Consider the following initial-boundary value problem for $u = u(x,t),$

$$u_t - a u _{xx} = f(x,t) \text { for } 0 < x < L \text { and } 0 < t< T$$ along with bunch of initial and boundary conditions.

If we want to employ finite difference method to solve the above parabolic problem in 1D, then under explicit Euler method, we use forward difference approximation in time and under implicit Euler method, we use backward difference approximation in time.

I'm slighty confused about the terminology for finite difference method for option valuation. I was reading Paolo Brandimarte's Numerical Methods in Finance and in chapter 9 Option Pricing by finite difference method, to solve Black-Scholes PDE, under explicit scheme, he approximates the derivative with respect to time by a backward difference and under implicit scheme, he uses forward difference approximation in time. Isn't that the exact opposite of the previous definition?

Does it have something to do with terminal conditions in Black-Scholes PDE rather than initial condition and we go backward in time ?

  • $\begingroup$ Could it be the case that Brandimarte is not using calendar time $t$ but time to maturity, $\tau \equiv T-t$? $\endgroup$ Jan 15 '21 at 5:28
  • $\begingroup$ @Kermittfrog No. The author is using $t$ and not $\tau.$ $\endgroup$
    – Rhombus
    Jan 15 '21 at 6:50

An explicit (resp. implicit) finite difference scheme means you do not need to (resp. have to) solve a linear system of equations to find the solution at each intermediate time step. Whence their names!

Consequently, it really depends on if you have an initial, versus a terminal condition in time, and in what 'axes' you express and how you discretise the PDE you are solving.


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.