KantAcademy_PerpetualAmericanPutOption_1710974920

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 4

Very complex Trading and Quant interview

question on Perpetual American Put Option

Interview Question :
What is the value of a perpetual American put option (in-
finite time horizon)?

Solution :
The task of valuing American options presents more complexity compared to their Euro-
pean counterparts, with put options posing a particularly greater challenge than calls. The
difficulty for an American put option stems from its potential for optimal early exercise at
virtually any point, unlike American call options which are typically optimally exercised only
around certain times (just before dividend distributions). This unique aspect of American
puts contributes to their pricing difficulty.

Let V denote the value of a perpetual American put option on a stock. Let S denote the
stock price. Let X denote the strike price. Assume that the stock pays continuous dividends
at rate ρ. Let σ and τ denote the volatility of stock returns and the riskless interest rate
respectively. The Black-Scholes PDE is given by :

∂V 1 2 2 ∂ 2V ∂V
+ σ S + (r − ρ)S − rV = 0.
∂t 2 ∂S 2 ∂S
For a perpetual put, time decay must be zero. Thus, the PDE becomes :

1 2 2 ∂ 2V ∂V
σ S + (r − ρ)S − rV = 0.
2 ∂S 2 ∂S
Let S denote the lower exercise boundary (this is how low the stock has to go before exercise of
the put becomes optimal - it has to be determined). Then, we have the boundary conditions

V (S = S) = X − S,
∂V
= −1,
∂S S=S
V (S) ≤ X − S.
All of this ODE’s solutions may be represented as a linear combination of any two linearly
independent solutions. It follows that

V (S) = A1 V 1 (S) + A2 V 2 (S),

where A1 and A2 are constants, and V 1 and V 2 are linearly independent solutions of the
ODE. We need to guess a form of solution (tricky part but it is a classic approach) and we
suppose V 1 = S λ1 , and V 2 = S λ2 for some constants λ1 , and λ2 . Substitution of V i into the
ODE yields (for i = 1, 2, and for S ≤ S):
 
1 2
σ λi (λi − 1) + (r − ρ)λi − r S λi = 0.
2

Rearranging and collecting terms in λi , we get for i = 1, 2:


 
1 2 2 1 2
σ λi + r − ρ − σ λi − r = 0.
2 2

This is a quadratic formula, with solutions for λi :

1 2
 q 2
− r − ρ − 2σ + r − ρ − 12 σ 2 + 2σ 2 r
λ1 = 2
,
σ
1 2
 q 2
− r − ρ − 2σ − r − ρ − 12 σ 2 + 2σ 2 r
λ2 = .
σ2

The solutions for λi can be seen to satisfy λ1 > 0 if r > 0, and λ2 < 0 if r > 0. Let us
now consider the behavior of the general solution we have derived: V (S) = A1 S λ1 + A2 S λ2 .
First of all, with λ1 > 0, and λ2 < 0, then

lim A1 S λ1 + A2 S λ2 = ±∞, if |A1 | > 0.



S→+∞

However, the boundary conditions put both upper and lower finite bounds on the value of
the put. Therefore, A1 = 0, and V (S) = A2 S λ2 . Now, the first boundary condition tells us
that
V (S) = A2 S λ2 = X − S,

X−S
so it follows that A2 = S λ2
, which yields:
   λ2
X −S S
V (S) = S λ2 = (X − S) .
S λ2 S
To determine the solution precisely, we must find the value for the lower exercise boundary,
denoted as S. Applying the second boundary condition, which necessitates that the derivative
of V with respect to S at S equals −1, we find:
 λ2 −1 
∂V S
= λ2 (X − S) ,
∂S S λ2
∂V (X − S)
⇒ = λ2 = −1,
∂S S=S S
⇒ λ2 (X − S) = −S,
λ2 X
⇒ S= .
λ2 − 1
λ2 X
Thus, for S ≥ S ≡ λ2 −1 , the value of the perpetual American put is determined as
 λ2   λ
S X (λ2 − 1)S 2
V (S) = (X − S) = ,
S 1 − λ2 λ2 X
1 2
 q 2
− r − ρ − 2σ − r − ρ − 12 σ 2 + 2σ 2 r
λ2 = .
σ2
1
Conversely, it can be demonstrated using similar reasoning that a perpetual American call
option, within the range of 0 ≤ S ≤ (λλ11−1)
X
≡ S̄, holds a value given by
 λ

(λ1 − 1)S 1
X
V (S) = ,
λ1 − 1
λ1 X
1 2
 q 2
− r − ρ − 2σ + r − ρ − 12 σ 2 + 2σ 2 r
λ1 = .
σ2

This was a tough interview question, there is another variant of this interview question fo-
cusing on European Calls and Puts and it is easier to discuss. Theoretically, a perpetual
European call equates to the stock price itself, while a perpetual European put is valued at
zero, as seen when examining the limit behavior within the Black-Scholes model.

Black and Scholes’ 1973 model assesses the value of European-style puts and calls. In sce-
narios where a stock does not distribute dividends, the valuation of a European call aligns
with that of an American call, as the lack of dividends removes the benefit of exercising the
option early. Consequently, the Black-Scholes model is applicable for pricing American calls
on stocks that do not issue dividends. However, the situation becomes more complex with
the introduction of dividends.
1
The solution is derived from ’Heard On The Street - Quantitative Questions from Wall Street Job Interviews’ and supplemented
by resources from Kant Academy. This exact question was presented to one of our students in a Quantitative Research internship
interview.
Nonetheless, for American calls associated with dividend-paying stocks, both an approxi-
mate solution (identified by Black in 1975) and precise pricing methods (developed by Roll
in 1977, Geske in 1979, and Whaley in 1981) have been established, as detailed in Hull’s
1997 work, specifically in Chapter 11. The scenario for American puts, however, exhibits
additional complexity. The dividend factor plays a less significant role for puts compared to
calls, with the primary incentive for early exercise of a put being driven by the strike price
rather than dividend considerations. While a definitive pricing formula for American puts
remains elusive, several approximation techniques have been proposed.

You might also like