?
aa
Equity Focus
Active Management: Three Arguments for Stock Selection
by Joseph M. Graham, CFA, Lord Abbett Investment Strategist
IN BRIEF
1 The demand far high tracking error, “satel
equity strategies is growing. However, these strategies
often come with risk-factor exposures that may outweigh any edge gained from stock selection,
1 We believe there are three reasons why investors should be aware of the portion of a manager's returns
derived from stock selection versus factor exposure:
1a Factor exposures are no longer worth active management fees.
‘Active managers are better at stock selection than they are at factor timing.
1 Stock selection is more diversifiable than factor exposures.
The key takeaway: Strategies that focus on diversifiable, repeatable, stock-specific risk may offer a
‘multi-manager platform more consistent risk-adjusted returns,
‘The recent landmark US, Department of Labor ruling
oon fiduciary standards is just the latest example of a
pervasive trend of increasing serutiny applied to allo~
ators’ portfolio choices, In many cases, the scrut
legal consequences is motivating a shift to
and risk ofl
passive and index-based investment strategies. But in
4 post-quantitaive easing, low-rerurn environment
for most asset classes, the need for active rear has,
arguably, never been greater. To balance these needs,
many allocators have turned to barbell approach,
allocating very litle ative risk vo che usually pasive
core portion of atypical core-satlliteportiolio—and
4 lot of active risk to the satellite portion. Active
‘managers, in turn, have adopted an “active a¢ ll
costs” approach to ft this mode, targeting high
tracking error as a necessary strategy differentiator
‘We think both the passive and active sides of a barbell
strategy have merit for many investors. But confining
active management to a high-risk functionality limits
its reach and effectiveness, Often, the intention of
high tracking error managers isto isolate the impact
research and stock-selection efforts by not confining a
‘manager’ choices to index constituents and by allow-
ing heavy concentrations in individual stocks, sectors,
and styles, While well-intentioned, this approach to
portfolio construction can lead to uncompensated risks.
Specifically, the lack of diversification in “conviction”
portfolios gives rise co rsk-factor exposures that may
outweigh any edge gained from security selection. Chart
1 shows this counterintuitive phenomenon.' As the
funds depicted in the chart exhibit more tracking erro,
less ofthat tracking error comes from stock seleetion.
‘CHART 1, SOURCES OF VARIANCE AMONG THE LARGEST MID- AND LARGE-CAP VALUE FUNDS.
a
Fund
itregreeson
‘% of Variance from Stock- Specific Ris
Wack Ero
Source: Morningstar, Asam, Lipper, ndLord Abel Aso December 21,205. Data points represent values othe 10 largest mtu funds in
tw Lipper categories: Mistap Value and Large-Cap Value Fo stale puoeses On ha does Pa regreset any pro manages by Lar el.
Uranyparicuat mvestnert.Why should anyone care what portion of a manager's returns come from stock selection? We believe there are
three reasons:
1) Factor exposures are no longer worth active management fees.
This is relatively new phenomenon. Twenty yeats ago, all but the most sophisticated investors needed active
managers to gain efficient and consistent exposure to various investment styles like value, quality, and momentum.
“The emergence of smart bea strategies has changed that. While smart beta funds usually don’ offer anything new
to the investment world, what they do bring tothe table i targeted exposure to retura sourees, and they ean doi
cheaply. Because thei strategies are rules-based, smart beta managers don’t have to employ researchers or analysts
and can pass those savings on to investors in the form of low management fcs.
As smart beta adoption expands, exposure to factor returns becomes a commodity, » fortuitous development for
{fee-conscious investors and allocators. We believe active management's role needs to shift to providing returns that
cannot be replicated by smart beta factor exposures. In a meaningful way, exposure to factor returns is no longer
an “active” exposure, and the market for active management is adjusting to this new paradigm, This is illustrated in
Chart 2, which shows how active manager foes decline as more of the active variance is explained by factors?
CHART 2. ACTUAL FEES AND PREDICTED FEES VERSUS THE FRACTION OF MANAGERS’ ACTIVE VARIANCE (R'),
ANUARY 2010-DECEMBER 2014
10 aa A Fee
“ Managers in this
range are at risk
7
oe
Management Fee
22
° 02 04 06 os
Rio a Mutitctor Model
aman oray 10h Tmt aetna reyesvanson sa mar esac remareal
2) Managers are better at stock selection than they are at factor timing.
For most investors this statement is intuitive. Its well known that experts’ predictions of maero factors such as
the price of oil or interest rates are rarely correct. The most compelling empirical proof that stock selection is 8
more reliable source of returns than factor betting is the active share literature pioneered by K.J. Martijn Cremers
and Anti Petaisto’ Sorting managers on relative active share and tracking error allowed the researchers to approx-
imate skill in stock seletion (an activity associated with
higher active share) versus skill in factor betting (an activity
associated with higher tracking error). The group of managers with higher active share and lower tracking error,
which the researchers called diversified stock pickers, outperformed the other manager groups by a significant
margin, Cremers and Petajisto noted that “most active stock pickers have enough skill to outperform theiz bench-
marks even after fees and transaction costs, In contrast, fands focusing on factor bets seem to have zero to negative
skill, which leads to particularly bad performance after fees.”
Paradoxically, the active share literature has been used to justify 2 “more is better” approach with regard to activeness,
when the researcher's findings were actually more nuanced, finding that a certain kind of activity (stock selection)
‘was beneficial while another kind (factor betting) was not. Cremers and Petajisto are not the only researchers to confirm
his finding. (We will offer a comprehensive review of research supporting the superiority of diversified stack picking
as a source of alpha in our upcoming white paper, "The Uncommon Virties of Security Selection.”)3) Stock selection is more diversifiable than factor exposures.
This is true a a portfolio level, and when combining portfolios in a multi-manager structure. Ac the portfolio level,
stock selection can offer thousand: of independent chances to apply an investment team’s skill, compared to just &
few oppormnities fora factor-timing manager. The point of factor identification is o label common determinants of
large postions of security returns. So factor bets, by their nature, are powerful bets that ae hard to diversify. Again,
empirical evidence supports the notion that lower tracking-error strategies typified by diversified stock pickers
generate more efficient excess rerarns than higher tracking ervor approaches typified by large factor bets. (We've
replicated and updated a comprehensive study on this subject in our forthcoming white pape)
[Ar the muli-manager level, common factor loadings do not diversify, while idiosyncratic risk does. So, combining
ortfolios with large weights on common factor risks will not yield much in the way of diversification benefits
‘Combining portfolios with different factor risks ean be more successful, as distinct factor correlations are often very
low over multi-decade return horizons. However, over shorter horizons, such as the one-year rolling time periods
shown in Chart 3, factor correlations can be erratie# Excess returns to factor exposures are similarly erratic, meaning
allocators who rely on manager factor exposures for outperformance must be prepared to endure extended and
sometimes simultaneous periods of underperformance
‘CHART 3, VALUE AND MOMENTUM RISK PREMIA: ROLLING ONE-YEAR CORRELATION BASED ON MONTHLY DATA,
DECEMBER 1927-MARCH 2016
10
Correlation
10
197 1948 1963 1981 2000 2016
oy den nt repent porte managed by ard AA
THE IMPLICATIONS FOR ALLOCATORS
Without a multi-decade evaluation horizon, an allocator may well abandon higher-risk managers during times of
sunderperformance. Indeed, this occurs frequently. In a 2008 study of plan sponsors’ hiring and fring decisions
lover 20 years, researchers Amit Goyal and Sunil Wabal found that in the years after the decision, recently fired
managers produced relatively higher excess returns compared to managers who were hired.’
“Those results underscore the uncertainty associated with higher eracking-error strategies, limiting the usefulness of
these approaches co an allocator. Managers with moderate ative risk—typitied by well-diversifed stock pickers—
can offer more consistent returns, helping to mitigate this uncertainty and, consequently, to minimize suboptimal
hiring and firing decisions.
The idiosyncratic stock selection performed by diversified stock pickers constitutes true activeness in a world
cegy focusing on
diversifiable, repeatable, stock-specific risk may offer a mulki-manager platform more consistent and diversifiable
reruns over the long term,
where factor exposures are commoditized. The research we've examined here suggests that a s" Meio In, ns 2d Lor Abel.
" anal. Katn and Mich Lem, "Te Ase Mares amma Haw art Betas Disrupt the Ivestent Manage nus” acl atv
(anceryfebuay 206,
> KJ. Marin Cras and Ant Pts, Hom Acie Fund Manage’? ANow Meas Ta Press Prtemane The Sei fr hac Stes. 708,
“emt Fee Data ira mak dartnuth a
* ei aya an Suni Wel” The Seton and Ternabon ef inestme: Marae Fis by ln Spor ha of ance gt)
Apa the eur on ar evesmet that sin ces te expected eu, gh ivsient ee rik ra porta seve. hs exes tu ated to
the poo mangers This imps that requis sume fr of active ranean suchas seeing eure lection alga eing amare sate rae
lass coc ing aba
‘Aetve shares a meas of ave pti managment wich resets the sare of par hngs that ferro be breaking, snd tb
de nermation abut au ptt regis echt nde, Active sare an bese tee wh reklama came
pike facie neering sens a tigi tees tuck stn ad at ig
Betas the roth rtf ean tha can atid tthe mt a whl, as oppose teu tha ae trust ports marae’ sil ltl
‘Smart eta fe aleratvenanes at suck oper vata cap we gh nds by waging th edge tar equal ub dames, sch as
sas rite,
Clans taste meas hat fates ow css erties, seus nes, parton] moe in reat ta each ot, craton of indetes
trata moe pr downy anes mated ya oven te same retin thee. Acarelaton of -1.Uindizates tata moe wpe wm wl e mated yma nthe
‘pps deco, cater ofOinats thal eo eis even tattal tons.
Excess retunstos ow profs etomane congas ith bat fis echmakoera stale pt fi
“Tracking errs anda deviation the eres eto and return and its echarcine re,is ameasue the way oa fund eu ines
af ts teh.
IMPORTART INFORMATION
sks to Consider The vue of iesensin qu secures wl cat espns ogee esreiccondtons an to changes inthe aspect of artar
‘omaras arr seri the cna, Midas pcg tc tr tbe mae wll and mayb es ld han age cp cary cs. Mid an nal
‘ap compares tpl xan ihr kf athe age cep companies. Smal ap omparis ray aso fave are iit pec nes, markets, aca
resneesand pal egerence higher ik of failure tas large compares, hie guw sacks rset tthe daly vs ad dows fh stk mare, er ong
tern poten aswel asthe voltliy cen be subst. Yleinesting nest isk a the ate mayo recy hat seus rude, and thy
ayo apts as ance. No ivesing sale cn ovecome al mare ally gate ature ress
ihr Sersteatin ar ase alsatn can quran ptr ptt aga sn acting warts.
Performance quoted represents pest prfrmance Pas performances nota relable indicator or a guarantee of fate ests, Ty Ptashne
chara frist purposes ny and teen ay Lat Ae mul under ary ariel investment,
apis psn ts mall ae a he lof ulation asco canes subeue deve’, and myeloma we.
ate st ine Be eed yp 25 rcs ese, rms aie, a3 acmmantn fet uy os ny seus tai ay instant
stag as tite tt or dept te perma of ay masien Paes shuld a sure tat ines incampais, seis, sects andr mats
estes ware aril be fale Inesing aes ik. iclig pst as of paca. Ts uments prepare hse einai rd Al deo le;
‘haw La Aides tua rary comple fara. vests hal canal tha ani aise pro kina vest en,
Ce 205 Le Ae a,c Da EA sed. Formate information visits at rcabbett.com