[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: Re: st: RE: RE: Bootstrap and Technical analysis

From   "Eva Poen" <>
Subject   Re: Re: st: RE: RE: Bootstrap and Technical analysis
Date   Wed, 20 Aug 2008 15:59:31 +0100


what you say doesn't make a lot of sense to me. This may be due to my
lacking background in time series data and bootstrap techniques;
others may be better qualified to comment. However, some list members
have strongly discouraged you from applying this kind of procedures to
paired time series data.

Your approach does not seem to be particularly well thought through,
so maybe you want to go over some statistics texts before asking for
more advice from the list. Please also consult the online help files
and the manuals for simple syntax questions.

Bootstrap is a method for learning more about the distribution of an
estimator in the situation where standard formulas don't apply. The
resulting bootstrap sample is used to compute a standard error, for
example. Its individual observations are not meant to be used as data
in subsequent procedures, as far as I know.

2008/8/20 Mahmoud Abd-El-Aal <>:
> I use reg var1 var2 to get the residuals which are saved in e(rss)

This is not correct. e(rss) is a scalar and contains the residuals sum
of squares. I am not sure how this information will help you in what
you want to do.

> I need to save this matrix into a variable (var4), i do not know how to do
> that

As I mentioned, e(rss) is a scalar not a matrix, and it is unnecessary
to save it as a variable. If you want to retain the residuals from a
regression, you can do so using -predict-; see -help predict-.

> The following steps for the problem i mentioned just in the previous
> thread i think i can solve them throught the following
> -bsample _N, var4

This is not legal Stata syntax. See -help bsample-, or even better the
manual entry, to learn more about the bsample command.

> This will produce a new set of returns , then all i want to do is
> bootstrap these returns to see their meeans

Again, why you would want to take random samples with replacement from
your data, then do some bootstrap on these to investigate means is
beyond me. I am sorry but I don't think I can comment any further on
your problem.

Best regards,
*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index