Suppose that I have an EGARCH(1,1) model, an APARCH(1,1) model and a GJR-GARCH(1,1) model all with GED errors. Are the following three measures for assessing the magnitude of news impact on volatility correct for all three models?
a) Intensity of Positive Shocks: α1+γ1
b) Intensity of Negative Shocks: −α1+γ1
c) Degree of Asymmetry: (α1+γ1)/α1
If you have any theoretical/empirical references on how to obtain the three measures for the above stated asymmetric models, please provide them here.