Asymmetric MIDAS

For technical questions regarding estimation of single equations, systems, VARs, Factor analysis and State Space Models in EViews. General econometric questions and advice should go in the Econometric Discussions forum.

Moderators: EViews Gareth, EViews Moderator

Bosupeng
Posts: 3
Joined: Thu Oct 21, 2021 3:08 pm

Asymmetric MIDAS

Postby Bosupeng » Thu Oct 21, 2021 3:49 pm

Hello everyone I am running an asymmetric MIDAS model with a monthly dependent (y1) variable and 2 daily high frequency (pos and neg) variable (positive and negative). If i run the MIDAS model like this for 60 lags

Low freq : y1
High Frequency: page\pos page\neg

I get an "OVERFLOW" error

However if I reverse the order like this:

low freq: y1
High frequency: page\neg page\pos

The model converges easily but the slopes and the betas are almost the same with little asymmetry. Please help. I have to start with the page\pos first according to the paper that I am following.

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13307
Joined: Tue Sep 16, 2008 5:38 pm

Re: Asymmetric MIDAS

Postby EViews Gareth » Thu Oct 21, 2021 3:56 pm

Could you provide the workfile?
Follow us on Twitter @IHSEViews

Bosupeng
Posts: 3
Joined: Thu Oct 21, 2021 3:08 pm

Re: Asymmetric MIDAS

Postby Bosupeng » Thu Oct 21, 2021 4:12 pm

Thanks @Eviews Gareth.

The workfile has the equations I want to run. I tried re-running the models but I get "OVERFLOW". Beta weighting function, restricted endpoints.
Attachments
fupdatedeviewsworkfile.wf1
(871.25 KiB) Downloaded 127 times

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13307
Joined: Tue Sep 16, 2008 5:38 pm

Re: Asymmetric MIDAS

Postby EViews Gareth » Thu Oct 21, 2021 5:42 pm

Try switching the optimization method to BFGS
Follow us on Twitter @IHSEViews

Bosupeng
Posts: 3
Joined: Thu Oct 21, 2021 3:08 pm

Re: Asymmetric MIDAS

Postby Bosupeng » Thu Oct 21, 2021 5:49 pm

I did but I get very different results and not similar to the authors output.

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13307
Joined: Tue Sep 16, 2008 5:38 pm

Re: Asymmetric MIDAS

Postby EViews Gareth » Thu Oct 21, 2021 5:58 pm

Not sure I can help with that. The order of the high frequency variables shouldn't (and doesn't) matter.
Follow us on Twitter @IHSEViews


Return to “Estimation”

Who is online

Users browsing this forum: Google [Bot] and 18 guests