文档介绍：华中科技大学
硕士学位论文
FIGARCH模型及其对中忆性研究
姓名:黄永强
申请学位级别:硕士
专业:数量经济学
指导教师:王少平
20060509
摘要
金融波动的持续性是金融市场中的一个重要现象之一,这种持续性是否具有长
记忆性对投资组合的选择,资产风险的管理等诸多问题有很大帮助。在有效市场假
说的框架下,信息对市场波动的影响都会通过市场完善而迅速的反馈机制及时地反
映到价格中去,也就是说当前信息对未来影响很小。然而,诸多的事实说明并非如
此,往往金融市场的波动具有持续性并且还具有长记忆性。这种记忆性的存在与否
会对金融市场产生重大的影响。本文运用长记忆性的 FIGARCH 模型对中忆性进行实证研究。在实证分析中,首先对我国上证和深证股市日收盘价
格的对数收益率基本统计分析,结果表明其具有尖峰肥尾特征,初步断定具有 ARCH
效应,进而再进行了 ARCH 效应和长记忆性的检验,最后进行 FIGARCH 模型估计。通
过这些分析和检验,其结果表明我忆性。

关键词:分整的广义自回归条件异方差模型长记忆 KPSS 检验拟极大似
然估计

I
Abstract
The persistence in financial volatility is one of the important phenomena,whether
the persistence remains long-term memory or not will be helpful to the choice of
portfolio,the management of asset risk,and so the frame of the Efficient Market
Hypothesis, the influence of innations to the market volatility will always reflect on the
market price in time through the feedback mechanism of the market which is perfect and
quick,in another word ,the innations at present will poorly influence on the
,many facts about them were not really so ,they often showed persistence
and long-term the financial volatility measures the risk ,whether the
market have long-term memory or not will influence on it greatly. The article uses
FIGARCH model which shows long-memory dependence for testing the persistence and
long-term memory on China Stock the substantial evidence and analysis,
first,we analyze the basic statistics characteristic about the logarithms rate of return of
stock market day closing price,and it show a peak and fatty tail,and it is considered that
there exists an ARCH effect,then we do the ARCH effect test and the long-term memory
test, and carry on the estimate of the FIGARCH a result,they suggest that there
exists the long-term memory in the stock market of our country.

Keywords:FIGARCH Long-term Memory KPSS test quasi-maximum