I am a complete beginner and I am using EVIEWS,
I have monthly information on portfolio returns, as well as returns on the ftse 100, and other risk factors such as value and size.
I am wondering whether I should include these external variables in the mean equation, together with "portfolioreturns c ar(1)", or the in variance equation as variance regressors.. what method should I use to decide?
Thank you for your help
By external, I presume you mean exogenous, endogenous and predetermined variables.
The returns equation can be specified with lagged effects that include exogenous and endogenous variables, all of which would be predetermined according to the lag structure.
The inclusion of any predetermined variables in the conditional variance equation is arbitrary, but such variables are typically found to be statistically insignificant.
By external, I presume you mean exogenous, endogenous and predetermined variables.
The returns equation can be specified with lagged effects that include exogenous and endogenous variables, all of which would be predetermined according to the lag structure.
The inclusion of any predetermined variables in the conditional variance equation is arbitrary, but such variables are typically found to be statistically insignificant.
Basically, by considering the variable's significance. However, consider that not always the statistical singnificance translates into an economic significance.
best
- Badges
- Science topic
- Similar topics
- Economics