*Monte Carlo Estimation of sales-ads-price model in Stata
*The file Monte Carlo Stata.dta contains 50 observations of
*invented data on ads and prices
*First make 50 random draws from a normal distribution.  
*We assume a standard deviation of 10, but you can make it
*whatever you like.
 drawnorm e, n(50) sds(10)
*Create your Y variable, assuming "true" parameters 
*of 200, 5, and -5:
 gen y=200+5*ads-5*prices+e
*Now do an ols regression to see how close you come to 
*retrieving the "true" model parameters
 reg y ads prices
*Stata remembers, so to repeat this in Stata, you will 
*first have to drop e and y, then re-generate them
 drop y
 drop e
*Now repeat the gen and reg commands above.
*You can run this again and again and compare results.