【作者】 张亚楠；

【导师】 王春峰；

【作者基本信息】 天津大学 ， 管理科学与工程， 2010， 博士

【摘要】 过度自信作为有限理性投资者最主要的心理特征之一,通过作用于投资者对于信息结构的处理方式及其交易行为对证券市场中的资产价格行为起着决定作用。因此,在金融市场微观结构理论框架下,基于过度自信及信息流动机制对证券市场资产价格行为的系统研究便格外重要。本文在过度自信视角下基于金融市场微观结构理论和经典信息模型系统的研究过度自信对交易的驱动作用及其对证券市场资产价格行为的影响,从理论和实证角度分析了过度自信对证券市场中资产价格行为的决定作用。全文主要从以下5个方面进行探讨:1.基于非对称信息和过度自信的交易驱动机理研究。首先从信息经济学的角度,区别于以往的研究,利用Nyholm模型度量超高频知情交易概率,重点研究了交易活跃程度和非对称信息之间的超高频特性,更加透彻地分析交易持续期间、交易量、知情交易变化的原因,对每一笔交易过程的深入研究为各变量之间的相互影响关系提供有力支持;其次,基于经典信息模型和理性预期框架,本部分在引入公共信息考虑信息结构环境的前提下,建立状态依赖过度自信模型,考察当市场中不存在噪音交易者时信息结构是如何作用于过度自信的投资者,并通过影响其对信息的处理方式从而对交易策略和市场的均衡资产价格产生影响的,从信息流动机制和市场微观结构角度分析历史收益和交易量之间的内在关系,为交易驱动研究提供全新的视角和解释。2.投资者过度自信与我国证券市场量价关系研究。本文通过构建适用于我国证券市场的计量模型,考察了整个证券市场、行业和个股的过度自信效应。首先从整个市场的角度,通过市场行情数据研究了投资者行为以及过度自信效应,并实证检验了市场收益和市场交易量之间的领先滞后关系;其次进一步分析了市场收益和个股交易量之间的跨期动态关系,并将不同行业股票作为整体来分别对其过度自信效应和处置效应进行研究,这样不仅可以使我们综合考察我国整个股市周期中过度自信效应的表现特征,更是可以让我们从行业和个股的层面深入探究过度自信效应的截面差异性,揭示了各个行业过度自信效应异同的经验原因。3.投资者过度自信与证券市场波动性关系研究。本文不同于以往学者对于我国证券市场过度自信效应的检验思路和方法,主要使用市场和个股的行情数据,首先对我国上证和深证两个A股市场进行整体综合考察,从整体市场角度分析过度自信在交易量之谜这个金融异象上的解释力;其次在行业和个股层面上深入探究过度自信和波动性关系的截面差异,随后更是进一步按照公司规模和交易活跃程度分组讨论过度自信你和波动性之间的相互作用,不仅使我们从整体市场的角度对过度自信与波动性之间的作用机制有所把握,而且深入细致的探究了不同行业、个股特征下过度自信与波动性之间关系的异同,并进行了合理的理论和经验阐释。4.投资者过度自信与市场质量。本文不仅基于市场指数高频数据系统构建了包含市场深度和市场有效性在内的动态指标体系,而且在此基础上深入探讨过度自信与市场质量之间的联系,揭示出了过度自信与波动性、流动性、市场深度和有效性等表征市场质量指标之间的动态关系,首次对我国证券市场过度自信与市场质量之间的动态关系进行整体综合考察,在市场和个股的角度探究过度自信与市场质量之间的内在作用。不仅拓宽了国内学者基于市场微观结构对行为金融理论研究的视角,而且为有关机构制定政策提供合理的科学依据。5.投资者过度自信和股票动量策略的交叉套利研究。本文首先基于信息模型研究框架,通过拓展引入特定信息结构环境和投资者过度自信参数构建了可以直接度量市场中过度自信交易概率的序贯交易理论模型,并运用合理方法对模型进行参数估计,突破性的解决了过度自信难以度量的问题;其次,利用估计所得时变的过度自信指标,在此基础上进一步构建投资组合和交叉策略,对我国市场中过度自信与动量效应的关系展开研究,首次从实证检验的角度从投资者有限理性的角度对我国动量效应的产生原因进行解释。更多还原

【Abstract】 Overconfidence is one of the defining psychological traits of irrational investors. It affects their information processing and trading behavior, and thus influence assets’prices in the security market. Therefore, it is important to study assets’price behavior based on human overconfidence and mechanism of information flow within the framework of the theory of financial market microstructure. This paper is based on the theory of financial market microstructure and the classic information model. It is a systematic analysis of how overconfidence drives transactions and how it affects asset price behavior both theoretically and empirically. This analysis consists of five main parts:1. The research is about the driving force of transactions based on asymmetric information and over-confidence. Firstly, it’s different from traditional economic researches in that it measures ultra-high-frequency probability of informed trading with the Nyholm model, studies ultra-high-frequency characteristic between transaction activity and asymmetric information in particular, and analyses more thoroughly the reasons for changes of transactions’duration, volume, and informed trading. Those in-depth researches on each transaction clearly show the relations between all variables. Secondly, this paper takes the environment of information structure into consideration, referring to public information. It sets up a state -dependent overconfidence model based on classic information model and the framework of rational expectation. It studies how information structure influences overconfident investors’information processing, hence their trading strategy and the equilibrium prices of assets in the market when there’s not noise traders. It studies the intrinsic relations between historical yield and volume of transactions from the perspective of information flow and microstructure, which is a new way to explain the driving force of transactions.2. This paper continues to study overconfidence and price volume relations in China’s security market. It sets up a quantitative model to measure the effect of overconfidence upon different industries, stocks and the security market as a whole. Firstly, it looks at the whole market, observes investor behavior and overconfidence effect based on data collected from the market. It empirically tests the time relations between yield and transaction volume in the market. Secondly, it further analyzed the dynamic relations between yield and transaction volume of individual stocks. It treats stocks in different industries as a whole and studies the effect upon it of overconfidence and disposition respectively. This enables us not only to observe the traits of overconfidence effect in a whole cycle of security market in China, but also to study the cross-sectional differences of overconfidence effect in individual industries and stocks, thus to find empirical reasons for such differences.3. Relations between investor overconfidence and volatility in the security market. Measures taken to study overconfidence effect in this paper is different from those traditionally used. It collects data of individual stocks and the market as a whole and 1) observes exchanges of Shanghai and Shenzhen– the two A share markets in China– to study the extent to which overconfidence can be applied to explain the mysterious financial phenomenon of unknown transaction volume. 2) It studies the cross-sectional differences of the relations between over-confidence and volatility based on different industries and stocks, then continue to discuss the interactions between over-confidence and volatility by groups divided according to company size and transaction dynamism. In this way, we can not only understand the interaction between over-confidence and volatility, but also do in-depth research of the differences and similarities in different industries and stocks of the relations between the two, thus explain the mechanism and our experience.。4. Investor overconfidence and market quality. This paper sets up a dynamic index system which includes variables such as market depth and efficiency based on the high frequency market index data. It also discusses deeply the relations between overconfidence and market quality, reveals the dynamic relations between different quality indices such as overconfidence, volatility, liquidity, market depth and efficiency. This is the first comprehensive study of the dynamic relations between over-confidence and market quality in China’s security market. It studies in the intrinsic interactions of the two from the perspective of both the market and individual stocks. It broadens domestic scholars’perspective in term of theoretical research in Behavioral Finance based on microstructure. It also builds reliable scientific grounds for governments’policy making.5. Investor overconfidence and cross-arbitrage using momentum strategy in stock trading. 1) The paper is based on the framework of information model. It builds up a theoretical model of sequential trading that can directly measure the probability of overconfidential transactions in the market, using parameters of overconfidence and specific information structure environment. It estimates parameters in a rational way and thus enables the measurement of overconfidence. 2) It builds up investment portfolio and cross-over strategy upon the estimated overconfidence index, studies the relations between overconfidence and the momentum effect in China’s market, and for the first time explains the reasons for momentum effect in China through the perspectives of both empirical tests and irrational investors.更多还原