Show simple item record

AuthorCharfeddine, Lanouar
Available date2021-04-29T08:10:44Z
Publication Date2016
Publication NameEconomic Modelling
ResourceScopus
ISSN2649993
URIhttp://dx.doi.org/10.1016/j.econmod.2015.12.009
URIhttp://hdl.handle.net/10576/18384
AbstractThe aim of this paper is to propose an empirical strategy that allows the discrimination between true and spurious long memory behaviors. That strategy is based on the comparison between the estimated long memory parameter before and after filtering out the breaks. To date the breaks, we use the probability smoothing of the Markov Switching GARCH model of Haas et al. (2004). Application of this strategy to the crude oil, heating oil, RBOB regular gasoline and the propane futures energy with the one, two, three and four months maturities show strong evidence for the presence of long range dependence in all futures energy prices volatility. 11As proxy of volatility, we use the squared returns. Others proxies are used in the empirical studies such as the absolute value of returns. Our empirical studies show there are strong differences between these two proxies. time series. This result of long range dependence in the volatility is confirmed by the superiority of the FIGARCH and FIEGARCH models compared with the Markov switching GARCH models in terms of out-of-sample forecasting and value at risk (VaR) performances. Moreover, we show that the proposed empirical strategy is robust to different data frequency. Practical implications of the results for market participants are proposed and discussed.
Languageen
PublisherElsevier B.V.
SubjectEnergy futures time series
Fractional GARCH-class of models
Long range dependence
Markov switching GARCH model
Out-of-sample forecasting and VaR
TitleBreaks or long range dependence in the energy futures volatility: Out-of-sample forecasting and VaR analysis
TypeArticle
Pagination354-374
Volume Number53
dc.accessType Abstract Only


Files in this item

FilesSizeFormatView

There are no files associated with this item.

This item appears in the following Collection(s)

Show simple item record