文档介绍：华中科技大学
硕士学位论文
我国股票市场分整模型的实证研究
姓名:汪力
申请学位级别:硕士
专业:数量经济学
指导教师:刘凯
20041107
摘要
经过十几年的发展我国的股票市场经历了一个从不成熟到逐渐走向成熟的过
程我国股票市场收益率的变化有什么特点和国外的成熟市场有什么区别是本
文研究的主要问题
本文主要运用ARMA类模型和 ARCH类模型对上证综指和深证成指的日收益率
进行分析首先介绍了我国股票市场的发展历程定性的分析了日收益率的性质
认为应选取 1998 年之后的数据进行定量分析然后介绍了几种常用的 ARMA 模型
和 ARCH 模型其次重点介绍了近几年刚刚发展的分整模型即 ARFIMA 模型和
FIGARCH 模型并对其进行比较分析指出这些模型的不足和适用范围最后运用
上述模型对上证综指和深证成指的日收益率进行实证研究得出的主要结论如下
1)我国股票市场还没有达到弱型有效市场我
忆性和长期记忆性股票市场的历史信息将对股市收益产生一定的影响对投资者
来说可以利用技术分析获得超额利润
2)我国股票市场价格的波动呈现集群性持续性过去的波动扰动对市场未来波
动有着正向而减缓的影响较大幅度的波动后面一般紧接着较大幅度的波动较小
幅度的波动后面一般紧接着较小幅度的波动并且波动扰动影响时间长以双曲线
形式递减
3)我国股票市场存在明显的杠杆效应股市对正的和负的股票收益率产生不对称
的反应股市中负价格变动比相同规模的正价格变动导致更高的波动性
关键词分整模型上证综指深证成指
IV
ABSTRACT
After more than ten years’ development, the stock market of China experiences a
process from juvenility gradually to maturity. The characteristic of the returns of Chinese
stock market and the difference between Chinese stock market and foreign mature market
are the main problems this thesis investigates.
ARMA-type models and ARCH-type models are used to analyze the daily returns of
shangzhengzongzhi and shenzhengchengzhi. Firstly, the developmental process of
Chinese stock market is introduced and the conclusion that the data after 1998 should be
used in quantitative analysis was drawn after qualitative analysis of the daily returns.
Secondly, mon ARMA models and ARCH models are introduced in detail,
and fractionally integrated models, such as ARFIMA and FIGARCH, which are developed
in recent years, are introduced, too. Finally, above-mentioned models are applied to stock
index return data, and follow conclusions can be drawn:
Chinese stock market has not achieved weak-form efficient market.
Volatility clustering and persistence in variance are found in our stock market.
There is distinct lever effect in our stock market.
Key words: fractionally integrated model ARFIMA FIGARCH
V
独创性声明
本人声明所呈交的学位论文是我个人在导师指导下进行