discussion

Pick one of them.

Contributions Carosa

Passive Investing: The Emperor Exposed?

by Christopher Carosa, CTFA

Christopher Carosa, CTFA, a member of the adjunct

faculty ofMedaille College in Buffalo, New York, has

been involved in the investment industry for more than

20 years. He is currently engaged in researching the

impact of behavioral economics on investment theory.

Mr. Carosa also serves as president of Carosa. Stanton

& DePaolo Asset Management LLC, and president and

chairman of the Bullfinch Fund Inc. a family of no-load

mutual funds.

F
or some time, the academic world
and the financial services industry
appear to have been converging on

the virtues of passive investing versus
active management. John C. Bogle, a highly
regarded spokesman of indexing, conci.sely
summarizes the current standard belief
\\ hen he states “that the market portfolio is
the mo.st sensible decision. It takes the need
for JLidgment out ot \

But u hat it Bogle’s presumption is
false? .’\ftcr all, most investors don’t in\’est
to avoid niakiny decisions. Most investors
don’t invest purely to reduce costs. Most
investors don’t invest merely to maximize
tax-efficiency. Taken alone, these factors
can produce irratmnal results in one’s
everyday life. Kor example, one can always
reduce costs by l)uying the least expensive
product. But what of reliabilitv? l.s it worth

Executive Summary

• Traditional studies of the passive

versus active management debate

appear to contain two flaws that can

dramatically affect results.

• The snapshot-in-time anomaly creates

period dependency, leading to incon-

sis

tent results.

• The equal-weighted anomaly produces

results that while statistically accurate,

fail to accurately reflect the results

experienced by actual investors.

• An analysis using rolling 12-month

returns appears to reduce, if not

eliminate, the snapshot-in-time phe-

nomenon, leading to more consis-

tent results.

• An analysis using asset-weighted per-

formance data to more accurately

it to buy a ‘IA’ for half price that needs to
be replaced three times over the life of the
alternative? More importantly, would yo

u

buy the cheapest |>ar:ichutc even though it
successfully opens oiil_\’ nine out ot tun
times? The same could be said of tax-effi-
ciency. The best way to avoid pa\’inu ta.xcs
is to not earn any money. Using this logic,
one uoiild never work, (jreat. You might
never pay taxes, but you’d never be able to
buy a TV—even the cheapest one!

So, choosing investments—just like
making any other purchase decision—
cannot focus on these “common sense”
rules. What, then, represents the critical
factor? Quite simply, it is the ability^)r,
more importantly, the likelihood—of one

reflect the actual behavioral patterns

of investors appears to produce more

significant results.

An analysis of investment return data

from January 1975 through June 2004

shows active investors in U.S. equity

funds performed better than the 5&

P

500 two-thirds of the time and by an

average of 2 percent annually.

Using both modern portfolio theory

and behavioral finance measure-

ments, the investors in active funds

appear to have taken less actual risk

than the index.

These results have broad implications,

not only for financial planners, but for

public policy issues such as ERiSA and

Social Security reform.

achie\’ing one’s lifetime goals, or “goal-

History

Along these lines. Bogle’s fundamental
assumption embraces the iiiea that index
mvesting provides returns that either equal
or exceed active investing. Indeed, this
assumption is well founded by both histori-
cal data and academic research. What

Journal of Financial Planning/October 2005 www.j ournaifp.net

Carosa Contributions

iilmost js ;i snuill (l)ut vcr\’ contni-
\crsial) afterthought in 197.^ when Charles
Illlis in;itter-ot-factly concluded that the
commonly held tenet that pn)tes,sio[i;il
si-ciirity analysts can manage portfolios that
consistently outperform the market
“\i|)pears to he false,” hlossoined into an
outright creed wlied, in 1’>S.?, William S.
Cjray III wrote “the median experience of
acti\’L’K [nanagetl eijiiity portfolios has
been well helow (I to 2 percent) the S&P
ÎKI in most years during the I’̂ T̂Os and

l9S()s.”” Most sit;nif1cantK’, tliis latter refer-
ence appears in a hook that is rc(|Liired
reading t(H” candidates for the (Chartered
I inancial .\rialyst ((^I-.\) designation—the
\er\” people w ho try to make a Ii\ iny l)y
[licking stocks!

Perhaps the defining work was piih-
lished h\ IJnrton .\lalkiel in 1995. Malkiel
summarizes the current generally accepted
academic principle w hen he concludes,
‘•.Most in\’esrors would he considtrahlv
lietter oft hy |>iirchasing a low expense
index fund than l)\- tr\ ing to select an
active fund manager who appears to pos-
sess a ‘hot hand.’ Since active management
generall}’ fails to provide excess returns and
tends to generate greater tax hurdens for
investors, the ad\ antage of passive manage-
ment holds.'”

Hut IJogle himseit pro\ idei.1 insight as to
the hiisie flaw in these types ot studies
u hen he pointed out that “each and every
comparison we see is period-dependent.”
Ave. there’s the ruh!

Basic Hypothesis and Source
of Data

The idea, then, is to test whether the
market index beats actively managed port-
tohos with enough consistency to justify
index investing u.sing a mcthodolog\’ that
incorporates actual investment decisions as
well as a practical financial planning tech-
niqtie into the traditional academic analy-
sis. To accomplish this, we’ll use the S&P

500 performance (as provided Uv Barra) as
our market proxy. While this does not rep-
resent the total market, since 1933 “the !2.

2

percent annual return of the SikP 5iK) has
l)een exactly rhc same as the return of the
total st))ck market.”” The S&P 500 holds
the further advantage of heing perhaps the
most popular indexing choice for investors.

On the other hand, sophisticated aca-
demics can riglufully express concern
al)out the potential for an apples-to-oranges
comparison hetueen the two data sets. To
address this concern, iliscussions were held
w ith other researchers u ho indicated there
is no statistical difference hetween the
Barra SiS:P 511(1 returns and the (Center for
Research in Security Prices ((JRSP”) total
stock database tor the |>eriod in (]uestion
(January 1975 through June 20(14). There-
fore, the results ofthis study are likely to
he identical whether using S^iP 500
returns or total market returns. (CRSP is a
proprietary, subscription-only database
generally available just to universit}’
researehers.)

U,sing monthly return data, we’ll meas-
ure the rolling 12-month return results for
the index against similar performance data
for actively managed U.S. equity mutual
funds. V\e chose rolling 12-nionth periods
rather than calendar years lieeause, as all
tlnancial advisors know, real investors—
especially those «”ho regularly contribute
to company retirement plans—do not
limit their investment decision-making to
December 3 1 of every year. Bv examining
rolling 12-nionth periods, we can more
accurately assess the near-term relative
performance between the index and the
average mutual fund. This addresses
Bogle’s concern by redueing—it not out-
right eliminating—any period-dependent
t)ias in the study.

Data regarding actively managed U.S.
C(]uity mutual funds was provided by
Lipper Analytics, the nation’s oldest and
largest provider of mutual fund perform-
ance data. Since Barra S&P 500 data w ere
available only from January 1975, we chose

that as our first month. .MonthU’ returns
were available through June 2004, To best
measure the aetively managed data. Upper
provided asset-weighted return data in
addition to the more commonly used
e(]ual-weighted return data. (The former
a\”oids any skev\ ing toward smaller funds
and hctrcr represents the inxestment deci-
sions (if actual in\estors.) The Lipjierdata
include funds that ha\’e been closed or
merged, and therefore contain no sur\ i\or
bias, meaning both acti\e funds anil mac-
tive funds were included in this study. Nei-
ther the Lipper tlata nor the Barra data
include loads, commissions, or investment
management fees. The Upper data reflect
in\estment pertormanee net of expenses
and I2h-I tees, while the Harra data, heint;
index data, do not. Both funds include total
return performance—that is, both capital
appreciation anti rein\’estcd dividends.
I’inally, the Upper data include index
funds, so the Upper results will skew to
some deirree {not meastirahle due to the
manner in w hieh the data w as provided)
toward Barra results. This means the actual
\’ariance hetween actively managed funds
and the index is prohal)ly greater (albeit at
some unknown significance) than the ntim-
l)ers reported here.

Note for those interested in duplicating
tliis research: The Barra data are publicly
a\’ailable on their Weh site (www.Harra.
com). I he I.ipper data are available only
l)y subscription, but Upper has at times
(this case is one example) offered
researchers a tjrant to obtain a limited
amount of data. For the purposes ofthis
study, it w as requested that Lipper provide
monthly return data for its “US Diversified
i-“.t|uity (Jroup” (USDE(iroup) of mutual
funds from December .’I, 1974 (this
enahles you to calculate the return for the
month of January 1975), through June 30,
2004. Lhe USDL Group represents data
points aggregated from 15 Lipper fund
class itlcations:

1. Large-Cap (Lil) Cirowth
2. LCCore

www.j oiirnalfp.net Journal of Financral Planning/October 2005

Contributions Carosa

3. LC Value
4. Mid-Cap (MC) Growth
5. MC (.\>re
6. MC Value
7. Multi-Cap (MLC) Growth
H. MLC Core
9. MLC Value

lO.Sniall-Cap(SC) Growth
l l . S C C o r c
12.SCValuc
l3.Iit]Uity Income
i4.S&P 500 Index Objective
15,SpeeiaIty Diversified Kquicy Funds

Becau.se only a ^ e g a t e data were made
available, it was impossible to perform fur-
ther anatyties on the results (this might
best be eondueted in future researeh). Also,
the Lipper terminology for the two formats
is “Average” (Lipper’s term for “equal
weighted”) and “Dollar Weighted Average”
(Li|)per\s term for “asset weighted”). Sinee
Lipper’s databases are usually survivor-
biased, it is critically important that the
retjuest specifically asks that all active and
inactive funds be included.

The Results

The period from January 1975 through
June 2004 represents 29 1/2 years. During
this period, the equal-ueighted average
annual return of all U.S. equity mutual
funds—the traditional measurement tech-
nique that gives smaller funds the same
weight as larger funds—was ] 3.93 per-
cent. Returns for the S&P 500, on the
other hand, were slightly smaller: 13.73
percent. Immediately, the casual reader
might hastily conclude that all previous
studies purporting to show the dominance
of passive investing to be in error. On the
other hand, this result should not surprise
those familiar with recent history in that,
on average, actively managed equit)’ funds
have beaten the S&P 500 during the last
five years. This phenomenon is just an
example of Bogle’s period-dependency.
For example, in the 25-year period fro

m

January 1975 through December 1999, the
equal-weighted average return of all U.S.
e(|uity mutual funds was 16.99 percent,
lagging the index return of 17.26 pereent.
So, while a longitudinal study ending in
1999 favors passive investing, a similar
study ending in June of 2004 favors active
investing.

But, as the astute planner recognizes,
reliance on equal-weighted data betra} s the
actual investment decisions made by and
on behalf of clients. Think of it this way:
Let’s say we did an analysis on ten mutual
funds. If only one mutual fund made
money, the equal-weighted average return
would probably be negative. On the other
hand, let’s say all investors invested in the
one fund that made money, and not in any
of the other nine funds. In this case, the
asset-weighted average return would prob-
ably produce positive results. What does
this lead us to conclude? In this very hypo-
thetical case, the asset-weighted average
return suggests investors (as measured by
the amount of money they invest) tend to
accurately recognize and reward the better-
performing fund by Investing in that fund.
The equal-weighted average return sug-
gests the mutual fund families were willing
ro create many different kinds of equity
funds, some of which will ultimately per-
form poorly and fail to attract investors.

Equal-weighted returns, though meas-
uring investment decisions of mutual fund
portfolio managers, emphasize the business
decisions of mutual fund corporate man-
agement. Asset-weighted returns, on the
other hand, while also measuring the
investment decisions of mutual fund port-
folio managers, emphasize the investment
decisions of shareholders and their profes-
sional advisors. Asset-weighted returns,
therefore, tend to reflect the practical real-
ity of the financial planning environment.
As a result, an analysis of asset-weighted
average returns, conducted in the same rig-
orous manner of past aeademic studies,
may be more meaningful to practitioners in
the finaneiai services industry.

During the period from January 1975
through June 2{X}4, the geometric asset-
weighted average annual return of all U.S.
equit)’ funds was 15.85 percent—nearl_\’ 2
pereent greater than the geometric average
annual return of the S&P 500 and the
equal-weighted return of all U.S. equity
mutual funds for the same period. Surpris-
ingly, contrary to our expectations, this
was not a period-dependent result. During
the period from January 1975 through
Deeember 1999, the geometric asset-
weighted annual return of all U.S. equity
funds was 19.11 percent—again nearly 2
percent better than the ec]uiva!ent

S&P 500

return and greater still than the et|ual-
weighted return of the same fund data for
the same period.

This suggests the actual decisions made
by investors and their advisors—as meas-
ured by the total dollars they invested—
tended tu produce better returns than
either investing in the index or investing
equal amounts in all U.S. equity mutual
funds. This inference, as further analysis
will soon reveal, contradicts the long-stand-
ing belief that active deeision-making adds
too little value compared with the amount
of risk it introduces. As policy-makers
deeide how to privatize Soeial Security,
this conclusion means any rules that
remove or reduce active decision-making
on the part of investors may actually
impede-—not protect—^those investors.

Roiling 12-Month Returns

Of course, legislative concerns aside, the
practical problem of the above return
analysis is that it assumes clients invested
all their funds In January 1975. Would that
all finaneiai planners had time machines!

Planners and their clients tend to invest
periodically throughout the entire year. T o
better acknowledge reality, let’s take a look
at the investment returns of every 12-
month period from January 1975 through
June 2004. There are 342 12-month periods

Journal of Financial Planning/October 2005 www.journalfp.net

ConlrihLilioiis Corosa

T A B L E 1

Percentage of 12-Month Periods
Where the Asset-Weighted
Return of U.S. Equity Mutual
Funds Outperformed the
S&P 500 in the Similar Period
Index ^ No Fee [ 64.33%

Expense 20 Basis Points ^ 66.37%
^ ^^^ –

Ratio 150 Basis Points I 68.13%

during that 29 !/2-year span yielding 342
different returns. For U.S. et]uit\’ funds,
the arithmetic average of those 342 asset-
weightetl returns was 1 7.84 percent. (We
must use the arithmetic average instead of
the geometric average heeause these return
[leriods overlap.) Likewise, the arithmetic
average of the S&P 500 in those same 342
periods uas 15.55 percent. Of interest, the
standard deviation of these two samples
was nearly identical—1 7.46 percent for the
as.set-\veighted U.S. cquit\’ mutual fiuid
returns and 17.54 percent for the S&:P 500.

For advocates of modem portfolio
theory, these three data points—lM)th the
arithmetic anti the geometric average
return a,s well as the staudari! deviation—
iuiply that, w liile the universe of U.S.
c(|iiity mutual funds exhibits the same risk
(that is, standard (.ie\iation) us the S \ P
500, the uni\erse of U.S. eiiiiity mutual
funds appears tu possess a sindificant
retiim |>remium. This further implies that
the more efficient jiorttolio is not the Si^P
500, but the asset-v\ eiglued collection of all
U.S. ei.|uity mutual tuiuls.

For those u ho prefer using the upside/
dww nsiJe analysis more eommon In behav-
ioral ectmomics, the sanu- ileductioii can be
draw n. I he worst return for any given
period is lower for the Sc^P 500 (-30.49
percent) versus the a.sset-weiglited average
of U.S. equity mutual funds (-27..^3 per-
eent). Similarly, the best return for any
given period is higher for the U.S. ei|uity
funds (-1-71.30 percent) versus the SivP 500
(-1-61.65 percent). These facts suggest the

equit)’ funds, in aggregate, have a lower
dounside aiui a higher upside for 12-month
periods. Again, in using asset-weigh ted
returns, the conclusion applies to the real-
life investment decisions people make, not
in the hypothetical instance that every
person invests the identical amount of
mone_\- in every U.S. equity mutual fund.

But wait, there’s more! We know the
mutual fund return data automatieally
account for the expense ratio of those
funds. I he S&P 500 data, being actual
index returns, fail to account for the actual
expenses incurred by index funds. ‘Fhe
inclusion of this expense ratio only further
extends the outperforniance of U.S. etjuity
funds relative to index returns. We relate
this in Table I, where we indicate the per-
centage of periods where the U.S. equity
funds beat the index returns in three differ-
ent scenarios:

1. I he raw index returu, which iiieludes
no expenses

2. The scenario where index fund returns
mimic the uidex returns but include a
20-basis-point expense ratio

3. The scenario w here index fund returns
mimie the index returns but include a
50-basis-point expense ratio

Table 1 indicates, for any given 12-
montb holding period, the asset-weighted
return of all U.S. equity mutual fluids
beats the S&P 500 roughly two-thirds of
the time. The table also suggests that, in a
manner consistent w ith the conclusions of
iMalkiel, investors shouki be mindfiil w hen
buying shares of iniiex funds and trv to
buy those with the lowest expense ratio.
(I he data provided by l.i[iper were not
sufficientk- detailctl for us to make any
conclusions regarding expense ratios of
U.S. equity funds in general.)

Given the above data, we do not mean
to suggest that U.S. equity fund returns
are not correlated w ith S&P 500 returns.
Indeed, a statistical analysis show s a posi-
ti\ e correlation coefficient of 0.H7. There
appears to be, however, a tendencv for
the U.S. eijuit}’ funds to either lead or lag

in discrete time periods. l*Or example.
U.S. equity funds tended to lead in the
late 1970s and early 1980s; again in the
late I9H0S and early 1990s; and, finally, iu
the most reeent five-year span. Cin the
other hand, the S&P 500 tended to consis-
tently lead for several short intervals in
the mid-l9H0s and again in the mid-late
1990s. We attempted to correlate these
periods with the general movement of the
market, but the analysis showed only a
slight correlation (0.28). A cursory visual
review* of the market during the nearly
30-year duration (see Figure ! on page 60)
hints there is a tendency for the S&P 500
to lead in late-stage bull markets and a
tendency for U.S. equit\’ mutual funds to
lead in bear markets.

The best we can eouclude is that some-
times it’s better to be in aetivel\- managed
funds and .sometimes it’s better to l)e in
passively managed S&P 500 index funds.
Unlike previous researeh. we clearly canno

i

conclude—or even imply—that the evi-
dence show s index lnvestint; offers consis-
tently better investment returns compared
w irb returns offered by the entire asset-
weighted U.S. equity mutual fund uni-
verse.

FinalK- eomes tbe question all finaneiai
planners need to know : What does this
study uncover about an in\’estor’s abi!it\’ to
meet a specific goal-oriented target
(GOT)? (The C;OT represents that spe-
cific rate ol return rei|uired for an in\esror
to attain a particular financial goal.) How-
do these series of 12-month returns stack
up against a variety of (;O Fsr Fable 2
(page 60) demonstrates how both the S&P
500 and rhe equity mutual funds eompare
with a \ariety of common (iO’Fs.

As ‘Fable 2 clearly indieates, the asset-
weighted average 12-month return for U.S.
equity funds consistently betters the S&P
500 returns for similar periods. This differ-
ence becomes more exaggerated as one’s
G O T increases. For example, when the
G O I IS merely to not lose any principal
(that is, 0 percent). U.S. equit_\’ mutual

Journal of Financial Planning/October 2005 www.jou rna If p. n et

Contributions Carosa

F I G U R E 1

Rolling 12-Month Returns for S&P 500
Areas Shaded in Blue Show Periods When S&P 500 Returns Lagged U.S. Equity Fund Returns

s
s
>

•

1

s
m

•

s
:i

S

m
•

‘J

s
m
•

3
•
•

m
m

( ^

s
• •

•
s
m
•

hi

m
m

5
ii
S

m
•
S
•
•
•

•
—
:i
m
m

Z
•

•

:i
•

2
Z
m

i

M

s
• •
•
•
•

5
:i
m
m

•

u
‘J
a
•

•

s

«

•

2
m
•

(^
•
p

•
•

I
m

•
m
m
«

kj

•
•
•
u
P
•
P
•

T A B L E 2

Percentage of 12-Month Periods
When Investment Returns Met
or Exceeded the Indicated

GOT

GOT

0%

4 %

6%

8%

10%

12%

S&P 500

80.12%

74.56%

71.05%

68.42%

65.50%

62^7%
14.875% 53.22%

16.5% 1 49.42%

U.S. Equity
Mutual Funds

83.63%

78.6S%

76.02%

72^1%
69.88%

66.96%

62.28%

57.60%

funds art’ just 4 percent more likely to
achieve that G O T compared with the
S&P 500. On the other hand, if the G O T
is to double the investment every five
years (that is, I4.S75 percent annually),
U.S. equity mutual funds are 17 percent
more likely to achieve that ( J O T com-
pared with the S&P 500. In the end, it is
more important for most investors to
attain a certain goal-oriented target than
to simply “beat the S&P 500.” This par-
ticular portion of the analysis reflects
how people really invest, not an arbitrary
target based on a market index. As a con-

sequence, the results reveal that investors
have been better served with equity
funds across a l)road range of CJOTs
based on how they and their financial
advisors implemented their actual invest-
ment decisions.

How statistically significant are all these
results? A simple paired two-sample t-test
for means yields a t-stat of-2.62 and a P-
value of 0.91 percent. (The eritical values at
the 5 percent significance level are
±1.966705.) This shows we can reject the
null hypothesis (that is, the hypothesis that
claims there is no statistically significant
differenee between the two series of
returns) at most standard significance levels
(ineiuding as low as 1 percent). This cer-
tainty level is large enough to suggest the
possibility of a statistically significant dif-
ference between the U.S. equity mutual
fund returns and the S&P 500 returns.

Conclusion

It has long been believed that actively man-
aged portfolios underperform tbe market.
This study indicates rhe more aceurate
conclusion would be that there exist
extended periods \\ hen the market outper-

forms aetivel}’ managed U.S. e(|uity funds
and extended periods when the aetively
managed U.S. equity funds outperform the
market. As a result, our re-examination of
the passive-versus-active investing debate
appears to confirm what others have
hypothesized—that previous studies pur-
porting to show the dominance of passive
investing may have reflected the snapshot-
in-time anomaly commonly found in meas-
uring investment performance. The study
can reach no statistically significant eonelu-
sion regarding the potential correlation
between these periods of overperformance
and underperformance with bull and bear
markets. We can anecdotally conclude
there appears to be a tendency for tbe S&P
500 to beat U.S. equity funds at the top of
bull markets and a tendeney for U.S.
equity funds to beat the S&P 500 during
bear markets.

Furthermore, and more significantl}’,
this study concludes, w ithin most standard
significance levels (including as low as I
percent), that for investment returns in
rolling 12-montb periods from January
1975 through June 2004, uhen looking at
actual investing patterns and not merely a
hypothetical equal weighting of all mutual
funds, U.S. equity funds have historically

Journal of Financial Planning/October 2005 www.journalfp.net

Contributions Carosa

Suggested Web Sites

www.Barra.com

www.LipperLeaders.com

beat the S&P >()<) roughly tuo-thirds of the time. In addition, U.S. e(jiiity funJs, on avLTiigf, Lire more Hkely to meet or exceed ;in iin'esror's retiini t;irgct acrns.s ;i hroad ninge of CIO I s, and hence are more hkeiv tn earn ;i sufficicnr rate of return to fin;ince an investor's ideal litest\le. "l"hese results directly contradict the prevailing view that investors and their financial advisors are better off merely in\ esting in the market and not taking the time to researcii mutual fund investment options. I here are both practical and public policy implications of this t.li,sc<)\ery.

I’lrst. the financial ser\ ices i[idustr\’ and
mutual fund shareholders appear to have
fxhibiteil a eonsistent track record of
investing in U.S. equity funds that are
niore likely to outperform the S&P .̂ 00.
Recognition ofthis broad effort—and the
fact that it has added value for investors—
has not received much cnvcraye in aea-
deniic studies. Now it looks as if u e have
empirical evidence that implies the results
of aetne decision-making have reaped
rewards for investors. The study docs not
intend to conclude that pnor investment
decisions are not made. Also, it is be\ ond
the scope ofthis study to suygcst which
types of U.S. et[uity funds mav or may not
inerease the likelihood of the investor beat-
ing the S&P 500. It is the hope ofthis
author that other studies, using the behav-
ioral analytics described here, might fur-
ther break down performance w ithin spe-
cific fund categories.

Second, and perhaps ot more long-term
conset]ucncc. are the public policy ramifi-
cations of this study. 1 lere we refer U>
both the current regulatory compliance
practices of KRISA plan trustees and fidu-

eiaries. as w eil as tbe burgeoning debate
on the privatization of Social Seeuritv.

Regarding the lormer, the eonckisions
ofthis study, in contradicting previous
studies, might cause regulatory bodies like
the Department of Labor to reconsider
how they define “generally accepted indus-
try- practices.” For example, in question-
ing—if not outright rejecting—the statisti-
cal dominance nt passive investing, plan
trustees and other fiduciaries might uant to
avoid relying solely on passive vehicles.

Regarding the latter, the [>ri\ati/.ation of
Social Security, this study has profound
repercussions. Witii the idea to give work-
ers ownership of some p(jrtion of their
Social Security savings, the goxernment
finds itself in the same position as that of a
401(k) plan trustee. This stud)’ reveals,
l)oth in terms of the standard-deviation risk
analysis of modern |iortfolio theory and the
upside/dounside analysis of beha\-ioral eco-
nomics, U.S. equity funds, as a whole, offer
better returns with either the same or less
risk than investing in the market. Again,
the study does not intend to say each and
every U.S. equity fund has these character-
istics. Rather, the study concludes the
aggregate investment decision-making of
investors and their financial advisors has
resulted in better returns than passively
investing in the S&P .̂ 00. This holds true
both over time and in nearly two-thirds of
tbe .’42 12-month periods from January
1975 through June 2004. It would be diffi-
cult for lawmakers to justify taking any
action that would, contrary to this studv,
place workers’ retirement assets in harm’s
way by unnecessarily restricting in\est-
ment choices based on a prevailing \\-isdom
that could find itself turned on iis head in
the near future.

Passive investing may not yet be naked,
but it certainl) has fewer clothes than
thought.

Endnotes

1. John C. Hogle, “The Telltale f:hart,”
keynote speech before tbe Morningstar
Investment Forum. Chicago, II., June
26, 2002; available online from the
Hogle Financial Markets Research
(Center.

2. (Jharles D. FUis, “The Loser’s (Same,”
l-innncial AnnlysTs Journal, Ju\y/August
]975: 19-26.

3. William S. CJray III, 19K3, “Portfolio

Construction: Fipiity,” Munaying

Investment Portfolios, John L. Maginn
and Donald L. luttlc, editors, 19^3:
402.

4. Burton (i. Malkiel, “Retui-ns from
liu’esting in Fquit\- Mutual Funds, 1971
to \99]” Jinirnal of Fin-jncc, 50 (1995):

549-572.

5. Bogle, ibid.
6. Bogle, ibid.
7. These two expense ratios appear to best

represent the range of actual expense
ratios based on a review of publicly
available index funds.

REPRINTS
EPRINTS
PLAQUES
POSTERS

877- 652-5295 and allow
our reprint coordinator to assist

you with some proven
marketing ideas.

Journal of Financial Planning/October 2005 www.journalfp.net

January/February 2016 www.cfapubs.org 9

Viewpoint is an occasional feature of the Financial Analysts Journal. This piece was not subjected to the peer-review process.
It reflects the views of the author and does not represent the official views of the FAJ or CFA Institute.

V I E W P O I N T

John C. Bogle
Founder and Former Chief Executive, The Vanguard Group
President, Bogle Financial Markets Research Center

The Index Mutual Fund: 40
Years of Growth, Change, and
Challenge

Some 40 years ago, a new concept of investing—
one broadly regarded by the financial community
as trivial and likely to fail—began its path toward
implementation. On 18 September 1975, at the board
meeting of the one-year-old Vanguard Group, the
directors approved my proposal that our firm form
the world’s first index mutual fund. Proud of our
new creation, we named it First Index Investment
Trust—now the Vanguard 500 Index Fund.

Ultimately, index funds would earn remark-
able acceptance from individual and institutional
investors alike. But all that glitters isn’t (necessar-
ily) gold. Almost a decade passed before the second
index fund was formed, and it took two full decades
(until the mid-1990s) for index funds to begin to earn
broad acceptance. Much of that acceptance, however,
reflected a major departure from the original “buy-
and-hold” concept.

The Triumph of Indexing
Despite a faltering start for that first index fund,
mutual fund investors themselves have come to
accept the value proposition offered by indexing.
Assets of index funds have risen dramatically over
the years—from $11 million in 1975 to $511 million
in 1985 to $55 billion in 1995, leaping to $868 billion
in 2005 and now standing at $4 trillion. From 4% of
equity mutual fund assets in 1995, the market share
of index fund assets grew fourfold to 16% in 2005
and then more than doubled to a record high of 34%
in 2015 (see Figure 1).

An even more dramatic story is told by the cash
flow into (and out of) equity mutual funds. Since 2008,
index funds have accounted for 160% of net cash flows into
equity mutual funds. During those seven years, inves-
tors have made net purchases of almost $1 trillion in

passively managed index funds even as they have
liquidated, on balance, $600 billion of their holdings
in actively managed equity funds—a remarkable $1.6
trillion swing in investor preferences. Such a dramatic
turnabout in investment strategy is, I believe, without
precedent in the mutual fund industry.

The fundamental principles established by that
first index fund are simple: Buy virtually the entire
US stock market and hold it intact “forever,” elimi-
nate advisory fees, and minimize both operating
costs and portfolio turnover. These simple principles
have won the day. But in 1993, almost two decades
after the creation of the original index fund, a new
form of index fund—originally designed for stock
traders and speculators—came into existence. That
change and that challenge, little noted in financial
history, will be long remembered.

Hired and Fired
Let’s start at the beginning. My inspiration for
the idea of the index fund came in 1951 with my
Princeton University senior thesis, “The Economic
Role of the Investment Company.” In it, I noted that
mutual funds “could make no claim to superiority
over the market averages”—a prescient, if imprecise,
harbinger of what was to come. In the years that
followed, my own investment experience unequivo-
cally reinforced that conclusion.

That Princeton thesis led me to a job in the
mutual fund field. Following my college gradua-
tion in 1951, I was hired by Walter L. Morgan, the
founder of the Wellington Fund, then managing
$145 million of assets. There I observed firsthand
our frustrating—and ultimately fruitless—battle for
significant performance superiority over our “bal-
anced fund” peers. So, early in my career, I learned
a vital lesson: This is one tough business!

In 1965, the conservative Wellington Fund (a
balanced fund holding both stocks and bonds) was
falling totally out of favor with investors and bro-
kers. Mr. Morgan named me to lead his firm and to
“do whatever is necessary” to reverse Wellington’s
sagging fortunes. This was the era of the hot “go-go
fund,” an era that would finally, well, go by 1969.

But in 1966, I put aside my skepticism and
merged our firm with a far smaller investment
advisory firm run by the young managers of the

Editor’s note: This article was reviewed and accepted
by Executive Editor Stephen J. Brown.

Author’s note: The views expressed in this essay are
solely those of the author and do not necessarily reflect
the opinions of Vanguard’s present management.

Financial Analysts Journal

10 www.cfapubs.org © 2016 CFA Institute. All rights reserved.

aggressive Ivest Fund. That fund would flourish
for a few years but then falter and fail, now con-
signed to the dustbin of history. Under the aegis of
our new money managers, the performance of the
Wellington Fund and most of our other funds also
plummeted, and in January 1974, my new partners
fired me—a perverse outcome indeed! My pros-
pects were bleak.

Just eight months later, I attempted to salvage
my career by founding a new firm. (The saga of
its convoluted creation is too long and complex to
recount in this article.) It would be the industry’s
first mutual mutual fund organization, owned and
controlled by its fund shareholders—not, as under
the conventional industry structure, by an external
management company. It would operate on an “at-
cost” basis. On 24 September 1974, right before the
bottom of the 1973–74 bear market, we were incorpo-
rated with the name that I had chosen, “Vanguard.”

Our vision was to become the world’s lowest-
cost provider of mutual funds in an industry in
which “costs matter.” Yet none of our peers sought
to compete on costs. Facilitated by our mutual struc-
ture and our later decision to move to a “no-load”
distribution network and eliminate stockbroker
sales of our funds, achieving that “lowest-cost” goal
would be easy—but only if we survived those dismal
early days. (Our first business plan was simply to
get through the day!) By the early 1980s, our nega-
tive cash flows had turned positive. We had at last
turned the corner.

Friendly Persuasion
“Strategy follows structure” was my central tenet.
Given the minuscule costs of index funds, it quickly
became apparent that our unique structure almost
demanded that we form the first index fund, because
the low-cost index fund was the key to capturing
almost the entire return of the US stock market.
Vanguard’s structural cost advantage over our
competitors virtually necessitated that index funds
become a major thrust of Vanguard’s strategy.

But what would it take to persuade the members
of Vanguard’s board of directors—conservative to
a fault—to approve an unprecedented innovation?
Such a fund belied the deeply embedded notion
that the professional management of Other People’s
Money—the mutual fund industry’s core selling
point—was necessary. (Ignore, if you can, the obvi-
ous reality that for equity fund managers as a group,
the high costs of management make beating the mar-
ket index over the long term virtually impossible.)

Two strong pillars supported moving my con-
cept of indexing from vision to reality. The first
pillar was intellectual support. I found this pillar
of support from none other than eminent MIT pro-
fessor Paul A. Samuelson. Just a year before our
1975 board meeting, this Nobel laureate had laid
out his case in the inaugural issue of the Journal of
Portfolio Management.1 In his paper (“Challenge to
Judgment”), Dr. Samuelson pleaded for some insti-
tution to set up a portfolio that tracked the S&P 500
Index. In my presentation to the board, his paper
was marked “Exhibit A.” His name and reputation
gave credibility to my proposal.

The second pillar was analytical support. Did the
data support my case that passive index funds could
outperform their actively managed counterparts? I
compared the annual returns of the S&P 500 with
those of the average equity mutual fund over the
previous three-plus decades. In my second exhibit,
I reported my findings to the directors. (In those
days, nearly all equity funds were invested largely
in blue-chip stocks, making the comparison with the
large-cap-dominated S&P 500 a fair one.)

The results were impressive. From year-end 1945
through mid-1975, the S&P 500 earned an average
annual return of 11.3%, an advantage of 1.6 percent-
age points (pps) over the return of 9.7% earned by the
average fund—a compounded 30-plus-year advan-
tage of 863 pps. In other words, the cumulative value
of a $1 million initial investment in the S&P 500 and
in the average equity fund would have grown to
$25,020,000 and $16,390,000, respectively—an advan-
tage of $8,630,000 for the index. Persuaded both by

Figure 1. The Growth of the Index Fund

Index Fund Assets ($ millions)

0%
4%

16%
34%

$868bn

$55bn

$511m

Index Fund
Assets

Equity Index
Fund Market

Share

Equity Index Fund Market Share (%)

10,000,000

1,000,000

100,000

10,000

100

1,000

10

40

30

20

10

0
76 1585 95 05

$4.0trn

Sources: Morningstar; Strategic Insight Simfund.

The Index Mutual Fund

January/February 2016 www.cfapubs.org 11

the Samuelson endorsement and by this compelling
evidence, the board unanimously approved my pro-
posal. First Index Investment Trust was organized
on 31 December 1975.

Repeating the Comparison 40 Years
Later
During the summer of 2015, as our Vanguard 500
Index Fund’s 40th anniversary approached, I decided
to calculate what a comparable study would show
for the three-plus decades then ending. Eerily, the
results almost precisely matched those of the study
that I had prepared four decades earlier (see Table 1).
The index advantage over the equity fund was iden-
tical: 1.6 pps a year. The S&P 500 annual return was
11.2% and the annual return of the large-cap funds
was 9.6%, for a cumulative advantage of 946 pps.

The virtual identity of the two returns for the
first and second periods was, of course, little more
than a random statistical freak. But the consistent
spreads in favor of the S&P 500—based largely on
the drag of the “all-in” costs of actively managed
funds—have been verified in virtually all longer-
term periods. Further, the remarkably high correla-
tions (R2) confirm the validity of the comparison.

In the original three-decade period, 96% of the
fund returns were explained by the returns of the
S&P 500 (R2); the corresponding number for the
recent period was 99%. In both periods, risk (stan-
dard deviation of returns) was only slightly higher
for the S&P 500. In both periods, the S&P 500 earned
significantly higher risk-adjusted returns as mea-
sured by the Sharpe ratio: 0.42 versus 0.38 in the
earlier period, and 0.48 versus 0.39 in the later one.

The returns earned on stocks in both periods
were well above the (100-year) historical annual
norm of 9.0% and far above my own reasonable
expectations for stock returns during the next 10
years.2 But the consistency of the index advantage,
the high correlations, the comparable volatility, and

the superior Sharpe ratios over six decades reaffirm
that the sound principles of the broad-market, low-
cost index fund have met the test of time.

Is Turnabout Fair Play?
Forty years ago, predicting the eventual domina-
tion of the mutual fund industry by index funds that
has now come to pass (illustrated by their massive
fund cash flows since 2008, described earlier) would
have seemed like madness. At the outset, the 1976
IPO for First Index Investment Trust was a flop; a
planned underwriting of $150 million (big for those
days) produced only $11 million. Describing the
new fund as “Bogle’s Folly” reflected the Wall Street
consensus. That reaction was captured by a large
poster published by one brokerage firm: a harried
Uncle Sam feverishly stamping “UNAMERICAN”
on index fund stock certificates. “Help Stamp Out
Index Funds,” the poster read. “INDEX FUNDS ARE
UNAMERICAN!”

What a difference 40 years make! Yes, index
funds continue to be the investment of choice for
Main Street investors—whom they were primar-
ily designed to serve—soon joined by a wide array
of pension funds and thrift plans. But index funds
have now become the darlings of Wall Street. To
a remarkable degree, they have taken a different
form: exchange-traded (index) funds (ETFs) that
can be—and frequently are—“traded all day long,
in real time” (an excerpt from an early ETF adver-
tisement), held largely by financial institutions and
mostly used for speculation, hedging, arbitrage, or
other short-term purposes.

Today, the assets of equity ETFs total $1.7
trillion—fully half of the $3.4 trillion of total assets
invested in equity index mutual funds. TIFs account
for the other half. (TIF is an acronym that I created
to identify traditional index funds, such as that origi-
nal broad-market, low-cost, no-load index fund,
designed to be bought and then held “forever.”)

Table 1. Déjà Vu? Three Decades of Index Superiority—Twice

1945–1975 1985–2015

Average Equity Fund S&P 500 Index Average Large-Cap Funda S&P 500 Index
Annualized return 9.7% 11.3% 9.6% 11.2%
Index advantage — 1.6% — 1.6%
Cumulative return 1,539% 2,402% 1,548% 2,494%
Index advantage — 863% — 946%
Standard deviation 16.4% 18.6% 16.8% 17.3%
Sharpe ratio 0.38 0.42 0.39 0.48
R2 0.96 1.00 0.99 1.00

Note: The starting dates are 31 December 1944 and 31 December 1984; the ending dates are 30 June 1975 and 30 June 2015.
aBy the start of this period, equity fund portfolios had become far more diverse. This series represents the average return of the
Lipper large-cap funds category, the most appropriate comparison for the large-cap-dominated S&P 500 Index.

Sources: Wiesenberger Investment Companies; Morningstar.

Financial Analysts Journal

12 www.cfapubs.org © 2016 CFA Institute. All rights reserved.

ETFs: Three Distinctly Different
Approaches
Today, there are only about 12 US broad-market
equity ETFs, similar in portfolio construction to that
original S&P 500 Index fund. Their assets currently
total $337 billion, about one-fifth of total assets of
equity ETFs. The remaining 1,800 ETFs have a far
narrower focus—stocks of single nations, industry
sectors, “smart beta” strategies (don’t ask), and
casino-like ETFs in which investors can speculate
(i.e., bet) on whether the stock market will rise or
fall, with as much as triple leverage. (I’m not mak-
ing this up.)

Both kinds of ETFs (broad or narrow in focus—
even those with portfolios tracking the S&P 500,
identical to that first index fund) are traded at a
frenzied pace. Through September 2015, shares of
the 100 largest ETFs, valued at $1.5 trillion, were
turning over at an annualized volume of $14 trillion,
a turnover rate of 864%.

By way of comparison, the annualized turnover
volume of the 100 largest stocks, valued at $12 tril-
lion, is running at $15 trillion for the same period,
a turnover rate of 117%. Trading in the 100 largest
ETFs thus represents about 89% of such stock trad-
ing, up from a mere 7% 15 years ago. Given these
powerful data, it is hardly unfair to describe today’s
ETFs—as a group—as the modern way to speculate
in the stock market.3

Assets of US ETFs, including bond ETFs, now
total $2 trillion. But the variety of objectives and
strategies of ETFs—many of which are index funds
in name only—is wide indeed. The level of ETF
turnover also varies widely. Some ETF managers
focus on attracting high-turnover institutional
traders; others strive to serve largely individual
investors. Table 2 shows some of these differences.
For instance, State Street Global Advisors appears

most focused on institutional trading (some 90%
of the shares of State Street’s S&P 500 SPDR are
held by institutions), whereas Vanguard seems most
focused on individual investors and their brokers
and advisers.

But even the highest institution-dominated turn-
over rate pales in comparison with the turnover rates
of most ETFs designed to attract investors (really
speculators) who like to gamble. Their selling propo-
sition: “Bet on which way the market is going, and
if you’re confident, do so with triple leverage.” In
2015, the average annualized turnover of the shares
of four of the most active ETF providers (though
relatively small and held almost entirely by individu-
als) ranged from 953% to 10,308%.

It is important to recognize the differences
among three distinct classes of ETF investors:
• Financial institutions that use ETFs as trading

vehicles—sometimes to speculate, sometimes
to hedge, sometimes to arbitrage, and some-
times to implement other short-term strategies.
These institutions are by far the largest traders
of ETF shares.

• Individual investors who trade rapidly—easily
done with ETFs on a momentary basis at rela-
tively low (but not zero) cost—the next largest
segment.

• Individual investors and their financial advis-
ers who prefer to use brokerage platforms and
trade only moderately and strategically—as
well as a growing cadre of “robo-advisers,”
whose computer-based allocations for their cli-
ents tend to be fairly stable but who may trade
often in order to harvest tax losses. Also, certain
withdrawal strategies (e.g., generating income
in retirement) can be easier to implement with
ETFs than with TIFs, although ETF trading costs
are generally higher than those of no-load TIFs.
This third segment appears to be the smallest.

Table 2. Assets, Institutional Ownership, and Turnover: ETFs of Various Sponsors, 30 August 2015

Total Assets
($ billions)

Institutional Ownership
(%)

Annualized Turnover
($ billions)

Annualized Turnover
(%)

Largest ETF sponsors
BlackRock 809 62 4,910 606
Vanguard 469 43 908 193
State Street 409 63 8,692 2,122

1,687 56 14,510 859

Most active ETF sponsors
PowerShares 97 40 928 953
ProShares 25 12 873 3,444
Direxion 9 5 506 5,551
VelocityShares 3 7 299 10,308

134 16 2,606 1,936

Sources: Morningstar; NASDAQ.

The Index Mutual Fund

January/February 2016 www.cfapubs.org 13

Which Strategy Wins?
That first index mutual fund—the TIF model—was
designed to serve long-term investors. The ETF
model, however, seems designed to provide a new
way for institutional and individual short-term spec-
ulators to trade to their hearts’ content—and thus a
new opportunity for Wall Street to make profits from
index funds. That fact alone helps explain why the
index fund is no longer that “unAmerican” pariah
of 1976, needing to be “stamped out” before it led
to a pandemic.

Most of today’s 1,800 ETFs are less diversified,
carry greater risk, and are used largely for rapid-fire
trading—speculation, pure and simple. “Follow the
money” is a useful aphorism to explain why Wall Street
has jumped on the bandwagon of indexing, another
surprising turnabout from the original concept.

But logic more than stubbornness (I think!) con-
vinces me that the model of that first index mutual
fund—tracking the S&P 500; operating with minus-
cule all-in costs (now as little as 0.05% of assets),
nominal portfolio turnover, tiny (if any) transaction
costs, and high tax efficiency; and designed to be
held “forever”—remains the optimal way for inves-
tors to earn their fair share of whatever returns, good
or bad, our stock market delivers.4

Conclusion: The Chief Cornerstone
The index fund’s dominance of the mutual fund
industry belies its discouraging initial reception
40 years ago. Perhaps we should not be surprised.
Similar turnabouts have happened before. As the

Good Book reminds us, “The stone that the build-
ers rejected became the chief cornerstone.” And the
index fund has become just that, the chief corner-
stone of the mutual fund industry.

Being the “first mover” in the index fund
arena—the stone that was rejected four decades
ago—and then creating the first broad lineup of
index funds, all the while preaching the gospel
of indexing from a “bully pulpit,” have sparked
Vanguard’s unarguable leadership in the field. That
original index fund, which began with $11 million
of assets, has become a major part of a huge index
fund family, now accounting for more than $2 tril-
lion of Vanguard’s $3 trillion asset base. Yes, index-
ing has proved to be not only a winning investment
strategy for its shareholders but also a winning
business strategy for its creator.

Don’t take my word for it. In his foreword to my
first book, Bogle on Mutual Funds: New Perspectives
for the Intelligent Investor, Paul Samuelson wrote that
Vanguard “has changed a basic industry in the opti-
mal direction” (Samuelson 1993), and his conviction
never wavered. David Swensen, Yale University’s
endowment fund chief, has concluded that “a pas-
sive index fund managed by a not-for-profit invest-
ment organization represents the combination most
likely to satisfy investor aspirations” (Swensen 2005).
Warren Buffett has also endorsed indexing, directing
that 90% of the assets of the trust he has established
for his wife be invested in a low-cost S&P 500 Index
fund. For investors who think they can pitch against
this “Murderers’ Row” of three great batters and
strike them out, I can say only, “Good luck.”

Notes
1. Paul A. Samuelson, “Challenge to Judgment,” Journal of

Portfolio Management, vol. 1, no. 1 (Fall 1974): 17–19.
2. In the interim period, 30 June 1975 to 31 December 1985, the

annual return of the S&P 500 was 13.3%, bringing its return
to 12.3% over the full 1945–2015 period—a remarkable era in
stock market history. It would be unwise in the extreme to
expect such a generous past return to be prologue to the future.

3. The SPDR (Standard & Poor’s Depositary Receipt, or “Spider”)
S&P 500 constitutes the largest ETF, which is, day after day,
the most widely traded stock in the world. With $150 billion
of assets, the dollar volume of its annualized trading volume
in 2015 was running at an annual rate of $6 trillion, a turnover
rate of 3,300%.

4. In the spring of 1991, the late Nathan Most of the American
Stock Exchange visited me in my Valley Forge, Pennsylvania,
office. He had designed a novel approach to indexing—what

would become known as the ETF—and he wanted to partner
with us, using our Vanguard 500 Index Fund as the trading
vehicle. He was a fine gentleman and took with equanimity
my on-the-spot decision—without much reflection and with
no consultation with my staff—to relinquish the opportunity
for yet another Vanguard “first” that would have combined
Most’s innovative structure and Vanguard’s largest index
fund. But I believed then—and I believe now—that buying
and holding a broad US stock market index fund is investing
that will serve long-term index investors well. I also believe
that trading any index fund “in real time” is speculation that
will, finally, poorly serve short-term index traders. Many who
know this story describe it as a major failure on my part. I
consider it a major success and have no regrets. Standing on
principle is the right course of action.

References
Samuelson, Paul. 1993. “Foreword.” In Bogle on Mutual Funds:
New Perspectives for the Intelligent Investor. By John C. Bogle. New
York: McGraw-Hill.

Swensen, David. 2005. Unconventional Success: A Fundamental
Approach to Individual Investment. New York: Free Press.

Copyright of Financial Analysts Journal is the property of CFA Institute 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.