Modeling discontinuous periodic conditional volatility: Evidence from the commodity futures market

Taylor, Nick James 2004. Modeling discontinuous periodic conditional volatility: Evidence from the commodity futures market. Journal of Futures Markets 24 (9) , pp. 805-834. 10.1002/fut.20114

Full text not available from this repository.

Abstract

This paper examines a wide variety of models that allow for complex and discontinuous periodic variation in conditional volatility. The value of these models (including augmented versions of existing models) is demonstrated with an application to high frequency commodity futures return data. Their use is necessary, in this context, because commodity futures returns exhibit discontinuous intraday and interday periodicities in conditional volatility. The former of these effects is well documented for various asset returns; however, the latter is unique amongst commodity futures returns, where contract delivery and climate are driving forces. Using six years of high-frequency cocoa futures data, the results show that these characteristics of conditional return volatility are most adequately captured by a spline-version of the periodic generalized autoregressive conditional heteroscedastic (PGARCH) model. This model also provides superior forecasts of future return volatility that are robust to variation in the loss function assumed by the user, and are shown to be beneficial to users of Value-at-Risk (VaR) models.

Item Type:	Article
Date Type:	Publication
Status:	Published
Schools:	Schools > Business (Including Economics)
Publisher:	Wiley-Blackwell
ISSN:	02707314
Last Modified:	25 Jun 2017 03:56
URI:	https://orca.cardiff.ac.uk/id/eprint/40985

Citation Data

Cited 21 times in Scopus. View in Scopus. Powered By Scopus® Data

Actions (repository staff only)

Edit Item