On Black-Scholes Equation, Black-
Scholes Formula and Binary Option
Price
Chi Gao
12/15/2013
Abstract:
I.
Black-Scholes Equation is derived using two methods: (1) risk-neutral measure; (2)
-
hedge.
II.
The Black-Scholes Formula (the price of European call option is calculated) is calculated
using two methods: (1) risk-neutral pricing formula (expected discounted payoff) (2)
directly solving the Black-Scholes equation with boundary conditions
III.
The two methods in II are proved to be essentially equivalent. The Black-Scholes formula
for European call option is tested to be the solution of Black-Scholes equation.
IV.
The value of digital options and share digitals are calculated. The European call and put
options are be replicated by digital options and share digitals, thus the prices of call and
put options can be derived from the values of digitals. The put-call parity relation is given.
1.
The derivation(s) of Black-Scholes Equation
Black Scholes model has several assumptions:
1.
Constant risk-free interest rate: r
2.
Constant drift and volatility of stock price:
3.
The stock doesn’t pay dividend
4.
No arbitrage
5.
No transaction fee or cost
6.
Possible to borrow and lend any amount (even fractional) of cash at the riskless rate
7.
Possible to buy and sell any amount (even fractional) of stock
A typical way to derive the Black-Scholes equation is to claim that under the measure that no
arbitrage is allowed (risk-neutral measure), the drift of stock price equal to the risk-free interest
rate
. That is (usually under risk-neutral measure, we write Brownian motion as
, here
we remove Q subscript for convenience )
(1)
Then apply Ito’s lemma to the discounted price of derivatives
, we get
[
]
[(
)
]
(2)
Still, under risk-neutral measure we can argue that
is martingale. It should have zero
drift. So
(3)
This is the Black-Scholes equation for the price of any derivatives on the underlying
, under the
Black-Scholes model.
Now I am going to use another method (
-hedge method) to derive the Black-Scholes equation.
This method avoids to directly use the claim (1).
The price of stock
follows
(4)
Apply Ito’s lemma to derivative on :
(
)
(5)
Now construct a portfolio called
-hedge portfolio, which shorts one derivative above and holds
shares of stock at time
. The value of this portfolio at time is
(6)
Before applying Ito’s lemma on (6), there is one thing that needs to be emphasized:
is
considered to be constant, that is
and
, although varies with time . This result
comes from the fact that
is determined by
at time
and thus should not be considered to be a
time-dependent variable when we calculate the change/differential of our portfolio. The
differential of our portfolio at time
is
(
)
(7)
The Brownian motion term has vanished! This is a portfolio with riskless return rate of
(
) . Now we use the assumption of no arbitrage, which requires
(8)
This yields
(
) (
)
(9)
Here comes the conclusion of Black-Scholes equation
(10)
Note that (10) doesn’t use the risk-neutral measure. Because (10) (or (3)) is a deterministic PDE,
it will hold regardless of which measure is used. However, we can see that the use of risk-neutral
measure does greatly simplify the derivation.
2.
The European vanilla call/put option price
The typical way to derive the European (vanilla) call/put option in many textbook is to calculate
the expected discounted payoff of option, in other words to integrate the discounted payoff the
risk-neutral measure. This procedure has nothing to do with the Black-Scholes equation we got in
(3) or (10). Below I will follow this procedure to get the price
of a call option on stock
at time
. The call option will mature at time with striking price .
The price of stock
at time will be
(
)
(11)
The expected discounted payoff of the call option (which is also the price of the call option, from
the assumption of no arbitrage) is
[
]
∫
(12)
Let
. Here
√
(
)
, where
is a normal distribution
with zero mean and variance
. So
∫
∫
(
)
√
(
)
(13.a)
∫
(13.b)
(
)
√
(13.c)
∫
∫
(
)
√
(
)
(13.d)
(
)
√
(13.e)
So the price of call option at time
is
(14)
Equation (14) is also called Black-Scholes formula for vanilla call option, because it can also be
derived from Black-Scholes equation (10) with appreciated boundary conditions:
(15.a, 15.b, 15.c)
By the change of variable transformation:
(
)
(16.a, 16.b, 16.c)
The Black-Scholes equation (10) becomes the diffusion equation with initial condition
(
{ }
)
(17.a, 17.b)
The solution for diffusion equation is
√
∫
[ ] [
]
(18)
After some math, we have
(19)
Changing the variables of
{ } back to { } yields the Black-Scholes formula (14).
3. Does the Black-Scholes formula satisfy Black-Scholes equation?
The first method used to derive Black-Scholes formula (14) doesn’t use the Black-Scholes
equation (10). But it so “happens” to give the solution of Black-Scholes equation (10). This is
the “good” property of call/put options: the expected discounted payoff of option is exactly the
solution of the Black-Scholes equation.
This property
can be
extended to other derivatives with different forms of payoffs. For example,
if you have a call option on the square of a log-normal asset (like stock price),
.
What equation does the price satisfy? The answer is still Black-Scholes equation, as long as the
derivative price is a function of the current time
and stock price
. If we derive the price using
expected discounted payoff, this price will also satisfy the Black-Scholes equation, i.e. the price
from expected discounted payoff is also a solution of Black-Scholes equation.
That is, the solution of Black-Scholes equation for the price any derivative
is:
[
]
(20)
The mathematical reason behind this is,
first of all needs to satisfy the Black-Scholes
equation (10):
We can transform this equation into typical diffusion equation
with change of variables
(
)
The solution of this diffusion equation with initial boundary condition
is
given in (18):
√
∫
[ ] [
]
Changing the variables of
{ } back to { } yields
∫
√
[
]
∫
√
[
]
∫
(21)
This is exactly the expected discounted payoff as defined in (20)!
So the price of any derivative
on
will satisfy Black-Scholes equation, and the solution (Black-Scholes formula) can be
calculated from expected discounted payoff (with much easy math).
Now I am going to show in straightforward method that Black-Scholes formula of the price of
vanilla call option really satisfies Black-Scholes equation. Recall the price of such call option
is
Define
√
(22.a)
Then
(22.b)
Also we have
√
(
)
√
(
)
√
(22.c, 22.d, 22.e)
Calculate
√
√
√
√
√
√
√
(22.f, 22.g, 22.h)
After some math, we have
4. Binary option (also called Digital option)
A binary option pays a fixed amount ($1 for example) in a certain event and zero otherwise.
Consider a digital that pays $1at time
if . The payoff of such a option is
{
(23)
Using risk-neutral pricing formula
[ ]
(24)
here
and
are same as defined in (13.b, 13.e).
It is not difficult to check that (24) satisfies Black-Scholes equation (10).
There are 4 kinds of digitals (if we consider dividend
)
Name
Definition
Value
digit call
Option that pays $1 when
digit put
Option that pays $1 when
share call
Option that pays 1 share when
share put
Option that pays 1 share when
For non-zero dividend,
are slightly different from previous definition (13.c, 13.e)
(
)
√
(
)
√
(25.a, 25.b)
With these four digitals, we can easily recover the price of European call and put options. For
European call option, use the definition of
in (23), the payoff of this call can be written as
(26)
This is equivalent to one share call minus K digital call. The combined price of this call option
will be
(27)
Similarly, a European put option is equivalent to K digital put minus one share put. The price
of the European put option is
(28)
The put-call parity is
(29)
This parity follows from the fact that both the left and the right-hand sides are the prices of
portfolios that have value
at the maturity of the option.
5. Greeks; hedging; hedging
Dostları ilə paylaş: |