FINM3003 ANU Lecture 2
FINM3003 ANU Lecture 2
FINM3003 ANU Lecture 2
Lecture 2: Options
Lecturer: Fei Huang
2.1 Introduction
An option is a contract between 2 parties the buyer (=hedger) and seller (=writer) to purchase or provide
an asset in the future.
Call options give holders the right (but not the obligation) to purchase an asset at time T for price K (’strike
price’) agreed at time 0. Put options give holders the right (but not the obligation) to sell an asset at time
T for a price K agreed at time 0. Options may be European type, which can be exercised only at maturity
T ; or American type, which can be exercised any time in [0, T ].
St is not known at times t ≤ T , so we find ct , 0 ≤ t ≤ T , based on a certain probablity model for St . The
value of an European option turns out to be a certain kind of expectation of the payoff of the option- but
NOT the expecatation under the real world distribution of the stock price. Valuation is again based on
no-arbitrage principles: seting up a risk-free hedge. American optioins are a bit more complicated.
The notations we are going to use are as follows.
2-1
2-2 Lecture 2: Options
Assume call and put have the same strike price, expiry date, then we have
ct − pt = St − Ke−r(T −t) ,
for 0 ≤ t ≤ T .
Proof. Suppose
ct − pt < St − Ke−r(T −t)
then
ct + Ke−r(T −t) < St + pt
At time t, sell 1 put and short 1 share, purchase call and invest Ke−r(T −t) in bonds. The value of this
portfolio at time t is positive (> 0). The value of the portfolio at time T is
So there is no liability at time T but there is positive income at time 0, which is a risk free profit. That is
impossible. Thus, we have
ct − pt ≥ St − Ke−r(T −t)
Similarly, we can get an inequality in the reverse direction. Thus proving
ct − pt = St − Ke−r(T −t) ,
for 0 ≤ t ≤ T .
Put-call parity bounds for American options with no dividends are as follows:
• Lower bound: Ct − Pt ≥ St − K
Proof. Since
Ct − Pt = ct − Pt ≤ ct − pt = St − Ke−r(T −t) ,
we have the upper bound. To prove the lower bound, consider the portfolio at time t:
Let Dt be the present value of dividends receivable up till time T . The put-call parities are as follows.
It may be optimal to exercise an American call early when dividends are paid. If so, it will only be at
their immediately prior to an ex-dividend date. (After the dividend, stock price decreases, making call less
valuable.)
If the stock has a continuous dividend yield of q% p.a. The put-call parities are as follows.
We can derive these from elementary no-arbitrage arguments. Or from put-call parity (Hull p227)
• Upper bound: ct ≤ St
Proof. If ct > St , at time t, we sell call and buy stock. Left with ct − St > 0. At time T, the portfolio is
worth
− max (ST − K, 0) + ST = min (ST , K) ≥ 0,
which is impossible. So we have the upper bound.
Follow from put-call parity:
ct − pt = St − Ke−r(T −t) ,
which implies ct ≥ St − Ke−r(T,t) and of course ct ≥ 0. So we have the lower bound.
• Portfolio 1: Borrow K, exercise option and receive stock. At time T, the portoflio is worth ST −Ker(T −t)
, possibly < 0.
• Portfolio 2: Keep option to maturity. At time T , the portfolio is worth max (ST − K, 0)
max (K − St , 0) ≤ Pt ≤ K,
which are obvious. Whether we exercise early or not depends on circumstances. Put will be exercised early
if deep in money. For example, if St = 0, exercising put, realise K; deferring exercise gives St chance to rise
higher, lessening profit.
Since American put may be worth more, we have
Pt ≥ pt .
An option lasts for time T and is dependent on a stock. We assume the current stock price is S0 and the
current option price is f . We also assume during the life of the option, the stock price can either move up
from S0 to S0 u (u > 1), or down from S0 to S0 d (d < 1). If the stock price moves up to S0 u, we suppose
the payoff from the option is fu ; otherwise, we suppose the payoff from the option is fd .
We imagine a portfolio consisting of a long position in ∆ shares and a short position in the derivative. Then
the portfolio value is:
fu − fd
∆=
S0 u − S0 d
2-6 Lecture 2: Options
The portfolio value at time T is S0 u∆ − fu (which is equal to S0 d∆ − fd ); and the portfolio value today is
∆S0 − f0 . Risk-free portfolio mush earn riskless rate, so we have
p is the probability that gives a return on the stock equal to the risk-free rate
The value of a derivative is then its expected payoff in a risk-neutral world discounted at the risk-free rate.
f = PV(Ê(fT )) (2.1)
Notes: When we are valuing an option in terms of the price of the underlying asset, the probability of up
and down movements in the real world are irrelevant.
We extend the one-step Bimomial model to two-step Binomial trees, see Figure 2.4. During each step, the
stock price either moves up from S0 to S0 u or moves down from S0 to S0 d. We assume the risk-free rate is
r and the length of the time step is ∆t years. Apply 1-step Binomial tree recursively, we get:
er∆t −d
where p = u−d . Substituting from Equations 2.3 and 2.4 into 2.2, we have
When we extend to n-step Binomial trees, the same fomula 2.1 applies, where now the risk-neutral distribu-
r∆t
tion is Binomial(n, p), with p = e u−d−d and ∆t = Tn . Thus, for a European call option, we have
n
−rT
X n i
c0 (n) = e p (1 − p)n−i max(S0 ui dn−i − K, 0)
i=0
i
n
X n i
= e−rT p (1 − p)n−i (S0 ui dn−i − K)
i=a
i
n n
X n i X n i
= e−rT p (1 − p)n−i S0 ui dn−i − e−rT p (1 − p)n−i K
i=a
i i=a
i
where
ln(S0 /K) + (r + 21 σ 2 )T
d1 = √ ,
σ T
√
d2 = d1 − σ T ,
r∆t
which is the Black-Scholes formula. The choice of u, d (and p = e u−d−d ) is to match the parameters of the
binomial model with the geometric Brownian motion (Continuous time model) for the stock price (see later).
S0 ui dn−i > K.
Next look at
B c (an , n, pn ) = P (Bin(n, pn ) ≥ an )
= 1 − P (Bin(n, pn ) < an )
Bin(n, pn ) − npn an − npn
= 1 − P( p <p )
npn (1 − pn ) npn (1 − pn )
By the CLT
Bin(n, pn ) − npn D
p → N (0, 1), as n → ∞
npn (1 − pn )
We have to work out what happens to
an − npn
p ,
npn (1 − pn )
as n gets large.
Lecture 2: Options 2-9
Similarly
B c (an , n, p0n ) → N (d1 ), as n → ∞
It follows that
c0 (n) → S0 N (d1 ) − Ke−rT N (d2 ), as n → ∞
In other words, the approximate call price based on an n-step binomial approximation converges to the B-S
price as the number of steps in the binomial tree tends to infinity.
Note that we must choose n large enough(so ∆t is small enough) to keep p between 0 and 1.
Since √
er∆t − e−σ ∆t
p= √ √ <1
eσ ∆t − e−σ ∆t
√ √ r2
iff er∆t − e−σ ∆t
> 0, iff σ ∆t > r∆t,iff ∆t < σ2 , we need
T σ2
n>
r2
for this. So, for example if σ = 0.2, r = 0.1, T =1yr, this means we must take n > 4, which is not a very
stringent requirement; we need n > 30 or so for the normal approximation.
Example 2.4.1. Consider S0 = 50, K = 52, ∆t = 1yr, T = 2yrs, u = 1.2, d = 0.8, r = 0.05. Price the
American put option at time 0. [Hint: write down the binomial tree and work backwards. Test at each node
to see if exercise at that time is optimal.]