Partial measures of time-series interdependence - Collections

＜technical report＞
Partial measures of time-series interdependence

Creator	Creator Name Hosoya, Yuzo 細谷, 雄三ホソヤ, ユウゾウ
Creator	Author PID K002710 Creator Name Takimoto, Taro 瀧本, 太郎タキモト, タロウ Affiliation Affiliation Name 九州大学大学院経済学研究院経済工学部門
Language	English
Publisher	Faculty of Economics, Kyushu University
Publisher	九州大学大学院経済学研究院
Date	2013-11
Source Title	Faculty of Economics Discussion Paper
Publication Type	Accepted Manuscript
Access Rights	open access
Related DOI	Discussion Paper ; 2013-09
Related DOI	http://www.econ.kyushu-u.ac.jp/
Related URI	http://www.econ.kyushu-u.ac.jp/
Relation	Discussion Paper ; 2013-09
Relation	http://www.econ.kyushu-u.ac.jp/
Abstract	The paper deals with a statistical analysis aiming at quantitative characterization in the frequency domain of the strength of one-way effects and reciprocity between a pair of series in the presence ...of a third series, suggesting a unified frequency-domain method of statistical estimation and testing for the proposed partial measures of interdependency. In particular, the paper provides an estimation procedure of those measures based on a numerical canonical factorizaation method of spectral densities and proposes Monte Carlo Wald tests for those measures. The method is applicable to data generated by the stationary multivariate ARMA process.show more

File	FileType	Size	Views	Description
DP2013-09	pdf	2.12 MB	524

Record ID	1397625
Peer-Reviewed	Unrefereed
Subject Terms	Canonical factorization
	Granger non-causality
	Measure of one-way effect
	Measure of reciprocity
	Partial interdependency
	Prediction error
	Vector ARMA model
	Wald statistics
Notes	The research is partially supported by JSPS Grant-in-Aid for Scientific Research (C) 22530211and JSPS Grant-in-Aid for Young Scientists (B) 22730180, respectively.
Notes	JEL classification: C12, C13, C32, C53
Type	Discussion Paper
Created Date	2013.12.11
Modified Date	2021.12.13