prices at time 0 of a European call option on a non-dividend-paying
stock and a European put option on a non-dividend-paying stock are:
c = S0 N (d1 ) KerT N (d2 )

(15)

p = KerT N (d2 ) S0 N (d1 )

(16)

d1 and d2 are defined as:

d1 =

ln(S0 /K) + (r + 2 /2)T

and

d2 =

ln(S0 /K) + (r 2 /2)T

= d1 T
T

N (x) = cumulative probability distribution for standardized Normal distribution (0, 1)

Deriving Black-Scholes formulas
One way is by solving the differential Eq. (13) subject to the boundary condition
Another approach is to use risk-neutral valuation
To provide an interpretation of the terms in Eq. (15), we note that it can be written:
c = erT [S0 N (d1 )erT KN (d2 )]

(17)

The expression N (d2 ) is the probability that the option will be exercised in a risk-neutral world,
so that KN (d2 ) is the strike price times the probability that the strike price will be paid
S0 N (d1 )erT = expected value (risk-neutral) of variable equal to ST if ST > K and 0 otherwise
Never optimal to exercise early an American call option on a non-dividend-paying stock
Eq. (15) is the value of an American call option on a non-dividend-paying stock
No analytic formula for value of American put on a non-dividend-paying stock
Black-Scholes in practice
In practice, interest rate r is set equal to the zero-coupon risk-free rate for maturity T
Time measured as [# of trading days left in option life] [# of trading days in 1 year]
Properties of the Black-Scholes formulas
When the stock price S0 becomes very large, a call option is almost certain to be exercised
It becomes very similar to a forward contract with delivery price K
From Eq. (5.5), we expect the call price to be S0 KerT
This is the call price given by Eq. (15) because when S0 becomes very large, both d1 and d2
become very large, and N (d1 ) and N (d2 ) become close to 1.0
When the stock price becomes very large, the price of a European put p approaches zero
Consistent with Eq. (16) because N (d1 ) and N (d2 ) both 0 in this case
Consider next what happens when the volatility approaches zero:
Because the stock is virtually riskless, its price will grow at rate r to S0 erT at time T and the
payoff from a call option is: max(S0 erT K, 0)
Discounting at rate r, call value today = erT max(S0 erT K, 0) = max(S0 KerT , 0)
Consider first the case where S0 > KerT . This implies that ln(S0 /K) + rT > 0:
As 0, d1 , d2 +, and N (d1 ), N (d2 ) 1.0 Eq. (15) becomes: c = S0 KerT
When S0 < KerT , it follows that ln(S0 /K) + rT < 0:
As 0, d1 , d2 , and N (d1 ), N (d2 ) 0 Eq. (15) gives: c = 0
The call price is therefore always max(S0 KerT , 0) as tends to zero
Similarly, the put price is always max(KerT S0 , 0) as tends to zero
77

Cumulative normal distribution function

In calculating the cumulative normal distribution function N (x), we can use:
(i) Tables for N (x), (ii) The NORMSDIST function in Excel, and (iii) A polynomial approximation
Warrants and employee stock options
Warrants and employee stock options are different from regular call options in that exercise leads to the
company issuing more shares and then selling them to the option holder for the strike price
As the strike price is less than the market price, this dilutes the interest of existing shareholders
How should potential dilution affect the way we value outstanding warrants/employee stock options?
The answer is that it should not! Assuming markets are efficient, the stock price will reflect
potential dilution from all outstanding warrants and employee stock options
Consider company contemplating a new issue of warrants (or employee stock options)
Company interested in the cost of issue assuming there are no compensating benefits
Assume that the company has N shares worth S0 each and the number of new options contemplated
is M , with each option giving the holder the right to buy one share for K
Company value today = N S0 . This value does not change as result of warrant issue
Suppose that without warrant issue, share price will be ST at warrants maturity With/without
warrant issue, the total value of the equity and the warrants at T will be N ST
If warrants are exercised, cash inflow from strike price increasing this to N ST + M K. This value
+M K
is distributed among N + M shares Share price immediately after exercise becomes: N SNT+M
Therefore the payoff to an option holder if the option is exercised is:
N
N ST + M K
K =
(ST K)
N +M
N +M
Thus, the value of each option is the value of N/(N + M ) regular calls c on the companys stock
The total cost of the options is M times this, i.e., M N c/(N + M )
Example
A company with 1 million shares worth $40 each is considering issuing 200,000 warrants each giving
the holder the right to buy one share with a strike price of $60 in 5 years
Interest rate = 3%, volatility = 30% per annum. The company pays no dividends
From Eq. (15), the value of a 5-year European call option on the stock is $7.04
In this case, N = 1, 000, 000 and M = 200, 000, so that the value of each warrant is:
1, 000, 000
7.04 = $5.87
1, 000, 000 + 200, 000
The total cost of the warrant issue is 200, 000 5.87 = $1.17 million
If market perceives no benefits from warrant issue, we expect stock price to decline by $1.17
Implied volatilities
The one parameter in BSM that cannot be directly observed is the volatility of the stock price
It can be estimated from a history of the stock price
In practice, traders usually work with what are known as implied volatilities
These are the volatilities implied by option prices observed in the market
Suppose that a European call is worth 1.875 when S0 = 21, K = 20, r = 0.1, and T = 0.25
The implied volatility is the value of that, when substituted into Eq. (15), gives c = 1.875
An iterative search procedure can be used to find the implied
Similar procedure with binomial trees used to find implied volatilities for American options
Implied volatilities are used to monitor the markets opinion about the volatility of a particular stock
Whereas historical volatilities are backward looking, implied volatilities are forward looking
Traders often quote the implied volatility of an option rather than its price
This is convenient because the implied volatility tends to be less variable than the option price

78

 The VIX index

CBOE publishes indices of implied volatility. The most popular index (SPX VIX) is an index of
the implied volatility of 30-day options on the S&P 500 calculated from wide range of calls/puts
Trade involving futures/options on S&P 500 is a bet on both future level of S&P 500/its volatility.
By contrast, a futures/options contract on VIX is a bet only on volatility
Example
A trader buys an April futures contract on VIX when futures price is 18.5 ( 30-day S&P 500
volatility of 18.5%) and closes out contract when futures price is 19.3
The trader makes a gain of $800 (one contract is on 1,000 the index)
Dividends
Modified Black-Scholes model to take account of dividends
Assume amount/timing of dividends during option life can be predicted with certainty
For long-life options, assume that the dividend yield rather cash dividend payments is known
The date on which the dividend is paid should be assumed to be the ex-dividend date. On this
date the stock price declines by the amount of the dividend
For tax reasons, stock price may go down by somewhat less than the cash amount of dividend
We need to interpret the word dividend in the context of option pricing as the reduction in
the stock price on the ex-dividend date caused by the dividend
European options
European options analyzed by assuming that stock price = sum of 2 components:
(i) Riskless component for known dividends during option life and, (ii) Risky component
The riskless component, at any given time, is the PV of all the dividends during the life of the
option discounted from the ex-dividend dates to the present at the risk-free rate
By time option matures, dividends have been paid and riskless component no longer exists
Thus, BSM correct if S0 is equal to the risky component of the stock price and is the volatility
of the process followed by the risky component
In theory, this is 6= volatility of stochastic process followed by whole stock price
The volatility of the risky component is approximately equal to the volatility of the whole
stock price multiplied by S0 /(S0 D), where D is the PV of the dividends
However, adjustment necessary only when volatilities estimated using historical data
An implied volatility is calculated after the PV of dividends have been subtracted from
the stock price and is the volatility of the risky component
Operationally, Black-Scholes can be used provided that stock price is reduced by PV of all dividends
during option life, with risk-free discounting from ex-dividend dates
American options
Can only be optimal to exercise American call immediately before stock goes ex-dividend
Assume n ex-dividend dates anticipated, at times t1 , , tn with t1 < < tn . The dividends
corresponding to these times are D1 , , Dn
Consider the possibility of early exercise just prior to final ex-dividend date (i.e., at tn )
If the option is exercised at time tn , the investor receives S(tn ) K
If the option is not exercised, the stock price drops to S(tn ) Dn
The value of the option is then greater than: S(tn ) Dn Ker(T tn )
If S(tn ) Dn Ker(T tn ) S(tn ) K, i.e.:
h
i
Dn K 1 er(T tn )
Not optimal to exercise at tn
(18)
On the other hand, if:
h
i
Dn > K 1 er(T tn )

Optimal to exercise at tn for high value of S(tn )

(19)

The inequality in Eq. (19) will tend to be satisfied when the final ex-dividend date is fairly
close to the maturity of the option (i.e., T tn is small) and the dividend is large
79

 Consider next time tn1

If the option is exercised immediately prior to time tn1 , the investor receives S(tn1 ) K
If the option is not exercised at time tn1 , the stock price drops to S(tn1 ) Dn1 and the
earliest subsequent time at which exercise could take place is tn
Hence, lower bound to option price if not exercised at tn1 = S(tn1 ) Dn1 Ker(tn tn1 )
It follows that if S(tn1 ) Dn1 Dn Ker(tn tn1 ) S(tn1 ) K, or:
h
i
Dn1 K 1 er(tn tn1 )
Not optimal to exercise immediately prior to tn1
Similarly, for any i < n, if:
h
i
Di K 1 er(ti+1 ti )

Not optimal to exercise immediately prior to ti

(20)

The inequality in Eq. (20) is approximately equivalent to Di Kr(ti+1 ti )

Assuming that K is fairly close to current stock price, inequality is satisfied when dividend
yield on the stock is less than risk-free rate of interest (often the case)
We can conclude that, in many circumstances, the most likely time for the early exercise of an
American call is immediately before the final ex-dividend date tn
Blacks approximation
Black suggests an approximate procedure for taking account of early exercise in call options:
Calculate prices of European options that mature at T and tn , and then set the American price
equal to the greater of the two (approximation works well in most cases)
Example
Consider an American option with dividends D1 = D2 = 0.5 in two and five months, S0 = 40,
K = 40, r = 0.09, t1 = 2/12, and t2 = 5/12
Since K[1 er(t2 t1 ) ] = 40(1 e0.090.25 ) = 0.89 > 0.5, it follows from Eq. (20) that the
option should never be exercised immediately before the first ex-dividend date
Since K[1 er(T t2 ) ] = 40(1 e0.090.0833 ) = 0.30 < 0.5, it follows that, when sufficiently
deep in the money, option should be exercised immediately before 2nd ex-dividend date
Blacks approximation to value the option
PV of 1st dividend is 0.5e0.16670.09 = 0.4926 Value of option, assuming it expires just
before final ex-dividend date, can be calculated using BSM with S0 = 40 0.4926 = 39.5074,
K = 40, r = 0.09, = 0.30, and T = 0.4167. It is $3.52
PV of the two dividends is 0.5e0.16670.09 + 0.5e0.41670.09 = 0.9711 Value of option
exercised at the end of 6 months can be calculated using BSM with S0 = 400.9711 = 39.0259,
K = 40, r = 0.09, = 0.30, and T = 0.5. It is $3.67
Blacks approximation involves taking the max of these 2 options American call = $3.67
Two reasons for differences between Binomial Model (BM) and Blacks approximation (BA):
(i) The timing of the early exercise decision, and (ii) The way volatility is applied
The timing of the early exercise decision tends to make BM greater than BA
BA: Holder has to decide today whether option will be exercised after 5/6 months
BM: Decision on early exercise at 5-month point depends on stock price at that time
The way in which volatility is applied tends to make BA greater than BM
In BA, when exercise after 5m, volatility is applied to stock price less PV of 1st dividend
When exercise after 6m, volatility is applied to stock price less PV of both dividends

80

Black: Using the holes in Black-Scholes

The formula
Using the Black-Scholes formula, the value of the option increases with increases in the stocks price,
the interest rate, the time remaining until the option expires, and the stocks volatility
Except for volatility, which can be estimated several ways, we can observe all other factors
The stocks expected return doesnt appear in the formula
A higher expected return on the stock means a higher expected return on the option, but it doesnt
affect the options value for a given stock price
This feature of the formula is very general
How to improve the assumptions
The original derivation of the Black-Scholes formula involves the following assumptions:
The stocks volatility is known, and doesnt change over the life of the option
The stock price changes smoothly: It never jumps up or down a large amount in a short time
The short-term interest rate never changes
Anyone can borrow or lend as much as he wants at a single rate
An investor who sells the stock or the option short will have the use of all the proceeds of the sale
and receive any returns from investing these proceeds
There are no trading costs for either the stock or the option
An investors trades do not affect the taxes he pays
The stock pays no dividends
An investor can exercise the option only at expiration
There are no takeovers or other events that can end the options life early
Volatility changes
The volatility of a stock is not constant
Changes in volatility may impact certain options, especially far-out-of-the-money options
One measure of that uncertainty is the volatility of the volatility
Formula with volatility includes both current/expected future levels of volatility
Other measure about future volatility: Relation between future stock price/its volatility
A decline in the stock price implies a substantial increase in volatility, while an increase in the
stock price implies a substantial decrease in volatility
Cox/Ross have formulas that account for relation between future stock price/its volatility
Cox-Ross formula implies lower values for out-of-the-money call options than Black-Scholes
But putting in uncertainty about future volatility often implies higher values for these options
We cant tell how the option values will change when we put in both effects
You should buy volatility if you think volatility will rise, and sell volatility if you think it will fall
To buy volatility, buy options. To sell volatility, sell options
Instead of buying stock, you can buy calls or buy stock and sell calls
Strongest position on volatility: Add long/short position in straddles to existing position
To buy pure volatility, buy both puts/calls in ratio giving no added exposure to the stock
To sell pure volatility, sell both puts and calls in the same ratio
Jumps
Up jumps and down jumps have different effects on option values than symmetric jumps
Merton has a formula that reflects possible symmetric jumps
Compared to the Black-Scholes formula, his formula gives higher values for both in-the-money and
out-of-the-money options and lower values for at-the-money options
The differences are especially large for short-term options
Short-term options show strikingly different effects for up jumps and down jumps
An increase in the probability of an up jump will cause out-of-the-money calls to go way up in
value relative to out-of-the-money puts
An increase in the probability of a down jump will do the reverse
81

 If you expect a symmetric jump, buy short-term out-of-the-money options

Instead of stock, you can hold call options or more stock plus put options
Or you can sell at-the-money options
Instead of stock, you can hold more stock and sell call options
Pure play on symmetric jumps: Buy out-of-the-money calls/puts, and sell at-the-money calls/puts
For up jumps, use similar strategies that involve buying short-term out-of-the-money calls, or selling
short-term out-of-the-money puts, or both. For down jumps, do the opposite
Interest rate changes
If future changes in the interest rate are known, we can just replace the short-term rate with the yield
on a zero-coupon bond that matures when the option expires
When the stocks volatility is known, Merton has shown that the zero-coupon bond yield will still work
At any point in time, use the zero-coupon bond yield at that time for the short-term rate
Effects of interest rate changes on option values not nearly as great as effects of volatility changes
Better off with direct positions in fixed-income securities vs. options to play interest rate changes
Interest rate changes may affect your decisions to buy or sell options
Higher interest rates mean higher call values and lower put values
If you think interest rates will rise more than market thinks, be more inclined to buy calls, and
more inclined to buy more stocks/sell puts, as a substitute for a straight stock position
If you think interest rates will fall more than market thinks, these preferences should be reversed
Borrowing penalties
Rate at which one can borrow, even with collateral, is higher than rate at which one can lend
Also, margin requirements/restrictions put on by lenders may limit amount one can borrow
High rates and limits on borrowing may cause a general increase in call option values, since calls provide
leverage that can substitute for borrowing
The interest rates implied by option values may be higher than lending rates
If this happens and you have borrowing limits but no limits on option investments, buy calls
If you can borrow freely at lending rate, get leverage by borrowing rather than by buying calls
When implied interest rates high, conservative investors buy puts/sell calls to protect a portfolio
Fixed-income investors might even choose to buy stocks and puts, and sell calls, to create a synthetic
fixed-income position with a yield higher than market yields
Short-selling penalties
Short-selling penalties are generally even worse than borrowing penalties
An investor must go to the expense of borrowing stock if he wants to sell it short
Part of his expense involves putting up cash collateral, generally at interest below market rates
Also, investors may have to put up margin with brokers in cash, and may not receive interest
For options, the penalties tend to be much less severe
An investor need not borrow an option to sell it short
Investor loses much less interest income in selling an option short than in selling a stock short
Penalties on short selling that apply to all investors will affect option values
When even professional investors have trouble selling a stock short, we will want to include an
element in the option formula to reflect the strength of these penalties
Sometimes we approximate this by assuming an extra dividend yield on the stock, in an amount
up to the cost of maintaining a short position as part of a hedge
If you want to short a stock but face penalties/restrictions:
You can short it indirectly by holding put options
Or by taking a naked short position in call options
When facing penalties in selling short, you face rewards for lending stock to those who want to short it
In this situation, strategies involving holding stock/lending it out may dominate
E.g., you might create a position with a limited downside by holding a stock and a put on the
stock, and by lending the stock to those who want to short it

82

Trading costs
Trading costs make it hard to create an option-like payoff by trading in the underlying stock
They can also make it hard to create a stock-like payoff by trading in the option
Sometimes trading costs can increase an options value, and sometimes they can decrease it
We cant tell how trading costs affect an options value They create a band of possible values
Within this band, it will be impractical for most investors to take advantage of mispricing by selling
the option and buying the stock, or by selling the stock and buying the option
Trading costs make options useful if you want to shift exposure to the stock after it goes up/down
To shift your exposure to the market as a whole, rather than to a stock, options even more useful
It is often more costly to trade in a basket of stocks than in a single stock
But you can use index options to reduce your trading in the underlying stocks or futures
Taxes
The

The

very existence of taxes will affect option values

A hedged position giving the same return as lending may have a tax 6= tax on interest
fact that investor tax rates differ will affect values too
Tax authorities adopt a variety of rules to restrict tax arbitrage
Rules to limit interest/capital loses deductions, rules to tax gains/losses before position closed out
E.g., most US index option positions are now taxed each year, partly as short-term capital gains
and partly as long-term capital gains, whether or not the taxpayer has closed out his positions
If you pay taxes on gains and cannot deduct losses, you may want to limit the volatility of your positions
and have the freedom to control the timing of gains and losses
This will affect how you use options, and may affect option values as well
Hard to predict, though, whether it will increase or decrease option values
Investors who buy a put have a capital gain/loss at end of year, or when option expires
Investors who simulate the put by trading in underlying stock sell after decline, buy after rise
By choosing which lots of stock to buy and which lots to sell, they will be able to generate a series
of realized capital losses and unrealized gains
The tax advantages of this strategy may reduce put values for many taxable investors
Similarly, tax advantages of a simulated call reduce call values for most taxable investors

Dividends and early exercise

Dividends reduce call option values and increase put option values
Dividends make early exercise of a call more likely, and early exercise of a put less likely
Several ways to change the BS formula to account for dividends:
1. Assume dividend yield constant for all possible stock price levels/all future times
2. Assume issuer has money set aside to pay dollar dividends due before option expires
3. Assume dividend depends in a known way on stock price at each ex-dividend date
Cox, Ross and Rubinstein tree approach is more flexible than BS formula
Use dividend that the firm will pay for each possible future stock price at each future time
We can also test, at each node of the tree, whether an investor will exercise the option early
Option values reflect markets belief about stocks future dividends and likelihood of early exercise:
If dividends higher than market thinks, buy puts or sell calls, other things equal
If option holders exercise too early/late, sell options to profit from opportunities the holders create
Takeovers
If A takes over B through exchange of stock, options on Bs stock will become options on As stock
We will use As volatility rather than Bs in valuing the option
If firm A takes over firm B through a cash tender offer, there are two effects:
1. Outstanding options on B will expire early Reduced values for both puts and calls
2. Bs stock price rises through tender offer premium Increased call values/decreased put values
But when market knows of a possible tender offer from A, Bs stock price higher than normal
All these factors work together to influence option values
83

 The chance of a takeover will make an options value sometimes higher and sometimes lower
For short-term out-of-the-money call, chance of takeover increases the option value
For short-term out-of-the-money put, chance of a takeover reduces the option value
Effects of takeover probability can be dramatic for these short-term out-of-the-money options
Portfolio insurance
Any strategy where you reduce your stock positions when prices fall, and increase them when prices rise
Some use option formulas to figure how much to increase/reduce positions as prices change
However, assumptions behind BSM affect portfolio insurance strategies that dont use the formula
The higher your trading costs, the less likely you are to create synthetic options
Sometimes, futures are priced against the portfolio insurers
This makes all portfolio insurance strategies less attractive
Portfolio insurance using synthetic strategies wins when market jumps big, but without much volatility
It loses when market volatility is high, because an investor will sell after a fall, and buy after a rise
The investor loses money on each cycle
But the true cost of portfolio insurance is a factor that doesnt even affect option values:
It is the mean reversion in market: Rate at which expected return on market falls as market rises
Mean reversion is what balances supply and demand for portfolio insurance:
High mean reversion will discourage portfolio insurers because it will mean they are selling when
expected return is higher and buying when expected return is lower
For the same reason, high mean reversion will attract value investors or tactical asset allocators
who buy after a decline and sell after a rise
Value investors use indicators like P/E ratios and dividend yields to decide when to buy/sell
They act as sellers of portfolio insurance
If mean reversion were zero, more investors would want to buy portfolio insurance than to sell it:
People have a natural desire to try to limit their losses
But, on balance, there must be as many sellers as buyers of insurance
What makes this happen is a positive normal level of mean reversion
The October 1987 crash
During 1987, investors shifted toward wanting more portfolio insurance
This increased the markets mean reversion
Because of mean reversion, market rise in 87 caused sharper-than-usual fall in expected return
But investors didnt see this at first
They continued to buy, as their portfolio insurance strategies suggested
Around October 19, the full truth of what was happening hit investors:
They saw that at existing levels of market, expected return was much lower than assumed
They sold at those levels. Market fell, and expected return rose, until equilibrium restored
Mean reversion and stock volatility
Mean reversion and volatility estimates
If good estimate of volatility, stocks expected return wont affect option values
Neither will mean reversion
But mean reversion may influence your estimate of the stocks volatility
With mean reversion day-to-day volatility will be higher than month-to-month volatility, which
will be higher than year-to-year volatility
Your volatility estimates for options with several years of life should be generally lower than
your volatility estimates for options with several days or several months of life
Using your views of mean reversion for investing
If mean reversion higher than markets, buy short-term options/sell long-term options
If you think mean reversion is lower, you can do the reverse
If you are a buyer of options, you will favor short-term options when you think mean reversion is
high, and long-term options when you think it is low
84

 If you are a seller of options, you will favor long-term options when you think mean reversion is
high, and short-term options when you think its low
Effects most striking in stock index options, but also in individual stock options
Trend followers act like portfolio insurers, but they trade individual stocks rather than portfolios
When the stock rises, they buy. When it falls, they sell
They act as if the past trend in a stocks price is likely to continue
In individual stocks, as in portfolios, mean reversion should normally make implied volatilities
higher for short-term options than for long-term options

85

86

Hull - Ch. 15: Options on stock indices and currencies

Options on stock indices
Several exchanges trade options on stock indices
Some indices track movement of whole market, others based on performance of particular sector
One index option contract is on 100 times the index, index options are settled in cash
On exercise, holder of call receives (S K) 100 in cash and writer pays this amount in cash,
where S = index value at close of trading day and K = strike price
Similarly, holder of put option (K S) 100 in cash, paid by writer of the option
Portfolio insurance
Portfolio managers can use index options to limit their downside risk:
Suppose that value of an index today is S0 , consider well-diversified portfolio ( = 1.0)
Assuming portfolio dividend yield = index dividend yield, we expect:
[% changes in portfolio value] [% changes in index value]
The portfolio value is protected against index falling below K if, for each 100 S0 dollars in the
portfolio, the manager buys one put option contract with strike price K
When the portfolios beta is not 1.0
If portfolios 6= 1.0, put options must be purchased for each 100S0 dollars in portfolio
Suppose a $500,000 portfolio has = 2.0 instead of 1.0, and S&P 500 index = 1,000
The number of put options required is: 2.0 500, 000/(1, 000 100) = 10
To calculate the appropriate strike price, the CAPM can be used
Suppose risk free rate = 12%, dividend yield on both index/portfolio is 4%, and protection is
required against portfolio value dropping below $450,000 in next three months
Calculation of expected value of portfolio when the index is 1,040 in three months and = 2.0
Value of index in three months:
Return from change in index:
Dividends from index:
Total return from index:
Risk-free interest rate:
Excess return from index over rf :
Expected excess return from portfolio over rf :
Expected return from portfolio:
Dividends from portfolio:
Expected increase in value of portfolio:
Expected value of portfolio:

1, 040
40/1, 000, or 4% per three months
0.25 4 = 1% per three months
4 + 1 = 5% per three months
0.25 12 = 3% per three months
5 3 = 2% per three months
2 2 = 4% per three months
3 + 4 = 7% per three months
0.25 4 = 1% per three months
7 1 = 6% per three months
$500, 000 1.06 = $530, 000

Strike price for options purchased = index level corresponding to required protection level
Two reasons why the cost of hedging increases as the of a portfolio increases:
1. More put options are required
2. They have a higher strike price
Currency options
Currency options are primarily traded in the OTC market
The advantage of this market is that large trades are possible, with strike prices, expiration dates,
and other features tailored to meet the needs of corporate treasurers
Example - European call to buy 1M with USD at 1.2000 USD per
If actual exchange rate at maturity = 1.25, payoff = 1, 000, 000 (1.25 1.20) = $50, 000
Similarly, example of European put option giving the holder the right to sell 10M Australian dollars
for USD at an exchange rate of 0.70 USD per AUD
If actual exchange rate at maturity = 0.67, payoff is 10, 000, 000 (0.70 0.67) = $300, 000
To hedge FX exposure, foreign currency options are a good alternative to forward contracts:
If due to receive at known future time, hedge risk with a put on that mature at that time
Hedging strategy guarantees that exchange rate applicable to will not be < strike price, while
allowing company to benefit from favorable exchange-rate movements
87

 Similarly, a company due to pay sterling at a known time in the future can hedge by buying calls
on sterling that mature at that time
This hedging strategy guarantees that the cost of the sterling will not be greater than a certain
amount while allowing the company to benefit from favorable exchange-rate movements
Forward contract locks in exchange rate for future transaction, option provides type of insurance
This is not free. It costs nothing to enter a forward, but options require a premium up front
Range forwards
Range forward: Variation on a standard forward contract for hedging FX risk:
Consider a US company that knows it will receive 1M in three months
Suppose that the three-month forward exchange rate is 1.9200 dollars per pound
Company can lock in this rate by entering a short forward to sell 1M in 3 months
This would ensure that the amount received for the 1M is $1,920,000
Short range forward contract
Alternative: Buy a European put with strike price of K1 and sell a European call with strike
price K2 , where K1 < 1.9200 < K2 . See payoff in Fig. 1(a), both options are on 1M
If the exchange rate in three months proves to be less than K1 , the put option is exercised and
as a result the company is able to sell the 1M at an exchange rate of K1
If the exchange rate is between K1 and K2 , neither option is exercised and the company gets
the current exchange rate for the 1M
If the exchange rate is greater than K2 , the call option is exercised against the company with
the result that the 1M is sold at an exchange rate of K2
The exchange rate realized for the 1M is shown in Fig. 2
(b) Long range-forward contract

Exchange rate realized with range-forward contract

Payoff

Payoff

(a) Short range-forward contract

K2
K1

Asset
Price

- Short range-forward contract used to hedge a future foreign currency inflow

- Long range-forward contract used to hedge a future foreign currency outflow

K2
K1
K2

Asset
Price
K1

Exchange rate in market

K1

Figure 1: Payoffs from short/long range-forwards

K2

Figure 2: FX rate realized with range-forwards

Long range forward contract

If company knew it was due to pay rather than receive 1M in 3 months, it could sell a
European put with strike price K1 and buy a European call with strike price K2 [Fig. 1(b)]
Exchange rate paid for 1M = rate received for 1M in short range-forward example
In practice, range forward set up so that price of put = price of call (select K1 , K2 such that p = c)
It costs nothing to set up the range forward ( costs nothing to set up regular forward)
As strike prices of call/put become closer in a range forward, it becomes a regular forward
A short range forward contract becomes a short forward contract
A long range forward contract becomes a long forward contract
Options on stocks paying known dividend yields
Dividends cause stock prices to reduce on the ex-dividend date by amount of dividend payment
The payment of a dividend yield at rate q therefore causes the growth rate in the stock price to be
less than it would otherwise be by an amount q
If, with a dividend yield of q, the stock price grows from S0 today to ST at time T , then in the
absence of dividends it would grow from S0 today to ST eqT at time T
Alternatively, in the absence of dividends it would grow from S0 eqT today to ST at time T
88

 We get the same probability distribution for the stock price at time T in each of the two cases:
1. The stock starts at price S0 and provides a dividend yield at rate q
2. The stock starts at price S0 eqT and pays no dividends
When valuing European option lasting for T on stock paying known dividend yield q, we reduce current
stock price from S0 to S0 eqT and then value the option as if the stock pays no dividends
Lower bounds for option prices
Substituting S0 eqT for S0 in Eq. (9.1), a lower bound for the European call c is:
c S0 eqT KerT

(1)

We can also prove this directly by considering the following two portfolios:
Portfolio A: One European call option plus an amount of cash = KerT
Portfolio B: eqT shares with dividends being reinvested in additional shares
To obtain a lower bound for European put, we similarly replace S0 by S0 eqT in Eq. (9.2) to get:
p KerT S0 eqT

(2)

This result can also be proved directly by considering the portfolios:

Portfolio C: One European put plus eqT shares with dividends reinvested in more shares
Portfolio D: An amount of cash equal to KerT
Put-call parity
Replacing S0 by S0 eqT in Eq. (9.3), put-call parity for option on stock paying dividend yield q:
c + KerT = p + S0 eqT

(3)

This result can also be proved directly by considering the two portfolios:
Portfolio A: One European call option plus an amount of cash = KerT
Portfolio C: One European put plus eqT shares with dividends reinvested in more shares
Both portfolios are both worth max(ST , K) at time T
They must be worth the same today, and the put-call parity result in Eq. (3) follows
For American options, the put-call parity relationship is:
S0 eqT K < C P < S0 KerT
Pricing formulas
By replacing S0 by S0 eqT in Black-Scholes, Eqs. (13.15) and (13.16), we obtain the price c of a
European call and the price p of a European put on a stock paying a dividend yield q as:
c = S0 eqT N (d1 ) KerT N (d2 )

(4)

p = KerT N (d2 ) S0 eqT N (d1 )

(5)

Since ln(S0 eqT /K) = ln(S0 /K) qT , d1 and d2 are given by:
d1 =

ln(S0 /K) + (r q + 2 /2)T

and

d2 =

ln(S0 /K) + (r q 2 /2)T

= d1 T
T

Dividend should, for the purposes of option valuation, be defined as the reduction in the stock
price on the ex-dividend date arising from any dividends declared
If the dividend yield rate is known but not constant during the life of the option, Eqs. (4) and (5)
are still true, with q equal to the average annualized dividend yield during the options life
Differential equation and risk-neutral valuation
When including a dividend yield of q in analysis of Ch. 13, the equation (13.13) becomes:
f
f
1
2f
+ (r s)S
+ 2 S 2 2 = rf
t
S 2
S

(6)

89

 Like Eq. (13.13), this does not involve any variable affected by risk preferences
The risk-neutral valuation procedure can be used
Risk-neutral world: Expected growth rate must be r q Risk-neutral process for stock price:
dS = (r q)Sdt + Sdz

(7)

To value a derivative dependent on a stock that provides a dividend yield equal to q, we set the
expected growth rate of the stock equal to r q and discount the expected payoff at rate r
When expected growth rate in stock price is r q, expected stock price at T = S0 e(rq)T
The expected payoff for a call option in a risk-neutral world is then:
e(rq)T S0 N (d1 ) KN (d2 )
Discounting at rate r for time T leads to Eq. (4)
Valuation of European stock index options
The index can be treated as an asset paying a known yield
Eqs. (1) and (2) provide a lower bound for European index options
Eq. (3) is the put-call parity result for European index options
Eqs. (4) and (5) can be used to value European options on an index
The binomial tree approach can be used for American options
In all cases, S0 is equal to the value of the index, is equal to the volatility of the index, and q is
equal to the average annualized dividend yield on the index during the life of the option
Calculation of q should include only dividends whose ex-dividend date occurs during option life
If absolute amount of dividend paid on stocks underlying index (rather than dividend yield) is known,
Black-Scholes can be used with initial stock price reduced by PV of dividends
Difficult to implement for broad stock index (requires dividends on each stock underlying index)
Forward prices
Define F0 as the forward price of the index for a contract with maturity T :
As shown by Eq. (5.3), F0 = S0 e(rq)T European call/put price c/p in Eqs. (4)/(5) are:
c = F0 erT N (d1 ) KerT N (d2 )
rT

p = Ke
d1 =

rT

N (d2 ) F0 e

ln(F/K) + 2 T /2

(8)

N (d1 )

and

d2 =

(9)
ln(F/K) 2 T /2

The put-call parity relationship in Eq. (3) can be written:

c + KerT = p + F0 erT

F0 = K + (c p)erT

(10)

If, as common in exchange-traded markets, pairs of puts/calls with same strike price are traded
actively for a maturity date, Eq. (10) can estimate the forward price of index for that maturity
Once forward prices of the index for different maturities have been obtained, the term structure
of forward rates can be estimated, and other options valued using Eqs. (8)/(9)
Advantage: The dividend yield on the index does not have to be estimated explicitly
Implied dividend yields
If estimates of the dividend yield are required (e.g because an American option is being valued),
calls and puts with the same strike price and time to maturity can be used. From Eq. (3),
q=

1 c p + KerT
ln
T
S0

For a particular strike price and time to maturity, the estimates of q calculated from this equation
are liable to be unreliable. But when the results from many matched pairs of calls and puts are
combined, a clear picture of the dividend yield being assumed by the market emerges
90

Valuation of European currency options

For currency options, S0 = spot exchange rate. S0 = value of one unit of foreign currency in USD
A foreign currency is analogous to a stock paying a known dividend yield:
Owner of foreign currency receives a yield = risk-free interest rate rf in foreign currency
Eqs. (1) and (2) with q replaced by rf provide bounds for the European call/put price c/p:
c S0 erf T KerT

p KerT S0 erf T

and

Eq. (3) with q replaced by rf provides the put-call parity result for currency options:
c + KerT = p + S0 erf T
Finally, Eqs. (4) and (5) provide pricing formulas for currency options when q replaced by rf :
c = S0 erf T N (d1 ) KerT N (d2 )
rT

p = Ke

rf T

N (d2 ) S0 e

(11)

N (d1 )

(12)

ln(S0 /K) + (r rf + 2 /2)T

ln(S0 /K) + (r rf 2 /2)T

and d2 =
= d1 T
T
T
Both the domestic interest rate r and the foreign interest rate rf are the rates for a maturity T
Put and call options on a currency are symmetrical in that a put option to sell currency A for currency
B at strike price K is the same as a call option to buy B with currency A at strike price 1/K
Using forward exchange rates
Because banks and other financial institutions trade forward contracts on foreign exchange rates
actively, foreign exchange rates are often used for valuing options
Eq. (5.9): Forward rate F0 for maturity T is: F0 = S0 e(rrf )T Eqs. (11) and (12) simplify to:
d1 =

c = erT [F0 N (d1 ) KN (d2 )]

(13)

p = erT [KN (d2 ) F0 N (d1 )]

(14)

d1 =

ln(F0 /K) + 2 T /2

and

d2 =

ln(F0 /K) 2 T /2

= d1 T
T

The maturity of forward/futures contract must be the same as maturity of European option
American options
Binomial trees can be used to value American options on indices and currencies
As with American
options on non-dividend-paying stock, u determines size of up moves and is set

equal to e t , where = volatility and t = length of time steps

The parameter determining the size of down movements d is set equal to 1/u, or e t
For a non-dividend-paying stock, the probability of an up movement is:
p=

ad
ud

where

a = ert

For options on indices and currencies, the formula for p is the same, but a is defined differently. In
the case of options on an index, with q as the dividend yield on the index: a = e(rq)t
In the case of options on a currency, with rf as the foreign risk-free rate: a = e(rrf )t
In some circumstances it is optimal to exercise American currency options prior to maturity:
Thus, American currency options are worth more than their European counterparts
In general, call options on high-interest currencies and put options on low-interest currencies are
the most likely to be exercised prior to maturity
Reason: A high-/low-interest currency is expected to depreciate/appreciate

91

92

Hull - Ch. 16: Futures options

Nature of futures options
A futures option is the right to enter into a futures contract at a certain futures price by a certain date
Futures options are generally American, can be exercised any time during contract life
A call futures option is the right to enter into a long futures contract at a certain price
If a call futures option is exercised, holder acquires a long position in underlying futures contract
plus cash amount equal to [most recent settlement futures price] [strike price]
Effective payoff from call futures option = [futures price at time of exercise] [strike price]
A put futures option is the right to enter into a short futures contract at a certain price
If a put futures option is exercised, holder acquires a short position in underlying futures
contract plus cash amount equal to [strike price] [most recent settlement futures price]
Effective payoff from put futures option = [strike price] [futures price at time of exercise]
Example
Suppose it is August 15 and an investor has one September futures call option contract on copper
with a strike price of 240 cents per pound. One futures contract is on 25,000 pounds of copper
Suppose that the futures price of copper for delivery in September is currently 251 cents, and at
the close of trading on August 14 (the last settlement) it was 250 cents
If the option is exercised, the investor receives a cash amount of 25, 000(250240) cents = $2,500
plus a long position in a futures contract to buy 25,000 pounds of copper in September
Position in futures contract can be closed out immediately: This leaves investor with $2,500 cash
payoff plus 25, 000 (251 250) cents = $250 reflecting change in futures price since last settlement
The total payoff from exercising the option on August 15 is $2,750, which equals 25, 000(F K),
where F is the futures price at the time of exercise and K is the strike price
Expiration months
Futures options referred to by delivery month of underlying futures, not by expiration of option
Options on interest rate futures
The most actively traded interest rate options offered by exchanges in the US are those on Treasury
bond futures, Treasury note futures, and Eurodollar futures
Treasury bond futures option (traded on CBOT):
One Treasury bond futures contract is for the delivery of $100,000 of Treasury bonds
1
Price of T-bond futures option quoted as % of face value of underlying T-bonds to nearest 64
Option on Eurodollar futures (traded on the CME):
When the Eurodollar futures quote changes by 1 basis point, or 0.01%, there is a gain or loss
on a Eurodollar futures contract of $25
Similarly, in the pricing of options on Eurodollar futures, 1 basis point represents $25
Interest rate futures option contracts work in the same way as other futures options contracts
In addition to cash payoff, holder of call option obtains a long position in the futures contract
when the option is exercised and option writer obtains a corresponding short position
The total payoff from the call, including the value of the futures position, is max(F K, 0),
where F is the futures price at the time of exercise and K is the strike price
Interest rate futures prices increase when bond prices increase (i.e., when interest rates fall).
They decrease when bond prices decrease (i.e., when interest rates rise)
Example - Eurodollar futures options
It is February and the futures price for the June Eurodollar contract is 93.82 (corresponding
to a 3-month Eurodollar interest rate of 6.18% per annum)
Price of call option on the contract with strike price of 94.00 is quoted as 0.1, or 10 bp
Option attractive to investor who feels that interest rates are likely to come down
Suppose short-term rates drop by 100 bp and investor exercises the call when Eurodollar futures
price is 94.78 ( 3-month Eurodollar interest rate of 5.22% per annum)
The payoff is 25 (94.78 94.00) 100 = $1, 950. The cost of the contract is 10 25 = $250
The investors profit is $1,700
93

 Example - Treasury bond futures options

9
In August, futures price for December T-bond is 96-09 (or 96 32
= 96.28125)
An investor who feels that the yield on long-term government bonds will fall by December
might choose to buy December calls with a strike price of 98
4
Assume that the price of these calls is 1-04 (or 1 64
= 1.0625% of the principal)
If long-term rates fall and the Treasury bond futures price rises to 100-00, the investor will
make a net profit per $100 of bond futures of 100.00 98.00 1.0625 = 0.9375
Since one option contract is for the purchase or sale of instruments with a face value of $100,000,
the investor would make a profit of $937.50 per option contract bought
Reasons for the popularity of futures options
People choose to trade options on futures rather than options on the underlying asset because a futures
contract is, in many circumstances, more liquid and easier to trade than the underlying asset
Furthermore, a futures price is known immediately from trading on the futures exchange, whereas
the spot price of the underlying asset may not be so readily available
Futures on commodities are also often easier to trade than the commodities themselves
An important point about a futures option is that exercising it does not usually lead to delivery of the
underlying asset, as in most circumstances the underlying futures contract is closed out prior to delivery
Futures options are therefore normally eventually settled in cash
This is appealing to many investors, particularly those with limited capital who may find it difficult
to come up with the funds to buy the underlying asset when an option is exercised
Other advantage: Futures and futures options are traded in pits side by side in the same exchange
This facilitates hedging, arbitrage, and speculation and make the markets more efficient
Futures options also tend to entail lower transactions costs than spot options in many situations
European spot and futures options
The payoff from a European call option with strike price K on the spot price of an asset is max(ST K, 0)
where ST is the spot price at the options maturity
The payoff from a European call option with the same strike price on the futures price of the asset is
max(FT K, 0) where FT is the futures price at the options maturity
If futures contract matures at same time as option, then FT = ST and the two options are equivalent
Similarly, a European futures put option is worth the same as its spot put option counterpart when
the futures contract matures at the same time as the option
European futures options can be used to value corresponding European spot options
Put-call parity
Consider European call/put futures options, both with strike price K/time to expiration T
Portfolio A: European call futures option plus an amount of cash = KerT
Portfolio B: European put futures option + long futures contract + cash = F0 erT
In portfolio A, the cash can be invested at the risk-free rate r and grows to K at time T
Let FT be the futures price at maturity of the option
If FT > K, the call option in portfolio A is exercised and portfolio A is worth FT
If FT < K, the call is not exercised and portfolio A is worth K
The value of portfolio A at time T is therefore max(FT , K)
In portfolio B, the cash can be invested at the risk-free rate to grow to F0 at time T
The put option provides a payoff of max(K FT , 0)
The futures contract provides a payoff of FT F0
The value of portfolio B at time T is therefore F0 + (FT F0 ) + max(K FT , 0) = max(FT , K)
Because portfolios have same value at T and no early exercise, they are worth the same today
The value of portfolio A today is: c + KerT where c is the price of the call futures option
Marking-to-market ensures that futures contract in portfolio B is worth zero today. Portfolio B is
therefore worth p + F0 erT where p is the price of the put futures option. Hence,
c + KerT = p + F0 erT

(1)
94

 The difference between this put-call parity relationship and the one for a non-dividend-paying stock in
Eq. (9.3) is that the stock price S0 is replaced by the discounted futures price F0 erT
When futures contract matures at same time as option, European futures/spot options are the same
Eq. (1) gives a relationship between price of call option on spot price, price of put option on spot
price, and futures price when both options mature at same time as futures contract
For American futures options, the put-call relationship is:
F0 erT K < C P < F0 KerT

(2)

Bounds for futures options

Because the price of a put p cannot be negative, it follows from Eq. (1) that:
c (F0 K)erT

(3)

Similarly, because the price of a call option cannot be negative, it follows from Eq. (1) that:
p (K F0 )erT

(4)

Prices of European call/put options are close to lower bounds when options are deep in the money:
Call option deep in the money Put option deep out of the money p is very close to zero
Difference between c and its lower bound equals p Price of call option very close to lower bound
Because American futures options can be exercised at any time: C F0 K and P K F0
Lower bound for American option price is always higher than lower bound for European option price,
because always some chance that an American futures option will be exercised early
Valuation of futures options using binomial trees
Key difference between futures-/stock options: No up-front costs with a futures option
Consider futures price starting at F0 and rising to F0 u or declining to F0 d over time period T
We consider an option maturing at time T
Suppose that its payoff is fu if the futures price moves up and fd if it moves down
Riskless portfolio: Short position in one option combined with long position in futures contracts:
=

fu fd
F0 u F0 d

The value of the portfolio at time T is then always: (F0 u F0 ) fu

Denoting the risk-free interest rate by r, the portfolio value today is: [(F0 u F0 ) fu ]erT
Another expression for the PV of the portfolio is f , where f is the value of the option today
Thus: f = [(F0 u F0 ) fu ]erT . Substituting for :
f = erT [pfu + (1 p)fd ]

where

p=

1d
ud

(5)

Eq. (5) gives the risk-neutral probability of an up movement

Multistep trees
Multistep binomial trees are used to value American-style futures
options like stock options

The parameter u defining up movements in the futures price is e t , where is the volatility of
the futures price and t is the length of one time step
The probability of an up movement in the future price is [Eq. (5)]: p = (1 d)/(u d)
Drift of a futures price in a risk-neutral world
In a risk-neutral world, a futures price behaves in the same way as a stock paying a dividend yield at
the domestic risk-free interest rate r. Clues:
p in a binomial tree for futures price is like for stock paying a dividend yield q when q = r
The put-call parity relationship for futures options prices is the same as that for options on a stock
paying a dividend yield q when the stock price is replaced by the futures price and q = r
95

 Calculation of the drift of a futures price in a risk-neutral world:

Define Ft as the futures price at time t
If we enter into a long futures contract today, its value is zero
At time t (the first time it is marked to market) it provides a payoff of Ft F0
If r is the very-short-term (t-period) interest rate at time 0, risk-neutral valuation gives the value
t F0 ] where E
= expectations in a risk-neutral world
of the contract at time 0 as: ert E[F
rt

Hence, e
E[Ft F0 ] = 0, showing that E[Ft ] = F0
2t ] = Ft , E[F
3t ] = F2t , . . . E[F
T ] = F0 for any time T
Similarly, E[F
The drift of the futures price in a risk-neutral world is therefore zero. From Eq. (15.7), the futures price
behaves like a stock providing a dividend yield q equal to r
Usual assumption made for process followed by futures price F in risk-neutral world is ( constant):
dF = F dz

(6)

Differential equation
The differential equation satisfied by a derivative dependent on a futures price is:
f
1 2f 2 2
+
F = rf
t
2 S 2

(7)

Same form as Eq. (15.6) with q r. This confirms that, for valuing derivatives, a futures price
can be treated like a stock providing a dividend yield at rate r
Blacks model for valuing futures options
Assuming that the futures price follows (lognormal) process in Eq. (6), European call/put price c/p for
futures option are given by Eqs. (15.4)/(15.5) with S0 replaced by F0 and q = r:
c = erT [F0 N (d1 ) KN (d2 )]
rT

p=e

(8)

[KN (d2 ) F0 N (d1 )]

(9)

ln(F0 /K) + 2 T /2
ln(F0 /K) 2 T /2

and d2 =
= d1 T
T
T
When the cost of carry and the convenience yield are functions only of time, the volatility of the futures
price is the same as the volatility of the underlying asset
Blacks model does not require the option contract and the futures contract to mature at same time
Using Blacks model instead of Black-Scholes
Futures-/spot options are equivalent when option/futures contract matures at same time
Eqs. (8) and (9) allow to value European options on spot price of an asset
The variable F0 in Eqs. (8) and (9) is set equal to either the futures or the forward price of the
underlying asset for a contract maturing at the same time as the option
Eqs. (15.13) and (15.14) show Blacks model being used to value European options on the spot
value of a currency: They avoid the need to estimate the foreign risk-free interest rate explicitly
Eqs. (15.8) and (15.9) show Blacks model being used to value European options on the spot value
of an index: They avoid the need to estimate the dividend yield explicitly
Blacks model useful to imply term structure of forward rates from actively traded index options
The forward rates can then be used to price other options on the index
The same approach can be used for other underlying assets
d1 =

American futures options vs. American spot options

Traded futures options are in practice usually American
Assuming risk-free rate r > 0, it may be optimal to exercise an American futures option early
American futures options are worth more than their European counterparts
It is not generally true that an American futures option is worth the same as the corresponding American
spot option when the futures and options contracts have the same maturity
96

 Assume a normal market with futures prices consistently higher than spot prices prior to maturity
Common with most stock indices, gold, silver, low-interest currencies, and some commodities
When futures option are exercised early, they provide a greater profit to the holder
American call futures option must be worth corresponding American spot call option
Similarly, American put futures option must be worth corresponding American spot put option
If there is an inverted market with futures prices consistently lower than spot prices, as is the case with
high-interest currencies and some commodities, the reverse must be true
The later the futures contract expires the greater the differences tend to be
Futures-style options
These are futures contracts on the payoff from an option
A futures-style option is a bet on what the payoff from an option will be
Traders who buy/sell a futures-style option post margin like on a regular futures contract
The contract is settled daily, and the final settlement price is the payoff from the option
If interest rates constant, futures price in a futures-style option = forward price in forward contract on
the option payoff Futures price for a futures-style option = price paid for the option if payment were
made in arrears It is the value of a regular option compounded forward at the risk-free rate
Defining d1 and d2 as in Eqs. (8) and (9):
The futures price in a call futures-style option is: F0 N (d1 ) KN (d2 )
The futures price in a put futures-style option is: KN (d2 ) F0 N (d1 )
Formulas ok for futures-style option on futures contract and futures-style option on spot value of asset
In the first case, F0 is the current futures price for the contract underlying the option
In the second case, it is the current futures price for a futures contract on the underlying asset
maturing at the same time as the option
The put-call parity relationship for a futures-style options is: p + F0 = c + K
American futures-style option can be exercised early Immediate settlement at options intrinsic value
It is never optimal to exercise an American futures-style options on a futures contract early because
the futures price of the option is always greater than the intrinsic value
American futures-style option can be treated like European futures-style option

97

98

Asset-Liability Management

BKM - Ch. 16: Managing bond portfolios

Introduction
A passive investment strategy takes market prices of securities as set fairly
Rather than attempting to beat the market by exploiting superior information or insight, passive
managers act to maintain an appropriate risk-return balance given market opportunities
Special case of passive management: Immunization strategy against interest rate risk
In contrast, active investment strategy tries to achieve returns greater than commensurate with risk
In the context of bond management this style of management can take two forms:
1. Either use interest rate forecasts to predict movements in the entire bond market
2. Or use intramarket analysis to identify particular sectors of the market/bonds mispriced
Interest rate risk is crucial to formulating both active and passive strategies
Interest rate risk
An inverse relationship exists between bond prices and yields
As interest rates rise (fall), bondholders experience capital losses (gains)
These gains/losses make fixed-income investments risky, even if coupon/principal guaranteed
Interest rate sensitivity - Malkiels rules
Percentage change in bond price

150

100

Bond
A
B
C
D

Coupon
12%
12%
3%
3%

-2

Maturity
5 years
30 years
30 years
30 years

Initial YTM
10%
10%
10%
6%

50

D
-4

-50

Change in Yield to Maturity (%)

Figure 1: Change in bond price as a function of change in YTM

1.
2.
3.
4.

Bond prices/yields inversely related: As yields increase, bond prices fall, and conversely
Increase in bonds YTM results in smaller price change than decrease in YTM of same magnitude
Long-term bonds more sensitive to interest rate changes than short-term bonds
The sensitivity of bond prices to changes in yields increases at a decreasing rate as maturity
increases. I.e., interest rate risk is less than proportional to bond maturity
5. Interest rate risk is inversely related to the bonds coupon rate: Prices of low-coupon bonds are
more sensitive to changes in interest rates than prices of high-coupon bonds
6. Sensitivity of bonds price to change in its yield is inversely related to current YTM
Maturity is a major determinant of interest rate risk, but other factors as well
Higher-coupon-rate bonds have higher fraction of value tied to coupons vs. final par payment
Portfolio of coupons more heavily weighted toward earlier, short-maturity payments Lower
effective maturity Malkiels rule #5 - Price sensitivity falls with coupon rate
Similarly, for rule #6 - Price sensitivity falls with YTM: Higher yield reduces PV of all bonds
payments, but more so for distant payments
At higher yield, higher fraction of bonds value due to its earlier payments (lower effective
maturity/interest rate sensitivity) Overall sensitivity to changes in yields is lower
Duration
To deal with the ambiguity of the maturity of a bond making many payments, we need a measure
of the average maturity of the bonds promised CFs to serve as a useful summary statistic
Macaulays duration
The effective maturity concept is called the duration of the bond
101

 Macaulays duration = wtd avg. of times to each coupon/principal payment

Weight associated with each payment time related to importance of that payment to bond
value: It is the proportion of bond total value accounted for by that payment
I.e., weight wt associated with CF at t = [PV of payment] [bond price]:
wt =

CFt /(1 + y)t

Bond price

where y is the bonds YTM

weights = 1.0 because the sum of the CFs discounted at YTM is the bond price
Macaulays duration formula
D=

T
X

t wt

(1)

t=1

Duration is a key concept in fixed-income portfolio management for at least three reasons:
1. It is a simple summary statistic of the effective average maturity of the portfolio
2. It turns out to be an essential tool in immunizing portfolios from interest rate risk
3. Duration is a measure of the interest rate sensitivity of a portfolio
The duration measure enables us to quantify the relationship that long-term bonds are more sensitive to interest rate movements than are short-term bonds
When interest rates change, proportional change in bonds price related to change in YTM y:


P
(1 + y)
= D
(2)
P
1+y
Proportional price change = proportional change in [1 + bonds yield] [bonds duration]
Modified duration D
D =

D
1+y

P
= D y
P

D =

1 dP
P dy

(3)

% change in bond price = [modified duration] [change in bonds YTM]

Because the % change in the bond price is proportional to modified duration, modified duration
is a natural measure of the bonds exposure to changes in interest rates
What determines duration?
If we speculate on interest rates, duration tells us how strong a bet we are making. Conversely, if we
remain neutral on rates, and simply match the interest rate sensitivity of a chosen bond-market
index, duration allows to measure that sensitivity and mimic it in our portfolio
Rule 1 for duration: The duration of a zero-coupon bond equals its time to maturity
Rule 2 for duration: At constant maturity, bonds duration lower when coupon rate higher
This property corresponds to Malkiels fifth relationship and is attributable to the impact of
early coupon payments on the weighted-average maturity of a bonds payments
The higher these coupons, the higher the weights on the early payments and the lower is the
weighted average maturity of the payments
Rule 3 for duration: Holding coupon rate constant, bonds duration generally increases with
time to maturity. Duration always increases with maturity for bonds selling at par/premium
Duration need not always increase with time to maturity
For deep-discount bonds (3% coupon in Fig. 2), duration can fall when maturity increases
For coupon bonds, duration increases < 1-yr with 1-yr increase in maturity (slope 1.0)
Rule 4 for duration: All else equal, duration of coupon bond higher when bonds YTM lower
While a higher yield reduces the PV of all of the bonds payments, it reduces the value of
more distant payments by a greater proportional amount At higher yields, higher fraction
of bonds total value lies in earlier payments, thereby reducing effective maturity
102

Duration (years)

30

on

Co
ro-

o
up

nB

Ze
20

15% Coupon
YTM = 6%

3% Coupon
YTM = 15%
10

15% Coupon
YTM = 15%

0
0

10

15

20

Maturity (years)

Figure 2: Bond duration vs. bond maturity

Rule 5 for duration: The duration of a level perpetuity is:
Duration of perpetuity =

1+y
y

Maturity and duration can differ substantially

(4)

Notice from Fig. 2 that as their maturities become ever longer, the durations of the two coupon
bonds with yields of 15% both converge to the duration of the perpetuity with the same yield
Convexity
Duration rule P/P = D y for impact of interest rates on bond prices is only approximate
The relationship between bond prices and yields is not linear
Duration rule: Good approximation for small changes in bond yield, less accurate for larger changes
The duration linear approximation always understates the value of the bond:
It underestimates the increase in bond price when the yield falls
It overestimates the decline in price when the yield rises
This is due to the convex curvature of the true price-yield relationship
The curvature of the price-yield curve is called the convexity of the bond
Convexity = rate of change of slope of price-yield curve, as a fraction of bond price
Convexity =

1 d2 P
P dy 2

Formula for convexity of a bond with maturity of T years making annual coupon payments:
Convexity =


T 
X
1
CFt
2
(t
+
t)
P (1 + y)2
(1 + y)t
t=1

Where CFt = coupon payment before maturity or final coupon plus par value at maturity
Convexity allows to improve the duration approximation for bond price changes. Eq. (3) becomes:
P
= D y +
P

1
2

Convexity (y)2

(5)

Bond with positive convexity: 2nd term 0, regardless of whether yield rises/falls
Duration rule always underestimates new value of a bond following a change in its yield
For small y, the linear approximation given by duration rule is sufficiently accurate
Convexity is more important as a practical matter when potential rate changes are large
Why do investors like convexity?
Bonds with greater curvature gain more in price when yields fall than they lose when yields rise
If interest rates are volatile, this is an attractive asymmetry that increases the expected return on
the bond, because bond will benefit more from rate decreases/suffer less from rate increases
103

Bond B (less convex) - Actual price change

Bond A - Actual price change

Percentage change in bond price

150

Bond A - Duration approximation

100

50

D
-4

-50

-2

Change in Yield to Maturity (%)

Bond Price

Figure 3: Bond price convexity

Of course, if convexity is desirable, it will not be available for free
Investors pay higher prices and accept lower YTM on bonds with greater convexity
Duration and convexity of callable bonds
When interest rates high, price-yield curve for callable bond is convex (as for straight bond)
But as rates fall, there is a ceiling on price: Bond cannot be worth more than its call price
As rates fall, the bond is subject to price compression
Its value is compressed to the call price
In this region, price-yield curve lies below its tangency line, and has negative convexity
In the region of negative convexity, the price-yield curve exhibits an unattractive asymmetry
Rate increases result in larger price decline than price gain for rate decreases of equal magnitude
Asymmetry arises from the fact that bond issuer has an option to call back the bond
Investors are compensated for this unattractive situation when purchasing the bond: Callable
bonds sell at lower prices (higher initial yields) than comparable straight bonds
The effect of negative convexity is highlighted in Eq. (5)
When convexity negative, 2nd term on RHS is necessarily negative Bond price performance
will be worse than would be predicted by the duration approximation
However, callable bonds/bonds with embedded options difficult to analyze with Macaulays
duration. Because of such options, future CFs from the bonds no longer known
Region of
Negative Convexity
Call
Price

Region of
Positive Convexity

5%

10%

Interest Rate (%)

Figure 4: Price-yield curve for a callable bond

For bonds with embedded options, the convention is to compute the effective duration
Effective duration: % change in bond price per unit change in market interest rates:
P/P
(6)
r
Effective duration is not computed relative to a change in bonds own YTM because for bonds
with embedded options, the YTM is often not a relevant statistic
Effective duration =

104

 Effective duration formula relies on pricing methodology that accounts for embedded options
The effective duration will be a function of variables that would not matter to conventional
duration, e.g. the volatility of interest rates
In contrast, modified/Macaulay duration comes directly from promised bond CFs/YTM
Example of effective duration calculation
Suppose that a callable bond with a call price of $1,050 is selling today for $980
If the yield curve shifts up by .5%, the bond price will fall to $930
If it shifts down by .5%, the bond price will rise to $1,010
To compute effective duration, we compute:
r = Assumed increase in rates Assumed decrease in rates = .5% (.5%) = 1%
P = Price at .5% rate increase Price at 5% rate decrease = 930 1,010 = -$80
Then the effective duration of the bond is: (80/980)/0.01 = 8.16 years
The bond price changes by 8.16% for a 1% point swing in rates around current values
Duration and convexity
Biggest market where call provisions are important is market for mortgage-backed securities
Such securities are called pass-throughs because CFs from borrowers are first passed through to an
agency (Fannie Mae/Freddie Mac) and then passed through to ultimate purchaser of MBS
Example
Suppose that ten 30-yr mortgages, each with principal of $100k, are grouped into $1M pool
If the mortgage rate is 8%, then the monthly payment on each loan would be $733.76
Owner of the MBS receives $7,337.60 = total payment from the 10 pooled mortgages
The right to prepay the loan is precisely analogous to the right to refund a callable bond:
The call price for the mortgage is simply the remaining principal balance on the loan
The MBS is best viewed as a portfolio of callable amortizing loans
Mortgage-backs are subject to the same negative convexity as other callable bonds:
When rates fall and homeowners prepay mortgages, repayment of principal is passed through
Rather than capital gains on investment, investors receive the principal balance on the loan
Differences between the mortgage-backs and callable corporate bonds:
Mortgage-backs are commonly selling for more than their principal balance because homeowners do not refinance as soon as rates drop MBS exhibit negative convexity at low rates, but
implicit call price (principal balance) is not a firm ceiling
Credit enhancement: Common feature of the asset-backed market: The credit risk of the underlying borrower is enhanced by the guarantee of Freddie or Fannie that any mortgage default
will be treated from the point of view of the investor as if the mortgage had been prepaid
Simple mortgage-backs have also given rise to a rich set of mortgage-backed derivatives that can
be used to help investors manage interest rate risk:
E.g., a CMO (collateralized mortgage obligation) further redirects the CF stream of the MBS
to several classes of derivative securities called tranches
Tranches designed to allocate interest rate risk to investors most willing to bear that risk
The following table is an example of a very simple CMO structure
The underlying mortgage pool is divided into three tranches, each with a different effective
maturity and therefore interest rate risk exposure
Suppose the original pool is subdivided into three tranches as:
Tranche A
Tranche B
Tranche C

$4 million principal
$3 million principal
$3 million principal

Short-pay tranche
Intermediate-pay tranche
Long-pay tranche

In each period, each tranche receives interest owed based on promised coupon/principal
But initially, all principal payments (prepayments and amortization) go to tranche A, tranches
B and C receive only interest payments until tranche A is retired
Once tranche A is fully paid off, all principal payments go to tranche B
Finally, when B is retired, all principal payments go to C

105

 This makes tranche A a short-pay class, with the lowest effective duration, while tranche C
is the longest-pay tranche Simple way to allocate interest rate risk among tranches
In essence, the mortgage pool is treated as a source of CFs that can be reallocated to different
investors in accordance with the tastes of different investors
Passive bond management
Passive managers take bond prices as fair and only seek to control only risk of their fixed-income portfolio
Two broad classes of passive management are pursued in the fixed-income market:
1. An indexing strategy that attempts to replicate the performance of a given bond index
2. Immunization techniques shield from exposure to interest rate fluctuations
Although indexing and immunization strategies are alike in that they accept market prices as correctly
set, they are very different in terms of risk exposure
A bond-index portfolio has same risk-reward profile as bond market index to which it is tied
In contrast, immunization seeks to establish a virtually zero-risk profile (no impact from rate moves)
Bond-index funds
In principle, bond market indexing is similar to stock market indexing: The idea is to create a
portfolio that mirrors the composition of an index that measures the broad market
Indexes of the broad bond market: (i) Lehman Aggregate Bond Index, (ii) Salomon Smith Barney
Broad Investment Grade (BIG) Index, and (iii) Merrill Lynch US Broad Market Index
All are market-value-weighted indexes of total returns
All three include government, corporate, mortgage-backed, and Yankee bonds in their universes
All three indexes include only bonds with maturities greater than 1 year
Problems in the formation of an indexed bond portfolio
Indexes include more than 5,000 securities Difficult to replicate
Many bonds thinly traded Difficult to identify owners/buy securities at fair market price
Bond-index funds also face more difficult rebalancing problems than do stock-index funds
Bonds are continually dropped from the index as their maturities fall below 1 year
As new bonds are issued, they are added to the index Securities used to compute bond
indexes constantly change. As they do, the manager must update/rebalance portfolio
Bonds generate considerable interest income that must be reinvested Further complication
In practice, a stratified sampling/cellular approach is pursued to replicate broad bond indexes:
First, the bond market is stratified into several subclasses
Criteria (maturity, issuer, bonds coupon rate, issuer credit risk) are used to form cells
Bonds falling within each cell are then considered reasonably homogeneous
Next, the percentages of the entire universe (i.e., the bonds included in the index to be matched)
falling within each cell are computed and reported
Finally, the portfolio manager establishes a bond portfolio with representation for each cell
that matches the representation of that cell in the bond universe
Characteristics of portfolio for maturity, coupon rate, credit risk, industrial representation match
index characteristics, and portfolio performance should match the index
Immunization
Many institutions try to insulate their portfolios from interest rate risk altogether:
1. Some institutions, such as banks, are concerned with protecting the current net worth or net
market value of the firm against interest rate fluctuations
2. Other investors (pension funds) may face an obligation to make payments after a given number
of years and are more concerned with protecting the future values of their portfolios
Immunization techniques refer to strategies used by such investors to shield their overall financial
status from exposure to interest rate fluctuations
Banks/thrift institutions have natural mismatch between asset/liability maturities
Bank liabilities are primarily deposits from customers, short-term in nature (low duration)
Bank assets by contrast are composed largely of outstanding loans or mortgages. These assets
are of longer duration than deposits, and more sensitive to interest rate fluctuations
106

In periods when interest rates increase unexpectedly, banks can suffer serious decreases in net
worth, their assets fall in value by more than their liabilities
Similarly, a pension fund may have a mismatch between the interest rate sensitivity of the assets
held in the fund and the PV of its liabilities - the promise to make payments to retirees
In some recent years pension funds lost ground despite excellent investment returns
As interest rates fell, value of their liabilities grew even faster than value of their assets
The idea behind immunization is that duration-matched assets and liabilities let the asset portfolio
meet the firms obligations despite interest rate movements
Two offsetting types of interest rate risk: (i) Price risk and (ii) Reinvestment rate risk
Increases in interest rates cause capital losses but at same time, reinvested income grows faster
If the portfolio duration is chosen appropriately these two effects will cancel out exactly
For horizon equal to portfolios duration, price risk/reinvestment risk exactly cancel out
Duration matching balances the difference between the accumulated value of coupon payments
(reinvestment rate risk) and the sale value of a bond (price risk)
We can also analyze immunization in terms of present as opposed to future values
Even with immunization, bond value at the horizon can be 6= obligation
This happens because the bond and the obligation have different convexities
The bond and the obligation are not duration-matched across interest rate shifts
Importance of rebalancing immunized portfolios
As interest rates and asset durations change, a manager must rebalance the portfolio of fixedincome assets continually to realign its duration with the duration of the obligation
Moreover, asset durations will change solely because of the passage of time
Duration generally decreases less rapidly than does maturity
Even if obligation immunized at the outset, as time passes, durations of the asset/liability fall
at different rates Without portfolio rebalancing, durations become unmatched
Immunization is a passive strategy (does not involve attempts to identify undervalued securities)
However, immunization managers still actively update and monitor their positions
Constructing an immunized portfolio
An insurance company must make a payment of $19,487 in 7 years. The market interest rate
is 10%, so the PV of the obligation is $10,000. The portfolio manager wishes to fund the
obligation using 3-year zero-coupon bonds and perpetuities paying annual coupons
Duration of liability: Single-payment obligation with duration of 7 years
Duration of assets: The duration of the perpetuity is 1.10/.10 = 11 years. With a weight w
invested in the zero, the asset duration is: w 3 years + (1 w) 11 years
Asset mix such that assets and liabilities have same duration: w = 1/2
Fully fund the obligation: Purchase $5,000 of zero-coupon bond and $5,000 of perpetuity
Rebalancing
Suppose that 1 year has passed, and the interest rate remains at 10%. The portfolio manager
needs to reexamine his position: Is the position still fully funded? Is it still immunized?
First, examine funding:
PV of obligation = $11,000 (1 year closer to maturity)
Managers funds = $11,000: Zero-coupon bonds have increased from $5,000 to $5,500 with
the passage of time, and perpetuity paid its annual $500 coupons and remains worth $5,000
The obligation is still fully funded
Portfolio weights must be changed, however: Zero-coupon bond now has duration of 2 years,
while perpetuity duration remains at 11 years. Obligation now due in 6 years
The weights must now satisfy: w 2 + (1 w) 11 = 6 which implies that w = 5/9
To rebalance the portfolio and maintain the duration match, the manager now must invest a
total of $11, 000 5/9 = $6, 111 in the zero-coupon bond
Requires that entire $500 coupon payment be invested in the zero, with an additional
$111.11 of the perpetuity sold and invested in the zero-coupon bond

107

 Rebalancing portfolio entails transaction costs, so one cannot rebalance continuously

In practice, compromise between desire for perfect immunization (continual rebalancing) and
need to control trading costs (less frequent rebalancing)
Cash flow matching and dedication
If we follow the principle of CF matching we automatically immunize portfolio from interest rate
movement because bond CF and obligation CF exactly offset each other
Dedication strategy: Cash flow matching on a multiperiod basis
In this case, the manager selects either zero-coupon or coupon bonds that provide total cash
flows in each period that match a series of obligations
Advantage of dedication: Once-and-for-all approach to eliminating interest rate risk
Once CFs are matched, no need for rebalancing: Dedicated portfolio provides the cash necessary to pay the firms liabilities regardless of the eventual path of interest rates
CF matching is not more widely pursued because of constraints it imposes on bond selection:
Immunization/dedication are appealing to firms that do not wish to bet on general movements
in interest rates, but they may want to use bonds perceived as undervalued
Cash flow matching, however, places so many more constraints on the bond selection process
that it can be impossible to pursue a dedication strategy using only underpriced bonds
Firms looking for underpriced bonds give up exact and easy dedication for the possibility of
achieving superior returns from the bond portfolio
Sometimes, cash flow matching is simply not possible:
Pension fund obligated to pay out perpetual flow of income to current/future retirees would
need to purchase fixed-income securities with maturities ranging up to hundreds of years
Such securities do not exist, making exact dedication infeasible
Other problems with conventional immunization
Duration in Eq. (1) strictly valid only for flat yield curve (all payments discounted at same rate)
If yield curve not flat, then definition of duration must be modified and [CFt /(1+y)t ] replaced with
[PV of CFt ] calculated by discounting with the appropriate spot interest rate from the zero-coupon
yield curve corresponding to the date of the particular CF
Even with this modification, duration matching immunizes only for parallel shifts in yield curve
Multifactor duration models have been developed to allow for tilts and other distortions in the
shape of the yield curve, in addition to shifts in its level
Added complexity does not pay off in terms of substantially greater effectiveness
Finally, immunization can be an inappropriate goal in an inflationary environment:
Immunization is essentially a nominal notion and makes sense only for nominal liabilities
No sense to immunize an obligation that will grow with price level using nominal assets (bonds)
Active bond management
Sources of potential profit
Two sources of potential value in active bond management:
1. Interest rate forecasting
Tries to anticipate movements across the entire spectrum of the fixed-income market
If interest rate declines anticipated, managers increase portfolio duration (and vice versa)
2. Identification of relative mispricing within the fixed-income market
An analyst might believe that the default premium on one particular bond is unnecessarily
large and therefore that the bond is underpriced
These techniques generate abnormal returns only if insight superior to the market
Interest rate forecasters have a notoriously poor track record
Portfolio rebalancing activities can be characterized as one of five types of bond swaps
In 1st two swaps, investor believes that yield relationship between bonds/sectors is temporarily out
of alignment When aberration eliminated, gains realized on underpriced bond
The period of realignment is called the workout period

108

1. The substitution swap

Exchange of one bond for a nearly identical substitute
Substituted bonds: Essentially equal coupon, maturity, quality, call features, SF provisions
Swap motivated by a belief that the market has temporarily mispriced the two bonds, and that
the discrepancy between the prices of the bonds represents a profit opportunity
2. The intermarket spread swap
The yield spread between two sectors of the bond market seems temporarily out of line
E.g., if current spread between corporate/government bonds is considered too wide and is
expected to narrow, investor will shift from government bonds into corporate bonds
Consider carefully whether there is a good reason that yield spread seems out of alignment
3. The rate anticipation swap
Pegged to interest rate forecasting
If investors believe that rates will fall, they will swap into bonds of longer duration
Conversely, when rates are expected to rise, they will swap into shorter duration bonds
4. The pure yield pickup swap
Pursued not in response to mispricing, but to increase return by holding higher-yield bonds
When yield curve is upward-sloping, yield pickup swap entails moving into longer-term bonds
5. Tax swap
A swap to exploit some tax advantage
E.g., an investor may swap from one bond that has decreased in price to another if realization
of capital losses is advantageous for tax purposes
Horizon analysis
Horizon analysis is one form of interest rate forecasting
Select a particular holding period and predict the yield curve at the end of that period
Given a bonds time to maturity at the end of the holding period, its yield can be read from the
predicted yield curve and its end-of-period price calculated
Then, add coupon income/prospective capital gain to obtain the total return over holding period
Example of horizon analysis
A 20-yr maturity bond with a 10% coupon rate (annual) sells at a YTM of 9%. An investor
with a 2-yr horizon needs to forecast the total return on the bond over the coming 2 years
In 2 years, the bond will have an 18-yr maturity. The analyst forecasts that 2 years from
now, 18-year bonds will sell at YTM of 8%, and that coupon payments can be reinvested in
short-term securities over the coming 2 years at a rate of 7%
Calculation of the 2-year return on the bond:
1. Current price = $100 Annuity(9%, 20 years) + $1,000 PV(9%, 20 years) = $1091.29
2. Forecast price = $100 Annuity(8%, 18 years) + $1, 000 PV(8%,18 years) = $1,187.44
3. The future value of reinvested coupons will be ($100 1.07) + $100 = $207
4. The 2 -year return is [$207 + ($1, 187.44 $1, 091.29)]/1, 091.29 = 27.8%
Contingent immunization
Contingent immunization: Mixed passive/active strategy suggested by Liebowitz/Weinberger
E.g., the manager wishes to pursue active management but is willing to risk losses only to the
extent that the terminal value of the portfolio would not drop lower than $11 million
Manager can risk some losses at outset and start off with active strategy vs. immunizing
Key: Calculate funds needed to lock in FV of $11M at current rates, via immunization
If T = time left until horizon date, and r = market interest rate at any particular time, then
the value of the fund necessary to guarantee the min. acceptable terminal value is $11M/(1 +
r)T : This portfolio size, if immunized, will grow risk-free to $11M by the horizon date
This value becomes the trigger point
If and when the actual portfolio value dips to the trigger point, active management will cease
Contingent upon reaching the trigger, an immunization strategy is initiated instead, guaranteeing that the minimal acceptable performance can be realized

109

110

Feldblum: Asset liability matching for P&C insurers

The asset liability matching problem
If a life insurers writings are large and stable, and its investment returns are steady, then its profits will
depend primarily on internal pricing and external competition
Complications arise when interest rates change:
If rates decrease, insurers II may be insufficient to satisfy policy obligations
If rates increase, insureds may take policy loans or lapse their policies to obtain higher returns
elsewhere Insurer may be forced to sell bonds at capital losses to meet cash needs
Life insurers have responded to these uncertainties in three ways:
1. Write term vs. permanent policies Short duration of term policies reduces interest rate risk
2. Shift investment reward/risk to p/h through variable/universal insurance policies/annuities
3. Match durations/CFs of liabilities/assets to mitigate effect of interest rate fluctuations
Characteristics of P&C insurers
First-party coverages:
Insurers pay claims soon after the accident date
Such P&C policies are similar to term life insurance
Long-tailed commercial liability lines
GL, PL, Medmal, CAL, CMP
Average loss paid about four years after accident date
Moreover, investment risk on assets supporting loss reserves can not be shifted to p/h
Insurers seek ways to match their investment and insurance portfolios
Crucial differences between the characteristics of P&C and life insurance operations:
1. Nominal vs. inflation sensitive
Traditional life insurance and pension fund liabilities are expressed in nominal terms
P&C loss determined at settlement Inflation from accident/settlement dates impacts liability
In other words, P&C liabilities are inflation sensitive
2. ALM involves holding asset portfolio whose duration = duration of liabilities
P&C: Short duration assets whose returns vary directly with inflation (T-bills, CP)
But with normal, upward-sloping yield curve, these assets have lower returns than long duration
assets (corporate bonds) Balance benefits of immunization with overall portfolio yield
3. Many actuaries/financial analysts ascribe long durations to common stocks
True if one examines only CFs resulting from current and expected dividends, but not price
changes or expected dividend changes due to changes in interest rates
However, common stock prices are inflation sensitive, just as insurance liabilities between the
acquisition/occurrence and the disposal/payment dates
Asset/liability matching risk stems from factors other than interest rate changes
For common stocks: Systematic stock market fluctuations
For insurance liabilities: Contagion risks and changes in legal interpretation of coverage
4. P&C insurers do not face the same disintermediation problems that life insurers face
Regardless of interest rate changes, P&C insurers expect a steady stream of premium inflow
Moreover, P&C insurers do not segment funds. Investment returns must be sufficient for the
company as a whole, not for any given block of policies
5. Measurements of asset risk often concentrate on nominal returns
Long-term bonds with amortized book values show high/steady nominal returns
But except for SAP, P&C depends on real returns. By this measure, long term bonds are risky
P&C insurers have a difficult balancing act
Loss reserves vary directly with interest rates
Immunization theory for fixed income assets recommends holding short-term bonds
But these securities have lower yields than long-term bonds
Moreover, interest rates have a minor influence on insurance CFs, since premium income does not
vary greatly with investment returns for most LOBs
111

 As a result, insurers invest in long-term bonds, which are risky when marked-to-market
Yet insurers do not want to add investment risks to fluctuations of insurance u/w cycle
A common solution is to diversify into equity investments, such as stocks and real estate
Real estate holdings are limited by state statutes, illiquid, and require investment expertise
Common stocks reported at market values on AS Book values fluctuate more than bonds
Long-term bonds therefore are the investment of choice for P&C industry
Were bonds reported on AS at their market values, instead of amortized values, their actual riskiness
would be apparent, and insurers would invest more heavily in common stocks
Accounting rules influence security selection as much as operating income does
Federal income tax laws also influence financial portfolios
Tax rate on long term capital gains was lower than rate on net investment income
Current law taxes both equally
Growth stocks have lost their tax advantage over income stocks/corporate bonds
The reduced tax exemptions for municipal bond/corporate dividend income reduces the tax
advantages of these securities over corporate bonds, federal bonds, and growth stocks
The strengthening of the Alternative Minimum Tax rules reduces the tax advantages of municipal bonds and income stocks over corporate bonds, federal bonds, and growth stocks
Nominal versus inflation sensitive liabilities
Asset liability matching theory is grounded on two characteristics of conventional life insurance policies:
(i) Nominally valued liabilities and (ii) Disintermediation
1. Traditional life liabilities stated in nominal terms, but funded (partially) by investment returns
If the assumed interest rate used for pricing the policy is conservative enough, i.e. if it is
sufficiently below actual investment returns, interest rate changes pose little risk
But investment returns have become more volatile, and competitive pressures have forced
insurers to be less conservative in their interest rate assumptions
Interest rate changes have a large effect on life insurance profitability
2. Life insurers faced strong disintermediation risks in the 1970s
No disintermediation in P&C because policy terms are short/reserves do not accumulate
When a change in new money interest rates affects the market values of liabilities and assets differently,
the liabilities and assets are mismatched
The greater the mismatch, the greater the interest rate risk
This is a speculative risk, not a pure risk: Insurer can gain/lose from interest rate changes
Aggressive insurer, confident in forecasting rate changes, may seek asset/liability mismatch
A conservative insurer would attempt to match assets and liabilities more closely
Matching techniques
Two common methods of ALM: (i) Cash flow matching and, (ii) Duration matching
Cash flow matching
Creates an asset/liability portfolio impervious to interest rate changes
The insurer forecasts net insurance cash flows from its book of business and buys fixed income
securities whose coupons and maturities provide the needed monies at the needed times
Exact cash flow matching can be cumbersome, inefficient, and costly
One bond may closely match liability CFs, but alternative bond may provide better yield
Exact cash flow matching is worthwhile only when the benefits of risk reduction outweigh the
costs of lower yields and administrative expenses (rare)
Duration matching
Hedges against small interest rate changes
A change in new money interest rates has two effects on bond prices:
1. Coupons reinvested at new money interest rate (higher returns when rates rise)
2. The bonds price declines when interest rates rise and rises when interest rates fall
The relative importance of changes in reinvestment returns and prices depends on the bonds
term, its coupons, and the date of the liability payment
112

 At a bonds expiration date, only the par value is received

Interim price changes have no effect on the proceeds
But reinvestment rate changes do influence the final wealth
Macaulay duration
Macaulay duration of the bond: The point at which a small change in interest rates just
balance the change in bonds market value
Formally, the Macaulay duration of a bond, or of any group of assets or liabilities, is the
weighted average of the cash flow dates, where the weights are the PV of each CF
Consider a 10-year 10% annual coupon bond issued on 1/1/88. CFs consist of $100 coupon payments each January 1, and the $1,000 principal repayment on 1/1/97
At a 10% current yield, the bonds duration is:
1 90.9 + 2 82.6 + + 10 424.1
= 6.76 yrs
90.9 + 82.6 + + 424.1
At a 5% yield, the duration is:
1 95.2 + 2 90.7 + + 10 675.3
= 7.29 yrs
95.2 + 90.7 + + 675.3
Duration of two bonds = wtd avg. of individual durations, weights = current market prices
More precisely, the weights are the sums of the discounted values of each bonds cash flows.
Barring marketplace imperfections, these are the market prices of the bonds
Match assets/liabilities by equating their (overall) durations
First determine the duration of the liabilities, either by examining total annual CFs or by
computing the weighted average of the durations of all policies
Then purchase FI securities whose overall duration matches duration of liabilities
Duration matching eliminates asset/liability mismatch risk only for small interest rate changes
As the rates change, the bond durations change too Even when asset/liability portfolios
have the same duration at one interest rate, they may have different durations at another
Fortunately, interest rates change slowly, giving insurer time to rebalance asset portfolio
Duration matching for P&C insurers
First determine loss payout patterns by line of business
Data are available either from internal company reports or from statutory AS
Loss reserve payout patterns for the Schedule P LOBs can be determined for 10 years
Loss payouts after 10 years may be estimated by an exponential decay model
Accident
Year
1983

1
16.1%

2
15.3%

3
14.6%

4
11.9%

Development Year
5
6
7
9.2% 7.0% 5.3%

8
4.2%

9
3.3%

10
2.6%

11+
10.5%

Loss reserve durations then determined by discounting nominal loss at appropriate investment rate
The timing of loss payments within each CY, as well as the pattern of payments after the tenth
CY, do not have a major effect on the duration
For simplicity, assume that all loss payments are made at mid-year, and that loss reserves still
held after 10 years are paid out evenly over the subsequent five years
The General Liability 1983 loss reserve duration is therefore:
(16.1)(0.5)(1.12)0.5 + (15.3)(1.5)(1.12)1.5 + + (2.1)(14.5)(1.12)14.5
= 3.2 yrs
(16.1)(1.12)0.5 + (15.3)(1.12)1.5 + + (2.1)(1.12)14.5
Interest rate
Risk free rate fine for discounting loss reserves Separate insurance/investment returns
For duration matching, must use same interest rate for liabilities and new investable assets
Payout pattern
For liability durations, Woll uses the payout pattern for loss reserves, not for incurred losses
113

 The two types of loss payment patterns are quite different

Many losses paid during year of occurrence Not shown as reserve liabilities on year-end
statement. Losses outstanding as of 12/31 are generally slower settling ones
The loss reserve payout pattern is usually slower than the incurred loss payout pattern
The proper payout pattern depends on the type of matching:
If match assets held/liability obligations as of 12/31, use loss reserve payout pattern
If match assets purchased/liabilities incurred, use loss incurred payout pattern
The apparent conclusion is that to duration match a GL liability portfolio, you should invest in
medium term bonds with an average duration of 3.2 years. This conclusion is misleading
Asset/liability matching mitigates the risk of interest rate changes
But procedure outlined does not model true effects of interest rate changes on loss payments
Inflation sensitive cash flows
ALM either: (i) Balances insurance and investment nominal cash flows or (ii) Balances insurance
cash flows with changes in investment cash flows plus capital gains/losses
If losses sensitive to inflation through settlement date, then reserve asset with 0 duration
To eliminate influence of interest rate changes on net worth, invest either in short term securities
(T-Bills, CP) or in securities inflation sensitive (common stocks, real estate)
In practice, most reserves are not fully inflation sensitive through the settlement date:
WC indemnity payments are largely fixed at the accident date
BI wage loss/medical bills determined soon after accident, often 1-2yr before settlement
Nevertheless, general damages which form the bulk of insurance payments in GL, PL, Medmal, CMP liability coverages, and BI are inflation sensitive through settlement date
Inflation is increasingly important for insurance liability losses through the settlement date
When the losses are not easily quantified, such as pain and suffering awards, juries are
influenced by the value of money at the settlement date, not at the accident date
Medical bills depend on time of treatment, which falls between accident/settlement dates
Even in disability cases, the plaintiffs attorneys usually incorporate the effects of inflation and
expected earnings changes in the demand for damages
Property reserves, on the other hand, are not inflation sensitive
Since they are paid out quickly, however, they have equally short duration
Short-term commercial paper has a duration similar to that of GL loss reserves:
If interest rates increase, the ultimate expected loss payment increases
Reinvestment returns from the commercial papers maturity payments increase similarly
Since the assets are short term, there is little change in market price
Assets (commercial paper) and liabilities (GL losses) change in the same direction
If LC trends and commercial paper yields change by approximately the same amount, the
magnitudes of the asset and liability changes are also equal
This is by no means a recommendation for investments in commercial paper
Short-term CP provides lower yields/incurs higher transaction costs than other FI securities
Equity durations
Introduction
Securities whose annual payments and maturity values vary directly with inflation rates and
interest rates, such as common stocks, are like casualty insurance liabilities
Macaulay durations are misleading for asset/liability matching
Asset and liability durations help quantify effects of interest rate changes on market values:
In general, [proportional change in market value] [duration] [change in interest rate]
Long-term bonds more sensitive to interest rate changes than CP and T-bills
Example
Consider two zero coupon bonds, with five and ten year terms, respectively
Each has a par value of $1,000, and each pays 10% per annum
The issue prices, therefore, are 1, 000 (1/1.10)5 = $621 and 1, 000 (1/1.10)10 = $386
114

 A zero coupon bonds duration equals its term The two durations are 5 and 10 years
Change in Price = 1 Duration Price Change in Interest Rate
Accordingly, % change in market price for the 10-yr bond is double that for the 5-yr bond
Common stocks
Traditional measurement of common stock duration uses the DDM of equity valuation
This model views a common stock as a perpetual bond that pays dividends for an infinite term
If dividends grow at G% per annum, and values are discounted at K%, PV[stocks dividends] is:
(Current dividend)

(1 + G)
(K G)

The duration of a fixed income security equals the negative of the derivative of the natural log of
its PV with respect to the discount rate For common stocks:

d{ln[Current Dividend (1 + G)] ln(K G)}

1
=
dK
K G

These traditional measures of equity duration give meaninglessly high figures

Market price changes
What happens to common stock prices when interest rates change?
Long duration implies that stock prices shift strongly in opposite direction of the change
This is true for long term bonds, but it is not true for common stocks
Interest rate and inflation rate changes affect common stock prices in several ways
1. Value of the firm
In theory, the real value of the firms major assets should not be affected by inflation
If inflation and interest rates accelerate, the nominal value of the firm should increase
accordingly, so that its inflation-adjusted value remains constant
2. Supply and demand
In practice, the value of a firm depends on its revenues and costs
When inflation/interest rates accelerate, supply costs increase, but demand may/may not
If inflation is demand-pull, there is excess demand
If inflation is supply-push, demand may be weak
Moreover, rising interest rates encourage households to save, not consume, further reducing
demand. In such cases, the values of the firm and of its common stock will decline
3. Investment strategy
When interest rates rise, investors often shift their holdings from common stocks to longterm bonds, to lock in the high rates
The lessened demand for common stocks reduces their market prices
The first effect is long-term; the next two are short-term:
When inflation/interest rates rise, common stock prices decline at first, but rise later
We quantify this phenomenon by correlating inflation rates with:
(i) Common stock capital appreciation
The correlation coefficient between the long term government bond capital accumulation series
and the CPI is -50% The market price of long-term bonds varies strongly and inversely with
inflation rates (and by implication, with new money interest rates)
If the long term government bond capital accumulation series is lagged one year, the correlation
is reduced but is still strongly negative (r = -25%)
There is no rebound from the initial decline in market values
Rather, reduction in correlation coefficient results from:
(a) Market values of bonds move towards par values as time to maturity shortens
(b) Newly issued bonds not affected by past changes in rates
115

(ii) Long-term government bonds capital appreciation

Correlation coeff. between the common stock capital accumulation series and the CPI is -19%
Market price of common stocks varies inversely but weakly with inflation rates
However, if common stock capital accumulation is lagged 1yr, correlation > 0 (r = +17%)
Presumably, initial acceleration of inflation hurts most firms, and their equity values decline
After several months, however, firms adjust their costs and prices, so nominal common stock
values vary directly with recent inflation rates
In sum, stock prices sensitive to inflation ( casualty reserves, both track real value of money)
The effect of inflation upon common stock prices depends upon the type of firm:
A retail store can quickly increase prices when inflation accelerates
A municipal utility must apply for rate changes
How should a P&C insurer divide assets among common stocks/long-term bonds/short-term bills?
Common stocks/short-term bills have relationship to rate changes as liability loss reserves
Long-term bonds have long durations and expose insurer to risk of interest rate changes
Common stocks and long-term bonds provide higher yields than short-term bills do, allowing an
investment return more commensurate with liability loss cost trends
Common stocks expose the insurer to systematic stock market risks
Short-term bills and long-term bonds protect against this risk
Finally, long-term bonds entail the lowest transaction
Expected yields, transactions costs, insurance cash flows, and risks
Expected yields
Asset/liability matching deals with speculative risk: A mismatched portfolio may provide either
greater or lesser net income than a matched portfolio provides
The investment analyst deals with expected returns: Shortening the duration of the bond portfolio
generally reduces the expected investment income
The yield differences can be great: High quality long-term corporate bonds usually offer 2-5 %
points more than short-term Treasury bills and commercial paper
Transaction costs
Transaction costs are equally important
Long-term bonds entail the least expenses, particularly if the bonds are held to maturity
Common stocks involve greater trading expenses and larger investment departments
ALM requires continual monitoring of new money rates/rebalancing of portfolio
Disintermediation
For life insurers:
Interest rates decline: Investment returns may fall below guaranteed yield in policy
Interest rates rise: Insureds may withdraw funds by means of policy loans (disintermediation), forcing insurers to sell bonds at capital losses
Either way, the insurer loses
This is misleading
Guaranteed interest rates in whole life policies are so conservative that there is little chance
that they will exceed investment returns over the long term
And when new money interest rates rise above policy loan interest rate, insurers receive II
above their expected return regardless of the extent of disintermediation
The withdrawal of funds by insureds and annuitants to obtain higher returns from other investments
does not have a significant adverse effect on economic worth
Cash flows
When new money interest rates decline:
The portfolio investment return exceeds the new money investment return
Liability LC trends decline as well
Net operating income increases

116

 When new money interest rates rise:

The portfolio investment return falls below the new money investment return
Liability LC trends rise above the expected levels
Net operating income decreases
The longer the duration of the portfolio, the greater the reduction in net operating income
Risk of statutory insolvency
Happens if the rise in interest rates depletes the insurers surplus
Amortization of long-term bonds on SAP statements reduces this risk (does not eliminate it)
To what degree are insolvencies caused by interest rate changes?
Financial portfolio with excessively long duration combined with rise in interest rates may
exacerbate an insolvent insurers negative net worth, but will rarely cause insolvency itself
Risks
Should the insurer invest in short duration common stocks or long duration bonds?
This question is not related to asset/liability matching. Each type of security presents its own risks
Common stocks expose the insurer to systematic market risks, such as market declines
Long-term bonds, if marked-to-market, expose an insurer to the risk of fluctuating inflation rates
The investment manager must balance these two risks in the financial portfolio
Conclusion
Liability reserves of P&C insurers are inflation sensitive
Similar to common stocks, other equity investments, commercial paper, and Treasury bills
CP and T-bills are highly liquid short-term investments, but yields are too low for large carriers
Real estate is limited by regulation and too risky for all but most experienced investment managers
Common stocks, therefore, are the apparent investment of choice
Interest rate changes do not pose a serious risk for stable P&C insurers
Long-term corp. bonds entail duration mismatch but offer higher yields than other FI securities
Asset/liability matching is but one investment concern. Traditional concerns remain paramount:
Maximizing expected returns
Ensuring safety of principal
Balancing the risks of each class of securities

117

118

Noris: Asset/Liability management strategies for P&C companies

Amount of cash flow

Market value of a P&C insurance company

Total market value of any firm derived from 2 sources: (i) Portfolio equity and (ii) Franchise equity
1. Portfolio equity = [value of currently booked assets] [value of currently booked liabilities]
2. The value of net income from business not yet booked by the insurance company is termed franchise
equity and is equivalent to the economically determined value for goodwill
The practical problems in measuring franchise equity preclude it from most analyses
Traditional emphasis on measuring the interest rate risk exposure of portfolio equity only
Portfolio equity, as proxy for the total value of the firm, does however represent an excellent target
account for interest rate risk management
The size of portfolio equity for a P&C company depends upon the difference between the cash flows
from currently booked assets (its investment account) and the amounts needed to pay insurance claims
on currently booked coverages (its loss profile)
To analyze portfolio equity: Visual representation of all CFs in the year they are to be received/paid
We can aggregate the assets of a P&C company into a profile, as in Fig. 1
Muni coupons
Muni maturities

Corp coupons
Corp maturities

Stock
T-Bills
Paid loss

30
20
10
0

'84

'88

'92

'96

'00

'03

Figure 1: XYZ Casualty assets and liabilities

Asset flows are shown by layering the coupon/maturity CFs from each holding
The liabilities are put in the same context by placing the paid loss profile over the asset profile
Book value accounting: Statutory surplus
Book value accounting: Discount most asset CFs at yields at which originally purchased
Stocks would be shown at their current market value (as would any assets in default)
On the other hand, because the P&C industry does not utilize discounting of reserves, the firms
liabilities as represented by the paid loss profile are, in effect, discounted at 0% interest
Portfolio equity value that results from this calculation (see table below) is the Statutory Surplus
Current value accounting: Current value surplus
Some argue: Book value accounting inappropriate for companies competing in market value world
Since 1981, Progressive includes in its annual report to shareholders financial statements on both
a GAAP basis and a basis called Current Value/Total Return accounting
Current value refers to values on the balance sheet
Total return refers to values on the income statement
Progressive values all assets at current market values and includes unrealized capital gains/losses
(reduced by long-term capital gains tax rate) in both balance sheet/income statement
Neither SAP nor GAAP recognizes unrealized gains/losses on book value assets, and unrealized
gains/losses on stocks are shown only on the balance sheet (adjustment to surplus)
Progressive states that total return accounting most accurately reflects real economic results
Current value accounting is more volatile than statutory accounting and would have shown the
industry to have dangerously low amounts of surplus in 1974 and 1981
119

 Economic accounting: Market value surplus

Current value accounting goes too far in placing a market value on the firms surplus account
While we have put a market value on all of the firms financial assets, we have still ignored the
time value of money in valuing the firms liabilities
In a true economic sense, any losses that are to be paid in the future should be discounted in
a manner similar to the way we discounted future cash flows on assets
Any positive rate of interest used in discounting these reserves would, of course, lower the
present value of the loss reserve and, thus, increase the reported amount of surplus
What rate of interest should be used in discounting P&C reserves?
Possibilities are: (i) Firms historical yield on asset portfolio (recommendation of GAO), (ii)
Yields for bonds in current investment market, (iii) Conservative rate of interest to account
for potential variation in loss experience, or (iv) No discounting at all (SAP)
Here, we use the spot yields for municipal bonds in the current investment market as a proxy for
the after-tax investment yield available on current investments
Even though discounting an insurers liabilities always causes market value surplus (MVS) to be
higher than current value surplus (CVS), and often higher than statutory surplus, there have been
periods (1980-81) when MVS has been lower than statutory surplus
If the industry continues its traditional investment practice of investing in long-term bonds, MVS
will continue to be highly volatile. There exists, however, an alternative: Company management
can choose to manage the volatility of MVS and help insure the profitable growth of the firm

Assets:

Liabilities:
Statutory Surplus:

Assets:

Liabilities:
Current Value Surplus:

Assets:

Liabilities:
Market Value Surplus:

Statutory Surplus (12/31/83)

Description
Discounting Discount Rate
Money Market
Yes
Cost
Taxable Bonds
Yes
Cost
Tax-exempt Bonds
Yes
Cost
Stock
N/A
Market
Total Assets
Loss Reserves
No
N/A
Current Value Surplus (12/31/83)
Description
Discounting Discount Rate
Money Market
Yes
Market
Taxable Bonds
Yes
Market
Tax-exempt Bonds
Yes
Market
Stock
N/A
Market
Total Assets
Loss Reserves
No
N/A
Market Value Surplus (12/31/83)
Description
Discounting Discount Rate
Money Market
Yes
Market
Taxable Bonds
Yes
Market
Tax-exempt Bonds
Yes
Market
Stock
N/A
Market
Total Assets
Loss Reserves
Yes
After-Tax

Value
$7,353
50,350
58,822
30,529
$147,054
$95,960
$51,094
Value
$7,353
43,765
52,056
30,529
$133,703
$95,960
$37,743
Value
$7,353
43,765
52,056
30,529
$133,703
$78,631
$55,072

Managing market value surplus

Reasons to use Market Value Surplus (MVS)
A firms MVS is a leading indicator of the future book value of the firm, since book value converges
to market value as asset and liability items mature
Managing the MVS enables company management to better serve the goals of the companys
owners, i.e., higher future stock prices for stockholders/higher future dividends for p/h
120

Surplus

Assets and liabilities at market (market value surplus - MVS)

Assets at market (current value surplus - CVS)
Assets at book (statutory surplus)

50
40
30
20
10
0
'72

'74

'76

'78

'80

'82

Figure 2: XYZ Casualty surplus values

Another reason for managing MVS may come from insurance regulators who, from time to time,
have become concerned with the capital adequacy of the industry
A 1979 report commissioned by NAIC did make a number of observations as to the appropriate
size and valuation of an insurers surplus account, concluding that statutory accounting is
appropriate only if asset/liability matching is adhered to:
. . . more attention must be paid to the matching of asset and liability maturities. If liabilities
are reasonably stated, and if matching is reasonably carried out, then amortized values will
not result in significant surplus distortions
. . . The required surplus for insurers who keep their asset and liability maturity schedules in
balance with each other should be less than for those who choose to invest in longer maturities,
thereby accepting the additional risk of market price fluctuations
Interest rate sensitivity of surplus: Duration gap
MVS net bond, albeit one with unusual series of positive/negative cash flows
These flows are represented by the amounts assets differ from liabilities
Like any bond, the MVS will have a price sensitivity to changes in interest rates: The longer the
bond is, the more sensitive its price will be to a given change in interest rates
Duration of MVS account based upon additive property of duration: Duration of combination of
instruments = sum of durations of individual instruments weighted by their market values
Denoting M VA the market value of insurer assets, and M VL the market value of its liabilities:
M V S = M VA M VL

(1)

We weight each components duration (D) by its market value (M V ) MVS duration:
Dmvs =

(Dmva M VA ) (Dmvl M VL )
MV S

(2)

Duration gap
The amount that MVS duration varies from zero is known as the firms duration gap
Insurers with a larger duration gap will have a market value surplus that is more susceptible
to changes in interest rates than firms with a gap of zero
Positive duration gap (assets longer than liabilities): Any rise in interest rates would lower the
absolute value of MVS (M VA would decline relatively more than M VL )
Negative duration gap: Rising interest rates would actually improve the MVS
The table below develops the duration gap for XYZ Casualty as of year-end 1983
Durations for FI assets (bonds and money-markets) are calculated using traditional method of
weighing their various CFs terms-to-maturity by their PVs
The duration of a consol bond (perpetuity) is 1/i, where i is its yield to maturity
Duration of common stock 1/d, where d is the current dividend rate
121

Assets:

Liabilities:
MVS:

Market Value Surplus (12/31/83)

Description
Discounting Discount Rate
Value
Money Market
Yes
Market
$7,353
Taxable Bonds
Yes
Market
43,765
Tax-exempt Bonds
Yes
Market
52,056
Stock
N/A
Market
30,529
Total Assets
$133,703
Loss Reserves
Yes
After-Tax
$78,631
$55,072

Duration (yrs)
1.00
7.33
8.41
23.26
11.04 yrs
2.51 yrs
23.22 yrs

XYZs positive duration gap of 23.2 years indicates that its MVS will have the interest rate sensitivity equal to that of a 23.2 year zero-coupon bond (very long term bond)
A rise in interest rates will cause a large decline in MVS
Any decline in interest rates will sharply increase MVS
Duration gap of target accounts
Rather than carry a large duration gap, which leaves the value of the firms MVS greatly dependent
upon level of interest rates (uncontrollable factor), insurer should manage this gap
By deciding when, and to what extent, to have a duration gap, company management can position
the firm to take advantage of any projected changes in interest rates or, in an extreme case, to
position the firm to have a MVS that is immune to changes in interest rates
Duration gap of surplus
Duration gap computed in Eq. (2) and the table above is the Duration Gap of Surplus (DGs )
in that it tells interest sensitivity of insurers market value surplus account
A natural target for managing DGs would be to achieve a DGs that is equal to zero, which
would mean that the value of MVS will be immune to changes in interest rates
Having DGs = 0 is valuable to most insurers since insurance regulators limit premium volume
to small multiple of reported surplus If surplus is managed so that it never declines from
changes in interest rates, then premium volume not constricted
However, DGs = 0 results in fluctuations in total earnings, depending upon interest rates
Duration gap of total return on surplus
Duration indicates the holding period over which a rate of return can be immunized
If duration desired holding period, then the initial promised return can be realized independent of changes in interest rates By setting a firms DGs equal to some holding period, the
net yield of MVS can be achieved over that period
Many insurers desire to manage annual returns-on-surplus so that they are always positive
Provided the net yield of MVS is positive, this goal can be achieved by setting DGs = 1
Immunizing a total rate of return over longer planning horizons can be achieved by having a
higher DGs (however, interim results may be above/below the immunized HPR)
The insurer must reach some compromise between holding-period return and stability in interim
results. In general, the Duration Gap of Total Return-on-Surplus (DGtrs ) is:
DGtrs = Ds H

(3)

Where Ds is the duration of surplus and H is the holding or investment period over which
management wishes to lock up the currently available return on surplus
By setting DGtrs = 0, the firm will immunize total return over the holding period
Duration gap of leverage
The ratio [MVS MVA] is another target account of interest for capital adequacy
Economic leverage: The [MVA MVS] ratio, the reciprocal of [MVS MVA] ratio
The economic leverage of a firm will remain unchanged only when both elements either remain
unchanged or change proportionately
Since the % change for the market value of any security (given a change in interest rates)
depends upon its duration, the duration of surplus must equal the duration of the assets to
immunize economic leverage against interest rate fluctuations
122

 To immunize economic leverage, set Duration Gap for Economic Leverage DGel = 0, where:
DGel = Dmvs Dmva

(4)

Miscellaneous issues
Nominal versus real return
Duration matching strategies above immunize a nominal return over some holding period
Targets are based upon expected dollar payment of claims and investment portfolio is then
constructed that will pay off at least that much in market value
But estimates of these expected claims incorporates only an expected rate of inflation
Unanticipated inflation may cause these reserves to be insufficient
Ideally, immunize a real rate of return that pays off in units of coverage/similar concept
Even if one can determine the correct rate of inflation to which returns should be indexed,
problem that traditional FI instruments, have flows denominated in fixed dollars
Unless there is some mechanism for these flows to increase with inflation, duration matching
strategies will not immunize a real rate of return
Indexed bonds, where either coupons/principal value change with inflation, may be issued in
the future and would be useful in immunizing real rates of return
To keep pace with inflation, buy investments with inflation dependent returns
Real assets, such as real estate, and common stocks come to mind
However, the problems of these assets volatility make their use as the core holding for an
insurers reserve position a difficult proposition
2nd method of keeping investment returns current with inflation: Investments that roll over often
If yields for similar investments in the new market environment are related to the inflation
rate, then rolling the portfolio into these new investments will achieve the inflated return
The problem with this strategy is that it requires a departure from the duration matching
principle, since shorter maturities are required if the portfolio is to be kept current
A far simpler method of assuring sufficient assets for inflated claims is to intentionally overestimate
the expected size or rapidity of the losses
The amount of the overestimation could be earmarked as a contingency reserve and would be
invested in the same manner as the market value of the estimated loss reserve
E.g., AAA/SOA recommendation is to apportion assets of a life insurer among:
1. Valuation Reserves: These reserves would be sufficient under expected circumstances,
but there is still some probability that additional reserves are required
2. Contingency Surplus: This is the amount of surplus (assets minus valuation reserves)
that is required to bring the probability of ruin down to an acceptably low level
3. Vitality Surplus: This remaining portion of assets = amount available for growth change:
Use the vitality surplus for riskier investment strategies
In practice, the hypothetical contingency surplus is held in form of short-term assets:
These short-term assets provide for CAT claims, and also for protection against unanticipated inflation as they are rolled over in differing interest rate environments

123

124

Panning: Managing interest rate risk

Summary
Objective of ALM: Measure/manage degree to which economic value of insurer is adversely exposed to
changes in interest rates ALM = component of ERM (ERM considers impact of more variables)
ALM fails to take into account franchise value: The economic value to the firm of future renewals
Franchise value is not recognized by accounting rules, but can be a significant portion of an insurers
total economic value, which is reflected in its market value
The problem is to identify a strategy that limits a firms exposure to interest rate risk while simultaneously limiting its exposure to accounting rules that could jeopardize its solvency or its ratings
Solution: Adopt a pricing strategy that controls interest rate exposure of future CFs from new business
Introduction
CEOs and CFOs believe that they are managing the value of the firm:
What they actually are doing is managing the portion of their firms value visible to them
Considerable portion of firms value invisible to management (not included in accounting numbers)
ALM can achieve its stated objective of protecting the value of the firm only if it:
Recognizes this invisible portion of a firms value
Makes it visible to senior management
Helps them to understand how to manage it effectively
Economic realities
Selling policies by direct marketing costs considerably more than selling through an agent, who
receives a sales commission (in fact, expected profit for directly-marketed policy < 0)
But by selling 1st policy at a loss, one get a customer highly likely to repeatedly renew
Renewals highly profitable (no marketing costs or agent commissions) Loss incurred with 1st
policy more than offset by future profits from subsequent renewals
From economic standpoint, prospective renewals are a valuable asset, called franchise value
But accounting rules typically do not recognize franchise value
An important exception occurs when a firm is sold for more than its book value, in which case
the excess is booked by the purchaser as an asset called goodwill
To accountants, programs return would be increased by reducing sales during the loss-producing
early years of the plan and increasing sales during the profitable later years of the plan
Economic model demonstrates conclusively that a better strategy is to grow as quickly as possible
in early years, despite accounting losses, so as to maximize profits from renewals later on
Several crucial implications for ALM
1. Franchise value is real
Policy renewals have real economic value even though accounting rules refuse to recognize it
Such franchise value is typically, although imperfectly, reflected in an insurers stock price
2. Because franchise value consists of the PV of expected future CFs from renewal business, it is
exposed to interest rate risk and as such should be recognized by ALM
3. Franchise value typically invisible to senior executives Unmeasured, unreported, and unmanaged
Until franchise value is recognized, measured and reported, ALM will remain incomplete
A simplified insurance firm
Financial models are indispensable:
(i) For creating understanding and, (ii) For applying that understanding to actual situations
For understanding, simple models are best. Transparency enables to appreciate their virtues/flaws
By contrast, the models we build to apply this understanding to actual situations are necessarily
far more complex, for they incorporate many more aspects of reality
Questions
How significant is franchise value as a component of a firms overall value?
How sensitive is franchise value to interest rate risk?
What strategies can firms use to protect franchise value from interest rate risk? Which is best?
125

 Characteristics of the model

The firm writes all of its business on January 1 of each year
It pays all expenses for the year on that same day
On 12/31 of each year, it learns the true value of losses (and LAE) on policies it wrote in January,
and pays those losses that same day AY, PY and CY are identical for this company
The firms expenses and expected losses are identical every year
If firm has made a profit, it immediately dividends that amount to shareholders. If it has incurred
a loss for the year, it immediately raises equity to restore its surplus to initial amount
Consequently, its surplus is identical every year
Ignore possible insolvency, taxes and the potential costs of raising capital
Assume, for convenience, that the term structure of interest rates is flat
All calculations below occur on 1/1, right after firm has written its new business for the year
Notations
P:
E:
L:
y:
S:
k:
cr:
F:
C:

WP on policies that the firm writes every year; P can vary yearly
Expenses, in dollars, that the firm pays each year; E is constant
Loss and LAE, in dollars, that the firm expects to pay each year; it is constant
Risk-free rate, to calculate income/discount future CFs; No default risk
Firms surplus, same every year (due to dividends and recapitalization)
Firms target return on surplus
Client retention, the % of clients who renew their policies from one year to the next
Firms franchise value, the PV of CFs from future renewals
Firms current economic value, the PV of surplus and business already written

The untaxed net income for this simplified firm is:

Net Income = P L E + y (S + P E) = kS
The fact that premiums can vary plays an important role in the firms exposure to interest rate risk
Firm varies its premiums from year to year so as to achieve a target dollar return on surplus (kS)
The firm will achieve its target return on surplus by setting the premiums it charges to:
P =

S(k y) + L
+E
1+y

The value of the firm

Firms current economic value C, the economic value of its current balance sheet on 1/1, is the value of
its current assets: Surplus + premiums expenses discounted expected losses
C =S+P E

L
1+y

Franchise value
The firms franchise value F is the PV of CFs from its future renewals, taking into account both
the time value of money and the firms client retention rate cr
If interest rates and the target return on surplus remain unchanged, then the values of P , L, and
E in a given year will be followed by the values P cr, L cr, and E cr in the subsequent year
Using the multiplier d = cr/(1 + y), the PV of future premiums = P (d + d2 + + dn )
As n , the PV of future premiums converges to: P d/(1 d) = P cr/(1 + y cr)
Similarly, the PV of future expenses associated with retained business = E d/(1 d)
Losses are paid a year later than premiums and expenses, so their PV = [L d/(1 d)]/(1 + y)
These three components of future renewals are combined to give the firms franchise value:
F = [P E L/(1 + y)] d/(1 d)
Add franchise value to the firms current economic value Total economic value of the firm. We
consider this to be identical to the firms total market value/market cap if publicly traded
126

 The market-to-book ratio varies with client retention. At high retention levels, the ratio climbs rapidly
The potential importance of franchise value is illustrated by its % of firms total market value (F + C)
When client retention 80%, franchise value 20% of the firms total market value
Franchise value is significantly affected by:
The level of interest rates
The firms target return on surplus
And, most important, by its pricing strategy
The interest rate sensitivity of franchise value
Franchise value sensitive to interest rate risk Calculate the duration of franchise value
Because the premium component of franchise value depends on the firms pricing policy (which can
depend on interest rates), we first describe how firms target return on surplus k is determined
The return on surplus k may be fixed or may depend on current interest rates
Here we assume that k = a + by, where a and b are constants for a given firm
y = spot interest rate corresponding to the maturity of firms liabilities (here, one year)
Setting fixed target return can be problematic: Interest rates may rise to exceed that level (early 80s)
A more pragmatic pricing policy may therefore be to set the target return as a risk-free rate of
interest plus some risk premium, so that b = 1 and a is the risk premium
Premiums may vary with the level of interest rates with pricing policies where b 6= 0
Given firms target return on surplus k, restate the firms franchise value as:
F =

cr S [a + (b 1)y]
(1 + y)(1 + y cr)

By definition, the duration of F with respect to changes in interest rates is the negative of the first
derivative of F with respect to y, as % of the current value of F :
D=

1 dF
ab+1
1

=
+
F dy
(1 + y)(a + by y) 1 + y cr

The dollar duration of franchise value is the product of premium present value and its duration, less the
comparable products for losses and expenses
The duration of franchise value is equal to its dollar duration divided by its present value

Premiums
Losses
Expenses
Total

Annual
Value
101.19
-75.00
-25.00

Present
Value P V
607.14
-428.57
-150.00
28.57

Duration D
7.85
7.62
6.67
17.62

Dollar
Duration P V D
4,768.71
-3,265.31
-1,000.00
503.40

The duration of future premiums is significantly higher than the duration of losses and expenses
The explanation for this is that premium cash flows are interest-sensitive
When interest rates rise, premium cash flows become smaller due to the particular pricing policy
we have assumed in our example (constant target return of 15%)
When premiums are interest-sensitive, a rise in interest rates has a double impact:
(i) PV of each dollar of future premiums decline, (ii) Volume of future premiums also declines
The first of these two effects is unavoidable when interest rates change. But the magnitude of the
second effect can be changed by adopting a different pricing strategy
Managing the interest rate risk of franchise value
Assume current assets (Surplus + WP expenses) are invested in a portfolio with duration = 1-yr
Liabilities also have a maturity of one year and a duration just less than one year
The duration of its current economic value (54.76) is one year
Taking franchise value of 28.57 into account means that firms total economic value = 83.33, or
52% larger than its current economic value. Franchise value has a duration of 17.62
Firms total economic value has a duration of (54.76 1 + 28.57 17.62)/83.33 = 6.70
127

 How can the firm go about reducing that duration?

One way, the traditional approach, would be to reduce the duration of its invested assets
Done either by changing the composition of the firms investment portfolio, or by purchasing
derivative securities that modify the firms asset duration
Suppose the firm reduces duration of invested assets to zero Reduce duration of total
economic value to 5.18, which the firm may still consider unacceptably high
If the firm had a higher client retention percentage, the problem would be even greater
Practical dilemma that has two aspects:
1. The greater the franchise value of a firm, the more difficult it is for that firm to manage the interest
rate risk of its total economic value by reducing the duration of its investment portfolio
2. Further problem with such strategy: Potential benefits of implementing it totally invisible to
regulatory authorities/rating agencies, who see only the accounting numbers of a firm
Using pricing strategy to manage total economic value
Adopting a pricing strategy that substantially alters the sensitivity of a firms total economic value to
changes in interest rates is a solution to the dilemma just posed
Suppose we alter the firms pricing policy by changing these parameters to a = 10% and b = 1
The target return on surplus remains at 15% (risk-free yield remains at 5%), but durations change
from 17.62 to 7.62 for franchise value, and from 6.70 to 3.27 for total economic value
Key insight: A firms pricing strategy can significantly affect the duration of its franchise value
and, consequently, the duration of its total economic value
More systematic approach to managing the duration of total economic value:
Find a combination of the strategy parameters a and b such that the return on surplus and the
duration of total economic value are both acceptable
Can be done either by systematic numerical search or by constrained optimization procedures
Managing duration of total economic value with pricing policies has limitations/advantages:
Limitation
Any desired combination of a target return on surplus and target duration of total economic
value can rigidly be maintained only for a rather narrow range of interest rates
Large changes in interest rates will necessarily disrupt the combination initially established
Same limitation virtually ubiquitous in ALM due to nonlinearity of prices vs. interest rates
Advantage
Avoids rating agency/regulatory risk associated with focus on managing duration of invested
assets to manage the risk to its franchise/total economic value
This key advantage results from the fact that implementing a pricing strategy is nearly as
invisible to these external audiences as the franchise value it is intended to protect
Conclusion
Firms exposure to interest rate risk depends on its assets/liabilities and interest rate volatility
Two additional insights:
Relying on traditional accounting rules to identify a firms economic assets and liabilities can blind
us to the importance of franchise value, the PV of an insurers future renewals
Firms exposure to interest rate risk also depends on the pricing strategy of the firm
Conventional understanding also recognizes that the way to manage a firms exposure to interest rate
risk is to select or alter the composition of its assets and liabilities
However, appropriately chosen pricing strategy avoids potential difficulties in protecting franchise
value, and can be flexible in achieving a targeted duration/return on surplus
Although pricing strategy has its limitations as a tool for ALM, these limitations arise from pricing
nonlinearities that likewise afflict the more conventional methods typically employed
Risk of equity portfolio can be altered by an appropriate dynamic investment strategy
Generalized to using dynamic pricing strategy to manage interest rate risk of franchise value

128

Additional Notes

129

Additional Notes

130

Additional Notes

131