复杂网络上的信息甄别、扩散与资产定价

发布时间：2017-12-31 16:03

本文关键词：复杂网络上的信息甄别、扩散与资产定价　出处：《天津大学》2013年博士论文　论文类型：学位论文

【摘要】：互联网和社会化媒体的发展从三个方面显著地改变了金融市场中个体的信息交互模式：1.产生了大量且影响广泛的“个体信息源”，个体决策行为面对着了从未有过的大量的信息渠道；2.市场信息表现出异质性和即时交互性，在信息传播速度得到提升的同时，也带来了虚假信息和市场信息操纵等问题，个体决策必须更多的考虑对信息真实性的甄别；3.个体拥有了大范围信息搜索能力，，在决策过程中可以随时改变获取信息来源，从而使金融市场信息交互网络结构呈现出动态演化的复杂特性。这些改变对个体利用信息做出投资决策的过程产生了重要影响，也更进一步地影响了资产价格的形成。面对由科技进步带来的信息扩散渠道和个体对信息处理方式的转变，现有的研究均只从某一个侧面反映了市场信息扩散过程的改变。本文利用计算实验建模的优势，自底向上地对个体信息甄别行为建模，并在个体信息交互层面加入复杂网络作为信息的扩散渠道。分别建立了具有固定信息扩散网络结构的人工股票市场和具有动态演化信息扩散网络结构的人工股票市场，研究异质信息源、动态信息扩散网络结构等多种要素对个体信息甄别能力和资产价格的影响。本文通过引入异质信息源和复杂网络结构，得到了与传统信息甄别模型不一致的结论。本文发现信息扩散网络结构和异质信息源可以影响非知情者的信息甄别能力和资产价格，在某些异质信息源和特定网络扩散结构条件下，非知情者的信息甄别能力将得到很大的提高。此外，本文进一步研究发现，异质信息源对非知情者甄别信息能力的提升具有“阀值效应”，只有保证市场中信息有足够的异质性，非知情者才能以较高概率准确甄别信息。本文还检测了信息精度、传播时间、信息搜索范围和信息源鉴别成本等因素对个体信息甄别能力及资产价格的影响，这些发现都为防范金融市场虚假信息扩散及市场信息操纵提供了必要的理论基础和政策建议。
[Abstract]:The development of Internet and social media has significantly changed the information interaction mode of individuals in the financial market in three aspects: 1. It has produced a large number of "individual information sources" with wide influence. Individual decision-making behavior is faced with a large number of information channels that have never existed before; 2. Market information shows heterogeneity and instant interactivity, which brings problems such as false information and market information manipulation while the speed of information dissemination is improved. The individual decision must consider more to the information truthfulness screening; 3. Individuals have the ability to search for information on a large scale, and they can change the sources of information at any time in the decision-making process. Thus, the information interaction network structure of financial market presents the complex characteristics of dynamic evolution, and these changes have an important impact on the process of making investment decisions by individuals using information. It also further affects the formation of asset prices. In the face of the information diffusion channels brought by the progress of science and technology and the individual transformation of the information processing mode. The existing research only reflects the change of market information diffusion process from one aspect. This paper makes use of the advantage of computational experimental modeling to model individual information discrimination behavior from bottom up. The artificial stock market with fixed information diffusion network structure and the artificial stock market with dynamic evolution information diffusion network structure are established respectively. The market. The effects of heterogeneous information sources, dynamic information diffusion network structure and other factors on individual information discrimination ability and asset price are studied. In this paper, heterogeneous information sources and complex network structures are introduced. It is found that the information diffusion network structure and heterogeneous information sources can affect the ability of information screening and asset prices. Under the condition of some heterogeneous information sources and specific network diffusion structure, the ability of information discrimination of non-insiders will be greatly improved. In addition, this paper further studies and finds out. Heterogeneous information sources have a "threshold effect" on the improvement of the ability of uninformed persons to identify information, only to ensure sufficient heterogeneity of information in the market. The information accuracy, propagation time, information search scope and information source identification cost are also examined to determine the impact of the factors on the ability of individual information screening and asset price. These findings provide necessary theoretical basis and policy suggestions for preventing the spread of false information and manipulation of market information in financial markets.
【学位授予单位】：天津大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：F830.91;F224

【参考文献】