Merton76Process

Posted by Penschke, Walter on
URL: http://quantlib.414.s1.nabble.com/Merton76Process-tp3338.html

Hi,

I was wondering why the Merton76Process does not really adhere to the
StochasticProcess interface. I do see that there is more to this process
than a drift and a diffusion. The fact, however, that the drift and
diffusion methods for the Merton76Process are implemented via QL_FAIL, makes
it impossible to use this process in the way, a stochastic process is
supposed to be dealt with.

Would it make sense to incorporate the jump part into the diffusion method?
The diffusion would be a random variable then but at least the interface of
StochasticProcess would be satisfied.

Alternatively, one might consider to break the inheritance and to build up
the Merton76Process separately. If it does not behave like a normal
StochasticProcess, why should it be one?

Comments and thoughts on this welcome,


wpe