RELATIONSHIP BETWEEN GRAIN PRICES AND INFLATION IN CHINA (1996-2008)——Nonlinear Causality Testing Method Based on Correlation Integral

ZHU Xin-kai,LU Jie

DOI: https://doi.org/10.3969/j.issn.1000-596x.2011.03.003

2011-01-01

Abstract:This study introduced and developed a Monte-Carlo nonlinear causality testing method based on correlation integral.This method does not use any presumed linear relations or any types of functions and it can check both causal intensity and time effect.In this paper,we investigated the model precision by using Monte-Carlo simulation for the first time in the world and defined a new concept – intensity of nonlinear causality.By using this model,we analyzed the nonlinear relations between the grain price and CPI in China,which showed that there existed bi-directional causality relations,but the time effect and the causal intensity are different.CPI has obvious causality effect on the grain price,and the effect lasted from lag=3 months until lag=7 months and the lag=3 months one has the strongest effect.On the other hand,the grain price also has causal effect on CPI,but only lasts for one month(lag=1month).Based on the theoretical results,we made suggestions on nation policies.