Comparative Study On The Quantitative Method Of Inflation Expectation And Rational Test Of Inflation Expectation

The current research on consumer inflation expectations are mainly about measuring inflation expectations, the nature of inflation expectations, the relationship between inflation expectations and the actual inflation and inflation expectations influence. But there is little literature research on comparison quantitative method of inflation expectations, to discuss the estimation of expected inflation rate can truly represent the level of inflation expectations of consumers. The traditional method to test whether inflation expectations rational is considering three necessary but not sufficient conditions, namely whether the mean of expected error is zero, whether or not the expected error is related; whether the expected error is unbiased and the strong or weak effectiveness of expected inflation. But there is a flaw, even if our test results show satisfying the above three conditions, we also cannot say that inflation expectations are rational. The definition of rational expectations is that consumers can make full use of all the information available to expect future, and do not make systematic mistakes. But my understanding is that rational expectations do not mean the full utilization of all obtainable information, but consider of the marginal cost and marginal benefit of information to neglect a part of information. That is to say that rational expectations means considering the information to satisfy the marginal benefit is bigger than or equal to the marginal cost.Firstly from the angle of feasibility and micro foundation of expected inflation, we analysis the method of quantification for inflation expectations which are current used, such as futures market price method, yield spread method, macroeconomic model method and survey data method. Then introduce the basic idea、 computing process and calculation of balance statistic method、C-P probability method and time-varying parameter method which are based on survey data. From the statistical characteristics——average error ME、absolute error MAE、mean square error RMSE and standard deviation, and new Keynes Phillips curve model to compare the expected rate of inflation based on three methods above. Finally, this paper puts forward a new method to test if inflation expectations are rational. If inflation expectations are better than inflation forecast in the three necessary conditions for the traditional rational expectations test, we can say that inflation expectations are rational enough.Through the above research, we can get the following conclusions:there is no significant difference between balance statistic method、C-P probability method and time-varying parameter method in the estimation of expected inflation rate. All can represent true inflation expectations of consumers in china. For simplicity’s estimate, we can easily take the simplest balance statistic method to estimate the expected rate of inflation in the future research. Inflation expectations and inflation forecast are all unbiased and strong effectiveness. The deviation of inflation expectations satisfying the mean value is zero and no auto correlation, while deviation of inflation forecast has self correlation. From the point of non selfcorrelation, inflation expectations better than inflation forecast. So we can think that consumers in our country are rational. The output gap of the NKPC regression model failed to pass t test. Therefore, China’s central bank cannot manage inflation by controlling the output gap.