Inflation Prediction Study Of China Based On Different Model

Since 1990s, the rate of inflation has gradually become an important index t o measure whether a country's economic development is steady or not. It is partic ularly important for the central bank to effectively control the price rising rate a nd to maintain the inflation rate. Since there is a certain degree of lag in the i mplementation of monetary policy, using of the monetary policy based on the cu rrent inflation rate can only affect the next period, which is unable to effectivel y control inflation.It is required to make a prediction of the future inflation rate. Accurately predicting the future inflation rate plays a key role in making and im plementing of the monetary policy. Through predictive inflation, central bank can predict future economy state, reduce the deviation effect of the time lag in mone tary policy, make macroeconomic regulation and control more accurate, and event ually improve the effectiveness of monetary policy.The importance of inflation prediction has attracted attention worldwide, and it also contributes to the inflation prediction fad in academia. International scholars have created a variety of prediction models for the inflation rate, among which t he data-driven model and the Traditional theoretical model are the two most co mmonly used models. Due to different development models and levels of each co untry, the prediction accuracy in different countries is different, and the relative m erits of these models have always been a controversial topic in the academia. Infla tion prediction development is still in its primary stage in China, and there has been nostudy on the comparison of the prediction effects of these two kinds of models inChina. For this, based on the experience of the foreign inflation predic tion models, this article will select data-driven models (AR model, MA model, A RMA model and VAR model) and theory-driven models (mixed Phillips curve mod el, four factors Phillips curve model and the four factors under open economy m odel), and to modify part of the models, according to China's special economic environment. And it will come to the conclusion that the prediction effects of d ata-driven models are superior to those of the theory-driven models in China, thr ough the prediction of inflation rate using the quarterly data from 1997 to 201 3, and the comparison of the predictive effects of each model according to the root mean square error RMSE. Several relevant policy recommendations are ma de in the end on how to improve the accuracy of central bank's inflation predic tion, such as improving the quality of statistical data, strengthening economic v ariables'market-oriented foundations, perfecting China's macroeconomic model, and improving monetary policy's transmission mechanism, etc.