35757845
35757845
35757845
S0 = 100
(1 - p)
S1 = 80
We value at-the-money European call
with a strike price, X of ₹ 100 and time
to expiry of one year.
The payoff of call with strike price at X =
100 is
At price of ₹ 125 : 125 -100 = ₹ 25,
At price of ₹ 80 : 0
Risk free return is 6% per annum.
.
We find the expected value of the call at t = 1.
Then we discount the value of the call at t = 0.
the probability of upside movement, p as 0.90.
Therefore, the probability of downside movement (1-p)
is 0.10.
Expected value of stock (at t = 1)
= 0.90 × 125 + 0.10 × 80 = ₹ 120.80
Expected value of the call (at t = 1)
= Higher of (expected value of the stock - exercise
price, 0)
= 120.80 -100.00 = ₹ 20.80
Value of the call today = 20.80/1.06 = ₹ 19.62
Estimating the probability and finding the
expected value of the option is fallacious
because we do not know the appropriate
discount rate.
The seemingly valid value of the call offers
arbitrage opportunity.
For another investor who assigns a different
probability to up move will have a different
value of the call.
With the value of the call at ₹ 19.62 one
arbitrageur can set up a portfolio of long five
shares and write (short) nine calls. The initial
investment would be:
Action Cash flow
Buy 5 shares = - 5 × 100 = - 500.00
Write 9 calls = + 9 × 19.62 = + 176.58
Cash outflow at t = 0 = ₹ 323.42
The portfolio can be set up with borrowing at
6%.
The liability would be 1.06 × 323.42 = ₹ 342.83
after a year.
With price at ₹ 125 With price at ₹ 80
Value of stock would be Value of stock would be
5 × 125 = ₹ 625 5 x 80 = ₹ 400
The liability on calls written The liability for the calls
would be 9 ×25 = ₹ 225. written is zero.
The net cash inflow is ₹ 400. Net cash flow is ₹ 400.
Portfolio Value at T =1
(All figures in Rupees)
- 20 % + 25%
80
- 20 %
64.00 0.00
T=0 T=1 T=2
From the values at T = 2 we work backwards to
value the call using risk neutral probabilities and risk
free rate.
Risk neutral probabilities are 0.5778 and 0.4222 for
up and down movements respectively.
Value of call option for Upper Node at T = 1
Value of call at T = 2: 0.5778 x 56.22 + 0.4222 x 0 = ₹ 32.48
Value of call at T = 1: 32.48/1.06 = ₹ 30.65
Value of call option for Upper Node at T = 1
Value of call at T = 2: 0.5778 x 0 + 0.4222 x 0 = 0
Value of call at T = 1: Zero
Value of call option for Node at T = 0
= 1/1.06(.5778 x 30.65 + 0.4222 x 0) = 17.71/1.06 = ₹ 16.71
To improve accuracy of the option pricing
Binomial model can be extended to multi-periods to
broaden the range of end-values.
For computational ease it is desired that prices at
each binomial tree should converge to the original
value. This is achieved by making up and down
movements reciprocal of each other.
If up move is 25% then the down move is 1/1.25 = 0.80
(25% rise and 20% fall).
As we increase the number of trees by shortening the
time interval the option price under binomial model
moves closer and closer to analytical model such as
Black Scholes.
An American call is no greater in value than an
equivalent European call as by exercising early
one gets the intrinsic value only and loses the
time value of the option.
We may use multi-period binomial model to
value options by:
Computing the risk neutral probabilities of up and
down movement, then
Using the probabilities and starting backwards, we find
the expected payoff of the option at preceding node, and
Discounting the payoff at risk free rate to find the
value at that node.
Proceeding with all steps till node T = 0 is reached.
For American call exercise of option before maturity does not make any
difference in valuation, and its value is same as European call.
We may use multi-period model to value American put by comparing the
value one gets by exercising with that of not exercising, and retaining the
higher of the two, at each node.
The process is summarized as below:
Computing the risk neutral probabilities of up and down movement,
then
Using these probabilities and starting backwards, we find the expected
payoff of the option at preceding node, and
Comparing risk-neutral values with the payoff if exercised at each
node, with higher of the two retained, and
Discounting the payoff at risk free rate to find the value at that node
The process is repeated till the last node is reached.
Binomial model can be extended to currency
options with small modification in the risk free
rate.
While valuing currency option the interest rate must
be considered net of foreign interest rate.
If the domestic risk free interest rate is 10% p.a.
and that in foreign currency is 4% p.a. then while
computing the risk neutral probabilities the interest
rate that must be used is 6% (10% - 4%); the
differential of the two.
Binomial model can also be extended to index
options again with small modification in the risk
free rate.
While valuing index option the interest rate
must be considered net of dividend yield on
the index.
If the risk free interest rate is 8% p.a. and the
dividend yield on index is 3% p.a. then while
computing the risk neutral probabilities the
interest rate that must be used is 5% (8% - 3%);
the differential of the two.
The current value of NIFTY is 4,500. In a period of 3 months it
can go up or down by 10%. Risk free interest rate is 8% and
dividend yield is 2%. Value of 3-m call and put with X = 4600
using single stage binomial model and 1point= ₹ 1.
Solution
3-m European call and put with X = 4,600 can be valued
with risk neutral valuation with net interest rate of 6%.
The risk neutral probabilities are:
rf t 0.06x3/12
e - (1 - d) e (1 - 0.10)
p = =
u-d 0.10 + 0.10
1.0151- 0.90
= = 0.5756 = 57.56%
0.20
1- p = 1 - 0.5756 = 0.4244
Risk Neutral Valuation of Index Options (Figures in Rupees)
T=3m Spot Call X = 4600 Put X = 4600
p= 0.5756 4,950 350 -
4,500
1- p = 0.4244 4,050 - 550
Value of call
At T = 3 m = 350 x 0.5756 + 0 x 0.4244 = 201.45
At T = 0: = 201.45 x e-0.06 x3/12 = 201.45 x 0.9851 = ₹ 198.45
Value of put
At T = 3 m: =0 x 0.5756 + 550 x 0.4244 = 233.44
At T = 0: = 233.44 x e -0.06 x 3/12 = 233.44 x 0.9851 = ₹ 229.96
Binomial model has following limitations:
Too unrealistic for determination of fair value.
The probability distribution seems far from reality.
The primary determinant of the option value i.e. the
volatility in the implied probability of the binary states.
However, in actual practice they are
overcome as follows:
By adding binomial periods usually around 30 small
intervals.
With 30 intervals we have 31 different terminal prices.
There would be 230 different paths to achieve these
31 terminal prices.
The probability of maximum price and minimum price is
1/230. The next level of prices would consist of 29 up
moves and one down move. It is possible to achieve it in
30 different ways i.e. 30C1. The probability iwill be 30C1 / 230.
Similarly we may find the probability of each of the 31
possible terminal prices. Thus probability distribution would
be far closer to reality than what one imagines in binary
state.
The way to overcome the objection of absence of volatility is
to incorporate the stock volatility in the binary model. It is
possible to equate the up or down move based on stock
volatility, σ.
For
n stage binomial trees,
j up moves of d, and
n – j moves of d, and
probability of up move of p
the call value is given by
n
n!
∑
0 (n - j)! j!
p j (1 − p)n− j Max (0,u j d n− j S − X)
c0 =
(1 + r) n