October 15th, 2014, 4:48 pm
You've a times series you are not sure you should better model as a gbm or a vacisek process and you're asking whether it exists some tool to discriminate between time series? A dilemna à la H.Geman's 'Is Mean Reversion Dead'?