Austin suggests working from first principles:
webuse grunfeld, clear
ren com id
ren year yr
tsset id yr
g c=d.mval
la var c "Change in mvalue"
bys id (yr): g sc=sqrt((c^2+l.c^2+l2.c^2)/3-((c+l.c+l.c)/3)^2)
* Optional correction (pop sd vs sample sd):
replace sc=sqrt(3/2)*sc
la var sc "SD of change in mvalue over prior 3 yrs"
and see -help tsvarlist- for more.
On 10/10/07, [email protected] <[email protected]> wrote:
> Hi,
>
> I know this should be simple but I just don't know what commands I
have to type in order to get Stata 9.2 to run a moving 'standard
deviation' for one variable of my panel (2892 companies, years:
1995:2005). Basically, the new variable should be measured as the
standard deviation of the change in 'earnings' over a three-year
period. I wish to measure the changes from year t-3 until year t.
Whenever (randomly, for some years) there are no data available, no
standard deviation is computed. That's all.
> I have been advised to use -rolling- but, after vain attempts to
follow Manual's suggestions (and not much help from searches on
previous similar threads, too), still I can't figure out the precise
commands and options I need.
> I shall really appreciate some help (an example, I hope) to make
the calculation. Thank you in advance.
>
You could always use -mvsumm- from SSC.
Kit Baum, Boston College Economics and DIW Berlin
http://ideas.repec.org/e/pba1.html
An Introduction to Modern Econometrics Using Stata:
http://www.stata-press.com/books/imeus.html
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
