The Nonparametric Test Of The Fisher Effect In China

The interest rate and inflation rate or the stock return rate and inflation rate are the important variables in the macroeconomics and financial economics. And it is meaningful to the manipulation of monetary policy and perfection of financial market to analyze the economic behavior of the variables and the interrelationship between them. The Fisher hypothesis was considered as one of the basic hypotheses in economics and accepted by many models in the macroeconomics and financial economics since the Fisher Effect was put forward. Many researchers tried to explain and test the Fisher Effect from the angle of theory and practice. So, the Fisher Effect is the hot topic in the field of economics. However, it is inconsistent between theory researches and practice testings; it is so called"the puzzle of the Fisher Effect". There were many literatures on how to test and explain"the puzzle of the Fisher Effect", but they weren't reached a consensus. Nonetheless, the research of Fisher Effect is irreplaceable for the financial market and monetary policy. How the interest rate and the stock return rate change accordingly and whether the Fisher Effect exists or not when inflation rate changes? That is the main purpose of the dissertation.The dissertation try to find whether there exist the long-run adjustment mechanism between the interest rate and inflation rate or the stock return rate and inflation rate by nonparametric unit root and cointegration tests based on the characteristics of nonlinear evolution between these variables. Moreover, we estimate the Fisher Effect coefficients by nonparametric local linear estimation with variant bandwidth (LLEVB). The conclusion is as follows:Firstly, based on nonparametric cointegration test, we can get that: (1) There exits the long run nonlinear cointegration link between nominal interest rate and inflation rate in China. We find our conclusions are better because that nonparametric test has better small sample properties, namely, higher test power and smaller size distortion and didn't overlook the nonlinear factors. In fact, Central Bank have raised the interest rate many times and changed from the prudent monetary policy to the moderately tight monetary policy in order to tighten the surplus liquidity and stabilize the expectation of inflation. The effect of the monetary policy has been appearance and the growth rate of inflation is slower that means the monetary policy of Central Bank is reasonable. It is as same as the test conclusion of the dissertation. In contract, the result of the Johansen cointegration test shows that there don't exist the long run cointegration link and then interest rate has no impact on the inflation rate. It diverges from the practice and we believe that it is wrong from either theory or practice. (2) There have the long run nonlinear cointegration link between the stock return rate and the inflation rate in China. Similarly, the result of the Johansen cointegration test based on linear parametric model shows that there doesn't exist the long run cointegration link between them and is wrong from either theory or practice. In practice, many factors, such as inflation rate, rate policy, output level and business cycle etc. have impact on the stock return rate. Thus, there don't exit a simple linear interrelationship but the complicated nonlinear impact mechanism between the stock return rate and the inflation rate. According to the above analysis, the dissertation find that the results get from nonparametric cointegration test are more reasonable and suitable to the nonlinear characteristics which as same as the data of stock return rate and inflation rate.Secondly, we make the conclusion by nonparametric local linear estimation with variant bandwidth (LLEVB) are as follows:(1)The Fisher Effect coefficients are all positive and less than 1 means that there only exits the weak Fisher Effect between nominal interest rate and inflation rate, in other words, the adjustment of interest has partly impact on the inflation or the inflation expectation. On average, the Fisher Effect coefficient is 0.4055 which means that the timing of Central Bank raised the interest rate to stabilize the expectation of inflation is properate and should have some effects. Furthermore, the monetary policy should turn to use the interest rate adjustment and promote the healthy development of Chinese economy(.2)The Fisher Effect coefficients estimated based on the sample of nominal stock return rate and inflation rate are different in each period and positive alternate with negative. It means that the changes of nominal stock return rate and the changes of inflation rate sometimes are co-directional alternate with opposite directional. There exists the complicated nonlinear adjustment mechanism that confirms the conclusion of the nonparametric cointegration test. From the definition and estimating result of the Fisher Effect in stock market, we find that there don't exist the Fisher Effect in Chinese stock market. Moreover, we apply the nonparametric estimation on the real stock return rate and inflation rate, and get the coefficients of the Fisher Effect are mainly negative, sometime positive and the coefficient exclude from the Fisher Effect are zero. It confirm that there don't exist the Fisher Effect in Chinese stock market again which consensus with most of the literatures on this field.From the above analysis, we make the conclusion are as follows: there exists nonlinear adjustment mechanism between nominal interest rate and inflation rate in China and exists the weak Fisher Effect. The monetary policy, such as raising the interest rate can stabilize the inflation expectation. However, there exists nonlinear adjustment mechanism between stock return rate and inflation rate in Chinese stock market, but don't exist the Fisher Effect, in other words, it still exists"the stock return-inflation puzzle"...