深圳综合指数长程相关性特征分析及其预测

本文从实证分析的角度讨论了深圳综合指数的长程相关性。与相同容量的随机序列相比,研究结果发现虽然深圳综合指数的收益率序列相关性较小,但收益率序列对于某种幂指数变化形式却有较强的长程相关性,收益率序列对于幂指数d的不同取值有不同强度的长程相关性,文章第二部分针对深圳综合指数的长程相关性特征提出了一种新的预测模型,该模型比传统的计量时序模型有较好的预测功能,它弥补了传统计量方法不能刻画股票指数长程相关性的不足。

The Shengzheng General index of the time-length interrelations characteristic analyzing and forecasting

This paper discusses the time-length interrelations of Shengzheng General index by analyzing cases. The research proves that the earnings rate order has much more interrelations for a profound index though the interrelations of earnings rate order of Shengzheng general index is quite small comparing with the random array of similar capacity. The earnings rate order has the different interrelations from different profound index d. Secondly, the paper puts forward a new forecasting model acording to the interrelations of Shengzheng general index, which has better calculating function than the old one. Thus this makes up the disadvantage that the old one cannot describe the time-length interrelations.