In order to study the effect of Shanghai-Hong Kong stock connect on the efficiency of Shanghai's stock market,this paper adopts Hurst index and DFA statistics to analyse Shanghai stock index from 2000 to 2015.The results show that Shanghai stock market has an obvious long-term memory,but after Shanghai-Hong Kong stock connect,the long-term memory and short-term memory decline significantly.Based on the corresponding relations between the long-term memory characteristics and the ARFIMA model,it is found that if long-term memory is remarkable,the ARFIMA model works;vice versa.This paper also conducts length of ten steps of the fractional order auto-regressive model predicts.The results show that the ARFIMA model for the overall forecast of Shanghai's stock market yields is better,while the latter prediction of Shanghai-Hong Kong stock connect is weaker than before.%为了研究沪港通对沪市股票市场有效性的影响，本文选取2000—2015年上证股指，运用 R／S 分析并结合 DFA 统计量得出：沪市股票市场存在明显的长期记忆性，但沪港通之后其长期和短期记忆性显著下降。基于 ARFIMA 模型的预测效果与长期记忆性特征之间得出对应关系：如果长期记忆性显著，则预测效果好；如果记忆性不明显，则预测效果差。本文对沪港通前后进行长度为10步的分数阶自回归模型预测，结果显示 ARFIMA 模型对沪指收益率整体的预测效果较好，但沪港通开通之后模型的预测效果却明显减弱。
展开▼
