中国股票市场的规模效应：理论与实证

发布时间：2018-01-04 15:24

本文关键词：中国股票市场的规模效应：理论与实证　出处：《复旦大学》2014年博士论文　论文类型：学位论文

【摘要】：本文以中国A股市场为样本,系统性地研究了规模效应的存在性及其原因。基于股票价格形成机制及规模因素在其中的作用,本文重点研究了四个方面的问题：第一,对规模效应本身存在性的检验；第二,从基本面的角度出发,研究规模因素在决定公司系统性风险和投资行为中的影响；第三,研究规模对市场信息因素的影响,以及是否由此产生规模效应；第四,研究规模对流动性的影响,并探讨流动性是否是规模效应的渠道。相应地,本文的第三章至第六章分别研究了这四个问题。本文是对股票横截面收益的解释,其研究方法沿袭该领域的常用手段。除了包括分组检验、Fama-MacBeth回归等经典方法外,本文引入了面板数据模型作为另一种主要的估计方法。面板数据模型不仅能在存在时间效应的情况下提供较好的结果,而且同样适用于个体效应的情况,并且对于一些存在扰动项相关性的情况仍然能够获得稳健的结果。本文较为确定的结论主要有三点：第一,在A股市场中规模效应是确定存在的,表现为小公司的预期收益总体上高于火公司。其中以总市值衡量时最为明显,并且这种效应通过了不同设定的稳健性检验。第二,信息因素是规模效应的可能来源,以分析师关注程度衡量的信息不完全能够降低规模效应的程度,小公司受到投资者的关注较少,因此相关的信息不完全的,从而投资者要求小公司有较高的预期收益。第三,流动性也能够部分地解释规模效应,这一点在股票横截面收益上并不明显,但在时间序列的角度上较为显著,流动性调整的CAPM能够使不同规模的股票组合的未被解释的成分没有差异。本文可能的创新之处有：首先,本文对股票收益实证研究的估计方法进行了较为完整的讨论,试图将面板数据模型引入到相关的研究中,并在规模效应的实证检验中充分运用了这一系列方法。因此,本文不仅获得了准确、稳健的估计结果,而且能够为类似的研究提供方法上的借鉴。其次,本文选择了基于生产的资产定价模型的典型理论,对股票收益背后的经济机制进行了尝试性的探索。以往对股票收益的决定因素的研究往往着眼于对“异象”的解释,而本文介绍的资产定价理论则将股票收益的规律性视为内生性的现象,有助于这方面后续研究的开展。第三,本文在股票价格形成机制的框架中分离出了基本面、信息、流动性三个环节,并结合相关理论,逐一讨论规模因素在其中的作用。此外,利用中国股票市场的样本进行实证分析,其结果为规模效应的来源提供了可能的解释。
[Abstract]:This paper systematically studies the existence and causes of scale effect based on the stock price formation mechanism and the role of scale factors in the Chinese A-share market. This paper focuses on four aspects: first, the test of the existence of scale effect itself; Secondly, from the angle of fundamentals, the paper studies the influence of scale factors in determining the systematic risk and investment behavior of the company. Third, study the influence of scale on market information factors, and whether to produce scale effect; 4th, study the effect of scale on liquidity, and explore whether liquidity is the channel of scale effect. The third to 6th chapters study these four problems respectively. This paper is an explanation of the cross-section return of stock, and its research methods follow the common methods in this field, except for the grouping test. Fama-MacBeth regression and other classical methods. In this paper, panel data model is introduced as another main estimation method. Panel data model can not only provide better results in the presence of time effect, but also be suitable for individual effect. And for some cases where there is correlation of disturbance terms, robust results can still be obtained. There are three main conclusions in this paper: first, in the A-share market, the scale effect is sure to exist. The performance is that the expected earnings of small companies are generally higher than those of fire companies. This effect is most obvious when measured by total market value, and this effect has passed the robustness test of different settings. Second. The information factor is the possible source of the scale effect. The information measured by the degree of concern of the analyst can reduce the degree of the scale effect completely, and the small company is less concerned by the investors, so the relevant information is not complete. Third, liquidity can partly explain the scale effect, which is not obvious in the cross-section return of stock, but is more significant in the perspective of time series. The fluidity adjusted CAPM can make the unexplained composition of the different size stock portfolio not different. The possible innovations of this paper are as follows: first of all. This paper discusses the estimation methods of stock return empirical research and tries to introduce the panel data model into the relevant research. And in the empirical test of scale effect, we make full use of this series of methods. Therefore, this paper not only obtains accurate and robust estimation results, but also can provide a reference for similar research. Secondly. This paper chooses the typical theory of production-based asset pricing model. This paper attempts to explore the economic mechanism behind stock returns. Previous studies on the determinants of stock returns have often focused on the explanation of "anomalies". The asset pricing theory introduced in this paper regards the regularity of stock returns as an endogenous phenomenon, which is helpful to the development of further research in this field. Third. In this paper, in the framework of the mechanism of stock price formation, we separate out three links: fundamentals, information, liquidity, and discuss the role of scale factors one by one. The results of empirical analysis on Chinese stock market provide a possible explanation for the source of scale effect.
【学位授予单位】：复旦大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：F832.51

【参考文献】