中国股票市场网络模型动态研究

发布时间：2018-05-19 00:07

本文选题：股票市场 + 动态复杂网络　；参考：《浙江工业大学》2012年硕士论文

【摘要】：股票市场是一个国家金融体系的重要组成部分.研究股票市场的规律特点,对金融市场管理者和投资者都具有十分重要的意义.复杂网络是研究金融股票市场的有力工具之一,本文用复杂网络和加权网络的理论研究了中国股票市场网络模型的动态性质,主要工作如下：一、用复杂网络的理论,以沪深300指数的样本股作为中国股票市场的代表,构建一系列连续的复杂网络,研究了中国股票市场的规律特点.研究发现,随网络总度数的增大,网络的幂律值按指数衰减.股指和网络总度数的波动几乎是一致的.另外,当股指出现剧烈波动时,网络平均聚集系数变大,特别是此时度的分布平均拟合误差变得非常大.进一步研究发现,它的变化和股指波动的变化是同步的,所以我们认为是市场的剧烈波动破坏了股票网络的无标度性.上述结论对收盘价网络更明显.另外,研究表明当股指剧烈波动时,大部分股票度的变化是一致的. 二、BBV模型并不能模拟高度相关的股票市场.在这种应用背景下,我们提出了FBBV模型,它是BBV模型的推广形式.在FBBV模型中，，当有新的节点加入到网络中时，不但会引起被连接节点和它邻居节点之间边权的改变，同时也会引起网络中其它边权的调整，但这种影响程度对不同的节点来说是不同的，相比于BBV模型，这更符合股票市场的特点。本文给出了它的演化算法以及点权、边权演化规律的理论推导,并对FBBV模型进行了数值模拟,模拟的结果和理论推导是一致的. 三、在复杂加权网络理论的基础上,用一种新的赋权方式,构建了沪深300样本股的加权网络.发现该网络的点权分布是幂律分布,这表明在这个网络中,点权很高的节点的数量较少,这些节点在整个加权网络中处于‘核心’地位.
[Abstract]:The stock market is an important part of a national financial system. Studying the characteristics of the stock market is of great significance to both the financial market managers and the investors. The complex network is one of the powerful tools for the study of the financial stock market. In this paper, the Chinese stock market network is studied with the theory of complex network and the theory of weighted network. The dynamic properties of the collaterals are the main work as follows:
First, using the theory of complex network, a series of continuous complex networks are constructed with the sample stock of the Shanghai and Shenzhen 300 index as the representative of the Chinese stock market. The characteristics of the Chinese stock market are studied. It is found that the power law value of the network decreases with the increase of the total network degree. The fluctuation of the index and the total network degree is almost identical. In addition, when the stock index fluctuates sharply, the average aggregation coefficient of the network becomes larger, especially the average fitting error of the distribution of the time is very large. Further research shows that the change of the stock index fluctuation is synchronized with the change of the stock index, so we think that the violent fluctuation of the market has destroyed the scale of the stock network. The price network is more obvious. In addition, the study shows that when the index fluctuates violently, the degree of change of most stocks is the same.
Two, BBV model can not simulate highly related stock market. Under this application background, we propose a FBBV model, which is the extension of the BBV model. In the FBBV model, when a new node is added to the network, it will not only cause the change of the edge between the connected node and its neighbor nodes, but also cause the others in the network. The adjustment of edge weight is different, but the degree of influence is different to the different nodes. Compared to the BBV model, this is more in line with the characteristics of the stock market. This paper gives its evolutionary algorithm, the theoretical deduction of the point weight, the boundary weight evolution law, and the numerical simulation of the FBBV model. The results of the simulation are consistent with the theoretical deduction.
Three, on the basis of the complex weighted network theory, a weighted network of Shanghai and Shenzhen 300 samples is constructed with a new weighting method. It is found that the point weight distribution of the network is a power law distribution, which shows that the number of nodes with very high point weight is less in this network, and these nodes are at the core status in the whole weighted network.
【学位授予单位】：浙江工业大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F832.51;F224

【相似文献】