【作者】 彭丽娜；

【导师】 杨湘豫；

【作者基本信息】 湖南大学 ， 概率论与数理统计， 2007， 硕士

【摘要】 随着我国基金业的蓬勃发展,基金已经成为普通老百姓投资的一个重要的渠道,特别是开放式基金相比封闭式基金具有更高的流动性、更好的市场运作与管理机制,已经成为国际基金市场的主流产品。对于开放式股票型基金而言,它以股票投资为主,投向于股票市场,其市场风险是基金管理公司所面临的最主要的风险,并且市场风险也是最基本、最常见、最难防范的风险,它的恶化往往导致其他风险(如流动性风险)的产生,它关系到我国开放式基金、乃至整个证券市场未来的持续健康地发展。因此本文选择研究开放式股票型基金的市场风险度量问题具有一定的理论和实践意义。对于市场风险的研究不仅可以帮助基金管理人更有效、更科学地管理好所管理的基金;对投资者来说,也可以使用投资模型选择适合自身要求的基金,进行组合投资,以此做到分散风险、提高收益,使投资行为更加理性化。本文以现代金融风险理论为基础,采用定性和定量相结合,以定量分析为主,对开放式基金,主要是对股票型基金投资组合的市场风险度量的VaR方法和非线性分形理论进行深入研究,最后进行实证分析。在实证部分,本文选择华安创新证券投资基金(040001)这只股票型开放式基金作为研究对象,它是我国第一只获准公开发行的开放式基金,对其及其组合中重仓股的收益率序列进行分析。研究发现,股票及其组合收益率不能用正态分布来拟合,收益率序列具有显著的尖峰厚尾性。通过R/S分析,求出该基金的Hurst指数大于0.5,遵循有偏随机游动,平均循环周期长度约为16个交易日。本文引入分形分布对收益率序列的分布进行拟合,并给出了分形分布下VaR的求解方法,同时通过半参数法计算出了基金组合VaR的近似置信区间,并结合边际VaR、成分VaR对基金组合VaR进行了分解。本文借鉴现有的风险度量研究成果,在前人的基础上力图有所突破,主要有以下一些创新:1)提出应用基于分形理论的Hurst指数来度量基金市场风险。2)在VaR的计算中引入分形分布,并估计出分形分布的各个参数,由此给出基于分形分布的VaR参数计算法。在此基础上,采用条件均值估计法对基金组合VaR进行有效分解。更多还原

【Abstract】 Along with the prosperous development of China’s mutual fund industry, mutual funds have become an important investing channel for common people. Especially open-end funds have become a mainstream product in the international fund market, for higher liquidity, better management and operation of the market mechanism compared to closed-open funds. For stock open-end funds, which invest mainly in the stock market, market risk is the largest risk which fund management companies in the face of. Also market risk is the most basic, common and the hardest risk to keep away. Its deterioration leads to other risks (such as liquidity risk) and it is vital to our open-end funds, as well as the stabilization and healthy development of the entire securities business. So this thesis researches on measuring the market risk of stock open-end funds, which is of theoretical and practical signification. Research on market risk can help not only fund managers more effectively and scientifically manage the fund, but also investors rationally choose their own investment requirement of the fund, the portfolio so as to diversification of investment and improvement of returns. Based on the theory of modern financial risk, this thesis uses methods of qualitative and quantitative, especially quantitative method to measure the market risk of stock open-end funds. We make use of Value-at-risk and nonlinear fractal theory to lucubrate. Finally we do a demonstrative study.On the demonstrative study, we choose Hua’an Innovation fund (040001)as the object of study, which is China’s first open-end fund. Through R/S analysis, we compute Hurst exponent of portfolio is greater than 0.5, following biased random walk process, and average cycle length of it is 16 business days. We induct fractal distribution to fit returns which have fat tails and excess kurtosis, and give the equation of VaR based on fractal distribution. Using a new semi-parametric approach, we obtain an approximate VaR confidence interval and make use of M-VaR and C-VaR to analyze the total risk of portfolio.This thesis uses the result of risk measurement research for reference, in order to achieve a breakthrough about it. These are some innovations: 1) Give the Hurst exponent to measure the market risk of stock open-end funds based on fractal theory. 2) Introduce fractal distribution in the calculation of VaR and estimate four parameters of the fractal distribution. Decompose portfolio VaR by condition mean eatimation.更多还原