Professional Documents
Culture Documents
The Black Scholes Model-2
The Black Scholes Model-2
Scholes Equation
By: A V Vedpuriswar
2
Discrete and Continuous time Stochastic
processes
3
Markov Process
6
Illustration
7
Generalized Wiener Process
8
Generalized Wiener Process (Continued)
a1 1T a2 2T , T T
2
1
2
2
=
a1 a2 ( 1 2 )T , ( 12T 22T )
S0 X1 + X2 follows a generalized wiener process with drift rate
1+ 2 and variance rate , 12 + 22 11
(b) In this case the change in the value of X1 + X2 in a short
interval of time t has the probability distribution:
( 1 2 )t , 12 22 2 1 2 )t
If 1, 2, 1, 2 and are all constant, the change in a longer
period of time T is
( 1 2 )T , 12 22 2 1 2 )T
The variable, X1 + X2, therefore follows a generalized Wiener
process with drift rate 1+ 2 and variance rate 12 + 22 +
2 1 2
12
dz.
For the first four years, = 2 and = 4; for the next four
years = 3 and =5.
If the initial value of the variable is 5, what is the
probability distribution of the value of the variable at the
end of year 8?
14
Brownian Motion
In the coin tossing experiment, the expected winnings
after any number of tosses is just the amount we
already hold.
This is called the Martingale property.
The quadratic variation of a random walk is defined by
[(S1 S0)2 + (S2 S1)2 + + (Si Si-1)2]
For each toss, the outcome is + $1 or - $1. So each
of the terms in the bracket will be (1)2 or (-1)2 i.e.,
exactly equal to 1.
Since there are i terms within the square bracket, the
quadratic variation is nothing but i.
15
Brownian Motion ( Contd)
17
Geometric Brownian Motion
The most widely used model of stock price behaviour is
given by the equation:
ds/s = dt + dz
is the volatility of the stock price
is the expected return.
This model is called Geometric Brownian motion.
The first term on the right is the expected return and the
second is the stochastic component.
The return on a stock price between now and a short period
of time, t later is normally distributed with mean t and
std devn, t.
Over a long time, the return will be normally distributed
with mean, ( - 2/2) ( T) and standard deviation, T
18
Illustration
Suppose a stock has a volatility of 20% per annum
and provides an expected return of 15% per annum
with continuous compounding. If the time interval = 1
week = .0192 years and the initial stock price is 50,
what will be the process for a short period of time?
The process for the stock price can be written as:
ds/s = .15dt + .20dz
or s/s = .15 t + .20 z
or s/s = .15 t + .20 t
. s = 50 (.15 x .0192 + .20 .0192)
= .144 + 1.3856
19
Problem
Stock A and stock B both follow geometric Brownian
motion. Changes in any short interval of time are
uncorrelated with each other. Does the value of a portfolio
consisting of one of stock A and one of stock B follow
geometric Brownian motion?
Let SA, SB, A , B and A, B be stock price, expected return
and volatility for stocks A, B respectively..
Then SA = A SA t + ASAAt and SB = B SB t +
BSBBt
23
The long run
24
Geometric Brownian Motion
We need to first understand that volatility tends to
depress the returns below what the short term returns
suggest.
Expected returns reduce because volatility jumps do not
cancel themselves.
A 5% jump multiplies the current stock price by 1.05; A
5% fall multiplies the amount by .95.
If a 5% jump is followed by a 5% fall or vice versa, the
stock price will reach 0.9975, not 1!
In general, if a positive return x is followed by a negative
return x, the price will reach (1+x) (1-x) = 1- x2
How do we estimate the value of x?
Consider a random variable x. We can calculate the
variance of x as follows: 25
2 = E [x2] {E[x]}2 = E [x2] (assuming E[x] = 0, ie.,
ups and downs in x cancel out)
Thus the expected value of x2 is the variance.
Amount by which the returns are depressed when a
positive movement of x is followed by an equal negative
movement is x2.
For two moves, the depression is x2.
So we could say that the average depression per move is
x2/2.
But the expected value of x2 is 2 .
So we can write 2 /2 as the expected value of the
amount by which the returns fall from the mean.
That is why we write (-2/2) and not .
26
A simple example to explain (-2/2)
27
Geometric Brownian Motion
Can we make some prediction about the kind of distribution followed by
the stock price under the assumption of a Geometric Brownian Motion?
Let us begin with the assumption that the stock returns are normally
distributed. 1 S T
ln
T t0 S t0
Annualised return from t0 to T =
ST = future price St0 = current price, T-t10 is expressed
1 in years.
lnST lnSt0
T t0 T t0
Annualised return = 1 1
lnST lnS t0
T t0 T t0
Let random variable X =
Let us define a 1new random variable
lnS t0
X+ T t0
Also X+ =
or (T-t0) X + ln St0 = ln ST
28
Geometric Brownian Motion
The mean return on S from time t to time T is (T-t) (r-2/2), while the std
devn is T-t
The return on S from time t to T = ln ST /St
The random variable is normally distributed with mean = 0 and std
devn = 1.
Suppose a call option on the stock with strike price, K is in the money at
expiration.
We want to estimate the probability of the stock price exceeding the
strike price.
ST K
ST/St K/St
/S ) ln
2
S 2
lln
n T (S ) (K/St)
K
(T
T tt )( r ln (T t )(r )
St 2 St 2
T t T t
29
Geometric Brownian Motion
St 2
ln (T t )(r ) 2
ST 2 St
ln (T t )(r (Taking
) the negative
K 2
T t T t
K S ST S
of both sides and noting thatln ST
ln t
K
ln
St
ln t
ST
The probability of the stock price exceeding the strike price can
be written as:
P (ST K) = N [ (l nS t / K (T t )( r 2
/ 2)
]
( T t )
30
Geometric Brownian Motion : Concluding notes
This expression reminds us of the Black Scholes formula!
Indeed, GBM is central to Black Scholes pricing.
GBM assumes stock returns are normally distributed.
But empirical data reveals that large movements in stock price
are more likely than a normally distributed stock price model
suggests.
The likelihood of returns near the mean and of large returns is
greater than that predicted by GBM while other returns tend to
be less likely.
There is also evidence that monthly and quarterly volatilities are
higher than annual volatility.
Daily volatilities are lower than annual volatilities.
So stock returns do not scale as they are supposed to.
31
Problem
A stock price follows geometric Brownian motion with
an expected return of 16% and a volatility of 35%.
The current price is $38.
(a) What is the probability that a European call option
on the stock with an exercise price of $40 and a
maturity date in 6 months will be exercised?
(b) What is the probability that a European put option
on the stock with the same exercise price and
maturity will be exercised?
33
Problem
An exchange rate is currently 0.8000. The annualised volatility of the
exchange rate is quoted as 12% and interest rates in the two countries
are the same. Using the log normal assumption, estimate the
probability that the exchange rate in 3 months will be
(f) greater than 0.9000. Based on the volatility smile usually observed
in the market for exchange rates, which of these estimates would you
expect to be too low and which would you expect to be too high?
Ref : John C Hull, Options, Futures and Other Derivatives 34
Solution
An exchange rate behaves like a stock that provides a dividend
yield equal to the foreign risk-free rate. Whereas the growth
rate in a non-dividend-paying stock in a risk-neutral world is r,
the growth rate in the exchange rate in a risk-neutral world is r
rf.
In this case the foreign risk-free rate equals the domestic risk-
free rate (r=rf).
The expected growth rate in the exchange rate is therefore
zero. If ST is the exchange rate at time T its probability
distribution is given by :
lnST~(lnS0 2T/2, T)
0.3567 0.2249
N
= N (-2.1955)
0.06
This is 1.41%
b) In 0.75 = -0.2877. The probability that ST < 0.75 is the same
as the probability that lnST < -0.2877. It is
0.2877= 0N.2249
(-1.0456)
N
0.06
This is 14.79%.
The probability that the exchange rate is between 0.70 and 0.75
is therefore 14.79 1.41 = 13.38%.
36
Cont..
(c) In 0.80 = -0.2231. The probability that ST < 0.80 is the same
as the probability that lnST < -0.2231. It is
0.2231
= N (0.0300)
0.2249
N
0.06
This is 51.2%. The probability that exchange rate is between .75
and .80 is 51.20 14.79 = 36.41%.
(d) In 0.85 = -0.1625. The probability that ST < 0.85 is the same
as the probability that lnST < -0.1625. It is
0.1625
= N0.(1.0404)
2249
N
0.06
This is 85.09%.
The probability that the exchange rate is between 0.80 and 0.85
is therefore 85.09 51.20 = 33.89%
37
Cont..
(e) In 0.90 = -0.1054. The probability that ST < 0.90 is the same
as the probability that lnST < -0.1054. It is
0.1054
= N0.(1.9931)
2249
N
0 .06
(f) The probability that the exchange rate is greater than 0.90 is
100 97.69 = 2.31%.
38
Itos lemma
Let us move closer to the Black Scholes formula.
Black and Scholes formulated a partial differential equation
which they later solved, with the help of Merton by setting up
boundary conditions.
To understand the basis for their differential equation, we need to
appreciate Itos lemma.
Consider G, a function of x.
The change in G for a small change is x can be written as:
G
G = x
dx
We can understand this intuitively by stating that the change in
G is nothing but the rate of change with respect to x multiplied
by the change in x.
If we want a more precise estimate, we can use the Taylor series:
G = dG + 1 d 2G 1 d 3G
x 2 dx 2
(x) 2
3
( x ) 3
.....
dx 6 dx
Ref : John C Hull, Options, Futures and Other Derivatives
39
Itos lemma
G = x t 2 x 2
But (x)2 = b22 t as we just saw a little earlier.
It can be shown (beyond the scope of this coverage) that the expected
value of 2 t is t, as t becomes very small.
Thus (x)2 = b2t
41
Itos lemma and GBM
But dx = a(x,t) dt + b(x,t) dz
So we can rewrite:
dG =
G G 1 2G 2
(adt bdz ) dt b dt
x t 2 x 2
=
G G 1 2G 2 G
(a b ) dt b dz
x t 2 x
This is called Itos lemma.
2
x
42
Problem
Suppose that a stock price S follows geometric Brownian
motion with expected return and volatility : dS = Sdt
+ Sdz. What is the process followed by the variable S n?
If G(S,t) = Sn then G/S = nSn-1, and 2G/S2 = n(n-1)Sn-2.
Using Itos lemma: dG = [nG+ n(n-1) 2G]dt + nGdz
This shows that G = Sn follows geometric Brownian motion
where the expected return is n+ n(n-1) 2 and the volatility
is n.
The stock price S has an expected return of and the expected
value of ST is S0eT. 1 2
[ n n ( n1) ]T
n
The expected value of SnT is S e
0
2
44
The Black Scholes differential equation
Our aim is to create a risk free portfolio whose value does not depend on the
S, the stochastic variable.
Suppose we create a portfolio with a long position
f of shares and a short
position of one call option. s
45
The Black Scholes differential equation
But s = St+Sz
f f 1 2 f 2 2 f f
or = - st t s t sz ( st sz )
s t 2 s 2 s s
f f 1 2 f 2 2 f f f
= - St t s t Sz st sz
s t 2 s 2
s s s
or = - f
t
1 2 f 2 2
s t
t 2 s 2
= - f 1 2 f 2 2
s t
t 2 s 2
47
The Black Scholes formula
Let C be the value of the call, P that of the put, K the
strike price
d1 = [ln(S0/k) + (r+2/2)T] / T
d2 = [ln(S0/k) + (r-2/2)T] / T = d1 - T
As per put call parity,
C P = S0 Ke-rT
or P = C S0 + Ke-rT
Problem
Use Black Scholes to value the following call option:
Stock price =$200, Strike price=$210,
Time to expiration =156 days, Risk-free interest rate
= 11%
Variance of monthly stock returns = 0.02
S = 200, K = 210, T = 156/365 = 0.4274 year
r = 0.11, = [12x0.02] = 0.489898 / year
ln(S / K ) (r 2 / 2)T ln(200 / 210) (0.11 0.12)0.4274
d1
T 0.320275
= 0.1546
51
N(d1) = 0.5614
N(d2) = 0.4342
Ke-rT = 210e-(0.11)(0.4274) = (210)e-0.0470 =(210)(0.9541)
= 200.3556
C = SN(d1) - Ke-rTN(d2)
= (200) (0.5614) (200.3556) (0.4342)
= 25.2902
The equivalent portfolio consists of long 0.5614
shares of stock and borrowing $200.3556. 52
Consider an American call option on a stock. The stock price is $50, the
time to maturity is 15 months, the risk-free rate of interest is 8% per
annum, the exercise price is $55, and the volatility is 25%. Dividends of
$1.50 are expected in 4 months and 10 months. Calculate the price of
the option.
The present value of the dividends is
1.5e-0.3333x0.08 + 1.5e-0.8333x0.08 = 2.864
The option can be valued using the European pricing formula with:
S0 = 50 2.864 = 47.136, K = 55, = 0.25, r = 0.08, T = 1.25
John C Hull, Options, Futures and Other Derivatives
53
Problem
A company can buy an option for the delivery of 1 million
barrels of oil in 3 years at $25 per barrel. The 3-year
futures price of oil is $24 per barrel. The risk-free interest
rate is 5% per annum with continuous compounding and
the volatility of the futures price is 20% per annum. How
much is the option worth?.
The option can be valued using Blacks model. We use
futures/forward price, F0 instead of spot price, S0 of underlying.
The
John C Hull, Options,
value Futures
of the and Other
option Derivatives
to purchase one million barrels is
Problem
Use the Blacks model to value a 1-year European put option
on a 10-year bond. Assume that the current value of the
bond is $125, the strike price is $110, the 1-year interest
rate is 10% per annum, the bonds forward price volatility is
8% per annum, and the present value of the coupons to be
paid during the life of the option is $10.
In this case, F0 = (125 10)e0.1x1 = 127.09, K = 110, = 0.08, and
T = 1.0
d1 = {[ln(127.09/110) + .0064/2]}/.08 = 1.8456
d2 = d1 0.08 = 1.7656
The value of the option is
110e-0.1x1 N (-1.7656) 115N (-1.8456) = 0.12
Or $0.12
John C Hull, Options, Futures and Other Derivatives
55
Problem
Calculate the value of a 4-year European call option on a bond that
will mature 5 years from today using Blacks model. The 5-year cash
bond price is $105, the cash price of a 4-year bond with the same
coupon is $102, the strike price is $100, the 4-year risk-free interest
rate is 10% per annum with continuous compounding, and the
volatility for the bond price in 4 years is 2% per annum.
We use Blacks formula.
The present value of the principal in the four year bond is 100e -4x0.1
= 67.032.
The present value of the coupons is, therefore, 102 67.032
= 34.968.
So forward price of the five-year bond is : (105 34.968)e4x0.1 =
104.475
F0 = 104.475, K = 100, r = 0.1, T = 4, and = 0.02.
dC
John 2 = d1 Options,
Hull, 0.024Futures
= 1.0744
and Other Derivatives
56
N(d1 ) = .6439
N(d2 )= .4247
e-rt = .7789, r = .05, t =5.
C = 50 x .6439- 50 x.7789x .4247 = 15.66
57
Problem
A companys stock price is $50 and 10 million shares are
outstanding. The company is considering giving its
employees 3 million at-the-money 5-year call options.
Option exercises will be handled by issuing more shares.
Estimate the cost to the company of the employee stock
option issue. Asssume the value of an option is $ 15.66.
N = No. of existing shares, M, the no. of new options
The cost to the company of the option is[ NS + MK]/[N+M]
-K
= [Nx (S-K)]/(N+M) where, N=10, M=3, S-K = option value
= 15.66
= 10/[10+3} X 15.66 = $ 12.05 per option.
The total cost is therefore 3 million times this or $36.15
million.
58