MAT 3788 Financial Mathematics

Prof. Boyan Kostadinov

Fall 2014

Contents
Financial Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Financial Derivatives

In this course, we are interested in special tradables called financial derivatives. The financial derivatives
derive their values from some underlying tradable(s), which may be a stock, bond, index, interest rate swap,
currencies, etc. For the tradable assets underlying financial derivatives, we’ll use the notation St to represent
the price of this tradable as a function of time t and we shall assume that this function is continuous with
time, although in reality there are jumps in price between trading days. Financial derivatives are therefore
not independent securities and they may enforce obligations or rights.
The main financial derivatives that we’ll consider are forwards and options.

Forwards

A Forward Contract is an agreement that enforces an obligation to buy or sell the underlying asset for
a certain price specified in the contract, called the forward price, at a certain future date, also specified
in the contract called the maturity date. The word obligation here is key. If we enter the forward contract
to buy the underlying in the future, we are long forward and if we enter the opposite side of the contract
to sell the underlying in the future, then we are short forward. Entering a forward on either side costs
nothing. In contrast, a spot contract is an agreement to buy or sell the asset today. Forward contracts are
traded only in the OTC (over the counter) market and not on the regulated exchanges because of the high
counter-party risk.
Payoffs from Forward Contracts
If we enter today a long position in a forward, then at the maturity date T (some time in the future) we will
be obligated to buy the underlying asset for the forward price K (also called delivery price) specified in the
contract, instead of buying the asset at the prevailing market price ST at time T . Keep in mind that ST , the
price of the asset at the future maturity date T is not known today and so, it represents a random variable
when viewed from today. Therefore, the payoff FT from a long position in a forward contract, on one unit of
the asset is

FT = ST − K

For example, if K = 100 and, when maturity comes, if ST = 120 then if we are long a forward we will be
obligated to buy the asset for 100 instead of the higher market price of 120, so our payoff is FT = 120−100 = 20,
since we buy the asset cheaply. The payoff for the short position of the same forward contract is the negative
of our payoff, namely −(120 − 100) = −20, so in general the payoff for the short position of a forward contract
is K − ST .
The payoffs from the long and short positions in a forward contract as a function of the variable ST are linear
and the graphs are mirror images to each other (w.r.t. the ST axis).

1
 Long Forward, K=100 Short Forward, K=100

100

100
Payoff = K − S(T)
50

50
Payoff = S(T)−K

0
−50

−50
−100

−100
0 50 100 150 200 0 50 100 150 200

ST ST

It costs nothing to enter into a forward contract, because you get only obligations, so nobody would want to
pay a premium for having obligations as opposed to rights. That’s why the payoff at maturity is the total
gain or loss for both parties in a forward. It is a zero-sum game in the sense that the payoff of the short side
is the negative of the payoff of the long side.
Note that while the long forward side has a limited downside risk of −K, the short forward side has a
potentially unlimited downside risk, which makes the short side of a forward a very risky position.

Options

Options are traded both on Exchanges and in the OTC market. There are two main types of vanilla options:
call and put options.
Call Options give the holder of the option the right but not the obligation to buy the underlying
asset at a certain price by a certain date.
Put Options give the holder of the option the right but not the obligation to sell the underlying asset
at a certain price by a certain date.
Option Features:

• the price specified in the contract is called the strike price or the exercise price and we use the letter
K to represent it.
• the date specified in the contract is called the maturity date or the exercise date and we use the letter
T to represent it.
• By default, Call and Put options are traded without requiring delivery of the underlying asset at
maturity. This means that at maturity if the long side has a positive payoff, they get from the short
side the dollar amount of their positive payoff rather than the underlying asset that they have the right
to buy for the strike price K. Of course, whether they get the asset for K instead of the market price of
ST (at maturity T ) or the option payoff (if positive) are two equivalent ways of cashing the long option.
• European options allow you to exercise the option only at maturity T .

• American options allow you to exercise the option at any time up to maturity.

Most options traded on exchanges are American. In the exchange-traded equity option market, one option
contract is an agreement to buy/sell 100 shares of the underlying asset. The key difference between forwards
and options is that in the case of options we have the right but not the obligation to buy/sell the underlying

2
asset. This means that we would not exercise the option if it is not in our interest to do so. Another difference
with forwards is that we have to pay a price to acquire an option. We pay this option premium for the rights
that we get. Unlike the forward contract, where both sides are symmetric in that they both have certain
obligations to buy or sell, options have asymmetric sides. We distinguish between a long position and a short
position in an option:

• Long option position is when we buy an option. It costs money to enter a long option position. We pay
a premium for the rights we get by buying the option.
• Short option position is when we sell an option. By selling the option we receive money (the option
premium) from the long side of the contract. Unlike the long position, the short position does not have
any rights but only the obligation to fulfill the contract in case the long side decides to exercise their
rights. The short side gets compensated with the option premium for agreeing to this obligation.

Selling an option is also called writing an option.

Long Call Option Payoff
Compare the definitions of a long forward position and a long call option position.
Is the long call option payoff at maturity = ST − K, like a long forward?
The answer is no because the long call option comes with rights, so you would choose not to exercise if you
would get a negative payoff. This optionality will simply remove the negative payoff from the long forward
payoff graph to produce a nonlinear payoff. Mathematically, we keep the positive part of the payoff and we
set the negative part to 0.
More precisely, while we can still buy the underlying at maturity T for the strike price K instead of the
prevailing spot market price ST , the call option payoff is CT = ST − K, provided the difference is positive
and if it is negative, we will choose not to exercise the option because it would not be in our interest to do so,
in which case our payoff at maturity would be 0. We are ignoring for now the option premium that the long
side has to pay.
Here is the long call option payoff at maturity T :
CT = (ST − K)+ = max(0, ST − K)

Here is the long put option payoff at maturity T :

PT = (K − ST )+ = max(0, K − ST )

The payoffs of the short sides are the negative of the payoffs of the long sides.
Long Call Option, K=100 Short Call Option, K=100
100

0
Payoff = max(0,K−S(T))

−20
80
60

Payoff

−60
40
20

−100
0

0 50 100 150 200 0 50 100 150 200

ST ST

3
 Long Put Option, K=100 Short Put Option, K=100

100

0
Payoff = max(0,K−S(T))

−20
80
60

Payoff

−60
40
20

−100
0

0 50 100 150 200 0 50 100 150 200

ST ST

Questions:

1. Is it true that the price of the long call is equal to the price of the short call?
2. Is it true that the price of the long put is equal to the price of the short put?
3. If you are long call, do you pay or receive the option premium?
4. If you are short put, do you pay or receive the option premium?

Put-Call Parity

Now, consider a portfolio of a long call and a short put. The payoff of this portfolio is linear, which allows
for model-independent pricing. This observation is the foundation of the so called Put-Call Parity. We’ll
come back to this later, when we start discussing risk-neutral pricing.

Examples

1. Suppose you want to speculate on a rise in the price of a certain stock. The current stock price is $29
and a 3-month call with a strike of $30 costs $2.90. You have $5,800 to invest. Describe two alternative
investment strategies, one involving an investment in the stock and the other involving investment in
the option. Describe the payoffs from both strategies.

• the first strategy is to buy shares of the stock; we can buy ∆S = 5800
29 = 200 shares today. The payoff
in 3 months will be ∆S ST − 5800, where ST is the price of the stock in 3 months.

curve(5800/29*x-5800,25,40,lwd=3,col="blue",ylab="Stock Payoff",xlab=expression(S[T]))
grid()
abline(h=0,lwd=2)

5800/29*40-5800 # payoff when S[T]=40

## [1] 2200

• the second strategy is to buy options. Since we speculate on a rise in the stock price, we would buy call
options. We can buy a total of ∆call = 5800
2.9 = 2000 call options today. The payoff in 3 months will be
∆call (ST − 30)+ − 5800, where ST is the price of the stock in 3 months.

4
curve(5800/2.9*pmax(0,x-30)-5800,25,40,lwd=3,col="blue",ylab="Option Payoff",xlab=expression(S[T]))
grid()
abline(h=0,lwd=2)

5800/2.9*max(0,40-30)-5800 # payoff when S[T]=40

## [1] 14200

Notice that in the going long stock strategy, the break-even point is $29, but in the going long call options
strategy, the break-even point is $32.9.

5800/2.9*max(0,32.9-30)-5800 # should be exactly 0 but is not because of numerical errors

## [1] -2.728e-12

We can list the payoffs from the two strategies for a few values of the asset price at maturity:

5800/29*(33:40)-5800 # payoff from holding stock for S[T]=33:40

## [1] 800 1000 1200 1400 1600 1800 2000 2200

5800/2.9*pmax(0,33:40-30)-5800 # payoff from holding calls for S[T]=33:40

## [1] 200 2200 4200 6200 8200 10200 12200 14200

We can now say that it’s worth going long call options only if you strongly believe that the stock price will go
above $34 in 3 months. Of course, nobody can give you guarantees about what would happen in 3 months.

Bull Spread

This is an example of an option trading strategy. We hold a call option with exercise price K1 and, for the
same asset and expiry date, we sell a call option with exercise price K2, where K2 > K1. At the expiry date,
the value of the first option is max(S(T ) − K1, 0) and the value of the second is max(S(T ) − K2, 0). Hence,
the overall payoff at expiry is

Payoff = max(S(T ) − K1, 0) − max(S(T ) − K2, 0).

The corresponding payoff diagram at maturity is plotted below. This combination gives an example of a
bull spread. We see from the figure that the holder of such a spread benefits when the asset price finishes
above K1, but gets no extra benefit if it is above K2. Note that this is not the net payoff since we are not
accounting for the price of the long call we are paying for and the price of the short call we are receiving
from the long side.

K1 <- 20 # strike price

K2 <- 40 # strike price
curve(pmax(x-K1,0)-pmax(x-K2,0),10,50,lwd=3,col="blue",xlab="S(T)",ylab="Payoff",main="Bull Spread with
grid()
abline(h=0,lwd=1,lty=2)

5
 Bull Spread with K1=20 and K2=40
20
15
Payoff

10
5
0

10 20 30 40 50

S(T)
Let today’s price of the asset be S0 = $30. Then, the call option with strike K1 = $20 is in the money
because even today it has a positive payoff max(0, S0 − K1) = 10. This makes the option expensive. The
other call option with strike K2 = $40 is out of the money because its payoff today is max(0, S0 − K2) = 0.
This makes the option cheap. So, the call option CK1 with strike K1 = $20 must be more expensive than the
call option CK2 with strike K2 = $40. Let’s say that the price of the call option CK1 is p1 = $4 and the
price of the call CK2 is p2 = $2. We pay the option premium p1 because we are long the call CK1 but we
receive the option premium p2 because we are short CK2 , that is we sell it to the long side. The net payoff at
expiry is then:

NetPayoff = max(S(T ) − K1, 0) − max(S(T ) − K2, 0) − p1 + p2

We can now visualize the net payoff of the bull spread, taking into account the option prices.

K1 <- 20 # strike price

K2 <- 40 # strike price
p1 <- 4 # price of C(K1)
p2 <- 2 # price of C(K2)
curve(pmax(x-K1,0)-pmax(x-K2,0)-p1+p2,10,50,lwd=3,col="blue",xlab="S(T)",ylab="Net Payoff",main="Bull Sp
grid()
abline(h=0,lwd=1,lty=2)

6
 Bull Spread with K1=20 and K2=40

15
Net Payoff

10
5
0

10 20 30 40 50

S(T)
Question: Find the break-even point from the net payoff.
Mathematically, this is the same as finding the root of a function f (x) = 0. In this case, the function f (x) is
our net payoff function:

f (x) = max(S(T ) − K1, 0) − max(S(T ) − K2, 0) − p1 + p2

We can define a vectorized version of this function in R, with default values for K1, K2, p1 and p2:

netPay<- function(S,K1=20,K2=40,p1=4,p2=2){pmax(S-K1,0)-pmax(S-K2,0)-p1+p2}
sol<-uniroot(netPay,c(20,30)) # look for roots of netPay in (20,30)
sol$root # extract the root from the sol object

## [1] 22

Thus, the break-even point is 22, so if the asset price goes above $ 22 then we are making money with the
bull-spread, otherwise we are losing money.

Bear Spread

The current price of an asset is $35 but an investor believes that the price will go down in 1 month. So, the
investor decides to speculate on this belief that the price is going down in 1 month and she decides to buy
for p1 =$4 a put with a strike price of K1 =$35 and sells for p2 =$2 a put with a strike price of K2 =$30.
Visualize the net payoff of this option strategy and explain why this strategy implements the investor’s belief.
Find the break-even point of this strategy.
The option strategy is to go long a put with a high strike and go short a put with a low strike.

7
netPay<- function(x,K1=35,K2=30,p1=4,p2=2){pmax(K1-x,0)-pmax(K2-x,0)-p1+p2}
curve(netPay(x),20,40,lwd=3,col="blue",xlab="S(T)",ylab="Net Payoff",main="Bear Spread")
grid()
abline(h=0,lwd=1,lty=2)

Bear Spread
3
2
Net Payoff

1
0
−1
−2

20 25 30 35 40

S(T)

sol<-uniroot(netPay,c(30,35)) # look for roots of netPay in (30,35)

sol$root # extract the root from the sol object

## [1] 33

Thus, the break-even point is 33, so if the asset price goes below $ 33 then we are making money with the
bear spread, otherwise we are losing money. Like bull spreads, bear spreads limit both the upside profit
potential and the downside risk. We can create bear spreads with calls instead of puts.

Speculating on Volatility

A bull spread is a bet that the price of the underlying asset will increase and a bear spread is a bet that the
asset price will decrease. These are directional bets. Options can also be used to create positions that are
nondirectional with respect to the underlying asset. With a nondirectional position, the holder does not care
whether the stock goes up or down, but only how much it moves (either way).
Examples of such nondirectional or volatility bets are: straddles, strangles, and butterfly spreads.

Straddles Consider the strategy of buying a call and a put with the same strike price and time to maturity.
This strategy is called a straddle. The general idea of a straddle is simple: If the stock price rises, there
will be a profit on the purchased call, and if the stock price declines, there will be a profit on the purchased
put. Thus, the advantage of a straddle is that it can profit from stock price moves in both directions. The

8
disadvantage to a straddle is that it has a high premium because it requires purchasing two options. If the
stock price at expiration is near the strike price, the two premiums are lost.
Let’s visualzie the profit diagram for a 40-strike straddle. The initial cost of the straddle at a stock price of
$40 is $5: $3 for the call and $2 for the put. The payoff of the two options at maturity is:

Payoff = max(S(T ) − K, 0) + max(K − S(T ), 0)

netPay<- function(x,K=40,p1=3,p2=2){pmax(x-K,0)+pmax(K-x,0)-p1-p2}
curve(netPay(x),30,50,lwd=3,col="blue",xlab="S(T)",ylab="Net Payoff",main="40-Strike Straddle")
grid()
abline(h=0,lwd=1,lty=2)

40−Strike Straddle
4
2
Net Payoff

0
−2
−4

30 35 40 45 50

S(T)
The break-even points are $35 and $45 and if the asset price goes below $35 and above $45 we make money,
otherwise, we lose money.
The net payoff diagram demonstrates that a straddle is a bet that volatility will be high: The buyer of an
at-the-money straddle is hoping that the stock price will move but does not care about the direction of the
move. Because option prices reflect the market’s estimate of volatility, the cost of a straddle will be greater
when the market’s perception is that volatility is greater (higher volatility, higher option prices).
Thus, purchasing a straddle is really a bet that volatility is greater than the actual market’s assessment of
volatility, as reflected in option prices.

Strangles The disadvantage of a straddle is the high premium cost. To reduce the premium, you can buy
out-of-the-money options (which are cheaper) rather than at-the-money options. Such a position is called a
strangle.
For example, consider buying a 35-strike put and a 45-strike call, for a total premium of $1. These transactions
reduce your maximum loss if the options expire with the stock near $40, but they also increase the stock-price
move required for a profit. Here is the payoff of the two options at maturity:

9
 Payoff = max(S(T ) − 45, 0) + max(35 − S(T ), 0)

netPay<- function(x,K1=35,K2=45,p=1){pmax(x-K2,0)+pmax(K1-x,0)-p}
curve(netPay(x),30,50,lwd=3,col="blue",xlab="S(T)",ylab="Net Payoff",main="Strangle")
grid()
abline(h=0,lwd=1,lty=2)

Strangle
4
3
Net Payoff

2
1
0
−1

30 35 40 45 50

S(T)
In this case, we can find the break-even points by finding the two roots of the netPay function.

sol1<-uniroot(netPay,c(30,35)) # look for roots of netPay in (30,35)

sol1$root # extract the root from the sol object

## [1] 34

sol2<-uniroot(netPay,c(45,50)) # look for roots of netPay in (45,50)

sol2$root # extract the root from the sol object

## [1] 46

Written Straddle What if an investor believes that volatility is lower than the market’s assessment?
Because a purchased straddle is a bet that volatility is high (relative to the market’s assessment), a written
straddle, which is selling a call and put with the same strike price and time to expiration—is a bet that
volatility is low (relative to the market’s assessment).
A written straddle is exactly the opposite of the purchased straddle. The written straddle is most profitable if
the stock price is $40 at expiration, and in this sense it is a bet on low volatility. This is a risky bet because
of the potential for a big loss. A large change in the stock price in either direction leads to a large, potentially
unlimited, loss.

10
netPay<- function(x,K=40,p1=3,p2=2){pmax(x-K,0)+pmax(K-x,0)-p1-p2}
curve(-netPay(x),30,50,lwd=3,col="blue",xlab="S(T)",ylab="Net Payoff",main="Written Straddle")
grid()
abline(h=0,lwd=1,lty=2)

Written Straddle
4
2
Net Payoff

0
−2
−4

30 35 40 45 50

S(T)

Butterfly Spreads The straddle writer can insure against large losses on the straddle by buying options to
protect against losses on both the upside and downside. Buying an out-of-the-money put provides insurance
on the downside, protecting against losses on the at-the-money written put. Buying an out-of-the-money call
provides insurance on the upside, protecting against losses on the written at-the-money call.
Consider the straddle written at a strike price of $40, along with the two extra options to safeguard the
position: a 35-strike put and a 45-strike call. The net result of combining these three strategies is an insured
written straddle, which is called a butterfly spread. The payoff of the butterfly spread at maturity is:

Payoff = max(S(T ) − 45, 0) + max(35 − S(T ), 0) − max(S(T ) − 40, 0) − max(40 − S(T ), 0)

The net payoff diagram of the butterfly spread can be understood as the net payoff of buying a strangle and
selling a straddle:
Butterfly Spread = Buy Strangle and Sell Straddle

netPay<- function(x,K=40,K1=45,K2=35,p=1,p1=3,p2=2){pmax(x-K1,0)+pmax(K2-x,0)-pmax(x-K,0)-pmax(K-x,0)+p1
curve(netPay(x),30,50,lwd=3,col="blue",xlab="S(T)",ylab="Net Payoff",main="Butterfly Spread")
grid()
abline(h=0,lwd=1,lty=2)

11
 Butterfly Spread

4
3
Net Payoff

2
1
0
−1

30 35 40 45 50

S(T)
Comparing the butterfly spread to the written straddle, we see that the butterfly spread has a lower maximum
profit if the stock at expiration is close to $40 but the potential losses are limited in both directions.
We can compare the payoff diagrams of the written straddle (blue) and the butterfly spread (red).
4
2
Net payoff

0
−2
−4

30 35 40 45 50

S(T)

12
Box Spreads

A box spread is accomplished by using options to create a synthetic long forward at one price and a synthetic
short forward at a higher price. For example, we can simultaneously enter into the following option positions:

1. Buy a 40-strike call and sell a 40-strike put.

payoff1<-function(x,K1=40) pmax(x-K1,0)-pmax(K1-x,0)
curve(payoff1(x),30,50,lwd=3,col="blue",xlab="S(T)",ylab="Payoff",main="Long 40-Forward",ylim=c(-2,5))
abline(h=0,lwd=1,lty=1)
abline(v=0,lwd=1,lty=1)

Long 40−Forward
5
4
3
Payoff

2
1
0
−2 −1

30 35 40 45 50

S(T)

2. Sell a 45-strike call and buy a 45-strike put.

payoff2<-function(x,K2=45) -pmax(x-K2,0)+pmax(K2-x,0)
curve(payoff2(x),30,50,lwd=3,col="blue",xlab="S(T)",ylab="Payoff",main="Short 45-Forward",ylim=c(-2,5))
abline(h=0,lwd=1,lty=1)
abline(v=0,lwd=1,lty=1)

13
 Short 45−Forward
5
4
3
Payoff

2
1
0
−2 −1

30 35 40 45 50

S(T)

• What is the payoff of the total option strategy at maturity?

Payoff = max(S(T ) − K1, 0) − max(K1 − S(T ), 0) − max(S(T ) − K2, 0) + max(K2 − S(T ), 0)

payoff<-function(x,K1=40,K2=45) pmax(x-K1,0)-pmax(K1-x,0)-pmax(x-K2,0)+pmax(K2-x,0)
curve(payoff(x),0,100,lwd=3,col="blue",xlab="S(T)",ylab="Payoff",main="Box Spread",ylim=c(0,6))
abline(h=0,lwd=1,lty=1)
abline(v=0,lwd=1,lty=1)

14
 Box Spread
6
5
4
Payoff

3
2
1
0

0 20 40 60 80 100

S(T)

• Justify mathematically why we are getting the horizontal line y = 5.

The payoff at maturity of the long 40-forward is y1 = x − 40, and the payoff of the short 45-forward is
y2 = 45 − x. Thus, the total payoff of adding the long 40-forward and the short 45-forward is:

y = y1 + y2 = x − 40 + 45 − x = 5

• How much would you pay for this option strategy?

The payoff does not depend on the price of the underlying asset and it is therefore not a random variable
at maturity (when viewed from today) - it is a known today number; $5 in our case. So, the payoff is the
constant $5 at maturity, and since this is a constant (known today) number, we can discount it back to today
using the risk-free annual interest rate r. The discount factor for continuous compounding is e−rT , if T is the
time to maturity, in years. Thus, the fair price today of this option strategy would be:

Price = 5e−rT

Of course, we would not be able to do this if the payoff at maturity is a non-trivial function of the price of
the asset at maturity, S(T ), which is a random variable (when viewed from today) and therefore it would not
allow to use the “risk-free” discount rate to discount risky payoff.

• What happens if the net price (today) of the 4 individual options (2 calls and 2 puts) that make up
this option strategy does not add up to 5e−rT ?

We buy a 40-strike call and sell a 40-strike put for a price of: −call40 + put40
We sell a 45-strike call and buy a 45-strike put for a price of: call45 − put45

15
The total price of this strategy is: −call40 + put40 + call45 − put45 and if it happens that

−call40 + put40 + call45 − put45 6= 5e−rT

then we say that there is an arbitrage in the market and the good traders would take advantage of this
arbitrage to make risk-free money.

• Box Spreads will lead to losing money if you use American options rather than European and many
exchanges trade only American options. One or both of the options that you sell maybe exercised
against you almost immediately if they get into the money, which is allowed if they are American options,
but this will destroy your constant payoff at maturity and you are losing the otherwise guaranteed $5
at maturity.
• In the past, Box Spreads were used to cheat on taxes but the IRS updated their outdated tax laws to
prevent this from happening. The tax loophole was related to the fact that bonds and options were
taxed differently but investors can synthetically create bonds using options, which creates tax problems
if bonds and options are taxed differently. In fact, the Box Spread is an example of a synthetic creation
of a zero-coupon bond (you pay something now to get a certain amount of (more) money later, without
any intermediate payments).

In-Class Exercises:

1. Visualize the net payoff at maturity of the following options strategy:

• buy 50-strike call, whose price is $2.

• buy 45-strike put, whose price is $3.
• What’s the name of this option strategy?

• Is this a directional or volatility bet?

• What kind of beliefs about the price of the underlying asset at maturity you need to have in order to
put together this kind of strategy in the hope of making money?

2. Visualize the payoff at maturity of the following options strategy:

• buy one 35-strike call.

• sell two 40-strike calls.
• buy one 45-strike call.

• What’s the name of this kind of strategy?

• Is this a directional or volatility bet?
• What kind of beliefs about the price of the underlying asset at maturity you need to have in order to
put together this kind of strategy in the hope of making money?

16
Homework:

1. Insert the word ‘rise’ or ‘fall’ to complete the following sentences:

• The holder of a European call option hopes the asset price will . . .
• The writer of a European call option hopes the asset price will . . .
• The holder of a European put option hopes the asset price will . . .
• The writer of a European put option hopes the asset price will . . .

2. The current price of a stock is $94 and a 3-month European call options with a strike price of $95
currently sell for $4.70. An investor who feels that the price of the stock will increase is trying to decide
between buying 100 shares and buying 2000 call options (= 20 contracts). Both strategies involve an
investment of $9,400. What advice would you give? How high does the stock price have to rise for the
option strategy to be more profitable?

3. Convince yourself that max(S(T ) − K, 0) + max(K − S(T ), 0) is equivalent to |S(T ) − K| and draw
the payoff diagram for this so called bottom straddle.
4. Suppose that for the same asset and expiry date, you hold a European call option with exercise price
K1 and another with exercise price K3, where K3 > K1 and also write two calls with exercise price
K2 = K1+K32 . This is an example of a butterfly spread. Derive a formula for the payoff of this
butterfly spread at expiry and draw the corresponding payoff diagram.

Pension Fund growth under fixed compound interest

# Inputs
r <- 0.05 # Annual interest rate, assumed fixed for the future
term <- 30 # Forecast duration (in years)
period <- 1/12 # Time between payments (monthly, in years)
payments <- 1000 # amount deposited each period, beginning at the start of period 2
# Calculations
n <- floor(term/period) # (integer) number of payments made at the end of each period
pension <- 50000 # starting pension fund, at the start of period 1 (time 0)
for (i in 1:n) {
pension[i+1] <- pension[i]*(1 + r*period) + payments
}
time <- (0:n)*period # time vector in years
# Output
plot(time, pension,"l",col="blue",lwd=2)

17
 1e+06
6e+05
pension

2e+05

0 5 10 15 20 25 30

time
tail(pension)

## [1] 1028987 1034275 1039584 1044916 1050270 1055646

Loan Repayments

Suppose that a loan has an initial amount P , a monthly interest rate r, a duration of N months, and a
monthly repayment of A. The remaining debt after n months is given by Pn , where P0 = P and

Pn+1 = Pn (1 + r) − A

That is, each month you pay interest on the remaining balance, then reduce the balance of the loan by amount
A.

Pension Fund growth with random rate of return

We can modify this code to make the fixed interest rate a random rate of return, having a Normal distribution.
In particular, we want to make the interest rate r a random variable in the future, so that for the first
period r = 0.05 is still known but starting with period 2, r should be sampled from the Normal distribution
N (µ = 0.05, sd = 0.20), using the R code rnorm(1,mean=0.05,sd=0.2). We assume that this Normal
distribution remains the same over time but it’s possible to have the mean and the standard deviation
as functions of time. We’ll plot 100 realizations of our pension growth curve using the built-in function
replicate and the plotting function matplot. Finally, we’ll compute the summary of sample statistics of
our terminal pension value, at the end of the 30Y term, using the built-in function summary applied on the
sample of terminal pension values:

# Inputs
r <- 0.05 # Annual rate of return, assumed known only for the first period
term <- 30 # Forecast duration (in years)
period <- 1/12 # Time between payments (monthly, in years)

18
payments <- 1000 # amount deposited each period, beginning at the start of period 2
pension <- 50000 # starting pension fund, at the start of period 1 (time 0)
r.mean<-0.05 # expected value of rate of return
r.sd<-0.20 # volatility of rate of return
# Calculations
n <- floor(term/period) # (integer) number of payments made at the end of each period

pension.curve<-function(term,period,payments,pension,r,r.mean,r.sd){
for (i in 1:n) {
pension[i+1] <- pension[i]*(1 + r*period) + payments
r<-rnorm(1,mean=r.mean,sd=r.sd) # a single Normal sample used for the next period
}
return(pension)
}

# replicate this random exp. many times

pension.sample<-replicate(100,pension.curve(term,period,payments,pension,r,r.mean,r.sd))
time <- (0:n)*period # time vector in years
matplot(time,pension.sample,type="l")
1500000
pension.sample

500000
0

0 5 10 15 20 25 30

time
# to compute good sample stats it's better to have a large sample
large.pension.sample<-replicate(1e2,pension.curve(term,period,payments,pension,r,r.mean,r.sd))
term.sample<-large.pension.sample[n+1,] # terminal sample at the end of the term
summary(term.sample) # sample stats of terminal pension

## Min. 1st Qu. Median Mean 3rd Qu. Max.

## 560000 873000 997000 1040000 1210000 1870000

hist(term.sample,50,col="blue")

19
 Histogram of term.sample
5
4
Frequency

3
2
1
0

600000 800000 1200000 1600000

term.sample

Simulating Random Rate of Return

An investor is considering investing $7000 for 5 years. At the end of 5 years the expected value of the
investment is expected to be $12,000, and the 5Y investment value is assumed to have a Normal distribution
with σ = $2, 500. The investor would like to know the sample distribution of the implied annual rate of return
on this investment and its key summary statistics. The investor would also like to know the probability of
achieving a rate of return of less than 8%, as well as the probability of having a rate of return of at least 10%.
The annual rate of return r is defined as follows:

F = 7000(1 + r)5 ,

where F is the future, 5Y, value of the investment. We can solve for r to get:

  15
F
r= − 1,
7000

or we can use a different way of expressing the same thing:

1
  
F
r = exp log − 1,
5 7000

Thus, r is a random variable because F ∼ N (µ = 12000, σ = 2500). We can easily generate a sample of r
using a large sample of F , and compute all stats based on this r-sample:

set.seed(123)
Fsample<-rnorm(1e4,mean=12000,sd=2000) # a Normal sample of size 1 million
summary(Fsample)

20
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 4310 10700 12000 12000 13300 19700

rsample<-(Fsample/7000)^(1/5)-1 # all operations act elementwise on the vector Fsample

hist(rsample,25,col="blue") # freq. histogram

Histogram of rsample
800
Frequency

600
400
200
0

−0.10 −0.05 0.00 0.05 0.10 0.15 0.20

rsample

summary(rsample) # sample summary stats

## Min. 1st Qu. Median Mean 3rd Qu. Max.

## -0.0925 0.0878 0.1130 0.1110 0.1380 0.2300

mean(rsample<0.08) # probability r<0.08

## [1] 0.1959

mean(rsample>=0.1) # probability r>=0.1

## [1] 0.6457

If we increase the σ of F , then Fsample would have some small number of negative values, since the Normal
distribution of F would then have a wider spread and cross into negative territory. If we use the same code
for rsample rsample<-(Fsample/7000)ˆ(1/5)-1, when Fsample is negative we would get NaN’s (not a real
1
number) since (negative number) 5 leads to 5 complex roots. The good news is that one of these 5 complex
roots is actually a real root because the rational power 15 has an odd number in the denominator. If the
denominator of the rational power was even then all 5 roots would have been complex.
Questions:

21
 1
• List all 5 complex roots of (−1) 5 and then argue why you always get a real root, when the denominator
of the rational power is an odd number.
1 1
• Justify the claim that (−2) 5 has a real root of the form −2 5 , or more generally, if a < 0 (a negative
1 1
number), then the real root of a 5 is −|a| 5 .

With this understanding we can rewrite the code for rsample to make sure that we get real numbers, even
if Fsample happens to be negative. Notice that it is very important here that we are dealing with an odd
number of 5 years.

set.seed(123)
Fsample<-rnorm(1e4,mean=12000,sd=5000) # much bigger sigma leading to negative values of F
summary(Fsample)

## Min. 1st Qu. Median Mean 3rd Qu. Max.

## -7230 8660 11900 12000 15400 31200

# to get real numbers and avoid NaN's

rsample<-ifelse(Fsample>=0,(Fsample/7000)^(1/5)-1,-(abs(Fsample)/7000)^(1/5)-1)
hist(rsample,25,col="blue") # freq. histogram

Histogram of rsample
4000
3000
Frequency

2000
1000
0

−2.0 −1.5 −1.0 −0.5 0.0 0.5

rsample

summary(rsample) # sample summary stats

## Min. 1st Qu. Median Mean 3rd Qu. Max.

## -2.0100 0.0435 0.1130 0.0832 0.1700 0.3490

22
mean(rsample<0.08) # probability r<0.08

## [1] 0.363

mean(rsample>=0.1) # probability r>=0.1

## [1] 0.5578

Exercises

1. An investor is considering investing $7000 for 5 years. At the end of 5 years the investment is assumed
to be uniformly distributed with expected value of $12,000 and a spread around the expected value
of ± $5000. The investor would like to know the sample distribution of the implied annual rate of
return on this investment and its key summary statistics. The investor would also like to know the
probability of achieving a rate of return of less than 8%, as well as the probability of having a rate of
return of at least 10%.

2. An investor is considering the purchase of rental property. The initial cost of the property is $1,000,000.
The annual income from rental is estimated to have an expected value of $200,000 and the resale value
of the building in 10 years is expected to be $2,000,000. The rental income is not certain and can
be modeled as a N (µ = 2e5, σ = 5e4). Similarly, the resale value of the building is estimated to be
Normally distributed N (µ = 2e6, σ = 8e5). The investor uses a discount rate of 10%.

• Create a freq. histogram for a sample of size 1e6 of the present value of the investment.
• What is the expected present value of the investment?
• What is the standard deviation of the investment’s present value?
• Set a 95% confidence interval on the present value of the investment.
• What is the probability that the investment will be profitable?

3. Consider an investment requiring an initial investment of $1000. After 2 years the investment will be
worth F dollars. Assume that F is a uniform random variable F ∼ U (1200, 1600).

• Visualize the sample distribution of the rate of return of this investment by plotting the density
histogram of a large sample of the rate of return.
• Compute analytically the pdf of the rate of return and plot the pdf over the density histogram you
previously created. If your work is correct the analytical pdf should closely follow the shape of the
density histogram.
• Compute the expected rate of return both analytically and using simulation.
• Compute the rate of return using the expected cash flows.

23