风险度量中的信息熵方法研究

发布时间：2018-01-13 19:29

本文关键词：风险度量中的信息熵方法研究　出处：《北京交通大学》2015年硕士论文　论文类型：学位论文

【摘要】：受经济全球化和金融一体化的影响,各国的投资市场迅猛发展的同时,投资风险也在随着悄然增长。投资者总是期望获取稳定的收益,规避不必要的风险,因此如何利用合理科学的方法来准确地度量投资中的风险,不仅是投资者关心的问题,而且已经是现代金融学研究的一项重要的课题。在此背景下本文综述了一系列度量风险的方法,尤其强调近几年来兴起的利用信息熵理论来度量风险,分章介绍如下：第一章首先介绍了风险度量方法产生的背景、国内外研究现状以及本文主要内容和安排。第二章综述了几种传统经典的风险度量方法,分别回顾了其发展历程,给出了模型并对其分析了各自的优缺点。第三章介绍了熵的起源与发展、信息熵的定义与性质以及最大熵与最小相对熵原理,这为第四章提供了必要的理论支持。第四章是本文的重点。在第二章的基础上,我们介绍了几种用信息熵来度量风险的模型,从理论上给出了模型的具体形式并分析了模型的特点与适用范围。第五章采用实证分析的方法,选取了50只较好的股票作为分析的样本,通过Matlab编程计算比较了熵模型和传统的风险度量模型的异同,最后得出了熵模型在风险度量中的优势。最后一章是本文的总结以及对后续工作的展望。
[Abstract]:Under the influence of economic globalization and financial integration, with the rapid development of investment markets in various countries, the investment risk is also increasing quietly. Investors always expect to obtain stable returns and avoid unnecessary risks. Therefore, how to use reasonable and scientific methods to accurately measure the risks in investment is not only an issue that investors are concerned about, but also an important subject in modern finance research. In this context, this paper summarizes a series of risk measurement methods, especially emphasizes the use of information entropy theory to measure risk in recent years. The first chapter introduces the background of risk measurement, the current research situation at home and abroad, and the main content and arrangement of this paper. In the second chapter, several classical risk measurement methods are reviewed, their development process is reviewed, models are given and their advantages and disadvantages are analyzed. The third chapter introduces the origin and development of entropy, the definition and properties of information entropy and the principle of maximum entropy and minimum relative entropy, which provides the necessary theoretical support for Chapter 4th. Chapter 4th is the focus of this paper. On the basis of the second chapter, we introduce several models to measure risk by information entropy, give the concrete form of the model theoretically and analyze the characteristics and application scope of the model. Chapter 5th uses the empirical analysis method, selects 50 good stock as the analysis sample, through the Matlab programming computation has compared the entropy model and the traditional risk measurement model similarities and differences. Finally, the advantage of entropy model in risk measurement is obtained. The last chapter is the summary of this paper and the prospect of future work.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：F830.59

【参考文献】