Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: ADF simulation


From   "Sascha O. Becker" <[email protected]>
To   [email protected]
Subject   st: ADF simulation
Date   Fri, 9 May 2003 11:43:02 +0200 (MEST)

Dear statalisters,

I have written some code as a basis for simulating the properties of ADF
test statistics (following the burn-in method, as Jeff Pitblado called it a day
ago) and am getting blind in understanding why I get an error message. When I
run the following code (i.e. doing a 1-replication-"simulation" by invoking
the below program just once without using the simul command), I get as error
message: 

( invalid name

When I leave out the brackets round r(Zt), I get the message

post r not found

So, it seems Stata does not recognize the expression r(Zt).
Strangely though, when I run a true simulation, i.e. I type

simul adfsim3, reps(2000)

I get no error message...

Does the post stuff only work together with simul or should it also post the
DF statistics when just invoking adfsim3 "by hand"? Even if this presumption
were true, why is the error message  "( invalid name" respectively "post r
not found" and not something else?

Thanks in advance for any suggestion you might have on this.

Ciao, Sascha



**************
BEGIN OF CODE
**************

program define adfsim3
   version 7.0
   if "`1'"=="?" {
          global S_1 "dfull"
          exit
   }
   /* args b1 b2 keep burn sigma */
   *args keep burn
   /* read in: b1    /* coefficient on time trend */
                b2    /* coefficient on AR(1) term */
                keep  /* #obs to be kept after all, i.e. length of the time
series after simulation */
                burn  /* #obs to be burnt after all */
                sigma /* standard error */ */
   local keep 50
   local burn 100
   local obs = `keep' + `burn'
   * drop all variables in memory
   drop _all
   * SET NUMBER OF OBSERVATIONS
   set obs `obs'
   * DRAW RANDOM NUMBERS FROM N(0,sigma) */
   * generate double mynormal= `sigma' * invnorm(uniform())
   generate double mynormal= se * invnorm(uniform())
   generate time=_n - `burn'
   tsset time
   * use the first entry in mynormal as a y(0) and the rest as epsilons
   generate double y=mynormal in 1 
   * NOTE: replace works recursively !!
   replace y=1+b1*time+b2*l.y+mynormal in 2/`obs' 
   * keep the last `keep' obs, i.e. the desired length of the time series 
   keep in -`keep'/l 
   * perform Dickey-Fuller: constant is used as default, otherwise option
"noconstant"
   dfuller y, lags(1) trend
   post `1' (r(Zt))
end

adfsim3

**************
END OF CODE
**************


Dr. Sascha O. Becker - Center for Economic Studies
Department of Economics - University of Munich
Schackstr. 4 - 80539 Munich, Germany

-- 
+++ GMX - Mail, Messaging & more  http://www.gmx.net +++
Bitte l�cheln! Fotogalerie online mit GMX ohne eigene Homepage!

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index