基于半参数LM-ARMAX模型的股价波动成因分析  

Stock Price Volatility Analysis Based on Semi-Parametric LM-ARMAX Model

作　　者：陈春春[1] 胡日东[1] 

机构地区：[1]华侨大学经济与金融学院,福建泉州362021

出　　处：《华侨大学学报（自然科学版）》2012年第3期330-336,共7页Journal of Huaqiao University(Natural Science)

基　　金：国家软科学计划项目(2008GXS5D130);教育部科学技术研究重点基金资助项目(209148)

摘　　要：选择沪市1991-2010年所有上市公司的数据,建立LM-ARMAX模型来实证股票价格波动的决定因素;然后,根据模型半参数估计的结果,进行基于半参数估计的广义似然比检验和基于Wild Bootstrap的Smirnov检验.研究结果表明:市帐率和成交量是股票价格波动的主要因素,而净资产收益率对股票价格波动的影响不显著;相比起"指数研究"和"样本替代研究"而言,实证的数据精确度更高,说服力更强.We selected all listed companies data of 1991-2010 in Shanghai stock exchange,established long memory-autoregressive moving average with exogenous variables（LM-ARMAX） model to study the determining factors of stock price volatility,and then,had a generalized likelihood ratio test based on the semi-parametric estimation and Smirnov test based on Wild Bootstrap according to the results of semi-parametric estimation of the model.The results show that the major factors of the stock price volatility are the book ratio and the volume,not the income ratio of net assets;compared to ＂index research＂ and ＂sample alternative research＂,the empirical data is more accurate and more convincing.

关 键 词：股票价格 LM-ARMAX模型 波动率 半参数估计 非线性 

分 类 号：O212.7[理学—概率论与数理统计] F830.91[理学—数学]