Journal
PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS
Volume 444, Issue -, Pages 985-995Publisher
ELSEVIER SCIENCE BV
DOI: 10.1016/j.physa.2015.10.080
Keywords
Wavelet; Time series; Co-movement
Categories
Funding
- National Natural Science Foundation of China [71173199]
- Fundamental Research Funds for the Central Universities [35732015060]
Ask authors/readers for more resources
The nonlinear relationships among variables caused by the hidden frequency information complicate the time series analysis. To shed more light on this nonlinear issue, we examine their relationships in joint time frequency domain with multivariate framework, and the analyses in the time domain and frequency domain serve as comparisons. The daily Brent oil prices, London gold fixing price and Shanghai Composite index from January 1991 to September 2014 are adopted as example. First, they have long-term cointegration relationship in time domain from holistic perspective. Second, the Granger causality tests in different frequency bands are heterogeneous. Finally, the comparison between results from wavelet coherence and multiple wavelet coherence in the joint time frequency domain indicates that in the high (1-14 days) and medium frequency (14-128 days) bands, the combination of Brent and gold prices has stronger correlation with the stock. In the low frequency band (256-512 days), year 2003 is the structure broken point before which Brent and oil are ideal choice for hedging the risk of the stock market. Thus, this paper offers more details between the Chinese stock market and the commodities markets of crude oil and gold, which suggests that the decisions for different time and frequencies should consider the corresponding benchmark information. (C) 2015 Elsevier B.V. All rights reserved.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available