Content uploaded by Laurence B. Siegel
Author content
All content in this area was uploaded by Laurence B. Siegel on Oct 24, 2014
Content may be subject to copyright.
56 FIVE MYTHS ABOUT FEES SPRING 2006
Ofthe three dimensions of investment man-
agement—return, risk, and cost—investors
have direct control over only cost. Cost
includes transaction costs and investment
management fees. Wefocus hereon fees.
While return, risk, and even transaction costs have
been widely studied, fees are poorly understood, and there
is little literature on them. Yet they are critically impor-
tant: The present value of fees in a long-term investment
relationship represents the transfer of a significant fraction
of the investor’scapital to the manager.Moreover, the
incentives provided bythe fee structurehave a strong influ-
ence on the manager’s strategy, particularly on the fund’s
volatility, the mix of alpha and beta bets, and the fund’s size.
Investment management fees areatimely topic
because of three trends in the investment landscape:
•
Investors increasingly look to separate alpha from beta.
•The cost of beta has dropped to very low levels.
•An explosion of new alpha providers, including
hedge funds, private equity firms, and even other-
wise traditional managers, use unconventional fee
structures.
Notably, the notion that focuses on separating alpha
from beta also places strong emphasis on paying active fees
only for the alpha portion of any investment and on looking
closely at costs.1Very briefly, the literature says that:
•Investors can obtain beta at very low cost through
index funds, exchange-traded funds (ETFs), futures,
and swaps. Thus beta is one of life’s great bargains,
if you believe that the market payoff for beta risk
will be attractive.
Five Myths About Fees
The truth behind analyzing fees, in the context of investment goals.
Ronald N. Kahn, Matthew H. Scanlan, and Laurence B. Siegel
RONALD N. KAHN
is the global head of
Equity Research at
Barclays Global Investors
in San Francisco.
ron.kahn@barclaysglobal.com
MATTHEW H.
SCANLAN
is the head of Americas
Institutional Business at
Barclays Global Investors.
matthew.scanlan@
barclaysglobal.com
LAURENCE B. SIEGEL
is the director of research,
Investment Division, at
The Ford Foundation in
New York City.
l.siegel@fordfound.org
For more information please
visit www.iijpm.com.
Copyright © 2007, Institutional Investor. Limited permission granted to CFA Institute.
SPRING 2006 THE JOURNAL OF PORTFOLIO MANAGEMENT 57
•Alpha is scarce (because active management is a zero-
sum game), difficult to find, and very valuable. It is
expensive, and should be.
•The beta and alpha decisions are separate. An investor
can build a portfolio of alpha sources from any mix of
asset classes, and then add or subtract beta exposures
as desired.
Our perspective in this article is that of the client,
but we must also understand the manager’s perspective.
Fee negotiation is a game; sometimes client and manager
interests are aligned, and sometimes they are opposed. To
understand the game, we must identify the motivations
of all the players.
The investor’s goal is to maximize expected returns
subject to a risk budget constraint. For most investors,
this involves maximizing expected alpha after fees. This
isn’t easy.
It’s difficult enough to maximize expected alpha
before fees. Managers deliver alpha with great uncertainty.
It takes time to distinguish winnersfrom losers, or (among
winners) to distinguish the truly skillful from the merely
lucky. And even time can never eliminate all such ambi-
guity. Yet investors must rise to this challenge to rationally
allocate risk to active managers.
Incentive fees make the challenge of estimating
expected alpha after fees even tougher. These fees both
depend on performance and can influence the underlying
strategy.And clients must often decide whether their
expected alpha after fees is higher with an incentive fee,
or the moretraditional fixed or ad valorem fee.
Our goal is to provide some guideposts for clients
seeking to maximize expected alpha after fees. Towardthat
end, we’ll identify, and correct, a number of popular myths
regarding fees. Along the way, we’ll describe key elements
of the fee negotiation game, and determine conditions
under which the client should prefer fixed or incentive fees.
Let’s start with a list of popular myths about fees.
Weaddress the issues raised by each myth to analyze the
truth behind them:
Myth 1: Fees should be as low as possible.
Myth 2:
Incentivefees are always better than fixed fees.
Myth 3: High water marks always help investors.
Myth 4: Hedge funds are where the alpha is.
They deserve their high fees.
Myth 5: You can always separate alpha from beta,
and payappropriate fees for each.
Let’s examine these one at a time.
MYTH 1—
FEES SHOULD BE AS LOW AS POSSIBLE
This baseline fee myth makes sense. Most people
understand that they should pay the lowest possible price
for a given good. For a given good is the tricky part, however.
We’ve already noted that index fund fees are very
low. They range as low as 0.01% for very large accounts
managed to track highly liquid indexes, and cap out at
around 0.20%, except in a few difficult-to-trade asset
classes. Swaps, futures contracts, exchange-traded funds,
and other ways of achieving beta exposure are also rela-
tive bargains. (We quote fees as annual rates charged as a
percentage of assets.)
Compared to index fees, typical fees for active man-
agement seem toweringly high. And, the zero-sum nature
of active management means that on average clients waste
these fees. It is difficult to put a number on typical active
management fees as the products vary so widely, but on
average for equities, these are roughly 0.50% for tradi-
tional long-only investments and 1.35% for retail accounts.2
Alternative investments, such as hedge funds and pri-
vate equity funds, charge annual fixed fees of 1% to 2% of
assets under management, plus an incentivefee equal to 20%
or more of performance above some benchmark.
To provide additional perspective, consider the fee
as a transfer of capital to the manager over the course of
asomewhat typical ten-year holding period.3A0.10%
fee on a retail index fund transfersabout 1% of capital,
while the 0.50% and 1.35% activeinstitutional and retail
fees transfer 5% and 13.5% of capital. Since clients pay
fees with certainty for the expectation of uncertain alpha,
these are significant sacrifices to make in the hope of alpha.
Goetzmann, Ingersoll, and Ross [2003] use an
option pricing model to estimate the value of hedge fund
management contracts. Using reasonable inputs, including
estimates of the rate at which investors exit the fund and
thus stop paying management fees, they find a contract
is worth 10% to 20%, and even as much as 33%, of the
amount invested. A permanent allocation to a portfolio
of hedge funds involves quite a large transfer of capital
from the investor to the population of managers.
So activefees aremuch higher than index fees, and
involve significant transfers of capital to managers. This
would seem to imply investorsshould tryto minimize
their fees by hiring index managers and the lowest-cost
active managers. But institutional and retail investors each
hire activemanagersfor upward of 70% of their assets.
Does this make sense?
58 FIVE MYTHS ABOUT FEES SPRING 2006
fees. Exhibit 1 captures the spirit of his approach and
results, which closely agree with the utility analysis.
Take,as an example, a manager (Manager 1) who
takes active risk of 5%, and who the investor expects to
deliver 4.2% alpha before fees. Assuming normal distri-
butions, the investor expects this manager to deliver pos-
itive alpha before fees with 80% probability. But as fees
rise, this probability of positive after-fee alpha falls dra-
matically.Note that a 50% probability of positivealpha cor-
responds to zero expected alpha—the annual active return
is as likely to be positive as negative.
One obvious lesson from Exhibit 1: Fees must
remain significantly below the expected alpha. But
Exhibit 1 includes a more subtle lesson as well. Consider
Manager 2, with the same expected alpha before fees, but
with higher active risk (10%) and hence a lower infor-
mation ratio. Exhibit 1 implies that investors would pay
higher fees to the more consistent (higher IR) manager.
This agrees with our analysis based on utility.
Beyond this analysis of utility and probability of
outperformance,moreconsistent managers have an addi-
tional advantage: Investors have higher confidence in their
skill.
So what is the truth about keeping fees as low as
possible? Our fundamental analysis has shown that investors
should be willing to pay higher fees to managers with
certain characteristics, especially the ability to consistently
deliver strong alpha and high information ratios. Ascer-
taining those characteristics is very challenging. Still, what
mattersis not the fee level, but the manager’s ability to
deliver utility after fees.
As Waring et al. [2000] have discussed,
hiring active managers makes sense only under
two conditions. One, the investor believes that
active management is possible, i.e., that there
are managers who will produce alpha on
average in the future. Second, the investor must
be able to identify those—presumably rare—
skilled managers.
Any investor satisfying those conditions
should rationally aim to maximize expected
alpha after fees, not just minimize fees. Achieving
this objective means sharing the alpha with the
manager. For any given level of expected alpha,
the investor should try to minimize the fee. But
in equilibrium, the investor must share a sub-
stantial fraction of the alpha with the manager,
because alpha is rareand valuable.
So what is the right level of active man-
agement fees? The market provides one answer, in the
prices we describe above. For example, according to the
market, 0.50% is about the right price for a traditional
long-only active equity product.
But the market maynot be right, and it is certainly
wrong on average. What about a more fundamental
approach to determining the right fee level? Let’s start by
considering the utility offered by different managers. We
will measureinvestor utility as:
(1)
where Equation (1) includes alpha net of fees, the investor’s
risk aversion,
λ
,and active risk,
ω
.Risk-averse investor
utility falls short of the net alpha due to the penalty for
risk. If twomanagersprovide the same gross alpha, but
different risk levels, the lower-risk manager provides higher
utility to the investor.
Equation (1) supports twoimplications. First, man-
ager fees should fall significantly not only below gross
alpha, but also below gross utility. Second, in the case of
two managers with identical gross alpha, investors should
be willing to payhigher fees to the lower-risk manager.
Note that the lower-risk manager has the higher infor-
mation ratio (IR), where:
(2)
Ennis [2005] explores similar territory, trying to
identify plausible ranges of fees, working from the impact
of fees on the likelihood of achieving positive alpha after
IR ≡α
ω
U=−αλω
2
EXHIBIT 1
Probability of Positive Net Alpha
(given expected before-fee alpha of 4.2%)
SPRING 2006 THE JOURNAL OF PORTFOLIO MANAGEMENT 59
MYTH 2—INCENTIVE FEES ARE
ALWAYS BETTER THAN FIXED FEES
Incentive fees have many advantages over fixed fees,
but they have disadvantages as well. The better choice
will depend on circumstances. For example, we construct
asimple model showing that as investors become more
able to choose skillful managers, their preference moves
from incentive to fixed fees.
But first, let’s describe how incentive fees work, and
the advantages and disadvantages of both fixed and incen-
tive fees. The simplest incentive fees include a base fee plus
apercentage of the return above some performance bench-
mark. More complicated structures add caps, high water
marks, and other features to the calculation of the sharing
amount. When managers offer investors the choice of
either a fixed or an incentive fee, the base fee should lie
below the fixed fee, and the expected total incentive fee
(base plus expected performance share) should exceed the
fixed fee alternative. A fixed (certain) fee should equate
to a higher but uncertain fee.
With that basic structure in mind, let’s discuss the
pros and cons of each fee structure. The pros and cons
of fixed fees arise mainly because they extract a fixed
amount for variable performance. The certainty associ-
ated with fixed fees benefits both clients and managers.
The disconnect between fees and performance raises sev-
eral issues, some benefiting clients and some benefiting
managers.
The certainty of fixed fees allows clients to budget
accurately for these costs, and provides managers with
low-volatility revenue. This in turn facilitates invest-
ment in the manager’s business—additional research,
product improvements—of benefit to managers and
clients.
The disconnect between fixed fees and variable per-
formance raises two main issues. First, the fee in a given
year or over time may be too high or too low. This can
advantage the manager at the expense of the client, or
vice versa, at least in the short run. In the longer run,
paying the wrong fee causes problems for both manager
and client. If the fee is too high for the alpha delivered,
amanager maybenefitfor a while,until the client ter-
minates the manager. Exacerbating the damage, this sit-
uation sometimes leads clients to keep poorly performing
managers too long, in the hope of earning back the fee.
If the fee is too low, the client benefits until the manager
neglects the product, under-resources it, dumps the client,
or gathers too much in additional assets.
This brings up the second issue arising out of the
disconnect between fees and performance, in particular
the interaction of fees with the different interests of clients
and managers. Clients want high returns. Managers want
high profits. With fixed fees, the manager maximizes
profits through extensive asset gathering, even if asset
gathering weakens performance.
All active strategies have capacity limits. As assets
grow, trading costs rise, and the manager has more and
more difficulty implementing insights in the portfolio.
Expected returns fall (see, for example, Kahn and Shaffer
[2005]). Capacity constraints create conflicts of interest
between the client and the manager.
So what about incentive fees? They address the struc-
tural problem of fixed fees by directly connecting pay and
performance. This seems like an unambiguous improve-
ment except that performance can arise out of skill or
luck, and this raises a different issue.
But first, incentive fees do address the two issues
concerning fixed fees. By connecting fees to performance,
they avoid years when fees and performance are out of bal-
ance. And incentive fees also help align the different inter-
ests of clients and managers. Theymotivate managersto
deliver strong performance, and to avoid raising assets to
the detriment of performance. They even motivate the
keyinvestment professionals to focus on investing, not
asset gathering. Managersand clients can both prosper
from these aspects of incentive fees.
Incentive fees even have some related side benefits.
Paying only for performance can facilitate investing with
unorthodox or more risky managers. It can also lead to
better pools of managers, by eliminating the temptation
to stick with poorly performing managers to try to earn
back fees already paid.
On the negative side, the volatility associated with
performance fees causes problems for clients and man-
agers, for the same reasons that the certainty of fixed fees
creates benefits. Clients can’t budget as easily for incen-
tive fee costs. Managers face volatile revenue streams.
The newissue raised byincentive fees follows from
the observation that managers receive the same fee
whether performance comes from skill or luck. And,
given that incentivefees havean option-like character
(especially in their payment for positive performance
without a symmetric penalty for negative performance),
they become more valuable with increasing volatility of
alpha. Managers can therefore increase incentive fee value
by adjusting the investment strategy. This is not in the
interest of the client.
60 FIVE MYTHS ABOUT FEES SPRING 2006
Beyond the temptation to increase
volatility, incentive fees offer more general
gaming opportunities. As Black [1976,
p. 217] notes:
When things go badly, some people
react by doubling their bets. They
increase their exposure to risk in
hopes of recouping their losses. . . .
When things go well they may reduce
their exposure to risk so they can’t
lose what they have won. It’s a very
common gambling strategy and it’s a
very common philosophy of life.
Unfortunately, while incentive fees
create this temptation for managers, the
resulting behavior does not correspond to how clients
want their money managed. Note that the various embell-
ishments to incentive fees—high water marks, longer
measurement periods—do not eliminate these issues.
So each type of fee has advantages and disadvan-
tages. And either can be a reasonable way to compensate
amanager. So why might a particular client prefer one over
the other? In part, this will depend on howaclient weighs
the particular advantages and disadvantages wehave dis-
cussed. Beyond that, preferences will depend on the ability
to pick skillful managers.
Assume that of the population of active managers,
20% areskillful enough to deliver an alpha of 1.5% per
year before fees. The remaining 80% cannot beat their
benchmark and thus deliver an alpha of −0.20% before
fees.4The assumption that 20% of managers are skillful is
more favorable than the most optimistic persistence-of-
performance studies would imply (see, for example,
Grinold and Kahn [2000, p. 566]).
Wewill further specify that there are only two
possible fee schedules: a flat 0.30% fee, or an incentive
fee of 0.20% plus 20% of the positivealpha. With the
incentive fee, skillful managers receive 0.20% + 20% ×
1.50% = 0.50% on average, while unskillful managers
receive0.20%.
Aclient with no ability to identify skillful managers
has a 20% chance of success, since skillful managers make
up 20% of all managers. Before fees, the client’s expected
alpha is:
(3)
Eα
{}
=+−⇒20 1 50 80 0 20 0 14%( . %) %( . %) . %
With fixed fees, the client loses 0.16% on active
management. What about using incentive fees? The
expected incentive fee in this case is:
(4)
So, with incentive fees, the client loses 0.12% on
activemanagement. While neither case looks attractive—
and active management should not look attractive to
investors with no ability to pick managers—the incen-
tivefee looks better than the fixed fee.
With perfect skill in picking active managers, on
the other hand, the client will prefer fixed fees. For skillful
managers, the fixed fees are 0.30%, while the incentive
fees average 0.50%.
Between these extremes, there is some point of indif-
ference between the two types of fee schedules. Exhibit 2
shows how this point depends in this example on the
investor’s skill in picking managers.
Exhibit 2 identifies three important regions,
depending on skill in hiring managers. Below a 29% prob-
ability of success, investorsshould not pursue activeman-
agement. The expected alpha after fees is negative.
Between a 29% and a 34% probability of hiring skilled
managers, investors should prefer incentive fees to fixed
fees in this model. Above a 34% probability, investors
would prefer fixed fees.
For comparison with required skill in other areas of
activemanagement, we can convert this to a required
information coefficient, or IC. The IC, the correlation
E fee
{}
=+⇒20 0 50 80 0 20 0 26%(.%) %(.%) .%
EXHIBIT 2
Indifference Analysis Between Fixed and Incentive Fee
SPRING 2006 THE JOURNAL OF PORTFOLIO MANAGEMENT 61
of forecast and realized returns, measures active manage-
ment skill. With no skill, IC = 0; with perfect skill, IC = 1.
Skillful stock-pickers exhibit ICs around 0.05 to 0.10.
For skillful asset allocation managers, or market timers, ICs
range from 0.10 to 0.20 at best. In our simple model,
investors require an IC of 0.11 to achieve positive alpha
net of costs, and an IC of 0.18 to prefer fixed to incen-
tive fees.
The specific ranges change as we change model
assumptions. In general, as the fixed fees increase, investors
increasingly prefer incentive fees. As manager skill
increases, investors increasingly prefer fixed fees.
MYTH 3—
HIGH WATER MARKS ALWAYS HELP INVESTORS
To makeincentivefees more palatable to investors,
many firms offer high water mark provisions. Such a pro-
vision calculates the incentive fee based on the highest pre-
viously achieved net asset value (NAV). This prevents an
investor from paying twice for the same performance. Say
that a fund experiences the returns shown in Exhibit 3,
and that the incentiveshareis 20%.
Without a high water mark, the incentive fee is $2
in year 1 and $3 in year 3, for a total of $5. With the high
water mark provision, the manager collects no fee on the
increase in value from $15 back to the old high of $20.
The fee in year 3 is only $2, for a total fee of $4.
What could be fairer? With the high water mark, the
investor avoids paying twice for the travel from $15 to $20.
In fact, high water marks do help investors in that
theyreduce the overall fee for a given pattern of invest-
ment returns. Unfortunately, they also introduce perverse
incentives that can alter future return patterns.
Consider the predicament of the manager in our
example after year 2. Any gain lower than $5 produces no
incentive fee. This increases the manager’s motivation to
takeadditional risk, whether the investor wants to or not.
Specifically, the manager may favor bets that add at least
$5 to NAV, preferring higher but less probable returns to
the lower but steadier returns preferred by clients.
If the probability of returning to the high water
mark within a reasonable time is too low, the manager
may close the fund, and start up a new fund, with a new
high water mark. The investor, then, also faces a new
high water mark, with a new manager. The investor thus
pays twice for the same travel, although by different
managers.
So high water marks help investors only when the
decline in NAV does not motivate the manager to increase
risk or to close the fund. This may correspond to a narrow
range of outcomes. We would caution investors to mon-
itor the behavior of managers with high water marks
carefully when they are losing money.
MYTH 4—
HEDGE FUNDS ARE WHERE THE ALPHA IS;
THEY DESERVE THEIR HIGH FEES
Let’s start with the evidence that many investors
believehedge funds arehigh-alpha, and then consider the
more complex truth behind this myth.
Institutional investors, pension plans in particular,
are todayin desperate need of alpha. At the peak of the
technology stock bubble,most plans were fully funded
or even overfunded, and the search for alpha was a fun but
not strictly necessary part of the job. But equity markets
and interest rates have dropped since then, and most plans
arenow significantly underfunded. Along with increased
contributions, they need alpha to deliver on promises to
beneficiaries. The demand for alpha has never been higher.
Consistent with this demand for alpha, as Exhibit 4
shows, we have seen large asset flows into hedge funds.
Since hedge funds promise pure alpha returns for the most
part, assets flowing into hedge funds are almost entirely
assets in search of alpha.
Exhibit 4 also shows a large increase in the number
of hedge funds. This provides a reasonable proxy for the
flow of investment managers into the hedge fund arena.
Finally, we seem to have seen a significant rise in
average hedge fund fees. T
en yearsago,almost all hedge
funds charged 1% of assets, plus 20% of performance above
abenchmark. Now many hedge funds charge 2% of assets
and/or incentive shares above 20%. Almost no funds
charge less than 1% of assets, or 20% incentive shares.
And, over these past ten years, we have also seen growth
in hedge funds of funds, with fund of fund fees layered
EXHIBIT 3
Incentive Fee Calculations
62 FIVE MYTHS ABOUT FEES SPRING 2006
on top of the hedge fund fees. We can’t exactly quantify
the average fee paid per dollar invested in hedge funds
today, but with many investors paying significantly more,
and basically none paying less, average fees have clearly
grown over the past ten years.
We have observed strong and increasing demand for
alpha, confronting its limited supply. In response, prices
and supply have increased. Unfortunately, the increase in
supply is an increase in the supply of hedge fund managers
offering alpha, not necessarily any increase in actual alpha.
So what is the truth here? First, arehedge funds
where the alpha is? Structurally,hedge funds offer two
distinct advantages over more traditional investments.
They avoid constraints, like the long-only constraint, that
can hinder investment performance. And they have the
flexibility to invest in manynon-traditional assets, from
private equity to distressed debt to derivatives. Clarke, de
Silva, and Thorley [2002] modify the fundamental law of
active management (Grinold [1989]) to say:
(5)
The information ratio of an investment product
depends on the information coefficient (a measure of
manager skill), the breadth, BR (a measureof opportu-
nity), and the transfer coefficient, TR (a measure of how
efficiently the manager’s ideas impact the portfolio). Struc-
turally, hedge funds can offer greater breadth (through
the availability of more assets) and higher transfer coeffi-
cients (through lack of portfolio constraints) than more
traditional products.
IR IC BR TR=
Beyond structure, what about talent?
Exhibit 4 demonstrates the flow of managers
into hedge funds. This is not surprising.
Beyond just responding to the increasing
demand for alpha, in which environment
would you rather work?
•Alarge and traditional firm owned by
someone else, where you spend consid-
erable time marketing and asset gath-
ering, you manage other people’s money
versus a benchmark, and you charge
0.50% and 0%.
•Your own business, where you spend
most of your time on investing, you
manage most of your liquid net worth
alongside your investors, you ignore
benchmarks, and you charge 2% and
20%.5
Of course hedge funds are not all wine and roses
for managers. They fail much more quickly than institu-
tional funds, because (like most entrepreneurial efforts)
they are usually undercapitalized and forced to take risks
that more established managers can avoid. And, the per-
ceived need to invest one’s own money in the fund makes
running a hedge fund even riskier.6
Still, there is no question that hedge funds have
attracted many investment managers, including many
leaving traditional investment firms.
While the structural advantages clearly exist, and
manyinvestment managers have moved from traditional
firms to hedge funds, beware the idea that hedge funds
are where all the alpha is. First, Sharpe’s [1991] arithmetic
of active management shows that aggregate alpha must
be zero. The increase in the number of hedge funds can’t
alter that. Aggregate alpha was zero ten years ago, and it’s
still zerotoday. Second, the many advantages of hedge
funds listed above appeal to both skilled and unskilled
managers. Both haveflowed into hedge funds. Unfortu-
nately, it isn’t easy to tell these two groups apart.
Third, traditional investment firms—particularly
those focused on institutional clients like pension plans—
have not stood still as demand for alpha, and hedge funds,
has grown. Most now offer products with the same struc-
tural advantages as hedge funds, plus the transparency and
institutional quality long demanded by these clients. They
have recognized the work environment advantages of
hedge funds, and at least started to address the issues most
EXHIBIT 4
Hedge Fund Universe 1990-2005
important for attracting and retaining key investment staff.
Institutional clients are desirable clients, due to their size,
sophistication, and typically longer commitment to prod-
ucts. As long as traditional firms can retain their institu-
tional clients, they should also be able to attract retain key
investment staff.
Finally, at least so far, traditional firms are the main
sources for lower-turnover strategies designed specifically
for the institutional investor need for alpha in bulk.
So hedge funds are not where all the alpha is. They
haven’t created any alpha in aggregate, and there are many
good reasons for investors to continue to use more tradi-
tional investment firms. But at the same time, there are
many talented hedge fund managers. Do they always
deserve their high fees?
The simple answer is no. No manager is great inde-
pendent of fees. At some price,amanager is just not worth
it. This was at least part of the motivation behind some
Harvard alumni vowing in 2004 to stop giving to the uni-
versity while it paid its employees as much as $30 million
for alpha delivered. A hedge fund decision should always
include an analysis of the impact of manager fees on the
net performance delivered to clients. This is always part
of hiring traditional managers, and should be part of hiring
hedge fund managers as well.
MYTH 5—
WE CAN ALWAYS SEPARATE ALPHA FROM
BETA, AND PAYAPPROPRIATE FEES FOR EACH
As we have seen, fees for alpha dramatically outpace
fees for beta. You should never pay alpha fees for beta
performance. Separating alpha from beta makes this rule
completely transparent.
In some cases, investors already do purchase sepa-
rated alpha and beta. Many products—including index
funds, ETFs, futures, and swaps—offer low-cost, cleanly
separated beta. A fewproducts, including pure market-
neutral equity funds (beta = 0) offer appropriately priced
purealpha. Beyond long-short, an active, long-only equity
manager who carefully adheres to style, capitalization,
industry, and factor neutrality delivers an essentially pure
alpha active return.
But most active products today deliver a combina-
tion of alpha and beta. Furthermore, there are challenges
to cleanly separating the two in many such products. Some
managers deliver beta that does not correspond to any
readily available index. Some managers deliver alpha
through timing of beta exposures.
SPRING 2006 THE JOURNAL OF PORTFOLIO MANAGEMENT 63
Consider, for example, a sector rotation manager.
When does a position represent beta, and when does it
represent alpha? Many international managers under-
weighted Japan for all of the 1990s. Was that a tactical
position, or just their choice of beta?
Adifferent problem arises in some asset classes like
real estate or private equity, where there are no pure beta
instruments to facilitate indexing, benchmarking, or
hedging.
Mixed (alpha and beta) products pose a danger for
investors of paying alpha fees for beta performance. Con-
sider a long-biased equity hedge fund with an average
beta of 0.6. In a given year, the equity market rises 16%
above the risk-free return, and the hedge fund delivers 11%
above risk-free. A standard 1% and 20% fee arrangement
would lead to a fee of 3.2%. But we might expect that
fund to return9.6% aboverisk-free just due to the average
beta. That would imply a true alpha of only 1.4%, and a
more appropriate fee of 1.28%. Investors in such a product
should understand its sources of return, and at a min-
imum try to pay, on average, alpha fees only for alpha
performance.
So you can’talways separate alpha from beta. This
doesn’t mean you will necessarily overpay for such prod-
ucts. It does mean you must carefully analyze what pro-
portions of alpha and beta the product delivers, and pay
appropriately for the combination.
CONCLUSIONS
Weshow in Myth 5 that while some investment
products offer pure alpha or pure beta, most active prod-
ucts offer a combination not easily separated into those
pieces. So, an investor who cares about fees above all else,
and who thus only wants to purchase alpha and beta sep-
arately, could do so. In fact, some institutional funds do
invest completely in beta, and in principle at least, others
could invest only in beta products plus equity market-
neutral funds.
But for most investors, restricting investments to
only separated alpha and beta products is too limiting.
There are many talented managers whose insights appear
only in mixed alpha and beta products. Whole asset classes
with distinct beta, like real estate and private equity, are
available almost exclusively as mixed products. The oppor-
tunity costs are simply too high to ignore such products.
Our goal has been to focus on the importance of
fees. Too often, investors consider fees only after already
deciding on an investment product. That’s too late.
64 FIVE MYTHS ABOUT FEES SPRING 2006
At the same time, fees should not be the overriding
single concern. For example, don’t invest only in per-
fectly separated alpha and beta products just because of the
fee transparency.
In the end, we return to the three dimensions of
active management: return, risk, and cost. High-fee prod-
ucts are worthwhile if they deliver sufficiently high returns
and low risk. Some high-return products have fees that
make them poor investments. Investors must analyze fees
in this overall context to manage their portfolio appro-
priately.
ENDNOTES
1See, for example, Kneafsey [2003], Leibowitz and Bova
[2005], Thomas [2005], or Waring and Siegel [2003].
2The institutional average comes from the eVestment
Alliance 2005 Fee Study,and an institutional product review
for the third quarter of 2005 from the firm, Casey, Quirk and
Associates. The retail numbers are based on third-quarter 2005
data for domestic stock mutual funds (excluding institutional
share classes) in the Morningstar Principia database.
3Webase the ten-year holding period on research bythe
firmCasey, Quirk and Associates showing that plan sponsors
typically turn over 10% of their investment manager pool per
year.
4For those worried about active management as a zero-
sum game, we can assume the skillful managers have somewhat
smaller asset size,so that the size-weighted alpha is zero, but we
ignorethis issue for our current purposes.
5Thanks to Elizabeth Hilpman of Barlow Partners for this
example.She originally presented this in “Hedge Fund Man-
agement,”aspeech at the AIMR Financial Analysts Seminar,
Evanston, Illinois, July 26, 2001.
6Brown, Goetzmann, and Ibbotson [1999] estimate an
annual attrition rate of 20% per year for established funds, with
apresumably higher rate for new funds.
REFERENCES
Black, Fischer, “Why and How Does an Investor Diversify?”
Quoted in Perry Mehrling, Fischer Black and the Revolutionary
Idea of Finance.Hoboken, NJ: John Wiley & Sons, 2005.
Brown, Stephen, William N. Goetzmann, and Roger G.
Ibbotson. “Offshore Hedge Funds: Survival and Performance,
1989-1995.” Journal of Business,January 1999, pp. 91-118.
Clarke, Roger, Harindra de Silva, and Steven Thorley. “Port-
folio Constraints and the Fundamental Law of Active Man-
agement.” Financial Analysts Journal,September/October 2002,
pp. 48-66.
Ennis, Richard M. “Are Active Management Fees Too High?”
Financial Analysts Journal,Vol. 61, No. 5 (September/October
2005), pp. 44-51.
Goetzmann, William N., Jonathan Ingersoll, and Stephen A.
Ross. “High-Water Marks and Hedge Fund Management Con-
tracts.” Journal of Finance,August 2003.
Grinold, Richard C. “The Fundamental Law of Active Man-
agement.” The Journal of Portfolio Management,Vol. 15, No. 3
(Spring 1989), pp. 30-37.
Grinold, Richard C., and Ronald N.Kahn. ActivePortfolio Man-
agement,2nd ed. New York: McGraw-Hill, 2000.
Kahn, Ronald N., and J.Scott Shaffer.“The Surprisingly Small
Impact of Asset Growth on Expected Alpha.” The Journal of
Portfolio Management,Vol. 32, No. 1 (Fall 2005), pp. 49-60.
Kneafsey, Kevin. “Solving the Investor’s Problem.” Investment
Insights,August 2003, Barclays Global Investors, San Francisco.
Leibowitz, Martin L., and Anthony Bova. “Allocation Betas.”
Financial Analysts Journal, July/August 2005.
Sharpe, William F. “The Arithmetic of Active Management.”
Financial Analysts Journal, January/February 1991.
Thomas, Lee R. III. “Engineering an Alpha Engine.” The
Journal of Investing,Winter 2005, pp. 23-32.
Waring, Barton, Duane Whitney, John Pirone, and Charles
Castille. “Optimizing Manager Structure and Budgeting Man-
ager Risk.” The Journal of Portfolio Management,Vol. 26, No. 3
(Spring 2000).
Waring, M. Barton, and Laurence B. Siegel. “The Dimensions
of Active Management.” The Journal of Portfolio Management,
Spring 2003, pp. 35-51.
To order reprints of this article, please contact Dewey Palmieri at
dpalmieri@iijournals.com or 212-224-3675.
