We have a BM $X_t$ with $dX_t=\sigma dB_t$ ($X_0$ not necessarily zero!) under the risk neutral measure $\Bbb Q$. Given upper barrier $U$, lower barrier $L$, "strike" $K$ such that $L<X_0<U, L<K < U$, rebate $b$, maturity $T$, and define $m:=\min_{0\le t\le T}X_t$ and $M:=\max_{0\le t\le T}X_t$. Suppose the terminal payoff function is $$|X_T - K|I(L\le m \text{ and } M\le U) + bI(\text{otherwise})$$

Suppose in addition a constant discount rate $r>0$. Is an analytical formula for this double barrier out option's price, i.e. $$e^{-rT}\Bbb E^{\Bbb Q}\left[|X_T - K|I(L\le m \text{ and } M\le U) + bI(\text{otherwise})\right]$$ possible? Thanks in advance.

EDIT Looks like Brownian Bridge is a good start. At least I can see it lead to explicit integral forms.


This is a well tackled problem in the GBM case. See

Geman/Yor (1996), Pricing and Hedging Double-Barrier Options: A Probabilistic Approach. Mathematical Finance, 6(4), p. 365-378

among other references. Though in practice finite differences or MC would be used to deal with discrete dividends, local and/or stochastic volatility, etc.

  • $\begingroup$ Thanks. However in my case the underlying follows a standard Brownian motion instead of a GBM. So they might be quite different. $\endgroup$ – Vim Jul 4 '19 at 9:41
  • 1
    $\begingroup$ Sorry I didn't see it was for a BM rather than for a GBM. However the barrier in the $\mu$-drifted GBM case translates into a log barrier in the $\mu - \sigma^2/2$ drifted BM case so a similar approach should work, perhaps following that outlined in Pelsser Pricing double barrier options using Laplace transforms core.ac.uk/download/pdf/19187200.pdf. $\endgroup$ – Antoine Conze Jul 4 '19 at 10:14
  • $\begingroup$ Thanks. Yeah the approach provided in your paper also works nicely for Bachelier model and leads to even simpler forms. The results are exponentially fast decaying Fourier series, and I think they should already be good enough for my use. $\endgroup$ – Vim Jul 5 '19 at 2:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.