【作者】 赵云立；

【导师】 陈守东；

【作者基本信息】 吉林大学 ， 数量经济学， 2004， 博士

【摘要】 中国证券市场一直是中国市场经济的热点和难点问题，在其在发展过程中，引起了理论界与实务界的广泛关注，如何全面、客观地评价中国证券市场，充分发挥其在经济发展中的作用，提高其优化资源配置的效率水平，不仅关系到宏观经济的持续稳定发展、金融发展和金融全球化下的金融稳定，也关系到中国经济体制改革的成效以及中国现代企业制度的建设的成败。有效市场理论一直是金融经济领域中最具争议和最重要的研究课题之一，有效市场理论的产生对金融经济领域的影响是重大和深远的。本文以证券市场有效性作为研究对象，以实证研究的方法，对中国证券市场效率现状加以评价，在分析中国证券市场效率形成的历史根源和制度根源的基础上，从金融改革及金融创新的角度提出改善中国证券市场效率的对策建议。 论文从有效市场理论入手，从经济学及金融学的角度对有效市场理论的概念、起源、分类、发展及其实践意义进行了系统的回顾和总结，分析了有效市场理论存在的缺陷及其在实际应用中的局限性，在兼顾其他学派的理论观点及争论的同时，指出了实证分析是检验有效市场理论的方向，对证券市场中资产价格(收益)序列分布特征的研究、均衡正常收益率模型的建立以及正确定价模型的选择是有效市场理论发展的热点问题及发展趋势，为本文的写作奠定了坚实的理论基础，为中国股票市场有效性的实证分析提供了理论依据。 随后论文从计量经济学的角度归纳了有效市场理论实证分析的方法、模型的构造、数据的处理、检验的方法，总结了国内外学者的实证研究经验，系统地描述了有效市场理论与随机游动理论以及有效市场实证检验三者之间的关系，为本文的实证研究提供了方法论。 在对中国股票市场有效性的实证分析上，首先全面回顾了十余年来国内学者对中国股票市场所做的实证研究，比较他们所使用的理论方法、实证模型、数据选择以及实证研究结果。从市场的波动入手，结合自相关分析、ADF检验对中国股票市场的全貌进行统计描述，观察到中国股票市场收益率波动具有明显的阶段性，具有右偏态及尖峰厚尾的分布特征，市场总体表现为上升趋势，涨多跌少；上证指数与深圳成指的收益率序列具有基本相同的统计特征。针对以上波动特征，以历史上的制度事件为时点，将市场划分为四个阶段，使用自相关检验、分位数作图及GARCH模型等方法，分段研究市场的波动特征，得到了各阶段的GARCH模型，并据此将市场划分为具有不同波动特征的三个阶段。随后采用Q统计量、CJ统计量、游程及方差比的方法对各阶段的市场进行有效性检验，数据表明，各阶段的统计量在数值上具有比较明显的差异，通过对比分析，认为中国股票市场尚未完全达到弱式有效，但已接近弱式有效，虽时间的推移，市场有效的程度在不断提高。论文在考察市场有效性阶段变化的同时，通过不同的阶段划分方法，还检验了中国股票市场的制度有效性，为政策评价提供了新的方法。 在实证检验了中国股票市场的有效性之后，论文采用事件研究法，以增发事件为例检验中国股票市场的半强式有效性。论文以2000至2002三年中在沪深两市增沪阂脚户动场才羊艾砂解夕汀 发的公司作为研究样本，对比分析了全部样本、沪市样本、深市样本、大公司样本、 小公司样本的累积异常收益，得出以卜结论:从累积异常收益来石，市场没有对增 发事件做出应有的事前反映，虽然己有向卜的趋势，但在统计上并不显著，表明中 国股票市场没有达到半强式有效，但从事后累积异常收益的相对平稳性来石一，市场 效率在逐步得到改善;深圳市场的有效程度要比上海市场低，但差异井不显著。计 算结果还明显检侧到了小公司效应。 作为中国股票市场有效性研究的佐证，文章还研究了卢万多因子模型在中国股票市场的适用性及其实证检验结果对中国股票市场有效性的解释。根据国外文献的惯例，选取】998年，月到2002年12月60个月的月度上证180指数的样本股和深证100指数的样本股收盘数据作为样本，以每年6月末数据为准，对所有的样本股票按照市值(材E)和账面市值比(月刀彻咙)划分出6个投资组合。通过实证分析验证了声不，多因子模刑同样话用于中国股票市场，且二种结果都给出了很高的拟合优度;利用厂F二因子模型捕捉到了中国股票市场存在的小公司效应，市值因子对模则的解释力度要比账面市值比的解释力度大。 最后从我国的实际情况出发，对中国证券市场低效率的制度根源及历史根游进行挖掘，井针对市场的现状，提出改善中国证券市场效率的对策建议。 本文借鉴国内外学者的研究成果，对有效市场的相关理论进行了系统分析与总结，力求在实证分析的方法和对象以及效率改进方面取得一些创新，土要包括以卜儿个方面: 第一，系统地问顾和总结了有效市场理论的产生和发展，指出了有效市场理论的发展方向。 第一，从市场的波动入手，使用白相关检验、ADF检验、分位数作图及6月尺CH模型等方法，分段研究市场的波动特征，以市场的波动特征为依据确定市场的阶段划分，采用Q统计量、CJ统计量、游程及方差比的方法检验各阶段市场的有效性，通过对比分析，较好的描述更多还原

【Abstract】 Stock market has been the most important issue addressed by either theoretical or empirical study during establishing market economy in China, starting with argument relevant to its quality, status; following diverge about its function. How to evaluate it reasonably and improve its function of distribution for resources not only relates to stable development of macro-economy, globalization of finance but also will be the key for economic reform and establishing modern corporate governance in China. Efficient Market Hypothesis (EMH) remains a highly contentions issue in financial field, holding its serious influence. The aim of this thesis is to evaluate efficiency in Chinese stock market using empirical methods and to give some advices to improve its efficiency in the term of financial innovation through investigation its history and change of trading policies.The motivation of this thesis is to present evidence of empirical study in Chinese stock market from investigating concept, classification, origin, development and purport of Efficient Market Hypothesis (EMH). Compared with other theories, some limitation has, then, been pointed out. Empirical study, relevant to distribution nature of return series and choosing reasonable return equilibrium model, would be the direction for development of EMH.Summarizing systematically national and abroad scholars’ results, discussed the relationship between EMH, random walk theory and empirical statistical test. The following partitions described methods and data for empirical study in the term of Econometrics.For empirical study of EMH, this thesis, firstly, reviewed all investigation of Chinese stock market during last decade, comparing the methods, models and data that they used and results. Staring with analysis of volatility in the market, to choose relationship analysis and ADF test to draw the full statistical picture of Chinese stock market. The results showed that return in Chinese stock market had significantly seasonal volatility with right bias, high kurtosis and thick tail. The market generally increased with more rise than fall. There were phenomena of rise and fall dramatically during its history. Return of Shanghai Stock Exchange Index and Shenzhen Stock Exchange Index performed similar statistical characteristics. So we chose Shanghai Stock Exchange Index as sample, divided into three periods by change of trading policies. The empirical study tested market’s period volatility based on relationship test, histogram analysis and GARCH model. Q-statistic, CJ ratio, Run Test and Variance Ratio had been employed to test whether or not Chinese stock market reached Efficiency in three periods. The testing results showed that Chinese stock market did not perform Weak Form Efficiency but quite near. Improving of market efficiency had been found through time moving. Furthermore, market trading system efficiency and policy efficiency were tested and provide a way toevaluate trading policies by classification of period.After test of Weak From Efficiency in Chinese stock market, Event Study was employed, which tested influence for market efficiency by public information announcement to exam whether or not Chinese stock market reached Semi-strong From Efficiency, choosing seasoned equity offerings (SEO) as sample. Companies quoted on Shanghai Stock Exchange and Shenzhen Stock Exchange releasing news of SEO within 2000 and 2002 were chosen for sample. Comparing entire observation with different size observation’s cumulative abnormal return (CAR), the results showed that market did not react before its announcement, which usually happened in overseas stock markets. In spite of drop it is not statistically significant, which means Chinese stock market reached Semi-strong Form Efficiency within the period chosen. By stationary CAR after announcement, market efficiency was improved. The degree of efficiency in Shenzhen Stock Exchange was lower than in Shanghai. The difference between two markets was not significant. The size effect has clearly been found in tes更多还原