Hello. I am using the PROC VARMAX procedure to run DCC GARCH. As per userguide: https://documentation.sas.com/doc/en/etscdc/14.2/etsug/etsug_varmax_gettingstarted10.htm
Code below. I am trying to find either a table/vector output or a plot of the conditional correlation series that is generated by the process. I have tried Plots=All in the proc statement and printall in the model statement to no avail. It's not clear that ODS has this chart/table either?
Thank you
/*--- DCC ---*/ proc varmax data=indices outest=oedcc outcov; model rDJIA rSP500 / noint; garch p=1 q=1 form=dcc outht=ohdcc; run;
Hello,
What version of SAS are you using?
Are you using SAS/ETS 14.2 (your help link is for 14.2)?
After SAS/ETS 14.2 a plot of the dynamic correlations was added to the VARMAX procedure (and also you can output the dynamic correlations to a data set via an ODS OUTPUT statement).
This was still experimental in SAS/ETS 15.2 (SAS 9.4 M7) , but I guess it's no longer experimental in SAS VIYA (or even SAS 9.4 M8).
In PROC VARMAX , put plots=(CONDCORR).
produces dynamic conditional covariance plots. This option is available only when the DCC GARCH model is specified.
The ODS output dataset is :
DCCCorrConstant | Unconditional correlation matrix in the DCC GARCH model | CORRCONSTANT=EXPECT with FORM=DCC |
If you are indeed on SAS/ETS 14.2, you can reach out to SAS Technical Support in your country.
They have a program to calculate (and plot) all this starting from the (time-varying) conditional covariance matrix :
outht=ohdcc;
Cheers,
Koen
Thank you. Very good point. I'm on an older version - 9.4 M3 (I think). This code delivers an error message. Perhaps even slightly before this was even in test phase?
Hello,
I see CONDCORR is still experimental, even in the latest SAS release (which does not mean that it doesn't work properly of course).
See here :
https://go.documentation.sas.com/doc/en/pgmsascdc/v_033/etsug/etsug_varmax_syntax02.htm
So, you have 2 possibilities :
- upgrading your SAS
- reaching out to SAS Technical Support. They have a data step that calculates what you need based on outht= data set.
Good luck,
Koen
Registration is now open for SAS Innovate 2025 , our biggest and most exciting global event of the year! Join us in Orlando, FL, May 6-9.
Sign up by Dec. 31 to get the 2024 rate of just $495.
Register now!
Learn how to run multiple linear regression models with and without interactions, presented by SAS user Alex Chaplin.
Find more tutorials on the SAS Users YouTube channel.