Font Size: a A A

Smooth Breaks, Stationary And Causality

Posted on:2011-11-01Degree:MasterType:Thesis
Country:ChinaCandidate:K ShiFull Text:PDF
GTID:2189360305488813Subject:Statistics
Abstract/Summary:PDF Full Text Request
It is a controversial issue whether the output and productivity follow a trend-stationary or difference-stationary process in modern macroeconomic theory. The precise answer to this problem is significant to the analysis of the short term effect of macroeconomic policies and long-run influence of unexpected internal and external shocks.As for the modern macroeconomic theory, Unit Root Revolution extends the theory of Business Cycles. However, ignoring possible structural changes can mistake the conclusion of the commonly used unit root tests. Hundreds of literatures on the unit root test under breaks show that structural changes and unit root test entangle with each other. The answer to either of them relies on the prior determination about the other. Under subjective selection of break numbers, traditional tests do not characterize China's economy very well.First and foremost, many relative literatures concerning the structural change unit-root test are reviewed and sorted out systematically, and besides, by means of comparative research, four general misspecification problems existed in the literatures are suggested carefully.Next, in view of the difference of our economic reform from Eastern Europe, smooth breaks characterizing the gradual reform of China's economy. Thus, by means of the flexible Fourier Transformation approximating the unknown breaks, we utilize a new nonparametric stationary test to investigate the features of China's macroeconomic. The results show that half of the variables, including Total Output, Output per Capital, Bank Credits, Wage etc. are stationary. Moreover, thanks to the weighted average of local dependence, the general nonlinear Granger causality among variables is under discussion using Diks-Panchenko test. Different from other nonlinear Granger causality tests, Diks-Panchenko test amends the problem of over-rejection under the null hypothesis, thus the conclusions are more robust and dependable.Last but not least, through the perspective of Real Business Cycle, this article proposes some new suggestions about economic reform and collaborations of fiscal and monetary policies, in order to response to the Sub-prime Crisis and stabilize macroeconomic. More importantly, some enlightenments are acquired for the future research.
Keywords/Search Tags:Smooth Breaks, Nonparametric Test, KPSS Stationary Test, Nonlinear Granger Causality, Diks-Panchenko Test, Real Business Cycle
PDF Full Text Request
Related items