Call Options

Call options
n Gives its owner the right
q But not the obligation
n To buy something
q The underlying
n At a specified price
q The exercise or strike price
n On or before a specified date
q The maturity or expiration date
n European options can only be exercised at maturity
n American options can only be exercised any time before maturity

FIN 206, Binomial Pricing Prof. Giulio Trigilia 2

 Put Options
Put options
n Gives its owner the right (but not the obligation)

n To sell the underlying

n At the strike

n On or before the expiration date

n Calls are like:

q Long forwards
q Where you refuse delivery
n Puts are like:
q Short forwards
q Where you choose not to deliver

FIN 206, Binomial Pricing Prof. Giulio Trigilia 3

 Example
n On January 11, 2003, Intel stock price was S = 17.42
n Intel calls maturing on 2/03 with a strike of K = 17.50
were trading at 1.15
n Puts with the same strike and maturity were trading at
1.30
n Terminology:
q Party that holds the option to exercise is long the option
n Long side always has value
n Long puts have a short exposure to the underlying price risk
q Counterparty “writes the option,” has a short position
n Is paid to do so
n Collects the “option premium”

FIN 206, Binomial Pricing Prof. Giulio Trigilia 4

 “Moneyness”
An option is
n In-the-money (ITM) if it’s worth something if exercised
today.
n Out-of-the-money (OTM) if it would costs something to
exercise today
n At-the-money (ATM) if S = K

n Intrinsic value:
q Value if the maturity date was right now

FIN 206, Binomial Pricing Prof. Giulio Trigilia 5

 Payoff diagrams

FIN 206, Binomial Pricing Prof. Giulio Trigilia 6

 Binomial Option Pricing

FIN 206, Binomial Pricing Prof. Giulio Trigilia 7

 Key Intuition (B-S)
Suppose you own a call option
■ Then if the underlying goes up (right now) you
make money
■ So you can short some stock, so that you’re
indifferent to whether the underlying goes up or
down
q Over the next instant
q If you sold the exact right amount
■ Two Questions:
q What is the exact right amount?
q Once you’ve “hedged,” what’s your expected return?

FIN 206, Binomial Pricing Prof. Giulio Trigilia 8

 Binomial pricing
Key intuition
■ Option payoffs are non-linear in the underlying

■ But are approximately linear in the underlying

for small moves

q Call goes up about 10x as much

q We’ll string lots of these together to get a final price

n Small moves only, so we can limit our attention to “ticks”
q I.e., assume the stock price can only go “up” or “down”
q Since we’re going to string a lot of moves together, we’ll be able
to get almost any stock price we want at the end

FIN 206, Binomial Pricing Prof. Giulio Trigilia 9

 Binomial vs. normal
Mean zero, standard deviation of one

4 steps 16 steps

64 steps 256 steps

FIN 206, Binomial Pricing Prof. Giulio Trigilia 10

 The “Tree” possible
underlying prices
at option’s
maturity
full tree consists of
lots of one-period
initial
trees
underlyin
g price

more “paths”
to prices in
the middle of
the
distribution

■ So we’ll look at 1-period trees first

■ Still need to address where the tree comes from
q We’ll hold off on this for a bit

FIN 206, Binomial Pricing Prof. Giulio Trigilia 11

 Pricing on a 1-period tree

■ Stock price tree

q 1-period simple discount rate is r

FIN 206, Binomial Pricing Prof. Giulio Trigilia 12

 Pricing on a 1-period tree
■ Call option tree
Note: there’s
nothing special
about the call here;
“c” could be a
generic derivative
(“contingent claim”)

■ What about an arbitrary portfolio of stocks and

bonds?
q What does it look like on the tree?
n Long “Δ” shares of stock financed with “D” dollars worth of debt
q Can it replicate the call payoffs?

FIN 206, Binomial Pricing Prof. Giulio Trigilia 13

 An example

■ Suppose S = 60, u = 1.5, and d = 0.5.

■ Also, r = 0.2.
q A dollar invested today risk-free is worth $1.20
“tomorrow”

FIN 206, Binomial Pricing Prof. Giulio Trigilia 14

 One-period ATM call

■ Call option tree

■ Now consider the following portfolio

q Long half a share of stock
q Financed (partly) with $12.50 of borrowing

FIN 206, Binomial Pricing Prof. Giulio Trigilia 15

 The “synthetic”

■ Portfolio payoffs

■ Exactly replicates the calls payoffs!

■ So what’s the call worth?
q 60/2-12.5 = $17.50, or else there’s an arbitrage!
n What if you saw the call trading at $16.50? Or 18.50?

FIN 206, Binomial Pricing Prof. Giulio Trigilia 16

 Why this works

■ There’s always a replicating portfolio

q Option values are non-linear in the underlying
q But by assuming only two possible stock prices,
we’ve made the payoff linear
n Technical jargon: the stock and bond span all possible
payoffs tomorrow
q At least in a binomial world

■ Can we see this graphically?

FIN 206, Binomial Pricing Prof. Giulio Trigilia 17

 Payoffs
Su

Underlying CK(T)
Payoff

Sd

K
S(T)

FIN 206, Binomial Pricing Prof. Giulio Trigilia 18

 Binomial assumption
Su
Su

Underlying CK(T)
Payoff

Sd Cu
Sd

Cd

Sd K Su S(T)

FIN 206, Binomial Pricing Prof. Giulio Trigilia 19

 Spanning
Su
Su

Underlying CK(T)
Payoff

Sd Cu
Sd

Cu - Cd
Cd

(1+r)D Sd
Su - Sd
Su S(T)

FIN 206, Binomial Pricing Prof. Giulio Trigilia 20

 Replication more generally

■ In order to replicate the call options “payoffs”

c= ·S+L

■ Need
uS (1 + r )D = cu

dS (1 + r )D = cd

FIN 206, Binomial Pricing Prof. Giulio Trigilia 21

 ThedSstock
(1 + rposition
)D = cd

■ This implies
cu cd cu cd
uS dS = cu cd ) = =
Su Sd uS dS

q “Hedge ratio” = slope of call price w.r.t underlying

■ Important: comes from no-arbitrage

q Does not depend on any probabilities!

FIN 206, Binomial Pricing Prof. Giulio Trigilia 22

 cu cd cu cd
uS dS = cu
The cash position
cd ) =
Su Sd
=
uS dS

■ Then

(1 + r )D = cu )
uS
✓ ◆
cu cd dcu ucd
(1 + r )D = uS cu = · uS cu =
uS dS u d

dcu ucd
D=
(1 + r )(u d)

FIN 206, Binomial Pricing Prof. Giulio Trigilia 23

 D=
(1 + r )(u d)
Option Leverage
cu cd
■ Call is a levered claim
= on
uS dS the
>underlying
0
cu cd
= >0
uS dS
q And
dcu ucd
D=
(1 + r )(u d)
max(duS dK, 0) max(udS uK, 0)
dc=u ucd 0
= (1 + r )(u d)
(1 + rA)(u d)
long stock position financed partly through borrowing
n

max(duS dK, 0) max(udS uK, 0) 24

= FIN 206, Binomial Pricing Prof. Giulio Trigilia
0
 Put Example

■ Intel (again), still priced at 20

q Will go to 18 or 23 next period, r = 0.08.

■ What’s the price of the 1-period ATM put?

q How much Intel is it short?

q How big is the cash position?

n So what’s it worth?

FIN 206, Binomial Pricing Prof. Giulio Trigilia 25

D= Put Example
dcu ucd
(1 + r )(u d)

■ In max(duS
order to replicate
dK, 0)the max(udS
put options uK,
“payoffs”
0)
= 0
(1 + r )(u d)

■ Need
uS (1 + r )D = pu

dS (1 + r )D = pd
■ Exactly as before, but with “p” instead of “c”

FIN 206, Binomial Pricing Prof. Giulio Trigilia 26

 Put Example
pu pd cu cd
p = =
Su Sd uS dS
dpu upd
Dp =
(1 + r )(u d)
■ Just like for the call
q Only now Δ ≤ 0 and D ≤ 0.
n Short stock, long bonds.

1 27
FIN 206, Binomial Pricing Prof. Giulio Trigilia
 Risk-neutral pricing
■ We typically don’t actually bother with
replication
q We use a “short cut” called r.n.-pricing
n It’s still based on replicating portfolios
n But combines the construction and pricing in one simple step

■Just substitute our formulas for Δ and D into the

replicating portfolio
✓ ◆
cu cd dcu ucd
c= S D= ·S
uS dS (1 + r )(u d)
q Now we need to do some simplifying algebra d

FIN 206, Binomial Pricing Prof. Giulio Trigilia 28

 cu cd dcu ucd
c= S D= ·S
Risk Neutral Pricing
uS dS (1 + r )(u d)
1+r d
+ u
cu
+ r )(cu cd ) (dcu ucd ) u d
✓ ◆ =
■ Rewrite
(1 u rusing
c+ c)(u
d a common
d) dcu denominator
ucd 1+r
D= ✓ ·S ◆
uS dS cu cd(1 + r )(u dc d)u ucd
c= S D= ·S
uS dS 1+r (1
d + r )(u u d)
1 r
+ r )(cu cd ) (dcu ucd ) u d cu + u d cd
= 1+r d
c + u 1 r
cd
(1 (1
+ r+)(u
r )(cud) cd ) (dcu ucd ) 1+ u rd u
=
(1 + r )(uq =
d)
1 + r = d
1+r
u d

u d
■ Defining q = 1 + r d
then yields
u d
You must qcu + (1 q)cd
know this! cu =
qc + (1 q)cd
c= 1+r
1+r
FIN 206, Binomial Pricing Prof. Giulio Trigilia 29
 1+r d
cu + u u 1 d r cd
“Risk-neutral probabilities”
(1 + r )(cu cd ) (dcu ucd ) u d
=
(1 + r )(u 1
d)+ r d= 1+r
q=
u d
■ Interpreting “q” as the 1“probability”
+r d of the up-
move, this just saysq = u d
qcu + (1 q)cd E Q [ct+1 ]
c= = Q
c=
1qc+ r
u + (1 q)cd
=
E 1[c+
t+1 ]r

q q is NOT the “objective”

1 + probability
r 1+r
n It’s the r.n. probability (why?)
 ✓ ◆ ✓ ◆
Q St+1 1+r d u 1 r
E = qu + (1 q)d = u+ d =1+r
St u d u d

n If up-prob. was q, then stock’s expected growth rate is the risk-

free
q And so is the calls, and every derivative on S

FIN 206, Binomial Pricing Prof. Giulio Trigilia 30

 Another interpretation of RNP

■ When we priced off the tree, probability of the

“up” move was completely irrelevant!
q At least for option pricing
n It does get into the current price of the stock,
n But, given the stock price, we never use it
■ What does this mean for options on two
different stocks, with
q The same trees, but
q Different expected returns?

FIN 206, Binomial Pricing Prof. Giulio Trigilia 31

■ Two stocks: A and B

pA = 0.6 pB = 0.8

■ What are the two stocks’ expected returns?

q Simple, one-period
■ What are the prices of ATM calls on the two?

FIN 206, Binomial Pricing Prof. Giulio Trigilia 32

■ We don’t actually care about the stock, per se
q Only its tree!
■ So if we can find a different stock– with the
same tree– that has options we can price
some other, easier way, then we absolutely
should
q If it has the exact same tree, then its options will all
have the exact same prices.

■ Where should we look for this other tree?

FIN 206, Binomial Pricing Prof. Giulio Trigilia 33

 A “risk-neutral world”!

■ This is a huge leap!

q We actually hypothesize a completely fictitious
world
q Where investors don’t care about risk
■ This is where we’ll price options
q Why?
■ On this world, don’t need to worry about
replication
q Assets are just priced as expected payoffs
n Discounted for the time-value of money

FIN 206, Binomial Pricing Prof. Giulio Trigilia 34

 Option pricing in a “risk-
neutral world”
■ Extremely powerful tool:
q Reduces option pricing to “integration”
q I.e., calculating the expected payoff
n Or adding up all the possible payoffs (the integration)
and averaging
q Remember: these are the payoffs in the fictitious
world!
■ Binomial trees, Black-Scholes techniques
(later), Monte Carlo (even later) really all just
ways to “do the integration”

FIN 206, Binomial Pricing Prof. Giulio Trigilia 35

 1+r d
q=
Expectedu payoffs?
d

■ To find these, need the probability of the up-

move qcu + (1 q)cd E Q [ct+1 ]
c= =
q In the risk-neutral
1 world
+r 1+r
Su = uS
q
S
✓ ◆ ✓ ◆
+1 1+r d u 1 r
= qu + (1 q)d = u+ d
t u d Sd = uS u d
■ Stock priced as discounted expected price:
quS + (1 q)dS 1+r d
S= ) q=
(1 + r ) u d

FIN 206, Binomial Pricing Prof. Giulio Trigilia 36

■ Use u, d and q to forecast possible prices
tomorrow
q In the fictitious world
■ Price any derivative using these forecasts
q To calculate the derivatives expected payoff

FIN 206, Binomial Pricing Prof. Giulio Trigilia 37

 State
qcu +prices
(1 q)cd E Q [ct+1 ]
c= =
1+r 1+r
■ One last way to think of the risk-neutral
 probability ✓ ◆ ✓ ◆
St+1q q/(1+r) is today’s price 1of+a rdollar u 1 r
d in the up-state
= qu + (1 q)d = u+ d =1
St n A dollar you get if the stocku goes
d up, and don’t get
u otherwise
d
n It’s also the price of the up-security (or “Arrow” up-security)
q Pays a dollar in the up-state, and nothing otherwise
quS + (1 q)dS 1 1+r d
S= ) q=
(1 +Aur ) u d
0
✓ ◆
u 1 d q
A = S D= S =
uS dS (1 + r )(u d) 1+r

FIN 206, Binomial Pricing Prof. Giulio Trigilia 38

 (1(1++rr)) u ud d

■ ✓(1-q)/(1+r) is◆today’s price of a dollar in

Similarly,
✓ 1 ◆ d q
= S theDdown-state
= 1 S d =
S q uSif thedS
= you get
DOne (1and
S up,
stock goes + rdon’t
)(u getd) 1=+
uS dS
otherwise (1 + r )(u d)
■ With state prices, pricing derivatives is trivial
q The call’s replicating Au + cdisAd
c = cu portfolio
u d
c = cu A + cd A
So its prices is
q
✓ ◆ ✓ ◆
q 1 q
c= cu + cd
✓ 1 + r◆ ✓1 + r ◆
q 1 q
c=
FIN 206, Binomial Pricing cu +
Prof. Giulio Trigilia cd 39
 State prices account for risk
✓ ◆ ✓ ◆
q 1 q
c= cu + cd
■ The stock price today
1 + r is NOT the
1 +expected
r price
tomorrow (discounted for time)
q It’s the discounted expected risk-adjusted price
tomorrow
E [Mt+1 St+1 ] pMu Su + (1 p)Md Sd
St = =
1+r 1+r
q 1 q
p· p · uS + (1 p) · 1 p · dS
=
1+r
n M = “stochastic discount factor” or “pricing kernel”
q Or marginal rate of substitution, or Radon-Nicodym derivative,
etc.

FIN 206, Binomial Pricing Prof. Giulio Trigilia 40