MSGARCH
MSGARCH copied to clipboard
Regime specific standard deviation of residuals by using Volatility function
The function conditional_volatility <- Volatility(fit) extracts the conditional volatility at each point. By data/conditional_volatility I get the residuals.
The residuals have close to unit standard deviation. But when I calculate the standard deviation of the residuals in each regime, they differ greatly, e.g. with one regime at 0.7 and the other at 1.6.
I'm not sure how the function Volatility works here. I imagine that they would also have unit standard deviation respectively, since the data in each regime is filtered by a GARCH model. This post might not be so relevant here, but could anyone help explain this?
Do you have a code snippet. Not sure I get entirely your questions. If residuals of the whole process have close to unit standard deviation, this is expected and shows that the model fits well.