【作者】 寻明辉；

【导师】 陈琦伟；

【作者基本信息】 上海交通大学 ， 金融学， 2006， 博士

【摘要】 在中国鼓励和培育机构投资者的政策背景和证券市场迅速发展的态势下,投资基金已经成为中国证券市场上备受关注的一类机构投资者。做为基金业创新发展的主流,开放式基金更是成为大众理财的重要工具。截止2006年1季度,中国已有基金管理公司53家,管理资产约5000亿元人民币,管理的开放式基金数量已达到146只,基金产品种类也在不断地丰富和完善。然而,我国资本市场发展的现状决定了开放式基金不得不面对高流动性的资金来源与低流动性的资产相匹配而引发的结构性矛盾,基金的流动性风险格外突出。怎样识别基金的流动性风险以及如何控制和防范流动性风险,成为金融学家和基金管理者无法回避的现实问题,对这些问题的回答和关注是本文研究的出发点。值得注意的是,现有关于流动性风险的文献大都基于既定的报价驱动机制,这和我国股票市场指令驱动的交易机制存在较大的差别,从而使传统的流动性度量指标失去了前提条件和理论支持。其次,关于流动性风险管理的研究,分散于风险管理的不同环节,且不同环节之间缺乏有机的联系。另外,从开放式基金管理实践出发进行流动性风险管理研究的研究成果也比较少见。这使得现有的研究结论对开放式基金的管理实践缺乏一定的现实指导意义。考虑到中国股票市场的指令驱动特性,以及开放式基金的管理实际,本文从风险识别、风险控制和风险防范角度对开放式股票投资基金流动性风险管理展开了系统研究。在理论研究方面,本文对股票价格冲击分类信息混合分布GARCH模型的研究,丰富和发展了现有的流动性度量方法和理论体系,同时,本文提出了能够使基金变现过程中的流动性风险和变现结束时的市场风险整体最优的流动性风险控制新思路;在实证研究方面,本文采用大量股票交易高频数据进行流动性相关测算,以实现对股票流行性的准确度量,而且,力求对文中所提出的每一量化结论都进行实证检验,以使研究结果能够更容易和更有效地应用于开放式基金管理实践。本研究的主要工作及由此形成的结论包括:一、定性分析了开放式基金流动性风险形成机制,总结了开放式基金流动性管理的实践历程和理论背景,确定本文研究方向和内容。二、综述了国内外流动性风险管理的相关研究成果,重点介绍了开放式基金规模变动、流动性影响因素、流动性风险的度量和控制等方面的研究成果和结论,并分析现有研究的不足,为论文的后续研究提供了理论基础。三、在股票流动性研究方面,采用标记法和报价法相结合的交易类型识别方法对成交指令流进行计算,在此基础上建立了成交量对价格冲击的一般微观结构模型,测算出上海A股市场30只成分股票的可变价格冲击系数和固定价格冲击系数,进一步,用交易所报价系统的指令簿信息代替混合分布GARCH模型方差方程的成交量作为解释变量,同时保持成交量在价格变动回归模型中的解释作用,构建了股票价格冲击分类信息混合分布GARCH模型,通过对样本股票的实证分析,证实该模型对股票价格冲击系数的测算是有效和可信的,而且该模型对股票价格冲击的解释能力明显强于一般线性价格冲击模型。这不仅丰富和发展了现有的流动性度量方法和理论体系,更重要的是为开放式基金流动性风险的识别奠定了基础。四、在开放式基金流动性风险识别方面,分别从外生流动性风险和内生流动性风险两方面展开分析。首先,在外生流动性风险方面分析了影响开放式基金规模变动的相关因素,提出和解释了收益指标、成本指标、风险指标、红利指标等因素对基金规模变动的作用;其次,通过量化指标对开放式基金内生流动性风险加以识别,在现有VaR模型的基础上,引入流动性指标和交易策略指标,构建一个流动性风险度量模型—L-VaR,通过Monte Carlo模拟法对L-VaR值进行测算,发现基金选择不同的变现股票、变现数量和不同的交易速度时会对流动性风险值产生明显的影响,这一结论为开放式基金流动性风险的控制研究提供了解决问题的出发点,具有重要的指导意义。五、从开放式基金流动性风险控制角度出发,构建了流动性风险控制的理论框架;给出连续时间框架下的流动性风险和市场风险的数学表达形式,对流动性风险最优控制策略进行求解,首先,利用目标规划方法确定需要变现股票的头寸,其次,利用随机动态规划方法确定股票在变现期内的最优变现策略;通过对流动性风险控制策略进行符值演算,发现组合中股票的最优变现策略同时受价格、流动性和波动性等因素的影响。六、在开放式基金流动性风险防范研究方面,主要从“均衡”管理、资金来源、费率结构和制度设计几个方面进行基金流动性风险防范分析,并结合国际先进经验,得到了一些有力于风险防范的措施。本研究的创新性体现在以下三个方面:一、提出了股票价格冲击分类信息混合分布GARCH模型,来分析股票价格与交易量之间的变化关系,并利用股票交易高频数据计算得到单个股票的价格冲击系数,通过实证研究分析,确认该模型优于价格冲击的一般微观结构模型,能够为基金流动性风险识别提供更有力的支持。二、提出L-VaR模型对开放式基金流动性风险进行定量评价,并设计出基金的均匀变现策略、流动性优先变现策略和累计价格冲击最小变现策略,得出不同变现策略对应的流动性风险具有显著差异的结论。三、构建了综合考虑基金变现过程中的流动性风险和变现结束时的市场风险的流动性风险控制新框架体系,通过目标规划和动态规划方法得到了开放式基金流动性风险控制的最优策略。总的来看,本文针对开放式股票投资基金构建了一个包括风险识别、风险控制和风险防范的完整流动性风险管理体系,不仅从理论上对流动性管理理论本身进行了丰富和发展,而且从我国开放式基金流动性风险管理实际出发,做了大量的实证研究,得到了一些有价值的结论。本文研究对于开放式基金更有效地进行风险管理、保护投资者既得收益、提高基金管理水平,促进基金业健康稳定发展具有重要的理论和现实意义。更多还原

【Abstract】 With the policy of encouraging and fostering institutional investors and the rapid development of stock market in China, mutual fund has become one kind of institutional investor that draws a great deal of attention. As the mainstream of the innovative development of fund industry, open-end fund has become the important financial means for the people. By the first quarter of 2006, there had been 53 investment fund companies, managing approximately 500 billion RMB, while the open-end fund had amounted to 146. Meanwhile, the variety of the financial products has also been continuously enriched and updated. However, under the current state of capital market, open-end fund has to be confronted with the structural conflict resulted from the incompatibility between the capital with a high liquidity and the assets with a low liquidity, which highlights the liquidity risk of the fund. How to identify, control and prevent the liquidity risk of the fund has been a realistic question that both financial experts and fund managers can not evade, and therefore is the focus of this dissertation.What should be pointed out is that the study liquidity risk in the current literature presupposes the price-driven mechanism while China’s stock market is a typical order-driven market. As a result, the traditional measurement of liquidity risk lacks the presupposition and theoretical foundation. Furthermore, With regard to liquidity risk management, most studies only focus on one or another stage of liquidity risk management, while ignoring the organic relations between different stages. Finally, the study of liquidity risk management departing from the practice is comparatively scarce, the conclusion of which can not be readily applied to the practice of open-end fund management.Take account of the order-driven nature of China’s stock market and the practice of open-end fund management, this dissertation conducts a systematic research on the liquidity risk management of open-end equity fund in terms of risk identification, risk control and risk prevention. Theoretically, the study of impacts on stock price through mixed distribution classified information GARCH model enriches and further enhances the methodology of liquidity measurement as well as the theoretical system; At the same time, this research put forward a new way of liquidity risk control, that is, optimizing integrally the liquidity risk in the fund-cashing process and the market risk at the end of fund-cashing. Empirically, this study utilizes the high frequency data of stock trade with a view to ensure the correct measurement of stock liquidity; Also, this study strives to test empirically every qualitative conclusion, so as to apply the conclusion to the fund management practice more easily and effectively.Main tasks and conclusions of this dissertation:Firstly, this study analyzes qualitatively the liquidity risk formation mechanism of open-end fund, sums up the practice and theoretical background of open-end fund liquidity management and then set the purpose and content of this research.Secondly, this study summarizes the relative researches on liquidity risk management domestically and abroad, with the emphasis on the open-end fund scale change, the factors affecting liquidity, measurement and control of liquidity risk, etc., and furthermore points out the limitation of the present researches and set the theoretical foundation for the further research.Thirdly, with regard to stock liquidity, this research combines the two approaches to identifying trade type, that is, tack method and quote method, and calculates the trading order flow, based on which this research establishes a general microstructure model of trade volume‘s impact on stock price and calculates the changing price impact coefficient as well as the fixed impact coefficient of 30 index-component stocks of Shanghai A-stock market; Furthermore, this research substitutes the information from order books for the trade volume in the variance equation of mixed distributional GARCH model to explain the variants, while meanwhile maintain trade volume in the regress equation. In this way, a stock price impact classified information mixed distribution model is constructed. The empirical analysis of sample stocks proves the validity and reliability of this model in terms of measuring and calculating the price impact coefficient. It is also found that this model can better explain the stock price’s impact than the common linear price impact model. Thus, the existing methodology of measuring liquidity and theoretical system are greatly enriched and enhanced, and more importantly, a solid foundation is set for identifying the liquidity risk of open-end fund.Fourthly, regarding liquidity risk identification, this research considers both the exogenous liquidity risk and endogenous liquidity risk. Firstly, as for exogenous liquidity risk, this research analyzes the relevant factors affecting the scale change of open-end fund and account for the impact of some factors, including income index, cost index, risk index, dividend index, and etc; Secondly, this research identifies the endogenous liquidity risk of open-end fund through quantitive index. Based on the existing VaR model, a new liquidity risk measuring model, namely, L-VaR is established. Tested through Monte Carlo simulation, it is found that cashing stock, cashing quantity and trading speed will have an obvious influence on the value of liquidity risk, which offers a starting point for the research on liquidity risk control of open-end fund and is therefore highly inspiring.Fifthly, as to liquidity risk control, this research establishes the theoretical framework of the liquidity risk control and, to calculate the optimum liquidity risk control strategy, give an mathematic expression representing the liquidity risk and market risk in continuity. We will first decide the quantity of the necessary cashing stock by means of objective-planning and then make sure the optimum cashing strategy in the cashing process by means of randomly-dynamic-panning; Through empirical testing, it is found that the optimum cashing strategy of the portfolio is affected simultaneously by price, liquidity, fluctuation and etc.Sixthly, with regard to liquidity risk prevention, this research bases its analysis mainly on the aspects of balanced management, capital source, fee structure, institution design and so on, and then taking reference from the advanced experience abroad, concludes some helpful measures to prevent risk.The innovations of this dissertation are as follows:Firstly, this study establishes the stock price impact classified information mixed distribution GARCH model to investigate the correlation between stock price and trade volume, and calculates the price impact coefficient based on the high frequency data of stock trade. The empirical testing proves the superiority of this model compared the general microstructure model, thus offering a strong support for the identification of liquidity risk.Secondly, this study conducts a quantitive analysis of the liquidity risk of open-end fund by means of L-VaR model, and designs three cashing strategies of the fund, namely, even-distributed cashing strategy, liquidity-first cashing strategy and minimizing-accumulative-price-impact cashing strategy. It is found that the liquidity risk corresponding to each cashing strategy is significantly different. Thirdly, this research establishes a new liquidity risk control framework by taking account of both the liquidity risk during the cashing process and the market risk at the end of the cashing process. The optimum liquidity risk control strategy of open-end strategy is achieved by means of both objective-planning and dynamic-planning.To sum up, this dissertation sets up a comprehensive system of liquidity risk management of open-end equity fund, taking risk identification, risk control and risk prevention into account, which not only enriches and advances the liquidity management theory, but also conducts a great quantity of empirical researches departing from the liquidity risk management practice of china’s open-end fund industry and draws some valuable conclusions. Therefore, this research has important theoretical and practical value for managing liquidity risk, safeguarding the vested interest of investors, enhancing fund management and promoting the healthful development of the fund industry.更多还原