CPI-PPI Granger Causal Dynamics In China

As two important indicators of inflation,the consumer price index(CPI)and the producer price index(PPI)have been the focus of researches in the economic field and provide important economic basis for government policy formulation.The existing theoretical studies generally suggest that there exists complex interactive relationships and transmission channels between CPI and PPI.Most of the researches can prove PPI as a leading indicator of the CPI,but the transmission from CPI to PPI remains unclear.This paper studied the time-varying Granger casual dynamics between CPI and PPI by linear and nonlinear models,which is helpful to explain the macroeconomic phenomenon and help to achieve the goal of price stability.Firstly preliminary investigations of the time-varying casual dynamics between CPI and PPI by using linear vector autoregressive model(VAR),Co-integration theory,Granger causality test and Rolling-window Granger causality test were made.The empirical results of linear vector autoregressive model(VAR)test showed that PPI is the one-way Granger cause of CPI.However,the Granger causality test assumes that the VAR model's parameters are constant while so that such test cannot model the relationship among economic variables well as there could be structural breaks in reality.Moreover,the non-linear information test for the residuals of the linear model also indicated that there exist nonlinear features of CPI and PPI.Given the facts above,the rolling window model was employed to mitigate the structural instability of the parameters.Its results showed that the causality relationship between CPI and PPI is time-varying.In order to model such relationship between CPI and PPI,the Markov-regime switching vector autoregressive model(MS-VAR),the Markov-regime switching vector error correction model(MS-VECM)and the Markov-regime switching vector error correction model with monetary factors were employed.The state variables of these models,reflecting the variation characteristics between different regimes,allowed different parameters under different states.And the transition probabilities were used to capture the dynamic causality relationship between CPI and PPI.The empirical results indicated that there is a time-varying causal relationship with causality running in both directions between CPI and PPI in China,which suggested that the changing of supply factors and the feedback of demand factors jointly influence China's inflation level at the same time.