Butler UniversityDigital Commons @ Butler University

Scholarship and Professional Work - Business College of Business

2011

Off the Rack Versus Savile Row: The Value ofCustom Tailoring for Equity InvestorsSteven D. DolvinButler University, sdolvin@butler.edu

John Gonas

Follow this and additional works at: http://digitalcommons.butler.edu/cob_papers

Part of the Finance and Financial Management Commons

This Article is brought to you for free and open access by the College of Business at Digital Commons @ Butler University. It has been accepted forinclusion in Scholarship and Professional Work - Business by an authorized administrator of Digital Commons @ Butler University. For moreinformation, please contact fgaede@butler.edu.

Recommended CitationDolvin, Steven D. and Gonas, John, "Off the Rack Versus Savile Row: The Value of Custom Tailoring for Equity Investors" (2011).Scholarship and Professional Work - Business. Paper 245.http://digitalcommons.butler.edu/cob_papers/245

Contributions D O L V I N I G O N A S

Off the Rack Versus Savile Row;The Value of Custom Tailoringfor Equity investorsby S teven D o l v i n , Ph .D. , CFA, a n d John Gonas , Ph.D.

Steven Dolvin. Ph.D., CFA, a finance professor at

Butler University since 2004. teaches primarily in

the area of investments and oversees student-run

portfolios in both public and private equity. He has

published more than 25 articles, with an emphasis in

initial public offerings and venture capital.

John Gonas, Ph.D., a finance professor at Belmont

Universiiy since 1998, primarily teaches corporate

finance, investments, economics, and an applied

portfolio management course. His research interests

include institutional investment management,

initial public offering pricing, and the economics of

charitable giving.

AS consumers, we must oftenchoose between a mass-produced item and one that

is more customized. For example, agentleman looking to buy a suit couldvisit the shopping mall and select froma number of "off the rack" garments orvisit Savile Row, a London street famousfor its high-quality custom tciilors. Asyou might expect, the price difference isquite large.

As with suits, investors—and financialadvisers—have the choice between amass-market product (such as a mutualfund) and a more customized approach(like a separately managed account[SMA]). The customized approach istypically more costly; however, it allowsfor customization in tax planning,investment holdings, and other key ele-ments. Although this trade-ofï is usually

Executive Summary

Separately managed accounts(SMAs) generally carry a higherfee structure than standard mutualfunds, but managers tout the abilityto customize accounts as beingworthy of this higher cost. This cus-tomization may increase returns, or itmay simply allow for more personal-ized tax management or control overother unique circumstances.Very little research exists on therelative return benefit of SMAscompared with actively managedmutual funds. We fill this gap byexamining firms that offer concur-rently managed funds—SMAs aswell as matching mutual funds runby the same manager(s) and follow-ing the same general strategy.We find that large-cap SMAsprovide no significant improvement

accepted, the more debatable issue iswhether this customization producesa higher return (is a higher-qualityproduct).

Because customized portfolio man-agement is costly, most brokerage firmstraditionally maintained minimumaccount sizes of $1 million to $5 mil-lion. Yet competition in the investmentadvisory industry and technological

in performance compared with themutual funds concurrently offeredby these managers.In contrast, we find that there doesappear to be a positive relativereturn (at least on average) forsmall-cap SMAs compared withtheir mutual fund counterparts.We also document that the SMAs(particularly small cap) most likely tooutperform mutual funds are thoseoffered by older firms and those thathave experienced larger SMA assetgrowth through returns; those withlarge net inflows from new accountstend to underperform.Although customization may benecessary for some, for mostinvestors an "off the rack" approachoffered by mutual funds may be thesimplest tactic.

advances in information managementhave resulted in a dramatic decline inSMA fees and minimum asset amounts.Foi" example, brokerage firms nowoffer SMA services to investors with aslittle as $50,000, and average SMA feesdeclined from 2.03 percent in 2000 to1.65 percent in 2005.'

However, grovrtb in SMA popularityresulting from the decline in minimum

54 JOURNAL OF FINANCIAL PLANNING I August 2011 www.FPAnet.org/Journal

D O I V I N I G O N A S Contributions

account sizes, management costs, andtrading fees has also had an adverseeffect on investment advisory firms'ability (or willingness) to customizeportfolios to investor-specific objec-tives. For example, in exchange for anadvisory firm's willingness to manageSMA portfolios with minimum accountsizes as low as $50,000, investors areoften restricted to target allocationsconsistent across all portfolios under thefirm's management.

The historical growth in the popular-ity of SMA accounts and the technologi-cal changes that allow an advisory firmto manage smaller account sizes moti-vate the two specific research questionsaddressed in this paper: (1) Do SMAsoffer higher returns than traditionalactively managed mutual funds concur-rently managed by the same firm? and(2) Are differences in SMA and mutualfund performance related to an advisoryfirm's ability to offer customized SMAportfolio management?

To answer these questions we analyzereturns of domestic equity SMAs andmutual funds (MFs). In addition, wefocus on a sub-set of SMAs and MFsconcurrently managed within thesame advisory firm. By examining theperformance of concurrently managedaccounts, we directly test whetherSMAs outperform MFs, holding theeffects of the asset manager constant.Our initial findings suggest that thereturns of large-cap SMAs and MFs arevirtually the same; there is no return (orquality) benefit from choosing a custom-ized approach. Thus, investors (or theirfinancial advisers) may only wish tofollow a customized route if specifictax-planning or similar issues are ofcritical importance. In contrast, we findthat the returns of small-cap SMAs aresignificantly higher, on average, thanthe returns from MFs. Thus, paying forcustomization at the small-cap level maybring multiple benefits.

Although small-cap SMAs usually

outperform MFs, this result is notconsistent across all firms. Thus, wealso examine which firm characteristicstend to be associated with increasedperformance. We find that older firmstend to provide a greater return fromcustomization, as do those experiencinga greater increase in relative assetsunder management (where SMA assetsare growing faster than MF assets).However, we find that an increase in netinfiow reduces return, suggesting thatgrowth through return is more favorablethan asset growth from new accounts,as an increased number of accountsmay prevent managers from offeringthe same level of customization (andassociated benefits).

Differences Between Managed Accounts andOpen-Ended Mutual FundsAlthough some confusion exists in theinvestment consulting industry as towhat actually constitutes a managedaccount, the industry generally classi-fies accounts into five program types,with the primary differences beingwho actually manages the portfolio,the types of investment products heldin the portfolio, and the degree ofcustomization available to an investor.These product types include open-ended actively managed mutual funds,fee-plus-commission SMA accounts,fee-only (wrap) SMA accounts, fee-onlymutual fund (wrap) accounts, andbroker-managed discretionary fee-onlyaccounts. Of relevance to this study,consultant separately managed accountprograms (the traditional SMA) includeall discretionary portfolios managedby unaffiliated investment advisersfor investors whose assets are held incustody at a broker-dealer.

An SMA typically has two differentcost structures. First, an investor canhave a fee-plus-commission arrange-ment in which the asset managercollects a percentage of assets undermanagement (usually paid quarterly)

and the broker-dealer receives a per-trade commission. Alternatively, underan all-inclusive wrap-fee arrangement,the investor pays for asset managementand broker-dealer fees simultaneously.The wrap-fee arrangement is mostcommonly used with retail-orientedbrokerage accounts, and the fee-plus-commission arrangement is primarilyused with relatively larger institutionalSMAs (minimum portfolio size of $5million to $10 million).

In the 1970s and 1980s, because ofthe higher transaction costs associ-ated with these accounts, fee-basedindividually managed accounts wereonly accessible to institutions, pensionplans, and high net-worth individuals.However, the development of newtechnology and trading software in theearly to mid-1990s allowed professionalasset managers to individually managefee-based portfolios in a fraction ofthe time and at significantly reducedcosts. For example, the developmentof specialized software now allowsbrokerage firms to disseminate a largetrade among hundreds of accounts withno manual bookkeeping.

One of the more obvious differencesin MFs and SMAs is that a traditionalopen-ended MF offers shareholdersa percentage ownership or stake in aco-mingled pooled account, whereas anSMA provides ownership of securitiesin custody at a broker-dealer. As aresult, the MF is commonly perceivedas more restrictive, given that investorscannot specifically request particulartrading guidelines. Conversely, SMAsoffer investors greater control overaccount management, particularly theability to restrict the manager's tradingactivity. The ability to set tradingparameters can be important forinvestors sensitive to incurring realizedcapital gains and losses.

Although MF managers may adhereto a tax-sensitive trading discipline interms of incurring realized capital gains.

www.FPAnet.org/Journal August 2011 I JOURNAL OF FINANCIAL PLANNING 55

Contributions D o L V I N 6 O N A S

they cannot meet specific capital gainsrestrictions for individual shareholders.Furthermore, new shareholders caninherit embedded capital gains realizedprior to purchase.^ In contrast, mostSMAs claim to offer asset managementguidelines that include client-specifictax-planning strategies. For example, ifa client wishes to withhold a particularsecurity from sale, the SMA assetmanager can set aside that position.

The last important difference betweenMFs and SMAs is the number of positionsusually held. A typical equity MF willhold in excess of 75 common or preferred

Í Í A traditional open-endedMF offers shareholdersa percentage ownershipor stake in a co-mingledpooled account, whereasan SMA provides ownershipof securities in custody at abroker-dealer.JJ

stocks, whereas a typical SMA will holdbetween 40 and 60 positions. EvenMFs concurrently mcinaged with SMAaccounts will commonly hold two to threetimes as many positions as their SMAcounterparts, diversifying among andwithin more sectors and industry groups.

Prior Research and BeyondPortfolio customization and investorcharacteristics aside, prior researchsuggests that actively managed SMAscan achieve superior performanceover co-mingled, open-ended MFs.For example, Padgette (1998) findsthat large-cap, domestic equity SMAportfolios outperformed similaropen-ended retail MFs from 1988 to

1997. However, Padgette only comparesgross returns, and thus his analysis doesnot control for the higher fees (trading,management, and administrative costs)charged to SMA investors. In this study,we compare net SMA returns to net MFreturns, adjusting for management andtrading costs.

Moreover, prior studies primarilyexamine SMAs as a whole versus MFsas a whole. Although this approachprovides some insight, it does not allowfor a more concrete comparison ofSMAs and MFs. Thus, we concentrateon investment managers that concur-

rently offer both MFs andSMAs, particularly thoseidentified as having the samebasic strategy, personnel, andinvestment approach. Thisspecific comparison allowsus to effectively control fordifferential managementand determine whether anystructural differences existbetween SMAs and MFsthat would affect investmentperformance.Advances in technology

in the early to mid-1990senabled advisory firms toconcurrently manage SMA

portfolios and MFs without losingefficiencies associated with trading andasset allocation. By comparing eachproduct within an advisory firm, wetarget portfolio and firm characteristicsthat can help explain potential dispari-ties in performance. In addition, we testfor differences in both large-cap andsmall-cap products.

We also consider whether activelymanaged portfolios (SMAs or MFs)offer risk-adjusted abnormal returns. Ifan adviser's management of a portfolioadds vcJue beyond the performance of apassively managed market portfolio, sheor he will generate positive risk-adjustedabnormal performance. SMA and MFrisk-adjusted abnormal returns (alphas)

are tbe intercepts estimated from thefollowing model:

FRETURN,, = a, + B,,RMRF,, + B,,SMB,,

B|,HMLi, -I- B,,PR1YR,, + e,,

whei-e FRFTURN is the quarterlyreturn of an SMA or an MF in excessof the risk-free rate. RMRF is the excessreturn of the Center for Research andSecurities Prices (CRSP) value-weightedmarket index. SMB, HML, and PRIYRwere obtained from Ken French andMark Carhart and represent factor-mimicking portfolio returns adjusted forsize, book-to-market, and momentum,respectively (see Fama and French 1993,and Carhart 1997). This approach allowsus to estimate the firm's alpha (a) todetermine whether any excess, risk-adjusted return is being earned.

V/e also consider whether firm char-acteristics pertaining to asset growth,reputation, and the decision to acceptsmaller SMA account sizes can explainperformance differences. For example,advisory firms that concentrate on man-aging relatively lcirger SMA portfolios(minimum account sizes in excess of$500,000) are assumed to offer greatercustomization. Conversely, advisoryfirms that participate in broker-dealersub advisory programs, and in exchangeare forced to accept account sizes aslow as $50,000, are less likely to tailorindividual portfolios to a client's specificinvestment objectives.

Data and Empirical MetiiodologyWe collected quarterly returns for asample of 115 pairs of SMA accountsand open-ended mutual funds forthe period 1998 to 2003.^ Our datasouixes are Prima Capital for SMAreturns, the CRSP for MF returns,and Morningstar for other MF data.We required that each MF and SMAfollow industry reporting standards inorder to minimize differences in returncalculation methodology, management

(1)

56 JOURNAL OF FINANCIAL PLANNING I August 2011 www.FPAnet.org/Journal

D O L V I N G O N A S Contributions

style categorization, and the treatmentof management and brokerage fees. Tocompare SMA returns with MF returns,we made three adjustments to theindividual returns. First, we adjustedSMA returns to reconcile a discrep-ancy between SMA and MF returncalculation methodology. Second, wereconciled differences in how SMAs andMFs define their investment objectives.Third, we reconciled differences in howreturns may or may not include specificmanagement or brokerage costs.

SMA returns include quarterlychanges in the market values ofportfolios within a single SMA productof an investment advisory firm. Incalculating returns, advisory firms usetheir own return calculation methodor the return calculation methodologysanctioned by the CFA "̂ Institute'sPerformance Presentation Standards(PPS).To minimize inconsistencies and

biases related to differences in SMAreturn calculation methods, we onlyinclude SMA accounts that adhere toPPS and thus best match standard MFreturn calculation methods.

MFs use a standardized calculationmethod to capture the change in netasset value (NAV) from one period tothe next. In contrast, SMA returns mustaccommodate cash fiows within specificportfolios. Because SMA accountsare conglomerations of individuallymanaged portfolios, calculating theirperiod returns presents two challenges:(1) accounts must be asset-weighted,where portfolio weightings shouldreflect meirket values as well as in-periodcash fiows, and (2) accounts mustaccommodate time-weighted cash flows,where sub-period returns are computedbetween cash flows and then geometri-cally linked.''

To meet these challenges, CFA

Institute Performance PresentationStandards suggest a process for calculat-ing a time-weighted composite return.First, each portfolio must accommodatean investor's cash withdrawals andcontributions by computing a weightingfactor for each cash flow. Second, thecomposite's period return is calculatedby capturing the cash-flow-weighte3change in the account's beginning mar-ket value. If an advisory firm does notadhere to the PPS calculation standards,its individual portfolio cash flows willhave an effect on distorting compositeSMA performance. Therefore, in orderto create a like SMA-MF comparison,we only compare MF returns to SMAreturns in compliance with PPS.

A second inconsistency across theSMA and MF accounts is the differencein the definition of a product's styleor management objective. Advisoryfirms follow a variety of self-prescribed

Have you struggled with communicatingmarket trends to your clients? Do you needhelp finding topics that are relevant and timely,or do you just want to build your reputation as aknowledgeable resource in your community?Let FPA's FINRA reviewed Financial PlanningPerspective articles (FPPs) help!

Get four timely and rele\ant articles lor allof your eommunication needs delivered eachmonth to your inbox.

Sign up and stai*jjsi(ig.)(out.a4ia?s today!

Learn more at www.FPAnet.org/FPP

Ayearot FINRAreviewed articles for$120members/S220

nonmembcrs.

FPArFINANCIAL PLANNING ASSOCIATION

www.FPAnet.org/Journal August 2011 I JOURNAL OF FINANCIAL PLANNING 57

Contributions D O L V I N I G O N A S

equity investment disciplines. AlthoughMorningstar is known to categorizemany MFs, the investment managementobjectives of most SMAs follow nostandardized categorization. Thus, wematch SMA and MF style descriptionsby pairing benchmarks self-prescribed byan advisory firm. As a result, our sampleonly includes domestic equity SMAcomposites and MFs benchmarked to anidentical list of self-proclaimed indices.̂

ííBased on the results, wefind that only two of thelarge-cap firms offer SMAproducts that significantly (ata 5 percent level or greater)outperform their concurrentlyoffered MFs. In fact, we findthat there are more firms(three) whose MF returns afc^significantly higher.55

Third, SMAs and MFs have differentmechanisms for how investors pay forbroker-dealer services and advisory firmasset management. The open-ended MFhas an internal management fee (paidquarterly as a percentage of the fund'smarket value), trading commissions,and a selling commission (paid at thetime of purchase 2ind/or sale for "loaded"funds). Selling commissions only applyto loaded funds sold by third parties(banks, brokerages, etc.), but are tieredbased on an investor's purcheise amount.Therefore, larger retail and institutionalloaded MFs offer reduced or no sellingcommissions when an investor's purchaseamount reaches break points.

Conversely, an SMA typically offersinvestors the opportunity to pay forall broker-dealer and asset manage-ment costs with one fee. To reconciledifferences between MF multiple fees(internal, commissions, and loads) withthe SMA all-inclusive fee structure, wecompare SMA returns net of manage-ment and trading costs to MF returnsnet of management and trading costs,but not selling commissions.

Three observations motivatedthe decision to exclude sellingcommissions from MF returns:(1) many MFs in our samplescire no-load and thus do notcarry a selling commission,(2) most loaded MFs in oursamples are offered throughbroker-dealer fee-onlyprograms that enable investorsto purchase them at NAV (noload) while paying a separatefee based on assets under man-agement, and (3) most loadedMFs offer reduced loads whena shareholder invests a certainamount within the MF's fundfamily. As a result, loaded MFsusually offer a variety of sellingcommission arrangementsbased on criteria specific toinvestor characteristics.

Prior to comparing SMA and MFreturns, we also adjusted SMA returnsfor management and trading costs.Unfortunately, SMA managers do notreport management and trading costinformation. Thus, we adjusted the SMAreturns using the average managementfee prevailing in the market based ona survey of five representative broker-dealers. We assumed that competitionfor assets from sophisticated SMA inves-tors ensures that SMA managementfees are roughly consistent with theseestimates.'' For example, the averagewrap fee on the smallest accounts is2.4 percent of AUM, and the largestaccounts (greater than $5 million) have

average wrap fees of 1.1 percent. Wenote that the range between the maxi-mum and minimum wrap fee declinesas account size increases, indicating thatcompetition among broker-dealers is themost intense for large accounts.

We adjusted SMA returns for theaverage annual wrap fee as follows:

where NR,, is the net-of-fee quarterlySMA return for fund i at quarter t, W,,,,is the? average annual SMA wrap feecorresponding to the minimum accountsize m for fund i, and GRi, is the gross-of-fee quarterly SMA return for fund i atquarter t. All analysis is conducted usingSMA net-of-fee returns.

Lastly, we excluded from the analysisall funds having names that indicateconcentrations in convertible securities,REITs, or sector funds. In addition, weexcluded all SMAs or MFs that follow apassive index management style, hold amajority concentration in internationalsecurities, or include "all-cap" or "mid-cap" in their names.'

Empiricai Resuib: Comparing PerformanceTable 1 reports the summary statisticsfor the 115 concurrently managedSMA and MF products. We compareSMA and MF return performance andportfolio characteristics. For return,we report the difference between theSMA and MF products offered bythe investment managers—for netquarterly return as well as for alphas, asdescribed in Equation 1. For portfoliocharacteristics, we include the differ-ence!S in portfolio turnover, averagenumber of positions, cash position,assets under management, and averageannual six-year net inflow.' Turnoverand the average number of positionsare directly controlled by the invest-ment adviser who has discretion overthe management £md allocation of the

58 JOURNAL OF FINANCIAL PLANNING | August 2011 www.FPAnet.org/Journal

D O L V I N | G O N A S Contributions

fund's portfolio. The decision of whereto set an SMA's or MF's minimuminitial investment is directed by thefund's advisory firm, whereas a fund'sassets under management and annualnet inflow are market-driven variablesthat capture investor sentiment.

Examining Panel A in Table 1, wefind that the quarterly return of SMAsis about 17 basis points higher thanthat of the comparable set of MFs, andthis difference is significant at the 10percent level. However, examiningPanel B reveals that this performancedifference is completely driven bysmall-cap SMAs, as the small-cap SMAsoutperform their concurrent MFs byabout 28 basis points per quarter. Thisresult is even more pronounced withthe alpha difference, as across theentire sample the risk-adjusted returnis insignificantly different betweenSMAs and MFs, whereas the alphas forsmall-cap SMAs are significantly higherthan those of the comparable set of MFs.Thus, it appears that SMAs may providesome performance enhancement, butprimarily for smcdl-cap funds.

Continuing the analysis, we find thatportfolio turnover is lower in SMAaccounts, suggesting that trading isless aggressive; however, this result isprimarily confined to large-cap funds.The number of positions held in SMAsis significantly smaller across all fundtypes, likely as a result, in part, of thediversification requirements imposedon mutual funds. But we find that thereis no difference in the amount of cashheld, suggesting that allocation strategymay be similar across products. Onaverage, the assets under managementwithin an advisory firm's SMA portfoliosare significantly larger than the assetsunder management within a firm's MFs.

The net inflow difference is insignifi-cant across tbe whole sample; however,there appears to be an offsettinginfluence in the large-cap and small-cap samples. Specifically, annual net

Table 1 : Summary Statistics, Concurrently Managed SMAs and MFs

Panel A: All SMAs and MFs fljjt

All(n=n5)

Quarterly Return Difference (%)

Alpha Difference (%)

Annual Portfolio Turnover Difference (%)

Number of Stocks in Portfolio Difference

Average Cash Position Difference (%)

Product AUM a/o 2003 Difference (Sbillions)

Annual Net Inflow Difference (%)

Panel B: Large and Small-Cap SMAs and MFs

Large-Cap (n=S8)

Quarterly Return Difference (%)

FFC Alpha Difference (%)

Annual Portfolio Turnover Difference (%)

Number of Stocks in Portfolio Difference

Average Cash Position Difference (%)

Product AUM a/o 2003 Difference (Sbillions)

Annual Net Inflow Difference (%)

Small-Cap (n=S7)

Quarterly Return Difference (%)

FFC Alpha Difference (%)

Annual Portfolio Turnover Difference {%)

Number of Stocks in Portfolio Difference

Average Cash Position Difference (%)

Product AUM a/o 2003 Difference (Sbillions)

Annual Net Inflow Difference (%)

Mean

0.166(5.122)0.758

(0.975)-3.254

(29.698)-13.292(35.997)

0.423(3.428)1.797

(4.474)2.602

(169.23)

Mean

0.056(6.32)-0.084(1.148)-9.724

(30.026)-12.810

(5.536)0.475

(3.235)2.828

(6.056)

-15.628(50.53)

Mean

0,278(3,488)0.238

(0.735)3.446

(28.068)-13.800

(22.323)0.371

(3.643)0.766

(1.244)

19.345(229.242)

Median

0.100

0.900

3.500

-5,000

0.855

0.514

-0.523

Median

0.004

-0.033

-2.500

-3.500

0.595

0.840

-3.395

Median

0.122

0.200

6.000

-7.000

0.967

0.377

-0.062

r-Statistics

1.71*

0.83

-1.17

-3.93***

1.32

4.29***

0.15

^ ^ ^ ^ ^

(Statistics

0.33

-0.55

-2.46**

-2.14**

1.12

3.52***

-2.07**

(-Statistics

2.94***

2.44**

0.92

-4.58**

0,76

4,65***

0.59

Note: f-Statistics are the results from significance tests (means different from zero). Standarddeviations are in parenthesis.*,**,and *** significance at a 10%, 5% and 1% level., respectively.

inflow into the large-cap SMAs appearsto be lower than the respective mutualfunds, but it is higher (though insig-nificant) for small-cap SMAs comparedwith their respective mutual funds.This may be a function of the periodunder review.

Evaluating Risi(-Adjusted PerformanceEach individual firm's concurrentlymanaged SMA and MF products werecompared to determine whether asignificant difference in risk-adjustedreturns exists (the model given inEquation 1 was estimated for each SMA/

www.FPAnet.org/Journal August 2011 I JOURNAL OF FINANCIAL PLANNING 59

Contributions D O L V I N I G O N A S

MF pair in the sample).' Based on theresults, we find that only two of thelarge-cap firms offer SMA productsthat significantly (at a 5 percent levelor greater) outperform their concur-rently offered MFs. In fact, we find thatthere are more firms (three) whoseMF returns are significantly higher,suggesting that as a general rule, large-cap customization does not provide anincreased return.

The small-cap sample reveals 13firms with significant abnormalperformance, almost a quarter of thesample. In addition, we find that thereis no occurrence of the MF productoutperforming the SMA, To determinewhether this difference is significant forthe group as a whole, we estimate theaverage consolidated alpha, finding thatit is significcmtly different from zero,with a t-statistic of 2,43. Thus, althoughall small-cap SMAs do not outperformthe concurrently offered mutucil funds,our evidence suggests that a significantnumber of them do.

Evaluating Firm CharacteristicsIn an attempt to help investors andfinancial advisers determine whichfirms are most likely to have SMA prod-ucts that outperform MFs, we extendour analysis by examining underlyingfirm characteristics. In particular, wefocus on the following nine variables:

• D.NETINFLOW, the average annualdifference in net infiows betweenconcurrently managed SMAs andMFs

• D.LNAUM, the average annualdifference in SMA and MF assetsunder management

• D.AUMGROWTH, the averageannual difference in percentagegrowth in SMA and MF assetsunder management from 1998to 2003

• ACCTGROWTH, the annual aver-age percentage growth in size for allSMAs managed by an investment

management firm• FFGROWTH, the average annual

percentage growth in size for allSMAs managed by an investmentmanagement firm categorized aseither an endowment or foundation

• MORNINGSTAR, the averagefive-year Morningstar rating forall mutual funds managed by aninvestment management firm

• AGF, the number of years aninvestment management firm hasbeen in existence (as of 2003)

• MINIMUM, a binary variable thatcaptures whether an advisory firm'sSMA minimum account size is lessthan $500,000, zero otherwise

• SUBADV, a binary variable thatcaptures whether an advisory firmmanages retail SMA portfolioswithin a broker-dealer program,zero otherwise

We use these variables in an attemptto capture an advisory firm's transpar-ency in the marketplace as well as thelevel of investor monitoring of its funds.Industry experts suggest that greatermarket visibility is correlated to agreater influx of new capital. As a result,we test whether an advisory firm's abil-ity or decision to acquire more accountsis related to an inability to customizeits SMAs, therefore leading to less ofa performance difference between itsconcurrently managed SMAs and MFs.

It is possible that some of thevariables (particularly the final six) arerelated and, therefore, not capturingunique characteristics. However, weconduct a correlation analysis, findingthat there is no significant correlationamong any of the variables. In fact, mostare at 0.2 or lower, with the one excep-tion being the relationship betweenaccount growth and a sub-advisory rela-tionship. Even this, however, is generallybelow 0.5. The low correlations amongthe variables indicate that the variablesare likely capturing different aspects ofthe firm.

T(3 examine the relative influenceof each variable on firm performance,we 1 egress the differences in abnormalperformance on the nine variablesidentified above. We report the resultsof this analysis in Table 2. Although ourprior results suggest that any significantdifference in return is concentratedprimarily among small-cap firms, we dofind two variables that are significant inall regressions. Specifically, differentialgro^vth in assets under management ispositively related to outperformance ofSM.̂ s versus MFs; however, in contrsist,differences in net inflow are negativelyrelated. Therefore, the greater thedispEurity in an advisory firm's SMA-MFassets under management, the greaterthe division in how the two producttypes are managed.

These findings may also suggest thatpas': performance (which would leadto an increase in relative assets undermanagement) is a key factor for finan-cial advisers to focus on, and increasedinfiow from new investors should raiseconcern as it likely prevents managersfrom continuing to implement corestrategies and provide the same levelof customization. The other significantvariable, but only in the small-capsample, is age. The results suggest thatolder firms are generally associatedwith increased performance differences(SA4As are more likely to outperformtheir MF counterparts).

ConciusionAs with a gentleman choosing betweena custom-tailored suit and one off therack, financial advisers are often facedwith the choice of using separatelymanaged accounts or actively managedequity mutual funds. There is littledebate that a customized SMA offersthe opportunity to manage accountswith taxes (or other issues) in mind.However, there previously existed littleevidence as to the return benefit of suchcustomization.

60 JOURNAL OF FINANCIAL PLANNING I August 2011 www.FPAnet.org/Journal

D O L V I N I G O N A S Contributions

We address this particular issue by

focusing on SMAs and MFs offered by

the same management firms, thereby

allowing us to control for the infiuence

of management. We find no difference

in performance for large-cap funds;

however, we do find that a significant

portion of small-cap SMAs outperform

their MF counterparts.

We further examine which firm

characteristics are related to this out-

performance. Our findings suggest that

financial advisers looking for an SMA

should focus on those firms that are older,

have positive growth in assets under

management (particularly SMAs relative

to MFs), and have moderate amounts of

net new inflow. We surmise that relative

growth of assets under management

causes firms to place more emphasis on

the management of SMAs (because that

is where the majority of growth exists);

however, increased net inflow into these

accounts prevents them firom continuing

to offer the extent of customization that

produced the performance.

As a whole, we believe our results

suggest that for most investors, SMAs

are like a Savile Row suit—nice to have,

but an off-the-rack mutual fund would

serve the same purpose. The one excep-

tion might be those clients who have

unique tax (or related) issues. However,

should a customized route be chosen,

we believe that greater return benefit is

likely achieved on the small-cap side.

Table 2: Multivariate Results, Return Differences

Variable

Endnotes1. Cerulli Associates.

2. For example, assuming a fund pays its capital

gains distribution.? to its shareholders in Novem-

ber, a new shareholder who purchases the fund

in October (prior to its ex-dividend date) will

fully participate in such a distribution.

3. Data availability is restricted to 1998-2003

because of availability of proprietary data as well

as limited access to professionals who concur-

INTERCEPT

D.NETINFLOW

D.LNAUM

D.AUMGROWTH

ACCTGROWTH

EFGROWTH

MORNINGSTAR

AGE

MINIMUM

SUBADV

Number of Observations

Adjusted R-Square

1 Differenc

Total

-0.004(-0.51)

-0.013***(-3.33)-0.001

(-0.37)

0.012***(3.39)0.001(0.03)

-0.001(-0.28)

0.001(0.39)0.0002

(0.87)-0.004

(-1.50)0.003(0.76)

115

:es

Large-Cap

-0.013(-1.04)-0.014*»

(-2.45)-0.001

(-0.52)0.018**

(2.10)-0.009

(-0.42)-0.009

(-1.28)0.003

(1.17)0.0002(0.32)

-0.006(-1.08)

0.004(0.59)

58

Small-Cap

0.005(0.70)

-0.021**(-2.42)

0.001(0.87)

0.018**(2.55)

-0.008**(-0.68)

0.003(1.06)-0.002

(-0.95)9E-4**(2.53)

-0.001(-0.29)-0.004

(-1.41)57

0.113 0.19 0.096

Note: f-statistics are in parenthesis and are results from significance tests (coefficients differentfrom zero). *, **, and *** significance at a 10%, 5% and 1 % level, respectively.

rently manage both SMAs and open-ended MFs.

4. CFA Institute Guidance Statement of Calcula-

tion Methodology

5. We categorize MFs and SMAs as large-cap funds

if they track one of the following indices: Russell

1000, Russell 1000 Growth, Russell 1000 Value,

S&P 500, S&P 500 Barra Growth, S&P 500

Barra Value, Wilshire Target Top 750, Wilshire

Target Large Company Growth, or Wilshire

Target Large Company Value. Small-cap funds

track one of the following indices; Russell 2000,

Russell 2000 Growth, Russell 2000 Value,

S&P Small Cap Barra 600, S&P Small Cap

600 Growth, S&P 600 Small Cap Barra Value,

Wilshire Target Small Company Growth, or

Wilshire Target Small Company Value.

6. We conducted separate analysis using the

mid-point, minimum, and maximum fee and

found that the results are virtually unchanged.

7. Parameters used in Ennis and Sebastian (2002)

and Del Guercio and Tkac (2002).

8. The portfolio turnover, number of positions,

and AUM reflect year 2003 only. These

characteristics are unavailable for earlier years.

9. To conserve space, we have elected not to

report the 115 regression results; however,

these are available upon request.

ReferencesCarhart, M. 1997. "On the Persistence in Mutual

Fund Performance." Journal of Finance 50:

57-82.

Del Guercio, Diane, and Paula Tkac. 2002. "The

Determinants of the Flow of Funds of Managed

Portfolios: Mutual Funds Versus Pension

Funds." Joumal of Finance and Quantitative

Analysis 37: 523-557.

Ennis, Richard, and Michael Sebastian. 2002. "The

Small-Cap Alpha Myth: No Evidence That Active

Small-Cap Managers Add Value Consistently."

Joumal of Portfolio Management 28:11-17.

Fama, Eugene F., and Kenneth R. French. 1993.

"Common Risk Factors in Returns on Stocks and

Bonds." Joumal of Financial Economics 33: 3-56.

Money Management Institute. 2010. 2010-2011

Industry Guide to Managed Investment Solu-

tions. 200.

Padgette, Robert. 1998. "Separate Account

Manager Versus Mutual Fund Performance."

The Mobius Strip 5: 1, 6-12.

www.FPAnet.org/Journal August 2011 I JOURNAL OF FINANCIAL PLANNING 61

Copyright of Journal of Financial Planning is the property of Financial Planning Association and its content

may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express

written permission. However, users may print, download, or email articles for individual use.