中国股市的波动性研究

【作者】 陈泽忠；

【导师】 周惠彬；

【作者基本信息】 西南财经大学 ， 数量经济学， 2001， 硕士

【摘要】 股价波动是股票市场最基本的特征之一。中国股市经过十余年的发展，取得了举世瞩目的成就，但与国外成熟股市相比，仍然属于新兴市场，其显著特征之一便是股价呈剧烈、频繁的波动。尽管这种剧烈、频繁的波动对推动中国股市的高速发展起到过积极的作用，但不可否认的是，股价的过度波动扭曲了股市的价格机制，致使股价不能较为真实地反映上市公司的内在价值，从而阻碍了股市优化资源配置这一核心功能的有效发挥，这一不利影响在中国股市中已有充分的体现。正是基于这一原因，平抑股价的过度波动以维护股市的平稳运行一直是中国股市监管部门的主要职责之一。本文研究的目的意义就在于揭示出造成中国股市过度波动的根本原因以及股市波动的若干数量特征，以期为监管部门的监管工作提供一些有益的决策依据。本文共分为四章，其中第一章对中国股市波动做一个概括性的制度分析，在此基础上，第二章、第三章和第四章分析中国股市波动的若干数量特征。本文的主要内容和观点概述如下：第一章运用新制度经济学理论分析中国股市过度波动的形成原因以及平抑过度波动的途径。股市的运行状态取决于股市参与者的行为模式，不同的行为模式会产生不同的市场运行特征。从这个角度出发，我们不难发现，正是由于中国股市参与者的大量不规范行为直接导致了股价的过度波动。新制度经济学理论认为，市场的无序源于制度性缺陷所导致的市场参与者的机会主义行为。相应地，中国股市的过度波动——市场失序的一种表现——源于股市的制度性缺陷所导致的参与者的不规范行为，即在现有的制度框架下，股市参与者从个人理性出发将倾向于采取不规范行为。因此，制度性缺陷是造成中国股市过度波动的根本原因，显然，制度的完善也就是平抑股市过度波动的根本途径。制度的完善过程就是制度的不断创新或制度变迁过程。依据新制度经济学理论，制度变迁是人与人之间互动的结果，新制度对旧制度<WP=3>的代替采取的是一种渐进的方式。同理，中国股市的制度变革是具有利益冲突的股市参与各方力量对比或者说讨价还价的结果，因而也必然采取渐进的方式进行。当然，政府可以在协调市场各方利益的基础上加速制度的变革，但如果政府过多地从自身意图出发，利用行政力量进行强制性的制度变革，由于这种强制性变革得不到大多数股市参与者的同意，将使得新制度在执行中遭到曲解甚至违背，从而形成无效的制度供给。因此，中国股市制度的完善不可能一蹴而就，而是一个渐进式过程，这就决定了中国股市在未来一段时期内仍将呈现出比较剧烈、频繁的波动特征。第二章分析股市中各种政策和消息等随机冲击对股市波动影响的持续性。中国股市被形象地称为“政策市”和“消息市”，足见政策和消息对股市的影响程度有多大。因此，有必要对这种影响做一些数量上的界定。VAR分析中的脉冲响应函数和预测误差方差分解能够反映出随机冲击对VAR系统中各变量的影响程度，因此，可以利用这两种分析工具来考察政策和消息等随机冲击对中国股市波动影响的持续性。将沪、深股市波动纳入一个一阶VAR模型做脉冲响应分析和预测误差方差分解，结果表明沪、深股市的随机冲击对两市波动的影响均具有大约6-7个月的持续性。因此，得出的结论是，政策和消息等随机冲击对中国股市波动的影响要经过大约6-7个月时间的衰减才会消失，换言之，一次随机冲击将使股市波动加剧，如果不再出现新的冲击，股市波动在受到冲击之后大约6-7个月才能恢复到冲击之前的水平。第三章考察中国股市波动与宏观经济波动之间的关系，以期从波动的角度考察中国股市是否具备宏观经济“晴雨表”的功能。股票定价的现价模型指出，当前股价等于投资者预期的未来现金流的贴现值，由于公司未来的现金流与未来的宏观经济状况密切相关，因此，未来宏观经济不确定性的增加将使当前的股价波动加剧。在这个意义上讲，股市波动的变化在时间应先于宏观经济波动的变化，从预测的角度讲，可以利用当前的股市波动预测未来的宏观经济波动，用计量经济学术语讲，股市波动与宏观经济波动之间具有Granger因果关<WP=4>系。选取工业增加值、广义货币M2和通货膨胀率三个宏观经济变量，首先将这三个宏观经济波动序列与股市波动序列做配对因果关系检验，再将三个宏观经济波动与股市波动纳入VAR模型做因果关系检验。所有检验结果都表明，中国股市波动与宏观经济波动之间不具备任何显著的Granger因果关系，这就从波动的角度证实了中国股市尚不具备宏观经济“晴雨表”的功能，其根本原因在于股价的过度波动扭曲了股市的价格机制，股价不能较为真实地反映上市公司的内在价值。第四章考察中国股市波动的时变特征，包括股市波动的平稳性、股市波动与收益率的关系以及股市波动与随机冲击之间的非对称关系等三方面问题。包括股票在内的大多数金融资产收益率都具有时变波动特征，即在某些时段的波动十分剧烈，而在另一些时段的波动又相对较为平稳。ARCH模型及其各种扩展形式对刻画股市波动的时变特征是一种非常有效的工具。对上证综合指数收益率和深证成份指数收益率做ARCH效应检验，检验结果表明中国股市波动具有非常显著的时变特更多还原

【Abstract】 Throughout a decade development of China Stock Market ,the stock price fluctuates violently and frequently .Though we can not deny the positive effect of the high volatility in the early development of China Stock Market ,the negative effect of the extremely high volatility should draw more attention .In fact ,the government has always been trying to keep the level of stock market volatility in a reasonable range .This paper is aimed to delve the basic factor that induces the high volatility of China Stock Market and reveal some quantitative characteristics of China Stock Market volatility.This paper is divided into four chapters .The content and viewpoints are summarized as follows:Chapter One is a brief analysis of the volatility of China Stock Market from the perspective of new institutional economics .The stock price fluctuation is determined by the behavior pattern of participants of the stock market ,so we can attribute the extremely high volatility of China Stock Market to the abnormal behavior of participants of China Stock Market .According to the new institutional economics ,the abnormal behavior of participants of stock market results from institutional drawbacks of stock market ,which is the basic factor to cause the extremely high volatility of China Stock Market ,and so institutional innovation is the basic way to control the level of volatility in a reasonable range.Chapter Two uses the Impulse Response Function and Forecast Error Variance Decomposition in VAR analysis to determine the persistence of the effect of stock market shocks on the volatility .The volatility of Shanghai Stock Market and Shenzhen Stock Market are put in a first order VAR model to estimate the Impulse Response Function and decompose the forecast error variance .The estimated Impulse Response Function converges to 0 and the Forecast Error Variance Decomposition converges to a constant value in 6 to 7 months or so ,which indicates that the effect of a stock market shock on the volatility would die out about 6 to 7 months after the shock happened.Chapter Three Examines the relationship between Stock Market volatility and the volatility of a set of macroeconomic variables .According to the present value model of stock price ,that is ,the current stock price is the discounted present value of expected future cash flows ,we can infer that a change in the level of uncertainty about future macroeconomic conditions would cause a proportional change in stock volatility .From the perspective of econometrics ,macroeconomic volatility is Granger caused by the stock market volatility .First ,the pairwise Granger cause test between the stock market volatility and the volatility of Value-added of Industry and broad money M2 and Inflation is conducted .Second ,the stock market volatility and the three macroeconomic volatility are put in VAR models to conduct Granger cause test .All the tests show that there is no significant Granger cause relationship between the stock market volatility and the macroeconomic volatility .So ,information contained in the past history of stock market volatility can not help to forecast the macroeconomic volatility.<WP=7>Chapter Four focuses on a particular characteristic of the stock market volatility .Most of financial asset returns including stock return have time varying volatility ,which can be well depicted by the ARCH model and its various extended forms .First ,LM test and Q statistic test are conducted for the Composite Index of Shanghai Stock Exchange and the Component Index of Shenzhen Stock Exchange ,the null hypothesis of constant volatility is rejected ,so the volatility of China Stock Market is not constant over time .Second ,GARCH(1,1) model ,IGARCH(1,1) model ,GARCH-M model and EGARCH model are estimated respectively to examine the stationarity of stock market volatility ,the relation of stock market volatility with stock return and the asymmetric effect of stock market volatility .The estimated GARCH(1,1) model indicates that the volatility of China Stock Market is wide-sense stationa更多还原