I wouldn't use -nl- to estimate this rate. On the face of it, it is more
natural to take logs and then fit a regression in each panel. You can do
that using -statsby- to pick up the regression slopes. Then -merge- back
to your original dataset.
Nick
[email protected]
Quang Nguyen
I have the following sample data set:
Household x y t
1 60 120 7
1 120 100 30
1 100 300 60
1 50 250 900
...
2
2
2
...
I would like to estimate the discount rate for each household based
on the fomular: x = y*exp(-rt), where r is the interest rate. I know
that in Stata we can use: nl (x=y*(exp(*-{r}*t) for each household. My
questions are:
1. How we can ask State to do the the above estimate for all household
at the same time instead of one by one?
2. How can we save the ESTIMATED discount rates as an variable that
can be used for latter analysis? For instance, on emay study the
relationship between the household's discount rate and the educational
level.
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
