MA 592: Mathematical Finance

Module 7: Forwards and Futures

Prof. Siddhartha Pratim Chakrabarty

Department of Mathematics
Indian Institute of Technology Guwahati
Forward Contract
1 Recall: A forward contract is an agreement to buy or sell an underlying
asset on a fixed date, in the future, called the delivery time, for a price
specified in advance, called the forward price.
2 The party to the contract which agrees to sell the asset is said to be
taking a short forward position.
3 The other party to the contract which agrees to buy the asset is said to
be taking a long forward position.
4 Let us denote the time of agreement on the forward contract as 0 and the
delivery time as T .
5 Let F (0, T ) be the corresponding forward price.
6 The price of the underlying asset at any time t is denoted by S(t).
7 No payment is made by either party at time t = 0, when the forward
contract is exchanged.
Forward Contract (Contd ...)
1 At delivery, the party with long forward position will benefit if
F (0, T ) < S(T ).
(A) They can buy the asset for F (0, T ) and sell for S(T ) to make a
profit of S(T ) − F (0, T ).
(B) The party holding the short forward position will suffer a loss of
S(T ) − F (0, T ).
2 If F (0, T ) > S(T ) then the situation will be reversed.
(A) Accordingly, the payoff at delivery are S(T ) − F (0, T ) (< 0) for a
long forward position and F (0, T ) − S(T ) (> 0) for the short
forward position, respectively.
3 Note: If the contract is initiated at time t < T rather than 0, then we
shall write F (t, T ) for the forward price, the payoff at delivery being
S(T ) − F (t, T ), for a long forward position and F (t, T ) − S(T ), for a
short forward position.
Forward Price for Stock Paying no Dividends
1 We begin with the simplest case of a stock paying no dividends. For a
stock paying no dividends, the forward price is:

F (0, T ) = S(0)e rT ,

where r is a constant risk-free interest rate, under continuous

compounding.
2 If the contract is initiated at time t ≤ T , then the forward price is:

F (t, T ) = S(t)e r (T −t) .

Forward Price for Stock Paying no Dividends: Proof
Suppose that F (0, T ) > S(0)e rT . In this case:
1 At time t = 0:
(A) Borrow an amount S(0), at rate r , till time T .
(B) Buy one share for S(0).
(C) Enter into a short forward contract with forward price F (0, T ), at
time T .
2 At time t = T :
(A) Sell the stock for F (0, T ).
(B) Pay S(0)e rT to clear the loan, with interest.
This will bring a profit of F (0, T ) − S(0)e rT > 0, in violation of the
no-arbitrage principle.
Forward Price for Stock Paying no Dividends: Proof (Contd ...)
Suppose that F (0, T ) < S(0)e rT . In this case:
1 At time t = 0:
(A) Short sell one share for S(0).
(B) Invest the proceeds at the risk free rate r , for time T .
(C) Enter into a long forward contract with forward price F (0, T ).
2 At time t = T :
(A) Cash the risk free investment with interest, collecting S(0)e rT .
(B) Buy the stock for F (0, T ) using forward contract.
(C) Close out the short position in stock by returning it to the owner.
This will bring a profit of S(0)e rT − F (0, T ) > 0, in violation of the
no-arbitrage principle.
Forward Price for Stock Paying no Dividends: Proof (Contd ...)
1 Thus we can only have:

F (0, T ) = S(0)e rT .

2 In a similar way, one can derive:

F (t, T ) = S(t)e r (T −t) .

3 Finally, under periodic compounding, the forward price is given by:

 r mT
F (0, T ) = S(0) 1 + .
m

Forward Price for Stock Paying Dividend

The forward price of a stock paying dividend div at time t, where 0 < t < T is:

F (0, T ) = S(0) − div × e −rt e rT .

 
Forward Price for Stock Paying Dividend: Proof
Suppose that F (0, T ) > S(0) − div × e −rt e rT .
 

1 At time t = 0:
(A) Enter into a short forward contract, with forward price F (0, T ), and
delivery time T .
(B) Borrow S(0), and buy one share.
2 At time t = t:
(A) Cash the dividend div , and invest it at risk free rate r , for the
remaining time T − t.
3 At time t = T :
(A) Sell the share for F (0, T ).
(B) Pay S(0)e rT to clear the loan with interest.
(C) Collect div × e r (T −t) .
The final balance is F (0, T ) − S(0)e rT + div × e r (T −t) > 0, which is in
violation of the no-arbitrage principle.
Forward Price for Stock Paying Dividend: Proof (Contd ...)
Suppose that F (0, T ) < S(0) − div × e −rt e rT .
 

1 At time t = 0:
(A) Enter into a long forward contract, with forward price F (0, T ), and
delivery at time T .
(B) Sell short one share, and invest the proceeds S(0), at the risk free
rate r .
2 At time t = t:
(A) Borrow div , and pay the dividend to the stock owner.
3 At time t = T :
(A) Buy one share for F (0, T ), and close out the short position in the
stock.
(B) Cash the risk free investment with interest, collecting the amount
S(0)e rT .
(C) Pay div × e r (T −t) to clear the loan with interest.
The final balance is −F (0, T ) + S(0)e rT − div × e r (T −t) > 0, which is in
violation of the no-arbitrage principle. Thus:

F (0, T ) = S(0) − div × e −rt e rT .

 
Forward Price for Stock Paying Dividends Continuously
The forward price of a stock paying dividends continuously at rate rdiv is:

F (0, T ) = S(0)e (r −rdiv )T .

Forward Price for Stock Paying Dividends Continuously: Proof

Suppose that F (0, T ) > S(0)e (r −rdiv )T .
1 At time t = 0:
(A) Enter into a short forward contract.
(B) Borrow the amount S(0)e −rdiv T , to buy e −rdiv T shares.
2 Between time 0 and T collect the dividends paid continuously. At time T ,
you will have short position in 1 share.
3 At time t = T :
(A) Sell the share for F (0, T ), closing out the short forward position.
(B) Pay S(0)e (r −rdiv )T , to clear the loan with interest.
The final balance is F (0, T ) − S(0)e (r −rdiv )T > 0, which is in violation of the
no-arbitrage principle.
Forward Price for Stock Paying Dividends Continuously: Proof (Contd ...)
Suppose that F (0, T ) < S(0)e (r −rdiv )T .
1 At time t = 0:
(A) Take a long forward position.
(B) Short sell a fraction e −rdiv T , of a share investing the proceeds
S(0)e −rdiv T , at risk free rate r .
2 Between time 0 and T , the short position in the stock will increase to 1
share, at time T .
3 At time t = T :
(A) Buy one share for F (0, T ), and return it to the owner, closing out
the long forward position, and the short position in the stock.
(B) Receive S(0)e (r −rdiv )T , from the risk free investment.
The final balance is S(0)e (r −rdiv )T − F (0, T ) > 0, which is in violation of the
no-arbitrage principle. Thus,

F (0, T ) = S(0)e (r −rdiv )T .

Value of a Forward Contract
1 Every forward contract has the value zero, when initiated. As time passes,
the price of the underlying asset may change.
2 Consequently, the value of the forward contract will keep changing and
will no longer be zero (as was the case initially).
3 At the time of delivery, the value of a long forward contract will be
S(T ) − F (0, T ), which could either be positive, zero or negative.
4 Suppose that the price for a forward contract for time T , initiated at time
t, is denoted by F (t, T ), where 0 < t < T .
5 Then the net gain/loss of the investor with the long position is
F (t, T ) − F (0, T ), as compared to an investor entering into a new long
forward contract, at time t, with the same delivery date T .
6 To find the value of the original forward position at time t, one needs to
discount this gain back to time t.
7 This discounted amount would be received (or paid, if negative) by the
investor with a long position should the forward contract initiated at time
0 be closed out at time t, which is earlier than the agreed delivery date T .
Value of a Forward Contract (Contd ...)
For any t such that 0 ≤ t ≤ T , the time t value of a long forward contract
with forward price F (0, T ) is given by:

V (t) = [F (t, T ) − F (0, T )] e −r (T −t) .

Value of a Forward Contract: Proof

Suppose that V (t) < [F (t, T ) − F (0, T )] e −r (T −t)
1 At time t:
(A) Borrow the amount V (t) to enter into a long forward contract, with
forward price F (0, T ) and delivery date T .
(B) Initiate a short forward position with forward price F (t, T ), at no
cost.
2 At time T :
(A) Close out the forward contracts collecting the amounts
S(T ) − F (0, T ), for the long position and −S(T ) + F (t, T ), for the
short position.
(B) Pay back the loan with interest, amounting to V (t)e r (T −t) in total.
The final balance F (t, T ) − F (0, T ) − V (t)e r (T −t) > 0, is the arbitrage profit.
Value of a Forward Contract: Proof (Contd ...)
Suppose that V (t) > [F (t, T ) − F (0, T )] e −r (T −t) .
1 At time t:
(A) Receive the amount V (t) to acquire a short forward contract with
forward price F (0, T ), and delivery date T (borrow and pay V (t), if
negative).
(B) Initiate a new long forward contact with forward price F (t, T ) at no
cost.
2 At time T :
(A) Close out both forward contracts receiving the amounts
F (0, T ) − S(T ) and S(T ) − F (t, T ), respectively (pay the amounts,
if negative).
(B) Collect V (t)e r (T −t) from the risk-free investment, with interest.
The final balance −F (t, T ) + F (0, T ) + V (t)e r (T −t) > 0, will be the arbitrage
profit.
Futures Contract
1 We assume that time is discrete with steps of length τ , typically a day.
2 Just like a forward contract, a futures contract involves an underlying
asset (say stock) with prices S(n) for n = 0, 1, . . . and time T , say.
3 In addition to the usual stock prices, the market also dictates the so called
futures prices f (n, T ), for each step n = 0, 1, . . . such that nτ ≤ T .
4 These prices are unknown at time 0, except for f (0, T ), and accordingly,
we shall treat them as random variables.
5 As in the case of a forward contract, it costs nothing to initiate a futures
position.
6 The difference lies in the cash flow during the lifetime of the contract.
7 A long forward contract involves just a single payment S(T ) − F (0, T ), at
delivery. A futures contract involves random cash flows, known as marking
to market.
8 This means that, at each time step n = 1, 2, . . . such that nτ ≤ T , the
holder of a long futures position will receive the amount
f (n, T ) − f (n − 1, T ), from the holder of the short futures position.
Futures Contract: An Example
1 Suppose that the initial margin is set at 10%, and the maintenance
margin at 5%, of the futures price.
2 The table below shows a scenario with “long” futures prices f (n, T ).
3 The columns labeled “Margin 1” and “Margin 2” show the deposit at the
beginning and at the end of each day, respectively.
4 The “Payment” column contains the amounts paid to top up the deposit
(negative numbers) or withdrawn (positive numbers).

n f (n, T ) Cash Flow Margin 1 Payment Margin 2

0 140 Opening 0 −14 14
1 138 −2 12 0 12
2 130 −8 4 −9 13
3 140 +10 23 +9 14
4 150 +10 24 +9 15
Closing 15 +15 0
Futures Contract: An Example (Contd ...)
1 On day 0, a futures position is opened and a 10% deposit paid.
2 On day 1, the futures price drops by 2, which is subtracted from the
deposit.
3 On day 2, the futures price drops further by 8, triggering a margin call,
because the deposit falls below 5%. The investor has to pay 9 to restore
the deposit to the 10% level.
4 On day 3, the forward price increases, and 9 is withdrawn, leaving a 10%
margin.
5 On day 4, the forward price goes up again, allowing the investor to
withdraw another 9. At the end of the day, the investor decides to close
the position, collecting the balance of the deposit.
6 The total of all payments is 10 (= −14 + 0 − 9 + 9 + 9 + 15), which is
equal to the increase in the futures price between day 0 and 4.
Theorem
If the interest rate r is constant, then f (0, T ) = F (0, T ).

Proof
For simplicity, suppose that the marking to market (for a futures contract) is
done at only two intermediate time points t1 and t2 , such that
0 < t1 < t2 < T . The argument given below can easily be extended to more
frequent marking to market.
Proof: Strategy Using Forward Contract
1 At time t = 0:
(A) We take a long forward position (at no cost) with the forward price
F (0, T ).
(B) Invest an amount of e −rT F (0, T ) in a risk-free account.
2 At time t = T :
(A) We receive an amount F (0, T ).
(B) Close the forward position by buying a share for F (0, T ).
(C) Sell the share for the market price S(T ).
Thus the final wealth is S(T ).
The idea is to replicate this payoff, S(T ), by using futures contracts, which we
examine in the next slide
Proof: Strategy Using Futures Contract
1 At time t = 0:
(A) We take a fraction e −r (T −t1 ) of a long futures position (at no cost).
(B) We invest an amount e −rT f (0, T ) in a risk-free account (this will
grow to v0 := f (0, T ) at time T ).
2 At time t = t1 :
(A) We receive (or pay if negative) the amount
e −r (T −t1 ) [f (t1 , T ) − f (0, T )] as a result of marking to market.
(B) We invest (or borrow if negative) e −r (T −t1 ) [f (t1 , T ) − f (0, T )] (this
will grow to v1 := f (t1 , T ) − f (0, T ) at time T ).
(C) We take long futures position for e −r (T −t2 ) of a contract (at no cost).
Proof: Strategy Using Futures Contract (Contd ...)
1 At time t = t2 :
(A) We receive (or pay if negative) an amount
e −r (T −t2 ) [f (t2 , T ) − f (t1 , T )] as a result of marking to market.
(B) We invest (or borrow if negative) e −r (T −t2 ) [f (t2 , T ) − f (t1 , T )] (this
will grow to v2 := f (t2 , T ) − f (t1 , T ) at time T ).
(C) We take long futures position for 1 of a contract (at no cost).
2 At time t = T :
(B) We receive an amount v0 + v1 + v2 = f (t2 , T ) from the risk-free
investments.
(B) We close the futures position by paying f (t2 , T ).
(B) We sell the stock for S(T ).
The final wealth level will be S(T ), as before. Thus, in order to avoid arbitrage
we have,
e −rT f (0, T ) = e −rT F (0, T ) ⇒ f (0, T ) = F (0, T ).
Hedging with Futures
One can hedge an exposure to stock price variations by entering into a forward
contract. However, an appropriate forward contract might not be easily
available, not to speak of the default risk. Instead one could hedge using the
futures market.

Hedging with Futures: An Example

1 Let S(0) = 100 and let the constant risk-free rate be r = 8%.
2 Assume that the marking to market takes place once a month, the time
1
step being .
12
3 Suppose that we intend to sell the stock after three months.
4 To hedge the exposure to stock prices, we enter into a futures contract,
with delivery in three months.
5 The payments from marking to market attract risk-free interest. The
results for two such stock price scenarios are given below.
Hedging with Futures: An Example (Contd ...)
 
n 3
n S(n) f , m2m Interest
12 12
0 100 102.02
1 102 103.37 −1.35 −0.02
2 101 101.68 +1.69 +0.01
3 105 105.00 −3.32 0.00
Total −2.98 −0.01

(A) In this case, one can sell the stock for 105.00, but marking to market
causes loss, bringing the sum to 105.00 − 2.98 − 0.01 = 102.01.
(B) If the marking to market did not attract interest, then the realized sum
would be 105.00
 − 2.98 = 102.02, which is exactly equal to the futures

3
price of f 0, .
12
Hedging with Futures: An Example (Contd ...)
 
n 3
n S(n) f , m2m Interest
12 12
0 100 102.02
1 98 99.32 +2.70 +0.04
2 97 97.65 +1.67 +0.01
3 92 92.00 +5.65 0.00
Total 10.02 +0.05

(A) In this case, one can sell the stock for 92.00 and along with the amount
for marking to market and interest accrued, receive a final amount of
92.00 + 10.02 + 0.05 = 102.07.
(B) Without the interest the final amount would be92.00 
+ 10.02 = 102.02
3
which is exactly equal to the futures price of f
0, .
12
A Few Observations
1 Some limitations arise because of the standardized nature of futures
contract. As a result there are difficulties in matching the terms of the
contract to the specific needs of the individual.
2 The exercise dates for futures are typically certain fixed days in a year, for
example, third Friday of March, June, September and December.
3 If we want to close out our investment at the end of April, then we need to
hedge with futures contracts with a delivery date after April, such as June.
Basis and Optimal Hedge Ratio
1 The difference between the spot price, and the futures price is called the
basis, and is given by:

b(t, T ) := S(t) − f (t, T ).

2 The basis converges to zero as t → T , since f (T , T ) = S(T ).

3 We consider the problem of designing a hedging strategy.
4 Suppose that we wish to sell the asset at time t < T .
5 In order to hedge ourselves against a potential fall in the price of the
asset, we short a futures contact with futures price f (0, T ) at time t = 0.
6 Accordingly, at time t we receive S(t) from selling the asset, in addition
to a cash flow of f (0, T ) − f (t, T ), resulting from marking to market (we
neglect any intermediate cash flow).
7 The total cash flow thus is:

f (0, T ) + S(t) − f (t, T ) = f (0, T ) + b(t, T ).

Basis and Optimal Hedge Ratio (Contd ...)
1 The price f (0, T ) is known at time 0, so the risk involved with the
hedging position is reflected by the basis.
2 The hedger seeks to minimize the risk associated with the basis. In order
to determine an optimal hedge ratio, the hedger enters into N futures
contract, where N need not necessarily be the number of units of the
underlying asset.
3 In order to determine the optimal N, we compute the risk as measured by
the variance of the basis bN (t, T ) := S(t) − Nf (t, T ), that is,
2
Var (bN (t, T )) = σS(t) + N 2 σf2(t,T ) − 2NσS(t) σf (t,T ) ρS(t),f (t,T ) .

4 The variance is a quadratic in N and attains a minimum at,

σS(t)
N = ρS(t),f (t,T ) ,
σf (t,T )

which is the optimal hedge ratio.