基于VAR风险度量的Markowitz投资组合模型-应用数学专业论文.docx

摘 要 本文将信用风险中的 VaR 方法引入到股票投资中度量股票的风险，给出了 股票 VaR 风险值的计算公式，并分别从上海证券交易所和深圳证券交易所各随 机选取了 5 只股票（2010.1.4 至 2010.6.30 日之间），分别对两交易所的 5 只股票 的收盘价建立了 GARCH 模型。大量的实证发现 GARCH 模型的残差服从自由度 为 4 的标准学生 t-分布，在此条件下，利用极大似然估计，借助 EVIEWS 6.0 进 行参数估计，计算出了各支股票在置信度为 95%的前提下的 VaR 风险值，并比 较了采用方差计算的风险值,，结果显示：在股票的投资组和中用 VaR 风险度量 法比方差法对风险的刻画更直接，更实用。 从 上海 证 券 交易 所 和深 圳 证 券交 易 所各 随 机 选取 5 只 股票 ， 用 经典 的 Markowitz 模型（方差刻画风险）和用 VaR 刻画风险建立的模型进行比较，结果 显示：在组合期望收益率和置信度一定的前提下，VaR 风险度量法计算出的组合 收益率更高。 关键词：VaR 风险度量；GARCH 模型；方差风险度量；投资组合模型 海南师范大学硕士学位论文 Abstract In this paper, we introduced the VaR method of the credit risk into the risk measure of shares, and given the calculation formula to the stock value at VaR risk, and randomly selected five stocks from the Shanghai Stock Exchange and Shenzhen Stock Exchange (from 2010.1 .4 to 2010.6.30), respectively, we established five GARCH models of the two exchanges on the stocks closing price. A large number of empirical find that the residuals of GARCH model are subject to degrees of freedom in the standard 4 students t-distribution,under this condition,using the maximum likelihood estimation and EVIEWS 6.0, we calculated the VaR risk value at the confidence level of stocks 95% of value and compared the calculated risk values using variance method.The result showed: in the investment of stock group VaR risk measurement method is more direct and practical than variance characterization of risk. We randomly selected five stocks from the Shanghai Stock Exchange and Shenzhen Stock Exchange, in the portfolio expected rate of return in the portfolio and under the premise of a certain degree of confidence, compared the classical Markowitz model (variance characterization of risk ) into the model with VaR risk measurement method,The results showed: The VaR risk measurement methods is used to calculate the rate of return higher. Keywords: VaR risk measure; GARCH model; variance risk measure; portfolio model 独 创 性 声 明 本人声明所呈交的学位论文是本人在导师指导下进行的研究工作及取得的研究成果。 除文中特别加以标注和致谢的地方外，论文中