流动性与股票组合投资管理研究.docx

1、流动性与股票组合投资管理研究上海交通大学博士学位论文流动性与股票组合投资管理研究 姓名:姚亚伟申请学位级别:博士 专业:金融学指导教师:杨朝军20090201摘要流动性与股票组合投资管理研究摘要在传统的均值.方差模型中,“市场流动性是充分的”假设使投资者忽略了流动性在组合 投资管理中的重要性。流动性作为金融资产的三大属性之一,体现并作用于组合投资管理整 个过程中。本论文尝试性的将流动性凶素引入到传统均值.方差模型下最优组合的构建过程 中,从一种新的视角将组合选择的思维从二维空问拓展至三维空间,从而丰富了现代资产组 合选择理论。论文的研究围绕着流动性和组合投资展开,从探讨流动性的股票特征到流动性

2、 指导组合投资构建,从理论角度和实证角度对流动性在组合投资管理中的作用进行了研究。 论文的主要内容及结论如下:首先,对流动性股票特征的认识(面向组合选择的对象。在对流动性内涵认识的基础 上,提出股票流动性的本质是股票与现金之间的相互转化能力,满足的是投资者正常的交易 需求。论文探讨了:(1股票的流动性、公司规模及成交金额之间的变化关系,结合中国证 券数据进行实证,研究结果表明:公司规模一定,可实现的成交金额越多,流动性越好;成 交金额一定,公司规模越大,流动性越好; (2股票价格与股票流动性之间的关系,实证 研究结果表明:股票的价格与股票流动性之间具有显著的正相关关系,即流动性越好,股票 价格

3、越高; (3流动性水平的稳定性,实证研究表明,市场处于稳定上涨或下降状态时, 市场的流动性相对稳定,当市场处于波动时,流动性水平波动较大且相对不稳定,但从中长 期看,流动性水平在一定时期内是具有稳定性的,并利用均值.回归模型检验了流动性水平 的稳定性。其次,如何将流动性引入到组合投资管理中(面向组合的构建。主要讨论了:(1 组合的规模问题。基于流动性的内涵,组合投资越分散,组合的流动性越好,但投资者对组 合中股票的流动性并非同时或持续有需求,组合规模没有必要过度分散化。进而从分散风险 的角度,利用连续T检验的方法,科学的测算了组合中合理的股票数目; (2基于流动性 的内涵,从对流动性在组合投资

4、中的事前和事后认识,讨论了流动性引入到均值-方差模型 的方法。具体包括:理论角度。考虑单期投资下,流动性对组合投资的最终作用结果就组 合变现时的执行成本,从期末期望财富效用最大化的角度,讨论了流动性和流动性风险与期 末期望财富效用之问的函数变化关系,并探讨了流动性引入对期末财富效用的影响,研究结 果表明:考虑流动性并不必然降低期末期望财富的效用;实证角度。将流动性作为对收益 的影响因子间接引入和作为独立的维度直接引入到传统的均值.方差模型中。流动性间接引 入模型下,通过构建流动性影响因子,对股票的收益率进行调整,并依据调整后的收益率在 传统的均值.方差框架下进行组合最优化的求解;而流动性直接引

5、入模型下,主要思想是考上海交通大学博士学位论文虑到组合管理过程中投资再平衡的需要,从流动性过滤、流动性约束、流动性偏好下的效用 最大化将流动性直接引入到均值.方差模型,论文给出了这三个模型取得有效边界时的有效 解集集合,验证了流动性、收益和风险实现均衡时的有效边界是一个曲面。从实务投资的 角度,提出了引入投资政策约束、行业配置比例约束下修正的流动性一均值.方差模型。 第三,流动性作用下的组合投资管理(面向组合的管理过程及结果。主要基于流动性 引入均值.方差的实证模型,分别从直接引入和间接引入下组合投资的有效边界进行实证研 究,主要内容包括:在直接引入模型下,通过选取A股市场市值最大的10只股票

6、作为研 究对象,通过比较考虑流动性的价格冲击影响效应和未考虑流动性价格冲击影响效应的组合 有效边界,实证结果表明:收益经流动性调整后,组合的有效边界向左上推移,这意味着:一方面投资者可以在承受更低的风险下实现一定的收益;另一方面,考虑流动性对收益的影 响,在同一风险水平下,投资者可以获得比不考虑流动性影响时更高的收益。在直接引入 模型下,通过选取由上证50指数成份股的股票构建的组合(高流动性组合和从未入选上 证50指数成份股的股票构建的组合(低流动性组合及两者共同构建的组合(中流动性组 合,用流动性过滤、流动性约束和流动性偏好下的目标函数效用最大化三种流动性.均值一 方差模型分别考察了流动性在

7、组合投资选择时的作用。主要研究结论包括:在流动性过滤 下的均值.方差模型下,研究结果表明:在组合投资过程中,同时配置高流动性和低流动性 的股票,可以实现在风险一定条件下实现比仅投资高流动性和仅投资低流动性的股票组合更 高的收益,这体现了流动性对组合投资资产配置策略的影响作用;在流动性约束下的均值 .方差模型下,研究结果表明:在流动性水平约束一定的条件下取得最优组合时,三个组合 的收益均随着风险的增加而增加,但组合的夏普比率随风险的增加呈倒“V”型;在风险水 平一定下取得最优组合时,组合的收益和夏普比率随着流动性水平约束的增加均呈倒“V” 型,但变化很不规则:尽管高流动性组合和中流动性组合在达到

8、最优组合时组合的收益均随 着流动性约束的提高而下降,且风险越高,组合的收益越高,但在低流动性组合中,在面临 不同的流动性水平约束时,并不能在遭受较高风险时获得较高的回报。同时,风险水平一定 时,组合夏普比率的变化也表明,在不同的流动性约束水平下,风险较高的组合不一定能实 现较高的夏普比率。而巾流动性组合则弥补了高流动性组合和低流动性组合在收益随风险或 流动性变化的不规则问题。在流动性偏好下的期望效用最大化模型下,研究结果表明:高 流动性组合与低流动性组合尽管在夏普比率等指标变化上具有一致性,但在风险厌恶、流动 性偏好下不同组合的收益、风险及非流动性之问仍存在较大差异。这可能在于高流动性股票 和

9、低流动性股票在流动性水平和风险上存在着系统性的明显差异。第四,流动性、行业效应与组合投资(面向组合的应用。从行业收益之间的相关性和 差异性、流动性之间的相关性和差异性探讨了行业筹异的表现特征,研究结果表明:我国股 票市场存在着明显的行业效应;对国内开放式基金的行业配置集中效应及基金行业配置与市 场上行业构成比较进行了统计分析,统计结果表明:我国基金投资的行业集中度在50%左摘要右,基金的行业偏好之间存在着差异,而且基金的行业配置与不同行业占有市场的比例相比 除采掘业、金融保险业、传播文化业和综合业超配外,其它行业均处于低配; (3考虑流 动性、行业效应与组合投资的关系,比较了在有无投资政策约束

10、时行业配置对组合投资的影 响,研究结果表明:不考虑投资政策约束将会导致行业集中持有风险;利用国内证券市场的 A基金的实际持有股票组合情况,对在流动性约束、投资政策约束及行业配置约束下组合最 优解进行了实证研究,研究结果表明:组合的风险和收益随着非流动性水平的提高先降后升, 但组合的单位风险收益会随着非流动性水平的提高先升后降,这意味着当组合的非流动性水 平达到一定水平时,组合的风险将加速恶化,从而导致单位风险的收益水平下降。关键词:组合管理,流动性,行业效应上海交通大学博士学位论文Research o n the Liquidity and Portfolio InvestmentMa nag

11、ement of StocksAB STRACTIn the traditional mean-variance model,the assumption ofrhe market liquid时if sufficient” leads investors to ignore the importance of liquidily in the portfolio management.As one of three attrmutes of financial products,liquidly reflects and effects on the whole process of por

12、tfolio investment management.The paper tried to introduce the effect of liquidity to the optimal portfolio construction process under the traditional m以lllvariance mod“and explored the portfolio choice thought厅om two dimensions to threedimensions for new viewpoint,enriched the modern portfolio manag

13、ement.The outline of this paper toward liquidity and portfolio investment, from exploring the stock character of liquidity to used liquid时to comtruct portfolio,from the theory and empirical point of view to research the effect of liquidity in the portfolio management. The main contents and conclnsio

14、ns as follows:First,understand the stock character of liquidity(for the objectives of portfolio choice.On the recognized of eSsenee of liquidity,we thought the essence of liquidity Was the multual-convert ability between stock and cash,and liquidly was satisf圮.xl for the normal trading demand of inv

15、estors.We explored:(1the multi-relationship among liquidity,company size and trade dollar, and used stock market data of China to give empirical analys is,the results showed:under certain company size,if the stock could afford higher trading dollar,the better liquidity;and under certain trading doll

16、ar,the larger of the company size,the better liquidity.(2the relationship between stock priceand liquidity,the empirical results showed that:the higher the stock price,the better liquidity.(3the stability of liquidity,the results showed that,when the market WaS at increasing or decreasing conditions

17、,the liquidity Was stable,but if the market was violate at volatility stage,the liquidity Waq not stable,but in the mediumand longterm,liquidity had stabflity character; furthermore,we used the meanreverse model to robust test the stabillty character of liquidity. Second,introduce liquidity into por

18、tfolio management(for the construction of portfolio.The paper main discussed:(1the numbers of stock in one portfolio.Based on the essence of liquidity, the more divers ify,the better liquid ity,but the investors had no continuously or simultaneously demand stocks generally,SO it Was not necessary fo

19、r over-diversify.From the diversify the risk,we used the continuously T test,estimate the proper numbers ofstock in one portfolio scientific;(2 based on the essence of liqu id ity,from the ex ante and ex post to recognize liquid ity,we discussedthe methods ofintroducing liquidi妙intomeanvariance mode

20、ls,which included:from the theory point,we considered oneperiod investment,the final effect of liquidity to portfolio investmentAbstractwas executing cost,assumed the maximization the expected utility of end wealth,we discussed the function relation among liquidi吼liquidity risk and expected utilily

21、of endperiod wealth,and discussed the results when introduced the liquidity into the expected utiIity of endperiod wealth, the results showed that:considering liquidivy may not reduce the expected utility of end-period wealth.From the empirical poinL we introduced the liquidity to meanvariance model

22、 indirectly and directly.Under the indirectly method,the paper constructed a liquidiIy adjustment coefficient, used the return after liquidity adjustment to construct optimal portfolio under the mean-variance model;under the directly method,the main point was considering the demand of portfolio reba

23、lance,we used liquidityfilter,liquid时constraint and the maximization util时under taking the liquidiIy into account to introduce liquidity into meanvariance model dkect鼽the paper gave the efficient solution sets of the three models,and checked efficient frontier was surface when return,risk and liquid

24、时reached equil南rium;(3from the practical investment point,we considered the liquid时一mean-variance model under investmem policies constraint and industry constraint.Th砌,portfolio investment management under liquid时effect(for the management process and results of portfolio.Based on the empirical model

25、 under liquidity,we gave empirical analysis of the efficient frontier of portfolio by indirectly and directly method.The main contents included: from indirectly poirIt,we chose the top 10largest market value listed companies,compared the efficient frontier under or with no considering the liquidity

26、effect,the results showed that after the liquidity adjustment,the efficient frontier moved to left-top,this meaned the investors could get return under bearing lower risk,meanwhile,the investors could get higher return under certain risk with the liquid时effect.From the directly po缸,we chose the comp

27、osite stocks of Shanghai 50 index as higher liquidiIy group,and chose the same nmabers of stock which never was Shanghai 50index randomly as lower liquidily group,the stocks in both higher and lower liquid姆groups constructed medium liquidivy group.We used the liquidityfilter,liquidity constraint and

28、 the maximization utility under taking the liquidity into account to analyze the effect of liquidity on portfolio choice,the main results included:under the liquid磅filter model the resultsshowed to allocate high and low liquidity stocks in portfolio could get higher return than only high or low liqu

29、idity stocks portfolio under some certain risk level this reflected the effect of liquidivy in the asset allocation strategieschoice.Under liquidity constraint,the results showed that under the some certain liquidity constraint,the returns of the optimal portfolios will increase as the risk,but the

30、sharpe ratio change V character as the risk increase.Under the some certain risk level the return and sharpe ratio was V character as the liquidity constraint increase,but changer irregularly, aRhough the return of high and medium portfolio decrease as the liquid ity constraint increase,for low port

31、folio,when suffering different liquid ity constraints,it could not get high return even suffering higher risk.Meanwhile,under cegain risk levels,the change of sharpe ratio showed that under the different liquid ity constraint leve Is,the higer risk portfolio may not get the higher sharpe ratios.Unde

32、r the maximization expected util睁with liquidil,preference,the results showed that 第v页上海交通大学博士学位论文thechange of sharpe ratios of highand low porffoho had the some direction,but under certain risk aversion and liquidi哆preference,the return,risk and illiquidity still had large differences.This may became the high and low liquid时stocks have systematic differences at liquidity levels and risk levels.Fourth,liquidib,industry effect and portfolio(for the appliance of portfolio.The paper analyzed the difference of industries by analyzing the correlation coefficients and difference of return and liq