[R-SIG-Finance] Irregularly Spaced Time Series

Brian G. Peterson brian at braverock.com
Sat Mar 10 03:45:04 CET 2007


On Friday 09 March 2007 20:18, elton wang wrote:
> Why not add zero returns for businessdays inbetween?
> If you really want to fit to the irregular dates,how
> are you going to use it to predict?

As Gabor Grothendieck has already pointed out, the zoo class may be 
reindexed or aggregated into a regular series for the purposes of GARCH 
using several methods to regularize the series.  Missing observations do 
not necessarily mean zero return for the period.  

Regards,

   - Brian
 
> > On 3/9/07, Rich Ghazarian <rich7804 at yahoo.com>
> >
> > wrote:
> > > Irregularly Spaced Time Series:
> > >
> > > I am trying a TS data that has the following
> >
> > format,  but I am not able to import it as a TS
> > frame?  I want to fit the data to a GARCH, but I can
> >
> > > not do that unless it is indexed to the
> >
> > appropriate series.
> >
> > > Date                Rt
> > > 1/12/2003            0.1
> > > 1/13/2003            0.5
> > > 1/14/2003            0.3
> > > 1/17/2003            0.1
> > > 1/18/2003            0.32
> > > 1/20/2003            0.276
> > > 1/21/2003            0.28
> > > 1/22/2003            0.284
> > >
> > > I search and could not find an answer to the
> > question.



More information about the R-SIG-Finance mailing list