Font Size: a A A

Research On The Relationship Between China's Inflation Rate And Inflation Uncertainty Based On Long Memory Process

Posted on:2011-01-15Degree:MasterType:Thesis
Country:ChinaCandidate:X G ZhangFull Text:PDF
GTID:2189330338976594Subject:Quantitative Economics
Abstract/Summary:PDF Full Text Request
During the period of high inflation, the inflation expectations of the public can not be accurately formed, that may result in the impact of the inflation uncertainty. This study aims to test the relationship between China's inflation rate and inflation uncertainty as a empirical research. Indicators for measuring the rate of inflation and the inflation uncertainty are appropriately selected. Then, Granger causality test is used to determine the relationship between these two time series.In this thesis, through pre-processing of China's monthly chain CPI data from January 1990 to December 2009, the inflation rate series is obtained. After the process studies of this sequence, the property of long memory can be discovered. Thus, the long memory process ARFIMA(0, d, 0) is used for its modeling, and the phenomenon of volatility clustering can be found in the residuals of the fitting model. Then, GARCH model is applied to describe the conditional heteroskedasticity. And the conditional heteroskedasticity is more appropriate and access means as a measure of inflation uncertainty.According the estimated model of ARFIMA-GARCH, the conditional second-order moment of the inflation rate may have the property of long memory. Further, the FIGARCH model is used to describe the inflation uncertainty. And the research suggests the result show a better fit, with the assumption of the skewed-t distribution for the residual items. Conditional heteroskedasticity sequence is obtained from the model of inflation rate series, using double long memory processes of ARFIMA(0, d, 0)-FIGARCH(1, d, 1).On the studies of the relationship between the rate of inflation and inflation uncertainty, the ARCH-M model is brought to the ARFIMA(0,d,0)-FIGARCH(1, d, 1) model to determine significant factors, and the Granger causality test is done both in binary VAR model and VEC model. The research concludes that the relationship of China's inflation rate and inflation uncertainty has a two-way impact, while supporting the hypothesis of the Friedman-Ball and Cukierman-Meltzer hypothesis. Thus, it asserts requirements for the policy makers of the monetary authorities to make stable inflation expectations and to keep price stability.
Keywords/Search Tags:inflation uncertainty, long memory, ARFIMA model, FIGARCH model, Granger causality test
PDF Full Text Request
Related items