基于多重分形理论的我国行业股价波动性研究

发布时间：2018-01-15 14:27

本文关键词：基于多重分形理论的我国行业股价波动性研究　出处：《南京信息工程大学》2014年硕士论文　论文类型：学位论文

【摘要】：近年来,证券市场波动性已成为国内外学术界研究的一个新热点。众所周知,我国股票市场作为一个新兴市场,市场波动性较大,且不同行业的股价波动总是与市场的波动行为息、息、相关。传统金融理论无法很好地描述实际金融市场复杂的波动特征,多重分形理论的诞生,从非线性的角度为金融市场的研究开辟了新的视野。本文从系统科学的角度出发,基于多重分形理论对我国行业股价波动特征进行研究,论文的主要工作和创新成果如下： (])借鉴信息熵概念对投资期限结构的演变特征进行了定量分析,揭示了投资期限结构的演变对资产价格波动和市场稳定的影响并有效验证了分形市场假说。研究发现：当市场上由不同投资期限的投资者组成时,投资期限结构的信息熵和均衡度较大,投资期限结构的一致性差,价格波动平稳,市场稳定性强；而当投资者的投资期限趋于一致时,投资期限结构的信息熵和均衡度变小,价格波动剧烈,市场稳定性变弱。 (2)以2007年10月17日为转折点,利用多重分形去趋势波动分析法(MF-DFA)比较分析了危机前后沪深300十大行业指数的奇异性特征。结果表明：危机前后各行业指数都具有多重分形特征；与其它行业相比,危机前期电信、工业、可选和信息行业的谱宽度更宽,波动更剧烈,危机后期金融、能源行业的谱宽度更宽,波动更剧烈；与危机前期相比,能源和金融行业危机后期的谱宽度变宽,波动变剧烈,而其它行业危机后期的谱宽度变窄,波动变平稳；就危机前后谱宽度的变化来说,能源、工业、可选、信息、消费和电信行业比其它行业变化幅度大,受危机的影响更显著。 (3)以2005年4月8日至2009年12月31日道琼斯和沪深300十大行业指数的日收益率为样本,在检验美国股市和我国行业长期相关性的基础上,利用多重分形去趋势相关分析法(MF-DXA)并结合滑动窗口技术,比较分析了不同经济时期道琼斯和沪深300十大行业收益率序列交叉相关的多重分形特征,揭示了美国股市对我国行业股价波动的影响。实证结果表明：危机前期,道琼斯和电信、工业、消费和可选行业交叉相关的谱宽度较宽,多重分形特征较强,相关关系较复杂；危机初期,道琼斯和消费、可选、金融和信息行业交叉相关的谱宽度较宽,多重分形特征较强,相关关系较复杂；危机后期,道琼斯和金融、医药、消费、能源行业交叉相关的谱宽度较宽,多重分形特征较强,相关关系较复杂。与危机初期相比,危机前后期道琼斯和除消费以外的各大行业交叉相关的谱宽度更宽,多重分形特征更强,相关关系更复杂,我国行业股价波动受美国股市影响更显著。以上结论不仅有利于投资者估计行业风险、合理进行投资,还有利于各行业企业自身的发展,同时也有利于市场监管者对市场进行有效调控,保证股市和经济的健康发展。
[Abstract]:In recent years, the volatility of the securities market has become a new focus of academic research at home and abroad. As we all know, the stock market of our country is a new emerging market, the market volatility is large. And the volatility of stock price in different industries is always related to the volatility behavior of the market interest, interest, correlation. Traditional financial theory can not describe the complex volatility characteristics of the actual financial market, the birth of multifractal theory. From the point of view of nonlinearity, it opens up a new field of vision for the study of financial market. From the point of view of system science, this paper studies the characteristics of stock price volatility in China industry based on multifractal theory. The main work and innovative results of the thesis are as follows: (]) using the concept of information entropy for reference, the paper makes a quantitative analysis of the evolution characteristics of the term structure of investment. It reveals the influence of the evolution of investment term structure on asset price fluctuation and market stability and effectively verifies the fractal market hypothesis. The information entropy and equilibrium of the investment term structure are large, the consistency of the investment term structure is poor, the price fluctuation is stable, and the market stability is strong; When the investment term of the investor tends to be consistent, the information entropy and equilibrium degree of the investment term structure become smaller, the price fluctuates violently, and the market stability becomes weaker. Take October 17th 2007 as the turning point. Using multifractal detrend fluctuation analysis method. The singularity characteristics of Shanghai and Shenzhen 300 ten industry indexes before and after the crisis are compared and analyzed. The results show that: before and after the crisis, every industry index has multifractal characteristics; Compared with other industries, the spectrum width and fluctuation of telecommunications, industry, optional and information industry in pre-crisis period are wider and more intense, and the spectrum width of finance and energy industry is wider and more volatile in post-crisis finance industry. Compared with the early period of the crisis, the energy and financial industries' spectrum width became wider and more volatile in the latter stage of the crisis, while the latter period of the other industries' crisis became narrower and more stable. In terms of changes in spectrum width before and after the crisis, the energy, industry, options, information, consumer and telecommunications industries have changed more rapidly than other industries, and are more affected by the crisis. From April 8th 2005 to December 31st 2009, the daily yields of the top 10 Dow Jones and Shanghai and Shenzhen 300 industry indices were taken as samples. On the basis of testing the long-term correlation between American stock market and Chinese industry, the multifractal detrend correlation analysis method (MF-DXA) and sliding window technique are used. The multifractal characteristics of cross-correlation between Dow Jones and Shanghai and Shenzhen 300 industry yield series in different economic periods are analyzed. The empirical results show that in the pre-crisis period, the cross-correlation spectrum of Dow Jones and telecom, industry, consumption and optional industries is wider and multi-fractal features are stronger. The correlation relation is complex; At the beginning of the crisis, Dow Jones and consumption, optional, financial and information industry cross-correlation spectrum width is wider, multi-fractal features are stronger, the correlation relationship is more complex; After the crisis, Dow Jones and finance, medicine, consumption, energy industry cross-correlation spectrum width is wider, multi-fractal features are stronger, the correlation is more complex, compared with the initial crisis. Before and after the crisis, Dow Jones and the major industries other than consumption have wider spectrum width, stronger multifractal features and more complex correlation, and the volatility of Chinese industry stock price is more significantly affected by the U.S. stock market. The above conclusions are not only helpful for investors to estimate the industry risk, reasonable investment, but also conducive to the development of enterprises in various industries, but also conducive to the effective regulation and control of the market by market regulators. To ensure the healthy development of the stock market and the economy.
【学位授予单位】：南京信息工程大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F832.51;F224

【参考文献】