A note on the solution to the standard normal Stein equation

Abstract

For $z\in\mathbb{R}$ we reconsider the solution $f_z$ to the standard normal Stein equation, corresponding to the Kolmogorov distance for distributions on the real line. More precisely, we prove the Lipschitz continuity of the related function $x\mapsto ρ_z(x):=xf_z(x)$ and explain how this observation may help to remove some unnecessary technical conditions from recent Berry-Esseen bounds in the literature around the Malliavin-Stein method. We further show that the Lipschitz constants of the functions $ρ_z$ are yet not uniformly bounded in $z$ and discuss consequences of this negative result.