Moving Kurtosis & Skewness for Panel Data

[Ostap Teplyy]

I have a panel data for 20000 stocks over a sample period from 1950 to 2018. What I need to do is to compute for each of this stocks a moving skewness and kurtosis over the previous 5 year. My data is on a monthly basis, therefore what it should be done is to compute those measures based on the first 60 observations of a given stock, then with a step of 12 month compute it again using the previous 5 years, and so on as long as the data is still available. Then it should be repeated for each stock.
Any idea on how to do this? I got stack in this problem since a while and have no idea ho to go on.

Ps. I'm still a beginner in Stata and for unknown reason I cannot download any external package. Would appreciate if you can present me any possible solution without those packages

[Nick Cox]

for unknown reason I cannot download any external package

People often know why they can't. If you don't know why then you should ask local IT Support or study

help netio

to see whether you have told Stata about your set-up.

I used to find that these three needed to be set in my profile.do


set httpproxy {on|off} [, init]


Set proxy host name

set httpproxyhost ["]name["]


Set the proxy port number

set httpproxyport #

PS Ideally my answer is to use rangestat from SSC.

If you really can't access that or anything similar, rolling with summarize might help seem indicated, but that may prove sloowww.