现代金融学理论的问题和解决的途径分析
Modern finance theory analysis the question and the solution ways
金融是现代经济的核心，它是推动经济发展的原动力。以金融为研究对象的金融学也经历了从传统金融学、标准金融学到行为金融学的历史演进。传统金融学和标准金融学的关键区别体现在分析范式上，传统金融学强调的是制度分析，主要是以定性分析为主。标准金融学是强调数理分析或定量研究，从定性到定量研究是金融学发展的一个革命性飞跃。然而，行为金融学却是在对标准金融学的质疑和挑战中逐渐形成的学说和理论体系。从金融学的发展历程看，标准金融学与行为金融学明显存在着重大分歧，行为金融学运用大量的实证研究成果对标准金融学提出了严峻的挑战，在推翻标准金融学的理论假设后，行为金融学的影响达到了前所未有的高度。但是，行为金融学理论并没有形成一个完整的理论体系，并没有建立一个比标准金融学更有实际应用价值的学说体系。甚至从行为金融学的角度也无法找到解释他们取得巨大成功的理论依据，其应用性甚至远远落后于被其否定的以资本资产定价模型、套利定价模型以及期权定价理论等为代表的标准金融学理论。看来行为金融学无法对金融市场上的种种现象给予一致性的解释，事实终将证明它就是金融学理论发展史上的一颗流星。本文提出市场有效周期理论并对其进行实证研究，对于解决标准金融学与行为金融学的重大分歧甚至对现代金融学的发展应该大有裨益。
Finance is the core of modern economy, it is the motive force to promote the development of economy. With financial as the research object of finance also experienced from traditional finance, standard financial learn behavioral finance historical evolution. Traditional finance and embodied in the analysis of the key differences between standard finance paradigm, traditional finance emphasizes the system analysis, mainly based on qualitative analysis primarily. Standard finance is to emphasize mathematical analysis or the quantitative research, from the qualitative to the quantitative research is a revolutionary leap in the development of finance. Behavioral finance, however, is in doubt and challenge to standard finance gradually formed the doctrine and theory system. Looked from the development of finance, the standard finance and behavioral finance clear there are major differences, behavior finance using a large number of empirical research results puts forward serious challenges on standard finance, after the overthrow of standard finance theory hypothesis, the influence of the behavioral finance reached unprecedented heights. However, behavioral finance theory has not formed a complete theoretical system, not to build a more than the standard finance theory system of practical application value. Even from the perspective of behavioral finance theory explain their success cannot be found, even lags far behind its application by its negation to the capital asset pricing model, arbitrage pricing model and option pricing theory represented by the standard finance theory. Behavioral finance can't seem to give consistency to a variety of phenomena in financial markets, the facts will prove that it is a meteor in the history of the development of finance theory. Proposed in this paper the market effective cycle theory and carries on the empirical research, to solve the significant differences of standard finance and behavioral finance even should be helpful to the development of modern finance.
一、标准金融学与行为金融学的分歧
A, the differences of standard finance and behavioral finance
标准金融学是建立在有效市场假说、资本资产定价模型、套利定价模型以及期权定价理论的基础上，以数量经济学为基础的科学。主要是针对复杂的金融产品定价问题进行定量研究，其标志是1952年Markowitz发表的《证券组合选择》一文，成为标准金融理论学说的开端，也标志着标准金融学理论的诞生，Markowitz本人也因此获得了1990年的诺贝尔经济学奖，这是对其理论价值的最好证明。
Standard finance is based on the efficient market hypothesis, capital asset pricing model, arbitrage pricing model and option pricing theory, on the basis of on the basis of quantity economics of science. Mainly to quantitative research on complex financial products pricing issues, the logo is Markowitz in 1952 published "portfolio selection", as the beginning of the theory of standard financial theory, also marks the birth of the standard finance theory, Markowitz himself also won the 1990 Nobel Prize in economics, this is the best proof of the value of the theory.
从标准金融学本身的发展历程看，它创建于20世纪50年代，发展于60年代，成熟于70年代，并最终成为金融学主流理论。有效市场假说最早是由Samuelson(1965)和Fama(1965)提出，并由Fama进行了全面阐述。继Markowitz的组合投资理论诞生之后，1970年，Fama发表了《有效资本市场：理论和实证研究回顾》一文，对前面有效市场假说(EMH)的研究作了系统的总结，提出了有效市场假说完整的理论框架。有效市场是指市场由大量的理性投资者构成，这些投资者基于市场上可以充分流动的信息，对于证券的未来市场价值作出判断，为实现自身利益最大化而相互竞争。证券市场上的有关信息对每个投资者都是均等的，而且每个投资者都能根据其掌握的信息及时进行理性的投资决策，那么，任何投资者都不能获得超常收益，则证券市场是有效的①。
Looked from the development of standard finance itself, it was founded in the 1950 s, the development in the 60 s, mature in the 70 s, and eventually become a mainstream finance theory. The efficient market hypothesis, the earliest is by Samuelson (1965) and Fama (1965), and comprehensively elaborated by Fama. Following Markowitz's portfolio theory was born, in 1970, published his "review of effective capital markets: theory and empirical research, the study of the efficient market hypothesis (EMH) in front of the summary of the system, the efficient market hypothesis complete theoretical framework is proposed. Effective market refers to a market made up of lots of rational investors, the investors can make the best of the information flow, based on market value judgement for the future of the securities market, competition for maximizing their own interests. Information about securities market is equal to every investor, but every investor can according to the information in a timely manner to make rational investment decisions, so, no investors can gain abnormal returns, the stock market is effective.
有效市场假说的核心内容是证券价格总是可以充分反映可获得的信息，基于对信息集的不同反映把市场有效性划分为三个层次：弱式有效、半强式有效和强式有效。该理论隐含的前提和结论可概括为：证券市场上每个投资者都能根据其掌握的信息及时进行理性的投资决策，任何投资者都不可能获得超常收益，则说明证券市场有效。由于所有的信息均充分反映在证券价格中，市场服从随机游走，每日的价格变动均与前日的价格无关。概括起来说，有效市场假说假定所有投资者对新信息都会立刻作出反应，所有信息都充分反映到证券价格中，没有任何一个行为主体具有信息上的优势，证券价格的未来与过去或现在无关。
The efficient market hypothesis is the core content of the security price can always fully reflect available information, based on different reflect the market validity of information set is divided into three levels: weak form efficient, half strong type of effective and strong type of effective. The theory of implicit premises and conclusions can be summarized as: every investor can according to its securities market information in a timely manner to make rational investment decisions, no investors can gain abnormal returns, then effective securities market. Because all the information are fully reflected in securities prices, market obey random walk, daily price movements are nothing to do with the price of the day. In summary, the efficient market hypothesis assumes that all investors react to new information will be immediately and fully reflect all information into the price of securities, there is no a behavior main body has the advantages of information on stock price's future has nothing to do with the past or now.
Sharpe(1964)、Lintner(1965)和Black(1972)构建资本资产定价模型(CAPM)，用来描述资本市场的价格机制。EMH与CAPM是一致且是相互关联的，后者提供了一套检验前者的方法。随后，大量的实证研究对有效市场假说进行了检验并对该学说提供了有力的支持。后续的研究，又有Ross(1976)的套利定价模型(APT)、Black－Scholes(1973)的期权定价模型(OPT)等。在20世纪70年代中期，以有效市场假说(EMH)为基础，以资本市场定价理论和现代资产组合理论为基石的标准金融理论确立了其在金融经济领域的地位。标准金融学理论所作的所有研究都是围绕两个基本假设进行的：即理性人和无风险套利。换句话说，投资者行为是理性的，可以准确地对影响证券价格的所有信息作出准确的反映，套利没有任何风险。
Sharpe (1964), Lintner (1965) and Black (1972) to build the capital asset pricing model (CAPM), is used to describe the capital market pricing mechanism. EMH and CAPM is consistent and is interrelated, which provides a testing method of the former. Then, a large number of empirical studies to test the efficient market hypothesis and the theory provides a powerful support. Follow-up studies and Ross (1976) of the arbitrage pricing model (APT), Black - Scholes option pricing model (1973) (OPT), etc. In the mid - 1970 - s, on the basis of the efficient market hypothesis (EMH), in the capital market pricing theory and modern portfolio theory as the cornerstone of standard financial theory has established its status in the field of financial and economic. Standard finance theory for all research is carried out around two basic assumptions: the rational and risk-free arbitrage. In other words, investors behavior is rational, can affect stock prices accurately make accurate reflect all the information, no risk arbitrage.
然而，在标准金融学理论发展过程中，许多实证研究已经得出诸多的“未解之谜”。学者们认为标准金融学忽视了对投资者实际决策行为的研究。他们通过大量的实证研究，揭示出各种违背有效市场假说的诸多异象，如“一月效应”、“周末效应”、“封闭式基金折价之谜”、“羊群效应”、“阿莱斯悖论”等。传统的“理性人”假定已经无法解释现实人的经济生活与行为，预期效用理论也遭到怀疑。行为金融学开始成为金融理论发展的主流，2002年诺贝尔经济学奖授予行为金融学领域的两位美国学者Kahneman和Smith，是这一学说发展水平的最好诠释。行为金融学是将行为科学、心理学和认知科学的成果运用到金融市场分析中，“有限理性”与“有限套利”是其两大支柱。Kahneman和Tvetsky(1979)通过实验对比发现，大多数投资者并非是标准金融投资者而是行为投资者，他们的行为不总是理性的，也并不总是风险回避的②。行为金融学具有代表性的模型主要包括BSV模型(Barberis, Shleffer, and Vishny, 1998)，认为人们进行投资决策时存在选择性偏差和保守性偏差，BSV模型是从这两种偏差出发来解释投资者决策模型最终导致证券的市场价格变化偏离效率市场③；DHS模型(Daniel, Hirsheifer and Subramanyam, 1998)，将投资者分为有信息和无信息两类，无信息的投资者不存在判断偏差，有信息的投资者存在着过度自信和有偏的自我归因④；Fama(1998)认为DHS模型和BSV模型虽然建立在不同的行为前提基础上，但二者的结论是相似的⑤；HS模型(Hong and Stein, 1999)把作用者分为“观察消息者”和“动量交易者”两类。观察消息者根据获得的关于未来价值的信息进行预测，“动量交易者”则完全依赖于过去的价格变化，模型认为最初由于“观察消息者”对私人信息反应不足的倾向，使得“动量交易者”力图通过套期策略来利用这一点，而这样做的结果恰好走向了过度反应⑥。可以这样认为，行为金融学推翻了标准金融学所构造的理论体系，但行为金融学也没有建立起一个足可以取代标准金融学地位的理论框架，其应用价值甚至远远落后于标准金融学。从最简单的道理上讲，有效市场假说与行为金融学的分歧核心问题是市场是否有效，两者的理论观点截然相反，两个理论至少有一个是不完美的，甚至说存在着严重的错误。
In the standard finance theory development process, however, many empirical studies have concluded that many of the "mystery". Scholars think that standard finance neglected the actual decision-making behavior of investors to study. They through a large number of empirical studies, reveals all kinds of violation of the efficient market hypothesis of vision, such as the "January effect", the "weekend effect", "discount of closed-end fund puzzle" and "flock effect", "the allais paradox", etc. Traditional "rational man" assumption have been unable to explain the reality of economic life and behavior, the expected utility theory has also been suspected. Behavioral finance began to become the mainstream of financial theory development, Nobel Prize for economics in 2002 two American scholars in the field of behavioral finance Kahneman and Smith, is the best interpretation of the theory development level. Behavioral finance is a behavioral science, psychology and cognitive science results apply to the analysis of financial markets, "bounded rationality" and "limited arbitrage" are the two pillars. Kahneman and Tvetsky (1979) by comparison with experiment, found that most investors but behavior of investors is not the standard financial investors, their behavior is not always rational, do not always (2) risk aversion. Behavioral finance representative model mainly includes the BSV model (Barberis, Shleffer, and Vishny, 1998), says that people existing in investment decision selection bias and conservative bias, BSV model from these two kinds of deviation decision model to explain the investors eventually lead to efficiency of securities market price changes deviating from the market (3); DHS model (Daniel, Hirsheifer and Subramanyam, 1998), investors should be divided into two types of information and information, investors there is no judgment deviation without information, the information of investor overconfidence and biased self attribution (4); Fama (1998) argue that although DHS model and the BSV model based on the different behavior of the premise, but both are similar conclusion (5); HS model (Hong and Stein, 1999) the role is divided into "observation" and "momentum traders" two categories. Observing messages according to obtain information about the future value of the forecast and "momentum traders", is wholly dependent on the price changes in the past, models that initially due to "watch" news tend to react to private information is insufficient, make "momentum traders" tried to through hedging strategies to take advantage of this, but the result just overreact to 6. Can think so, behavior finance to overthrow the standard finance the constructed theoretical system, but the behavioral finance does not establish a sufficient theoretical framework can take the place of standard finance status, its application value and even far behind the standard finance. From the most simple sense that the differences of the efficient market hypothesis and behavioral finance core question is whether the market is effective, both opposite theory view, two theories of at least one is not perfect, even said that there is a serious mistake.
20世纪六七十年代以来，对EMH的实证检验主要基于两个思路：一是证券价格总是可以充分体现可获得的全部信息，即信息反映的即时性和准确性。“即时”要求价格对信息的调整时间越短越符合市场有效性的要求；二是证券价格总是等于其内在价值，在没有影响证券基本价值的信息变化时，证券价格应保持不变。“即时”和“准确”是有效市场假说的致命缺陷，它所强调的就是在每一个时点上市场都有效，即要求市场在任何一个时点上都应绝对有效。行为金融学恰恰就抓住了这一点，通过大量的实证检验，揭示证券市场的各种异象。
Since the sixties and seventies of the 20th century, the empirical test of the EMH is based on two ideas: one is the security price can always fully reflect all available information, which reflect the real-time and accuracy of information. "Instant" request price adjustment of information time more short more accord with the requirement of market efficiency; Second, securities prices are always equal to its intrinsic value, basic value has no effect on securities information changes, stock prices should remain the same. "Real-time" and "accurate" is the fatal flaw of the efficient market hypothesis, it is emphasized in every point in time the market, which requires the market at any one point in time should be absolutely effective. Behavioral finance is caught it, through a large number of empirical test, reveals various visions of the stock market.
对有效市场理论实证检验方面的质疑主要有：Shiller-Summers模型中的非理性泡沫(Shiller, 1981)、过度反应(De Bondt和Thaler, 1985)、反应不足(Jegadeesh和Titman, 1993)和无基本面变化股价变动的研究结论(Harris and Gurel, 1986; Shleifer, 1986)。都对有效市场假说提出了具体且明确的质疑。虽然Fama(1991，1998)把上述现象归结为劣模型和偶然现象，甚至还给出了Fama－French三因子模型(Fama和French，1992)来解释部分异象的存在是联合检验问题。但是，行为金融学的影响却日益扩大，至少说明这种解释还是不被金融学界所接受。丁志国、苏治的实证研究发现，证券价格对信息变化的反映不是及时、准确的，存在价值偏离，市场不会达到Fama意义下的有效⑦。
The empirical test of the efficient market theory question mainly has: Shiller - the irrational bubble in the Summers model (Shiller, 1981), an overreaction (De Bondt and Thaler, 1985), reaction (Jegadeesh and Titman, 1993) and no fundamental change of share price movements research conclusion (Harris and Gurel, 1986; Shleifer, 1986). For the efficient market hypothesis put forward specific and clear. Although Fama (1991199) attribute the phenomenon to substandard model and accidental phenomenon, and also gives a Fama - French three factor model (Fama and French, 1992) to explain the existence of partial vision is joint inspection issue. However, the influence of behavioral finance is growing, at least show that explanation is not accepted by the financial community. Ding Zhiguo, Sue for empirical research found that stock prices reflect the information change is not timely, accurate, being value deviation, the market will not achieve effective Fama sense 7.
有效市场假说与行为金融学的分歧的核心就是市场是否有效。经典标准金融理论之所以能成为主流金融学理论而受到广泛的关注，就在于其使用简洁的数学公式建立了统一的理论来解答金融市场上的各种金融问题，在这一点上行为金融学至今无法超越。行为金融学的贡献仅仅是否定了标准金融学的理论假设，甚至可以说推翻了标准金融学的理论前提。但是，在金融学新理论体系的建立方面却仍然显得束手无策。
Differences of the efficient market hypothesis and behavioral finance is at the core of the market is valid. Classic standards can become mainstream finance theory and financial theory has been widely attention, is that they use simple mathematical formula to solve the establishment of a unified theory of all kinds of financial problems in the financial markets, behavioral finance on this point never surpassed. Behavioral finance is simply denied the contribution of the standard finance theory hypothesis, or even overthrow the standard finance theory premise. However, in the establishment of the new theory system of finance in still seemed helpless.
二、市场有效周期理论的创建
Second, the market create effective cycle theory
在解释行为金融学和标准金融学分歧的问题上，我们提出市场有效周期理论。它可以有效地解释标准金融学和行为金融学的分歧。所谓市场有效周期，就是必然存在着一个足够长的时间周期，使得证券价格可以充分反映该时间段内的所有信息，证券市场即定义为有效市场。市场有效周期理论和有效市场假说的本质区别体现在以下方面：
In behavioral finance and standard finance divisions, we put forward effective market cycle theory. It can effectively explain the gap between standard finance and behavioral finance. Alleged market cycle effectively, it is inevitable there is a long enough period, makes the stock prices can fully reflect all the information within the time period, the securities market is defined as the efficient market. Market cycle theory and the essential difference between the efficient market hypothesis is embodied in the following areas:
市场有效周期理论认为，市场在一个时间段内有效，即市场在一定时间段内的平均价格充分反应该时间段内所有信息即为市场有效；有效市场假说则认为，证券市场在每个时点上都有效，即在任何一个时点上证券价格都充分反应所有信息。市场有效周期理论认为证券市场对信息的反应是一个过程，而非迅速实现。高鸿桢、林嘉永(2005)认为，市场信息的传递和价格反应是有条件的而且也是需要时间的，现实的证券市场由于投资者的非完全理性信息的传递和价格反应不会瞬间完成⑧。
Effective cycle theory, market in valid for a period of time, namely the market average price during a certain period of time to fully effective against all information should be within the time period is the market; The efficient market hypothesis is that the securities market at each point in time, or at any point in time security prices fully reflect all information. Effective market cycle theory is that the securities market's response to information is a process, rather than rapid implementation. Douglas d. Davis &charles a. holt, nasser Lin (2005) argues that market information transmission and price reaction is conditional and also need time, reality not entirely rational as investors in the securities market information delivery and price response didn't finish at once today.
如果把有效市场假说的“时点有效”换成“时期有效”，即证券市场存在有效周期。从理论上讲，行为金融学实证研究所揭示的大量异象仅仅是否定了时点有效性，并没有否定时期有效性。如果把有效市场假说的市场有效性定义为时期有效或称存在市场有效周期，则解决了这一理论上的重大分歧。
If the efficient market hypothesis "effective" point to "effective period", namely the effective cycle in security market. In theory, behavioral finance empirical study revealed a large number of vision is simply denied validity time, and have no validity period. If the efficient market hypothesis of market efficiency is defined as the period of valid, or there is market cycle effectively, has solved this theoretically significant differences.
有效市场假说的核心是理性预期和完美套利，有效的市场必须充分反映所有信息，它的基本要求是证券价格对信息反映的即时和准确。“即时”和“准确”所要求的就是在每一个时点上市场都有效，这样的市场并不现实。有效市场假说的致命缺陷是检验时点有效，即在任何一个时点上证券价格都即时、准确地反映所有信息，这一理想化的约束条件不仅使该理论的应用价值大打折扣，也使其存在着巨大的理论缺陷。行为金融学恰恰抓住了有效市场假说的这一纰漏，对其提出了质疑和挑战。市场有效是一个时间周期，而不是一个时点。市场在某一时点上的绝对有效是偶然的，时点的非有效性才是市场的常态。基于此我们提出市场有效周期理论，并提出以下重要创新性概念范畴：
The efficient market hypothesis is the core of rational expectations and perfect arbitrage, effective market must fully reflect all information, it is the basic requirement of stock price to reflect the real-time and accurate information. "Real-time" and "accurate" is required at each point in time the market, this market is not reality. Is the fatal defect of the efficient market hypothesis test point, namely the securities price at any point in time are real-time, accurately reflect all information, this ideal constraints to undercut the application value of the theory not only, also there is a great theoretical defect. Behavior finance has grabbed the efficient market hypothesis this omission, the questions and challenges. Market effectively is a time period, rather than a point. Absolutely effective on the market at some point is incidental, point of effectiveness is the market norm. Based on this we put forward effective market cycle theory, and put forward the following innovative conceptual category:
1.均值有效和平均理性：均值本身就是证券的内在价值，均值是反映一定时间长度范围内市场对该证券价格的承认。在某一个时点上证券价格变动具有极强的随机性，而均值却具有较强的稳定性特征，它具有熨平证券价格随机波动的作用，但并不会完全消除证券价格变动的随机性；均值选择的时间周期越长，稳定效果会越好。平均理性要从横向和纵向两个角度来进行研究。从横向上看，每一个个体可以不理性，但一个群体不理性的可能性就会变得越来越小，一般来说，平均价格会趋于理性；从纵向上看，一个个体在某一个时点上可以不理性，但从一个较长的时间段上看，平均起来还会趋于理性，或者说至少比时点理性的概率大大增强。
1. The mean average effective and rational: average itself is the intrinsic value of securities and the mean is reflected within the scope of a certain length of time on the securities market price. At a certain point in time the securities price change with a strong randomness, and the average is has stronger stability characteristics, it has pressed random fluctuations in stock prices, but will not completely eliminate the randomness of stock price change; Mean to choose the longer the time period, stable effect will be better. Average rationality from the horizontal and vertical two aspects to study. From the horizontal point of view, every individual can be irrational, but a group the possibility of irrational will become more and more small, generally speaking, the average price will tend to be more rational; From the vertical point of view, an individual at a certain point in time can be irrational, but look from a longer time period, the average will tend to be rational, or at least better than point a rational probability increase greatly.
2.长期有效和长期理性：在某一个时点上投资者的行为可以不理性，市场可以无效，但从一个较长的时间段上看，不理性的行为和市场无效会在市场机制的作用下被自发地矫正。不过，矫正的行为也不会在任何时点上都实现得恰如其分，也会有偏离现象的发生，证券价格可能会高于或低于证券的内在价值，但从长期看，偏离内在价值的证券会向其内在价值回归，即均值回归。回归均值是一种必然现象，不过回归时间和空间上却具有随机性。
2. Long-term effective and long-term rational: at a certain point on the behavior of investors can not rational, the market can be ineffective, but look from a longer period of time, not rational behavior and market inefficiencies in the market mechanism under the action of correction was spontaneously. Correct behavior, however, will not at any point in time are appropriately implemented, there will be deviation, stock price may be higher or lower than stock's intrinsic value, but in the long run, deviate from the intrinsic value of securities will return to its intrinsic value, namely the mean reversion. Mean reverting is an inevitable phenomenon, but return on time and space is random.
因此，针对标准金融学提出的“理性人”和“无风险套利”，行为金融学提出的“有限理性”与“有限套利”；我们在这里提出“长期理性”和“长期无风险套利(即均值回归)”两个范畴。
According to standard finance, therefore, put forward the "rational man" and "risk-free arbitrage", proposed by behavioral finance "bounded rationality" and "limited arbitrage"; We proposed here long-term "rational" and "long-term risk-free arbitrage (., mean reversion)" two categories.
如果市场存在有效周期，即不是时点有效，而是在某一个时间段内均值有效，在此时间段的每一个时点上，市场无效也一定是市场的常态特征。既然是时点无效，均值有效，就必然发生在市场机制的作用下证券价格自动矫正偏差的现象，证券收益率一定呈均值回归特征。但是，证券价格偏离均值并不意味着立即回归，还可能在一个特定的时间段沿着背离的方向继续呈趋势性特征。均值回归的时间周期越长，市场越呈现时点无效特征，证券市场收益率的均值就应是市场平均利率加上风险溢价。均值有效应是市场有效的基本前提，否则就是对市场机制的否定。市场的自身功能——理性预期和套利行为可能会矫正时点无效，并形成市场有效周期。
If markets are efficient cycle, that is not point effectively, but at a certain time period the mean effective, each point in time in this period, the market is invalid and must be a normal feature of the market. Since time is invalid, the mean effective, it must be on the market mechanism under the action of stock price of automatic correcting deviation phenomenon, must mean reversion in stock yield characteristics. Stock price deviating from the average, however, does not mean return immediately, and may also be in a specific period of time continue to deviate from the direction trend features. Mean reversion, the longer the time period of the market, the present moment is invalid, the securities market is of average yield should market average rate plus a risk premium. The mean effect is the basic premise of effective market, otherwise, it is a kind of negation to the market mechanism. Market itself functions, rational expectations and arbitrage may correct point is invalid, and form effective market cycle.
三、市场有效周期理论研究的实证分析路径
An empirical analysis of the three effective cycle theory, market research paths
检验市场有效周期是均值回归理论及其实证研究方法。如果市场存在有效周期，就必然呈均值回归特征，过度反应和反应不足都会以反向修正来实现市场的有效性。如果存在均值回归，就说明市场存在有效周期，我们选择均值回归理论作为理论依据，选择均值回归的数量方法进行实证研究。
Test market effective cycle is mean reversion theory and empirical research method. If markets are efficient cycle, will mean reversion characteristic, excessive reaction and reaction will be to reverse correction to achieve the effectiveness of the market. If there is a mean reversion, means that markets are efficient cycle, we choose to the mean reversion theory as the theoretical basis, select the number of mean reversion method for empirical research.
运用该方法可以检验股票指数对信息公布引起股票价格上涨或下跌之后是否出现显著的反向修正现象，即选择不同的时间周期分析股价指数存在均值回归或呈趋势性特征。均值回归(Mean Reversion)是指证券价格无论高于或低于价值中枢(或均值)都会以很高的概率向价值中枢回归的趋势⑨。
Stock index can be tested by using this method to release information caused the stock price rise or fall after whether appear significant reverse correction phenomenon, namely choose different time cycle analysis stock index is mean reversion, or a trend features. Mean Reversion (Mean Reversion) refers to the stock price whether above or below the central value (or average) with a high probability to the value trend of central return pet-name ruby.
均值回归理论近些年有很大影响，尤其在发达国家引起了很多学者的重视。De Bondt与Thaler(1985)所作的经验研究发现，他们以1926年到1982年期间所有在NYSE上市的所有股票为样本，根据过去三年的累计非正常收益率进行排序，构造“赢家组合”与“输家组合”，比较两个组合收益率在随后三年的表现。在所考察的时间区间中，“输家组合”的收益率比“赢家组合”收益率平均每年高出约8％，表明股票收益存在明显的均值回归现象⑩；Dimitros Malliaropulos和Richard Priestley(1999)检验了7个东南亚国家或地区股票市场，结论是认为大量存在均值回归的证据(11)；Balvers和Gilliland(2000)运用对18个具有代表性股票市场1969—1996年数据进行实证研究，结果发现明显的均值回归特征(12)；Jeffrey Gropp(2004)对美国证券交易所、纽约证券交易所和纳斯达克市场进行实证研究的结果发现有明显的均值回归证据(13)；Nam、Pyun和Arize(2002)运用ANST－GARCH. 模型，选择1926年1月—1997年12月美国股票市场的月度数据进行研究，认为股票收益率呈均值回归，但回归是非对称的，负收益率的均值回归速度明显大于正收益率的均值回归速度(14)。
Mean reversion theory has a great influence in recent years, especially in the developed countries has attracted the attention of a great many scholars. De Bondt and Thaler (1985) made by experience, the study found they use the period of 1926 to 1982 all listed on the NYSE stock as sample, over the past three years, according to the cumulative abnormal return of sorting, tectonic combination "winner" and "loser portfolio", compare two combination yields during the next three years. All of the time interval, the loser portfolio returns than "the winner portfolio returns per year on average about 8% more, indicates that stock returns are obvious phenomenon of mean reversion attending; Dimitros Malliaropulos and Richard Priestley (1999) examined seven southeast Asia countries or regions in the stock market, the conclusion is that a lot of evidence of mean reversion (11); Balvers and Gilliland (2000) use of 18 representative stock market in 1969-1996 data for empirical research, the results show obvious characteristics of mean reversion (12); Jeffrey Gropp (2004), the American stock exchange, the New York stock exchange and nasdaq stock market carries on the empirical research results found to have obvious evidence of mean reversion (13); Nam, Pyun and Arize. (2002) using the ANST - GARCH model, select January 1926 - December 1997, the . stock market's monthly data for research, think a stock yields mean reversion, but regression are asymmetric, negative yield is bigger than the speed of mean reversion is yield rate of mean reversion (14).
四、实证研究方法选择
Choose four, empirical research method
目前，国外证券理论界对收益率均值回归问题已有很多研究，概括起来主要有两种数量分析方法，它们是方差比例检验和ANST－GARCH模型分析法。其中ANST－GARCH模型是研究均值回归对称性的分析方法，不是本文的研究范围。国外实证研究最可靠的方法是方差比检验，这也是本文要选择的研究方法。
At present, foreign securities and the theoretical circle of yields the mean regression problems has a lot of research, sum up the main has two kinds of quantitative analysis methods, they are the variance ratio test and ANST - GARCH model analysis method. The ANST - GARCH model is the study of mean reversion symmetry analysis method, is not this article research scope. Foreign empirical research is the most reliable method of variance ratio test, this also is this article to select research methods.
在这里我们使用方差比检验实证方法，对世界上最成熟的几个主要股价指数进行实证分析，主要包括道琼斯、标准普尔500指数、纳斯达克、日经225指数、英国金融时报指数、恒生指数(15)。找出均值回归的证据，并证明市场存在有效周期。
Here we use the variance ratio test empirical method, for the world's most mature several major stock index for the empirical analysis, mainly including the dow Jones and standard &poor's 500 index and nasdaq, the nikkei 225 index, the financial times index, the hang seng index (15). Find evidence of mean reversion, and prove the existence of market cycles.
如果VR(q)小于1，则表示短期回报存在负的自相关，说明短期价格过度波动，长期股票收益率呈均值回归；如果VR(q)大于1，则表示短期回报存在正的自相关，说明短期价格没有过度反应，长期呈均值回避的趋势性特征；当市场有效时，则价格将随机波动，故不存在自相关，即VR(q)等于1。VR(q)偏离1越远，则说明市场的时点有效性越低。也就是说，如果方差率显著不为1，就拒绝随机游走假设。
If the VR (q) is less than 1, the said short-term returns there exists negative autocorrelation, excessive short-term price fluctuations, long-term stock yields the mean regression; If the VR (q) is greater than 1, the said short-term returns exist positive autocorrelation, short-term price not overreact, long show mean inevitable trend; When the market is valid, the price will be random fluctuations, so there is no autocorrelation, namely the VR (q) is equal to 1. VR (q) deviation 1, lower the time validity of the market. That is to say, if the variance rate is not significantly 1, refused to the random walk hypothesis.
五、实证检验
Five, the empirical test
本文选取了从1970年1月到2010年12月之间的道琼斯、标准普尔、纳斯达克、日经225指数、英国金融时报指数、香港恒生指数作为分析样本数据。使用的统计与时间序列分析软件有、Excel 2007和Visual Basic，编程。实证研究中所使用的股价指数的对数价格，即指对数差分收益率(或称连续复利收益率)。
This paper selected from January 1970 to December 2010 between the dow Jones and standard &poor's, nasdaq, the nikkei 225 index, the financial times index, the hang seng index in Hong Kong as the analysis of sample data. Using statistics and time series analysis software has , Excel 2007 and Visual Basic, programming. Used in the empirical research of stock price index of logarithmic price, namely the logarithmic difference yields continuous compound interest (or yield).
用Excel作为辅助分析软件进行方差比检验，实证检验选择的时间间隔为q=1，2，3，4，5，6，12，18，24，30，36。当样本数据足够大时，｜Z(q)｜和｜Z(q*)｜都服从正态分布，当｜Z(q*)｜＞、｜Z(q*)｜＞时，认为在5％的置信水平下，拒绝随机游走假设，认为呈均值回归特征；反之，认为接受随机游走假设。
Use Excel as auxiliary analysis software to carry on the variance ratio test, the empirical selection of interval,2,3,4,5,6,12,18,24,30,36 for q = 1. When the sample is big enough, the | Z (q) (q *) | | and | Z all obey normal distribution, when the | Z | (q *) > , | Z | (q *) > , think under the confidence level of 5%, refused to random walk hypothesis, that assumes the mean reversion characteristic; On the other hand, accepting the random walk hypothesis.
从下表的实证结果看，当q=2时，道琼斯、标准普尔、纳斯达克、恒生和日经225指数的方差比值接近1，说明市场接受股票收益率随机游走假设，不存在自相关；从q=3开始，道琼斯、标准普尔、纳斯达克、恒生和日经指数的方差比值显著不为1，说明以上股票指数收益率均值回归；金融时报指数略有不同，在q=4以后都呈现均值回归特征。
From the empirical results in the table below when q = 2, dow Jones and standard &poor's, nasdaq, the hang seng and the nikkei 225 index variance ratio close to 1, shows the market accept stock yields random walk hypothesis, there is no autocorrelation. Starting q = 3, dow Jones and standard &poor's, nasdaq, the hang seng and the nikkei index variance ratio significantly is 1, that the above stock index returns mean reversion; Financial times index is slightly different, after q = 4 presents the mean reversion characteristic.
我们发现，道琼斯、标准普尔、纳斯达克、恒生和日经225指数，从q=3开始，金融时报指数从q=4开始，再也没有显著等于1的现象发生，说明长期均值回归具有绝对的稳定性。随机波动的时间最长的是金融时报指数，仅仅是4个月，市场时点无效的时间没有很长。这一点从股价指数的走势上也能看出，股市非理性的即脱离经济基本面的时间就是几个月。然后，就会回到理性的轨道。而且我们选择的检验时间周期已经足够长，最长使用了36个月的时间段，即3年周期进行的实证检验，不会有任何检验遗漏问题。
We found that the dow Jones and standard &poor's, nasdaq, the hang seng and the nikkei 225 average, since q = 3, the financial times index from q = 4, no significant phenomenon is equal to 1, the long-term mean reversion has absolute stability. Random fluctuations of the longest is the financial times index, is only four months, the market timing invalid time not very long. It also can see from the trend of the stock index, stock market irrational that time is a few months out of the economic fundamentals. Then you will return to the orbit of rational. And we choose the test cycle time is long enough, the longest use a period of 36 months, the three year period to carry on the empirical test, there won't be any inspection missing problem.
六、结语
Six, the concluding
我们的研究只是证明了股票市场存在有效周期，说明收益率偏离均值一定不是持久性的现象。它可以快速回归，也可以慢速回归，也可以暂时不回归，并不是收益率偏离均值就会立刻向均值回归，在一个特定的时间段内还可能向更严重的偏离均值的方向运行，只是概率越来越小，回归的概率越来越大。这个结论恰恰应对了股票短期走势符合随机漫步特征，回归的周期具有随机性，不过，时间周期越长，回归的稳定性越好。
Our study is just prove the existence of the stock market cycle effectively, that yield deviation from the mean is not the phenomenon of persistence. It can quickly return to, can also be slow to return, also can temporarily don't return, not yield deviation from the mean to mean reversion, immediately at a specific time period may also extend to a more serious deviation from the mean direction, only probability is more and more small, the probability of return is more and more big. This conclusion just deal with the stock short-term movements in line with the characteristics of random walk, randomness of regression cycle, however, the longer the time period, the stability of regression, the better.
市场有效周期理论在解决有效市场假说与行为金融理论的争论上取得了较大进展，对丰富和发展现代金融投资理论应该是一个有益的探索甚至伟大的尝试。均值回归理论实际上否定了市场时点有效。它既说明市场时点有效是偶然的，又说明市场有效周期的存在。证券价格只是缓慢地对信息调整，市场有效性的研究必须考察长期收益。许多研究提出市场无效率，特别是对信息的反应不足或过度反应，事实上是不自觉地支持了时点有效的偶然性，我们的研究着眼点应转向对市场有效周期的研究，而不是围绕时点有效的喋喋不休的争论。
Effective market cycle theory in solving the debate on the efficient market hypothesis and behavioral finance theory has achieved great progress, the rich and the development of modern financial investment theory should be a beneficial exploration great even try. Mean reversion theory actually denied market timing effectively. It shows that market timing is incidental effectively, and that the market cycle effectively. Stock prices only slowly to adjust information, market research must review the effectiveness of the long-term gains. Many study and put forward the market inefficient, particularly in response to the information insufficient or excessive reaction, in fact is to support the point of effective contingency, our research focus should be turned to the study of effective market cycle, rather than around point quarreling over it effectively.
因此，市场存在有效周期具有必然性，市场有效性的实质是证券价格反应决定其变动的所有信息。证券市场有效周期越短，证券市场越健康，短到一个时点时就达到了Fama意义的市场有效性；有效周期越长，市场对信息的反应越缓慢，证券市场健康程度越差。
As a result, markets are efficient cycle has the inevitability, the essence of market efficiency is all the information of stock price reaction to determine the change. Shorter valid period in the securities market, securities market is healthy, short to the significance of a point to reach his market efficiency; The longer the effective cycle, the market's response to information more slowly, the securities market level, the worse health.
