Presenters: Ken Vetzal and Peter Forsyth
**Please note attendence is limited to 25 for this course, please register early**
This course will focus on methods for pricing path dependent options using numerical PDE methods. Participants will receive detailed course notes (120 pages) and prototype Matlab software.
The course is organized so that morning sessions are lectures, while afternoons are "hands on" sessions which are devoted to using Matlab software for pricing options.
It will be assumed that the participants have a basic understanding of option pricing, and are also familiar with standard numerical algorithms, such as finite difference methods.
An overview of the topics covered includes:
Day 1
Morning
* Introduction, comparison of PDE, Monte Carlo, Lattice methods
* Basic discretization methods
* Stability
* Causes of oscillations in Delta and Gamma
* Positive coefficent discretizations
* Smoothing discontinuous payoffs (projection, averaging)
* American options: the penalty method
* Implicit vs. explicit handling of the American constraint
* Timestep selection
* Discretely observed barriers
* Discrete non-proportional dividends
Afternoon (Matlab Sessions)
* Non-uniform grids, convergence tests
* Fully implicit vs. Crank-Nicolson
* American options: use of the Penalty method
* Variable timesteps
* Discrete Dividends
* Discontinuous Payoffs
Day 2
Morning
* Nonlinearities: uncertain volatility, transaction costs
* Convergence to the viscosity solution
* Finite volume vs. forward/backward differencing
* Path dependent options: the augmented state variable approach
* Examples: Asian options, Shouts, Parisian barriers
* Software issues: a general framework
* Implementation on high-performance architectures
* Introduction to multi-factor option pricing
Afternoon (Matlab Sessions)
* Barrier options
* Fully implicit methods for uncertain volatility models
* Asian options: interpolation effects, comparison with forward shooting grid
* Parisian options
Fee: $400 -- includes lunch and and refreshments for both days. Space in this course is limited to 25, please register early.
Please note that this course is now full.