联系我们

论文查询结果	返回搜索

论文编号：	2093
作者编号：	2004005
上传时间：	2010/4/20 9:33:40
中文题目：	证券市场的时间效应问题研究
英文题目：	Security markets ;Time effect ; Weekday effect；
指导老师：	齐寅峰
中文关键字：	证券市场；时间效应；周内效应；月份效应；假日效应
英文关键字：	Security markets ;Time effect ; Weekday effect；Monthly effect；Holiday Effect
中文摘要：	证券市场的时间效应问题，自20世纪30年代人们被发现并开始研究以来，一直处于不断发展之中。近年来，全球证券市场的竞争日趋激烈，全球范围内金融市场的动荡加剧，对于市场时机的把握显得越来越重要，时间效应则是市场时机选择的重要内容之一。随着金融理论的发展以及研究侧重点的切换，时间效应问题急需进行全面深入的研究。本文对我国A股市场中的时间效应问题从理论、实证及应用三个层面进行了全面而深入的探讨。 在理论研究层面，主要论述了当前三个主要金融理论及其与时间效应的关系，并对时间效应问题的研究范畴进行拓展，为下一步的实证分析提供理论依据。在理论分析的同时，对时间效应的研究现状，特别是我国当前的研究动态进行概括。 在实证研究方面，我们以时间段由小到大、由短到长的顺序进行展开，从日内、日间两个层面进行分类，其中日间部分再分成周内、月内及月份间，最后是节假日效应等几个方面。日内部分，从未成交和已成交两个方面展开。在指标选取上，日内的未成交指标选择买卖价差及委托量两个指标，已成交方面采取了收益率及成交量两个指标；日间研究主要采取了收益率和成交量两个指标。无论日内指标还是日间指标，都从趋势性及波动性两个方面进行分析。 在日内时间效应的实证研究中，得出全天适合的交易时间在下午14点至14点50分之间，同时还得出，未成交指标可作为已成交指标的先导指标，以及收益率在临近收盘前有明显的上涨现象。 目前国内对周时间效应研究最多，所采用的样本期间差距较大。本文运用近期的数据对周效应重新进行了实证研究，得出了周二收益率最高、周四收益率最低，而成交量却是周四最大的结论，这与以往研究结论有较大差别。在月内时间效应的研究中，将一个月分成月初、月中、月末三部分进行分析，发现月初收益大多为正，一个月中的收益大部分在月初完成。 在月份时间效应的研究中，我们不仅对月份收益率进行了研究，而且分析了市场在一年之中高、低点在月份间的分布规律。实证结果显示，我国股市存在明显的上半年高收益、下半年低收益现象，高点往往出现在六月份，同时也是收益率最大的月份。 节假日时间效应的研究结果显示，节假日前的收益率高于普通交易日以及节假日后的交易日，说明我国投资者在节日前有着普遍的乐观预期，这可能与我国所处的经济快速发展阶段和股市的初创期有关。 以往关于时间效应的应用，主要作为验证市场有效性与否的证据，很少探讨时间效应研究结论如何在投资实践中应用。本文关于市场时间效应的应用研究是一个开创性的工作。通过本文实证研究的结论，对时间效应的应用加以探讨：一方面，监管层可以利用时间效应来及时发现市场的过度投机，维护市场的稳定；另一方面，投资者可以运用时间效应来更好地完成交易计划，争取更大的收益，通过特定的交易时间段市场的变化特征，来推断可能发生的重大消息。 关于市场时间效应的其它问题，如对日内分笔交易以及交易时间间隔的研究；如何扩展实证结果的应用范围，提升其应用价值等问题都值得进一步研究。
英文摘要：	Time effect of security markets has been studied for many years since 20th century 30’s. In recent years, along with stronger competition in global securities market and more intensified fluctuation in financial market, it is more and more important to grasp the market opportunity. Furthermore, the time effect becomes one of the most crucial elements to choose market opportunity. With the development of financial theory and the change of study emphasis, it is urgent to study this problem comprehensively and thoroughly. In this paper, I try to research time effect of security markets from three aspects of theoretical study, empirical analysis and application. From the aspect of theoretical study, I mainly discuss three current financial theories and their relationship with time effect. Then I extend the research scope of time effect and provide a theory basis for the next step study, i.e., empirical analysis. At the same time, I summarize the existing theories especially the study dynamics. From the aspect of empirical analysis, I investigate it in the order from small to big and from short to long. I classify the time period as “during the day” and “among days”. And “among days” is divided into during the week, during the month and among months. The final type is holiday effect. On the “during the day”, there are two points to study, these are trading and non-trading. For non-trading, I choose two indexes- buy-sell spread and commission quantity. Another two indexes for trading are yield and volume. To “among days”, the indexes are also earning yield and volume. Regardless of “during the day” or “among days”, I will work on two aspects – tendency and fluctuation. In the empirical study of “during the day” time effect, I conclude that the most suitable transaction time in one day is between 2:20pm and 2:40pm. I also show that non-trading index could be the forecast index for trading, and the earning yield would obviously rise before closing. At present, many domestic researchers have studied concerning “week time effect”. But there are great differences among choices of samples period. In this paper, I will adopt the newly data to demonstrate the “week time effect” again. I find that, earning yield on Tuesday is the highest and earning yield on Thursday is the lowest. At the same time, the biggest trading volume will appear on Thursday. Thus there exists a great difference from former conclusion. On the research of “during the month” time effect, I divide one month into three parts—the beginning, the middle and the end of the month. Then I find that the mostly earning yields are positive at the beginning of the month, and mostly earning yield during one month are also earned at the beginning of the month. During time effect research of “among months”, I not only study month earning yield and trading volume, but also analyze the distribution rules of high and low points among months. The demonstration showed that the obvious phenomenon in domestic stock market was that the higher earning yield appeared in the first half year, on the other hand, the lower earning yield appeared in the next half year. And the highest point commonly appears in June, simultaneously the highest earning yield also comes in the same month. The research conclusion about holiday time effect indicates that the earning yield before holidays is better than that of ordinary trading day and after holidays, which means investors have optimistic anticipation before holidays. The possible reason is that the economy is in the fast development stage or stock market is in the startup stage. Existing studies for the application of the time effect are mainly used for evidence to confirm whether the market is valid. These studies don’t discuss how to apply the research conclusions to the investment practice. In this paper, the application study of the market time effect is an innovational work. It discusses the application of the security markets’ time effect based on the conclusion of empirical analysis. With the help of time effect study, the security market supervisor could find the excessive speculation in time and maintain the market stability. On the other hand, the investor is able to complete the transaction plans and earn more by using the time effect. There are many other questions for the time effect that worth further studying, such as separate transaction during one day, time distance of transaction, stability of the empirical conclusion due to insufficient data length, how to extend the application extent of empirical analysis, how to enhance the application, and so on.
查看全文：	预览  下载(下载需要进行登录)