摘要
首次引入广义混合分布假说(MDH)理论,并利用中国股票市场数据检验其是否能够解释日收益波动与交易量的动态关系。结论显示:对数交易量的非预期成分是日信息流很好的工具变量;日收益波动序列可以分解为预期成分和非预期成分,非预期成分是由日信息流对市场的冲击产生的,预期成分主要由滞后的收益冲击所趋动;日收益波动包含传统GARCH模型不能反应的很大随机成分。
A generalized mixture-of-distributions hypothesis (GMDH) is introduced in this paper and is examined to interpret the relation between price volatility and trading volume. The conclusion shows that the unexpected component of daily log volume is a good instrument for the daily information flow;the variance of daily returns is decomposed into two distinct terms;an expected component,which reflects the lagged return shocks, and an unexpected component that relate to the daily information flow.Our analysis indicates the daily volatility contains a large stochastic component that is missed by conventional GARCH models. The findings in this paper have significant implications for research in estimating the dynamics of daily volatility.
出处
《现代财经(天津财经大学学报)》
CSSCI
2004年第9期38-42,46,共6页
Modern Finance and Economics:Journal of Tianjin University of Finance and Economics
基金
河北省教育厅人文社会科学研究计划(SO3206)
2003年度河北经贸大学校级青年项目。
关键词
广义MDH
ARCH效应
量价关系
EGARCH—M模型
Generalized Mixture Distribution Hypothesis
ARCH Effect
Price-Volume Relation
EGARCH-M Model