| The relationship between money supply and prices have been a hot research fieldof macroeconomic, well understanding of their relationship is beneficial to bothselection and implementation of macroeconomic policies. Especially in recent years,the relationship between changes in the money supply and price fluctuations has beensomewhat uncertainty, considering the time variability of their relationship, it ismeaningful to re-examine this relationship.Consumer price index is generally used as a measure of the price level; however,considering the limitations of CPI to portray the overall price level, we decided to useStock and Watson’s dynamic factor model to extract the common component of CPI,PPI and RPI, which can characterize the fluctuation of the price level. We found thatthe price index extracted from the model can reflect the overall trend of pricefluctuations in the sectors of production, consumption and retail. In the later section,we will use this price index as the proxy variable of overall level of prices in China tostudy the dynamic relationship between the money supply and price fluctuations.Continuous adjustment of the economic structure might cause some impact in thestructural relationship between money supply and price fluctuations, so we choose theform of rolling correlation coefficient and its expansion. According to the study, wefound that the structure of the associations between variables (M1, M2, M1/M2) andprice index have been changing. Comparing to other variables, narrow money supplykeeps higher relationship with price index. By the rolling form of Granger causalitytest, we found, the Granger causality of M1and price index is more obvious, M2Granger cause price fluctuations while is seems that price changes does not Grangercause M2.Both the correlation analysis and the causality test were preliminary analysis ofthe two variables. In fact, money supply, price, the domestic output gap and other variables affect each other which need to be considered. Taking into account thetime-varying relationships between variables, we choose VAR model withtime-varying coefficients to study the dynamic relationship of narrow money supplyand the price index. And this time, we introduced the output gap and interest rates intothe model in order to avoid their impact on empirical results. By comparing the valueof the impulse response of different points in time, we found that there exists apparenttime-varying relation between money supply and price fluctuations, significantlyweakening trend in the response of price level to money shocks has been found, whichindices greater difficulty when central bank controlling inflation with money policies.Comparing with previous studies, here we have a new definition and understanding ofimpulse response function in time-varying VAR model.It is more sensitive than thetraditional one when capturing the changes of the responses. By comparison of thesetwo types of the impulse response, we found that the time-varying structure is themost important factor when explaining the fading response of one variable to the other.But this time-varying impulse response function will change into powerless withrecent characterization of the impulse response, while the traditional one can.With the accelerated process of globalization and the expanding scale of China’sforeign trade, foreign inflation is likely to affect domestic price level through theimport and export trade, increasing foreign exchange will more or less affect theindependence of money policy. Based on this, we used VAR model with time-varyingcoefficients to discuss the impact of changes in foreign demand shocks, weconstructed a proxy for foreign demand shocks according to the imports scale ofseveral representative nations and weighted their output gaps. With the time-varyingimpulse response values of narrow money supply and price index, money supplyshows stronger responses to the shocks of foreign demand, and this implies that, underthe current exchange rate regime, foreign demand shocks significantly weaken theindependence of monetary policy. |