Dear colleagues,

I have been running threshold regressions on Stata 15 but it seems this program doesn't generate the standard error measure for the break. Does anyone knows how to solve this?