Research On The Dynamic Characteristics And Forecast Of China's Inflation

Inflation is a core factor in the healthy and steady development of the economy, and also is an important indicator of macroeconomic policies. Low inflation rate is a main object of the government. In this thesis we studied the advanced theories and methods, applied the quantitative economioc models, analyzed the dynamic characteristics of China's inflation empirically and deeply, and established forecasting models of inflation.The thesis is seperated into 5 chapters, and the structure and contents are as following.In chapter 1 I first introduce the motivation and importance the topic. Secondly I introduce the main concepts about inflation related to this thesis, including the clasification, measurement, causes and cost of inflation.Thirdly I review the history of China's infalation and policies of the government. Finally I give the structure and main results.Chapter 2 is thChina's output gap and inflation. Output gap and inflation have very close relation, and output gap is a main explainatory variable of forecasting inflation. Output gap is very important to measure the economic situation and to make the economic policies. In Chapter 2 I estimated China's potential output and output gap from 1982 to 2007 aplying the HP filter, and I also analysize the charasteristics of the relation between output gap and inflation.Chapter 3 is the research on China's inflation and inflation uncertainty. In this Chapter I generated an indicator of measuring China's inflation uncertainty applying Generalized Autoregressive (GARCH) model, and empirically analyzed the relation between China's inflation and inflation uncertainty from January 1983 to June 2007. I also apply the granger causality test and TGARCH model to test the causality of inflation and inflation uncertainty, as well as the different effect between the affects of inflation and deflation to inflation uncertainty.Chapter 4 is the estimation and application of China's core inflation. In this Chapter I estimated China's core inflation by two models. Firstly, I extended Quah and Vahey's two-variable Structural Vector Autoregressive (SVAR) model, and establish a three-variable SVAR model consisting Consumer Price Index (CPI), food price index and output. Based on the long-run restriction we estimated China's monthly core inflation from January 2001 to June 2007. Secondly, we estimated the weights of all kinds of goods used to combine the CPI, and then excluded some goods based on the kenel density estimator. Finally I estimated China's monthly core inflation from January 2001 to November 2007 using the remainning goods.Chapter 5 is the inflation forecasting model and its application based on the P-Star indicator. In this Chapter, we first assess the limitation of money indicator and output gap, and than combine the money factor and real economic activity to establishi the P-star indicator for inflation forecasting, and then established the P-Star model. The key idea of P-Star model is that: there is a long run price level, and the real price level moves around the long run price level. Therefore, one can forecast the real price level and inflation by the long run level.The main fruits are as the following.The Phillips curve reflect the close relation between output and inflation. HP filter is a often used methods to estimate potential output and output gap. However, HP filter has a serios shortcoming that is usually ignored, that is , it is not acurrate when dealing with the data on the head and tail of a time series. In this thesis I forecast the GDP series first, and then estimate China's GDP applying the HP filter.Since the end of 1980's, the research on the speed limit effect and asymmetry of the Phillips curve are two hot points. Speed limit effect means even the real output is lower than the potential output, the reduction of the mainus output gap can still lead to inflation. The asymmetry of Phillips curve means that the plus output gap will lead to a hihger inflation compared to the minus output gap lead to deflation. The results of this thesis show that the change of output gap affects inflation, but not output itself. On the other hand, the effect of contracted policy is 6 times of expansionary policy.Inflation and inflation uncertainty is a main aspect of inflation research. It is very rare in China. The resutls of the thesis show that In China high inflation will lead to high inflation uncertainty. However, inflation uncertainty does not affect inflation rate. Also, there is no significant difference between the affects to inflation uncertainty coming from the shock of inflation and that of deflation.In China the researchers mainly apply two variable SVAR model to estimate the core inflation, and they usually use the yearly data. In this thesis I extended Quah and Vahey's two-variable Structural Vector Autoregressive (SVAR) model, and establish a three-variable SVAR model consisting Consumer Price Index (CPI), food price index and output. Based on the long-run restriction we estimated China's yearly core inflation. On the other hand I apply the item excluding method based on the kenel density estimator to estimate China's core inflation. The core CPIs estimated by the two methods are very similar.The traditional inflation forecast model employ money or output gap as main explainatory variable. However, there is no stable relation between money and inflation, we can not use money as the only indicator of inflation forecast. On the other hand the relation between money and inflation is also not stable, the slope and the intercept vary from time to time. Therefore we can not forecast inflation based on the output gap. In this thesis I proposed the P-Star indicator and established a P-Star model,. This model connects the money factor and real conomic activity,and provide more accurrate information compared with money factor and real economic activity,.The P-Star indicator is a efficient indicator of forecasting inflation.