节点文献

基金申赎流量影响因素的实证研究

An Empiricial Study on Factors Affecting Mutual Fund Flows

【作者】 田耕

【导师】 李一军;

【作者基本信息】 哈尔滨工业大学 , 管理科学与工程, 2013, 博士

【摘要】 开放式基金具有份额的开放机制,投资者可以根据基金的业绩表现情况决定其申购和赎回行为。投资者的申购和赎回行为会对基金规模造成直接影响。通过对基金申赎流量的分析和研究,能够帮助我们了解基金投资人群体的投资行为特征,这些有利于基金公司提升其投资管理能力并做好基金销售相关的市场营销工作,清楚基金投资者的申赎行为特征能够为可能出现的大规模的基金申购和赎回做好事前防备,以便及时地采取相应的应对措施,从而能够及时有效地减少基金投资管理过程中可能面临的流动性风险;此外,通过对基金投资人行为特征的了解,基金营销人员能够做出正确的营销方式,减少因不正确的营销方式给公司到来的损失。本文在总结以往文献的基础上,构建了多个基金申购赎回流量指标,并分别从单基金、总基金、分类基金三个角度探讨了基金申购赎回流量的时间序列特征,并采用面板数据模型分析了市场因素、基金自身表现特性以及投资者情绪等对单基金申赎流量的静态分析,在此基础上,本文着重分析了市场因素、基金业绩和资金流动间的动态关系并阐释了各类基金的差异性。时间序列分析发现,首先,基金申购行为表现出惯性,基金当期的申购份额受到前期申购行为的影响明显。其次,长期趋势上看,各类基金赎回份额表现情况基本一致平稳,但各类基金申购份额变化情况存在明显差异,随着时间的延续,投资者更倾向于投资于股票型基金。但基金历史的资金流动只能一定程度上解释基金后期资金流动的趋势,而对于其波动方面的特征无法有效刻画。因此,后文的重点就是通过三大类影响因素对基金申赎流量进行解释,分别讨论了单基金、总基金和分类基金间的不同特征。单基金申赎流量分析能发掘市场因素和基金自身因素的影响,对基金微观管理具有实际价值。总基金申赎流量分析则更关注基金市场和股票市场的关联性,对于关注整个基金市场资金流动情况的证券市场监管层和投资者具有借鉴作用。分类基金的影响分析能反映不同类型基金的特性对于其流量的影响。三个层次的划分能够对我国证券投资基金市场作出整体、全面的分析,使得文章更具有系统性。通过面板模型对单基金申赎流量影响因素的分析,发现单基金申赎流量受到基金投资策略、投资者结构以及市场因素等方面的影响,首先,基金当期收益率越高,会导致基金资金的大量流出,这与部分学者已有的研究结论较为一致,基金收益越好反而导致净赎回,说明国内基金存在“赎回悖论”。这不利于发挥开放式基金“优胜劣汰”的机制,会在一定程度上制约基金市场的健康发展。其次,基金持有资产中股票比例越高越能促进基金资金的流入,说明国内投资者对收益的偏好强于对风险规避的意识。再次,基金现金比例越高同样能够促进资金的流入,说明投资者倾向于基金持有一定的现金比例以保证自己的正常赎回。最后,通过分析投资者申赎份额与其滞后期的统计分析发现,投资者的申赎行为均存在一定的惯性,即投资者前期的申赎行为会正向影响之后的申赎情况。此外,透过对基金投资者不同类型的划分,我们发现过多的个人投资者可能是造成处置效应和赎回悖论的根源。而运用VAR模型对总基金和分类基金中资金流动、基金业绩、市场收益率的动态分析发现:(1)基金市场存在较强的羊群效应,投资者的申购、赎回份额在受到自身脉冲冲击后均出现一个正向的增长,已有基金投资者的投资行为引导基金市场后续投资者的资金流动方向。(2)总体基金的流量受到基金历史收益率的影响,从分离的申购、赎回行为上,基金历史收益较好并未能引起投资者更大范围的申购,但能抑制已有的基金投资者的赎回。(3)市场收益率的影响并不总是正向的。(4)基金申赎流量、基金自身收益率和股票市场收益对自身的波动处于主导地位,但长期来看,基金申赎流量对基金自身收益率和股票市场收益他们的影响有所增加。而针对不同类型的基金,虽然整体上上述结论依然成立,但也存在一定的差异性。

【Abstract】 Investor can choose whether and when to purchanse or redeem any open-endfund.As mutual fund flow will have direct impacts on the size of each fund, so itis important for banks, mutual fund companies to analyze the pattern of mutualfund flow and factors affecting the fund flow.This paper consists of seven chapters, Chapter1is introduction andliterature review; Chapter2conclude that the individual mutual fund flow ismainly depended by market factors, fund characteristics and investment strategy,then build the index system of mutual fund flow; Time series analysis isconducted in Chapter3. From Chapter4-5, this paper focuses on the factorsaffecting mutual fund flow. As an innovation compared to other literature, thispaper uses the panel data model and VAR model to analyze the factors affectingthe fund flow of individual fund, total mutual fund, and total mutual fund withinspecific type.Time series analysis show that: first, the purchase behavior exhibitdependent tendency, which means that purchase amount is affected by previouspurchanse amount. Second, for the long term, the redeem amount of differenttype of mutual fund is stable, but highly dependent on the type of mutual fund.Investors tend to purchase stock type mutual fund other than other fund.This paper argues that the factos affecting the fund flow of individualmutual fund, total fund, total fund of specific type are different. Analyzing thefund flow of individual fund will discover the impact of fund specificcharacteristics on fund flow, and have practical value for fund managementcompanies; while analyzing the fund flow of total funds will focus on thecorrelation between fund market and stock market, and have implications forstock market supervision and investment; Analyzing the fund flow of specifictype of fund will discover the impact of investment type of fund in fund flow.These three dimension analysis help systematically understand mutual fund flow.Based on a panel-data model, Chapter4analyze the three factors affectingindividual mutual fund flow-market factors, fund characteristics and investment strategy, and we find the following results:(1) fund tends to flow out from themutual fund with higher current yield,which confirm that the redemptionanalomy existed in China market.(2) fund tends to flow into mutual funds withhigher ratio of stock in total portfolio, which implies that investor prefer highyield than risk aversion.(3) fund tens to flow into mutual fund with higher cashratio in fund portfolio, implying that investor tends to invest funds which couldbe easily redempted.(4) Previous purchase behavior will have significantimpacts in purchase behavior later.(5) Individual investors show higheriratiionality that institutional investors, which may be the reason of dispositioneffect and redemption anomaly in China.In Chaper5, VAR is used to analyze dynamic relationship among total fundflow, fund yield and stock market yield. Following results have been obtained:(1)there is existed significant herd effect in mutual fund market, the purchase andredemption show positive impulse feedback after a shock to the variable itself.Furthermore, the fund flow of most mutual fund show the same direction, alsocomfirm the herd behavior in mutual fund investment.(2) fund flow of totalmutual funds is affected by fund historic rate of return. Better performance in thepast could not induce greater purchase, but prohibit redemption.(3) the effect ofmarket yield on fund flow is heterogeneous, not always positive.(4) themovement of mutual fund flow, rate of return of mutual fund, market yield iscaused mainly by the variable itselves, but in the long run, the mutual fund flowplay an increasingly role in the movement of fund return and stock market return.We also conducted impulse feedback and variance decomposition to mutual fundflow of four different types of mtual fund in this chapter, finding that:(1) thefund flow of the four types of fund all have shock effect to later fund flow, butthe degree is different.(2) there existed difference among the performanceconsistence of the four types. Stock type and hybrid type mutual fund havepoorer performance consistence, so investor tend to redeem the fund when itsperformance is relatively higher as they expect the good performance could notcarry on. While, bond type fund has better performance consistence, so hasweaker disposition effect.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络