![3
1.Introduction
Since the firstpubliclylistedfundof ForeignandCapitalin1868,Investmenttrustshave played
a fundamental roleinfinancial intermediationacrossthe world.A type of mutualfund,anInvestment
trust raised funds by issuing shares on the stock market and then investing the proceeds into a
portfolio of assets. Significantlygrowth in the operating number of funds has meant fundmanagers
have hadtostrive todelivercompetitive returns.Ithasbeenwidelydocumentedthatinvestorsbenefit
from managerial skills as well as diversification, cost advantages and liquidity intermediationwhen
investinginInvestmenttrusts.These skillsare relatedtoselectivityabilityinpickingsuccessful stocks
and timing ability in accurately forecasting future market movements. Variability in such skills are
evidenced in fund manager’s strategy and asset allocation.
Most studies on US mutual funds suggest little or no superior performance, but stronger
evidence of underperformance (Lakonishok et al 1992, Grinblatt et al 1995, Cahart 1997). Similar
resultswere achievedonUK funds(Blake and Timmermann,1999; Blak et al 1999) Although,the UK
fundmanagementindustryisresponsible foranexcessof $5.5tr1
, to the bestof our knowledge most
research on selectivity and market timing ability has used out-of-date dataset. With little research
been done that surpasses 2010. On this account, the papers look to provide some fresh conclusions
on whethermanagersfromUKInvestmenttrustsgenerate positive alphasandthusshow evidence of
selectivityability.Whilstalsoaccountingformarket timingabilitythroughthe gamma termsupplied
by Treynor and Mazuy (1966). With these underlying objectives, this paper aims to discuss the
following research questions:
Do UK Investment Trusts generate positive abnormal performance relative to the market?
Do fund managers in UK Investment trust possess superior selectivity and market timing skills?
Is Investment ability more evident for fund managers who hold portfolios concentrated in a few
industries?
The structure of this paper is as follows. Firstly, an analysis of relevant literature on asset
pricingmodelsandmutual fundperformance.Thisisfollowedbya descriptionof the data and chart
analysis. The next sectionprovides an explanationof the empirical models and methodology. This is
followedbyempirical resultsonallestimatesmeasures.Section6presentsdiscussiononthe empirical
findingsinthe contextof other literature.Section7 discussesthe limitationsof the study and future
1 Annual Reports and Accounts year end 2015,The Investment Association (2016),
www.theinvestmentassociation.org .[accessed: 2/09/16]](https://image.slidesharecdn.com/1fa24288-281e-476d-a8c5-ee7aca926ec6-161129150440/85/MASTER-DISSERTATION-3-320.jpg)
![6
the CAPMmodel,and10 satisfythe three-factormodel.These empirical findingsare encouragingto
thispapers chosen methodology to use the single-index CAPM to model Investment trust returns.
Following the arrival of the three-factor model,Jegadeesh and Titman (1993) explained the
case of investorsutilisingastrategybasedonrecentmomentum5
togenerate abnormalreturns.Soon
after Carhart (1997) examinedthe persistence instockreturnsof mutual funds inUS equitymarkets
firstly using an augmented multifactor model to account for momentum. The limited literature
surrounding the validity of this factor mainlysupport its inclusion. L’Her et al (2004) concluded that
the four-factor model was valid in the Canadian Market. Likewise, Lam and So (2009) found the
momentumfactortobe significantforthe HongKongMarket.Lai andLau(2010) highlightthe relative
strength of the model in explaining mutual fund returns in Malaysia. Unlu (2012) found consistent
results for the Irish Stock Exchange.
2.2 Fund Performance measures
The evolution of fund performance measures has been parallel with the growth of asset
pricingmodels.The introductionof the CAPMbySharpe(1964) andLintner(1965) ledtothe arrival of
the “three indicies”fromSharpe (1966),Treynor(1965), and Jenson (1968). All three of these models
were derivedsimplyfromthe CAPMmodel;The Sharperatioisbasedonthe rewardtovolatilitytrade-
off andformulatesthe ratiobetweenaverage returnsearnedinexcessof the risk-freerate perunitof
volatility. The Treynor Ratio from Treynor (1965) is of close format to the Sharpe ratio, however
definesthereward-to-volatilityratioinrelationtoeachunitof the CAPMbetarisk.Incomparison,the
Jensen(1968) alphameasuresreferstothe interceptdeterminedfrom the CAPMregressionof excess
portfolio returns on the excess market returns.
Jensen’s alpha has been the predominant measures used in fund performance valuation.
Essentially,itholdsastable positionbecauseitrepresentsthe interceptwhenexcessfundreturnsare
estimated against either the market index, book-to-market ratio, size or momentum factors. In
relationtothe efficientmarkethypothesis,the alphatermindicatesmarketefficiencyonthe basisof
either out or under-performance in fund returns. According to the efficient market hypothesis this
alpha term should not be significantly different from zero.
2.3 Mutual Fund Performance
The UK and US have two of the most developed and largest fund management industries in
the world.Total U.S mutual fundassetsbeinginexcessof $15.7 trillion6
whilstUKfundstotallingover
5 This strategy involves buyingstocks which haveperformed well in the pastyear, whilstsellingrecentpoor
performing stocks.
6 2016 Investment Company FactBook, A review of Trends and Activities in the U.S Investment Company
Industry,56th edition. Investment Company Institute, [accessed 2/09/16]](https://image.slidesharecdn.com/1fa24288-281e-476d-a8c5-ee7aca926ec6-161129150440/85/MASTER-DISSERTATION-6-320.jpg)
![7
$7.3 trillion7
inmanagedfunds.Giventhe largeamountofdataavailableformeaningful analysis,much
of the academic literature hasfocusedon fundsin these market.Althoughmore recentlythere have
been new focus on European and Australian industries.
U.S Market
Studies on US mutual funds suggest evidence of little or no superior performance. Earlier
research in the seminal paper of Jensen (1968) tests the abnormal performance on 115 funds over
1945 to 1964, and found no significant abnormal performance.Malkiel (1995) analysed US equity
fundsusingthe singleindex model butoveralongerperiodthanJensen(1968).Resultsoverthe 1971
to 1991 period here show the average alpha is statistically insignificant from zero. Other literature
from the US has indicated minimal superior performance, but more evidence of underperformance
(Daniel et al, 1997; Chevailler and Elliso, 1999; Wermer, 2000, Baks et al, 2001).
Most of the abovementionedstudieshadusedstandardconventional statistical techniques,
however there has been a recent influx of new methods to measure fund performance more
accurately.Namely,Kosowskietal (2006) and FamaandFrench(2010) adoptthe bootstraptechnique
to calculate alpha and its corresponding test statistic. Thismethod aimsto separate managerial skill
from luck since the standard statistical technique does not account for presence of luck or the non
normalitypropertiesinalpha.Applyingthismeasure,Kosowski etal (2006) demonstrated thatonlya
minority from the analysed 2118 US mutual funds posses stock picking skills. Moreover, using FDR
(False Discoveryrate)8
approach,Scailletetal (2010) found75% of US fundsexhibitazeroalphabased
on returns, with a few showing evidence of genuine skill.
U.K Market
Research in the UK has been more limited than the US. This is because dataset providers in
the UK are more commercially motivated in that they only offer information on active funds. In
comparison, academics in the US have benefited from access to the CRSP9
database which holds
informationonbothdeadand live funds.Asa result,a large portionof UK studieshave beensubject
to survivor-bias10
samples.
7 Annual Reports and Accounts year end 2015,The Investment Association (2016),
www.theinvestmentassociation.org .[accessed: 2/09/16]
8 The FalseDiscovery Rate (FDR) is a measure to providea simpleway to calculatethe number and the
proportion of funds with truly positiveand negative performance in any portion of the tails of the cross -
sectional alphadistribution.
9 Center for Research in Security Prices – Provider of historical stock marketdata. Maintains someof the
largestand most comprehensive proprietary historical databases in stock marketresearch.Researchers rely on
the CRSP for accurate,survivor bias-freeinformation.
10 Survivorship bias refers to the tendency for failed companies to be excluded from performance studies
based on the fact that they no longer exist. This can causeskewness in results becauseonly successful
companies areincluded.](https://image.slidesharecdn.com/1fa24288-281e-476d-a8c5-ee7aca926ec6-161129150440/85/MASTER-DISSERTATION-7-320.jpg)
![13
4. Methodology
4.1 Overall Performance
This studyuses the unconditional single-factorperformance measure:Jensen’s(1968) alpha.
This measure is the intercept term estimated through the the regression model:
𝑅 𝑖𝑡 − 𝑅 𝑓𝑡 = 𝛼𝑖 + 𝛽𝑖( 𝑅 𝑚𝑡 − 𝑅 𝑓𝑡) + 𝜖𝑖 (1)
where 𝛼𝑖 isthe risk adjustedabnormal returnfromthe single index model, 𝑅 𝑖𝑡 isthe returnonfund i
overperiod t, 𝑅 𝑓𝑡 is the returnon the 5 year US Treasuryyieldadjustedtoconstant maturity, 𝑅 𝑚𝑡 is
the returnonthe FTSEALL-SHAREIndex, 𝛽𝑖 the factorsensitivityof differencesinfundreturnsandthe
risk free rate, and 𝜖𝑖 is the error term with the following properties: E(𝜖𝑖,𝑡) = 0 , Var(𝜖𝑖,𝑡) = 𝜎𝜖 𝑖,𝑡
2 ,
Cov( 𝜖𝑖,𝑡,𝑟 𝑚,𝑡) = Cov( 𝜖𝑖,𝑡,𝜖𝑗,𝑡) = 0.
The following hypothesis are tested with this model:
𝐻0: 𝛼𝑖 = 0
𝐻 𝑎: 𝛼𝑖 ≠ 0
where the null hypothesissuggests fund i does not out or under-performthe relative market proxy,
the alternate hypothesis indicates fundi either out or under-preformsto the relative market proxy.
Statistical significanceismeasuredusingcritical valuesusingatwo-tailedtestat5% significancelevel.
Althoughfordiscussionpurposeswe alsoconsidercoefficientsthatgenerate a hightest-statisticbut
are marginally insignificant.
The intercept term identifies whether fund managers have superior stock selection abilities.
Alphagenerationisachievedthroughselectingsecuritiesthatresultin 𝜖𝑖 > 0.A statisticallysignificant
positive alpha indicates the fund manager has the abilityto forecast future securityprices. An alpha
term that does not statistically differ from zero implies the manager mimics the composition of a
reference marketbenchmark.A statisticallysignificantnegative alphatermssuggestthe fundmanager
performs poorer than a naïve strategy of random selection. In the context of compensation, the
annualizedJensen’salphaisthe maximummountof moneyaninvestorshouldbe willingtopayafund
per year. Subsequently, this allows us to contribute to the literature surround pay performance in
mutual funds.
In addition,we alsoassessandcompare the riskof eachfund.UnderCAPMthe riskisdivided
intotwo components.These are formallyexpressedasproductsof the standard deviationinreturns:
𝜎𝑖
2
= [𝛽]2 𝜎 𝑚
2 + 𝜎 𝑢,𝑖
2](https://image.slidesharecdn.com/1fa24288-281e-476d-a8c5-ee7aca926ec6-161129150440/85/MASTER-DISSERTATION-13-320.jpg)
![14
The firstpart [𝛽]2 𝜎 𝑚
2 indicatesthe market (or systematic) risk.Thistype of riskarisesdue to
fluctuations in the market as a whole and cannot be eliminated through diversification. This is also
interpretedintermsof ahigh 𝑅2 of the regressionmodel(1).We payspecial attentiontothisvaluein
each funds model in our results analysis. The second part 𝜎 𝑢,𝑖
2
is referred to as specific (or non-
systematic) risk. This is often attributable to managerial competence with the fund. In contrast to
systematicriskthistypeof riskisidiosyncraticandthuscanbe eliminatedthroughsuccessfulportfolio
diversification. A well diversified portfolio containing a great number of assets with different
characteristics can cancel out such risk. We take the difference between 1 and 𝑅2, to estimate each
funds idiosyncratic risk.
Anothermeasure forrisk-adjustedreturnsthatwe use isthe Sharpe ratio. This performance
method was developed by Sharpe (1966) under its birth name, the reward-to-variability ratio. Over
the yearsithas gainedsignificantpopularityinthe financeandnow operatesan industrystandardfor
measuring risk-adjusted performance. Incorporating standard deviation of portfolio returns against
excess returns, the ratio has benefits in its comparability across fund categories as well its
independence of any choice of benchmark (Jagric et al 2007). The measure is given by:
𝑆𝑖 =
𝑅̅ 𝑖 − 𝑅̅ 𝑓
𝜎𝑖
where 𝑅̅ 𝑝 is the meanreturn to portfolio i duringthe evaluationperiod, 𝑅̅ 𝑓isthe meanrisk free rate
of return and 𝜎𝑖 is the standard deviation of the portfolio return. The higher the ratio, the better its
returns have been relative to the amount of investment risk. We replicate (Bal and Leger, 1996) in
estimatingthismeasureoverdifferenttime horizonstogauge itsvariabilityunderdifferenteconomic
conditions and to provide further analysis on fund manager selectivity.
4.2 Market Timing
Overall performance can be decomposed into fund managers stock selection and market
timing abilities. To measure whether fund managers have the ability to predict market movements,
we adopt an extended version of (1) from Treynor and Mazuy (1966) (TM hereafter). The model
includes the additionof a quadratic term to equation (1) in order to account for non-linearityin the
function of market return, and is as follows:
𝑅 𝑖𝑡 − 𝑅 𝑓𝑡 = 𝛼𝑖 + 𝛽𝑖( 𝑅 𝑚𝑡 − 𝑅 𝑓𝑡) + 𝛾𝑖( 𝑅 𝑚𝑡 − 𝑅 𝑓𝑡)
2
+ 𝜖𝑖 (2)
where 𝛾𝑖 is the risk adjusted measure of market timing ability of fund i, and 𝜖𝑖 is an error term with
the following properties: E(𝜖𝑖,𝑡) = Cov( 𝜖𝑖,𝑡,𝑟 𝑚,𝑡) = Cov( 𝜖𝑖,𝑡,𝑅 𝑚,𝑡
2 ) = 0
The followinghypothesisistestedusingthismodel:](https://image.slidesharecdn.com/1fa24288-281e-476d-a8c5-ee7aca926ec6-161129150440/85/MASTER-DISSERTATION-14-320.jpg)
![28
Treynor, J, and K Mazuy. "Can mutual funds outguess the market." Harvard business review 44.4
(1966): 131-136.
Treynor, J.L., 1965. How to rate mutual fund performance. Harvard Business Review, 43, pp.63-75.
Unlu, U., 2013. Evidence to support multifactor asset pricing models: The case of the Istanbul stock
exchange. Asian Journal of Finance & Accounting,5(1), p.197.
VanNieuwerburgh,SandVeldkamp,L,2005. InformationAcquisitionandPortfoliounder
Diversification. NYUWorkingPaperNo.FIN-04-025
Verheyden, T., De Moor, L. and Van den Bossche, F., 2015. Towards a new framework on efficient
markets. Research in International Business and Finance, 34, pp.294-308.
Wermers,R., 2000. Mutual fundperformance:An empirical decompositionintostock‐pickingtalent,
style, transactions costs, and expenses. The Journal of Finance, 55(4), pp.1655-1703.
Data Sources
Datastream.(2012) Thomson ReutersDatastream.[Online]. Available at:SubscriptionService
(Accessed:17/07/2016)
Lazard AssetManagement(2016). The Lazard World Trust FundAnnual Report(2016). Retrieved
fromwww.theworldtrustfund.com
Morningstar.(2016, August17) AtlantisJapanGrowthFund. RetrievedfromMorningstar
InvestmentResearchdatabase.
NewCityInvestmentManagers(2016). CQSNew CityHighYieldFundLtd InterimReport(2015).
Retrievedfromwww.ncim.co.uk
ShrodersUK. (2016). ShroderAsiaPacificFundplc.2015 Annual reportandAccounts.Retrievedfrom
www.shroders.co.uk
ShrodersUK. (2016). ShroderIncome GrowthFundplc. 2015 Annual reportandAccounts.Retrieved
fromwww.shroders.co.uk
ShrodersUK. (2016). ShroderJapanGrowth Fundplc.2015 Annual reportandAccounts.Retrieved
fromwww.shroders.co.uk
ShrodersUK. (2016). ShroderMid Cap fundplc.2015 Annual reportandAccounts.Retrievedfrom
www.shroders.co.uk
ShrodersUK. (2016). ShroderUK Gowth Fundplc.2015 Annual reportand Accounts.Retrievedfrom
www.shroders.co.uk](https://image.slidesharecdn.com/1fa24288-281e-476d-a8c5-ee7aca926ec6-161129150440/85/MASTER-DISSERTATION-28-320.jpg)
![38
APPENDIX FIGURES
Figure 1.
Plotsthe monthlyshare price of the 10 analysedInvestmentTrustsoverperiodJanuary1995 to June 2016.
(Source: Thomson Reuters DataStream [Accessed: 17/07/17])
0
100
200
300
400
500
600
700
800
900
1000
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
ATLANTIC JAP.GW.FD (A) SHRODER ASIA PACIFIC (B) ORYX INTERNATIONAL (C) SHRODER INCOME (D)
SHRODER JAPAN (E) SHRODER UK GROWTH (F) CQS NEW CITY HIGH YIELD (G) LAZARD WORLD TRUST (H)
LAZARD WORLD TRUST (I) SHRODER UK MID CAP (J)](https://image.slidesharecdn.com/1fa24288-281e-476d-a8c5-ee7aca926ec6-161129150440/85/MASTER-DISSERTATION-38-320.jpg)
MASTER DISSERTATION
- 1. 1 MSc Banking and Finance Research Project: 926N1 Candidate Number: 143347 Do Mutual Funds outperform the market? An analysis of the Stock-Picking and Market Timing Ability in Fund Managers from leading UK Investment Trusts Abstract This study examines the selectivity and timing performance of 10 UK Investment trusts over the period January 1995 to June 2016 using a combination of Jensen, Sharpe, and Treynor measure. Results show little evidence of outperformance against the FTSE All Share index. Only 1 fund showed evidence of superior stock selectivity, whilst no funds showed evidence of superior market timing. Consistent with other studies, this paper also highlights the positive association between portfolio concentration and performance in mutual funds. Acknowledgement I would like thank my dissertation advisor Dr Bruce Hearn of the School of Business and Management at the University of Sussex .
- 2. 2 Table of Contents 1.Introduction............................................................................................................................3 2.Literature Review...................................................................................................................4 2.1 Evolution of Asset Pricing Models .......................................................................................4 2.2 Fund Performance measures...............................................................................................6 2.3 Mutual Fund Performance ..................................................................................................6 U.S Market............................................................................................................................ 7 U.K Market ........................................................................................................................... 7 2.4 Selectivity and market timing inUK Investment Trusts ........................................................8 2.5 Fund concentration and performance .................................................................................9 3. Data description....................................................................................................................9 3.1 Share price analysis over time........................................................................................... 11 4. Methodology.......................................................................................................................13 4.1 Overall Performance......................................................................................................... 13 4.2 Market Timing.................................................................................................................. 14 5.Empirical Results..................................................................................................................15 5.1 Selectivity Performance ()............................................................................................... 15 5.2 Market Timing Performance ()......................................................................................... 16 5.3 Sensitivity to the market ().............................................................................................. 17 5.4 Market and Idiosyncratic risk............................................................................................ 17 5.5 Sharpe Ratio’s .................................................................................................................. 18 6. Discussion ............................................................................................................................18 6.1 Selectivity Performance.................................................................................................... 18 6.2 Market timing................................................................................................................... 20 7. Limitations & Future Research ...........................................................................................21 8. Investors Implications.........................................................................................................22 9. Conclusion............................................................................................................................23 Bibliography.............................................................................................................................24 Data Sources............................................................................................................................28 APPENDIX TABLES ...................................................................................................................29 APPENDIX FIGURES .................................................................................................................38
- 3. 3 1.Introduction Since the firstpubliclylistedfundof ForeignandCapitalin1868,Investmenttrustshave played a fundamental roleinfinancial intermediationacrossthe world.A type of mutualfund,anInvestment trust raised funds by issuing shares on the stock market and then investing the proceeds into a portfolio of assets. Significantlygrowth in the operating number of funds has meant fundmanagers have hadtostrive todelivercompetitive returns.Ithasbeenwidelydocumentedthatinvestorsbenefit from managerial skills as well as diversification, cost advantages and liquidity intermediationwhen investinginInvestmenttrusts.These skillsare relatedtoselectivityabilityinpickingsuccessful stocks and timing ability in accurately forecasting future market movements. Variability in such skills are evidenced in fund manager’s strategy and asset allocation. Most studies on US mutual funds suggest little or no superior performance, but stronger evidence of underperformance (Lakonishok et al 1992, Grinblatt et al 1995, Cahart 1997). Similar resultswere achievedonUK funds(Blake and Timmermann,1999; Blak et al 1999) Although,the UK fundmanagementindustryisresponsible foranexcessof $5.5tr1 , to the bestof our knowledge most research on selectivity and market timing ability has used out-of-date dataset. With little research been done that surpasses 2010. On this account, the papers look to provide some fresh conclusions on whethermanagersfromUKInvestmenttrustsgenerate positive alphasandthusshow evidence of selectivityability.Whilstalsoaccountingformarket timingabilitythroughthe gamma termsupplied by Treynor and Mazuy (1966). With these underlying objectives, this paper aims to discuss the following research questions: Do UK Investment Trusts generate positive abnormal performance relative to the market? Do fund managers in UK Investment trust possess superior selectivity and market timing skills? Is Investment ability more evident for fund managers who hold portfolios concentrated in a few industries? The structure of this paper is as follows. Firstly, an analysis of relevant literature on asset pricingmodelsandmutual fundperformance.Thisisfollowedbya descriptionof the data and chart analysis. The next sectionprovides an explanationof the empirical models and methodology. This is followedbyempirical resultsonallestimatesmeasures.Section6presentsdiscussiononthe empirical findingsinthe contextof other literature.Section7 discussesthe limitationsof the study and future 1 Annual Reports and Accounts year end 2015,The Investment Association (2016), www.theinvestmentassociation.org .[accessed: 2/09/16]
- 4. 4 research. Section 8 present a brief summary of the investor implications to corporate policy of our findings. Finally, this if followed by concluding remarks. 2.Literature Review 2.1 Evolution of Asset Pricing Models The origins of the CAPM stem from the work of Sharpe (1964) and Lintner (1965), whilst Markowitz (1952) and Tobin(1958) laid the model’s foundations through the mean-variance algorithm.Markowitz’smodelexplainshow aninvestorselectsaportfolioattime t-1thatproducesa stochastic return at t. While assuming investors are risk averse and only consider the mean and variance of their investment return. Thus, investors choose “mean-variance-efficient” portfolios, as giventhe expectedreturnandvariance,portfolio’sbothminimize the varianceof the portfolioreturn and maximize expected return. Sharpe (1964) and Lintner(1965) developedtwokeyassumptionstothe Marrkowitz’smean- variance framework. The first is complete agreement: Given market clearing prices at t-1, investors agree on the jointdistributionof assetreturnsfromt-1to t. The secondassumptionstatesthatthere is borrowingandlendingat a risk-free rate,whichisthe same for all investorsandis independentof amount.Combinedwiththe workof Black(1972) whoformedthe CAPM,explainedhow theexpected return on a stock is determinedbythe risk-free interestrate and a risk premium, whichisa function of the stock’s responsiveness to movements in the market. The latter is classified as the beta coefficient,arguablythe maincomponentwhichisheavilyusedamongfundmanagersinthe financial markets. A majorityof the earlierempirical testsof CAPMgive supportto its specificationthatbetais the only explanatory factor in explaining cross sectional portfolio returns (Lintner, 1965; Douglas, 1968). However, in later research support for the model has weakened. Fama and MacBeth (1973) showthe betacoefficientwasstatisticallyinsignificant.Blacketal (1972) usedtime seriesregression analysistoshowhowthe interceptissignificantlydifferentfromzeroanditstime varyingproperties, whichviolate marketefficiencyandthe original model.LaterRoll (1977) suppliedfurthercriticism,in claiming the proxies used to compose the market portfolio are not reflective of the portfolio of invested wealth.Thus using any other portfolio as opposed to the true market portfolio tests the efficiencyof the selectedproxyportfolio.More recently,BartholdyandPeare (2003) concludedthat any correctly used proxy will always generate biased estimates for expected returns. As empirical research documenting the flaws of CAPM grew, a wave of alternative asset pricing models arrived. The Arbitrage Pricing Theory (APT) by Ross (1976) presented a multi-factor model which allows an asset returns to have many systematic risk measures. These refer to
- 5. 5 macroeconomic risk factors which cannot be diversified against. Although the APT benefited the CAPMinbeinglessrestrictiveandexplaininggreaterproportionof securityreturns,several drawbacks exist.Unlike the CAPM,the APT doesnot reveal the identityof pricedfactors.Therefore,demanding users to reasonably estimate the factor sensitivities.Further studies on asset pricing have identified numerous variables beyond the market beta that explain stock returns, termed ‘anomalies’. These include market capitalization (Banz, 1981), earnings to price ratio (Basu, 1983) and book-to-market ratio (Rosenbeg et al, 1985). Fama and French (1992) confirm these anomalies explain returns, claimingmiss-specificationinthe CAPMbetween1963 and 1990. Similarly,there hasbeenevidence of thisforEuropeanandJapanese markets(Capaul etal,1993).Withinthe mutual fundliterature,the APT framework has been applied in studies including Connor et al (1991) and Fletcher (1997), who both conclude that on average trusts in the UK and US do not outperform the market benchmark. In response to the poor performance and anomalies of the CAPM, Fama and French (1993) developed the three-factor asset pricing model. In this model excess portfolio returns are explained bythree riskfactors.These factorsincludethe CAPM’sexcessmarketreturn,sizefactor2 andbook-to- market factor3 . Early tests of the three-factor model by Fama and French (1995) show that only market and size factors help explain returns, though B/Mrevealed no relation. Comparably, Porras (1998) found B/Minsignificant,althoughfoundsizetobe insignificantusing cross-sectionalregression analysis. Nonetheless,studies post the millennium have found the two additional factors significant in explainingreturnsacrossAustralia,Canada,Germany,Japan, the UK and US (Maroney et al, 2002; Drew et al, 2003). Given the popularityof the single index by financial professionals today,there is an ongoing controversial debate between the CAPM and the three-factor model. Blanco (2012) favoured the three-factor model with respect to explaining expected returns in the American Stock Market, providing support for the size and B/M factor inclusions. Similarly, Simpson et al (2008)4 found the relative meritof the three-factormodelisitsabilitytocapture informationrelatingtoawide range of economic indicators. On the other hand, numerous studies have yielded evenly sided results. Bartholdy and Peare (2005) show that CAPM explained on average 3% of stock returns, whilst the three-factorexplained5%.Likewise,Sourmere etal (2013) finds11 out of 28 companystocks satisfy 2 The sizefactor (SMB) is a zero-investment portfolio that is longon small capitalization stocks and shorton bi capitalization stocks. 3 The book-to-market factor (HML) is a zero-investment portfolio that is longon high book - to-market (B/M) stocks and shorton lowB/M stocks.
- 6. 6 the CAPMmodel,and10 satisfythe three-factormodel.These empirical findingsare encouragingto thispapers chosen methodology to use the single-index CAPM to model Investment trust returns. Following the arrival of the three-factor model,Jegadeesh and Titman (1993) explained the case of investorsutilisingastrategybasedonrecentmomentum5 togenerate abnormalreturns.Soon after Carhart (1997) examinedthe persistence instockreturnsof mutual funds inUS equitymarkets firstly using an augmented multifactor model to account for momentum. The limited literature surrounding the validity of this factor mainlysupport its inclusion. L’Her et al (2004) concluded that the four-factor model was valid in the Canadian Market. Likewise, Lam and So (2009) found the momentumfactortobe significantforthe HongKongMarket.Lai andLau(2010) highlightthe relative strength of the model in explaining mutual fund returns in Malaysia. Unlu (2012) found consistent results for the Irish Stock Exchange. 2.2 Fund Performance measures The evolution of fund performance measures has been parallel with the growth of asset pricingmodels.The introductionof the CAPMbySharpe(1964) andLintner(1965) ledtothe arrival of the “three indicies”fromSharpe (1966),Treynor(1965), and Jenson (1968). All three of these models were derivedsimplyfromthe CAPMmodel;The Sharperatioisbasedonthe rewardtovolatilitytrade- off andformulatesthe ratiobetweenaverage returnsearnedinexcessof the risk-freerate perunitof volatility. The Treynor Ratio from Treynor (1965) is of close format to the Sharpe ratio, however definesthereward-to-volatilityratioinrelationtoeachunitof the CAPMbetarisk.Incomparison,the Jensen(1968) alphameasuresreferstothe interceptdeterminedfrom the CAPMregressionof excess portfolio returns on the excess market returns. Jensen’s alpha has been the predominant measures used in fund performance valuation. Essentially,itholdsastable positionbecauseitrepresentsthe interceptwhenexcessfundreturnsare estimated against either the market index, book-to-market ratio, size or momentum factors. In relationtothe efficientmarkethypothesis,the alphatermindicatesmarketefficiencyonthe basisof either out or under-performance in fund returns. According to the efficient market hypothesis this alpha term should not be significantly different from zero. 2.3 Mutual Fund Performance The UK and US have two of the most developed and largest fund management industries in the world.Total U.S mutual fundassetsbeinginexcessof $15.7 trillion6 whilstUKfundstotallingover 5 This strategy involves buyingstocks which haveperformed well in the pastyear, whilstsellingrecentpoor performing stocks. 6 2016 Investment Company FactBook, A review of Trends and Activities in the U.S Investment Company Industry,56th edition. Investment Company Institute, [accessed 2/09/16]
- 7. 7 $7.3 trillion7 inmanagedfunds.Giventhe largeamountofdataavailableformeaningful analysis,much of the academic literature hasfocusedon fundsin these market.Althoughmore recentlythere have been new focus on European and Australian industries. U.S Market Studies on US mutual funds suggest evidence of little or no superior performance. Earlier research in the seminal paper of Jensen (1968) tests the abnormal performance on 115 funds over 1945 to 1964, and found no significant abnormal performance.Malkiel (1995) analysed US equity fundsusingthe singleindex model butoveralongerperiodthanJensen(1968).Resultsoverthe 1971 to 1991 period here show the average alpha is statistically insignificant from zero. Other literature from the US has indicated minimal superior performance, but more evidence of underperformance (Daniel et al, 1997; Chevailler and Elliso, 1999; Wermer, 2000, Baks et al, 2001). Most of the abovementionedstudieshadusedstandardconventional statistical techniques, however there has been a recent influx of new methods to measure fund performance more accurately.Namely,Kosowskietal (2006) and FamaandFrench(2010) adoptthe bootstraptechnique to calculate alpha and its corresponding test statistic. Thismethod aimsto separate managerial skill from luck since the standard statistical technique does not account for presence of luck or the non normalitypropertiesinalpha.Applyingthismeasure,Kosowski etal (2006) demonstrated thatonlya minority from the analysed 2118 US mutual funds posses stock picking skills. Moreover, using FDR (False Discoveryrate)8 approach,Scailletetal (2010) found75% of US fundsexhibitazeroalphabased on returns, with a few showing evidence of genuine skill. U.K Market Research in the UK has been more limited than the US. This is because dataset providers in the UK are more commercially motivated in that they only offer information on active funds. In comparison, academics in the US have benefited from access to the CRSP9 database which holds informationonbothdeadand live funds.Asa result,a large portionof UK studieshave beensubject to survivor-bias10 samples. 7 Annual Reports and Accounts year end 2015,The Investment Association (2016), www.theinvestmentassociation.org .[accessed: 2/09/16] 8 The FalseDiscovery Rate (FDR) is a measure to providea simpleway to calculatethe number and the proportion of funds with truly positiveand negative performance in any portion of the tails of the cross - sectional alphadistribution. 9 Center for Research in Security Prices – Provider of historical stock marketdata. Maintains someof the largestand most comprehensive proprietary historical databases in stock marketresearch.Researchers rely on the CRSP for accurate,survivor bias-freeinformation. 10 Survivorship bias refers to the tendency for failed companies to be excluded from performance studies based on the fact that they no longer exist. This can causeskewness in results becauseonly successful companies areincluded.
- 8. 8 ResultsonUK mutual fundshas tendedto yieldsimilarresultstothose discoveredin the US. Fletcher (1997) examines fund performance using Henriksson and Merton (1981) to decompose performance intostockpickingandmarkettiming.The resultssuggestthatonaverage managersfrom UK unittrust exhibitpositive stockselectivityandnegative markettiming.Usingthe same datasetin a laterstudy,Fletcher(1997) findsno significantevidence thatUK unit trust outperformthe market. Constant with this finding, Blake and Timmermann (1998) find evidence of under performance by equity and balanced managed UK funds. Moreover, Quigley and Sinquefield (2000) use both CAPM and three –factor model to analyse monthly returns on 752 UK equity based funds over a 20 year periodof 1978 to 1998. They show that UK managers netof expensesare unable to outperformthe market, thus coinciding with the US findings. Studies using measures beyond the conventional standard statistical techniques have also developedontoUK data.Cuthbertsonetal (2008) appliesthe approachof Kosowski etal (2006) to a survivor-biasfreesampleof 842UKequityunittrusts.Resultsfromthisbootstrappingtechnique show the average alpha of funds is negative but statisticallyinsignificant. These results lie consistentwith Blake and Timmermann (1998). In a subsequent study using the same survivor-bias free dataset, Cuthbertsonetal (2010) replicatesasimilarFDRmethodologyasScailletetal (2010).Resultssuggests that the number of UK funds with truly negative abnormal performance significantly exceeds the number of funds with truly positive abnormal performance. 2.4 Selectivity and market timing in UK Investment Trusts The literature closely linked with this paper have concludedthat UK Investment trusts have not on average beenable toout-performthe market.Bal andLeger(1997) analyse 92 fundsoverthe period 1975 to 1993 using Jensen’s alpha and the Sharpe Ratio. Even without correction for transactioncosts,fundsonaverage didnotgenerate significantalpha’s.Theyalsoshowthatthe choice of variance or covariance risk(Sharpe andTreynormeasures) mattersverylittle.Inaddition,theyfind evidence of perverse markettimingfromthe 90’s onwards.Thistrendintersectswiththe startof the dataset used in this paper, and thus will be compared against in later discussion. Correspondingly, Leger(1996) observedinsignificantalphasandnegative timingperformance forone inthree trustsof a sample of 72funds.Whilsthighlightingthe strongnegativecorrelationbetweenmanagertimingand selectivity.Bangassa(1999) addsfurther supportto these results,but identifiesfundstylesinJapan, North America and Europe generates significant perverse timing practices. Moreover, Cuthbertson (2009) showedthat only1% of fundsdemonstratedpositive markettimingat5% level,while 19%of fundsexhibitednegative timings.More recently,Bangassaetal (2012) examinedselectivityandtiming performance of 218 UK Investment Trusts. They conclude that international funds show some favourable selectivity ability, while domestic funds show better timing ability.
- 9. 9 2.5 Fund concentration and performance Positive fund performance can be attributed to a wide selection of factors related to managerial characteristics.Thoughtherehasbeenanavenueof researchonthe relationshipbetween portfolio composition and performance. Kacperczyk et al (2005) claimed investment ability is more evidentamongmanagerswhoholdportfoliosconcentratedinafewindustries.Likewise,otherrelated studieshave showedthatfocusedmanagersoutperformtheirmore broadlydiversifiedcounterparts (Baks etal, 2006; Hujj and Derwall,2011). Othershave attributedfocusedinvestmentstrategiesand outperformance to the case where managers exercise their informational advantages (Coval and Moskowitz,1999,2000; Nanda et al, 2004). Whereas, Sapp and Yan (2008) find no evidence that focused funds outperform diversified funds.These empirical findingssuggest that should there be presence of abnormal performanceinthispaper,one wouldexpectthe fundtobe concentratedona few industries. 3. Data description The data used in this study consistsof monthlyreturns calculatedas the percentage change inshare price of 10UK InvestmentTrusts.DatawasextractedfromThomsonReutersDataStreamover the examined period of 1 January 1995 to 1 June 2016. OftentermedasClosed-Endfunds,Investmenttrustssell afixedamountof unitstoinvestors at the time of offer.Theydonot issue additionalunitsinresponsetodemand,insteadtheyare listed on the stockmarketfor investorstopurchase.These listedunitsactessentiallyascompanystockand their prices are determined by demand and supply forces. Unlike other investment funds, shares in Investmenttrustscanbe purchasedandsoldatpricesabove orbelow the NetAssetValue(combined value of all assetsthe trustshold).A share price lowerthanthe NAV issaidtobe tradingatadiscount, in comparison to a price that is above the NAV, the shares trading at a premium. In selectingdata,we incorporate anapproachthatonlyaccountsforsurvivingfunds.Firstly,a search of UK investment trusts on the DataStream database was taken over the specifiedperiod. Secondly,all resultingfundswereimportedtoanexcel spreadsheet.Next,allnon-survivingfundsthat have ceased operations/terminated over the observationperiod are droppedand this was evidence by a constant share price. Fund attrition has been the consequence of poor fund performance over time or a judgementfrommanagementthatmarketvalue of funis sufficientlysmall andthusitis no longer rational to maintain the fund (Elton et al,1996). This process of fund selection generated a remainingsampleof 10UK InvestmentTrusts.The detailsof these trustscanbe foundinthe appendix (see table 2), which also outlines each trust Investment Objectives.
- 10. 10 Empirical research in Grinblatt and Titman (1989), Brown et al (1992), and Brown and Goetzmann (1994) on survivorshipbias may suggest implications in our study. For example, Malkiel (1995) findsthatsurvivingfundsconsistentlyhavehighermeanreturnsthannon-survivingfunds.We aimto analyse historical funddatawiththe intentionof providingfuture fundvalue prediction,which would not be relevant to terminated funds. Nonetheless, we consider this bias with caution our interpretations and conclusions. To construct an appropriate CAPMmodel forthe model we alsoretaindatasetfor the return on a risk-free assetandthe market return.The annualizedUS three-monthTreasuryBill rate is used torepresentthe risk-freereturnacrossthe observationperiod.Deductingthisvalue fromfundreturns providesuswithvaluesthatrepresenteachfundsexcessreturns. The marketproxyusedinthisstudy is the FTSE All-Share Index returns. This is considered the best performance measure of the London equitymarketandcaptures98%of UK’smarketcapitalization.Itisthe mostsuitableindexforanalysis on index tracking funds such as investments trusts, unit trusts and exchange-traded funds. Whilst being the predominant market benchmark in previous literature on UK Investments Trusts performance. In addition, and to gain a greater insight behind fund manager selectivityand performance, we retain information on the asset allocation of each fund.Data relating to portfolio composition is collectedfromthe AnnualReportsof correspondingInvestmenttrusts.Takingthe mostrecentannual report,we assume thatfundsmaintainanapproximatelyconstantindustryconcentrationthroughout their life, e.g each industry weighting remains roughly the same. The Morningstar Global Equity Classification Structure is used to categorise and define asset allocations into three major economic sectors; cyclical11 , defensive12 and sensitive13 . Within these ‘super sectors’ there the associated industrygroups(see Table 3). The applicationof thisclassificationinthisstudyenablesustoevaluate and compare each portfolio’s exposure to different sectors,whilst supporting understanding behind abnormal volatility and correlation with major economic events. 11 The cyclical super sector includes industries significantly impacted by economic shifts.When the economy is prosperous these industries tend to expand and when the economy is in a downturn these industries tend to shrink.In general, the stocks in these industries havebetas of greater than 1. 12 The defensive super sector includes industries thatarerelatively immune to economic cycles.These industries provideservices thatconsumers requirein both good and bad times, such as healthcareand utilities.In general,the stocks in these industries havebeta of less than 1. 13 The sensitivesuper sector includes industries which ebb and flow with the overall economy, but not severely so. Sensitiveindustries fall between the defensive and cyclical industries asthey are not immune to a poor economy but they also may not be as severely impacted by a poor economy as industries in the cyclical super sector. In general, the stocks in these industries havebetas that are closeto 1.
- 11. 11 The descriptive statistics are used to describe the basic features of the dataset and are presentedinTable 1. Data has beenseparatedintomonthlyreturnsandexcessreturns.Initial values showsthatforall fundsaveragemonthlyreturnswerepositive, butare all negativeformonthlyexcess returns. 3.1 Share price analysis over time Figure 1 showsaplotof the monthlyshare pricesof eachInvestmentTrustfromJanuary1995 to June 2016. Most of the assessedfundsfollow asimilarpattern,howevercertaintrustssharesshow high volatility level. Low performers include E and A funds. Medium performers includesfunds such as F, B, G and H. High performers where share price has grown at least six fold include J, D and C. In the yearsleadingupto2000 we can observe relative correlationbetweenamajorityof the funds.However,fundswithinvestmentobjectivesinAsiaandJapansufferedanepisode of decline in share price. Shroder Asia (B), Shroder Japan (E) and Atlantis Japan (A) were exposed heavily to the Asian financial crisis in July 1997 which saw several major companies including Nissan Mutual Life Insurance and Yaohan a Japanese retailer go into bankruptcy. Betweenthe monthsleadinguptothemillenniumand2001Lazard World (H/I) trustandmore evidentlyShroderUKMidCap (J) experiencedasteepspike intheirshare price’s.Thiscanbe justified by the speculative‘dotcombubble’whichsaw arapid rise inthe equitymarketsthroughexponential investment in internet-based companies. Both funds price peaked in December 2000 and then suffered a steepdecline until November 2001. Whereas, a majority of the remaining funds display a smaller change in their price movement. This implies that Shroder Mid Cap and Lazard World Trust are likelytohave heldaportfolioheavilyexposedtotechnologyandconsumerstocks,incomparison to the other funds. From 2003 all examined trusts show some recovery from the dot com crisis, and all share pricesdisplaysteadygrowthuntilmid-2007.Similarly,tothe spike seenin2000 ShroderMid-Capand Lazard World funds outgrow the other funds. However, ORYX International (C) and Shroder Income (D) bothfollowsimilargrowth.We detectadisparitybetweenthese fourfundsandthe remainingfive fundsduring2007 to 2008. The periodhasbeenlabelledbyfinancial commentatorsasthe ‘run-up’to the US housingbubble,fuelledbyextraordinarylow interestratesandthe reallocationof investment fromthe stockmarketintothe housingmarket.InAugust2007 the UK stockmarketsufferedextreme volatilityas a result heightened fears in the interbank market amidfears of exposure to high-risk US Mortgages.Thiswas followedbythe UKBank NorthernRock beingnationalisedinFebruary2008 and the collapse of LehmenBrothersinSeptember2008. Subsequently,assetsacrossall classesdeclined in value andthisis indicatedwiththe steepfall infundssharesbetween2008 and 2009. We observe
- 12. 12 the leastaffectedfundstothiscrisiswere AtlantisJapan (A) and ShroderJapan (E).Interestingly,the graphhighlightsthatinfactAtlantisJapanexperiencedadeclinewell beforethe crisisandotherfunds. Duringthe aftermathof the financial crisisof 2008 all fundsshare’sshow positive growthand a majorityrecovertotheirpre-crisisvalueby2012. Notably,we observe rapidgrowthinShroderMid- Cap (J) and Shroder Income (D) within this period. In contrast, a majority of the other funds such as CQS New City (G), Shroder Asia Pacific (B), Shroder UK (F) and ORYX International (C) follow similar patternbutat slowerrate.Interestingly,we alsoobserve the increasinggrowthof ORYXInternational (C) as well as Shroder Mid Cap (J) and Shroder Income (D).
- 13. 13 4. Methodology 4.1 Overall Performance This studyuses the unconditional single-factorperformance measure:Jensen’s(1968) alpha. This measure is the intercept term estimated through the the regression model: 𝑅 𝑖𝑡 − 𝑅 𝑓𝑡 = 𝛼𝑖 + 𝛽𝑖( 𝑅 𝑚𝑡 − 𝑅 𝑓𝑡) + 𝜖𝑖 (1) where 𝛼𝑖 isthe risk adjustedabnormal returnfromthe single index model, 𝑅 𝑖𝑡 isthe returnonfund i overperiod t, 𝑅 𝑓𝑡 is the returnon the 5 year US Treasuryyieldadjustedtoconstant maturity, 𝑅 𝑚𝑡 is the returnonthe FTSEALL-SHAREIndex, 𝛽𝑖 the factorsensitivityof differencesinfundreturnsandthe risk free rate, and 𝜖𝑖 is the error term with the following properties: E(𝜖𝑖,𝑡) = 0 , Var(𝜖𝑖,𝑡) = 𝜎𝜖 𝑖,𝑡 2 , Cov( 𝜖𝑖,𝑡,𝑟 𝑚,𝑡) = Cov( 𝜖𝑖,𝑡,𝜖𝑗,𝑡) = 0. The following hypothesis are tested with this model: 𝐻0: 𝛼𝑖 = 0 𝐻 𝑎: 𝛼𝑖 ≠ 0 where the null hypothesissuggests fund i does not out or under-performthe relative market proxy, the alternate hypothesis indicates fundi either out or under-preformsto the relative market proxy. Statistical significanceismeasuredusingcritical valuesusingatwo-tailedtestat5% significancelevel. Althoughfordiscussionpurposeswe alsoconsidercoefficientsthatgenerate a hightest-statisticbut are marginally insignificant. The intercept term identifies whether fund managers have superior stock selection abilities. Alphagenerationisachievedthroughselectingsecuritiesthatresultin 𝜖𝑖 > 0.A statisticallysignificant positive alpha indicates the fund manager has the abilityto forecast future securityprices. An alpha term that does not statistically differ from zero implies the manager mimics the composition of a reference marketbenchmark.A statisticallysignificantnegative alphatermssuggestthe fundmanager performs poorer than a naïve strategy of random selection. In the context of compensation, the annualizedJensen’salphaisthe maximummountof moneyaninvestorshouldbe willingtopayafund per year. Subsequently, this allows us to contribute to the literature surround pay performance in mutual funds. In addition,we alsoassessandcompare the riskof eachfund.UnderCAPMthe riskisdivided intotwo components.These are formallyexpressedasproductsof the standard deviationinreturns: 𝜎𝑖 2 = [𝛽]2 𝜎 𝑚 2 + 𝜎 𝑢,𝑖 2
- 14. 14 The firstpart [𝛽]2 𝜎 𝑚 2 indicatesthe market (or systematic) risk.Thistype of riskarisesdue to fluctuations in the market as a whole and cannot be eliminated through diversification. This is also interpretedintermsof ahigh 𝑅2 of the regressionmodel(1).We payspecial attentiontothisvaluein each funds model in our results analysis. The second part 𝜎 𝑢,𝑖 2 is referred to as specific (or non- systematic) risk. This is often attributable to managerial competence with the fund. In contrast to systematicriskthistypeof riskisidiosyncraticandthuscanbe eliminatedthroughsuccessfulportfolio diversification. A well diversified portfolio containing a great number of assets with different characteristics can cancel out such risk. We take the difference between 1 and 𝑅2, to estimate each funds idiosyncratic risk. Anothermeasure forrisk-adjustedreturnsthatwe use isthe Sharpe ratio. This performance method was developed by Sharpe (1966) under its birth name, the reward-to-variability ratio. Over the yearsithas gainedsignificantpopularityinthe financeandnow operatesan industrystandardfor measuring risk-adjusted performance. Incorporating standard deviation of portfolio returns against excess returns, the ratio has benefits in its comparability across fund categories as well its independence of any choice of benchmark (Jagric et al 2007). The measure is given by: 𝑆𝑖 = 𝑅̅ 𝑖 − 𝑅̅ 𝑓 𝜎𝑖 where 𝑅̅ 𝑝 is the meanreturn to portfolio i duringthe evaluationperiod, 𝑅̅ 𝑓isthe meanrisk free rate of return and 𝜎𝑖 is the standard deviation of the portfolio return. The higher the ratio, the better its returns have been relative to the amount of investment risk. We replicate (Bal and Leger, 1996) in estimatingthismeasureoverdifferenttime horizonstogauge itsvariabilityunderdifferenteconomic conditions and to provide further analysis on fund manager selectivity. 4.2 Market Timing Overall performance can be decomposed into fund managers stock selection and market timing abilities. To measure whether fund managers have the ability to predict market movements, we adopt an extended version of (1) from Treynor and Mazuy (1966) (TM hereafter). The model includes the additionof a quadratic term to equation (1) in order to account for non-linearityin the function of market return, and is as follows: 𝑅 𝑖𝑡 − 𝑅 𝑓𝑡 = 𝛼𝑖 + 𝛽𝑖( 𝑅 𝑚𝑡 − 𝑅 𝑓𝑡) + 𝛾𝑖( 𝑅 𝑚𝑡 − 𝑅 𝑓𝑡) 2 + 𝜖𝑖 (2) where 𝛾𝑖 is the risk adjusted measure of market timing ability of fund i, and 𝜖𝑖 is an error term with the following properties: E(𝜖𝑖,𝑡) = Cov( 𝜖𝑖,𝑡,𝑟 𝑚,𝑡) = Cov( 𝜖𝑖,𝑡,𝑅 𝑚,𝑡 2 ) = 0 The followinghypothesisistestedusingthismodel:
- 15. 15 𝐻0: 𝛾𝑖 = 0 𝐻 𝑎: 𝛾𝑖 ≠ 0 where the null hypothesisstatesthatmanagerinfund iexhibitsno positiveornegativemarkettiming ability.The alternate hypothesisstatesthatmanagerinfund i exhibitssignificantpositive ornegative timingabilities.Markettimingabilityisreflectedbygreatermarketexposurewhenthe excessmarket returns are higher and vice versa. A significant positive 𝛾𝑖 would indicate superior market timing ability. When 𝛾𝑖 does not deviate significantly from zero, the fund manager cannot outguess the market. A significantly negative 𝛾𝑖 implies perverse market timing from the manager. 5.Empirical Results 5.1 Selectivity Performance () Table 4 presents the results of applying the unconditional single-index model (1) to each individualfundinoursample overthe time period. The valuesforJensen’s (1965) alphaindicates the risk adjusted abnormal return in relation to the FTSE All Share market proxy, net of the manager’s timing ability. In testing the null hypothesis of no abnormal fund performance on individual Investment Trust, we apply t-distributioncritical values using 2-tailed laws. However, we also pay attentiontoall alpha’sthatcontainatest-statisticgreaterthan+-1.5andlie marginal tothe 5% critical value. Estimations show only three out of the ten funds generate a positive alpha term. Although, withinthis three onlyone fundgeneratesastatisticallysignificantpositivealphaata 5% level.FundB has an alpha value of 0.0097 with a t-statistic of 2.19. This suggest that the fund B has experienced positive abnormal performance relativetothe the FTSE All Share overthe observationperiod.Which providesevidence thatbasedon the unconditional single factormodel,the fundmanagerfromfund B exhibited superior stock selectivity over the period. Whereas the remaining positive funds F and J generate alphatermsthatdonosignificantlydifferfrom0at5% level,withalphavaluesof 0.0036 and 0.0079, respectively.Nonetheless,these valuesare relativelyhighand cause a marginal upheldinthe null hypothesis of no abnormal fund performance. This couldbe sensitive to changes in observation timeframe orusedmodel.Thus,we may infertoa smallerdegree thatfundmanagersinfundsF and J have shown stock selectivity skills over the observation period. The estimationsinTable1alsoshowsthatsevenouttenfundsgeneratenegative alphavalues. Unlike the positive alpha sample, all of the negative alpha are statistically insignificant at 5% level. Thus for fundsA, C, D, E, G, H, and I, the null hypothesisof noabnormal fundperformance isupheld. This impliesthatbasedon the unconditional single model,fundmanagersinthese sevenfundshave
- 16. 16 exhibited neither superior or inferior stock selectivity skills over the observation period. However, funds C, D and G generate relatively high t-statistics of -1.71, -1.6 and -1,51, respectively. Which suggestsamarginal upheldinthe null hypothesisthatissensitive tochangesin observationperiodor model design. This also implies with less strength that fund mangers in fund C, D, and G may have possessed inferior stock selectivity over the observation period. 5.2 Market Timing Performance () Table 5 displaysthe resultsinapplyingthe unconditional TM(1966) model to each individual fundover the observationperiod.The valuesforgamma indicatesthe fundmanagers markettiming ability. In testing the null hypothesisthat fund manager posses no market timing ability, we apply t- distribution critical values using a 2 tailed rule. However, we also pay notice to gamma values that generate a t-statistic that lies marginal to the 5% critical value. Furthermore, a comparison of alpha termsfromthe TM(2) model andthe singleindexmodel (1) ismade toassessthe impactof accounting for market timing. Coefficientestimatesshowthatonlytwo out of ten fundsgenerate a positive gamma value. Althoughthesecoefficientsare notsignificantata5% levelandhencethe null hypothesisof nomarket timingabilityis upheld.FundsDand F generate gammavaluesof 0.98 and 1.03, respectively,witht- statisticsof 1.82 and 1.91, respectively.Thisimpliesthatbasedonthe unconditionalTMmodel,fund managersinD and F exhibitneithersuperiororperverse markettimingabilitiesoverthe observation period. Howbeit, these values are marginal to the 5% critical value of 1.96. This may indicate tests againstthe null hypothesisare sensitivetoresearchdesign.Onthisbasis,resultsgivepartial evidence that fund managers in D and F have shown positive market timing skills. Subsequently,the majorityof fundsdisplaynegative gammavalues.Table 5showsthat eight from the tenfundsgenerate a negative gamma.Althoughof thiseight,onlythree are significanttoa 5% level.FundGyieldsagamma termof -2.85, FundH of -1.91, and FundI with -1.65. Notably,these are in fact all significantto a 1% level. Thismeans that the null hypothesis that fund managershave neither positive or negative market timing ability is decisively rejected. Hence,fund managers from fund’s G, H and I have shown evidence of perverse market timing against the market. Whereas the remaining five negative gamma funds of A, B, C, E and J show insignificant values to a 5% level. This implies that fund managers within these funds exhibit neither superior or perverse marking timing skills. Although,fund A has a gamma value which is marginal to the 5% critical value. Thus could be subject to choice of methodology and research design. In addition, the inclusion of the quadratic market term in the TM model highlights the potential bias present in coefficients in the unconditional single index model. In econometric sense
- 17. 17 thisis branded‘omittedvariable bias’,whichsuggestsestimatesmaycapture the effectof important omitted variables and thus provide a biased estimate. Table 5 shows that two alpha estimates have changedconsideratelyasa resultof the inclusionof markettiming. Firstly,the alphatermon FundB hasincreasedfrom0.0097 to0.011. Witha negative gammaestimate,one caninferthatthe omission of markettimingwascausingsome downwardbiasonthe stockselectivitymeasure (alpha).Secondly, the alpha value on Fund D has now become significant to a 5% level. The value has changed from - 0.004 to -0.005, implying that the omission of market timing was causing some upward bias on the selectivity measure in model (1). 5.3 Sensitivity to the market () Table 4 also shows the estimated β values of each individual Investment trust in relation to the FTSE All share market returns. This refers to the sensitivity of fund returns to market returns measure byFTSE ALL share excessreturns. Giventhe nature of Investmenttrustsandfundmanager’s objectivestotrackmarketindices, one wouldexpectβvaluesformostof the sample toapproximately equal 1. Which implies fund returns have a one-to-one relationship with the market returns. A beta value above 1 implies fund returns are more volatile than market, or termed by the financial professionals as ‘aggressive’ stock. A beta value between 0 and 1 implies fund returns are less responsive to market returns, generally termed ‘defensive’ stock. The estimated βvaluesinTable4indicate thatall the fundreturnsinoursamplemove closely withthe FTSE ALL share index.The fundbeta’soverthe observationperiodrange from0.665 forfund C to 1.29 for fund B. Notably, only three out of the ten funds generate a beta value greater than 1, these include Fund B, F and J. This indicates that these funds are more volatile than the market and behave similar to an ‘aggressive’ stock. Whereas the remaining sevenfunds of A, C, D, E, G, H, and I yielda betalessthan 1. These fundsare lesssensitive tomarketmovementsandbehave similartoa ‘defensive’ stock. In addition, the fund that tracks the FTSE All-Share the closest is fund I which generated an alpha value of 0.967 over the observation period. 5.4 Market and Idiosyncratic risk Table 1 alsocontaininformationregardingeachmodel’sexplanatorypowerinthe 𝑅2 values. These representthe percentageof Investmenttrustreturnsthatisexplainedby variationsinthe FTSE ALL share index. Statistically this is a measure of how well the data are to the fitted regression line, however in the financial context this measure is interpretedas systematic or market risk. This arises due to fluctuations in the market as a whole and cannot be eliminated by diversification. The percentage of Investment trust returns that is unexplainedby variations in the FTSE ALL share index is referred to as non-systematic or idiosyncratic risk. This can be attributable to managerial
- 18. 18 competence yet incomparisonto the systematicrisk,this type of risk can be eliminatedbyportfolio diversification. Values Table 4 and the unconditional single index model show that the highest 𝑅2 values is generated in Fund’s D, F, H and I. This implies that these funds contain the greatest market risk. Specifically,we observe the Fund F contains the highest market risk from the sample and hence the lowestidiosyncraticrisk.Thisisevidentthrougha 𝑅2 value of 0.7489. In comparison,FundCcontains the lowestmarketriskof the sample and hence the highestidiosyncraticrisk.Thisisevidentthrough a 𝑅2 value of 0.2335. 5.5 Sharpe Ratio’s Table 6 presentsthe calculationsforeach fundsSharpe ratio overseveral examinedperiods. Inthe periodof 1995 to2016 all fundsgeneratedanegativeratio.Thisimpliesthatthe average excess return over the 15 years was negative for all funds. Values show that Fund D had the most negative value of -0.6125 and thus its portfolio returned the greatest underperformance per unit of risk. Whereas, Fund J yielded the least negative value of -0.2795 and thus its portfolio returned the least underperformance perunitof risk.Examiningthe fundsacrossconsecutive inclusive5-yearperiodsof 2000 – 2004 and 2005 – 2009, showed that all funds generated negative Sharpe ratios over these separate periods. Notably, these alignwithourexpectationsgiventhe dotcome bubble in 2001 and the creditcrisisin 2008. Albeit,basedonthese ratio’s,the bestperformingfundsoverthe respective crises were Fund J and Fund B. This could be attributed to managerial competencies in each fund to minimise portfolio losses during economic downturns. Inthe mostrecent5-yearperiodof 2010to2014 inclusive,allfundsgeneratedpositive Sharpe ratios. These ranged from 0.0485 for fund I to 0.2194 for Fund C. The results for this period provide evidence that fund managers have become more positively rewardedfor taking additional risk, in comparison to the first few years into the millennium. 6. Discussion 6.1 Selectivity Performance The results in Table 4 provide evidence that a majority of fund mangers from the selected Investment trusts have not generated significant alphas over the observation period. This suggests mostfundshave notbeenable tooutperformthe market,andthusthe average fundmanageracross this sample does not possess stock selectivity skills. Howbeit, Fund B remains the anomaly to this in generatingasignificantalpha.These resultsshow consistencywithearlierstudiesthatdocumentthe insignificance inalphavaluesof UKInvestmentTrusts(Bal and Leger1996; Bangassa1999). Similarly, our resultsthatnine outof the tenanalysedfund’sgenerateinsignificantalpha’sfollowsCuthbertson
- 19. 19 et al (2010), who found 75% of UK mutual funds neither underperform not outperform their benchmarks. Nonetheless,theseresultsshowsome contradictiontostudiesthathave foundevidence of under-performance onariskadjustedbasisbythe average fundmanager(BlakeandTimmermann, 1998; Quigley and Sinquefield, 2000). The finding of one significantly positive alpha in our sample supports the view that fund managers contain superior selectivity skills relative to the market. The positive estimated alpha for only Fund B over the observationperiodshowsconsistencywithCuthbertsonetal (2008), who finds stock picking abilityin 5% to 10% of top performing UK equity mutual funds. This result also follows findingsfromUSstudiesof Kosowski etal (2006) and Barras etal (2005), whofindstrong evidenceof stock selectivity skills among top performing US funds. Moreover, this result for Fund B supports Banagassa et al (2012), that international funds show some favourable selectivity abilityand their commentary on the international diversification provides some relevance. They argue that international funds can benefit from international diversified portfolios and greater stock returns in global markets.ThiscouldpartiallyexplainFundB’sestimatedalpha,since the fundprimarilyinvests in equitiesof companiesacrossAsiaandfar easterncountriesborderingthe PacificOcean (see More recently, our results lend support to Verheyden and Moor (2015), who found only 6 out of 272 US equity funds generated a positively significant alpha from 2004 to 2014. Conversely, they attribute outperformance by fund manager’s ability to limiting losses in times of market inefficiency and by profiting from subsequent learning effects. Hence, we may infer that managers in Fund B have followed this strategy over the observation period. In addition,the estimates of the Sharpe ratio for Fund B providesfurtherevidence of strong managerial competency.Thisisshownforthe timeframeof 2005 to2009 inTable 6. Specifically,Fund B generated the smallest negative ratio over this examined period. This suggests that each unit of additional risk was associated with lower negative excess returns relative to all other funds in the sample.Therefore, wemayinferthatintermsof risk-adjustedreturns,FundBoutperformedthe other fundsthroughoutthe yearssurroundingthe 2008 Financial Crisis.Thismaypotential explainwhythe fundgeneratedtheonly significantpositive alpha. Thisisalsoevident of Fund’sFandJ,whobothhave positive alpha terms and have considerably smaller negative Sharpe ratios over this period. To this end, we can confidently assume that superior stock selectivity over the 15-year period was more apparent for firms with a smaller Sharpe ratio over the 2008 financial crisis. This also confirms the positive relationship between alpha and Sharpe ratio. Figure 2 presents Fund B’s asset composition. A majority of the fund’s investment are in industrieswiththecyclicalsupersector.Stockswithintheseindustriesare generallymore volatilethan
- 20. 20 the market,containingbeta’sgreaterthan 1. Linkingthiswiththe alpha term, we may inferthat the managerfromFundB has utilisedcyclical andvolatile stockstogenerateabnormal performance. This relationmaybe attributedtoluckorstockpickingskills,asobviousfromthe significantlypositive alpha estimate. Also, one can observe from Figure 2 that Fund B is the most heavily exposed to the real estate industry. Moreover, Fund B’s investment portfolio is more concentrated industries of technology, consumer cyclical, real estate and financial services. Such concentrationhas the potential to explain superior selectivity ability, as evidenced by a positive alpha. Fund managers may want to hold a concentratedportfolioif theypredictgrowthincertainindustriesoverothers,orif theyhavesuperior informationtoselectprofitable stocks14 .Thisinference issupportedby Kacperczyk etal (2005), who foundthatinvestmentabilityismore evidentamongmanagerswhoholdportfoliosconcentratedina few industries. Similarly,other studies have documented that focused fund managers outperform their more broadly diversified counterparts (Baks et al ,2006; Huij and Derwall, 2011). Nonetheless, our inference is less consistent with Sapp and Yann (2008), who do not support the view that fund mangers holding focused portfolios have superior stock picking skills. 6.2 Market timing The estimates fromthe unconditional TMmodel donot provide evidence insupportof fund managersexhibitingpositive markettiming overthe observationperiod.Resultsshow thatno funds generate significantly positive gamma, and three out ten fund generated a significantly negative gamma.ThiscoincideswithrecentresearchfromCuthbertsonetal (2009),whofoundthatonly1%of UK fundsdemonstrate positivetimingabilityand19% of fundsexhibitnegativetimingandonaverage miss-timethe market.These results alsosupportfindingsthatmarkettimingabilitiesinfundmanagers has diminished over time into the 90’s (Bal and Leger, 1996). In addition, our results follow Leger (1996), whofoundthat one inthree trustspossesnegative timingabilities.Likewise,Bangassa(1999) also found evidence of perverse timing practices in 72 UK Investment trusts over a 15-year period. Nonetheless,ourresultscontradicthisfindingsthatfundstylesassociatedwithJapan,NorthAmerica and Europe are the only funds to generate significant gamma’s. In our results, we observe no significant gamma values for Japanese styled funds such as Fund A (Atlantis Japan) and Fund E (Shroder Japan). Attemptsat market timinginportfoliomanagementcomprise tactical assetallocation,using financial derivatives,or rebalancing. Figure Evidence for significant perverse timing was found for 14 Levy and Livingston (1995) concludethat fund managers that have superior information should hold a relatively concentrated portfolio,under the mean-varianceframework. Van Nieuwerburgh and VeldKamp (2005) concludethat increasingreturns to scalein market learningshould causeoptimal under-diversification.
- 21. 21 fund’sG, andH. Figure 7 and Figure 8 show these fundscontainaheavyconcentrationof investment inthe financialservicessectorcomparedtoall otherfunds.Thismaysuggestthatthese fundmanagers have not beentactical withtheirassetallocationstrategiesoverthe observationperiod. Otherstudiesthat have founddifferencesinresultswhilstusingahigherfrequencydataset. Thisavenue of resultsshedspotential weaknessonourselecteddatatype inthisstudy.Banagassaet al (2012) find domestic funds show some favourable market timing ability. Bollen and Busse (2001) also highlight weakness in our measure of timing ability. They show mutual fundsexhibit significant timing abilitymore often in daily data. Goetzmannet al (2000) argue that monthly frequencymight fail tocapture the contributionof manager’stimingactivitiestofundreturnssincedecisionsregarding marketexposure are likelymade more frequentlythanmonths.Inlightof this,our resultsonmarket timing could be inaccurate. 7. Limitations & Future Research One of the mainissues withthis paperstemsfromthe dataselection.Thispaperuses surviving investmenttrusts monthlyreturns only andignoresfundsthathave terminatedoverthe observation period. As discussedearlier, this can lead to biases in estimatedcoefficients. Malkiel (1995) implied that the datasetused inthis studywill overstate the returnsto fund investors. AccordingtoGregory et al (2007) a problemincalculatingabnormal returnsusingfactormodel isthatit incorporateslook- ahead bias if funds are required to survive for a certain number of months. This empirical issue has also been argued in earlier research (Grinblatt and Titman, 1989; Brown et al, 1992; Brown and Goetzmann,1994). Though,a majorityof studieshave usedsamplesthatcontainsbothsurvivingand non-survivingfundstocounteractthisbias. Leiteetal (2009) arguedthatthisissue couldhave serious implicationsforstudiesusingasmallernumberof funds.Tothisend,the resultsinthisstudymustbe interpretedwith some precaution. An avenue of extension from this paper would be to record non- surviving funds intothe sample, withthe view of comparing non-surviving alphaswith the surviving fund alphas. Another limitation relatesto the frequencyof the dataset. This study collected the monthly share price of 10 UK InvestmenttrustfromJanuary 1995 to June 2016 to measure bothstock picking and market timing skills. However, empirical research in Bollen and Busse (2001) and Goetzmann et al (2000) support the use of dailyreturnswhenanalysingmutual fundperformance.Theyargue that manager’s decisions regarding market exposure are likely to happen more frequently than months. This issue hinders the accuracy of this papers estimations on fund managers markettiming skills. A responsive approachtothisissue andinthe contextof future researchinthisareaistouse funddaily returns to measure market timing skills.
- 22. 22 Additionally, the use of the unconditional Jensen (1965) and TM (1968) models to measure stock picking and market timing also pose limitations to this paper. These provide estimates that disregardinformationonthechangingnature of theeconomyandcanincorrectlymeasurealpha,beta and gamma.In reality,fundmanagersrespondtomarketinformationusingdynamicstrategieswhich often means varying alpha’s and beta’s. Ferson and Schadt (1996) shows that conditional models outperformunconditional ones.Thus,thispapersresultsof alpha and gamma shouldbe interpreted with some caution. Nevertheless, this study will be of value to a novice investor in due to its uncomplicated method’s used to analyse fund performance and managerial skills. Finally,the assumptionthatfundcompositionremainsapproximatelyconstantovertime islikely no to hold.Thisstudy usesthe most recentavailable Annual ReportsandMorningStar Trusts Data to retain each funds asset allocation. Subsequently, we draw some inferences in associating fund performance with the assumed asset allocation. However, portfolio holdings are likely change as managersmove away from riskyindustriesorwhena funappointsa new manager.On this account, the inferencesregarding stockpickingskills inthe concentratedfundsrelative tothe diversified funds will lack validity. 8. Investors Implications Testsof the performanceof mutual fundsare importantforinvestorschoosingbetween active andindex funds.Inthisstudy,resultssuggestonaveragemanagersfromUKinvestmenttrustare were not able to out perform the market. From an investors point of view, this provides discouraging information for the likelihoodof generating profit through holdingUK Investment trusts. Moreover, testsof performance ofmutualfundssupplyinformationonthe questionof thevalidityof theEfficient Market Hypothesisinthe InvestmentManagementIndustry.The resultsderivedinthispaper infact lendsupporttothe notionthatno fundmanagersare able tobeat the market.However,estimations show that Shroder Asia Pacicific fund (B) is an exception. In testing for the presence of managerial skills, this study providessome implications relatingto the controversial performance related pay that fund managers are rewarded with. Based on the results, on average fund managers from UK investment trust exhibit little or often perverse market timing. This result would also be of value to chairmen, committee and high shareholders in UK Investment trusts. Using these results, executives from both CQS New City High Yield Fund (G) and Lazard World Trust (H/I) may take actions to revaluate their technical andpredictive modelsthatare responsible for switching between asset classes.
- 23. 23 9. Conclusion Thisstudyevaluatestheperformanceof 10UKInvestmenttrustsoverthe periodJanuary1995 – June 2016. The unconditional model of Jensen’s (1968) alpha and Treynor and Mazuy (1966) are employed to investigate the presence of managerial stock picking and market timing skills. Results provide evidence thatInvestmenttrustsonaverage donot outperformthe market.Thisisevidenced by only 1 fund generating a truly positive alpha. Estimationsfor market timing indicate that 3 trusts exhibittrulynegative gammas.Resultswere consistentwithrelevantUKliterature inCuthbertsonet al (2008) andBangassa etal (2012) as well ascomplyingwithUSfindingsinKosowskietal (2006) and Barras et al (2005). Togetherthisadvocatesthatmanagerial skill ismore attributabletostockpicking as oppose to market timing abilities. In addition,Sharpe ratioswere estimatedfordifferenttime framesbetweenthe observation period detailed above. Results for this measure suggest the reward-to-volatility in Investment Trust was persistentlynegative between2000 and 2009. However, became positive in the period beyond 2010. Interestingly, we found that the funds with the lowest negative ratio (best performers) over 2008 crises period were the funds which generated positive alphas.Thirdly,the collection of fund’s asset allocation’senabled an insight into the relationshipbetween generated alphas and portfolio concentration. We observe thatShroderAsiaPacific(theonlyfundtogenerateasignificantlypositive alpha) holdsamore concentratedportfoliorelativetomostotherfunds.Thisassociationisconsistent with work of Kacperczyk et al (2005), Baks et al (2006) and Huij and Derwall (2011). These results propose the extending evidence that fund managers may exercise informational advantages in Shroder Asia Pacific Fund. Therefore, confirming that investment ability is more present in concentrated portfolios
- 24. 24 Bibliography Andre, P., Kooli, M. and L'her, J.F., 2004. The long-run performance of mergers and acquisitions: Evidence from the Canadian stock market.Financial Management, 33(4). Baks, K.P., Metrick, A. and Wachter, J., 2001. Should investors avoid all actively managed mutual funds? A study in Bayesian performance evaluation. The Journal of Finance, 56(1), pp.45-85. Bal,Y.andLeger,L.A.,1996.The performance of UKinvestmenttrusts.ServiceIndustriesJournal,16(1), pp.67-81. Bangassa, K., 1999. Performance of UK investment trusts: 1980–1994.Journal of Business Finance & Accounting, 26(9‐10), pp.1141-1168. Bangassa, K., Su, C. and Joseph, N.L., 2012. Selectivity and timing performance of UK investment trusts. Journal of international financial markets, institutions and money, 22(5), pp.1149-1175. Banz, R.W., 1981. The relationship between return and market value of common stocks. Journal of financial economics, 9(1), pp.3-18. Barras, L., Scaillet, O. and Wermers, R., 2005. False Discoveries in Mutual Fund Performance: Measuring Luck in Estimated Alphas, FAME Research Paper No. 163, University of Geneva. Barras,L.,Scaillet,O.andWermers,R.,2010.False discoveriesinmutualfundperformance:Measuring luck in estimated alphas. The Journal of Finance, 65(1), pp.179-216. Bartholdy,J. and Peare,P., 2003. Unbiasedestimationof expectedreturnusingCAPM. International Review of Financial Analysis, 12(1), pp.69-81. Bartholdy, J. and Peare, P., 2005. Estimation of expected return: CAPM vs. Fama and French. International Review of Financial Analysis, 14(4), pp.407-427. Basu, S., 1983. The relationshipbetween earnings' yield, market value and return for NYSE common stocks: Further evidence. Journal of financial economics, 12(1), pp.129-156. Blake, D. and Timmermann, A., 1998. Mutual fund performance: evidence from the UK. European Finance Review, 2(1), pp.57-77. Blake,David.&AlanTimmermann,1998, Mutual FundPerformance:Evidence fromthe UK,European Finance Review, 2: 57-77. Blanco, B., 2012. The use of CAPMand Fama and French Three Factor Model: portfolios selection. Bollen, N.P. and Busse, J.A., 2001. On the timing ability of mutual fund managers. The Journal of Finance, 56(3), pp.1075-1094. Brown, S.J. and Goetzmann, W.N., 1994, July. Attrition and mutual fund performance. In Journal of Finance (Vol. 49, No. 3, pp. 1055-1056). Brown, S.J., Goetzmann, W., Ibbotson, R.G. and Ross, S.A., 1992. Survivorship bias in performance studies. Review of Financial Studies, 5(4), pp.553-580.
- 25. 25 Capaul, C., Rowley, I., and Sharpe, W.F., 1993, “International Value and Growth Stock Returns”, Financial Analysts Journal, January/February, 27-36. Carhart, M.M. "On persistence inmutual fundperformance." The Journal of finance 52.1 (1997): 57- 82. Chevalier,J.andEllison,G.,1999.Are some mutual fundmanagersbetterthanothers?Cross‐sectional patterns in behavior and performance. The journal of finance, 54(3), pp.875-899. Connor,G. and R.A.Korajczyk(1986), `Performance MeasurementWiththe Arbitrage PricingTheory: A New Framework for Analysis', Journal of Financial Economics, Vol. 15, pp. 373±94. Coval, J.D. and Moskowitz, T.J., 1999. Home bias at home: Local equity preference in domestic portfolios. The Journal of Finance, 54(6), pp.2045-2073. Coval,J.D.andMoskowitz,T.J.,1999.The geographyof investment:Informedtradingandassetprices. Cuthbertson, K., Nitzsche, D. and O'Sullivan, N., 2008. UK mutual fund performance: Skill or luck?. Journal of Empirical Finance, 15(4), pp.613-634. Cuthbertson, K., Nitzsche, D. and O'Sullivan, N., 2010. The market timing ability of UK mutual funds. Journal of Business Finance & Accounting, 37(1‐2), pp.270-289. Daniel,K.,Grinblatt,M.,Titman,S.andWermers,R.,1997. Measuringmutual fundperformance with characteristic‐based benchmarks. The Journal of finance, 52(3), pp.1035-1058. Douglas,W.W., 1968. Stimulus‐secretioncoupling:the conceptand cluesfrom chromaffinandother cells. British journal of pharmacology, 34(3), pp.451-474. Drew,M.E., Naughton,T.,&Veeraraghavan,M.(2003).FirmSize,Book-to-MarketEquityandSecurity Returns:Evidence fromthe Shanghai StockExchange. Australian Journalof Management,28(2),119- 139. Elton,E.J., Gruber,M.J. and Blake,C.R.,1996. Survivorbiasand mutual fundperformance. Reviewof Financial Studies, 9(4), pp.1097-1120. Fama, E.F. and French, K.R., 1992. The cross‐section of expected stock returns. the Journal of Finance, 47(2), pp.427-465. Fama,E.F. and French,K.R.,1993. Commonriskfactors in the returnson stocksand bonds. Journalof financial economics, 33(1), pp.3-56. Fama,E.F.andFrench,K.R.,1995.Size andbook‐to‐marketfactorsinearningsandreturns. TheJournal of Finance, 50(1), pp.131-155. Fama, E.F. and French, K.R., 2010. Luck versus skill in the cross‐section of mutual fund returns. The journal of finance, 65(5), pp.1915-1947. Fama, E.F. and MacBeth, J.D., 1973. Risk, return, and equilibrium: Empirical tests. The journal of political economy, pp.607-636.
- 26. 26 Ferson,W.E.andSchadt,R.W.,1996. Measuringfundstrategyandperformanceinchangingeconomic conditions. The Journal of finance, 51(2), pp.425-461. Fletcher, J. (1995), `An Examination of the Selectivityand Market Timing Performance of UK Unit Trusts', Journal of Business Finance & Accounting, Vol. 22, pp. 143±56. Fletcher, J., 1997. An examination of UK unit trust performance within the arbitrage pricing theory framework. Review of Quantitative Finance and Accounting, 8(2), pp.91-107. Fletcher, J., 1997. An examination of UK unit trust performance within the arbitrage pricing theory framework. Review of Quantitative Finance and Accounting, 8(2), pp.91-107. Goetzmann,W.N.,Ingersoll Jr,J.andIvković,Z.,2000. Monthlymeasurementof dailytimers. Journal of Financial and Quantitative Analysis, pp.257-290. Grinblatt, M, S Titman, and R Wermers. "Momentum investment strategies,portfolio performance, and herding: A study of mutual fund behavior." The American economic review (1995): 1088-1105. Grinblatt, M. and Titman, S., 1989. Mutual fund performance: An analysis of quarterly portfolio holdings. Journal of business, pp.393-416. Henriksson,R.D.andMerton,R.C.,1981. Onmarkettimingandinvestmentperformance.II.Statistical procedures for evaluating forecasting skills.Journal of business, pp.513-533. Huij,J. andDerwall,J.,2008. “Hot Hands” inbondfunds. Journalof banking &finance,32(4), pp.559- 572. Jagric, T., Podobnik, B., Strasek, S. and Jagric, V., 2015. Risk-adjusted performance of mutual funds: some tests. South-eastern Europe journal of Economics, 5(2). Jegadeesh,N.andTitman,S.,1993.Returnstobuyingwinnersandsellinglosers:Implicationsforstock market efficiency. The Journal of finance, 48(1), pp.65-91. Jensen, M.C., 1968. The performance of mutual funds in the period 1945–1964. The Journal of finance, 23(2), pp.389-416. Jensen,M.C.,Black,F.and Scholes,M.S.,1972. The capital assetpricingmodel:Some empirical tests. Kacperczyk, M., Sialm, C. and Zheng, L., 2005. On the industry concentration of actively managed equity mutual funds. The Journal of Finance, 60(4), pp.1983-2011. Kosowski,R.,Timmermann,A.,Wermers,R.and White,H., 2006. Can mutual fund“stars” reallypick stocks? New evidence from a bootstrap analysis. The Journal of finance, 61(6), pp.2551-2595. Lai, M.M. and Lau, S.H., 2010. Evaluating mutual fund performance in an emerging Asian economy: The Malaysian experience. Journal of Asian Economics, 21(4), pp.378-390. Lakonishok,J,etal. "The structure and performance of the moneymanagementindustry." Brookings Papers on Economic Activity. Microeconomics 1992 (1992): 339-391. Lam, K., Li, F.K. and So, S., 2009. On the Validity of the Augmented Fama-French Four-Factor Model. Available at SSRN 1343781.
- 27. 27 Leger,L., 1997. On investmenttrusts:performance,timingandselectivity.AppliedEconomicsLetters 4, 207–210. Leite, P., Cortez, M.C. and Armada, M.R., 2009. Measuring fund performance using multi-factor models: Evidence for the Portuguese market. International Journal of Business, 14(3), p.175. Levy, A. and Livingston, M., 1995. Financial Markets, Institutions & Instruments: The Gains from Diversification Reconsidered: Transaction Costs and Superior Information. Blackwell. Lintner, J. "The valuation of risk assets and the selection of risky investments in stock portfolios and capital budgets." The review of economics and statistics (1965): 13-37. Malkiel, B.G., 1995. Returns from investing in equity mutual funds 1971 to 1991. The Journal of finance, 50(2), pp.549-572. Markowitz, H. "Portfolio selection." The journal of finance 7.1 (1952): 77-91. Maroney, N. and Protopapadakis, A., 2002. The book-to-market and size effects in a general asset pricing model: Evidence from seven national markets. European Finance Review, 6(2), pp.189-221. Nanda, V., Wang, Z.J. and Zheng, L., 2004. Family values and the star phenomenon: Strategies of mutual fund families. Review of Financial Studies. Porras, D. and Griswold,M., 2000. The value line enigmarevisited. Quarterly Journalof Businessand Economics, pp.39-50. Quigley, G. and Sinquefield, R.A., 2000. Performance of UK equity unit trusts.Journal of Asset Management, 1(1), pp.72-92. Roll, R., 1977. A critique of the asset pricing theory's tests Part I: On past and potential testability of the theory. Journal of financial economics, 4(2), pp.129-176. Rosenberg,B.,Reid,K.andLanstein,R.,1985. Persuasive evidence of marketinefficiency. TheJournal of Portfolio Management, 11(3), pp.9-16. Ross, S.A., 1976. The arbitrage theory of capital asset pricing. Journal of economic theory, 13(3), pp.341-360. Sapp, T. and Yan, X.S., 2008. Security concentration and active fundmanagement:do focused funds offer superior performance?. Financial Review, 43(1), pp.27-49. Sharpe, W. F. "Capital asset prices: A theory of market equilibrium under conditions of risk." The journal of finance 19.3 (1964): 425-442. Simpson, M.W. and Ramchander, S., 2008. An inquiry into the economic fundamentals of the Fama and French equity factors. Journal of Empirical Finance, 15(5), pp.801-815. Soumaré, I., Aménounvé, E.K., Diop, O., Méité, D. and N'sougan, Y.D., 2013. Applyingthe CAPMand the Fama–Frenchmodelstothe BRVMstock market. Applied FinancialEconomics, 23(4), pp.275-285. Tobin, J., 1958. Liquidity preference as behavior towardsrisk. The review of economic studies,25(2), pp.65-86.
- 28. 28 Treynor, J, and K Mazuy. "Can mutual funds outguess the market." Harvard business review 44.4 (1966): 131-136. Treynor, J.L., 1965. How to rate mutual fund performance. Harvard Business Review, 43, pp.63-75. Unlu, U., 2013. Evidence to support multifactor asset pricing models: The case of the Istanbul stock exchange. Asian Journal of Finance & Accounting,5(1), p.197. VanNieuwerburgh,SandVeldkamp,L,2005. InformationAcquisitionandPortfoliounder Diversification. NYUWorkingPaperNo.FIN-04-025 Verheyden, T., De Moor, L. and Van den Bossche, F., 2015. Towards a new framework on efficient markets. Research in International Business and Finance, 34, pp.294-308. Wermers,R., 2000. Mutual fundperformance:An empirical decompositionintostock‐pickingtalent, style, transactions costs, and expenses. The Journal of Finance, 55(4), pp.1655-1703. Data Sources Datastream.(2012) Thomson ReutersDatastream.[Online]. Available at:SubscriptionService (Accessed:17/07/2016) Lazard AssetManagement(2016). The Lazard World Trust FundAnnual Report(2016). Retrieved fromwww.theworldtrustfund.com Morningstar.(2016, August17) AtlantisJapanGrowthFund. RetrievedfromMorningstar InvestmentResearchdatabase. NewCityInvestmentManagers(2016). CQSNew CityHighYieldFundLtd InterimReport(2015). Retrievedfromwww.ncim.co.uk ShrodersUK. (2016). ShroderAsiaPacificFundplc.2015 Annual reportandAccounts.Retrievedfrom www.shroders.co.uk ShrodersUK. (2016). ShroderIncome GrowthFundplc. 2015 Annual reportandAccounts.Retrieved fromwww.shroders.co.uk ShrodersUK. (2016). ShroderJapanGrowth Fundplc.2015 Annual reportandAccounts.Retrieved fromwww.shroders.co.uk ShrodersUK. (2016). ShroderMid Cap fundplc.2015 Annual reportandAccounts.Retrievedfrom www.shroders.co.uk ShrodersUK. (2016). ShroderUK Gowth Fundplc.2015 Annual reportand Accounts.Retrievedfrom www.shroders.co.uk
- 29. 29 APPENDIX TABLES Table 1. Descriptive statisticsof monthlyreturnsandmonthlyexcessreturnsof eachInvestmentTrust,Risk-Freerate andFTSEAll Share index.ObservationPeriod1 January1995 to 1 June 2016. Monthly Returns Monthly Excess Returns Fund N Mean Std.Dev Min Max Skewness Kurtosis Mean Std.dev Min Max A 240 0.0066 0.088 -0.227 0.574 1.5684 11.2163 -0.0262 0.0936 -0.2983 0.5255 B 247 0.0082 0.082 -0.253 0.451 0.5845 7.5309 -0.0255 0.0873 -0.3208 0.3884 C 255 0.009 0.059 -0.297 0.3897 0.0035 12.746 -0.0256 0.0658 -0.3137 0.3846 D 255 0.008 0.044 -0.171 0.153 -0.0284 4.4616 -0.0265 0.0506 -0.2384 0.1396 E 257 0.0044 0.071 -0.174 0.333 0.4552 4.7102 -0.0304 0.0769 -0.2304 0.2709 F 257 0.0063 0.057 -0.212 0.2298 -0.1012 5.363 -0.0285 0.0632 -0.2588 0.2241 G 257 0.0069 0.0666 -0.389 0.2298 -1.1081 11.732 -0.02796 0.0709 -0.4258 0.2634 H 257 0.0055 0.055 -0.235 0.19 0.8412 5.7642 -0.0293 0.0604 -0.2723 0.1872 I 257 0.0056 0.05599 -0.2398 0.232 -0.707 6.1886 -0.0293 0.0617 -0.2723 0.2268 J 257 0.011 0.0857 -0.3101 0.462 0.3775 8.2159 -0.0243 0.0899 -0.3607 0.4059 US 3-month 258 0.03495 0.0241 0.0023 0.0736 -0.2646 1.4756 - - - - FTSE All Share 257 0.0068 0.04156 -0.13507 0.1215 -0.6252 4.0627 -0.028 0.0478 -0.1848 0.1166
- 30. 30 Table 2. Presentsthe code usedforthe 10 UK InvestmentTruststhatare analysedinthisstudy.Informationrelatingfundinvestment objectiveshasalsobeen extractedfromthe Annual ReportsandProspectus’of the funds. Code Fund Name Investment Objective A Atlantis Japan Growth Fund Investsinundervaluedgrowthcompaniesacrossthe marketcap range includingsome cyclical growth companiesthatcan do well overthe longertermsandthatlookcheapin termsof valuation.The Fund's objective istoachieve longtermcapital appreciationthroughanactivelymanagedportfolioof equityand equityrelatedinvestmentissuedbycompanieslistedinJapan. B Shroder Asia Pacific Growth Fund plc The Company'sprincipal investmentobjective istoachieve capital growththroughinvestmentprimarilyin equitiesof companieslocatedinthe continentof Asia(excludingthe Middle EastandJapan),togetherwiththe Far Easterncountriesborderingthe PacificOcean,withthe aimof achievinggrowthinexcessof the MSCIAll countriesAsiaexcludingJapanIndex inSterlingterms(Benchmark)overthe longerterm. C ORYX International Growth Fund The investmentobjective of the companyistoseekto generate consistentlyhighabsolute returnswhilst maintainingalowlevel of riskforshareholder.The companyprincipallyinvestsinsmall andmid-size quotedand unquotedcompaniesinthe UnitedKingdomandUnitedStates.The Investmentmanagertargetscompanies that have fundamentallystrongbusinessmodels,butwhere there maybe specificfactorswhichare constrainingthe maximizationorrealizationof shareholdervalue,whichmaybe realizedthroughthe persuitof an activistshareholderagendabythe InvestmentManager. D Shroder Income Growth Fund plc The Company'sprincipal investmentobjectivesare toprovide real growthof income,beinggrowthof income in excessof the rate of inflation,andcapital growthasa consequence of the risingincome. E Shroder Japan Growth Fund plc The Company's principal investmentobjective istoachieve capital growthfromanactivelymanagedportfolio principallycomprisingsecuritieslistedonthe Japanesestockmarkets,withthe aimforachievinggrowthin excessof the TSE FirstSectionTotal Returnoverthe longerterm. F Shroder UK Growth Fund plc The principal investmentobjective of the Companyistoachieve capital growthpredominantlyfrominvestment inUK equities,withthe aimof providingatotal return inexcessof the FTSE All-Share Index.The companyinvest ina relativelyconcentratedportfolioof between35 and65 stock principallyselectedfortheirpotential to provide shareholderswithattractivereturnsrelative tothe FTSEAll-Share Index.The portfolioisinvested
- 31. 31 primarilyinlistedUKequities.Itmayinclude convertible securities,andequity-relatedderivativesmaybe used for efficientportfoliomanagementpurposes.Stocksare predominantlyconstituentsof the FTSE350 Index. G CQS New City High Yield Fund Ltd To provide investorswithahighdividendyieldandthe potential forcapital growthbyinvestingmainlyinhigh yieldingfixedinterestsecurities. H Lazard World Trust Fund Seekstoachieve long-termcapital appreciationbyinvestingprimarilyincompanieswhose sharestrade ata discounttotheirunderlyingNetAssetValue.The Fundmeasuresitsperformance principallyagainstthe MSCI All CountriesWorldIndex,althoughLazardAssetManagementLLC(the 'Manager') seekstoachieve the highest possible risk-adjustedreturnsandthe allocationof the Fund'sassetswill normallydivergesubstantiallyfromthe Index,inparticularinrelationtoitsweightinginthe US marketswhichhistoricallyhasbeenrelativelylow. J Shroder UK Mid Cap Fund plc The Company'sinvestmentobjective istoinvestinMidCapequitieswiththe aimof providingatotal returnin excessof the FTSE 250 (ex-InvestmentCompanies) Index.
- 32. 32 Table 3. Presentthe MorningStarGlobal EquityClassification.Thisclassificationisusedtoproduce several portfoliosassetallocation. Super Sector Industry Definition Cyclical Basic Materials Companiesthatmanufacture chemical,buildingmaterialsandpaperproducts.Thissector alsoincludes companiesengagedincommoditiesexplorationandprocessing.CompaniesinhissectorincludeArcelot Mittal,BHP BillitonandRioTinto. Consumer Cyclical Thissector includesretailstores,autoandautoparts manufacturers,companiesengagedinresidential construction,lodgingfacilities,restaurantsandentertainmentcompanies.Companiesinthissectorinclude Ford Motor Company,McDonald'sand NewsCorporation. Financial services Companiesthatprovide financial serviceswhichincludesbanks,savingsandloans,assetmanagement companies,creditservices,investmentbrokerage firms,andinsurance companies.Companiesinthissector include Allianz,J.PMorganChase andLegg Mason. Real estate Thissector includesmortgage companies,propertymanagementcompaniesandREITs.Companiessinthis sectorinclude KimcoRealtyCorporation,VornadoRealtyTrustandWestfieldGroup. Defensive Consumer Defensive Companiesengagedinthe manufacturingof food,beverages,householdandpersonal products,packaging,or tobacco. Alsoincludescompaniesthatprovide servicessuchaseducation&trainingservices.Companiesin thissectorinclude PhilipMorrisInternational,Procter&Gamble andWal-Mart Stores. Healthcare Thissector includesbiotechnology,pharmaceuticals,researchservices,home healthcare,hospitals,long-term care facilities,andmedical equipmentandsupplies. Companiesinthissectorinclude AstraZeneca,Pfizerand Roche Holding. Utilities Electric,gas,and waterutilities.CompaniesinthissectorincludeElctricitede France,ExelonandPG&E Corporation. Sensitive Communication services Companiesthatprovide communicationservicesusingfixed-line networksorthose thatprovide wireless access andservices.Thissectoralsoincludescompaniesthatprovide internetservicessuchasaccess, navigationandinternetrelatedsoftwareandservices.CompaniesinthissectorincludeAT&T,France Telecom and VerizonCommunications. Energy Companiesthatproduce orrefine oil andgas,oil fieldservicesandequipmentcompanies,andpipeline operators.CompaniesinthissectorincludeBP,ExxonMobilandRoyal DutchShell. Industrials
- 33. 33 Companiesthatmanufacture machinery,hand-heldtoolsandindustrialproducts.Thissectoralsoincludes aerospace anddefence firmsaswell companiesengagedintransportationandlogisticsservices. Companiesin thissectorinclude 3M, BoeingandSiemens Technology Companiesengagedinthisdesign,development,andsupportof computeroperatingsystemsand applications.Thissectoralsoincludescompaniesthatprovide computertechnologyconsultingservices.Also includescompaniesengagesinthe manufacturingof computerequipment,datastorage products,networking products,semi-conductors,andcomponents.CompaniesinthissectorincludeApple,Google andMicrosoft. Table 4.
- 34. 34 Resultsof applyingthe unconditional single-index CAPMmodel (1) tosample of 10 UK InvestmentTrustsoverthe periodof 1 January1995 to 1 June 2016. * = significance level of 5%, ** = significance level of 1%. Fund alpha t(alpha) beta t(beta) 𝑅2 A -0.001 -0.18 0.94 8.71 0.2418 B 0.0097* 2.19 1.29 16.05 0.513 C -0.007 -1.71 0.665 8.78 0.2335 D -0.004 -1.6 0.824 19.8 0.6062 E -0.0037 -0.82 0.955 11.77 0.3519 F 0.0036 1.55 1.14 27.58 0.7489 G -0.0067 -1.51 0.759 9.51 0.2617 H -0.003 -1.03 0.939 17.73 0.552 I -0.002 -0.72 0.967 18.02 0.5601 J 0.0079 1.53 1.147 12.28 0.372
- 35. 35 Table 5. Resultsof applyingthe TMmodel (2) tothe sample of 10 UK InvestmentTrustsoverthe periodof 1January 1995 to1 June 2016 Fund alpha t(alpha) beta t(beta) gamma t(gamma) 𝑅2 A 0.002 0.38 0.764 5.37 -2.62 -1.88 0.253 B 0.011* 2.43 1.21 11.38 -1.22 -1.16 0.5153 C -0.006 -1.42 0.61 6.11 -0.81 -0.82 0.2356 D -0.005* -2.05 0.89 16.26 0.98 1.82 0.6129 E -0.0032 -0.69 0.93 8.66 -0.38 -0.36 0.352 F 0.002 0.98 1.21 22.24 1.03 1.91 0.753 G -0.003 -0.72 0.57 5.48 -2.85** -2.78 0.283 H -0.0007 -0.24 0.812 11.77 -1.91** -2.81 0.5655 I -0.0001 -0.05 0.858 12.21 -1.65** -2.38 0.5697 J 0.0096 1.78 1.055 8.56 -1.39 -1.14 0.3749 * = significantat5% level,**= significance at1%
- 36. 36 Table 6. Presentsthe resultsof applyingthe Sharpe ratiosacrossthe 10 InvestmentTrusts.The time framesexaminedinclude2000 to 2004, 2005 to 2009, and 2010 to 2014 as well asthe overall observationperiodestimatesof 1995 to 2016. These examinedwindowsrepresentedaninclusive5-yearduration. Fund 1995 - 2016 2000 - 2004 2005 - 2009 2010 - 2014 A -0.3222 -0.4908 -0.4926 0.1775 B -0.3262 -0.4689 -0.2570 0.0835 C -0.4398 -0.7718 -0.3735 0.2194 D -0.6125 -0.8132 -0.6071 0.1921 E -0.4303 -0.5986 -0.5312 0.1346 F -0.5026 -0.8735 -0.3444 0.1122 G -0.4212 -0.3816 -0.6439 0.1474 H -0.5355 -0.8025 -0.3863 0.0682 I -0.5242 -0.7993 -0.3635 0.0485 J -0.2795 -0.3359 -0.3143 0.1986
- 37. 37
- 38. 38 APPENDIX FIGURES Figure 1. Plotsthe monthlyshare price of the 10 analysedInvestmentTrustsoverperiodJanuary1995 to June 2016. (Source: Thomson Reuters DataStream [Accessed: 17/07/17]) 0 100 200 300 400 500 600 700 800 900 1000 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 ATLANTIC JAP.GW.FD (A) SHRODER ASIA PACIFIC (B) ORYX INTERNATIONAL (C) SHRODER INCOME (D) SHRODER JAPAN (E) SHRODER UK GROWTH (F) CQS NEW CITY HIGH YIELD (G) LAZARD WORLD TRUST (H) LAZARD WORLD TRUST (I) SHRODER UK MID CAP (J)
- 39. 39 Figure 2: Atlantis Japan Fund (A) Asset allocation Figure 3: Shroder Asia Pacific (B) Asset Allocation Cyclical Sensitive Defensive Basic Materials Consumer cyclical Financial Services Reals estate Communication Services Energy Industrials Technology Consmer defensive Healthcare Utilities Cyclical Sensitive Defensive Basic Materials Consumer cyclical Financial Services Reals estate Communication Services Energy Industrials Technology Consmer defensive Healthcare
- 40. 40 Figure 4: ORYX International (C) Asset allocation Figure 5: Shroder Income (D) Asset allocation Cyclical Sensitive Defensive Basic Materials Consumer cyclical Financial Services Reals estate Industrials Technology Consmer defensive Healthcare Cyclical Sensitive Defensive Basic Materials Consumer cyclical Financial Services Communication Services Energy Industrials Technology Consmer defensive Healthcare Utilities
- 41. 41 Figure 6: Shroder Japan (E) Asset allocation Figure 7: Shroder UK (F) Asset allocation Basic Materials Consumer cyclical Financial Services Reals estate Communication Services Energy Industrials Technology Consmer defensive Healthcare Utilities Cyclical Sensitive Defensive Basic Materials Consumer cyclical Financial Services Reals estate Communication Services Energy Industrials Technology Cyclical Sensitive Defensive
- 42. 42 Figure 8: CQS New City High Yield (G) Asset allocation Figure 9: Lazard World Trust (H/I) asset allocation Cyclical Sensitive Defensive Basic Materials Consumer cyclical Financial Services Reals estate Communication Services Energy Industrials Technology Cyclical Sensitive Defensive Basic Materials Consumer cyclical Financial Services Reals estate Communication Services Energy Industrials Technology Consmer defensive Healthcare Utilities
- 43. 43 Figure 10: Shroder Mid Cap (J) Asset allocation Figure toFigure 10 displayeachInvestmenttrustassetallocationcompiledfrominformationatthe MorningstarWebsite andAnnual Reportsof eachfund.Investmentsare arrangedthroughguidance fromthe MorningStarGlobal EquityClassification.The lefthandside chartsdisplaythe portfolios compositionintermsof supersectors.The righthandside charts displaythe portfolioscompositions interms of associatedindustries.Informationoncompositionisretainedthroughthe mostrecent and available assetallocationsource.Thispaperassumesthatfundamental compositionof each portfolioremainsapproximatelyconstantoverthe fundslife. Cyclical Sensitive Defensive Basic Materials Consumer cyclical Financial Services Communication Services Energy Industrials Technology Consmer defensive Healthcare Utilities
- 44. 44