Binomial Trees: RECAP

Lecture #: 3

MATH 4FM3: FINANCIAL MARKETS AND DERIVATIVES - Winter 2020

Dr. Anas Abdallah
Department of Mathematics and Statistics
Actuarial and Financial Mathematics

McMaster University
2
Contents

0.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
0.2 One-Period Binomial Tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
0.3 Risk-Neutral Pricing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
0.4 Two-Period Binomial Tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
0.5 The General Binomial Tree Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
0.6 Increasing number of steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

3
Binomial Option Pricing: Review of the basic concepts

0.1 Introduction

• In this lecture, we review our very first class of models for the price of the underlying asset, namely the
binomial tree model.
• The binomial model is the most famous discrete-time financial model for the evolution of the price of
a risky asset.
• The binomial option pricing model assumes that, over a period of time, the price of the underlying asset
can move up or down only by a specified amount.
• Given this assumption, it is possible to determine a no-arbitrage price for the option.
• Surprisingly, this approach, which appears at first glance to be overly simplistic, can be used to price
options.

1
2 CONTENTS

0.2 One-Period Binomial Tree

• In a one-period model, there are two time points: the beginning of the period is time 0 and the end of
the period is time h.
• One should think of time 0 as today while time h is in the future.

• A one-period binomial model is composed of only two assets: a risk-free asset and a risky asset (e.g.,
stock or index) with a known initial value and an unknown future value.
0.2. ONE-PERIOD BINOMIAL TREE 3

One-Period Binomial Tree, cont’d

• Binomial pricing achieves its simplicity by making a very strong assumption about the stock price.
• The evolution of the price of the risky asset S = {S0 , Sh } is such that S0 is a known quantity and Sh
is a random quantity.
• This random variable Sh is taking values in {Shu , Shd } with probability p and 1 − p respectively. By
convention and without loss of generality, we assume that Shu > Shd .
4 CONTENTS

One-Period Binomial Tree, cont’d

• In other words, Sh has the following probability distribution with respect to the real-world probability
measure P:

P (Sh = Shu ) = p and P Sh = Shd = 1 − p.

• The real-world probability measure contains the information of how investors view the market and
consequently p is the likelihood of the event {Sh = Shu }.
• One could think of the scenario {Sh = Shu } as a bull market scenario and {Sh = Shd } as a bear market
scenario.
0.2. ONE-PERIOD BINOMIAL TREE 5

One-Period Binomial Tree, cont’d

• We now consider an option written on the stock.

• The evolution of this asset over the period is denoted by V = {V0 , Vh }, where V0 is today’s value while
Vh is its value at the end of the period.
• As there are only two outcomes in this model, the payoff Vh can take only two values denoted by Vhu
and Vhd :

• Finally, for the call and put options which are very important cases to analyze, instead of V = {V0 , Vh },
we might also use the notation C = {C0 , Ch } and P = {P0 , Ph }, respectively.
6 CONTENTS

Computing the Option Price

Example 0.2.1. A stock that currently trades for 41$, will be worth either 60$ or 30$ in one year.
Now, we want to have the price of a 40-strike 1-year call. Consider two portfolios:
1. Portfolio A: buy one call option. The cost of this is the call premium.
2. Portfolio B: buy 2/3 of a share of the stock and borrow 18.462 at the risk-free rate of 8%.
The two portfolios have the same payoff one year from now.
0.2. ONE-PERIOD BINOMIAL TREE 7

Computing the Option Price, cont’d

• The idea that positions that have the payoff must have the same cost is the law of one price.
• There is an arbitrage opportunity if the law of one price is violated.
8 CONTENTS

The Binomial Solution

How?
In the preceding example, how did we know that buying 2/3 of a share of stock and borrowing 18.462 would
replicate a call option?

• We have two instruments to use in replicating a call option: shares of stock and a position in bonds
(i.e., borrowing or lending).
• To be specific, we wish to find a portfolio consisting of 4 shares of stock and a dollar amount B in
lending, such that the portfolio imitates the option whether the stock rises or falls.
• We will suppose that the stock has a continuous dividend yield of δ, which is reinvested in the stock.
0.2. ONE-PERIOD BINOMIAL TREE 9

The Binomial Solution, cont’d

• Let S0 be the stock price today . We can write the stock price as uS0 when the stock goes up and as
dS0 when the stock goes down:

• If the value of a period is h, the interest factor per period is erh .

• The problem is to solve for 4 and B such that our portfolio of 4 shares and B in lending duplicates
the option payoff.
10 CONTENTS

The Binomial Solution, cont’d

Theorem 0.2.2 (4 and B). The values of 4 and B that duplicate the option payoff are:

Vhu − Vhd
−δh
4 = e ,
S0 (u − d)
d u
−rh uVh − dVh
B = e .
u−d
Theorem 0.2.3 (Valuing the option). The values of the option is expressed as follows:
(r−δ)h (r−δ)h
 
e − d u − e
V0 = e−rh Vhu + Vhd .
u−d u−d
0.3. RISK-NEUTRAL PRICING 11

0.3 Risk-Neutral Pricing

Risk-Neutral Pricing

What About p?
In the preceding option price calculations, we did not make use of the probability that the stock price would
move up or down.

• The strategy of holding 4 shares and B bonds replicates the option payoff whichever way the stock
moves, so the probability of an up or down movement is irrelevant for computing 4S0 + B.
• Although we do not need probabilities to price the options, there is a very important probabilistic
interpretation of

(r−δ)h (r−δ)h
 
e − d u − e
V0 = e−rh Vhu + Vhd .
u−d u−d
12 CONTENTS

Risk-Neutral Pricing, cont’d

• Define

e(r−δ)h − d
q=
u−d
and

V0 = e−rh Vhu q + Vhd (1 − q) .

(0.3.1)
• Consequently, we can say that q is the risk-neutral probability of an up-move and that 1 − q is the
risk-neutral probability of a down-move.
• Equation (0.3.1) can be considered as the expected value of the discounted payoff if we use the new
probability weights q and 1 − q:

V0 = EQ e−rh Vh ,
 

where the superscript Q emphasizes the use of the risk-neutral probability weights q and 1 − q, instead
of the real-world weights p and 1 − p.
0.3. RISK-NEUTRAL PRICING 13

Risk-Neutral Pricing, cont’d

Real-World v. Risk-Neutral

• Real-World: Using probabilities p and 1 − p, we can compute the P expectation:

EP [Sh ] = pShu + (1 − p)Shd .

This is the real-world expected value of S1 , i.e., what is expected by financial analysts (agreeing on this
market model).
• Risk-Neutral: If we instead use the probabilities q and 1 − q, we get:

EQ [Sh ] = qShu + (1 − q)Shd .

• The two expectations represent two different quantities.
14 CONTENTS

0.4 Two-Period Binomial Tree

• We now extend the binomial tree to more than one period. Instead of one period of one year, let us
now consider two periods of one year.

Figure 0.4.1: Two-Period Binomial Tree

0.4. TWO-PERIOD BINOMIAL TREE 15

Option Pricing

The key insight to price option in two-period binomial tree is that we work backward through the tree.

• At the outset, the only period where we know the option price is at expiration (Period 2).
• Knowing the price at expiration, we can determine the price in Period 1.
• And having determined that price, we can work back to Period 0.

Example 0.4.1. Consider a European call option on a stock, with a $40 strike and 1 year to expiration. The
stock does not pay dividends, and its current price is $41. Suppose the volatility of the stock is 30%. The
continuously compounded risk-free interest rate is 8%. What is the price of this option using a two-period
binomial tree?
16 CONTENTS

Option Pricing, cont’d

0.4. TWO-PERIOD BINOMIAL TREE 17

Option Pricing, cont’d

18 CONTENTS

Option Pricing, cont’d

0.4. TWO-PERIOD BINOMIAL TREE 19

Option Pricing, cont’d

Once we know the prices of Period 1, we are in the same situation as the oneperiod tree...
20 CONTENTS

Option Pricing, cont’d

Some Considerations

• The option price is greater for the 2-year than for the 1-year option.

• We priced the option by working backward through the tree, starting at the end and working back to
the first period.

• The option’s 4 and B are different at different nodes. However, to get the price of the option, we do
not need to compute these.
0.5. THE GENERAL BINOMIAL TREE MODEL 21

0.5 The General Binomial Tree Model

• An obvious objection to the binomial calculations thus far is that the stock can only have two or three
different values at expiration.
• It seems plausible that to increase accuracy we would want to divide the time to expiration into more
periods, generating a more realistic tree.
• Fortunately the generalization to many binomial periods is straightforward.
22 CONTENTS

0.6 Increasing number of steps

Is it that simple in ”real life”?

The binomial model presented above is unrealistically simple. Clearly, a financial analyst can expect to
obtain only a very rough approximation to an option price by assuming that stock price movements during
the life of the option consist of one or two binomial steps.

• When binomial trees are used in practice, the life of the option is typically divided into 30 or more time
steps
• In each time step there is a binomial stock price movement. With 30 time steps there are 31 terminal
stock prices and 230 , or about 1 billion, possible stock price paths are implicitly considered.
• As the number of time steps is increased (so that ∆t becomes smaller), the binomial tree model makes
the same assumptions about stock price behavior as the Black– Scholes–Merton model, which will be
presented in later in this course.
• When the binomial tree is used to price a European option, the price converges to the Black–Scholes–Merton
price, as expected, as the number of time steps is increased
• The idea for the upcoming chapters is taking a tree with finer and finer ∆t’s to try to imagine a stock
price (or log of a stock price) following some random path where, at any moment of time t, it has a
normal distribution of possible values.