Advanced Finite Differences: Implicit, Crank-Nicolson, and More
In the previous chapter we built the explicit finite-difference method. It works, it is easy to code, but it has that annoying stability constraint: your time step must be tiny relative to your asset step. Chapter 78 shows how to remove that constraint and get better accuracy at the same time. The price? The code gets a bit more complicated. But as Wilmott says, the extra complexity is worth it.