Why FARIMA Models are Brittle
Preprint
Abstract
The FARIMA models, which have long-range-dependence (LRD), are widely used in many areas. Through deriving a precise characterisation of the spectrum, autocovariance function, and variance time function, we show that this family is very atypical among LRD processes, being extremely close to the fractional Gaussian noise in a precise sense. Furthermore, we show that this closeness property is not robust to additive noise. We argue that the use of FARIMA, and more generally fractionally differenced time series, should be reassessed in some contexts, in particular when convergence rate under rescaling is important and noise is expected.Keywords
All Related Versions
This publication has 7 references indexed in Scilit:
- A complete asymptotic series for the autocovariance function of a long memory processJournal of Econometrics, 2008
- Long‐range Dependence: Revisiting Aggregation with WaveletsJournal of Time Series Analysis, 1998
- Robustness of whittle-type estimators for time series with long-range dependenceCommunications in Statistics. Stochastic Models, 1997
- RATE OPTIMAL SEMIPARAMETRIC ESTIMATION OF THE MEMORY PARAMETER OF THE GAUSSIAN TIME SERIES WITH LONG‐RANGE DEPENDENCEJournal of Time Series Analysis, 1997
- Stable Non-Gaussian Random Processes: Stochastic Models with Infinite Variance.Journal of the American Statistical Association, 1995
- A Handbook of Fourier Theorems1987
- AN INTRODUCTION TO LONG‐MEMORY TIME SERIES MODELS AND FRACTIONAL DIFFERENCINGJournal of Time Series Analysis, 1980