st: RE: RE: RE: RE: Question on Dfactor and Gaps in Time Series

["Degas Wright" <dwright@cornerstoneadvice.com>]

Nick,
The data is weekly prices, volume and earnings yield, so missing a week
of data is not critical to the model. So I can delete the missing data
week and reorder the periods to assume that the missing data point never
existed as pointed out by Richard Gates.

Thanks,

Degas A. Wright, CFA
Chief Investment Officer
Decatur Capital Management, Inc.
250 East Ponce De Leon Avenue, Suite 325
Decatur, Georgia  30030
Voice: 404.270.9838
Fax:404.270.9840
Website: www.decaturcapital.com

-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Nick Cox
Sent: Monday, October 11, 2010 2:09 PM
To: 'statalist@hsphsun2.harvard.edu'
Subject: st: RE: RE: RE: Question on Dfactor and Gaps in Time Series

Sorry, no idea. But what's the answer to my question about your data? 

(A very good convention introduced by William Gould on this list is to
refer to commands using - -. Thus -var- is the Stata command of that
name. This is (a) easy on the eye (b) a convention that doesn't clash
often with other conventions.) 

Nick 
n.j.cox@durham.ac.uk 

Degas Wright

Nick,
Thanks for your response.  I have noted that the VAR command seems to
work better with the data than the dfactor model, i.e. the VAR command
runs fine over the same data. So if I conduct the following:  

var (D.(p v ep))

How do I obtain the in sample estimate of D.p?

Nick Cox

I'd look more closely at the data to see what gaps exist before -tsfill-
and whether it actually makes any difference. 

Weeks are awkward. Stata's weeks are defined so that week 1 of any year
begins on 1 January and there are always 52 weeks in any year, the last
being 8 or 9 days long. It may be those assumptions doesn't map
perfectly on to your data somehow. They don't often correspond to the
way that stock markets are reported, I guess. 

Either way, -dfactor- is bailing out very early, and the most natural
thing to do is to look at the data. I doubt that -dfactor- is being
dumb, or that its recent introduction to Stata is of consequence here. 

Degas Wright

I am using the dfactor command and have run into the gap in time series
error. My data is price (p), volume (v) and earnings yield (ep) and I am
trying to develop a dynamic factor model using the dfactor command. My
code is:

tsset
        time variable:  date, 2008w25 to 2010w40
                delta:  1 week

. dfactor(D.(p v ep)=,noconstant)(f=,ar(1/2))
gaps in the time series are not allowed
r(459);

. tsfill,full

tsset
        time variable:  date, 2008w25 to 2010w40
                delta:  1 week

. dfactor(D.(p v ep)=,noconstant)(f=,ar(1/2))
gaps in the time series are not allowed
r(459);

It appears that the dfactor command is not recognizing the tsfill, full
step.  The dfactor is a relatively new command and I did not see
anything regarding the time gap issue and dfactor on the FAQ.


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/