Re: Re: st: plot confidence intervals with time series

[annoporci <annoporci@gmail.com>]

Thanks Kit, your comment is very helpful!

> It is not obvious to me that kurtosis in the data carries over to the  
> distribution of the beta-hats in your regression model,
> which are after all asy normal, The finite-sample CI uses the t  
> distribution, which assumes symmetry, but allows for greater kurtosis
> than would z-scores.
>
> Kit

--
Patrick Toche.
*
*   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/