Stationary Bilinear Bayesian Model Of Inflation On China

Inflation affects every individual,industry,department,and other economic unit in the economy,which changes people's actual money income and the amount of goods and services that they can purchase.For the working classes,because of the existence of inflation,the purchasing power of each unit of currency will decline as the price rises and the standard of living will fall.For the savers,inflation will make the actual value of the deposit to devalue,which will reduce their willingness to save.Tempered inflation is benefical to investment,and the increase in investment promotes output which can improve the level of employment.In recent years,Chinese government has continuously strengthened the regulation of the macroeconomy.In particular,since China entered the new normal economy in 2012,its economic growth rate has declined slowly.Inflation has generally shown a downward trend.China's population base and the diffirence between rich and poor is relatively large.Inflation is related to the level of employment and people's living standards.Therefore,inflation issues are very important in China's macroeconomics,the study of it is quite significance.This paper systematically analyzes the dynamic characteristics of inflation in China and makes a more accurate prediction of inflation.The stationary bilinear model is a nonlinear model that introduces the cross-terms of autoregressive and moving average based on a mixed autoregressive moving average model.This model can better describe the time series process which depends on the previous impact and continues to change.Using stationary bilinear model to study inflation in China's macroeconomy,and using bayesian methods and Gibbs sampling for parameter estimation and prediction.Firstly,this article summarizes the meaning of inflation,analysis of the causes of inflation,and explain the current situation of China's inflation.Secondly,the origin and development of the bilinear model,the assumptions and evolution of the model are systematically introduced,and inflation is studied by using stationary bilinear model.Once again,estimating,inferring,and predicting the model,the prior distribution and posterior distribution are derived as well as the likelihood function and prediction function.The program is programmed by R softwareto estimation and simulation,and then the coefficient value and currency forecast value can be obtain in the stationary bilinear.The paper uses the monthly data of consume price index(CPI)for 27 years from February 1990 to February 2017 to construct and analyze the model.The research shows that: Firstly,the bilinear model is suitable for studying the time series process that is influenced by its own early time,which is highly conform with the study of inflation.Secondly,inflation in different periods in China is not independent of each other but is interreleated,we can think that inflation have inertia and continuity inflation.Inflation has a tendency to continue to circulate throughout the whole economy.Third,by comparison,it is found that the stationary bilinear model used in this paper is superior to the traditional autoregressive moving average time series model.