Premium
Modelling U.S. monthly inflation in terms of a jointly seasonal and non‐seasonal long memory process
Author(s) -
GilAlana L. A.
Publication year - 2005
Publication title -
applied stochastic models in business and industry
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.413
H-Index - 40
eISSN - 1526-4025
pISSN - 1524-1904
DOI - 10.1002/asmb.552
Subject(s) - long memory , inflation (cosmology) , seasonal adjustment , econometrics , series (stratigraphy) , seasonality , memory test , zero (linguistics) , economics , mathematics , statistics , geology , psychology , physics , mathematical analysis , philosophy , volatility (finance) , paleontology , cognition , linguistics , variable (mathematics) , neuroscience , theoretical physics
This article examines monthly U.S. inflation rates by means of fractionally integrated time series techniques. We use a procedure of Robinson ( J Am Statist Assoc 1994; 89 :1420) that permits us to simultaneously test the degrees of integration at both the zero and the seasonal frequencies. The results show that long memory takes place at both frequencies, the orders of integration being in both cases higher than 0 but smaller than 0.5, implying stationarity but long memory behaviour. In addition, the root at the long run or zero frequency seems to play a much more important role than the seasonal one, with shocks affecting the latter component decaying faster than in the former case. Copyright © 2005 John Wiley & Sons, Ltd.