Timeseries - mean and variance equations

[selina gravette]

hi,
I am using a GARCH (1,1) regression model to regress stock returns against specific exogenous variables as follows:

arch returns variable1 variable2 variable2, arch(1) garch(1)

1) I then want to check the effect of these variables on VOLATILITY instead of returns. From the stock market dataset I actually have a volatility variable, but im not sure what the GARCH function of CONDITIONAL VARIANCE shows.

what I mean is, do I do the following to see the effect:

arch volatility variable1 variable2 variable2, arch(1) garch(1)

2) OR, am I meant to predict the conditional variance and then regress this against the variables? I don't really understand what the conditional variance shows and if you can then regress this against the variables to see the effect on variance. What I mean is do I use the volatility variable in data set or predict the conditional variance and then regress as follows:

arch returns variable1 variable2 variable2, arch(1) garch(1)
predict variance
arch variance variable1 variable2 variable2, arch(1) garch(1)

3) Is this how to obtain the variance equation in Stata? What is the code to regress the variance equation? I think the first regression is the mean equation, so I don't know how to use Stata to regress the variance equation. please help! thank you xxx

If this doesn't make sense do let me know , also very happy to privately message as I am very stuck on my disertation. thank you so much for your time reading this.

[selina gravette]

any help appreciated! - im asking if the conditional variance should be regressed, or my variable for volatility.

[Raph DiGi]

Hello,

Did you find any solution to your question?
I have a similar problem where I want to add some regressors in my volatility equation of a GARCH(1,1) model. It would look like this :

However, I don't know how to insert those regressors in the variance equation on Stata.
Thank you!

[Raph DiGi]

Sorry, I don't know why but the picture did not upload correctly in my last post. Here is the image