Study On The Prices Of Rare Earth Products In Bayan Obo Based On Time Series Model

The fluctuations of the prices of rare earth resources non-renewable affect all aspects of politics and economics,and are related not only to the safety of national mineral resources but the vital interests of all participants involved in the rare earth industry.Therefore,quantitative analysis of the fluctuations of the prices of rare earth products is of great significance.Time series analysis includes unary time series which is widely used in the price prediction of energy,stock and agricultural products because of high prediction accuracy and multivariate time series which is widely used in the research of the correlation between the prices of different products because of good results.This paper made an empirical analysis to the sample that 120 monthly prices of major rare earth oxides in bayan obo mine from January 2008 to December 2017 using time series analysis method.The research content includes the predictions of the prices of rare earth oxides and the analysis of the correlation of them.First of all,the ARIMA model was set up based on the prices of neodymium oxide,lanthanum oxide,cerium oxide,praseodymium oxide respectively,the results of tests showed that the prediction errors within the interval were small,that's to say that the established models were effective to be used for predicting the future development trends of the prices of rare earth oxides,and forecast the movements of the prices from January2018 to March 2019 with those models,the results showed that the prices of the four kinds of rare earth oxides all fall first and then rise,but the overall trend was downward,and the standard errors were small.Secondly,the co-integration theory was used to analyze the long-term interaction among the prices of four rare earth oxides: neodymium oxide,cerium oxide,lanthanum oxide,praseodymium oxide,the results showed that there is a positive correlation among the prices of neodymium oxide,cerium oxide and praseodymiumoxide,and there is a long-term negative correlation between the prices of lanthanum oxide and neodymium oxide.Based on the VEC model,the short-term dynamic relationship among the fluctuations of the prices of them was explored by using impulse response function and variance decomposition method.The results showed that the prices of neodymium oxide and lanthanum oxide are influential mutually;there is a unidirectional influence from the prices of cerium oxide to the prices of neodymium oxide;there is no significant influence between the prices of neodymium oxide and the prices of praseodymium oxide;the prices of lanthanum oxide and cerium oxide,lanthanum oxide and praseodymium oxide,cerium oxide and praseodymium oxide affect each other;the variation of each variable has different interpretations of the changes in other variables.Then,neodymium oxide and europium oxide were selected as the representatives of light rare earth and medium rare earth respectively to explore the correlation between the fluctuations of their prices.The results showed that there is no mutual influence between the two.The forecast results of the models established in this paper for the prices of rare earth oxides can be used to provide reference for the investment and business decisions of all participants because of the high accuracy.The research results of this paper for the correlation between the prices of different rare earth products are instructive for enterprises to improve the current strategy and optimize the products mix.