原油价格波动对股票市场多时间尺度影响研究

发布时间：2018-01-05 04:01

本文关键词：原油价格波动对股票市场多时间尺度影响研究　出处：《中国地质大学(北京)》2017年博士论文　论文类型：学位论文

【摘要】：原油作为现代工业的基础原料与重要能源来源,其价格波动对股票市场存在显著性影响。本研究采取多领域方法交叉融合的时间序列分析框架,从多时间尺度的视角出发,以供给驱动型和需求驱动型原油识别为切入点,对原油价格波动对股票市场的影响及在股票市场中的传导进行研究。主要研究工作和创新贡献体现在以下几个方面:(1)将原油价格波动影响因素分析与多时间尺度分析理念融合,建立基于供给和需求影响因素的多时间尺度原油价格波动类型识别模型,突破传统原油价格波动与股票市场动态关系研究视角。首先在原始域下,识别原油价格波动主要影响因素为原油供给和需求。在多时间尺度下,通过多时间尺度分解及动态相关关系分析结合,将原油价格在多时间尺度下区分为供给和需求驱动型原油价格波动,结果显示基于不同的影响因素驱动型原油价格波动表现出不同的波动特征,为进一步分析原油价格波动对股票市场的影响提供了更为细致的观察角度并奠定了数据基础。(2)融合小波变换及向量自回归模型构建多时间尺度下变量间动态影响分析模型,分析供给和需求驱动型原油价格波动对股票市场的综合股指与行业股指的影响。在短时间周期下,供给和驱动型原油价格波动对全球综合股指与行业股指的影响方向均呈现随机性;在较长的时间周期下,在初始方向上均呈现为正,这表明供给和需求驱动型原油价格波动带来国际间财富转移使得原油进口国拥有更多的财富并投向国际金融市场,全球股票市场在这种情况下呈现出上升态势;供给驱动型原油价格波动对大部分行业股指的影响呈现负向影响,是通过提高生产成本等对行业股指产生负向影响,而需求驱动型原油价格对部分行业呈现正向影响对部分行业呈现负向影响,这主要取决于原油价格上涨带来的负向影响是否能够抵消行业自身的上升趋势。(3)融合小波相干、互相关函数及网络分析方法,构建多时间尺度下原油价格波动在行业股指间的传导网络模型,通过网络分析挖掘原油价格波动在股票市场中传导特征。首先,随着时间尺度的增加,原油价格波动在行业股指间传导时的领先滞后时间长度逐渐增长。交通、公用以及可选消费行业具有较高的概率滞后于其他行业股指发生波动。材料以及电信行业具有较高概率领先于其他行业股指在原油价格波动影响下发生波动。此外,根据不同时间尺度下领先滞后时间长度及传导路径的变化,本文推测,在短时间周期下,股票市场的复杂波动是由于在受到原油价格波动影响后,各行业股指同时发生波动而造成的;而在较长的时间周期下,因为传导时间差异造成股票市场的波动是来自于不同时间点上的原油价格波动冲击的叠加。因此,在短时间周期和长时间周期下,股票市场波动复杂性的引起原因不同。
[Abstract]:Crude oil as the basic raw material of modern industry and an important source of energy, its price fluctuations have a significant impact on the stock market. This research adopts time series cross fusion method of multi domain analysis framework, starting from the multi time scale perspective, driven and demand driven type crude oil identification as a starting point to study the impact of supply. The crude oil price fluctuations on the stock market and the conduction in the stock market. The main research work and contributions are reflected in the following aspects: (1) the factor analysis and multiple time scale impact of oil price fluctuation analysis of the concept of integration, the establishment of the impact of supply and demand of multiple time scale factors of crude oil price fluctuation type recognition model based on breakthrough the traditional crude oil price volatility and the stock market perspective. First, in the original domain, the main influence factors identification of fluctuations in crude oil prices for crude oil supply And demand. In the multi time scale, the multi time scale decomposition and dynamic correlation analysis, the price of crude oil in the multi time scale between supply and demand driven fluctuations in crude oil prices, the impact of different factors on the driven oil price fluctuations show different wave characteristics showed that for the further analysis of influence of crude oil the price fluctuation of the stock market to provide a more detailed view and the data base. (2) analysis model based on wavelet transform and vector autoregressive model to construct the multi time scale under dynamic impact effect between variables, analysis of supply and demand driven oil price fluctuations on the stock market composite index and industry index. In a short period of time, the direction of supply driven and oil price fluctuations impact on the comprehensive index and industry global stock index showed a random; in a long time Between the cycle, in the initial direction showed positive, indicating that the supply and demand driven oil price fluctuations of international wealth transfer makes the oil importing countries have more wealth and invest in international financial markets, the global stock market in this case showed a rising trend; the supply driven effect of crude oil price fluctuations on most the industry index showed the negative influence, is a negative impact by increasing the cost of production of the industry index, while the demand driven price of crude oil has a positive impact on the part of the industry has a negative effect on the part of the industry, which mainly depends on crude oil prices brought about the negative influence of the industry itself is able to offset the upward trend (. 3) fusion wavelet coherence, cross-correlation function and network analysis method, constructing multi time scale fluctuations in crude oil prices in the transmission network model of industrial stock price indices, by Network analysis and mining of crude oil price fluctuations in the stock market conduction characteristics. First, with the increasing time scale, crude oil price fluctuations in the stock index of the leading industry conduction time lag length increasing. The traffic, public consumption and optional lag probability industries with high fluctuations in other industries after the stock index and the telecommunications industry has high material. The probability ahead of other industry index fluctuations in the impact of oil price fluctuation. In addition, according to the change of different time scales lead lag time length and transmission path in this paper that in a short period of time, complex fluctuations in the stock market is due to the impact of fluctuations in crude oil prices, the industry index fluctuation occurred at the same time; while in the longer period of time, because the transmission time difference caused by the fluctuation of the stock market is from a different point in time In the short time period and the long period, the volatility of the stock market is caused by the superposition of the fluctuation of the price of crude oil.

【学位授予单位】：中国地质大学(北京)
【学位级别】：博士
【学位授予年份】：2017
【分类号】：F416.22;F764.1;F831.51

【相似文献】