中国股票市场泡沫的实证研究

发布时间：2018-03-06 14:05

本文选题：股市泡沫　切入点：持续期依赖　出处：《北方工业大学》2012年硕士论文　论文类型：学位论文

【摘要】：股票市场泡沫问题一直是金融理论界和实务界关注和争论的问题之一,泡沫的产生和破灭会对一个国家金融市场及经济环境产生重大影响,因此对股票市场泡沫的研究具有重大理论和现实意义。本文在国内外相关研究基础上,从实证角度对上海和深圳证券市场不同时期泡沫的存在性、泡沫的度量及泡沫的破灭预测三个问题进行研究。本文的研究内容主要分为三部分：第一部分对我国股票市场泡沫的存在性进行检验,在相关检验理论基础上,分别推导了游程持续期依赖分析模型和冲量门限自回归模型(MTAR),并利用这两种方法对上证综指和深证综指1997年至2011年的泡沫存在性进行了检验；第二部分内容基于相关泡沫度量理论及模型,对上证综指和深证综指1997年至2011年的泡沫大小进行度量,实证研究中使用两种方法,第一种方法是利用市盈率指标度量各年的泡沫大小,第二种方法是对剩余价值收益模型(F-O)进行修正,并用修正后的模型计算各年泡沫度指标,进而判断泡沫的大小情况;第三部分应用当前最新的金融物理学研究成果,对上证综指和深证综指泡沫的破灭时间进行预测,在相关理论基础上经过推导得到预测泡沫破灭时间的对数周期幂律模型(LPPL),然后对该模型进行一系列简化处理,并应用遗传算法估计模型中的未知参数,进而得到泡沫破灭的预测时间。根据本文的研究得出以下结论：第一,在本文的研究时间范围1997年至2011年间,曾经出现过三次泡沫,分别是1997年至2001年、2005年至2008年和2009年至2011年；第二,在对泡沫的度量方面,用市盈率指标和剩余价值收益模型得到的结论十分一致,2000年和2007年的股票市场泡沫比较大,特别是2007年的泡沫度达到历史最高点,同时实证结果表明2005年的股票市场出现最具价值的投资时机；第三,股票市场泡沫的破灭时间是可以预测的,前提是泡沫需要呈现出特定性质,即只有在股价走势符合对数周期震荡性质的情况下,利用对数周期幂律模型才能够较好地预测泡沫的破灭时间。
[Abstract]:The bubble of stock market has always been one of the problems concerned and debated by the financial theorists and practitioners. The emergence and burst of the bubble will have a great impact on the financial market and economic environment of a country. Therefore, it is of great theoretical and practical significance to study the stock market bubble. On the basis of domestic and foreign relevant research, this paper makes an empirical study on the existence of bubbles in Shanghai and Shenzhen stock markets in different periods. The measurement of bubbles and bubble burst prediction are studied. The research content of this paper is divided into three parts: the first part tests the existence of the stock market bubble in our country, on the basis of the relevant test theory, The model of run duration dependence analysis and impulse threshold autoregressive model are derived, and the bubble existence of Shanghai Composite Index and Shenzhen Composite Index from 1997 to 2011 are tested by using these two methods. The second part measures the bubble size of Shanghai Composite Index and Shenzhen Composite Index from 1997 to 2011 based on the relevant bubble measurement theory and model. The first method is to measure the bubble size of each year by using the price-earnings ratio index. The second method is to modify the surplus value income model (F-O) and calculate the bubble index of each year by using the modified model, and then judge the bubble size. The third part uses the latest financial physics research results to predict the bursting time of the bubble of Shanghai Composite Index and Shenzhen Composite Index. On the basis of relevant theories, the logarithmic periodic power law model for predicting bubble burst time is derived, and then a series of simplified processes are carried out to estimate the unknown parameters in the model by genetic algorithm. Then the forecast time of bubble burst is obtained. According to the research of this paper, the following conclusions are drawn: first, there have been three bubbles between 1997 and 2011 in the time range of this study, namely, 1997 to 2001, 2001 to 2001 and 2001 to 2011; second, In the measurement of bubbles, the results obtained by using the price-earnings ratio index and the surplus value return model are very consistent. The stock market bubbles in 2000 and 2007 were relatively large, especially in 2007, the bubble degree reached the highest point in history. At the same time, the empirical results show that in 2005, the stock market has the most valuable investment opportunity. Third, the time of bubble burst can be predicted, provided that the bubble needs to take on a specific nature. That is, only when the stock price trend accords with the property of logarithmic periodic oscillation, can the logarithmic period power law model be used to predict the bursting time of the bubble.
【学位授予单位】：北方工业大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F832.51;F224

【参考文献】