search

bashtage / arch

ARCH models in Python
Other
1.32k stars 246 forks source link

Why the sqrt(2/pi) factor in EGARCH? #620

Closed mikessut closed 1 year ago

mikessut commented 1 year ago

I'm not sure where to ask this question. Apologies if this is the wrong forum.

I'm trying to understand why there is a sqrt(2 / pi) factor in the EGARCH model.

I've been trying to dig through some references on EGARCH to understand it better. The brief discussion on Wikipedia doesn't show this factor nor is there a factor in this reference "Conditional Heteroskedasticity in Asset Returns: A New Approach".

https://github.com/bashtage/arch/blob/6e3325635fc0fd4cb6b7c4a305f610fc9b80a265/arch/univariate/volatility.py#L2504

mikessut commented 1 year ago

Of course I find a pretty good answer almost immediately after posting!

I think this answers it, correct?

https://quant.stackexchange.com/questions/63939/egarch1-1-mean

bashtage commented 1 year ago

That is the right reason. It makes the absolute value shock have mean 0. The other shock is already mean 0.

mikessut commented 1 year ago

Thanks!