Statalisters,
When we covered the cauchy distribution in my econometrics class, we were warned how nasty things can be if the world turned out to be bell shaped but Cauchy instead of normal--and how hard to tell would that be. People in finance also worry a lot about stock prices behaving Cauchy-ously.
For estimations purposes, however, robust estimation is needed since OLS would be do a poor job against Cauchy-distributed errors--because of its lack of high-order moments.
My professor suggested to use least absolute deviation estimates (LAD), or quartile regression (-qreg- in Stata) to verify if OLS was not being driven by cauchy-like problems.
Would the experts on this list agree with this?
-Javier
--
__________________________________________________________
Sign-up for your own FREE Personalized E-mail at Mail.com
http://www.mail.com/?sr=signup
*
* 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/
