Difference of GARCH(1,1) and GARCH(1,1)-M

[Tinoosh16259]

I wish to know why one country's spillover effects to my selected dependent country is "not significant" when I am considering standard deviation in the mean equation (GARCH-M), but the same country's volatility spillover shows "significant" effects through GARCH, not considering deviation/variance in the mean equation?
Thank you for your help!
Cheers
Tinoosh

[trubador]

I am not sure about the exact specification of your model, but if you have added the volatility of foreign country as a regressor to your mean equation, then it is (probably) because there is a significant co-movement between the two volatilities (i.e. domestic and foreign). In that case, the volatility of foreign country will not have any explanatory power in the presence of the volatility of domestic country. You can try adding it as a variance regressor, but as I mentioned before, the spillover effects are generally captured within a multivariate ARCH setting.

[Tinoosh16259]

Thank you for your reply. About specification of my model, lagged squared residuals of other countries (captured by application of GARCH-M for each country separately) are added to the volatility equation of the country which is under investigation, estimated coefficients of those other countries' lagged squared residuals gives me information about significance of spillover effects from those other countries to the country which is under investigation, one at the time.

I understand, as what you mentioned earlier, that MGARCH could be more appropriate to capture spillover effects, but I am interested on analysis of unidirectional spillover effects, since I am analysing spillover effects of other countries to 'Iran', having small financial markets. Also unidirectional analysis give my model more flexibility to include structural breaks in the mean equation and exogenous variables in the variance equation. In addition, I have done some preliminary empirical work showing that risk effects are significant in the mean equation in most cases. Therefore I selected GARCH-M for my analysis? Does that sound right?

and now I am also wondering whether I should include volatility of foreign country as a regressor to my volatility equation to capture volatility spillover or I should include lagged squared residuals of foreign country as a regressor to my volatility equation? and what is the difference?
Thank you for your help again!
Cheers
Tinoosh