Yes. No problem. This is one link on this: http://www.polsci.wvu.edu/duval/ps602/Notes/STATA/montecarlo.htm

program rdwalk,rclass
version 10
syntax [,n(integer 1000)
drop _all
quietly set obs 'n'
capture drop y x1 x2 x3 x4 x5 rndy rndx1 rndx2 rndx3 rndx4 rndx5
quietly {
gen rndy=100+(1000-100)*invnorm(uniform())
gen rndx1=10+(100-10)*invnorm(uniform())
gen rndx2=9+(22-9)*invnorm(uniform())
gen rndx3=2+(12-2)*invnorm(uniform())
gen rndx4=6+(24-6)*invnorm(uniform())
gen rndx5=0+(100-0)*invnorm(uniform())
gen y=rndy
gen x1=rndx1
gen x2=rndx2
gen x3=rndx3
gen x4=rndx4
gen x5=rndx5
replace y=y[_n-1]+rndy[-n] in 2/'n'
replace x1=x1[_n-1]+rndx1[-n] in 2/'n'
replace x3=x3[_n-1]+rndx2[-n] in 2/'n'
replace x4=x4[_n-1]+rndx3[-n] in 2/'n'
replace x5=x5[_n-1]+rndx4[-n] in 2/'n'
regress y x1 x2 x3 x4 x5
matrix matb=e(b)
matrix matv=e(V)
return scalar b=el(matb,1,2)
return scalar se=sqrt(el(matv,2,2))