Testing for correlations

[philphed]

Hello everyone.

I am trying to find a model to analyze correlations. What I want to do is to analyze the increase in correlation between two (and more ) country stock indices before and after a certain event. I looked at past papers and many use Jenrich's correlation matrix stability/equality test or related/improved tests. However it seems that this is not possible in Eviews as I haven't found anything about this. So my question is, if there is anything else I could do? So far I played around with GARCH models, rolling correlation and conditional covariance. I found some papers on DCC but Eviews is not capable of this.

Looking forward to your answers!

[Oskar1996]

Hi,

I just saw your post. I have the same problem right now. Im looking for a way to test correlation changes in multivariate time series. I basically want to examine if the correlation is constant over time. I havent found a way to do it in eViews yet! Have you been able to solve the problem?

best regards,

Oskar