基于对数周期幂律模型的中国股票市场泡沫破裂临界时间实证研究

发布时间：2017-12-31 17:00

本文关键词：基于对数周期幂律模型的中国股票市场泡沫破裂临界时间实证研究　出处：《东北财经大学》2013年硕士论文　论文类型：学位论文

【摘要】：传统的有效市场理论假定在市场上的每个人都是理性的经济人,股票的价格反映了这些理性人的供求的平衡,也充分反映该资产的所有可获得的信息,市场的竞争最终会使证券价格从旧的均衡过渡到新的均衡,而与新信息相应的价格变动是相互独立的或随机的,最终,市场是有效的,均衡是市场的常态。而目前世界上大多数国家股票市场的实践都证明股票收益率分布具有尖峰胖尾与长期记忆的特征,学者们在非线性分析思维的启示下,提出了与有效市场理论相对应的分形市场理论。分形市场理论认为市场是非线性的,由于不同投资者对信息的判断不同,所以信息的传播不是均匀扩散的。在任一时点,价格并没有反映所有已获得的信息,而只是反映了与投资期限相对应的信息的重要性,市场信息的反应不是传统假设的简单的线性关系,市场中存在正反馈与自相似特征,收益率具有长期记忆的效果,市场的波动并不是完全无序的,价格序列遵循的是有偏的布朗运动,所以市场中也一直存在大量的非均衡。对中国股市的收益率进行统计分析,可以得出其分布具有尖峰胖尾的特征,说明中国股票市场是非线性的,然后用R/S分析法可以得到hurst指数,其大于0.5可以得到中国股票市场具有分形特征。股票市场之所以有分形特征,收益率呈现幂律分布,主要原因是正反馈的存在,正反馈导致的羊群效应也是市场产生泡沫的主要原因,这意味着市场的失灵,而市场一旦失灵,我们也并非无从去了解这个市场的运行,市场泡沫的运行也有其自身的规律,正反馈导致的泡沫生长,从物理学的角度,属于自组织临界行为,自组织临界态是一个吸引域,即使改变初始条件,系统最终都会达到这一临界态。在临界态时,系统内事件大小与其频率之间是幂函数关系,这时系统不存在特征尺度,但是可以根据分形理论中标度不变性原理对市场做出一些描述。在股市中,异质投资者之间互相影响,直到最后的一致行为,导致了泡沫的破裂,这就是股票市场的自组织临界。对数周期幂律模型有两个特征,一是对数周期性振荡,二是幂律增长或衰减。股市的加速上扬就是泡沫过程符合这两个特点,所以可以用对数周期幂律模型拟合泡沫的演化过程,找出泡沫破裂的临界时间。首先根据市场中趋势持续时间的长短,找出股票市场存在的自组织状态,这一状态即为泡沫从积累到成熟最后到破裂的一个周期循环。之后用对数周期幂律模型对泡沫进行拟合,找出泡沫破裂的临界时间。通过对中国股票市场的实证研究,发现上证综指、深证成指、创业板指以及个股价格泡沫过程,均可以用对数周期幂律模型进行很好地拟合,得到的拟合临界时间都在泡沫破裂真实时间的误差范围之内。之后,文章对LPPL模型进行稳健性分析,结果显示不同时间段得出的临界时间都在误差接受范围之内,数据越长,越靠近崩盘时间,LPPL模型拟合得出的临界时间相对更精确一些。股市中出现的这些特征可以让我们对投资者行为,股市运行的内在机制等方面进行多角度,更细致的分析,这也可以让我们更准确的去理解市场,也为政策制定者制定政策的时候提供一些借鉴和参考。最后,针对中国股市出现的泡沫现象,提出三条政策建议：第一要保持宏观经济政策的稳定性,防止政策本身的大幅变动引发股市泡沫。第二要努力打造一个公开透明的市场环境,减少信息不对称引发的股市泡沫。第三要培养成熟的机构投资者,可以成为稳定市场的中坚力量,也要加强普通投资者的教育,培养理性投资理念。
[Abstract]:The traditional efficient market theory assumes that people in the market are rational, the price of the stock reflects the supply and demand of the rational balance of human, but also fully reflect all available information about the asset, the market competition will eventually cause the stock price from the old equilibrium transition to a new equilibrium, and are independent of each other or random, and the corresponding changes in the price of new information, the market is efficient, the market equilibrium is the norm. The current practice in most countries in the world stock market have proved that the characteristics of the distribution of stock returns have fat tails and long memory, the scholars in the nonlinear analysis of Enlightenment thinking and put forward the effective market theory and fractal market theory. The corresponding fractal market theory that the market is nonlinear, due to different investors' perceptions of different information, so the information dissemination is not uniform expansion Scattered. At any point in time, the price does not reflect all the information that has been obtained, which just reflects the importance of the investment period information corresponding to the simple linear relationship between the market information was not the traditional assumption, there is a positive feedback and self similarity in the market, has received long-term memory effect of interest rate, market the fluctuation is not completely disordered, price series is to follow a Brown movement, so the market has a large number of non - equilibrium.
Statistical analysis of Chinese stock returns, we can conclude that their distribution has the characteristics of fat tails, that China stock market is nonlinear, and then can get the Hurst index by R/S analysis method, which can get more than 0.5 Chinese stock market has fractal characteristics. The stock market has fractal characteristics, yields show a power-law distribution the main reason is that the positive feedback are main reasons of bubble herding caused positive feedback is the market, which means that the market failure and market failure, we do not have to understand the market, market bubble operation also has its own rules, resulting in the growth of foam positive feedback. From the point of view of physics, which belongs to the self organized critical, self organized criticality is a domain of attraction, even changing the initial conditions, the system will eventually reach a critical state in the pro. Bound states, between the system event size and frequency is the function of power, then the system does not have the characteristic scale, but according to fractal theory scale invariance principle to make some description on the market. In the stock market, heterogeneous investors affected each other, until finally the consistent behavior, lead to the bursting of the bubble, the stock is critical the market self organization. Log periodic power law model has two characteristics, one is the logarithmic periodic oscillation, two is power-law growth or decay. The stock market bubble is a process to accelerate the rise in line with these two characteristics, so we can use the evolution of log periodic power-law model fitting foam, the critical time to find out the bubble burst.
According to the market trend of the duration, find the self organization existing in the stock market, the state is the foam from the accumulation to mature and finally to a cycle of rupture. After the bubble was fitted by logarithmic power law model, the critical time to find out the bubble burst. Through the empirical research on the stock market China the discovery of Shanghai, Shenzhen, the gem index and stock price bubble process can be well fitted by log periodic power-law model, fitting the critical time are all within the scope of the real time bubble burst error within. Then, based on LPPL model robustness analysis, the results show that critical time different time periods that are acceptable within the error, the data is more long, the more close to the collapse time, critical time LPPL model fitting is more accurate in the stock market. These features are that we can on the behavior of investors, the stock market operation mechanism for multi angle, a more detailed analysis, this also allows us to more accurately understand the market, but also provide some reference for policy makers. When the policy after the bubble phenomenon in the stock market China and put forward three policy recommendations: first to maintain the stability of macroeconomic policy to prevent substantial changes in the policy itself caused the stock market bubble. Second to create an open and transparent market environment, reduce the information asymmetry caused by the stock market bubble. Third to cultivate a mature institutional investors, can become the backbone of the stability of the market, also want to strengthen the ordinary investor education, foster rational investment philosophy.

【学位授予单位】：东北财经大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F832.51;F224

【参考文献】