The RATS Software Forum

Dear Tom,
I am estimating GARCH BEKK model with asymmetries. I know the Ds are the asymmetric shocks but I also estimated negative shocks by adding the coefficients of the As and the Ds (i.e. A+D). Am not sure if this is the right way of measuring the negative shocks of a GARCH BEKK model. I have seen some articles that used Ds as negative shocks but from the definitions, I thought the Ds are indications of possible asymmetric effects which you have to then use to estimate negative shocks. I would like to ask if what I am doing is right. If so, how will I determine the significance of such shocks. What I am doing currently is to add the coefficients of the As and Ds to obtain negative shocks, and then add their corresponding p-values or t-ratios to determine their significance.
Thank you

The D's aren't the "negative shocks"; they're the coefficients on the negative shocks. And no, you can't add A and D in a BEKK model---the overall effects can only be seen by converting to a VECH form. I'm not sure what you're trying to do. Impact calculations are the usual way to look at how the asymmetry affects the variances and covariances.

Hi Tom,
Thank you for the reply. I am estimating an asymmetric GARCH BEKK model and I intend to investigate different types of shocks (i.e. whether negative shocks have higher effects than positive shocks). So I want to estimate a negative shock and compare it's effect to the effect of a positive shock. As I am not estimating overall effects, I am not sure if converting to a VECH form will produce the negative and positive shocks. You also suggested impact calculations but am not sure how this is done and also whether this will do what I am trying to achieve.
Thank you

What do you mean by "estimate a negative shock"? You need to take the time to figure out and accurately describe what you want.

The VECH form of the BEKK is what you need to forecasts, variance IRF's or news impacts.