Font Size: a A A

The Modeling Methods And Applications Of Panel Data Nonlinear Regression Model

Posted on:2013-02-04Degree:MasterType:Thesis
Country:ChinaCandidate:L ZhaoFull Text:PDF
GTID:2249330395484557Subject:Statistics
Abstract/Summary:PDF Full Text Request
Economics theory shows that many important macroeconomic variables and phenomenon have nonlinear character. If we analyze these problems without thinking about this, we will get wrong conclusions. So if relax the linear restriction, the results will be improved.This article first combed and summarized the OLS,2SLS,GMM and the Forward Orthogonal Deviation estimations methods of the static and dynamic panel data threshold regression model, and expounded how to use the Bootstrap method to test the significance of the threshold effect; Second, established the section data threshold Logit model, put forward the maximum likelihood estimation method (ML), structured Wald statistics to test the significance of the threshold effect, and studied the properties of Wald statistics; Again, put forward the panel data Logit threshold model and its conditions maximum likelihood estimation method (CML). At last, the paper raised panel data Markov regime switching regression model and its expectation maximization (EM algorithm) algorithm, Monte Carlo simulation method was used to confirm the robustness of EM algorithm.On the research part, mainly in the following analysis:(1) by using the dynamic panel data threshold regression model, Our results confirm that, to a certain extent, there is "double threshold effect" in the relationship between inflation and economic growth, the threshold values are3.2%and15.7%. Therefore, mild inflation rate that less than3.2%is beneficial to growth. However, inflation rate that more than3.2%, will produce "Anti-Tobin effect". Especially, the inflation hampers growth if it exceeds15.7%. Therefore, the optimal target range of the inflation rate is (0%,3.2%];(2) according to the "Fisher Effect", using the panel data threshold Logit model to reveal the relationship between the inflation rate and interest rate. The analysis shows that compared with the previous inflation rate, if the change of current inflation is less than1.16%, Central Banks tend to use interest rates instrument to fight against inflation.(3) based on the different economic situation in different economic regime, we applying the panel data Markov regime switching regression model to discuss the structure of investment growth, broad money growth and import growth and their effects on inflation, in addition,we put forward some enlightening measures to solve the inflation.
Keywords/Search Tags:Panel Data Threshold Regression Model, Panel Data Markov RegimeSwitching Model, Panel Data Threshold Logit Model, Monte Carlo Simulation, Bootstrap
PDF Full Text Request
Related items