In this paper we simulate 200 re-samples with
replacements to estimate the BTS standard errors and the tail probability
intervals at ˛= 5%. Of course, one can allow the sample size
of BTS to vary and then to examine the effects of such variation on
our measures of dispersion.