05-13-2017 04:47 PM
I am running Granger causality tests using Proc Varmax for the three volatility variables pertaining to three related securities.
My objective is to determine which one leads the volatility process. (leaders and determining followers)
In this sense, in some instances, i find some two-way relationships. As said, my objective is to understand the one which determines the others. How to interpret the strength of the relationship?
a) How do you interpret the "Variance Estimate for the Innovation"? does the magnitude of this tell anything meaningful?
b) In the context of Proc Varmax Granger causality, does a higher Chi-square mean a more stronger effect?
thanks in advance for your response