基于线性与非线性方法的中国股市量价关系实证研究
|
摘要
一个运转正常的股票市场的表现在一定程度上反映了宏观经济的状况,国民经济发展态势往往可以在股市中得以体现。股票市场中最基本的关系是量价关系。传统的理论认为,量价配合的市场是相对稳定的,而若二者发生背驰,则蕴含着一定的风险,并且在一般情况下,量在价先。近几年中国资本市场已经发生了重大改变,特别是在全球经济危机冲击过后,经典结论是否适合中国的情况,有必要结合中国证券市场运行的实践进行检验。因此,论文的研究有一定的理论价值和现实意义。
论文以2005年下半年以来上证指数和成交量数据为基础,采用中国资本市场这段特殊时期的最新数据,借助于向量误差修正模型、脉冲响应分析、非参数GARCH模型过滤方法、线性与非线性因果检验等现代计量分析技术,从线性与非线性两个角度检验了中国证券市场的牛市和熊市两种不同市场情况下的量价特征：沪市量价之间的基本关系、量价之间的因果性检验、相互间的影响力度、投资者使用量价分析进行投资决策是否可行等。实证对比研究,发现“量价关系”已与以往研究结果有很大差异。不同市场行情下,量价关系具有不对称性。具体表现为,牛市中,股票价格与成交量之间存在着很强的正相关关系,在线性意义下股票价格对股票成交量有着很强的单向拉动作用,而成交量却对股票价格没有解释力；而非线性因果检验发现成交量却对股票价格有一定的解释力,这证实了非线性方法相对于线性方法能够捕捉到更多的信息。熊市中,股票价格和成交量之间不存在明显相关性,也不具有因果关系。传统的“量在价先”投资理念不适合近期的中国证券市场,中国股市“羊群行为”在逐渐消失,日前的中国证券市场是部分有效的。
论文研究创新有：(1)充分考虑中国证券市场股价和成交量剧烈的周期性波动特征,从股价周期波动的不同阶段检验证券市场的运行特征,对不同运行趋势下的股票市场量价关系进行研究；(2)考虑到金融时间序列常表现出明显的非线性特性,从线性与非线性两个角度检验中国证券市场的牛市和熊市两种市场情况下的量价特征。采用非参数GARCH模型过滤方法,并首次引入国外最新非线性因果关系检验方法Diks and Panchenko (2006),对中国股市量价关系进行检验。
The performance of a normal functioning stock market can reflect the situation of macroeconomics to some extent. The developmental trend of national economy often can be reflected in the stock market. The most basic relationship in the stock market is volume price relationship. According to the traditional theories, the markets in which price coordinates with volume are relatively stable, and if the divergence between volume and price occurs, the market contains a certain amount of risks. Under the normal circumstances, and normally, quantity appears before the price. In recent years, China's capital markets have significant changes, especially after the impact of global economic crisis. Therefore, it is necessary for us to check out that classic conclusions whether are still suitable to the current conditions of China in considerations with the practices of securities market in China. Therefore, the thesis has a theoretical value and practical significance.
This thesis is base on the dates of the Shanghai index and turnover since the second half of 2005 and updated data of China's capital market in this special stage, applying the modern econometric analysis technologies, such as VECM model, impulse response analysis and non-parameter GARCH model filtering method, the linear and nonlinear causality test and so on to testify the features of quantity and price in bull and bear markets of China's securities markets from two perspectives of linear and non-linear, which includes the basic relationship between quantity and price in Shanghai stock exchange, causality test on relationship between volume and price, mutual influence strength, feasibility of investors using volume and price analysis to conduct investment decisions. At last, we can find ihat "volume-price relationship" has big differences with the previous research results. Under different market conditions, relationship between price and volume features with asymmetry. It specifically embodies in following aspects:in the bull market, stock prices and trading volume exists the strong positive correlation, and stock price has a strong one-way stimulating effect on the stock trading volume under the sense of linear, while the volume of the stock price has not explanatory power toward stock price; but nonlinear causality test showed that volume of turnover has explanatory power toward stock price which testify the method of non-linear can capture more information than linear method. In the Bear market, there is no significant correlation between stock prices and trading volume and nor the causality. Finally, the extended conclusion can be drawn:the traditional investment philosophy of "quantity before price" is not suitable for the Chinese stock market recently, the Chinese stock market "herd behavior" is gradually disappearing, and current securities market is partially efficient in China.
The innovations of this thesis mainly lie in two points:(1)conducting the research on relationship between volume and price under the different running trend of stock market, especially taking the features of dramatically periodic fluctuation of China's securities market into consideration to testify the running features of from different stages of stock periodic fluctuation;(2) taking the fact into account that the financial time series often show obvious nonlinear characteristics and testifying the characteristics of volume-price relationship under different market circumstances of bull and bear market of China's securities market from linear and non-linear perspectives. Through the comparative analysis on the causality test methods, the thesis introduces foreign latest analysis of nonlinear causality test method for the first time. Diks Panchenko (2006), and applies GARCH model by using nonparametric filtering methods to conduct research on the Chinese stock market price volume relationship.
论文以2005年下半年以来上证指数和成交量数据为基础,采用中国资本市场这段特殊时期的最新数据,借助于向量误差修正模型、脉冲响应分析、非参数GARCH模型过滤方法、线性与非线性因果检验等现代计量分析技术,从线性与非线性两个角度检验了中国证券市场的牛市和熊市两种不同市场情况下的量价特征：沪市量价之间的基本关系、量价之间的因果性检验、相互间的影响力度、投资者使用量价分析进行投资决策是否可行等。实证对比研究,发现“量价关系”已与以往研究结果有很大差异。不同市场行情下,量价关系具有不对称性。具体表现为,牛市中,股票价格与成交量之间存在着很强的正相关关系,在线性意义下股票价格对股票成交量有着很强的单向拉动作用,而成交量却对股票价格没有解释力；而非线性因果检验发现成交量却对股票价格有一定的解释力,这证实了非线性方法相对于线性方法能够捕捉到更多的信息。熊市中,股票价格和成交量之间不存在明显相关性,也不具有因果关系。传统的“量在价先”投资理念不适合近期的中国证券市场,中国股市“羊群行为”在逐渐消失,日前的中国证券市场是部分有效的。
论文研究创新有：(1)充分考虑中国证券市场股价和成交量剧烈的周期性波动特征,从股价周期波动的不同阶段检验证券市场的运行特征,对不同运行趋势下的股票市场量价关系进行研究；(2)考虑到金融时间序列常表现出明显的非线性特性,从线性与非线性两个角度检验中国证券市场的牛市和熊市两种市场情况下的量价特征。采用非参数GARCH模型过滤方法,并首次引入国外最新非线性因果关系检验方法Diks and Panchenko (2006),对中国股市量价关系进行检验。
The performance of a normal functioning stock market can reflect the situation of macroeconomics to some extent. The developmental trend of national economy often can be reflected in the stock market. The most basic relationship in the stock market is volume price relationship. According to the traditional theories, the markets in which price coordinates with volume are relatively stable, and if the divergence between volume and price occurs, the market contains a certain amount of risks. Under the normal circumstances, and normally, quantity appears before the price. In recent years, China's capital markets have significant changes, especially after the impact of global economic crisis. Therefore, it is necessary for us to check out that classic conclusions whether are still suitable to the current conditions of China in considerations with the practices of securities market in China. Therefore, the thesis has a theoretical value and practical significance.
This thesis is base on the dates of the Shanghai index and turnover since the second half of 2005 and updated data of China's capital market in this special stage, applying the modern econometric analysis technologies, such as VECM model, impulse response analysis and non-parameter GARCH model filtering method, the linear and nonlinear causality test and so on to testify the features of quantity and price in bull and bear markets of China's securities markets from two perspectives of linear and non-linear, which includes the basic relationship between quantity and price in Shanghai stock exchange, causality test on relationship between volume and price, mutual influence strength, feasibility of investors using volume and price analysis to conduct investment decisions. At last, we can find ihat "volume-price relationship" has big differences with the previous research results. Under different market conditions, relationship between price and volume features with asymmetry. It specifically embodies in following aspects:in the bull market, stock prices and trading volume exists the strong positive correlation, and stock price has a strong one-way stimulating effect on the stock trading volume under the sense of linear, while the volume of the stock price has not explanatory power toward stock price; but nonlinear causality test showed that volume of turnover has explanatory power toward stock price which testify the method of non-linear can capture more information than linear method. In the Bear market, there is no significant correlation between stock prices and trading volume and nor the causality. Finally, the extended conclusion can be drawn:the traditional investment philosophy of "quantity before price" is not suitable for the Chinese stock market recently, the Chinese stock market "herd behavior" is gradually disappearing, and current securities market is partially efficient in China.
The innovations of this thesis mainly lie in two points:(1)conducting the research on relationship between volume and price under the different running trend of stock market, especially taking the features of dramatically periodic fluctuation of China's securities market into consideration to testify the running features of from different stages of stock periodic fluctuation;(2) taking the fact into account that the financial time series often show obvious nonlinear characteristics and testifying the characteristics of volume-price relationship under different market circumstances of bull and bear market of China's securities market from linear and non-linear perspectives. Through the comparative analysis on the causality test methods, the thesis introduces foreign latest analysis of nonlinear causality test method for the first time. Diks Panchenko (2006), and applies GARCH model by using nonparametric filtering methods to conduct research on the Chinese stock market price volume relationship.
引文
① Karpoff,J.The relation between price changes and trading volume:A Survey[J]. Financial and Quantitative Analysis,1987(22 ):109-126
①Diks, C.and Panchenko, V. A New Statistic and Practical Guidelines for Nonparametric Granger Causality Testing[J]. Journal of Economic Dynamics & Control,2006(30):1647-1669
①Osborne,M.Brownian Motion in the Stock Market[J].Operations Research,1959(7):145-173
②Ying,C.C.Stock Market Prices and Volumes of Sales[J],Econometrica,1966(34):676-686
①Hiemstra C and Jones J D.Testing for linear and nonlinear Granger causality in the stock price-volume relation[J].Journal of Finance,1994(54):1639-1664
②张维,闫翼楠.关于上海股市量价因果关系的实证探测[J].系统工程理论与实践,1998年第6期：111-114
③钱争鸣,郭鹏辉.上海证券交易市场量价关系的分位回归分析[J]，数量经济技术研究,2007年第10期：141-158
① Milton Harris and Artur Raviv.Differences of opinion make a horse race[J].The review of Financial studies,1993(6):473-506
② Karpoff.J.M.Costly short sales and the correlation of returns with volume, Working paper,University of WA.,1985
①高辉.基于Granger检验的上海A股量价关系动态分析[J].河北经贸大学学报,2010年第4期：61-64
②孙克任,季先飞.经济危机前后上证A股量价关系实证分析[J].商业经济,2010年第5期：78-80
③李雪.牛市与熊市中股票价格和成交量互动关系对比分析—基于对上证综指的实证研究[J].现代商业,2010年第14期：30-31
① Lamoureux C, Lastrapes. W D.Heteroskedasticity in stock return data:Volume versus GARCH effects[J]. Journal of Finance, 1990(45):221-229
②史美景,邱长溶.基于极差的条件自回归极差模型[J].统计与决策,2007年第6期：148-149
③邓晓益,郭庆春.证券市场成交量对收益率波动性影响的实证分析[J].上海金融学院学报,2007年第3期：27-31
④杨炘，王邦宜.交易量与股价波动性对中国市场的实证研究[J].系统工程学报,2005年第20期：530-534
① Ray, Y, Chou. Forecasting financial volatilities with extreme values:the conditional autoregressive range(CARR) model[J]. Journal of Money, Credit and Banking,2005(37):561-582
②夏天.基于CARR模型的交易量与股价波动性动态关系的研究[J].数理统计与管理,2007年第5期：887-894
[1]马薇,协整理论与应用[M].天津：南开大学出版社,2004：45-122
[2]李子奈,高等计量经济学[M].北京：清华大学出版社,2002：155-200
[3]叶阿忠,非参数计量经济学[M].天津：南开大学出版社,2003：50-130
[4]高铁梅,计量经济学分析方法与建模[M].清华大学出版社,2007：145-183
[5]张晓峒,计量经济学基础[M].南开大学出版社,2008：282-341
[6]张维,闫冀楠.关于上海股市量价因果关系的实证探测[J].系统工程理论与实践,1998年第6期：111-114
[7]陈怡玲,宋逢明.中国股市价格变动与交易量关系的实证研究[J].管理科学学报,2000年第2期：62-68
[8]吴冲锋,王承炜,吴文锋.交易量和交易量驱动的股价动力学分析方法[J].管理科学学报,2002年第2期：1－11
[9]杨子晖,温雪莲.价格国际传递链中的“中国因素”研究—基于非线性Granger因果检验[J].统计研究,2010年第2期：87-93
[10]赵留彦,王一鸣.沪深股市交易量与收益率及其波动的相关性：来自实证分析的证据[J].经济科学,2003年第2期：57-67
[11]张永东,黎荣舟.上海股市口内波动性与成交量之间引导关系的实证分析[J].系统工程理论与实践,2003年第2期：19-24
[12]芮萌,孙彦丛,王清和.中国股票市场交易量是否包含预测股票收益的信息研究[J].统计研究,2003年第3期：54—59
[13j唐齐鸣,张学功.基于内幕交易下的中国股市量价因果关系分析[J].数晕经济技术经济研究,2005年第6期：95-100
[14]钱争鸣,郭鹏辉.上海证券交易市场量价关系的分位回归分析[J],数量经济技术研究,2007年第10期：141-158
[15]刘汉中.沪深股票市场回报率、波动率和交易量关系的实证研究.运筹与管理[J],2007年第6期：123-127
[16]张志鹏,杨朝军.成交量与收益、收益波动的动态关系[J].哈尔滨工业大学学报,2009年第4期：280-283
[17]何兴强.中国股市收益和交易量动态引导关系的实证分析[J].南方经济,2006年第6期：102-110
[18]李双成.中国股票市场量价关系的理论与实证研究[D].天津：天津大学管理学院,2006年：1-132
[19]李捷瑜.股票收益与交易量的动态关系研究：来自不同估计方法的证据[J].南方经济,2006年第1期：94-104
[20]段吉华.价量关系的理论背景与实证证据探索[D].四川：西南财经大学,2002年：1-67
[21]王明照,郭冰.沪深两市收益率与成交量因果关系的实证研究[J].数学的实践与认识,2006年第9期：23-27
[22]曾梅凤,张仁舰.上证综合指数与单日成交量的计量模型分析[J].统计与决策(理论版),2007年第3期：73-74
[23]郑志姣.深市股价变动与成交量关系的实证研究[J].现代商业,2009年第3期：26-27
[24]孙克任,季先飞.经济危机前后上证A股量价关系实证分析[J].商业经济,2010年第5期：78-80
[25]李雪.牛市与熊市中股票价格和成交量且动关系对比分析一基于对上证综指的实证研究[J].现代商业,2010年第14期：30-31
[26]邓晓益,郭庆春.证券市场成交量对收益率波动性影响的实证分析[J].上海金融学院学报,2007年第3期：27-31
[27]程希明,蒋学雷,陈敏,吴国富.中国股市板块羊群效应的实证研究[J].系统工程理论与实践,2004年第12期：34-48
[28]李双成.中国股票市场量价关系的理应与实践研究[D].天津：天津大学管理学院,2006年：1-132
[29]尹为醇.中国股市交易量波动率和交易量相关性关系的实证研究[J].世界经济情况,2004年第6期：66-79
[30]李双成,邢志安,任彪.基于MDH假说的中国沪深股票市场量价关系实证研究[J].系统工程,2006年第4期：77-82
[31]李双成,王红霞.中国股票市场交易量与价格波动关系实证研究[J].数学的实践与认 识,2008年第12期,1-10
[32]蒋祥林等.中国股市的信息流、流动性与交易量关系[J].系统工程理论方法应用,2005年第14期：6-10
[33]杨炘,王邦宜.交易量与股价波动性：对中国市场的实证研究[J].系统工程学报,2005年第20期：530-534
[34]夏天.基于CARR模型的交易量与股价波动性动态关系的研究[J].数理统计与管理,2007年第5期：887-894
[35]Karpoff,J.The relation between price changes and trading volume:A Survey[J]. Financial and Quantitative Analysis,1987(22):109-126
[36]Diks, C. and Panchenko, V. A New Statistic and Practical Guidelines for Nonparametric Granger Causality Testing [J]. Journal of Economic Dynamics& Control,2006(30):1647-1669
[37]Clark.P.K.A subordinated stochastic process model with finite variance for speculative prices[J].Econometrica,1973(41):133-155
[38]GRANGER,Clive W.J.,and Oskar MORGENSTERN.Spectral Analysis of New York Stock Market Prices.Kyklos,1963(16):1-27
[39]Hiemstra C and Jones J D.Testing for linear and nonlinear Granger causality in the stock price-volume relation [J].Journal of Finance,1994(54):1639-1664
[40]ANDERSEN,T GReturn volatility and trading Volume:An information flow interpretation of stochastic volatility [J].Journal of Finance,1996(51):169-204
[41]Tauchen,G.and M.Pitts.The price variability-volume relationship on the speculative markets [J],Econometrica,1983(51):485-505
①Diks, C.and Panchenko, V. A New Statistic and Practical Guidelines for Nonparametric Granger Causality Testing[J]. Journal of Economic Dynamics & Control,2006(30):1647-1669
①Osborne,M.Brownian Motion in the Stock Market[J].Operations Research,1959(7):145-173
②Ying,C.C.Stock Market Prices and Volumes of Sales[J],Econometrica,1966(34):676-686
①Hiemstra C and Jones J D.Testing for linear and nonlinear Granger causality in the stock price-volume relation[J].Journal of Finance,1994(54):1639-1664
②张维,闫翼楠.关于上海股市量价因果关系的实证探测[J].系统工程理论与实践,1998年第6期：111-114
③钱争鸣,郭鹏辉.上海证券交易市场量价关系的分位回归分析[J]，数量经济技术研究,2007年第10期：141-158
① Milton Harris and Artur Raviv.Differences of opinion make a horse race[J].The review of Financial studies,1993(6):473-506
② Karpoff.J.M.Costly short sales and the correlation of returns with volume, Working paper,University of WA.,1985
①高辉.基于Granger检验的上海A股量价关系动态分析[J].河北经贸大学学报,2010年第4期：61-64
②孙克任,季先飞.经济危机前后上证A股量价关系实证分析[J].商业经济,2010年第5期：78-80
③李雪.牛市与熊市中股票价格和成交量互动关系对比分析—基于对上证综指的实证研究[J].现代商业,2010年第14期：30-31
① Lamoureux C, Lastrapes. W D.Heteroskedasticity in stock return data:Volume versus GARCH effects[J]. Journal of Finance, 1990(45):221-229
②史美景,邱长溶.基于极差的条件自回归极差模型[J].统计与决策,2007年第6期：148-149
③邓晓益,郭庆春.证券市场成交量对收益率波动性影响的实证分析[J].上海金融学院学报,2007年第3期：27-31
④杨炘，王邦宜.交易量与股价波动性对中国市场的实证研究[J].系统工程学报,2005年第20期：530-534
① Ray, Y, Chou. Forecasting financial volatilities with extreme values:the conditional autoregressive range(CARR) model[J]. Journal of Money, Credit and Banking,2005(37):561-582
②夏天.基于CARR模型的交易量与股价波动性动态关系的研究[J].数理统计与管理,2007年第5期：887-894
[1]马薇,协整理论与应用[M].天津：南开大学出版社,2004：45-122
[2]李子奈,高等计量经济学[M].北京：清华大学出版社,2002：155-200
[3]叶阿忠,非参数计量经济学[M].天津：南开大学出版社,2003：50-130
[4]高铁梅,计量经济学分析方法与建模[M].清华大学出版社,2007：145-183
[5]张晓峒,计量经济学基础[M].南开大学出版社,2008：282-341
[6]张维,闫冀楠.关于上海股市量价因果关系的实证探测[J].系统工程理论与实践,1998年第6期：111-114
[7]陈怡玲,宋逢明.中国股市价格变动与交易量关系的实证研究[J].管理科学学报,2000年第2期：62-68
[8]吴冲锋,王承炜,吴文锋.交易量和交易量驱动的股价动力学分析方法[J].管理科学学报,2002年第2期：1－11
[9]杨子晖,温雪莲.价格国际传递链中的“中国因素”研究—基于非线性Granger因果检验[J].统计研究,2010年第2期：87-93
[10]赵留彦,王一鸣.沪深股市交易量与收益率及其波动的相关性：来自实证分析的证据[J].经济科学,2003年第2期：57-67
[11]张永东,黎荣舟.上海股市口内波动性与成交量之间引导关系的实证分析[J].系统工程理论与实践,2003年第2期：19-24
[12]芮萌,孙彦丛,王清和.中国股票市场交易量是否包含预测股票收益的信息研究[J].统计研究,2003年第3期：54—59
[13j唐齐鸣,张学功.基于内幕交易下的中国股市量价因果关系分析[J].数晕经济技术经济研究,2005年第6期：95-100
[14]钱争鸣,郭鹏辉.上海证券交易市场量价关系的分位回归分析[J],数量经济技术研究,2007年第10期：141-158
[15]刘汉中.沪深股票市场回报率、波动率和交易量关系的实证研究.运筹与管理[J],2007年第6期：123-127
[16]张志鹏,杨朝军.成交量与收益、收益波动的动态关系[J].哈尔滨工业大学学报,2009年第4期：280-283
[17]何兴强.中国股市收益和交易量动态引导关系的实证分析[J].南方经济,2006年第6期：102-110
[18]李双成.中国股票市场量价关系的理论与实证研究[D].天津：天津大学管理学院,2006年：1-132
[19]李捷瑜.股票收益与交易量的动态关系研究：来自不同估计方法的证据[J].南方经济,2006年第1期：94-104
[20]段吉华.价量关系的理论背景与实证证据探索[D].四川：西南财经大学,2002年：1-67
[21]王明照,郭冰.沪深两市收益率与成交量因果关系的实证研究[J].数学的实践与认识,2006年第9期：23-27
[22]曾梅凤,张仁舰.上证综合指数与单日成交量的计量模型分析[J].统计与决策(理论版),2007年第3期：73-74
[23]郑志姣.深市股价变动与成交量关系的实证研究[J].现代商业,2009年第3期：26-27
[24]孙克任,季先飞.经济危机前后上证A股量价关系实证分析[J].商业经济,2010年第5期：78-80
[25]李雪.牛市与熊市中股票价格和成交量且动关系对比分析一基于对上证综指的实证研究[J].现代商业,2010年第14期：30-31
[26]邓晓益,郭庆春.证券市场成交量对收益率波动性影响的实证分析[J].上海金融学院学报,2007年第3期：27-31
[27]程希明,蒋学雷,陈敏,吴国富.中国股市板块羊群效应的实证研究[J].系统工程理论与实践,2004年第12期：34-48
[28]李双成.中国股票市场量价关系的理应与实践研究[D].天津：天津大学管理学院,2006年：1-132
[29]尹为醇.中国股市交易量波动率和交易量相关性关系的实证研究[J].世界经济情况,2004年第6期：66-79
[30]李双成,邢志安,任彪.基于MDH假说的中国沪深股票市场量价关系实证研究[J].系统工程,2006年第4期：77-82
[31]李双成,王红霞.中国股票市场交易量与价格波动关系实证研究[J].数学的实践与认 识,2008年第12期,1-10
[32]蒋祥林等.中国股市的信息流、流动性与交易量关系[J].系统工程理论方法应用,2005年第14期：6-10
[33]杨炘,王邦宜.交易量与股价波动性：对中国市场的实证研究[J].系统工程学报,2005年第20期：530-534
[34]夏天.基于CARR模型的交易量与股价波动性动态关系的研究[J].数理统计与管理,2007年第5期：887-894
[35]Karpoff,J.The relation between price changes and trading volume:A Survey[J]. Financial and Quantitative Analysis,1987(22):109-126
[36]Diks, C. and Panchenko, V. A New Statistic and Practical Guidelines for Nonparametric Granger Causality Testing [J]. Journal of Economic Dynamics& Control,2006(30):1647-1669
[37]Clark.P.K.A subordinated stochastic process model with finite variance for speculative prices[J].Econometrica,1973(41):133-155
[38]GRANGER,Clive W.J.,and Oskar MORGENSTERN.Spectral Analysis of New York Stock Market Prices.Kyklos,1963(16):1-27
[39]Hiemstra C and Jones J D.Testing for linear and nonlinear Granger causality in the stock price-volume relation [J].Journal of Finance,1994(54):1639-1664
[40]ANDERSEN,T GReturn volatility and trading Volume:An information flow interpretation of stochastic volatility [J].Journal of Finance,1996(51):169-204
[41]Tauchen,G.and M.Pitts.The price variability-volume relationship on the speculative markets [J],Econometrica,1983(51):485-505