Bootstraping

[Michelle]

Dear friends,

I am working with time series of yield spreads on bonds that are all non-normally distributed, and are very skewed to the right (the skewness is greater than 2 in some of the cases). I therefore believe I should calculate bootstrapped p-values for the test statistics. Does anyone know whether EViews offers this option, like it does for BDS tests? If how could it be done manually?

Thanking you kindly in advance,

Michelle.

[trubador]

EViews has a resampling procedure for series/groups and a resampling function for vectors/matrices. With a little programming, I guess you can calculate what you want.