Please use this identifier to cite or link to this item: http://hdl.handle.net/1959.13/919674

OpenURL Link
Title
Parameter estimation of structural commodity price models
Author/Creator
Perez, Tristan;  Goodwin, Graham C.;  Godoy, Boris
Institution
The University of Newcastle. Faculty of Engineering & Built Environment, School of Electrical Engineering and Computer Science
Description
Commodity price modeling is normally approached in terms of structural time-series models, in which the different components (states) have a financial interpretation. The parameters of these models can be estimated using maximum likelihood. This approach results in a non-linear parameter estimation problem and thus a key issue is how to obtain reliable initial estimates. In this paper, we focus on the initial parameter estimation problem for the Schwartz-Smith two-factor model commonly used in asset valuation. We propose the use of a two-step method. The first step considers a univariate model based only on the spot price and uses a transfer function model to obtain initial estimates of the fundamental parameters. The second step uses the estimates obtained in the first step to initialize a re-parameterized state-space-innovations based estimator, which includes information related to future prices. The second step refines the estimates obtained in the first step and also gives estimates of the remaining parameters in the model. This paper is part tutorial in nature and gives an introduction to aspects of commodity price modeling and the associated parameter estimation problem.
Relation
15th IFAC Symposium on System Identification (SYSID 2009). Proceedings of the 15th IFAC Symposium on System Identification (Saint-Malo, France 6-8 July, 2009) p. 1429-1434
Publisher Link
http://dx.doi.org/10.3182/20090706-3-FR-2004.00238
Date
2009
Publisher
International Federation of Automatic Control (IFAC)
Keyword(s)
financial models;  regression analysis;  prediction error methods
Resource Type
conference paper
Identifier
http://hdl.handle.net/1959.13/919674
Identifier
ISBN:9783902661470
Identifier
ISSN:1474-6670
Reviewed
Reviewed
34 Visitors 35 Hits 0 Downloads
Save/E-mail Citation
Citation Format
E-mail Address
Subject
"Parameter estimation of structural commodity price models"
 
OR