Specification procedures for multivariate stable-Paretian laws for independent and for conditionally heteroskedastic data
Statistics Theory
2023-12-20 v1 Statistics Theory
Abstract
We consider goodness-of-fit methods for multivariate symmetric and asymmetric stable Paretian random vectors in arbitrary dimension. The methods are based on the empirical characteristic function and are implemented both in the i.i.d. context as well as for innovations in GARCH models. Asymptotic properties of the proposed procedures are discussed, while the finite-sample properties are illustrated by means of an extensive Monte Carlo study. The procedures are also applied to real data from the financial markets.
Keywords
Cite
@article{arxiv.2310.13445,
title = {Specification procedures for multivariate stable-Paretian laws for independent and for conditionally heteroskedastic data},
author = {Simos G. Meintanis and John P. Nolan and Charl Pretorius},
journal= {arXiv preprint arXiv:2310.13445},
year = {2023}
}