当前位置:主页 > 管理论文 > 货币论文 >

现代投资组合理论及其在中国基金业的实证研究

发布时间:2018-02-14 11:09

  本文关键词: 投资组合 跟踪比率 风险容忍参数 三步循环估计法 证券投资基金 出处:《重庆大学》2012年硕士论文 论文类型:学位论文


【摘要】:证券投资基金在国外已有一百多年的发展历史,证券投资基金已成为西方发达国家金融投资业的三大支柱之一。我国资本市场起步于二十世纪八十年代,经过十几年的发展,我国证券投资基金业取得了长足的发展,规模逐渐壮大,但我国证券投资基金业起步较晚,市场占有份额较低,其投资管理水平与西方发达国家相比还有较大差距。证券投资基金作为资本市场机构投资力量的生力军,对改善资本市场投资环境,引导先进投资理念,规范市场投资行为有着极其重要的作用。现代投资组合理论和方法在资产投资和风险分散方面有着举足轻重的地位,对我国证券投资基金的发展、对指导投资者理性投资有着重大的现实意义。因此,,引入和深入研究现代投资组合理论和方法,加强对宏观经济基本面的研究,大力发展金融(衍生)工具市场对推动我国证券投资基金的发展,指导完善我国资本市场,促进我国社会的稳定和经济繁荣,推动我国证券市场保持长期稳定的发展具有重大的理论和现实意义。 本文研究成果如下: ①提出了基于Copula—EGarch—GPD的投资组合模型。选用EGarch—GPD模型拟合单样本数据,在AIC值意义下确定出能最佳拟合数据联合分布函数的Copula函数,结合度量风险损失能力强的CVaR指标与最优化理论运用Monte Carlo模拟法计算出最优投资组合权重。 ②提出了基于期望效用函数极大化的投资组合模型。运用期望效用理论提出了基于期望效用极大化的投资组合模型,新模型的最优投资组合权重运用拉格朗日乘子法很易求出。 ③提出了一种新的基金业绩评价指标——跟踪比率(Tracking Error;简称TR)。从夏普比率和信息比率的定义可知,夏普比率只评价了资产管理者总的投资能力,却无法衡量管理者在不同行情(牛市或熊市)下超过大盘的投资能力,而信息比率尽管准确度量了资产管理者超过大盘的投资能力,却无法衡量管理者总的投资能力。跟踪比率很好地克服了夏普比率(SR)和信息比率(IR)的管理能力评价缺陷。 ④提出引入风险容忍参数讨论投资组合模型。文中得到风险容忍参数的选取并不是全部正实数,而是由历史数据的期望收益率、方差-协方差矩阵和大盘收益共同确定的正实数区间,由此区间确定的最低要求收益同时也是判别投资组合是否为有效组合的必备条件。 ⑤提出了一种新的参数估计法——三步循环估计法(Repeating MaximumLikelihood,简称RML)。EML参数估计法尽管估计出的参数值精确度非常高,但EML方法在实际应用中往往很难估计出参数值。IFM方法尽管计算速度快,但其没有EML方法估计精度高。三步循环估计法即比EML方法计算速度快,也比IFM方法更有效。 ⑥选取中国证券市场中具有代表性的标的指数对本文提出模型、方法以及所得结论运用Eviews、Matlab等数据处理软件进行实证分析,实证结论与理论推导结论完全吻合。
[Abstract]:The securities investment fund has a history of more than 100 years in foreign countries, securities investment fund has become one of the three pillars of the financial investment industry in western developed countries. China's capital market started in 1980s, after ten years of development, China's securities investment fund industry has achieved great development, gradually expand the scale, but China's securities investment the fund industry started late, the market share is low, the investment management level there is a large gap compared with western developed countries. The new securities investment fund as the capital market investment strength, to improve the capital market investment environment, and guide the advanced investment philosophy, plays an important role in the investment behavior to regulate the market. Modern investment portfolio theory and method of dispersion plays a decisive role in investment and risk, the development of China's securities investment fund, to guide Is of great practical significance to rational investment of investors. Therefore, the introduction and research of modern portfolio theory and methods, to strengthen the research of macroeconomic fundamentals, vigorously develop the financial market (Yan Sheng) tool to promote the development of China's securities investment fund, to guide the perfection of our capital market, promote China's social stability and economic prosperity is of great theoretical and practical significance to maintain long-term stable development and promote China's securities market.
The results of this study are as follows:
The proposed investment portfolio model of Copula EGarch based on GPD. The EGarch - GPD model of single sample data in the AIC value to determine the optimal Copula function can fit the data distribution function, CVaR index and combine the risk measurement ability of the loss theory using Monte Carlo simulation method to calculate the optimal investment the combined weight.
The portfolio model based on expected utility maximization. Using the expected utility theory in portfolio model based on expected utility maximization, the new model of the optimal portfolio weights using the Lagrange multiplier method is easy to calculate.
It presents a new fund performance evaluation index - tracking ratio (Tracking Error; TR). From the SHARP ratio and information ratio definition, SHARP ratio only evaluates the asset managers total investment capacity, but can not measure the managers at different prices (bull or bear) more than the market investment capacity but, although the exact information ratio to measure the asset managers than the market's ability to invest, but can not measure the management ability. The total investment ratio tracking methodovercome SHARP ratio (SR) and information ratio (IR) management ability evaluation of defects.
The proposed risk tolerance parameter on portfolio model. The risk tolerance parameter is not all positive real numbers in the text, but by the historical data of the expected rate of return, variance covariance matrix and the market returns jointly determine the positive real interval, the interval to determine the minimum income and investment portfolio is essential for discrimination the effective combination.
Put forward the method of three step cycle estimation method, a new parameter estimation (Repeating MaximumLikelihood, referred to as RML) method to estimate the.EML parameters while the estimated parameter values of accuracy is very high, but the EML method in the practical application is often difficult to estimate the parameter values of the.IFM method while the calculation speed is fast, but it does not have a EML estimation method high precision. Three step cycle estimation method is better than EML method has fast calculation speed and is more effective than the IFM method.
The selection of representative mark China in the stock market index model is proposed in this paper and the conclusion by using Eviews method, Matlab data processing software to carry out empirical analysis, empirical results consistent with theoretical conclusion.

【学位授予单位】:重庆大学
【学位级别】:硕士
【学位授予年份】:2012
【分类号】:F832.51;F224

【参考文献】

相关期刊论文 前10条

1 张波;陈睿君;路璐;;粒子群算法在投资组合中的应用[J];系统工程;2007年08期

2 吴如海,宋逢明;基金分离下中国股市交易量模型的实证研究[J];管理科学学报;2000年01期

3 陈学华;韩兆洲;唐珂;;基于VaR和RAROC的保险基金最优投资研究[J];数量经济技术经济研究;2006年04期

4 曲圣宁,田新时;投资组合风险管理中VaR模型的缺陷以及CVaR模型研究[J];统计与决策;2005年10期

5 徐绪松,杨小青,陈彦斌;半绝对离差证券组合投资模型[J];武汉大学学报(理学版);2002年03期

6 陈金龙,张维;CVaR与投资组合优化统一模型[J];系统工程理论方法应用;2002年01期

7 曹勇,唐小我,曾长修;国际组合证券投资有效边界研究[J];系统工程理论方法应用;1995年01期

8 马永开,唐小我;不允许卖空的多因素证券组合投资决策模型[J];系统工程理论与实践;2000年02期

9 屠新曙,王键;求解证券组合最优权重的几何方法[J];中国管理科学;2000年03期

10 王春峰,屠新曙,厉斌;效用函数意义下投资组合有效选择问题的研究[J];中国管理科学;2002年02期



本文编号:1510570

资料下载
论文发表

本文链接:https://www.wllwen.com/guanlilunwen/huobilw/1510570.html


Copyright(c)文论论文网All Rights Reserved | 网站地图 |

版权申明:资料由用户bd0f1***提供,本站仅收录摘要或目录,作者需要删除请E-mail邮箱bigeng88@qq.com