Premium
Assessing the value of Hermite densities for predictive distributions
Author(s) -
Mauleón Ignacio
Publication year - 2010
Publication title -
journal of forecasting
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.543
H-Index - 59
eISSN - 1099-131X
pISSN - 0277-6693
DOI - 10.1002/for.1160
Subject(s) - multivariate statistics , mathematics , hermite polynomials , nonparametric statistics , econometrics , edgeworth series , hermitian matrix , statistics , transformation (genetics) , interpretation (philosophy) , sample (material) , physics , computer science , mathematical analysis , pure mathematics , thermodynamics , chemistry , biochemistry , gene , programming language
The empirical behaviour of two Hermitian densities—the Edgeworth asymptotic expansion, and the Gallant and Nychka (1987) squared transformation—are assessed and compared to Hansen's (1994) asymmetric Student t . The comparison focuses on VaR measurements and overall probability fits. In and out of sample, Berkowitz's (2001) extended procedures are also implemented. Results are moderately favourable to the Hermitians, supporting the semi‐nonparametric interpretation, and multivariate developments. Copyright © 2009 John Wiley & Sons, Ltd.