基于时间序列模型的上海证券融券余额分析与预测

Analysis and Prediction of Shanghai Securities Margin Balance Based on Time Series Model

针对上海证券融资融券当日融资余额的分析与预测问题,建立ARMA自回归移动模型,搜集从2014-08-21到2015-06-16的数据,对其进行编号处理,运用Eviews软件,得到平稳序列,观察自相关、偏自相关系数及其图形,利用最小二乘法及信息准则,建立合适的ARMA模型,并充分提取序列的相关信息,确定模型阶数,求出模型的表达式,并利用ADF等多种检验方法检验模型是否正确,最终利用此模型求出未来5d的预测值。并与真实值进行比较,检验预测的准确性。

Aiming at the analysis and prediction problems of Shanghai securities margin financing bal-ance that day,we established ARMA auto regressive model.By using Eviews software for numbered data from 8/21/2014 to 6/16/2015,the stationary sequence was obtained,then appropriate ARMA model was established with observing autocorrelation coefficient, partial autocorrelation coefficient and graphics. The order and model expression were gotten through extracting related information of sequence and tested with ADF test methods.At last the forecast for the next 5 days was given and compared with real value to test the accuracy of prediction.