Hi Bruce, thanks!

I think that’s a case where you’d want to model variance directly (e.g. using GARCH). I touch a bit on that topic in the article, though with little detail.

Maybe I’ll dive a bit into that in a future article.

Cheers