st: programming for coskewness and cokurtosis

[natty.shen@gmail.com]

Dear STATA experts

Greeting all. I have data of stock return and market return in following format

month      A             B             C          .....................................       Market
1            7.2          10.4        3.6          ...................................         4.5
2            6.5          8.9          4.4          ...................................         5.6
.................        .......       .......       ......        ...................................         ......

I want to estimate coskewness and cokurtosis by:

1. Start first data set with data record # 1-60 of stock A
2. Regress stock return with Market Return, with equation Rit = a + b
(Rmt) with 60 data records
3. Calculate residual "e" of month #1 by e = r(1) - a - b (Rmt(1)),
and repeat for month 2-60
4. Calculate "em" of month #1 by em(1) = the different between market
return of month 1 and average of market return of month 1-60, and
repeat for month 2-60
5. Calculate CSK of month 61 =
Average{(e)*em)^2}/{SQRT(Average(e^2))*Average(em)^2)} of month 1-60
6. Calculate CKT of month 61 =
Average{(e)*em)^3}/{SQRT(Average(e^2))*Average(em)^3)} of month 1-60
7. Save "month", "Stock Name", "r", "CSK", "CKT", "e", "em" of month 61
8. Move data time window to t+1 (data record # 2-61) and repeat step 2- 7
9. Keep doing until all data of stock A and start for stock B until all stock

If STATA can do this, could you please give me some hint on how to?
Thank you very much in advance for your kind help.
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/