Newey command with bootstrap

[Chang Seok]

Hello,

I have one quick quesiton regarding newey-west and bootstrap.

As far as I know, stata command newey is to correct or heteroscedasticity and autocorrelation and bootstrap is resampling technique when the sample size is small.

I would like to know if I could you both newey and bootstrap command at the same time.

So I made an example as follows

clear

sysuse citytemp4

drop if heatdd == .

bys division: gen time = _n

keep if division == 1

tsset time

bootstrap: newey heatdd tempjan, lag(6)

Then I got an error message saying that

insufficient observations to compute bootstrap standard errors
no results will be saved

Is there any way I can implement this?

Thanks in advance.

[Rich Goldstein]

I virtually never have this type of data (i.e., w/autocorrelation); however, the traditional bootstrap is for iid data and that is, apparently, now what you have; you may be able to do something with a block bootstrap; see

help wildbootstrap