dear all,
could you help me out with respect to Kit Baums -zandrews- command:
Zivot-Andrews Unit Root test allowing for a single break in intercept and/or trend
zandrews [varname] [if exp] [in range], maxlags(#) trim(#) break() generate(varname) lagmethod() level(#) graph
my problem: for my analysis the [if] qualifier and [in] range fail to specify the sample. while -zandrews- works without constraint (structural break at obs 624 of 1055), it is only reporting structural breaks within the range 1/365 for any(!) constraint with the [if] or [in] terms.
an example:
zandrews jcs2 in 732/1055, maxlags(1) trim(0.15) break(both) graph
yields in a minimum t-statistic -3.574 at 27apr2007 (obs 117)
gen group=0
replace group=1 in 366/731
zandrews jcs2 if group==1, maxlags(1) trim(0.15) break(both) graph
yields in a minimum t-statistic -3.574 at 27apr2007 (obs 117)
gen zjcs2=jcs2
replace zjcs2=. in 1/731
zandrews zjcs2, maxlags(1) break(both) graph
yields in a minimum t-statistic -3.526 at 27apr2007 (obs 117)
that is, even if obs 117 is not in range, not qulified (binary=0), or missing=. obs 117 is still considered as a structural break. how come?
am i not seeing my fault, or is something wrong here?
i am looking at credit spreads for bonds. still using stata 10. my dataset is a tsset (tsset date, d) for 1055 obs. all other ts commands seem to work fine, i.e. dfuller, dfgls, kpss, clemao/io 1 and 2.
help is very much appreciated.
best wishes marie
Send instant messages to your online friends http://uk.messenger.yahoo.com
*
* 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/
