Chow predictive failure test
[fstat,f_pvalue,dfnum,dfden]=predfailin0(res,n1,np)
resulols = a results tlist from a first stage estimation
n1 = # of observations of the sub-period
fstat = value of the test statistic
f_pvalue = its p-value
dfnum = # dof of the numerator
dfden = # dof of the denominator
bounds('1964q3','1989q2') // column HE in table 9 of Hendry and Krozlig paper rols=ols('delts(lm1-lp)','lagts(lm1-lp-ly)','delts(lp)','rnet','delts(lagts(lm1-lp-ly))','cte') [fstat,f_pvalue,dfnum,dfden]=predfailin0(rols,50) //Useful mainly for programming purpose (since jbnorm does much more).