Valuing Commodity Options using Stochastic processes

[swoop[TR]]

Guys,
Any of you guys use a particular model to incorporate stochastic volatility in models used to value commodity options prices (especially Gold but not limited to).
I'm not sure which model I should use that would generate decent results. They are OTC options and hence cannot be recalibrated to the market, except for one. I'm using a modified BS model but it keep overvaluing the option. It is very probably due to the use of a static volatility initially calibrated at one delta. It seems we're having a sticky delta kind of problem.
The Gold market does have a certain skew/leptokurtic distribution.
Any books or advice?
Anyone use Jump diffusion techniques, or other models ?

 

[Reitberg]

Let me state the obvious.

When it comes to pit traded commodity options, your best valuation method is the "good old boy system" where you have a great relationship with a pit broker.

Those guys in the options pits will do special deals with each other, and it is really unlike anything that I have seen in the electronic equities options.

I understand that I'm not answering your question about theoretical valuation, but I do know that when it comes to pit traded commodity options, options vaulation models go out the window.

I hope this helps.