Dear Austin,
this is the result (for the first firm only) of the short program you kindly wrote to me:
. webuse grunfeld, clear
. g k0=int(uniform()*1000) if year==1935
(190 missing values generated)
. g k=.
(200 missing values generated)
. bys c(y): replace k=cond(l.k<.,l.k*(1-`d'),k0)+i
time-series operators not allowed
r(101);
. li com year k0 k inv in 1/10, noo sepby(com)
+-----------------------------------+
| company year k0 k invest |
|-----------------------------------|
| 1 1935 715 . 317.6 |
| 1 1936 . . 391.8 |
| 1 1937 . . 410.6 |
| 1 1938 . . 257.7 |
| 1 1939 . . 330.8 |
| 1 1940 . . 461.2 |
| 1 1941 . . 512 |
| 1 1942 . . 448 |
| 1 1943 . . 499.6 |
| 1 1944 . . 547.5 |
+-----------------------------------+
-----------------------------------------------
As you can see (I tried several times...), no new "k" has been generated by Stata 9.2. I don't understand the reason and I ask for your help again to find a solution. Thank you.
Damiano
---------- Initial Header -----------
