Jonathan Shore·Aug 26, 2021Nicely done, heteroskedasticity is indeed a problem for regression.Given that you point to applications in volatility as opposed to return direction prediction, I take it that the predicted mu (mean /…A response icon1A response icon1