Limit theorems for integral functionals of Hermite-driven processes (Q2040091)

The authors consider a moving average process \(X\) of the form \(X(t) = \int_{-\infty}^t \varphi(t-u) d Z_u\), where \(Z\) is a Hermite process and \(\varphi\colon \mathbb{R}^+ \to \mathbb{R}\) is sufficiently integrable. The paper investigates the fluctuations of integral functionals of the form \(t\to \int_{0}^{Tt} P(X(s)) d s\), where \(P\) is any given polynomial function and \(T\to\infty\). Such processes appear in the context of statistical estimation. The results are split in two cases: Theorem 1 answers the case of a fractional Brownian motion as the driving process \(Z\) while Theorem 2 deals with Hermite processes of order \(q\ge 2\). In both cases, functional limit theorems are provided. The proof relies on a chaotic decomposition of powers of \(X(t)\).