Are oil markets chaotic? A non-linear dynamic analysis

被引:94
|
作者
Panas, E [1 ]
Ninni, V [1 ]
机构
[1] Athens Univ Econ & Business, Athens 10434, Greece
关键词
petroleum products; correlation dimension; entropy; lyapunov exponent; non-linear dynamic; chaos;
D O I
10.1016/S0140-9883(00)00049-9
中图分类号
F [经济];
学科分类号
02 ;
摘要
The analysis of products' price behaviour continues to be an important empirical issue. This study contributes to the current literature on price dynamics of products by examining for the presence of chaos and non-linear dynamics in daily oil products for the Rotterdam and Mediterranean petroleum markets. Previous studies using only one invariant, such as the correlation dimension may not effectively determine the chaotic structure of the underlying time series. To obtain better information on the time series structure, a framework is developed, where both invariant and non-invariant quantities were also examined. In this paper various invariants for detecting a chaotic time series were analysed along with the associated Brock's theorem and Eckman-Ruelle condition, to return series for the prices of oil products. An additional non-invariant quantity, the BDS statistic, was also examined. The correlation dimension, entropies and Lyapunov exponents show strong evidence of chaos in a number of oil products considered. (C) 2000 Elsevier Science B.V. All rights reserved.
引用
收藏
页码:549 / 568
页数:20
相关论文
共 50 条