技术分析、有效市场与行为金融

发布时间：2018-05-07 01:11

本文选题：技术分析 + 市场有效性　；参考：《复旦大学》2014年博士论文

【摘要】：技术分析作为一种经典的投资分析工具,在现实中经久不衰,蒙受了大量投资者的青睐。但是由于理论依据的缺乏和方法上的主观性,加之与弱式有效市场假说的矛盾,技术分析在金融学术界中始终备受争议。理论与现实形成了鲜明的对比,这一反常现象至今尚未有充分合理的解释。本文采用了理论建模和实证检验相结合的研究方法,探寻了技术分析被广泛使用的深层次原因。通过建立三个简单的资产定价模型,本文从多角度解释了技术分析能够产生经济收益的可能原因,其中：信息发现模型假设市场中存在信息不对称以及信息传播上的摩擦,技术分析可帮助推测市场中与基本面价值相关的非公开信息；趋势追随模型假设市场中存在大批非理性的噪音交易者,技术交易规则捕捉到了他们的趋势交易行为,并确有可能获利；羊群效应模型假设大量技术交易者使用同质的交易规则,他们自身的羊群买卖行为推动了价格走势向着他们所期望的方向“自我实现”。基于2000-2012年中国A股市场的实证结果表明,双重移动均线和交易区间突破两类规则可产生最为显著的经济收益,年化超额回报率分别达到了5.18%和2.85%。而相对价格指数、亚历山大过滤、价格形态以及K线图等规则,在考虑了交易成本后,并不能产生稳健的超额回报率。本文采用了多种较为前沿的计量技术和先进的实证方法来解决数据过度挖掘和风险调整的问题。基于898家上市公司股票的面板数据回归分析结果进一步表明,信息不透明程度较高、技术交易者数量较少、投资者情绪较弱的公司股票,所录得的技术分析超额回报率较高,支持了信息发现模型的预测,但是与趋势追随模型和羊群效应模型的预测相抵触。这说明技术分析并非是盲目的噪音交易行为,而是市场暂时失效时,投资者获取信息的一种必要手段。
[Abstract]:As a classical investment analysis tool, technical analysis has been favored by a large number of investors for a long time in reality. However, due to the lack of theoretical basis, the subjectivity of methods and the contradiction with the weak efficient market hypothesis, technical analysis has always been controversial in the financial academia. There is a sharp contrast between theory and reality, which has not yet been fully explained. In this paper, theoretical modeling and empirical testing are used to explore the deep reasons why technical analysis is widely used. Through the establishment of three simple asset pricing models, this paper explains the possible reasons for the economic benefits of technical analysis from various angles. Among them, the information discovery model assumes that there is information asymmetry and information transmission friction in the market. Technical analysis can help to speculate on non-public information related to fundamental value in the market; trend following models assume that there are a large number of irrational noise traders in the market, and technical trading rules capture their trend trading behavior. The herding effect model assumes that a large number of technical traders use homogeneous trading rules, and their own herding behavior promotes the price trend to be "self-actualized" in the direction they expect. Based on the empirical results of China A-share market from 2000 to 2012, the results show that the double moving average and the trading interval break through the two kinds of rules can produce the most significant economic returns, the annualized excess returns are 5.18% and 2.85%, respectively. Rules such as relative price indices, Alexander filtering, price patterns and K charts do not produce robust excess returns when transaction costs are taken into account. This paper adopts several advanced metrological techniques and advanced empirical methods to solve the problem of data over-mining and risk adjustment. The results of panel data regression analysis based on 898 listed companies further show that the companies with higher degree of information opacity, less number of technical traders and weaker investor sentiment have higher abnormal returns on technical analysis. It supports the prediction of information discovery model, but contradicts the prediction of trend following model and herding effect model. This shows that technical analysis is not a blind noise trading behavior, but a necessary means for investors to obtain information when the market temporarily fails.
【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】