2
$\begingroup$

I asked a question some days back and got an answer which I understand and make sense: Probability of touching short call strike and not touching touching short put strike of a short strangle?

However, the answer to that question brought up another related question to my mind.

With the probability of that event happening given as P(A ∩ Bcomplement)=P(A)−P(A∩B):

Is that the probability for the event with or without time still left to expiration?

Or the probability of the event only without time left in the trade?

NOTE: Because if there is still time in the trade, it can still touch B before the trade expires even though it touched A first and satisfied the condition

$\endgroup$

1 Answer 1

1
$\begingroup$

Referring to one touch down no touch up (OTD-NTU) option with expiry $T$ as interpreted in your previous question, it might help to formalize the payoff stated in the respective answer.

The option pays rebate \$1 at expiry $T$ if

$$ \boxed{ \tau_L \leq T \; {\rm and} \; \tau_L < \tau^H }$$ where:

$$ \tau_L = \min \; \{t \geq 0 : S_t \leq L \} $$

and

$$ \tau^H = \min \; \{t \geq 0 : S_t \geq H \}, $$

with $L< S_0 < H$.

Its price amounts to calculating (under $Q$ probability measure)

$$ E^Q\left[e^{-rT}1_{\{\tau_L \leq T\} \cap \{\tau_L < \tau^H\}}\right]=e^{-rT} Q(\{\tau_L \leq T\} \cap \{\tau_L < \tau^H\}), $$

where $1_A$ is $1$ if event $A$ takes place, and $0$ otherwise, and $r$ is a flat risk-free discount rate.

The \$1 rebate pay can be made at touching (hitting) time $\tau_L$ too, in which case the price of the option is:

$$ E^Q\left[e^{-r\tau_L} 1_{\{\tau_L \leq T\} \cap \{\tau_L < \tau^H\}}\right]. $$

Its calculation is more complex (pay timing is random).

Back to your $A$ and $B$ events in the question, they are:

$$ A = \{\tau_L \leq T \}, \; B = \{\tau^H \leq T \}.$$

So, the payoff for which

$$Q(A\cap B^c) = Q(\{\tau_L \leq T\} \cap \{\tau^H > T\})$$

would be its price is the one that would pay \$1 at expiry $T$ if

$$ \boxed{ \tau_L \leq T \; {\rm and} \; \tau^H > T.} $$

$\endgroup$
4
  • $\begingroup$ Thanks for your reply, but it is a bit math-heavy for me but in summary you mean it would payoff only at the expiration? $\endgroup$
    – user56826
    Jun 11, 2021 at 20:03
  • 1
    $\begingroup$ For the last boxed payoff yes. $\endgroup$
    – ir7
    Jun 11, 2021 at 20:37
  • 1
    $\begingroup$ In general, no. For a NTD (no touch down) option there is only one pay date possible, expiry (or after expiry). For one touch down (OTD) option the pay date can be the hitting time ($\tau$) itself, that is, pay date is random. Same for OTD-NTU (in which case the option terminates immediately with 0 pay when the $H$ gets hit before $L$ does). This forces us to think in terms of the hitting time itself (its expectation, distribution, joint distribution with other hitting times etc.), not just POT probabilities of events like $\tau_L \leq T$ where $T$ is fixed, $Q(\tau \leq T)$. $\endgroup$
    – ir7
    Jun 11, 2021 at 20:37
  • 1
    $\begingroup$ (Note that last probability can be seen as $Q(\tau_L \leq T)=Q(\min_{t\leq T} S_t < L)$, the probability of the minimum of stock price over the life of the option being less than $L$) $\endgroup$
    – ir7
    Jun 11, 2021 at 20:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.