文本課件教程lecture10fund flows

1、Lecture 10: Fund flows and other fund characteristicsMN20026 Investment ManagementFlow of Funds Empirical strong relation between past performance and flow of funds: Sirri & Tufano (1998) Chevalier & Ellison (1997) Del Guercio & Tkac (2002) But why do fund flows follow past performance? Away from po

2、or performers To good performers Berk & Green (2004) model relationship between past performance and mutual fund flows Flow of funds is mechanism to ensure no performance persistence Also Gruber (1996) “Follow the smart money” Do fund flows predict outperforming funds?MN20026 Investment ManagementEv

3、idence on Performance Persistence in mutual funds Some empirical evidence on performance persistence Grinblatt & Titman (1992) Brown, Goetzmann, Ibbotson & Ross (1992), Brown & Goetzmann (1995) : persistence explained by survivorship biases Hendricks, Patel & Zeckhauser (1995) Tonks (2005) Bollen an

4、d Busse (2005): Short-term But Carhart (1997) shows that after controlling for FF-factors + Momentum & fees, no evidence of persistence, after 4-factor correction Except in poor performing firms fees explain persistence of net returns:- if fees are high, then funds that charge high fees will have pe

5、rsistently poor performance of net returnsMN20026 Investment ManagementSirri & Tufano (1998)MN20026 Investment ManagementNote denominator might also be inflated by returnSirri & Tufano (1998)AssessesFLOWSit =f(Returni,t-1, Riski,t-1, Expensesi,t-1, . . .)FLOWS are computed over a one-year horizon, a

6、nd Returns over the previous one-yearRanks funds on the basis of past performance, and forms quintilesPiecewise linear regressionFinds that fund flows are sensitive to past performance, but relationship is highly non-linearConvex relationship between flows and past performanceMN20026 Investment Mana

7、gementSirri & Tufano (1998)MN20026 Investment ManagementImplications of convex performance-flows relationshipAnnual tournament: top-performing fund gets all next years money Incentive for money managers to take perverse actions Such as managers of “l(fā)oser funds” taking more risks? Brown, Harlow and S

8、tarks (1996) examine riskiness of portfolios of the 50 per cent of investment funds who perform above average in any year (winner funds with riskiness 1W) and the 50 per cent of loser funds 1L. They expect:2L/2W 1L/1WExamine performance of 334 growth-oriented mutuals 1976-1991 and find mid-year lose

9、rs increase fund volatility in the latter part of an annual assessment period to a greater extent than mid-year winners See also Chevalier & Ellison (1997)MN20026 Investment ManagementMutuals vs PensionsDel Guercio and Tkac (2002) compare performance-flows sensitivity for mutuals & pension fundsThey

10、 find mutual fund flows are convex function of past performance, but pension fund flows are notThey suggest this is because pension fund sponsors are more sophisticated than mutual fund investorsSuggest pension fund managers are less likely to risk-shift, than but mutual fund managers have a strong

11、incentive to take risks to try to be in the top percentile. MN20026 Investment ManagementDel Guercio and Tkac (2002)MN20026 Investment ManagementMutual fundsPension fundsEvidence on flow-performance sensitivitiesMN20026 Investment ManagementSource: Goriev, Nijman and Werker (2008, J. Fin Mkts)Fund f

12、lows are most sensitive to past performance measure over previous 6 months:Top figure = 1990sBottom figure = 2000sBerk & Green (2004)Paper is motivated by asking why do fund flows follow past performance if there is no performance persistence?Distinguish between gross returns (Rt ) and net returns (

13、rt ): rt = Rt c(qt ) where c(qt) is cost function for fund management, and qt are assets under management c(qt ) = f + C(qt )/qt is the unit cost from active fund management and is made up of salary costs f which are a fixed proportion of assets under management so total salary costs = f*qt and C(qt

14、) is convex, reflects idea that it is more difficult to actively manage a large fund: trading costs of large price impacts etc. MN20026 Investment ManagementBerk & Green (2004)Allows for differential managerial ability such that Rt = + t,and differs across managersInvestors update on past performanc

15、e high return likely to be from better managersFunds flow into those mutual funds who demonstrate high returns q0, and out of those funds with poor performance q qt-1, but because of cost function convexity: performance will worsenUnderperforming funds will shrink in size qt qt-1, , making it easier

16、 for funds to improve performanceNotice that since fund manager fees are typically a percentage of AUM, the better managers will manage larger funds, and will receive higher fees (and salaries) commensurate with their higher skillsHence high-ability managers earn the “economic rents” from their scar

17、ce skills, but outside investors can only expect to earn normal rates of return on their investment fundsMN20026 Investment ManagementEvidence for Berk & GreenChen, Hong, Harrison and Kubick (2004, AER) find that size is detrimental to fund performancePollet and Wilson (2008, JF) find that net inflo

18、ws into funds are used to “scale up” existing strategies which is likely to incur scale diseconomiesBut Fama & French (2010) challenge the findings of Berk & Green and argue there is no evidence of managerial skillMN20026 Investment ManagementFund Manager TurnoverKhorana (1996) suggests outflows and

19、 managerial dismissal are alternative external and internal control mechanisms to discipline under-performing fund managers:Using sample of 339 funds that replaced their managers 1979-1992, and matching control group of 4,830 funds (with no change) finds -ve relationship between the probability of m

20、anagerial change and past performanceKhorana (2001) examines the effect of the change in manager on a funds subsequent performance. Using sample of 393 domestic equity and bond fund managers that were replaced over the period 1979-1991, he finds Underperforming funds subsequently improve post-replac

21、ement performance with abnormal performance improving from -2.4% in the year before replacement to 0.5% in the third year after replacement. Outperforming funds suffer deterioration in post-replacement performance, from 1.9% in the pre-replacement period to 0.4% in year +3 MN20026 Investment Managem

22、entSmart moneyDo fund flows predict outperforming funds?Can investors identify superior mutual funds? Gruber (1996, JF) and Zheng (1999, JF) find that funds that receive the largest net money flows subsequently outperform other fundsThis pattern was termed the “smart money” effect. However more rece

23、ntly research, however, Sapp and Tiwari (2004) find that after fund performance is adjusted for the momentum factor in stock returns, greater net flows no longer lead to better performance.See also Keswani and Stolen (2008, JF)MN50322 Investment ManagementFund management around the worldKhorana, Ser

24、vaes and Tufano (2005, JFE) attempt to explain the size of the mutual fund Industry in 56 countriesFinds mutual fund industry is larger in countries: With stronger rules, laws, and regulations, andBetter protection for mutual fund investors rightsThat are wealthier and have more educated populations

25、, Where the mutual fund industry is older, Where trading costs are lower and In which DC pension plans are more prevalent. Mutual fund industry is smaller in countries where barriers to entry are higher.These results indicate that laws and regulations, supply-side and demand-side factors simultaneou

26、sly affect the size of the fund industry.MN20026 Investment ManagementCosts of Fund Management Annual management charge (AMC)Percentage of assets under managementFront end fees Exit fees (particularly with hedge funds)Total expense ratio (TER)Also hidden costs of trading securitiesInvestment Managem

27、ent Association (IMA) estimates 481 billion in managed UK domiciled funds (unit trusts & OIECS) in 2008So 1% AMC generates approx. 5 billion p.a.MN20026 Investment ManagementPlexus IcebergExplicit (Hard) Costs (18%) Commissions9 bp Taxes & Fees Custodial ChargesPotential Hidden Costs (82%) Bid/Ask S

28、pread Market Impact12 bp Information leakage Opportunity Costs9 bp Delay Costs21 bpSource Plexus Group: Maggin et al (2007) Chapter 10: estimates for 2005Explicit (Hard) CostsFund management feesCommissionsTaxes (Stamp duty/Tobin tax)Custodial chargesAccording to GIPS not to be included in computati

29、on of net returnsPotential Hidden CostsAt tradeBid-ask spreadPre-trade:Market impactPre-hedgingInformation leakageShopping order flow to increase crossingAdvertising trades to attract liquidity (Indications of Interest, IOIs)Post-trade:Missed trade opportunity costs unfilled orders eg limit order th

30、at was not sufficiently aggressiveUndisclosed revenueDelay costs (slippage)Inability to complete the trade in desired time (VWAP)Waiting for crossingMarket exposure (?)Fund Management FeesTaken from Myners Report (2001) “Institutional Investment in the UK: A Review”, (HM Treasury)MN20026 Investment

31、ManagementFund Management FeesSource: Mercer (2007)Fund Management Fees % AUM Across Mandate Type by Size of Mandate (Median fees across managers for Segregated Portfolios) UK Investments (Pounds sterling) 25M 50M 100M 250M UK - Multi-Asset (ie Balanced) 0.49 0.43 0.35 0.29 UK - Equity All Cap 0.60

32、0.48 0.42 0.35 UK - Equity Small Cap 0.75 0.70 0.56 0.49 US Investments (US dollars)US Equity Large Cap/All Cap - Growth 0.65 0.58 0.53 0.46 US Equity Large Cap/All Cap - Value 0.64 0.56 0.51 0.45 US Equity Small Cap Growth 1.00 0.90 0.84 0.78 US Equity Small Cap Value 0.90 0.87 0.79 0.75 Canadian I

33、nvestments (Canadian dollars)Canada - Balanced 0.35 0.30 0.26 0.22 Canada - Small Cap Equity 0.63 0.60 0.55 0.51 International Investments (US dollars)International Global Equity - Growth 0.75 0.70 0.65 0.54 International Global Equity - Value 0.80 0.76 0.65 0.57 Emerging Markets Equity 1.00 0.95 0.

34、88 0.83 MN20026 Investment ManagementCosts for unit trustsMN20026 Investment ManagementCosts for International Unit TrustsMN20026 Investment ManagementComparison of Retail & Institutional Fund Management Fees % AUM by Size of Mandate UK Investments (Pounds sterling)Segregated mandates: 5M 10M 25M 50M 100M 250M UK - Balanced 0.49 0.43 0.35 0.29 UK - Equity All Cap 0.60 0.48 0.42 0.35 UK - Equity Small Cap 0.75 0.70 0.56 0.49 UK- Corporate Bonds0.400.290.240.21UK - Government bonds0