1 / 42
文档名称:

VaR参数和非参数法及其在中国股市中的应用.pdf

格式:pdf   页数:42
下载后只包含 1 个 PDF 格式的文档,没有任何的图纸或源代码,查看文件列表

如果您已付费下载过本站文档,您可以点这里二次下载

VaR参数和非参数法及其在中国股市中的应用.pdf

上传人:banana 2014/2/21 文件大小:0 KB

下载得到文件列表

VaR参数和非参数法及其在中国股市中的应用.pdf

文档介绍

文档介绍:华中科技大学
硕士学位论文
VaR参数和非参数法及其在中国股市中的应用
姓名:刘林春
申请学位级别:硕士
专业:数量经济学
指导教师:周少甫
20050429
摘要
是风险估值模型的简称是近年来国外兴起的一种金融
风险管理工具旨在估计给定金融产品或组合在未来资产价格波动下可能的或潜在
的损失的概念虽然简单但是它的度量是一个具有巨大挑战性的统计问题针
对的测算西方学者进行了深入的探讨近些年来国内学者们也开始引进
这一风险分析工具并对有关的理论问题做了初步的探讨在国内外学者的研究基
础上本文对这种新兴的风险测量模型进行了全面而深入的阐述并着重运用
我国证券市场的有关数据对风险测量模型在我国股票市场风险测量中的具体应
用作了实证分析旨在寻找一套符合我国国情的具有可操作性的证券市场风险测量
体系从而促进我国证券市场的健康发展
是一种基础的金融风险管理工具的定义是在给定的置信水平和目标时
段下预期的最大损失本文将简单的介绍的参数和非参数的五种估计方法及其在
实践中的优势进行比较然后我们针对中国上海的股票交易市场指数对以上五种
方法的统计效用进行了实证分析接着利用本文所选择的方法对上海市场指
数中的两个股票组合进行了预测和检验最后我们得出历史模拟法总是得到最好
的预测检验结果但是在波动性模型的基础上我们需要对参数方法做出更好的改


关键词风险价值历史模拟法半参数方法非条件检验
I
Abstract
VaR is the abbreviation of Value at Risk. It's a brand new tool of Finance Risk
Management rising from West Country and is used to estimate the possible and potential
loss of appointed financial products or portfolio according to the fluctuation of prices. The
concept of VaR is very simple, but the measurement method is a challenged item in
statistics. Surrounding the estimating method of VaR, some specialists and scholars in
West Country have made deeper research. In the recent years, some Chinese experts start
to apply the tool of VaR and try researching the relevant theory of it. Based on the theories
of the former experts, this article makes further statement about Value at Risk and applies
some relevant data of China stock market to analyze the application of VaR for risk
measurement in our stock market. It's in order to form the operation system of stock risk
measurement and to promote the development of our stock market more stable.
Value at Risk (VaR) is a fundamental tool for managing market risks. It measures the
worst loss to be expected of a portfolio over a given time horizon under normal market
conditions at a given confidence level. This dissertation will give brief introduction of five
parametric and nonparametric estimation approaches to VaR Calculation as well as their
relative advant