How does one go about finding out how their VaR estimator works under the hood? Overnight it changed from 6k to 31K with no obvious change in the underlyings ... I guess I need to finish writing my own monte-carlo var sim
Yes, I have John Hull's book and was just looking at the section on using Principal Component Analysis to calculate VaR. I'm more comfortable with monte-carlo however since I have a theoretical model that fits the implied and realized volatilities about as good as it can get.. also perhaps the 'user interface' is holding me back a bit.. I'd like to be able to plug in assumptions about the parameters of the model, or hypothesize about future changes in mean levels and reevaluate the risk under those changed assumptions