standard errors in tsayp591

Discussion of models with structural breaks or endogenous switching.

standard errors in tsayp591

Postby wardb » Tue May 20, 2008 12:04 am

I used the tsayp591.prg to reproduce the results for Markov switching GARCH example from Tsay's Analysis of Financial Time Series, Example 12.6, but cannot obtain the estimated standard errors of the estimated parameters for the two regimes and their differences. I'd be most grateful if anyone could assist me with code. Thanks in advance.
wardb
 
Posts: 15
Joined: Fri Nov 02, 2007 3:46 pm

Postby TomDoan » Tue May 20, 2008 11:12 am

Just add a STATISTICS instruction inside that final loop. That will compute the mean and variance of the MCMC draws for each of the parameters.

Code: Select all
do i=1,10
   stats stats(i) 1 nkeep
   density(type=histogram,counts,maxgrid=12) stats(i) 1 nkeep xx fx
   scatter(style=bar,hlabel=vlabels(i))
   # xx fx
end do i
TomDoan
 
Posts: 2720
Joined: Wed Nov 01, 2006 5:36 pm

tsayp591

Postby wardb » Thu May 22, 2008 3:11 pm

Thanks for the extra code, which was most helpful.
wardb
 
Posts: 15
Joined: Fri Nov 02, 2007 3:46 pm


Return to Structural Breaks and Switching Models

Who is online

Users browsing this forum: No registered users and 1 guest

cron