基于ARIMA模型对上证国债指数的预测研究

Study on Prediction of T-Bond Index of Shanghai Stock Exchange by ARIMA Model

债券价格变化受很多不确定因素影响,而各个因素之间的相关关系又错综复杂,所以从理论上完全弄清楚债券的变化机理是件冗杂的事情,也非常困难。然而,债券是一个运动的、特殊的系统,也存在着可以被探索到的规律。本文以上证国债指数为例,选取2008年7月1日至2009年7月31日的日收盘价作为数据,利用Eviews软件对该数据建立自回归移动平均模型,对上证国债指数进行预测,并对预测效果给予评价,提出采用一步向前静态预测方法对国债价格进行短期预测的方法,可为国债发行主体和投资者进行相关决策时提供参考。

There are many indefinite factors to influence the price of bonds. The correlation between each factor is complicated. Theoretically, it＇s difficult to entirely find the mechanism of change in bonds. However, bond is a special, active system, and there are laws that can be explored. Take Shanghai Stock Exchange T-Bond Index as an example, selecting day＇s closing price from July 1, 2008 to July 31, 2009 and predicting T-Bond Index of Shanghai Stock Exchange by modeling ARIMA with Eviews 5.0 software. Then, forecasting the short-term trend of T-bond price by method of ＂one step static forward＂, which could be helpful for the issuers and investors to make decision.