DCC-(R)GARCH add-in (DCC-GARCH and DCC-RGARCH model)

[MarF]

This thread is about the dcc_rgarch add-in that estimates the DCC-Range-GARCH and DCC-GARCH models.

The documentation file attached below (also available in the add-in).

[dakila]

Many thanks for providing the Eviews user community with the comprehensive add-in. Could you provide an example data?

[MarF]

Thank you for kind words.

An example workile is attached. Simply, load the workfile (data_dcc_rgarch_example.wf1) and run the add-in.
You may run GUI or the command line.


The command line example:

Code: Select all

dcc_rgarch(dep="aapl_rt amd_rt google_rt ibm_rt intel_rt", model="rgarch", dep_high="aapl_high amd_high google_high ibm_high intel_high", dep_low="aapl_low amd_low google_low ibm_low intel_low", smpl="02/01/2019 31/12/2020", show_out="yes")

This example estimates the DCC-RGARCH model for five time series representing stock data.
Notice, that the date representation is day/month/year, so it may need adjustments.

The output table for the DCC model should be like the following:
[img]

dcc_rgarch_example1.jpg (250.05 KiB) Viewed 37015 times

[/img]

[dakila]

Thanks Marcin.

[pascalle]

Thank you for providing the DCC-GARCH add-in.

May I know is there any problem if I obtained a negative significant DCC coefficient? If yes, how should I fix this problem?

Thank you

Pascalle

[MarF]

Actually, it depends. The matrix Q (the quasi-correlation) is guaranteed to be positive definite if both alpha and beta are all positive.
Notice, that negative alpha may not be and issue due to the fact that in most cases beta is quite high and alpha is low, respectively.
So depending on the situation and data it may not be an issue. The two parameters alpha and beta govern the speed of adjustment it is worth considering if there is some explanation to being alpha negative.
Hope that helps.

[pascalle]

Thank you for your reply. I am deeply appreciated with your help.
Many Thanks.

[paolo.zanghieri]

Thanks for this great add in. Is there a simple way to get the dynamic conditional betas in case of 3+ prices?

[paolo.zanghieri]

Hi, the add-in does not seem to work in Eviews 13. It starts estimating the model and then crashes. Could you have a look?
thanks

[AASA]

When trying to use the DCC two stage add-in, the Rho's are not populating. Where do I find documentation to understand if I am doing something wrong? I would also like to better understand the DCC(R) GARCH model.