|
The finite difference method is a powerful tool for numerically
solving partial differential equations. For valuing financial derivatives, it
offers many advantages over trees and the Monte Carlo method. Its popularity has
lagged behind those methods because it requires more specialized expertise. To
meet this need, Tavella and Randall have produced an excellent monograph.
Targeted to knowledgeable financial engineers, it assumes considerable knowledge
of stochastic calculus and derivatives pricing theory.
Two opening chapters review familiar concepts of financial engineering:
Ito's calculus, arbitrage pricing theory, etc. They discuss when a derivative
can be modeled with a PDE and techniques, such as change of measure or dimension
reduction, that make a problem more amenable to analysis by finite differences.
|
|
|
|
1. Introduction 1
2. The Pricing Equations
3. Analysis of Finite Difference
Methods
4. Special Issue
5. Coordinate Transformations
6. Numerical Examples
|
|
Chapter 3 covers how to construct finite difference
approximations. It focuses on stability issues and assesses distortions caused
by discretization. Chapter 4
addresses practical issues related to jump models, path dependence, boundary
conditions, etc. Chapter 5 covers coordinate transformations. The final chapter
covers a number of valuable case studies.
I highly recommend this book for practicing financial engineers. Prior
familiarity with finite difference methods, as might be obtained from Wilmott,
Dewynne and Howison (2000) or Seydel (2002),
will be valuable.
|