基于ARIMA模型对上证指数月度时间序列的分析和预测  

Analysis and Prediction of the Shanghai Composite Index Monthly Time Series Based on ARIMA Model

作　　者：崔远远[1] 文忠桥[1] 

机构地区：[1]安徽财经大学金融学院,安徽蚌埠233000

出　　处：《枣庄学院学报》2015年第2期102-106,共5页Journal of Zaozhuang University

基　　金：国家自然科学基金项目(项目编号:11YJA790162)

摘　　要：股票价格指数的影响因素错综复杂,现阶段影响我国股票价格的主要领域是银行储蓄、债券市场、期货市场、房地产,汇率等,从目前金融学发展的趋势和广大投资者对股票市场众多金融工具迫切的需求来看,通过建立恰当的时间序列模型可以达到对股票价格整体走势进行大致的预测的目的.本文选取了从2011年12月我国加入WTO至2014年7月以来的上证综合指数的月度数据,通过建立ARIMA模型采用一步向前静态预测的方法对我国股市2014年8月的上证综合指数进行了预测,发现我国2014年前两个季度以来整体股市呈现上升的趋势.本文的创新之处在于对样本数据取了对数,从而消除了时间序列中的自相关和异方差,同时使得预测值接近实际值,效果良好,希望对广大股民提供借鉴参考.The factors affecting the stock price index are very complicated. Currently,the main field that have impact on our country＇s stock price is bank deposit,bond market,futures market,real estate,exchange rate and so on. From the current development trend of finance and the urge needs of the many financial instruments of vast number of investors on through the establishment of time series model can achieve appropriate forecast of the purpose of the overall trend of stock price. This paper selected the Shanghai Composite monthly data from 2011 December when China joined the WTO to 2014 July. Through the establishment of ARIMA model by using the method of one step forward static prediction predicting the value of 2014 August China＇s stock market and finding the first two quarters over of stock market in China showed a rising trend. The innovation of this paper is to take the logarithm of sample data so that the time sequence of autocorrelation and heteroskedasticity can be eliminated. Meanwhile,China＇s stock market composite index was close to the actual value,which is a good result. Hope to provide reference for investors.

关 键 词：ARIMA(自回归单整移动平均模型) 上证综合指数 一步向前静态预测 B-J方法论1 

分 类 号：F832.5[经济管理—金融学]