From the Board
The Arithmetic of Active represents3 per cent of the value of the securities
in the market, a passive investor's portfolio will
Management have 3 per cent of its value invested in X.
Equivalently, a passive manager will hold the
by WilliamF. Sharpe,TimkenProfessorEmeritusof same percentageof the total outstanding amount
Finance,StanfordUniversity,and Chairman,William of each security in the market.2
F. SharpeAssociates * An activeinvestoris one who is not passive. His or
"Today's fad is index funds that track the Stan- her portfolio will differ from that of the passive
dard & Poor's500. True, the average soundly beat managers at some or all times. Because active
most stock funds over the past decade. But is this managers usually act on perceptions of mispric-
an eternal truth or a transitoryone?" ing, and because such perceptions change rela-
tively frequently, such managers tend to trade
"In small stocks, especially, you're probablybet- fairly frequently-hence the term "active."
ter off with an active manager than buying the
market." Over any specified time period, the marketreturn
will be a weighted average of the returns on the
"The case for passive management rests only on
securitieswithin the market,using beginning market
complex and unrealistictheories of equilibriumin
values as weights.3 Eachpassive managerwill obtain
capital markets."
precisely the marketreturn, before costs.4 From this,
"Any graduate of the __ Business School it follows (as the night from the day) that the return
should be able to beat an index fund over the on the average actively managed dollar must equal
course of a marketcycle." the market return. Why? Because the market return
must equal a weighted average of the returns on the
Statements such as these are made with alarming
passive and active segments of the market.If the first
frequency by investment professionals.' In some
two returns are the same, the third must be also.
cases, subtle and sophisticated reasoning may be This proves assertion number 1. Note that only
involved. More often (alas), the conclusions can only
simple principles of arithmetic were used in the
be justified by assuming that the laws of arithmetic
process. To be sure, we have seriously belaboredthe
have been suspended for the convenience of those
obvious, but the ubiquityof statements such as those
who choose to pursue careersas active managers.
quoted earliersuggests that such laboris not in vain.
If "active" and "passive" management styles are
To prove assertionnumber2, we need only rely on
defined in sensible ways, it mustbe the case that
the fact that the costs of actively managing a given
(1) before costs, the return on the average actively number of dollars will exceed those of passive man-
managed dollar will equal the return on the agement. Active managers must pay for more re-
average passively managed dollar and search and must pay more for trading. Security
(2) after costs, the return on the average actively analysts (e.g., the graduates of prestigious business
managed dollar will be less than the return on schools) must eat, and so must brokers, traders,
the average passively managed dollar. specialists and other market-makers.
Becauseactive and passive returnsare equal before
These assertions will hold for any time period. More-
cost, and because active managersbear greatercosts,
over, they depend only on the laws of addition, it follows that the after-costreturn from active man-
subtraction,multiplicationand division. Nothing else agement mustbe lower than that from passive man-
is required.
agement.
Of course, certain definitions of the key terms are This proves assertion number 2. Once again, the
necessary. Firsta marketmust be selected-the stocks
proof is embarrassingly simple and uses only the
in the S&P 500, for example, or a set of "small"
most rudimentarynotions of simple arithmetic.
stocks. Then each investor who holds securitiesfrom
Enough (lower) mathematics. Let's turn to the
the market must be classified as either active or
practicalissues.
passive.
Why do sensible investment professionalscontinue
* A passive investoralways holds every security to make statements that seemingly fly in the face of
from the market, with each represented in the the simple and obvious relationswe have described?
same manner as in the market.Thus if securityX How can presented evidence show active managers
beating "the market" or "the index" or "passive
1. Footnotes appear at end of article. managers"?Three reasons stand out.5
1991Gl7
FINANCIALANALYSTSJOURNAL/ JANUARY-FEBRUARY
The CFA Institute
is collaborating with JSTOR to digitize, preserve, and extend access to
Financial Analysts Journal ®
www.jstor.org
* First, the passive managers in question may not large-capitalizationstocks, but may exceed the
be truly passive (i.e., conformto our definitionof market's performance in periods when small-
the term). Some index fund managers "sample" capitalization stocks do well. In both cases, of
the market of choice, rather than hold all the course, the average actively managed dollarwill
securitiesin marketproportions.Some may even underperformthe market, net of costs.
charge high enough fees to bring their total costs
To repeat:Properlymeasured, the average actively
to equal or exceed those of active managers.
managed dollar must underperformthe average pas-
* Second, active managersmay not fully represent
sively managed dollar, net of costs. Empiricalanaly-
the "non-passive" component of the market in ses that appear to refute this principle are guilty of
question. Forexample, the set of active managers impropermeasurement.
may exclude some active holders of securities This need not be taken as a counsel of despair. It is
within the market (e.g., individual investors). perfectly possible for some active managers to beat
Many empirical analyses consider only "profes- their passive brethren,even after costs. Such manag-
sional" or "institutional"active managers. It is, ers must, of course, manage a minority share of the
of course, possible for the average professionally actively managed dollars within the market in ques-
or institutionallyactively managed dollar to out- tion. It is also possible for an investor (such as a
perform the average passively managed dollar, pension fund) to choose a set of active managersthat,
after costs. For this to take place, however, the collectively,provides a total returnbetter than that of
non-institutional, individual investors must be a passive alternative, even after costs. Not all the
foolish enough to pay the added costs of the managers in the set have to beat their passive coun-
institutions' active management via inferiorper- terparts, only those managing a majority of the in-
formance. Another example arises when the ac- vestor's actively managed funds.
tive managers hold securities from outside the An importantcorollaryis the importanceof appro-
market in question. For example, returns on priateperformance measurement. "Peergroup" compar-
equity mutual funds with cash holdings are often isons are dangerous. Because the capitalization-
compared with returns on an all-equityindex or weighted average performance of active managers
index fund. In such comparisons, the funds are will be inferior to that of a passive alternative, the
generally beaten badly by the index in up mar- former constitutes a poor measure for decision-
kets, but sometimes exceed index performancein making purposes. And because most peer-group
down markets. Yet another example arises when averages are not capitalization-weighted, they are
the set of active mangers excludes those who subject to additional biases. Moreover, investing
have gone out of business during the period in equal amounts with many managersis not a practical
question. Because such managers are likely to alternative.Nor, a fortiori,is investing with the "me-
have experienced especially poor returns, the dian" manager(whose identity is not even known in
resulting "survivorship bias" will tend to pro- advance).
duce results that are better than those obtained The best way to measure a manager'sperformance
by the average actively managed dollar. is to compare his or her return with that of a compa-
* Third, and possibly most important in practice, rablepassivealternative.The latter-often termed a
the summary statistics for active managers may "benchmark"or "normal portfolio"-should be a
not truly represent the performanceof the aver- feasiblealternativeidentified in advanceof the period
age actively managed dollar. To compute the over which performance is measured. Only when
latter, each manager'sreturnshould be weighted this type of measurement is in place can an active
by the dollars he or she has under management manager (or one who hires active managers) know
at the beginning of the period. Some compari- whether he or she is in the minority of those who
sons use a simple average of the performanceof have beaten viable passive alternatives.
all managers (large and small); others use the
performance of the median active manager. Footnotes
While the results of this kind of comparisonare, 1. The first two quotations can be found in the
in principle, unpredictable,certainempiricalreg- September3, 1990 issue of Forbes.
ularitiespersist. Perhaps most important,equity 2. When computing such amounts, "cross-holdings"
fund managers with smaller amounts of money within the market should be netted out.
tend to favor stocks with smaller outstanding 3. Events such as mergers, new listings and reinvest-
values. Thus, defacto,an equally weighted aver- ment of dividends that take place during the
age of active manager returns has a bias toward period require more complex calculationsbut do
smaller-capitalizationstocks vis-a-vis the market not affectthe basic principles stated here. To keep
as a whole. As a result, the "average active things simple, we ignore them.
manager" tends to be beaten badly in periods 4. We assume here that passive managers purchase
when small-capitalizationstocks underperform their securities before the beginning of the period
FINANCIAL ANALYSTS JOURNAL / JANUARY-FEBRUARY 1991 O 8
in question and do not sell them until after the agers' willingness to provide desired liquidity(at a
period ends. When passive managers do buy or price).
sell, they may have to tradewith active managers; 5. There are others, such as differentialtreatmentof
at such times, the active managers may gain from dividend reinvestment, mergers and acquisitions,
the passive managers, because of the active man- but they are typically of less importance.
Ten Commandments of that summarizesall the informationrelevant to mak-
ing an investment decision. A user must analyze
Financial Statement Analysis growth and leverage, among other factors, as well as
profitability.
by WilliamH. Beaver,JoanE. HorngrenProfessorof 5. Thou shalt not overlook the implications of what
Accounting,GraduateSchoolof Business,StanfordUni- is read. It is not sufficient simply to know that a
versity company is a high-growth firm or a highly leveraged
firm; one must also know that such characteristics
Individuals must pass a proficiency test before typicallyimply higher risk, as well.
obtaining a driver's license. By contrast, investors 6. Thou shalt not ignore events subsequent to the
need not pass any proficiencytest before tryingto use financial statements. Financial statements are not
financialstatements as part of their investment anal- forecasts of the future. The annual financial state-
ysis. Investorsare not requiredto have takena course ments report the financialcondition of the company
in accounting or financial statement analysis. They as of year-end. They do not purport to capture the
are not required even to have read or understood effects of events that occur after year-end. They thus
books written on the subject. Yet analyzing financial become increasingly out-of-date as the year
statements requires at least as much knowledge and progresses. The rate of deteriorationin timeliness is
skill as driving an automobile. Perhapseach financial related to many factors, including the growth rate of
statement should contain a warning to potential the firm.
users, similarto those found on many products. The 7. Thou shalt not overlook the limitationsof finan-
warning would include at least the following 10 cial statements. Financialstatements reporton only a
commandments. specified set of events, not all events or all possible
1. Thou shalt not use financialstatements in isola- financial effects of a single event. Financial state-
tion, but only in the broadercontextof other available ments do not generally represent estimates of the
information. The additional information includes market values of the reported assets and liabilities,
data on economy-wide conditions and industry-wide nor do they reflect changes in the market values of
conditions. those assets and liabilities.
2. Thou shalt not use financial statements as the 8. Thou shalt not use financialstatements without
only source of firm-specificinformation. There are adequateknowledge. Investorsshould be sufficiently
many other sources of information about the com- competent to read, understand and analyze financial
pany. Consider, for example, the popular financial statements. Otherwise, the investor cannot be called
press and periodicals, as well as analysts' reports. a user of financial statements in any meaningful
3. Thou shalt not avoid reading footnotes, which sense.
are an integral part of financialstatements. Financial 9. Thou shalt not shun professionalhelp. If unwill-
statements cannot be reasonably analyzed without ing or unable to attain adequate knowledge, the
reading and understanding the footnotes. By anal- investor should defer to someone who does have
ogy, a temperatureof 10 degrees is meaningless in such ability, such as a financialanalyst. If unwilling
isolation, unless one knows whether it is being mea- or unable to obtain help, the investor should hand
sured on the Celsius or Fahrenheitscale. In a given over a portion of the investment process (hence a
country, a uniform temperature scale may be as- portion of the investment decision itself) to a profes-
sumed. The same is not true of the accounting sional manager.
methods used under generally accepted accounting 10. Thou shalt not take unnecessary risks. If un-
principles. GAAP, for example, permits a variety of willing or unable to obtain professional help, the
inventory and depreciation methods. A description investor should undertakeinvestments where invest-
of a company's accounting policies is included as a ment risk is minimal, or where analysis of financial
statements is not an issue. Investment in U.S. Trea-
part of the footnotes.
4. Thou shalt not focus on a single number. The sury bills is one example.
Of course, there may be more than 10 command-
investor should read and understand all the material
ments for financial statement analysis, but these
presented in the financialstatements. Financialstate-
capture the primaryissues.
ments are not designed to be reduced to a single
number. Net income is not intended to be thenumber Concludedon page 18.
FINANCIAL ANALYSTS JOURNAL / JANUARY-FEBRUARY 1991 O 9
