hey,I calculated some exotic option prices based on either a Heston-Stochastic-Volatility, Merton-Jump-Diffusion or Bates model by means of Monte Carlo simulation. Now I want to say something about whether the obtained prices are significantly different from each other. As I understand it this works in the following way:- I look up the (e.g.) the 99.5% quantile of the standard normal distribution (2.576 in this case)- then I calculate for each model: mean +/- 2.576*stdev to obtain the 99% confidence intervall- if these intervalls are distinct for two models then the obtained prices (i.e. the means) are significantly different from each other at the 99% confidence niveauis this correct?thanks, bernd

QuoteOriginally posted by: BerndSchmitzhey,I calculated some exotic option prices based on either a Heston-Stochastic-Volatility, Merton-Jump-Diffusion or Bates model by means of Monte Carlo simulation. Now I want to say something about whether the obtained prices are significantly different from each other. As I understand it this works in the following way:- I look up the (e.g.) the 99.5% quantile of the standard normal distribution (2.576 in this case)- then I calculate for each model: mean +/- 2.576*stdev to obtain the 99% confidence intervall- if these intervalls are distinct for two models then the obtained prices (i.e. the means) are significantly different from each other at the 99% confidence niveauis this correct?thanks, berndNo, you should take the difference and test if it is different than zero. The variance of the difference is the sum of the two variances.

thanks! just to get this straightI calculateabs(mean(model1)-mean(model2)) - 2.576 * sqrt(variance(model1)+variane(model2))and if the values is larger than 0 then the models produce prices that are significantly different from each other at the 99% confidence niveauis this correct?bernd

QuoteOriginally posted by: BerndSchmitzthanks! just to get this straightI calculateabs(mean(model1)-mean(model2)) - 2.576 * sqrt(variance(model1)+variane(model2))and if the values is larger than 0 then the models produce prices that are significantly different from each other at the 99% confidence niveauis this correct?berndyeah, that should work