1.

Introduction to Derivatives
1. Defining Derivatives
(a) A derivative security is a financial security whose value depends on other, more fundamental,
underlying financial variables such as

 stock price
 interest rate
 index level
 commodity price
 exchange rate
The value of the derivative will change whenever the underlying variable changes. To get some intuition
for this, let’s pretend that the underlying variable is stock price of company ABC, and the value of the
derivative and the stock price are related by the following formula:

derivative value= {ABC stock0price−100 if ABC stock price<100

if ABC stock price ≥ 100
The value of the derivative does not equal the value of the underlying variable but is related to it.
(b) There are three basic types of derivative securities:

 Futures and forwards: Agreement to a trade on a specified future date

 Swaps: Agreement to a series of exchanges on specified dates in the future
 Options: The right to engage in a trade on a future date for the option buyer, and the obligation
for the option seller to take part in the trade if the option buyer exercises her right
The table below defined each derivative in greater detail and mentions some useful terminology:

Forwards/futures
Agreement between two parties (the buyer is “long”, the seller
“short”)
to buy or sell a specified good in a
specified quantity (the “underlying”)

at a specified price
(the “delivery price”)
on a specified future date
(the “maturity date”)

1
 Options
The right to buy or sell (Call option gives right to buy and Put
option right to sell; option buyer is
long, seller or “writer” is short)
a specified asset in specified quantity
(the “underlying”)
at a specific price
(the “strike” or “exercise” price)

on or before a specified date

(the “maturity” or “expiration” date)

Swaps
Agreement (e.g., cash flow exchanges based on
between two parties for periodic interest-rates, currencies, commodity
exchanges of cash flows prices, and equity prices)

on specified dates T 1 , T 2 , … , T 10

and calculated using specified rules (e.g., rules regarding frequency of

payments, semi-annual versus
quarterly, for each party)

(c) All derivatives have some relationship to agreements about the future. Futures, forwards and swaps
enable investors to lock in future cash flows.
Futures example: A company that needs crude oil in one month can lock in a future price by using a
futures contract.
Swap example: A company that has borrowed at floating interest rates can lock in fixed interest rate
payments by entering into a swap.
By locking in future cash flows, the company is reducing or “hedging” risk associated with changes in
crude oil prices or interest rates.
On the other hand, Options provide protection to the option buyer, in other words, they offer financial
insurance.
Option example: A company that needs crude oil in one month can buy an option giving it the right to
buy oil for $100 per barrel in one month. If prices rise above $100 per barrel, the company can use the
option, and if it falls below $100, the company does not have to use the option.
By having the option to buy oil for $100 per barrel in one month, the company has insured itself against
losses associated with price rising above $100 per barrel.

2
2. Payoffs for Forwards, Futures, and Options
(a) We have seen that the value of a derivative depends on the value of an underlying financial variable.
To understand this valuation, we first needs to understand the payoff a derivative will provide at its
maturity date.
(b)(i) Forward example 1: A investing company enters into a forward contract with a bank to buy 1000
shares of XYZ stock for a price of $100 per share in one month ( F=100). We say the company is “long”
and the bank is “short” with respect to the contract. Depending on what the stock price at maturity ST is,
the company may have gained or lost on the trade as a result of locking in a price of $100 per share. The
table below summarizes various scenarios for the payoff per share:
Forward payoff per share at maturity
Delivery price F=100
40

Stock Price in Payoff per share Payoff per share 30

one month to Long to Short Long
20 Short
70 -30 +30
80 -20 +20 Payoff per share 10
90 -10 +10
100 0 0 0
60 80 100 120 140
( ST =F ) -10
110 +10 -10
-20 ST = F
120 +20 -20
130 +30 -30 -30

-40
Stock price at maturity

The payoff in one month to long is equal to

payoff ¿ long (forward )= ⏟T

S − ⏟
F
(1)
stock price delivery price
at maturity date

and the payoff to short is equal to

payoff ¿ short =F−S T
We see that one party’s loss is the other party’s gain.

(b)(ii) Forward example 2: A US-based exporter forecasts that it will make €200 million in Europe. It
will be converting the sales to US dollars and would like to lock in an exchange rate to avoid cash flow
uncertainty. It enters into a forward contract with a bank to sell €200 million in six months for $1.3/€. It
will be able to buy
€200 million x $1.3/€ = $260 million,

3
and the bank will receive €200 million.
Depending on the exchange rate in six months, the company will gain or lose money as a result of the
forward contract. The table below summarizes the situation:
Exchanged rate in six Company (Selling Euros)
months USD for €200 Gain
$1.1/€ $220 +$40
(=€200 x 1.1) (=$260 - $220)
$1.2/€ $240 +$20
$1.3/€ $260 $0
$1.4/€ $280 -$20
$1.5/€ $300 -$40

(c) Given that entering into a forward contract could mean a gain or loss for either party, what would be a
fair price for the contract? It would be a price that would advantage neither of the parties, that is, a
breakeven price that would make the contract have zero expected value to both parties at the time of the
agreement.
(d)(i) Put option example: An investor plans to sell Cisco stock in a month’s time. The stock is currently
$25. The investor is worried that the price might fall below $23 in this period. To protect herself against
price declines, she buys a one-month put option with strike price of $23. Her payoff in one month will
depend on how the stock price moves. The table below summarizes various scenarios:
Put Option Payoff per Share at Maturity
Strike price K=23
4

Stock Price in 3
one month Payoff to Long Payoff to Short Long
2
20 +3 -3
Payoff per share

21 +2 -2 1
22 +1 -1
0
23 0 0
20 21 22 23 24 25 26
( ST =K ) -1
24 0 0 -2 ST = K
25 0 0 Short
26 0 0 -3

-4
Stock price at maturity

The payoff in one month to long is equal to

{
0 if ST ≥ K
payoff ¿ long= ⏟
K − S⏟T if ST < K
strike price sock price
at maturity date

This can be written in a single line as

4
 payoff ¿ long ( put )=max ( K −S T , 0 ) . (2)

The payoff to short is

payoff ¿ short =min ( S T −K , 0 ) .

(d)(ii) Puts provide sellers of the underlying insurance against declines in the underlying’s price. The
higher the strike or the longer the maturity, the greater the amount of insurance provided by the put. The
higher strike price means that the put option buyer will be able to sell the underlying for a higher price,
increasing the option’s value; and the longer maturity means that there is a higher probability that the
underlying’s price will move in a favourable direction, also increasing the option’s value.
(e) Call option example: Apple stock is trading at $220. An investor is planning to buy stock in a month’s
time, and is concerned that the price could rise sharply over that period. To protect herself against price
rises, the investor buys a one-month call on Apple with a strike price of $225. Her payoff in one month
will depend on how the stock price moves. The table below summarizes various scenarios:
Call Option Payoff per Share at Maturity
Strike price K=225
4

Stock Price in 3
one month Payoff to Long Payoff to Short 2 Long
222 0 0
Payoff per share

223 0 0 1
224 0 0
0
225 0 0
222 223 224 225 226 227 228
( ST =K ) -1
226 +1 -1 -2 ST = K
227 +2 -2 Short
228 +3 -3 -3

-4
Stock price at maturity

The payoff in one month to long is equal to

payoff ¿ long=
{ S T −K
0
if ST ≥ K
if S T < K

This can be written in a single line as

payoff ¿ long(call)=max ( S T −K , 0 ) . (3)

The payoff to short is

5
 payoff ¿ short =min ( K−ST , 0 ) .

(e)(ii) Calls provide buyers of the underlying insurance against increases in the underlying’s price. The
lower the strike or the longer the maturity, the greater the amount of insurance provided by the call. In the
case of a call, the lower strike price means that the option buyer will be able to buy the underlying for a
lower price, increasing the option’s value; and the longer maturity means that there is a higher probability
that the underlying’s price will move in a favourable direction, also increasing the option’s value.
(f) If the option buyer obtains insurance, what does the option seller (the “writer”) who provides the
insurance get? The writer receives a fee called the “option price” or the “option premium”. The charts
below illustrate the difference between payoffs and net payoffs (i.e., profits) for the call option example.

Payoff verus Net Payoff for Call Option

Exercise price E=225, Option premium = $1

Payoff Net Payoff (Profit)

4 2.5
3 2.0
Long 1.5 Long
2
1.0
Payoff per share

Profit per share

1 0.5
0 0.0
222 223 224 225 226 227 228 -0.5 222 223 224 225 226 227 228
-1
-1.0
-2 ST = K ST = K
Short -1.5 Short
-3 -2.0
-4 -2.5
Stock price at maturity Stock price at maturity

3. Comparison of Forwards, Futures, and Options

(a) The table below summarizes the differences between forwards and futures:

Criterion Futures Forwards

Buyer-seller
Via exchange Direct
interaction
Contract terms Standardized Customizable
Unilateral
Possible Not possible
reversal
Who bears Exchange Individual

6
 default risk? parties
How is default Margin
Collateral
controlled? accounts

We will look at futures markets in greater detail in the next note. “Unilateral reversal” of a futures
contract means that a trader can close out a futures position at any point up till contract maturity without
requiring approval from the counterparty. This is achieved by making a trade that is the reverse of the
trader’s position. For example, a trader would close out a futures contract to buy 1000 barrels of crude oil
by entering a futures contract to sell 1000 barrels of crude oil. The net effect of the trade reversal is the
trader receiving the value of the long position at the time of closeout. The position’s value could be either
positive or negative, and depends on comparing the delivery price to the current forward price (this will
be explained in more detail later).
(b) There is no cost associated with the forward or futures contract itself, as the contract has zero value at
inception for both the buyer and seller. On the other hand, there is an option cost or premium associated
with buying an option, which is paid to the seller. This doesn’t mean that trading in forwards and futures
is free. There will always be transaction costs.
(c) Long and short positions in forwards and futures are obligations, whereas the option buyer has a right,
but the options seller doesn’t. The option seller is obliged to engage in the trade when the option buyer
chooses to exercise her right.
(d) Forwards and futures have linear payoff profiles, whereas call and put option payoffs exhibit a
“hockey stick” shape.

4. Using Derivatives to Speculate

In addition using options to provide protection against price declines or rises (“hedging”), investors might
also use them to bet against the market in order to make a profit (“speculation”):
Call option (option to buy) for speculation example: IBM stock is trading at $150. John and Edison both
speculate that prices will rise to $160 in the next 3 months. John buys 100 call options with a strike price
$150 and Edison buys 100 shares of stock with the hope of making money. Each call option costs $5. The
table below compares their returns after 3 months depending on how the stock price moves.

Profit per call/share Return

Stock Price
in 3 months Call strategy Stock strategy Call strategy Stock strategy
-100% -2.7%
146 -5 -4
(¿−5/5) (¿−4 /150)
148 -5 -2 -100% -1.3%
150 -5 0 -100% 0.0%
152 -3 2 -60% 1.3%
154 -1 4 -20% 2.7%
156 1 6 20% 4.0%
158 3 8 60% 5.3%
160 5 10 100% 6.7%

7
The table above shows why using options for speculation is extremely risky: You may make a lot of
money, and you may lose everything. The risk is magnified even more if borrowed money is used to pay
for the strategy.

2. Pricing Forwards and Futures

1. Key concepts for pricing derivatives

(a) There are two key concepts we need to understand before talking about forward prices:

 No-arbitrage: You cannot make riskless profits, or, equivalently, obtain guaranteed net cash
flows. Any arbitrage opportunities can only be temporary, and cannot persist.
 Replicating portfolio: A portfolio of the underlying asset (and perhaps cash) that mimics the
cashflows of the derivative.

(b) If you can find a replicating portfolio, then by the principle of no arbitrage the value of the portfolio
and the derivative should be equal.

value of derivative=value of replicating portfolio .

2. Replicating portfolio and Forward Price

(a) Consider a forward contract for purchasing 1000 barrels of oil in one month’s time. You will pay a
price F for the asset to possess it in the future. Alternatively, you could borrow at the risk-free rate
today and buy the asset for the spot price S and hold onto it until you need it in a month’s time. Holding
onto the asset will cause you to incur holding costs such as storage and insurance. The two approaches
should have equal cash flow in present value terms:

PV ( forward price of asset )=PV ( spot price of asset +holding costs )

⇒ PV ( forward price of asset )=spot price of asset + PV ( holding costs )
Why should these two be equal? Suppose

PV ( forward price of asset ) > PV ( spot price of asset + holding costs ) .

You could short the forward contract and buy the asset today by borrowing at the risk-free rate. In a
month’s time you would sell the asset at the forward price, and the present value of your cash inflows—
PV(forward price of asset)—would be greater than the present value of your cash outflows—PV(spot
price of asset + holding costs). This is an arbitrage opportunity for making riskless profits.

Similarly, suppose

PV ( forward price of asset ) < PV ( spot price of asset + holding costs ) .

You could sell the asset at the spot price and invest this amount plus at the risk-free rate, and long the
forward contract. By selling the assets you would not incur holding costs, so your cash inflows would be
the spot price of the asset plus the holding cost savings. In a month’s time you would buy the asset at
the forward price, and the present value of your cash inflows—PV(spot price of asset + holding cost

8
savings)—would be greater than the present value of your cash outflows—PV(forward price of asset).
Once again, this is an arbitrage opportunity for making riskless profits.

Hence, we can write the following formula:

⏟
PV ( F )= S⏟ + M
⏟
pvof spot pv of net
forward price holding
price today costs

⏟
⇒ F=FV ( S+ M )
future value

(b) Let us consider the term PV ( F ) in the first formula, there are two common approaches used for
interest rates when discounting the forward price to compute the present value: continuous
compounding and money-market convention. We will first explain what continuous compounding is and
then apply it get an expression for the present value of the forward price.

Continuous compounding: The usual formula for the discount factor given an annualized interest rate r ,
a horizon of T years, compounded over N periods is as follows:

1
.
( )
N
rT
1+
N
For example, an annualized interest rate of 3%, a horizon of 3-months (0.25 years), compounded over 3
periods would have a discount rate of

1 1
= =0.99254
( ) ( )
3 3
0.03 ×0.25 0.03
1+ 1+
3 12
As the number of compounding periods becomes very large, we approach a discount factor that
depends on a continuously compounding interest rate. The formula for this discount factor includes an
exponential term:

1 1 −rT
lim = =e .

[ ]
N rT
N →∞ rT e
1+
N
Under continuous compounding the discount rate would be
−0.03 × 0.25
e =0.99253 .
The chart below shows how the discount factor gets closer to e−rT as the number of compounding
periods N increase.

9
 Comparison of Discount Factors: Discrete vs. Continuous

Both discount factors assume a 5% annual return (r =0.05) over a horizon of two years (T =1). The formula for the
1

[ ]
N
discrete discount factor is rT and the continuous discounting factor is e−rT .
1+
N
0.9526
0.9524
0.9522
Discount Factor

0.9520
0.9518
0.9516 Discrete
compounding Continuous
0.9514 compounding
0.9512
0.9510
0 10 20 30 40 50 60 70 80 90 100
Number of Compounding Periods N

Using continuous compounding, the present value of the forward price associated with a horizon of T
years must be

PV ( F )=e−rT F=S+ M ,
and the forward price will be

F=e rT ( S+ M ) . (1)

The interest rate used here is the risk-free rate, since there is practically no uncertainty regarding cash
outflows associated with borrowing money to hold an asset.

Money-market convention: the discount factor for a horizon of d days and an (annualized) interest rate l
is

1
d
1+ l×
360

10
The interest rate l is usually based on Libor, the London Interbank Offered Rate, a benchmark lending
rate between banks. For example, a 3-month USD Libor rate of 0.251% per year would have a discount
factor of

1
90
1+ 0.00251×
360
The present value of the forward price must therefore be

F S+ M
PV ( F )= =
d d
1+l× 1+l ×
360 360
and the forward price will be

F=( S+ M ) 1+ l×
[ d
360 ]
. (2)

(c) The table below summarizes this information

Interest rate Discount

approach factor PV ( F) F
1 −rT
Continuous =e
compounding e rT
e−rT (S+ M ) e rT ( S + M )

1
S+ M
Money-Market
convention
1+l×
d
360 1+ l×
d
360
[ 1+ l ×
d
360 ]
(S+ M )

3. Convenience Yields

(a) Sometimes there might be a benefit from holding onto the asset, especially when it is in short supply.
You would have the option of consuming the asset out of storage (oil is an important example of such an
asset). The “convenience yield” measures the value of this option.

(b) We modify the forward pricing formula by changing the discount factor to

e−(r −c)T .
This indicates that the cost of borrowing at the risk-free rate is offset by getting the convenience yield
from holding the asset. Consequently, the price of the forward is

F=e (r −c )T (S+ M ).

11
This price is less than a price without a convenience yield, and represents a lower bound for price. We
don’t in general have a good way of measuring the convenience yield, but, by no-arbitrage, the asset
price would have to be between the lower and upper bounds:

e (r −c ) T ( S+ M ) ≤ F ≤ erT ( S+ M ) .

4. Forward Price Behaviour

(a) We normally observe forward prices to be (1) greater than spot prices, (2) higher for longer
maturities. This is a direct consequence of the formula when r > c.

Occasionally, we observe the opposites. of (1) and (2). One way this is possible is when c >r . There are
special names given for the normal and abnormal scenarios, summarized in the table below.

Forward prices > spot prices Contango

Forward prices < spot prices Bakwardation
Longer maturity prices > shorter maturities Normal market
Longer maturity prices < shorter maturities Inverted
market

The charts below show how a higher convenience yield can cause backwardation and an inverted
market situation.
Contango vs. Backwardation Normal vs. Inverted

Spot price S0 =100, holding costs M =0 , r =0.05, Spot price S0 =100, holding costs M =0 , r =0.05,
T =0.05 . Forward price calculated using T =0.05 . Forward price calculated using
( r −c ) T ( r −c ) T
F=e ( S+ M ) F=e ( S+ M ).

102.0 103.0
Normal market
101.5 102.0
Forward (c = 0.02 < r)
101.0 Contango
Forward price

100.5 101.0
Price

100.0 100.0 Inverted

99.5 Spot (c = 0.08 > r)
99.0
99.0 Backward-
98.5 ation 98.0
98.0 97.0
0.02 0.04 0.06 0.08 0.2 0.4 0.6 0.8
Convenience yield Maturity

12
Crude oil futures may exhibit contango and inverted markets because of high convenience yields.
Holding onto crude oil may be especially beneficial during times of tight supply or market uncertainty.

5. Implied Repo Rate

(a) We have assumed that the interest rate connecting forward prices and spot prices is the risk-free
rate based on no-arbitrage considerations. If we simply observed forward prices and spot prices, we
could compute the interest rate implied by the forward price formula and check whether it matches the
risk-free rate. This implied interest rate is called the “implied repo rate”, and it represents the rate at
which one can borrow or invest using spot and futures/forward markets. Assume that holding costs and
convenience yield are zero for simplicity and rearrange the forward price formula to solve for r :
rT
F=e S
⇒ ln F=ln erT S=ln erT + ln S=rT + ln S
⇒rT =ln F−ln S
1
⇒ r = (ln F−ln S )
T
To include convenience yield or holding cost, we can simply replace r with r −c and S with S+ M . This
would give us

1
r −c= ( ln F−ln ( S + M ) )
T
1
⇒ r =c + ( ln F−ln ( S+ M ) ) .
T (3)

(b) This interest rate is called a “repo” rate because it is realized through sale and “repossession” of the
asset: (1) Sell the asset at spot and obtain a cash inflow of S (this is equivalent to borrowing S) and
enter into a futures contract to buy the asset at the futures price of F=e rT S after time T has elapsed;
(2) at maturity of the contract pay the futures price F (this is equivalent to making a payment of S plus
interest) and cover the short asset position. The trader has effectively borrowed at the rate r by selling
and buying back or “repossessing” the asset.

(c) The repo rate should theoretically be equal to the risk-free rate, but in practice, it may not. This could
give rise to arbitrage opportunity. How? Let the risk-free rate of borrowing be denoted r b and the risk-
free rate of lending r l . Let’s say the repo rate r was less than the risk-free rate you could lend at:

r <r l .

If this were true you could “borrow” at the repo rate by selling the asset (“borrowing amount S”) and
investing the proceeds at the risk-free rate, and also taking a long position in the forward contract. The

13
amount invested would grow to e r T S . You would then “pay off the loan” by buying the asset at the
l

forward price of F=e rT S (“repayment of borrowed S with interest”), and make a risk-free profit of

e r T S−F=er T S−e rT S=( e r T −e rT ) S> 0.

l l l

This situation could not persist under no-arbitrage. Similarly let’s say the repo rate r was greater than
the rate you could borrow at:

r >r b .

If this were true you could “lend” at the repo rate by borrowing at the risk-free rate r b and buying the
asset (“lending the amount S”), and taking a short position in the forward contract. You would then
“receive payment on the loan” by receiving the forward price of F=e rT S (“receiving payment of S with
interest”), and the amount you owed would be e r T S, giving you a risk-free profit of
b

F−e r T S=e rT S−e r T S=( erT −e r T ) S> 0.

b b b

(d) We have provided examples of how to “borrow” and “lend” at the implied repo rate. We don’t
actually borrow and lend, but the way assets and forwards are bought and sold creates cash flows that
resemble borrowing and lending activities, namely, receiving a cash inflow/outflow of S
(“borrowing/lending S”) and paying/receiving a cash outflow of e rT S (“making/receiving payment on
loan of S with interest”). This is an example of “synthetic” borrowing/lending, that is, creating cash flows
that resemble the cash flows associated with borrowing.

7. Unbiased Expectations Hypothesis (Optional)

(a) It is a commonly held belief that forward prices predict future spot prices. This is called the “unbiased
expectations hypothesis”, that is, forward prices are an unbiased prediction of future spot prices. Can
this be correct? If true, we are saying that

F=e rT S=E ( ST )

E (ST ) rT
⇒ =e
S
E ( ST ) rT
⇒ ln =ln e
S
⇒ ln
⏟ E ( S T ) −ln S = ⏟
rT
expression for expected stock risk free return
return with ¿horizon T
continuouscompounding

Asset returns are not risk-free, so the above expression must be false. If we approximate asset returns
using the CAPM, we would replace the term rT with r f + βE ( r M −r f ):

14
 ln E ( ST )−ln S=[r f + βE ( r M −r f ) ]T
⏟
risk premium

⇒ E ( ST )=e[
rf + βE (r M−r f )] T
S>e r T S=F .
f

This expression is greater than the forward price formula. Hence, the forward price is not a predictor of
future spot prices; in fact, it would underpredict the expected spot price because asset returns have a
positive risk premium (expected return above the risk-free rate). Only if asset returns had a risk
premium equal to zero would forward prices predict future spot prices.

8. Forward Value

(a) The forward price is the delivery price that would make the forward contract have zero value at
inception to both parties, the buyer and the seller of the asset.

(b) The spot price of the underlying asset may change between the inception of the contract and the
maturity date. A change in the spot price results a change in the forward price of new contracts with the
same maturity date. However, this would not change the delivery price of the original contract.

(c) As a result of the difference in the delivery price of the contract and the prevailing forward price, the
original contract may have gained/lost value relative to a new forward contract with the same maturity
date. The chart below illustrates these concepts:

15
 Forward Price vs. Spot Price vs. Delivery Price

The spot price on day 0 is S0 =100.The risk-free rate 0.04; the original contract’s horizon is 100days; the delivery price of
the original contract is computed as
( 0.04 ) ( 100
356 ) ; the forward price of a new contract on day
F 0=e × 100=101.13=K
n that matures on day 100 is computed as F =e (0.04 )( )
100−n

× S n.
356
n

115
F50 = 109.15 Payoff at
S50 = 108.54 maturity is
110 Forward positive
price Fn
Contract has
105 positive value
Spot
Price

price Sn

100 Delivery price K = 101.13

Contract has
95 negative value

90
0 10 20 30 40 50 60 70 80 90 100
Day n

(d) The value of the original forward contract at some intermediate point between inception n=0 and
maturity n=10 0 will depend on the difference between the delivery price K and the forward price of a
new contract F n. A positive difference would mean that the original contract had more favourable
pricing terms. The present value of this difference would give us the value of the contract:
100−n
−r ( )
356
value of long position on day n=PV ( F n−K )=e ( Fn −K)

−r ( 100−n
356 )
⇒ value of short position=PV ( K−F n )=e (K−F n ).

The long position has positive value whenever the forward price F n is greater than the delivery price,
otherwise it has a negative value. Even though at inception the forward price was fair to both sides, at
maturity the payoff to the long side is positive because the final spot price is above the delivery price
(106 vs 100). We also see that as we get closer to maturity, forward prices converge to spot prices. The
chart below shows how the value of the long position changes over time.

16
 Changing Value of Forward Contract

The spot price on day 0 is S0 =100.The risk-free rate 0.04; the original contract’s horizon is 100days; the delivery price of
the original contract is computed as
( 0.04 ) ( 100
356 ) ; the forward price of a new contract on day
F 0=e × 100=101.13=K
n that matures on day 100 is computed as
( 0.04 ) ( 100−n
356 )
F n=e × S n. The original contract’s value is calculated as

− ( 0.04) (100−n
356 )
PV ( F n−K )=e ( F n−K ) .

15

10
Contract value: PV(F-K)

Fn K

PV(Fn K)
0
0 10 20 30 40 50 60 70 80 90 100

-5

-10
Day n

9. Currency Forwards (Optional)

Forward on currencies need a slightly modified argument. Suppose you want to buy £1 in T years and
wish to lock-in a price today. You can achieve this by one of two strategies:

 Forward contract: Pay forward price $ F in at time T to buy £1. The cost of this strategy in USD
is given by
PV ( $ F )=PV ( F × $ 1 )=F × PV ( $ 1 ) .

 Spot or replicating strategy: Buy £ x today and invest it at the risk-free rate associated with £, so
that £ x grows to £1 in T years. This means that PV ( £ 1 ) =x. The £ in the expression is to
emphasize that the present value is taken with respect to the £ interest rate. The cost of this
strategy in USD will depend on the spot price of GBP in USD, that is, the spot exchange rate:

PV ( £ 1 ) × spot USD per GBP exchange rate=PV ( £ 1 ) × S .

By no-arbitrage, most strategies must have the same present value:

F × PV ( $ 1 )=PV ( £ 1 ) × S

17
 PV ( £ 1 )
⇒ F= ×S.
PV ( $ 1 )

18
The ratio of present values depends on whether we use continuous compounding or money-market
conventions.

Interest rate
approach PV ( $ 1 ) PV ( £ 1 ) F
Continuous −rT −dT (r −d)T
compounding e e e S

[ ]
d
1 1 1+l( $) ×
Money-Market 360
d d ×S
convention 1+ l($)× 1+ l( £ )× d
360 365 1+l( £) ×
365

3. Futures Markets
1. Introduction
(a) The key features distinguishing futures contracts from forward contracts are:

 standardization
 ability to unilaterally reverse positions
 use of margin accounts to control default by investors
(b) The Chicago Board of Trade (CBoT) was established in 1848 and mostly traded in grain futures. Up
till the 1970s futures markets have been commodity based (e.g., wheat, gold, oil). They were
subsequently expanded to incorporate currency futures (1972), followed by interest-rate futures (mid
1970s) and stock-index futures (early 1980s).
(b) Futures exchanges exist in different parts of the world. Some of the biggest names in the US are
Chicago Mercantile Exchange (CME), Chicago Board of Trade (CBOT), and New York Mercantile
Exchange (NYMEX).

2. Standardization
(a) Since buyers and sellers do not interact directly, contract terms must be standardized by the exchange.
This is essential for a couple reasons.

 Promoting liquidity: A large number of buyers and sellers are interested in the same contract
terms.
 Improving the quality of the hedge: Buyers and sellers know what to expect and are able to
manage risk more accurately.
(b) Standardization involves four components:

 contract maturity date

 quantity: size of the contract
 quality: specifying the standard deliverable grade

19
  delivery options: alternative deliverable grades and price adjustments associated with them

Commodity Futures Financial Futures

Size of Approx USD Size of Approx USD
Underlying Exchange contract market value Underlying Exchange contract market value
LME US Treasury CBoT
Copper London Metal 25 tonnes 175,000 Chicago Board of USD 100,000 100,000
Exchange Bonds Trade
CME Bund Eurex
Live Cattle Chicago Mercantile 40,000 lbs 34,000 German
Eurex Exchange
EUR 100,000 145,000
Exchange government debt
CBoT 5,000 bushels S&P 500 CME
Wheat 25,000 250 x Index 275,000
Chicago Board of
Trade
1 bushel ≈ 35 liters Index Chicago Mercantile
Exchange
COMEX Dow Jones
Gold 100 ounces 109,000 CBoT
The Commodity
Exchange Inc.
Index Chicago Board of 10 x Index 105,000
Trade
NYMEX 1,000 barrels
Light sweet CME
crude oil
New York Mercantile
Exchange
1 barrel ≈ 159 75,000 British
Chicago Mercantile GBP 62,500 100,000
liters pound Exchange

(c) Size of contract: The table below provides some examples of commodity futures and financial futures
(based on information circa. 2011).

20
(b) Example of price adjustment: Consider the following contract

Exchange CBoT
Maturity date December
Size of contract $100,000 in face value of US Treasury Bonds
Quality Coupon of 6% with at least 15 years to maturity
Delivery options Any other coupon can be delivered; cash flows from the delivered bond will be
discounted at 6% rate to obtain “conversion factor” for price adjustment

Suppose that the seller wants to deliver 20-year bonds with an 8% coupon, paid semi-annually. This
requires a price adjustment based on the present value of the cash flows based on a 6% discount:

PV ( 20 year 8 % bond )
conversion factor=
Face Value of Bond

¿
[ 4
+
4
1.03 1.03 2
+…+
4+100
1.03 40 ]
100
=1.2311 .
N = 40 (20 * 2)
I/Y = 6 (discount)
PMT = 4
FV = 100
P/Y= 2
C/Y = 2
CPT PV = 1.2311

The long position will pay the short position 1.2311 times the agreed-upon delivery price.
(c) Delivery options are provided to (1) enhance market liquidity (makes it easier for the seller to fulfil
the contract terms), and (2) make inappropriate exercise of market power (“corners and squeezes”) more
difficult.
The seller will choose an alternative delivery grade when it is cheaper to do so, which means it is the
“cheapest-to-deliver” grade. This change reduces the effectiveness of the price protection or hedge the
buyer hoped to achieve, because the hedge was calculated based on the standard delivery grade. The
quality of the buyer’s hedge is degraded.

21
3. Unilateral Reversal of Positions
(a) Holders of futures contracts can unilaterally reverse or “close out” their position by taking the
opposite position to the original. For example, they can reverse a long position in 10 COMEX gold
contracts for delivery in April by taking an offsetting short position in the same contracts. This could cost
lead to a loss, as explained in the example below.
Contract reversal example: Suppose a trader has a long position in 10 COMEX gold contracts for
delivery in April at the futures price of $1045 per oz. Each contract is for 10 oz. of gold. However, he
needs the gold in March, one month prior to the delivery date, so he decides to close out (or reverse, or
settle) the futures contract in March. The futures price at the time of closeout is now $1,037 per oz. This
means that the trader will see a net loss of $8 per oz, or $800 overall:
cash flow ¿ futures contract reversal
¿ short contract value−long contract value
¿ ( short price per oz .−long price per oz . ) × total oz .
¿ ( 1037−1045 ) × ( 10× 10 )
¿−8 ×100
¿−800.
This loss is settled through the margin account, which we discuss in the next section. The trader must also
buy the spot asset on the market commitment date. The reversal gain compensates for the change in spot
price, and the overall effect is to mitigate the risk associated with spot price changes. If reversals were not
possible, the trader would be forced to buy the gold at the spot price on the market commitment date, and
to buy it again on the delivery date at the original futures price, and sell it off on the same day. Risk from
changes in the spot price would also be mitigated in this scenario, but not to the same extent (see below).
(OPTIONAL) Let us consider three scenarios and compare how risky the cash flows of each are. Assume
the market commitment date is in 3 months, but the futures delivery date is in 4 months. The following
notation and assumptions are used:

 Sn is the spot price in month n ; e.g., S3 is the spot price in month 3 (market commitment date)
 F m ,n is the futures price in month m for delivery in month n ; e.g., F 3,4 is the futures price in
month 3 for delivery in month 4.
 The 3-month risk-free rate (not annualized) is equal to r . Assuming no holding costs, this means
that F 3,4=( 1+r ) S3 .
 Δ S n is the change in spot prices between two consecutive months; e.g., Δ S 4=S 4−S3
 Changes in spot prices are i.i.d. (independent and identically distributed), which means that
var ( Δ S n )=var ( Δ S m ), and cov ( Δ S n , Δ S m )=0 . A consequence of this is that cov ( Δ S n , Sm ) =0.

22
 Cash outflows Variance of cash outflows

No futures contract: (i) Purchase S3

3 var ( Δ S 3 )
spot in month 3 ¿ S0 + Δ S 1 + Δ S 2 + Δ S3
Futures contract without reversal:
(i) Enter futures contract at price for
delivery in month 4 ; (ii) purchase S3 + F 0,4−S 4 var ( Δ S 4 )
spot asset in month 3; (iii) purchase ¿ F 0,4− Δ S 4
asset for price F 0,4 in month 4; (iv)
sell the asset for S4 in month 4

Futures contract with reversal: (i) S3 + F 0,4−F 3,4

Enter futures contract at price for ¿ F 0,4−( F 3,4 −S 3 ) 2
delivery in month 4; (ii) reverse the r var ( S3 )
contract in month 3; (iii) purchase ¿ F 0,4−( ( 1+ r ) S3 −S 3 )
spot asset for S3 in month 3 ¿ F 0,4−r S3

The “no futures contract” scenario has 3 times the variance of the “futures contract without reversal”
scenario since var ( Δ S 3 )=var ( Δ S 4 ) (by i.i.d. assumption), so a futures contract reduces risk exposure
significantly, even if it cannot be reversed. Intuitively, futures with reversal should be less risky than
futures without reversal because the variance without reversal depends on two random variables, S4 and
S3 (since Δ S 4=S 4−S3 ), whereas the variance with reversal depends on only one random variable, S3.
Mathematically, we have to show that:

r 2 var ( S3 ) < var ( Δ S 4 ) .

2
var ( S 4−S3 )
⇒r < (1)
var ( S3 )
We now express the numerator of right-hand side in terms of correlation.

var ( S 4−S 3 )=var ( S 4 ) + var ( S 3 )−2 cov ( S4 , S3 )

¿ var ( S 3+ Δ S 4 ) + var ( S3 ) −2 cov ( S3 + ΔS 4 , S 3 )

¿ var ( S 3+ ΔS 4 ) + var ( S3 ) −2 ρ √ var ( S3 + Δ S 4 ) var ( S 3 )

This expression has a higher value than an expression not including Δ S 4, which means that

var ( S 4−S 3 )> var ( S3 ) + var ( S 3 )−2 ρ √ var ( S3 ) var ( S 3)

¿ 2 var ( S3 ) −2 var ( S 3 ) ρ

¿ 2 var ( S3 ) ( 1−ρ )

Condition (1) will be satisfied if

23
 2
2 var ( S 3 ) ( 1−ρ )
r <
var ( S3 )

⇒ r 2 <2 ( 1−ρ ) .
For ρ=0.99, which is conservatively large, this would require the risk-free rate to be less than:

√ 2 (1−0.99 )=0.14
This is will typically be true. Hence, we have shown that scenario 3, or futures with reversal, has the
lowest risk compared to all other scenarios.

4. Margin Accounts
(a) Since buyers and sellers interact with a third party, the futures exchange, the party losing value on the
contract due to changes in the spot price has an incentive to default.
(b) Futures exchanges use margin accounts to prevent default. “Margin” can be thought of as the
collateral that is put up in case of default.
(c) Margin levels that are too high discourage trading and reduce liquidity. Margin levels that are too low
increase the risk of default. Margin levels are not too high in practice, approximately between 2% and
10% of contract value.
(d) The margining procedure includes

 the initial margin: amount initially deposited by investor into the margin account
 marking-to-market: daily adjustments to reflect gains/losses from futures price movements over
the day
 the maintenance margin: floor level of margin account; if balance falls below this, customer
receives a “margin call” to reset the margin to the initial balance; if the margin call is not met,
account is closed out immediately
(e) Margin account example:
Contract details Margin requirements
10 futures contracts Initial margin: $878/contract
Each futures contract is for 5000 bushels of wheat ⇒ Total initial margin: 878 x 10 = $8780
Price of wheat $3.60 per bushel Maintenance margin: $650 per contract
Contract value = 5000 x 3.60 = $18,000 ⇒ Total maintenance margin: 650 x 10 = $6500
⇒Total value = $18,000 x 10 = $180,000

Margin Account Behaviour

The “settlement price” refers to the price that is used to determine the value of the position each day. It
is equal to the price of a futures contract with the same quantity, quality, delivery options, and maturity
date. The daily settlement enables the original futures contract to be “marked-to-market” each day.

Settlement price Gain (loss) settled Account Margin

Day per bushel per contract per contract total balance call

24
 1 3.58 17,900 (100) (1000) 7,780 NO
2 3.57 17,850 (50) (500) 7,280 NO
3 3.56 17,800 (50) (500) 6,780 NO
4 3.55 17,750 (50) (500) 6,280 YES
5 3.54 17,700 (50) (500) 8,780 NO
6 3.53 17,650 (50) (500) 8,280 NO
7 3.52 17,600 (50) (500) 7,780 NO

With the margin account the total gain is -4000 (sum of the fourth column in the table). If there were no
margin account, and the contract were closed out on day 7, the gain would be equal to:

gain=short contract value ( day 7 ) −longcontract value(day 0)

¿ ( short price per bushel−long price per bushel ) ×total bushels
¿ ( 3.60−3.52 ) × ( 5000 ×10 )
¿−4000

This gain would be a realized entirely on day 7. With the margin account, settlements occur on a daily
basis, resulting in a series of smaller losses from daily price movements. The investor cannot let losses
pile up and walk away. He must settle his losses daily, which reduces the size of the default risk and,
consequently, the desire to default.

(f) A mathematical formula for futures is difficult to derive because of the additional uncertainty
associated with (1) delivery options, and (2) the interim cash flows of margin accounts. Practically
speaking, the impact of this uncertainty is not very large, and forward prices are a good approximation of
what futures prices, especially short-dated ones, should be.

4. Hedging with Forwards and Futures

1. Introduction

(a) Hedging is perhaps the most important function served by futures and forwards contracts, because it
reduces the uncertainty in cash flows associated with market commitments, for example, cash flow
commitments associated with buying commodities, currencies, or making interest payments.

(b) Standardization creates some problems in using futures contracts as a result of two things.
 Commodity mismatch: The grade of the commodity underlying the futures contract may not be
the one being hedged. This is known as “commodity basis risk”.
 Delivery date mismatch: The standardized maturity dates of the futures contract may not match
the desired hedging dates. This is known as “delivery basis risk”.

2. Basis risk overview

(a) We used the term “basis risk” above. This has a specific meaning associated with the cash flow
uncertainty of a hedge. The two examples below explain how this uncertainty may arise.

Futures contract without basis risk: It is currently January and a gold-wire manufacturer anticipates a
need for 10,000 oz. of gold in April. The manufacturer has two hedging options:

25
 Option 1 Option 2
(1) Go long 10,000 oz. of April gold futures for (1) Go long 10,000 oz. of April gold futures for
delivery price F 0 per oz. delivery price F 0 per oz.
(2) Make the payment when the contract (2) Close out the futures position in April at the
matures and collect the physical gold. closeout price F T .
⇒Net cash flow = −10,000 F0 ⇒The futures settlement will be equal to 10,000
(F T −F 0 ).
(3) Buy 10,000 oz. of gold in April at the spot
price ST per oz.
⇒Net cash flow = 10,000 ( FT −F 0 )−10,000 ST

The manufacturer might opt for option 2 because it wants to avoid of collecting the physical gold from
the seller. It would rather buy it from its local supplier. The net cash flows from the second option can
be re-written as follows.

Net cash flow=10,000 ( F T −F 0 ) −10,000 ST

¿ 10,000 ( FT −ST )−10,000 F 0 .

There is no commodity mismatch since the underlying asset in futures contract is the same as spot asset
(both are gold). Also, there is no delivery date mismatch since the futures price F T and the market
commitment were both based on April. Thus
r0
F T =e ST =ST .

In the presence of delivery date mismatch, say the spot market commitment was for April but futures
contracts were available for May, then the delivery date of the futures contract would not be aligned
with the market commitment date. At closeout, the April futures price would have one month remaining
till contract maturity, and this would make its price different from the spot asset:

r ( 121 )
F T =e ST ≠ ST .

Returning to the scenario when there’s no mismatch, we have

r×0
F T =e S T =S T

⇒ Net cash flow=10,000 ( F T −S T ) −10,000 F 0

26
 ¿−10,000 F 0 .

The hedge is “perfect”: It is the same as going long a futures contract for the required underlying and
delivery date, as in option 1. The above scenario has no basis risk, that is, a mismatch between
commodities or delivery dates.

(b) Forward contract with basis risk: A US exporter might wish to lock-in the price of the Czech koruna,
but no active forward contract exists for USD/CZK. The company can use a related currency, the Euro, to
hedge against exposure to the Czech koruna: It would enter a forward contract for USD/EUR, close out
the contract on the delivery date, and buy the Czech koruna at the spot price. Commodity risk is
obviously present here.
(c) To define basis risk more precisely, we make use of the following notation

 S0 : spot price of asset being hedged

 F 0: current futures price
 T : date of market commitment
 Q : size of market commitment
 H : size of futures position
 ST : time-T price of asset being hedged
 F T : time-T futures price

Like in the example in (a), to buy at time T the strategy is:

(1) Go long H futures at price F 0 for delivery date that may or may not be the same as
the market commitment date T
(2) Close out futures at date T at price F T ⇒ Cash flow H ( FT −F 0 )
(3) Buy Q units of spot at date T at price −Q S T
ST ⇒ Cash flow
⇒ n et cash inflow=H ( FT −F 0 )−Q ST
⇒ net cash outflow=Q ST −H ( F T −F 0 )

The risk that we wish to minimize is the variance of the net cash outflow

basis risk=var [ Q S T −H ( F T −F 0 ) ] .

To minimize this risk we have to choose the combinations of Q and H . We will see how to do this in the
next section.

3. Minimizing Basis Risk via Minimum Variance Hedge

(a) In order to come up with a mathematical expression for minimizing basis risk, we will first re-write it
in terms of changes to the spot and futures prices. We do so by adding and subtracting Q S 0 to our
expression for the net cash outflow above.

net cash outflow=Q S T −H ( F T −F0 )

27
 ¿ Q S T +Q S 0−Q S0−H ( F T −F 0 )

¿ Q S 0 +Q ( S T −S 0 )−H ( F T −F 0 )
⏟ ⏟
ΔS ΔF

¿ Q S 0 +Q Δ S−H Δ F

⏟
¿ Q S 0 +Q(Δ S−H /Q Δ F)
h

¿ Q S 0 +Q ( Δ S−h Δ F ) (1)

The terms Δ S and Δ F represent the change in spot prices and futures prices over the hedging horizon,
and h=H /Q represents the “hedge ratio”, or the quantity of futures contract used in the hedge relative
to the quantity of spot. The variance of the net cash outflow is the definition of “basis risk”:

basis risk=var ( net cash outlfow )

¿ var [ Q S0 +Q ( Δ S−h Δ F ) ] .

We wish to minimized basis risk by choosing the appropriate value for h (Q and S0 are given, Δ S and
Δ F are random variables, and h is what we can choose). The variance is calculated as follows:
var [ Q S0 +Q ( Δ S−h Δ F ) ]=var [ Q ( Δ S−h Δ F ) ]

¿ Q var [ ΔS−h Δ F ]
2

basis risk=Q [ σ Δ S + h σ Δ F −2hcov ( Δ S , Δ F ) ]

2 2 2 2
(2)

To find the minimum value of this function as we vary h , we take its derivative with respect to h and set
it equal to zero.

d
dh
( Q2 [ σ 2Δ S +h2 σ 2Δ F −2 hcov ( ΔS , Δ F ) ])

¿ Q 2 [ 0+2 h σ 2Δ F −2 cov ( Δ S , ΔF ) ] =0
2
⇒ 2 h σ Δ F −2 cov ( Δ S , Δ F )=0

h σ 2Δ F −cov ( . , . )=0

28
 cov ( Δ S , Δ F )
⇒ h⏟¿ = 2
minimum
σ ΔF
variance
hedge ratio

¿
We want to write the minimum variance hedge ratio h in terms of correlation rather than covariance.
Correlation is an adjusted form of covariance:

cov ( Δ S , Δ F )
corr ( Δ S , Δ F )= ρΔ S Δ F =
σ ΔS σ ΔF

⇒ cov ( Δ S , Δ F )=ρ Δ S Δ F σ Δ S σ Δ F (3)

We can therefore write the minimum variance hedge ratio in terms of the correlation:

cov ( Δ S , Δ F ) ρ Δ S Δ F σ Δ S σ Δ F
⇒ h ¿= 2
= 2
σ ΔF σ ΔF

σ ΔS H¿
⇒ h ¿= ρΔ S Δ F = . (3)
σ ΔF Q

¿
(b) If we substitute h into the expression for the variance of the net cash outflow and simplify, we
obtain an expression for the minimum variance. To reduce notational clutter, ρ is used instead of
ρ ( ΔS , ΔF ).

variance of net cash outflow with hedge ratio h¿

¿ Q [ σ Δ S + ( h ) σ Δ F −2h cov ( ΔS , Δ F ) ]
2 2 ¿2 2 ¿

[ ( ) ]
2
2 2 σ ΔS 2 σ
¿ Q σ ΔS+ ρ σ ( Δ F ) −2 ρ Δ S ρ σ Δ S σ Δ F
σΔF σ ΔF

¿ Q [ σ Δ S + ρ σ Δ S−2 ρ σ Δ S ]
2 2 2 2 2 2

¿ Q [ σ Δ S− ρ σ Δ S ]
2 2 2 2

⇒ minimum variance of net cash outflow=Q σ Δ S ( 1−ρ ) .

2 2 2

(c) The minimum variance that is achievable depends on the correlation ρ between the changes in spot
price and changes in futures price. When spot and futures are perfectly correlated, that is ρ=1 or
¿
ρ=−1, the minimum variance will be equal to zero as the minimum variance hedge h =ρ σ Δ S / σ Δ F
results in the spot price changes Δ S being exactly offset by futures price changes Δ F . When correlation
is equal to zero, there is no benefit in entering into a futures contract, as it will only increase variance of
¿ 2 2
the net cash outflow, so h =0 , and variance is equal to the variance of the spot alone Q σ Δ S . These
results are summarized in the chart below.

29
 Correlation and Minimum Variance of Cash Outflow
The chart uses equation (2) to compute variance and equation (3) to compute the minimum variance hedge h¿ , rewriting
the covariance terms in terms of correlation, that is, cov ( Δ S , Δ F )=ρ Δ S Δ F σ Δ S σ Δ F. σ Δ S =1; σ Δ F =0.8 ; Q=1.

4
Cash outflow variance

ρ = -0.4
3
h* for ρ=0
different
2 values of ρ

ρ = 0.6
1

ρ=1
0
-1 -0.5 0 0.5 1 1.5 2
hedge ratio h
(d) We can see from the chart that the minimum variance is achieved at the minimum variance hedge
¿
ratio h . It is interesting to note that if we hedge one-for-one ( h=1), we may in fact end up with a
higher variance as a result of the hedge: the cash flow variance associated with ρ=0 and ρ=0.4 is
greater when h=1 versus h=0. Hedging one-for-one, though intuitive, is not the right approach.
¿
The extent to which h removes variance can be quantified by examining the percentage change as a
result of using the minimum variance hedge. This is given by:

min cash flow variance with hedge−cash flow variance without hedge
.
cash flow variance without hedge

Q σ Δ S ( 1− ρ )−Q σ Δ S
2 2 2 2 2
¿ 2 2
Q σ ΔS

¿−ρ2 .
Hence, the minimum variance hedge reduces the variance by ρ2. The table and chart below show how
much variance is removed based on the correlation ρ .

Correlation Var removed Variance removed vs. Correlation

2
ρ ρ
0.90 81%
0.80 64%
0.50 25%
0.20 4%

30
 1.0

Variance rmoved ρ2
0.8

0.6

0.4

0.2

0.0
0 0.2 0.4 0.6 0.8 1
Correlation ρ

4. Examples

(a) Cross hedging with currencies: Cross hedging is hedging risk using two distinct assets with positively
correlated price moments, as in the following example: A US exporter will receive NOK 25 million in
three months and would like to hedge this with a USD/EUR forward contract. The spot asset is NOK, but
the underlying asset in the forward contract is EUR. Hence, there is basis risk due to a mismatch in the
underlying asset.

The minimum-variance hedging ratio is given by

σ ΔS
h¿ =ρ Δ S Δ F .
σ ΔF

Suppose we are given

σ Δ S =0.005 , σ Δ F =0.025 , ρ=0.85 .

The minimum-variance hedging ratio is equal to

¿ 0.005
h =0.85 × =0.17 .
0.025
We can use this to determine the minimum-variance quantity of the underlying in the futures contract
H ¿:
¿ futures quantity H¿
h= =
spot asset quantity Q
¿
The quantity of spot asset is NOK 25,000,000. Which means that H is given by

H ¿ =h¿ Q
¿ 0.17 ×25,000,000=EUR 4,250,000.
This will be a long position since the investor wishes to buy the spot asset at the market commitment
date. The strategy will be to enter a long forward contract of EUR 4,250,000 with maturity date three
months, closeout the contract on the market commitment date, and buy NOK 25,00,000 spot.

(b) Cross hedging with equities: An investor manages a portfolio of S&P 100 index holdings worth
$80,000,000, with current index level 800. In order to hedge exposure to risk, the investor uses an S&P

31
500 index futures contract, with current index level 960, and one futures contract for 250 times the
index. There is basis risk because the underlying asset mismatch.

Suppose we are given

σ Δ S =60 , σ Δ F =75 , ρ=0.90 .

Then the minimum-variance hedge ratio is equal to

60
h¿ =0.90 × =0.72 .
75
¿
To compute the minimum-variance quantity of the futures contract H we need to determine the
quantity of spot Q . The investor’s spot asset holdings are worth $800,000,000 and the index level is 800,
which means that the number of units of the spot asset must be 100,000 (= 80,000,000 / 800).

⇒ H ¿ =h¿ Q=0.72 ×100,000=72,000 S∧P 500 index units .

Since one futures contract is for 250 units of the index,

72,000
Number of futures contracts= =288.
250
This will be a short position since the investor wishes to maintain the value of the underlying spot asset,
which can be thought of maintaining the selling price of the asset.

5. Implementation Considerations

(a) We don’t directly observe variance but have to estimate it from historical changes to price. If we
2
observe daily price changes, we can estimate daily variance σ δS. How do we convert the daily variance
estimate to a K day horizon variance estimate? If we assume price changes to behave like random
variables that are i.i.d. (independent and identically distributed), we can do this simply by multiplying
the daily variance by K :

K horizon variance=K × daily variance

2 2
σ Δ S =K σ δS .

⇒ σ Δ S=√ K σ δS .

For example, suppose we have estimate daily standard deviation of price changes to be σ δS =60, and we
to compute the monthly standard deviation σ ( ΔS ). Given 22 trading days per month, the monthly
standard deviation would be σ Δ S =√ K σ δS =√ 22× 60=281.42 .

Under the i.i.d. assumption, we can also show that K day horizon correlations are equal to daily
correlations:

ρ Δ S Δ F =ρδSδF .
(4)

32
(b) (OPTIONAL) To show that (4) is true, we first write the K day horizon change in terms of daily
changes:

cov ( Δ S , Δ F )
ρΔS ΔF=
σ ΔS σ ΔF

cov ( Δ S , Δ F )=cov ( δ S 1+ …+δ S K −E ( δ S1 +…+ δ S K ) , δ F 1+ …+ δ F K −E ( δ F 1+ …+δ F K ) )

¿ cov ([ δ S1−E ( δ S 1 ) ] +…+ [ δ S K −E ( δ S K ) ] , [ δ F 1−E ( δ F 1 ) ] +…+ [ δ F K −E ( δ F K ) ] )

Under the independence assumption this simplifies to the sum of covariances of all contemporaneous
pairs:

cov ( [ δ S1−E ( δ S 1 ) ] , [ δ F 1−E ( δ F 1 ) ] )+ …+cov ([ δ S K −E ( δ S K ) ] , [ δ F K −E ( δ F K ) ] )

Under the identically distributed assumption the covariance of all contemporaneous pairs is the same,
which means that the sum of K covariances is equal to

Kcov ( [ δS−E ( δS ) ] , [ δF−E ( δF ) ] )

¿ Kcov ( δS , δF )
cov ( Δ S , Δ F )
⇒ ρΔ S ΔF=
σ ΔS σ ΔF

Kcov ( δS , δF ) cov ( δS ,δF )

¿ = = ρδSδF
√ K σ δS √ K σ δF σ δS σ δF

(b) We could also use daily price change data to compute the minimum variance hedge ratio. The
original formula which is computed in terms of K day horizon correlation and standard deviations would
now be computed in terms of daily correlation and variances:

¿ σ ΔS σ δS
h =ρ Δ S Δ F=ρδSδF
⏟σ ΔF ⏟ σ δF (5)
K day horizon daily horizon
parameters parameters

¿
(c) Regression for computing h : The minimum variance hedge ratio can also be computed by running a
regression of daily spot price changes δS on daily futures price changes δF :

δS =a+bδF + ϵ
¿
The estimate of the coefficient b would be equal to h . This is because the least squares regression
estimate of b , denoted b^ , has the same value as h we require, since
¿

33
 ^ cov ( δS , δF )
b= 2
σ δF

cov ( δS , δF ) σ δS σ δF
¿ 2
×
σ δF σ δS σ δF

σ δS ¿
¿ ρδSδF =h
σ δF

(d) Hedging one risk with many futures contracts: We could use multiple futures contracts to hedge
against changes in the prices of the spot asset. For example a portfolio of junk bonds may be better
hedged by a combination of equity futures and interest-rate futures. We would then run a multivariate
regression of the form

δS =a+b 1 δ F 1 +b2 δ F 2+ …+bn δ F n +ϵ .

The regression estimates b^ 1 , b^ 2 , … , b^ n are precisely the minimum-variance hedge ratios h1 , h2 , … ,h n .

¿ ¿ ¿

(d) Hedging multiple risks with one futures contracts: Suppose that a company wishes to hedge exposure
to both USD/NOK and USD/CZK exchange rates, but will use the same USD/EUR futures contract to
hedge each currency. How should it choose the total size of the hedge?

 Compute the minimum-variance hedge size for each risk separately.

 Add them all up.

This results in the minimum-variance hedging strategy for the portfolio of risks.

(e) Tailing factor (Optional): We have treated futures contract as if it is a forward contract, marked-to-
market only once at the end of the contract. If we mark-to-market daily, the hedge ratio is modified
slightly. The adjusted hedge ratio is
¿
¿∗¿=g ( R , K ) × h ¿
h
R=per day gross interest rate
K=number of days ∈hedging horizon

1+ R+ R 2+ …+ R K
g ( R , K )= .
1+ R 2+ R 4 +… R2 K
The function g(R , K ) is called the “tail factor”. “Tailing” the hedge is not that important in practice for
short horizons and low to moderate interest rate levels. For example, for a 5% annualized interest rate
and K=100, the tail factor would be

0.05
R=1+ =1.000137
360
2 100
( ) 1+1.000137+1.000137 + …+1.000137
g 1.000137,100 = 2 4 200
=0.993 .
1+ 1.000137 + 1.000137 +…+ 1.000137

34
The chart below shows how the tailing factor varies with the number of days in the hedging horizon.

Tailing Factors vs. Number of Days in Hedging Horizon

1.00

0.99

0.98
Tailing factor g(R,K)

0.97

0.96

0.95

0.94

0.93
0 100 200 300 400 500 600 700 800 900 1000
Number of days in hedging horizon K

5. Interest-Rate Forwards and Futures

1. Definition of a forward rate agreements (FRA)

(a) A forward rate agreement (FRA) is an agreement to exchange interest calculated at a fixed rate for
interest calculated at a floating rate for a period [T 1 , T 2] in the future. The diagram below shows the
nature of this exchange:

fixed rate
k
Long FRA Short FRA
ℓ

floating
rate

The table below defines an FRA in greater detail and mentions some important points.

An FRA is an agreement…

…to exchange interest calculated at a (the long position pays k and receives
fixed rate k for interest calculated at the l (usually Libor-based); the short
a floating rate l … position pays l and receives k )

…on a specified principal amount P … (the principal P is never exchanged,

but interest payments are based on
it)
…over a specified horizon [T 1 , T 2] in
the future… (the times are stated in months, e.g.,

35
 a 3 ×6 FRA begins in 3 months’ time
and ends in 6 months’ time)
which is settled at time T 1 in
discounted form at the floating rate l (see below for settlement payoffs)

(b) To understand the payoff that is settled in discounted form at time T 1, let’s simplify and suppose
that the payoff is settled at time T 2without discounting. In this case, the long position receives l and
pays k :
payoff ¿ long at T 2
¿ floatinginterest received −¿ interest paid
d d
¿ P ×l −P ×k
360 360
d
¿ P × ( l−k ) .
360
Now if we consider that the settlement takes place at time T 1, we simply adjust the above payoff by
discounting it back at the floating rate l :
payoff ¿ longat T 1=PV ( payoff ¿ long at T 2 )
d
P × (l−k )
360
¿
d
1+l
360
d
(l−k )
360
¿P× . (1)
d
1+l
360

The payoff to the short is simply the reverse of the payoff to the long:
payoff ¿ short at T 1
d
( k−l )
360
¿P× .
d
1+l
360
(c) FRA payoff example: Consider 4 ×7 FRA has a spot date of March 15, with principal P=$ 5 million,
the fixed rate k =0.05. This means that the interest payments will begin in 4 months (July 15) and end in
7 months (October 15): a 92-day period. The payoff for the 3-month Libor l on July 15 is shown in the
table below for two scenarios: l=5.40 % and l=4.70 %.

36
 l=5.40 % l=4.70 %
92 92
( 0.054−0.05 ) × ( 0.047−0.05 ) ×
360 360
5,000,000 × 5,000,000 ×
92 92
1+0.054 × 1+0.047 ×
360 360
¿ 5,041.43 ¿−3,787.83

⇒Long receives from short 5,041.43 ⇒Long pays short 3,787.83

The chart below shows that whenever l>k , the long position receives payment (payoff > 0).

Payoff to Long Position vs. Floating Rate

The fixed rate is equal to 5%. Positive payoffs are interpreted as the long
position receiving payment from the short position, and negative payoffs as
the long position making payments to the short position.

56,000
Discounted Payoff to Long

36,000

16,000

-4,000
0 0.025 0.05 0.075 0.1
-24,000
Floating rate = Fixed rate
-44,000

-64,000
Floating rate

2. Hedging with FRAs

(a) In order to convert floating rate liabilities into a fixed rate, companies would undertake the following
steps.

T 1 CF T 2 CF
Enter into a long T 1 ×T 2 FRA d
( l−k )
today on a principal amount of P . 360
P×
d
1+l
360
Invest proceeds from T 1 at the d d
( l−k ) P × ( l−k )
prevailing Libor rate for maturity at 360 360
T2 −P ×
d
1+l
360

37
 At T 1, borrow the required
d
amount at the prevailing Libor rate P −P−P ×l
for payment at T 2
360

Net cash flow d

P −P−P × k
360

From the table above we see that the floating-rate borrowing at T 1 is being compensated by the FRA,
and the net cash flow is as if the company borrows the principal P and pays it with interest based on the
fixed rate k .

3. Valuing and Pricing FRAs

(a) Valuing an FRA: The value of an FRA would be equal to the present value of the payoff to long at time
T 1:

value of FRA =PV [ payoff ¿ long at T 1 ]

[ ]
d
( l−k )
360
¿ PV P ×
d
1+l
360

We will find a useful expression for the present value by first re-writing the expression for the
discounted payoff:

d
( l−k )
360
P×
d
1+l
360
d d
Pl −Pk
360 360
¿
d
1+l
360
Add and subtract P in the numerator:

38
 [ P+ Pl
d
360 ][
− P+ Pk
d
360 ]
d
1+l
360
d d
1+l 1+k
360 360
¿P× −P ×
d d
1+l 1+ l
360 360
d
1+k
360
¿ P−P × . (2)
d
1+ l
360

The first term represents a certain inflow of P , but the second represents an uncertain outflow, since l is
not known until T 1. If second term is invested at Libor at time T 1, its future value at time T 2 will be the
fixed amount:

[
P × 1+ k
d
360
.
]
Let B ( T 1 ) and B ( T 2 ) represent discount factors for cash flows at times T 1 and T 2. We can write the
present value of equation (2) as

( )
d
1+k
360
PV P−P ×
d
1+ l
360

( )
d
1+k
360
PV ( P ) −PV P ×
d
1+l
360

¿ P × B ( T 1 ) −P × 1+k
[ d
360 ]
× B (T 2)

{
¿ P × B ( T 1 )− 1+ k
[ d
360 ] }
B (T 2) (3)

(b) Pricing an FRA: To price a new FRA, we would need a fair value for the fixed rate k that will be
exchanged for floating rates. The rate should be one where both the long and short positions break even
after exchanging interest payments. This occurs when the present value of the payoff is equal to zero:

39
 PV [ payoff at T 1 ] =0

{
⇒ P × B ( T 1 )− 1+ k
[ d
360 ] }
B ( T 2 ) =0

⇒ B ( T 1 ) − 1+k
[ d
360 ]
B ( T 2) =0

⇒ B ( T 1 ) = 1+k
[ d
360 ]
B ( T 2)

B (T 1) d
⇒ −1=k
B (T 2) 360

¿
B ( T 1 ) −B ( T 2 ) 360
⇒k = × (4)
B( T2) d [T ,T ]
1 2

Thus, the price only depends on the discount factors B ( T 1 ) and B ( T 2 ) .

(c) Pricing a new FRA example: Suppose we are given the following data from which we need to
determine the price of a new 3 ×6 FRA:

 The current 3-month and 6-month Libor rates are 4.00% and 4.50% respectively.
 The first 3-month period contains 92 days, the second 91 days.

To determine the price, we use the formula:

¿
B ( T 1 )−B ( T 2 ) 360
k= ×
B (T 2) d [ T ,T ]
1 2

B ( 3 )−B ( 6 ) 360
¿ × .
B (6 ) 91

We calculate B(3) and B(6) first:

1 1
B (3 )= = =0.98988
d3 92
1+ l3 1+0.04 ×
360 360
1 1
B ( 6 )= = =0.97763
d6 (91+92)
1+l 6 1+0.045 ×
360 360
¿
Finally, we plug the values into the expression for k

¿ 0.98988−0.97763 360
k= × =4.96 %
0.97763 91

40
(d) Valuing an FRA example: Suppose we enter into a long 3 ×6 FRA and we wish to value it after one
month has elapsed. After one month, the FRA becomes a 2 ×5 FRA. We have the following information.

 Fixed rate of 4.96%

 P=25,000,000
 2-month Libor is 5.50%, 5-month Libor is 6.00%
 There are 61 days in the first 2-month period, and 91 days in the next 3-month period

The value of the FRA is given by

{ [
P × B ( T 1 )− 1+ k
d [ T ,T ]
1

360
2

] ( )}
B T2

{ [
¿ 25,000,000 × B ( 2 )− 1+ 0.0496 ×
91
360 ] }
B (5) .

We need to determine B ( 2 ) and B(5).

1
B ( 2 )= =0.99077
61
1+ 0.055×
360
1
B (5 ) = =0.97529
(91+61)
1+ 0.06 ×
360
Substituting these terms, we get the current value of the FRA

{
25,000,000 × 0.99077− 1+0.0496 × [ 91
360 ]
0.97529 }
¿ 81,150.

41