Measuring Nonlinear Granger Causality in Mean

Abstract

The full-text may be used and/or reproduced, and given to third parties in any format or medium, without prior permission or charge, for personal research or study, educational, or not-for-prot purposes provided that:• a full bibliographic reference is made to the original source • a link is made to the metadata record in DRO • the full-text is not changed in any way The full-text must not be sold in any format or medium without the formal permission of the copyright holders.Please consult the full DRO policy for further details. Measuring Nonlinear Granger Causality in Mean ABSTRACTWe propose model-free measures for Granger causality in mean between random variables. Unlike the existing measures, ours are able to detect and quantify nonlinear causal effects. The new measures are based on nonparametric regressions and defined as logarithmic functions of restricted and unrestricted mean square forecast errors. They are easily and consistently estimated by replacing the unknown mean square forecast errors by their nonparametric kernel estimates. We derive the asymptotic normality of nonparametric estimator of causality measures, which we use to build tests for their statistical significance. We establish the validity of smoothed local bootstrap that one can use in finite sample settings to perform statistical tests.Monte Carlo simulations reveal that the proposed test has good finite sample size and power properties for a variety of data-generating processes and different sample sizes.Finally, the empirical importance of measuring nonlinear causality in mean is also illustrated. We quantify the degree of nonlinear predictability of equity risk premium using variance risk premium. Our empirical results show that the variance risk premium is a very good predictor of risk premium at horizons less than six months. We also find that there is a high degree of predictability at horizon one-month which can be attributed to a nonlinear causal effect.

A Better Understanding of Granger Causality Analysis: A Big Data Environment

“…Theorem 3 : Let Assumption A holds. Under the null hypotheses ( 22), (26), and ( 27), respectively, if √ T /N → 0, then we have…”

“…The literature on Granger causality analysis is extensive and many tests and measures have been introduced to detect and quantify both linear and non-linear Granger causality; for review see Dufour and Taamouti (2010), Bouezmarni et al (2012), and Song and Taamouti (2018). The original definition of Granger (1969) that have been adopted in this literature implicitly assumes that all the relevant information is available and used for the causality analysis.…”

Measuring Granger Causality in Quantiles

“…However, once it has been established that a "causal relationship"exists, it is usually important to assess its strength. Few studies have proposed to measure Granger causality in mean between the variables of interest; refer to Geweke (1982Geweke ( , 1984, Dufour and Taamouti (2010), and Song and Taamouti (2018). However, these studies ignore or pay less attention to the causality that can occur at other levels or aspects of the conditional distribution, such as conditional quantiles.…”

“…Polasek (1994Polasek ( , 2002 show how causality measures can be computed using Akaike Information Criterion and a Bayesian approach. Song and Taamouti (2018) propose nonparametric measures of Granger causality in mean, extending the parametric approach of Geweke (1982Geweke ( , 1984 and Dufour and Taamouti (2010). Taamouti et al (2014) propose a nonparametric estimator and test for measures of Granger causality in distribution.…”

A Better Understanding of Granger Causality Analysis: A Big Data Environment

“…The literature on Granger causality analysis is extensive and many tests and measures have been introduced to detect and quantify both linear and nonlinear Granger causality; for review see Dufour and Taamouti (2010), Bouezmarni, Rombouts and Taamouti (2012), and Song and Taamouti (2018). The original definition of Granger (1969) that have been adopted in this literature implicitly assumes that all the relevant information is available and used for the causality analysis.…”