A Consistent Nonparametric Test for Granger Non-Causality Based on the Transfer Entropy

To date, testing for Granger non-causality using kernel density-based nonparametric estimates of the transfer entropy has been hindered by the intractability of the asymptotic distribution of the estimators. We overcome this by shifting from the transfer entropy to its first-order Taylor expansion n...

Full description

Saved in:
Bibliographic Details
Published in:Entropy (Basel, Switzerland) Switzerland), 2020-10, Vol.22 (10), p.1123
Main Authors: Diks, Cees, Fang, Hao
Format: Article
Language:English
Subjects:
Asymptotic properties
Causality
Entropy
financial time series
Granger causality
high frequency data
Hypotheses
Information theory
Nonparametric statistics
nonparametric test
Normality
Random variables
Taylor series
Time series
U-statistic
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:To date, testing for Granger non-causality using kernel density-based nonparametric estimates of the transfer entropy has been hindered by the intractability of the asymptotic distribution of the estimators. We overcome this by shifting from the transfer entropy to its first-order Taylor expansion near the null hypothesis, which is also non-negative and zero if and only if Granger causality is absent. The estimated Taylor expansion can be expressed in terms of a U-statistic, demonstrating asymptotic normality. After studying its size and power properties numerically, the resulting test is illustrated empirically with applications to stock indices and exchange rates.
ISSN:1099-4300
1099-4300
DOI:10.3390/e22101123