中国粮食价格与CPI的关系(1996—2008)——基于非线性关联积分的因果检验

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

当前,中国特色的市场化道路步入了发展的快车道,粮食市场与宏观经济的相互关系也表现得更为复杂。本文尝试引入一种不基于任何函数关系假定前提的非线性检验模型,对1996—2008年间中国粮食价格与CPI的关系进行检验和解释,以期更为准确地理解和阐释两者之间的关系,为制定宏观经济政策,实现国家宏观调控目标,提供可资借鉴与参考的理论解释。本文的检验结果表明,中国粮食价格和CPI之间存在双向因果关系,但在时效和强度上存在较大差异。CPI对粮食价格有半年左右的显著影响,粮食价格对CPI仅表现为滞后1个月的短期效应。最后在给出理论解释的基础上,本文提出了相关政策建议。

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.