Hello,
I've checked in a MC Basket pricing engine to the CVS. It uses the
multifactor path generator, but the design needs tidying up.
What is going on with Diffusion processes and Stochastic Processes? Is one
to be preferred over the other. I ended up taking all the info from the
BlackScholesStochasticProcess and creating a DiffusionProcess with it.
What is the recommended usage of the different processes?
I think that we need some sort of MultiDiffusionProcess object to contain
a vector of DiffusionProcesses and their correlations. This would also
allow us to enforce the same underlying RiskFreeTermStructure for baskets
of stocks in one currency.
As the variance is in each DiffusionProcess there would be less redundancy
if the Correlation matrix was stored, rather than the Covariance matrix.
The Correlation matrix would then be used to generate a vector
of correlated random numbers.
Comments,
Neil
---------------------------------------------------
Neil Firth
Brasenose College Oxford OX1 4AJ United Kingdom
Office: 01865 280616
[hidden email]
http://www.maths.ox.ac.uk/~firth---------------------------------------------------
Locked
