基于日涨跌率推动的股指序列建模和实证分析

Analysis and forecast of stock price based on daily rise and fall rate

为了更好地拟合原始股指序列,以信息流的驱使推进股指序列,在以交易时间或日历时间为推进进程的原始股指序列的基础上,重新构造基于日涨跌率的新的序列,并通过对其进行误差自回归分析来建立AR-GARCH模型。实证分析表明,相对于原始股指序列,用新序列预测的误差明显缩小,因此通过日涨跌率这一信息流维度变化思想重构股指序列的方法是可行、有效的。

In order to better fit the original stock index sequence,the information flow is driven as the reason for the stock index sequence advancement.Under the premise of trading time or calendar time as the original stock index sequence of the advancement process,a new sequence based on the daily rate is reconstructed,and the AR-GARCH model is established by error autoregressive analysis.Compared with the original stock index sequence,the empirical analysis shows that the error predicted by the new sequence is significantly reduced.This shows that it is feasible and effective to reconstruct the stock index sequence through the change of the information flow dimension of the daily rise and fall rate.