Monte Carlo Simulation for Bermudan Swaptions (Conference Proceeding 2001)
The presentation gives a survey of the different methods/models that are available for the pricing of American/Bermudan options using Monte-Carlo.
The interest in finding efficient ways to do that followed the introduction of the so-called Libor-Market Model – which by its nature requires Monte-Carlo simulation in order to price this type of instruments.
The presentation kicks-off by specifying the Free Boundary problem for this type of instruments. Next we look at the traditional ways to price these type of assets – the class of spot-rate models. With respect to the spot-rate models, we investigate the volatility structure.
Next we turn our attention to LMM, and specifies it in a stationary volatility structure framework.
The rest of the presentation looks at different ways to value Bermudan style assets using Monte-Carlo simulation, keywords are here: Non-recombining trees, Markov Chain Models, reduction of the exercise region, Stochastic Mesh methods, Longstaff and Schwartz and Andersen’s boundary optimization technique, Jaeckel’s method.
The presentation closes off with a short non-recombining tree approach example.
Articles