Binomial Tree

Lecture 13.
Binomial Trees
Introduction to Binomial Trees
Options, Futures, Derivatives 10/15/07 back to start 1

Binomial Trees
Binomial trees are useful tools for pricing options. Construct charts of possible movements of a
stock and price according to the movements.
We assume that the stock price is a random walk, i.e. at each time step there is a certain
probability of the stock moving in one direction or another (up or down).
Work first from example (artificial):
• Stock is priced currently at $20 and at the end of three months it will be either $22 or $18.
Value the European call option to buy the stock for $21 in 3 months.
• If the stock is $22 then the option will be worth $1, and if the stock is worth $18 then the
option is worth zero.
• Use arbitrage argument to value this situation.
– Long position in ∆ shares and a short position in the call option
– If the stock price moves from $20 to $22 then value of shares is $22∆ and the value of the
option is $1. Total value is 22∆ − 1.
– If the stock price moves down from $20 to $18 then the value of the shares is $18∆ and
the value of the option is zero. Total value is 18∆.
– Portfolio is risk-less if ∆ chosen so both situations agree. Then
22∆ − 1 = 18∆ =⇒ ∆ = 0.25

Example, cont.
– Risk-less portfolio is therefore long 0.25 shares and short 1 option.

– In either case the value is 22 × 0.25 − 1 = 18 × 0.25 = 4.5.
• Risk-less portfolios must earn the risk-free interest rate r . If it is 12% per year, cont. comp.
then the value of the option is
3
−0.12× 12
4.5e = 4.367
at the start time.
• The stock price is $20 today and the option price is f , then the value of the portfolio is
20 × 0.25 − f = 5 − f
• But
5 − f = 4.367 =⇒ f = 0.633.

Generalization
Consider now a stock with price S0and an option with current price f .
• Suppose that the option lasts for time T and during the life of the option, the price can go
either up from S0 to S0u (where u > 1) or down from S0 to S0d (where d < 1).
• The percentage increase in the stock price in up movement is u − 1. The percentage decrease
in the stock price in down movement is 1 − d.
• Let the option payoff for up movement be fu and the option payoff for down movement be fd.

Generalization
• Build a portfolio of a long position in ∆ shares and a short position in one option.
• On up movement the value of the portfolio is at the end of the option
S0u∆ − fu
• On down movement the value of the portfolio is at the end of the option
S0d∆ − fd
• Equality if
S0u∆ − fu = S0d∆ − fd (1)
or
fu − f d
∆=
S0 u − S0 d
Thus ∆ is the ratio of the change in the options prices to the change in the stock prices.
• Portfolio is now risk-less, and so earns risk-free rate r . Present value of the portfolio (due to
(1)) is
−rT
(S0u∆ − fu) e

Generalization
• Cost of setting up the portfolio is

S0 ∆ − f
• Since cost should equal present value (else arbitrage opportunity)
−rT
S0∆ − f = (S0u∆ − fu) e
or “ ”
−rT −rT −rT
f = S0∆ + (S0u∆ − fu) e = S0∆ 1 − ue + fu e
fu −fd
• Recall ∆ = S0 u−S0 d then
“ ” f −f
−rT u d −rT
f = S0 1 − ue + fu e
S0 u − S0 d
` rT ´ !
−rT e − u (fu − fd) fu (u − d)
=e +
u−d u−d
` rT ´ !
−rT e − u (fu − fd) fu − fd + fd (u − d)
=e +
u−d u−d
! !!!
rT rT
−rT e −d e −d
=e fu + fd 1 −
u−d u−d

Generalization
• Or
−rT
f =e [pfu + (1 − p)fd]
where !
rT
e −d
p=
u−d
Example: Recall from our example: u = 1.1, d = 0.9, r = 0.12, T = 0.25, fu = 1, fd = 0.

Then
3
e0.12× 12 − 0.9
p= = 0.6523
1.1 − 0.9
and so
−0.12×0.25
f =e (0.6523 × 1 + 0.3477 × 0) = 0.633
Remark: Pricing independent of the probability of the stock moving up or down! Calculating the
price in terms of the underlying stock (incorporated in the value of the stock price already).

Example
Example: A stock price is currently $40. It is known that at the end of 1 month it will be either
$42 or $38. The risk-free interest rate is 8% per annum with continuous compounding. What is
the value of a 1-month European call option with strike price of $39?
42 S u S d
Solution: Here we set S0 = 40, u = S0 = 40 = 1.05, and d = S0 = 38 40 = 0.95. We also
0 0
have r = 0.08 and K = 39. The payoff in the $42 case is fu = 42 − 39 = $3, whereas the
payoff in the second case is fd = 0.
Next we use ! 1
rT
e −d e0.08× 12 − 0.95
p= = = 0.5669
u−d 1.05 − 0.95
and
−rT 1
−0.08× 12
f =e [pfu + (1 − p)fd] = e [0.5669 × 3 + (1 − 0.5669) × 0] = 1.689
Therefore, the price should be $1.69.

Risk-Neutral Valuation
−d rT
• Natural to interpret p = e u−d as the probability of the upward movement. Then 1 − p is the
probability of downward movement.
• Then
pfu + (1 − p)fd
is the expectation (expected payoff = probability × payout).
• In a similar vein, if we consider the same probability of up and down movement, then the
expectation of stock value should be
!
rT rT
e −d e −d
pS0u + (1 − p)S0d = S0 u + 1 − S0 d
u−d u−d
erT − d u − erT
= S0 u + S0 d
u−d u−d
" #
rT rT
ue − ud + ud − de
= S0
u−d
rT
= S0 e

Risk-Neutral Valuation
• Therefore, a stock should grow at the risk-free rate on average. This is an example of a
risk-neutral valuation.
• A risk-neutral world is a market in which all individuals are indifferent to risk. Investors require
no compensation for risk. All stocks gain at the risk-free rate.
• Risk-neutral valuation says we can assume the world is risk-neutral when pricing options.

Re-evaluation one-step binomial trees
Evaluate via risk-neutral valuation. Consider
• Stock price S0 = 20, S0u = 22, S0d = 18 after three months. European option with strike
price 21 and expiration in 3 months. Risk-free rate r = 12%.
• Expected return in a risk-neutral world must be the risk-free rate of 12%. Thus
3
0.12× 12
22p + 18(1 − p) = 20e
or 3
0.12× 12
4p = 20e − 18.
Thus p = 0.6523.
Re-evaluation one-step binomial trees
• After three months the option has 0.6523 chance of being worth 1 and 0.3477 chance of being
worth 0. Expected value is
0.6523 × 1 + 0.3477 × 0 = 0.6523
• In risk-neutral world, this should be discounted at the risk-free rate. The value of the option
should therefore, be worth
3
−0.12× 12
0.6523e 0.633
which we derived earlier.

Real World vs. Risk-Neutral World
• The probability p of up movement is from a risk-neutral world. Then

rT
E(ST ) = S0e
• Suppose on the contrary the expected return is 16% and p∗ is the probability of up movement
in the real world. Then from our first example with S0 = 20, S0u = 22, and S0d = 18, then
3
0.16× 12 ˆ ∗ ∗ ˜
20e = 22p + 18(1 − p )
then 3
∗ 0.16× 12
4p = 20e − 18
or
∗
p = 0.7041
• We interpret this as the expected payoff from the option in the real world is
∗ ∗ ∗
p × fu + (1 − p ) × fd = p = 0.7041
since fu = 1 and fd = 0.
• However, it is riskier to own the option than the stock, so the discount rate should be higher
than 16% to compensate for this. How to quantify??
• Easier to use the risk-neutral world assumptions to calculate the prices.
Two-step Binomial Trees
We can extend the analysis of one-step trees to two-step trees. Consider a stock with a price at
$20. In each time step of 3 months, the price can go up 10% or down 10%. Suppose the risk-free
rate is 12% per annum. Assume the option price is $21.
We would like to compute the price of a call option given the above data.
Method? Work backwards!

Two-step Binomial Trees, cont.
• At node (D) with value $24.20 then option value fuu = 24.20 − 21 = $3.20.
• At node (E) with value $19.80 then option value fud = fdu = 0.
• At node (F) with value $16.20 then option value fuu = 0.

• At node (B) with value $22 then option value

−rT
fu = e [pfuu + (1 − p)fud]
erT −d 3
with p = u−d and u = 1.1, d = 0.9, r = 0.12 and T = 12 . Then
3
e0.12× 12 − 0.9
p= = 0.6523
1.1 − 0.9
then 3
−0.12× 12
fu = e [0.6523 × 3.20 + 0.3477 × 0] = 2.0256
• At node (C) with value $18 then option value
−rT
fd = e [pfud + (1 − p)fdd]
erT −d 3
3
e0.12× 12 − 0.9
p= = 0.6523
1.1 − 0.9
then 3
−0.12× 12
fd = e [0.6523 × 0 + 0.3477 × 0] = 0
• At node (A) with value $20 (initial point) then option value
−rT
f =e [pfu + (1 − p)fd]
erT −d 3
3
0.12× 12
e − 0.9
p= = 0.6523
1.1 − 0.9
then 3
−0.12× 12
f =e [0.6523 × 2.0256 + 0.3477 × 0] = 1.2822
Therefore, we price the option at $1.28.

Generalization
Price option for generalization. Set time scale to be ∆t.
• Price is
−r∆t
f =e [pfu + (1 − p)fd]
where
erT − d
p=
u−d
Generalization
• Again price is
−r∆t
f =e [pfu + (1 − p)fd]
• For the node at S0u price is
−r∆t
fu = e [pfuu + (1 − p)fud]
• For the node at S0d price is

−r∆t
fd = e [pfdu + (1 − p)fdd]
• Then substituting we get

−r∆t
f =e [pfu + (1 − p)fd]
h “ ” “ ”i
−r∆t −r∆t −r∆t
=e p e [pfuu + (1 − p)fud] + (1 − p) e [pfdu + (1 − p)fdd]
h i
−2r∆t 2 2
=e p fuu + 2p(1 − p)fud + (1 − p) fdd
• The probabilities p2, 2p(1 − p), (1 − p)2 are the probabilities of achieve final stock prices of
S0u2, S0ud, S0d2, respectively.
• Risk-neutral evaluation.
Put pricing
Consider a 2-year European put with strike price of $52 on a stock whose current price is $50.
Suppose two time steps of 1 year, and the stock price moves up either 20% or down 20%. Assume
the risk free rate is 5%.
So u = 1.2, d = 0.8, r = 0.05, ∆t = 1.
• First compute p via

erT − d e0.05×1 − 0.8
p= = = 0.6282.
u−d 1.2 − 0.8
• Note S0u = 50 × 1.2 = 60, S0d = 50 × 0.8 = 40.
• And S0u2 = 72, S0ud = 48, S0d2 = 32.
• Next fuu = 0, fud = fdu = 4, fdd = 20. We get a price for f as
h i
−2r∆t 2 2
f =e p fuu + 2p(1 − p)fud + (1 − p) fdd
h i
−2×0.05×1 2 2
=e 0.6282 × 0 + 2 × 0.6282 × (1 − 0.6282) × 4 + (1 − 0.6282) × 20
= 4.1923

Put pricing

Binomial Trees to Price American Options
So far we’ve evaluated prices for European options. We follow the following procedure to price
1. Final nodes are priced as for European options

2. Previous nodes we price via taking the greater of
(a) Value given by
−r∆t
f =e [pfu + (1 − p)fd] (2)
(b) Value given by early exercise
Redo the previous example: two-step binomial tree for American put. We have So u = 1.2,
rT −d
d = 0.8, r = 0.05, ∆t = 1. Then p = e u−d = 0.6282 and 1 − p = 0.3718.
• At S0u we have fuu = 0 and fud = 4. Then we first check fu via (2) then
−0.05×1
e [0.6282 × 0 + 0.3718 × 4] = 1.4147.
The value of early exercise is 52 − 60 = −8 < 0. Then we take
fu = max{1.4147, −8} = 1.4147

• At S0d we have fud = 4 and fdd = 20. Then we first check fd via (2) then
−0.05×1
e [0.6282 × 4 + 0.3718 × 20] = 9.4636.
The value of early exercise is 52 − 40 = 12 > 0. Then we take
fd = max{9.4636, 12} = 12
• At S0, starting time, we have fu = 1.4147 and fd = 12. Then first check f via (2) then
−0.05×1
e [0.6282 × 1.4147 + 0.3718 × 12] = 5.0894.
The value of early exercise is 52 − 50 = 2. Then we take
f = max{5.0894, 2} = 5.0894
• The value of the American put option should be $5.09.


Example 2
1
By hand. Let u = 1.05, d = 0.95, ∆t = 12 , r = 12%.
Price an American put option with three-step binomial tree, assuming 5%

volatility at each month.

Delta
Define Delta, ∆, to be the ratio of the change in the price of the stock option to the change in the
price of the underlying stock.
Number of units of stock that we should hold for each option shorted in order to create a risk-less
hedge.
• A risk-less hedge is referred to as delta hedging.
• Tells investor how to adjust the amount of stock in portfolio so that the portfolio remains
risk-less.
• Example: First example in the one-step binomial tree, fu − fd = 1 − 0 = 1 and

S0u − S0d = S0(u − d) = 20(1.1 − 0.9) = 20 × 0.2 = 4. Then ∆ = 14 = 0.25.

Delta
• Example 2: Consider the second example
– Delta hedge on first step is 2.0257−0

22−18 = 0.5064
– Delta for stock price movements over second time step are
3.2 − 0 0−0
= 0.7273 ↑ =0↓
24.2 − 19.8 19.8 − 16.2
for up and down movements, respectively.

Delta
• Example 3: Consider the second example
– Delta hedge on first step is 1.4147−9.4636

60−40 = −0.4024
– Delta for stock price movements over second time step are
0−4 4 − 20
= −0.1667 ↑ = −1.000 ↓
72 − 48 48 − 32
for up and down movements, respectively.

Delta
• Therefore, Delta changes over time.
• In order to maintain a risk-less hedge using an option and underlying stock, the stock holdings
must be adjusted periodically.

Matching Volatility with u and d
When constructing binomial trees, choose u and d to match the volatility of the stock price.
We present the following method:
• Let µ be the expected return on the stock (in the real world).
√
• Let σ be the volatility. The volatility σ of a stock price is defined so that σ ∆t is the
standard deviation of the return on the stock price in a short period of time of length ∆t.
More later.
• Let ∆t be the interval of time.
• Suppose the stock price starts at S0 and stock moves to either S0u or S0d at the end of ∆t.

• Probability of up movement is p∗. Then the expected return at the first time step is
µ∆t
S0 e
and the expected stock price at this time is

∗ ∗
p S0u + (1 − p )S0d
which should match the return. Then

∗ ∗ µ∆t
p S0u + (1 − p )S0d = S0e
• Therefore,
eµ∆t − d ∗
p =
u−d
√
• Now we use the volatility into the picture. Since the volatility σ ∆t is the standard deviation
of the return on the stock price over short periods of time ∆t, then σ 2∆t is the variance on
the return on the stock price over short periods of time.

• But the variance on the return is E(X 2) − (E(X))2 for X = u, d then

2 ∗ 2 ∗ 2 ´2
` ∗ ∗
σ t = p u + (1 − p )d − p u + (1 − p )d
Using
∗ eµ∆t − d
p =
u−d
then
2 ∗ 2 ∗ 2 `´2 ∗ ∗
σ t = p u + (1 − p )d − p u + (1 − p )d
h“ ” “ ” i
−2 µ∆t 2 µ∆t 2
= (u − d) e − d (u − d) u + u − e (u − d) d
h“ ” “ ” i2
−2 µ∆t µ∆t
− (u − d) e −d u+ u−e d
µ∆t 2µ∆t
=e (u + d) − ud − e
Can solve for u and d and get

√
σ ∆t
u=e
√
−σ ∆t
d=e

When we move from the real world to the risk-neutral world then
• The expected return on the stock changes

• The volatility remains the same (as ∆t goes to zero)
• Known as Giranov’s Theorem. More later.
Moving from one set of risk preferences to another is referred to as changing the measure.
• The real-world measure is the P-measure - p∗.

• The risk-neutral measure is the Q-measure - p.

Example: Consider the american put option where the stock price is $50, the strike price is $52,
the risk-free rate is 5%, the life of the option is 2 years, two time steps. Suppose the volatility is
30%.
0.3×1 1 rT 0.05×1
u=e = 1.3499 d= = 0.7408 e =e = 1.0513
1.3499
1.053−0.7408
Then p = 1.3499−0.7408 = 0.5097.
Value of the put is 7.43.

Compare with
Different from the value with u = 1.2 and d = 0.8 of 4.1923.

Increasing the number of steps
Too little information to get an accurate price of the option. In practice, the life of option divided
into many more time steps (as many as 30 steps for 230 ≈ 109 possible pathways).
Still many of the variables are the same including

√ √
σ ∆t −σ ∆t
u=e d=e
and the risk-free rate r and ∆t over the life of the binomial tree.

Homework
Due Oct. 24, 5PM.
• 10.7, 10.13, 11.7
• Graded: 10.21, 11.16, 11.17

Binomial Tree

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Binomial Tree

Uploaded by

Copyright:

Available Formats

Lecture 13.

Introduction to Binomial Trees

Options, Futures, Derivatives 10/15/07 back to start 1

Work first from example (artificial):

22∆ − 1 = 18∆ =⇒ ∆ = 0.25

Options, Futures, Derivatives 10/15/07 back to start 2

– Risk-less portfolio is therefore long 0.25 shares and short 1 option.

Options, Futures, Derivatives 10/15/07 back to start 3

Options, Futures, Derivatives 10/15/07 back to start 4

Options, Futures, Derivatives 10/15/07 back to start 5

• Cost of setting up the portfolio is

Options, Futures, Derivatives 10/15/07 back to start 6

Example: Recall from our example: u = 1.1, d = 0.9, r = 0.12, T = 0.25, fu = 1, fd = 0.

Options, Futures, Derivatives 10/15/07 back to start 7

Therefore, the price should be $1.69.

Options, Futures, Derivatives 10/15/07 back to start 8

Options, Futures, Derivatives 10/15/07 back to start 9

Options, Futures, Derivatives 10/15/07 back to start 10

Evaluate via risk-neutral valuation. Consider

0.6523 × 1 + 0.3477 × 0 = 0.6523

Options, Futures, Derivatives 10/15/07 back to start 12

• The probability p of up movement is from a risk-neutral world. Then

Method? Work backwards!

Options, Futures, Derivatives 10/15/07 back to start 14

Options, Futures, Derivatives 10/15/07 back to start 15

• At node (B) with value $22 then option value

Therefore, we price the option at $1.28.

Price option for generalization. Set time scale to be ∆t.

• For the node at S0d price is

• Then substituting we get

So u = 1.2, d = 0.8, r = 0.05, ∆t = 1.

• First compute p via

Options, Futures, Derivatives 10/15/07 back to start 20

Options, Futures, Derivatives 10/15/07 back to start 21

1. Final nodes are priced as for European options

The value of early exercise is 52 − 60 = −8 < 0. Then we take

fu = max{1.4147, −8} = 1.4147

Options, Futures, Derivatives 10/15/07 back to start 22

The value of early exercise is 52 − 40 = 12 > 0. Then we take

The value of early exercise is 52 − 50 = 2. Then we take

• The value of the American put option should be $5.09.

Options, Futures, Derivatives 10/15/07 back to start 23

Options, Futures, Derivatives 10/15/07 back to start 24

Price an American put option with three-step binomial tree, assuming 5%

Options, Futures, Derivatives 10/15/07 back to start 25

• A risk-less hedge is referred to as delta hedging.

• Example: First example in the one-step binomial tree, fu − fd = 1 − 0 = 1 and

Options, Futures, Derivatives 10/15/07 back to start 26

• Example 2: Consider the second example

– Delta hedge on first step is 2.0257−0

for up and down movements, respectively.

Options, Futures, Derivatives 10/15/07 back to start 27

• Example 3: Consider the second example

– Delta hedge on first step is 1.4147−9.4636

for up and down movements, respectively.

Options, Futures, Derivatives 10/15/07 back to start 28

• Therefore, Delta changes over time.

Options, Futures, Derivatives 10/15/07 back to start 29

We present the following method:

• Let ∆t be the interval of time.

Options, Futures, Derivatives 10/15/07 back to start 30

and the expected stock price at this time is

which should match the return. Then